The Neolithic Settlement of Knossos in Crete: New Evidence for the Early Occupation of Crete and the Aegean Islands [Illustrated] 9781931534727, 1931534721

Citation preview

The Neolithic Settlement of Knossos in Crete

Nikos Efstratiou, Alexandra Karetsou, and Maria Ntinou

The Neolithic Settlement of Knossos in Crete New Evidence for the Early Occupation of Crete and the Aegean Islands

Frontispiece. The city, the fortifications, the harbor, and the hinterland of Khandax (Herakleion) in the first half of the 17th century. Map by unknown cartographer, 17th c., Collezione Museo Civico, Padua. Vikelaia Municipal Library, Herakleion.

PREHISTORY MONOGRAPHS 42

The Neolithic Settlement of Knossos in Crete New Evidence for the Early Occupation of Crete and the Aegean Islands

edited by Nikos Efstratiou, Alexandra Karetsou, and Maria Ntinou

Published by INSTAP Academic Press Philadelphia, Pennsylvania 2013

Design and Production INSTAP Academic Press, Philadelphia, PA Printing and Binding Hoster Bindery, Inc., Ivyland, PA

Library of Congress Cataloging-in-Publication Data The neolithic settlement of Knossos in Crete : new evidence for the early occupation of Crete and the Aegean islands / edited by Nikos Efstratiou, Alexandra Karetsou, and Maria Ntinou. pages cm. -- (Prehistory monographs ; 42) Includes bibliographical references and index. ISBN 978-1-931534-72-7 (hardcover : alk. paper) 1. Knossos (Extinct city) 2. Neolithic period--Greece--Crete. 3. Land settlement patterns, Prehistoric--Greece-Crete. 4. Crete (Greece)--Antiquities. I. Efstratiou, Nicholas. DF221.C8N46 2013 939’.18--dc23 2013016076

Copyright © 2013 INSTAP Academic Press Philadelphia, Pennsylvania All rights reserved Printed in the United States of America

To the memory of Professor J.D. Evans, a gentleman of British Archaeology –Nikos Efstratiou

Table of Contents

List of Tables in the Text. ......................................................................................... ix List of Figures in the Text. ........................................................................................... xiii Preface, Alexandra Karetsou. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix Acknowledgments. ................................................................................................... xxiii Introduction, Nikos Efstratiou. ................................................................................. xxv 1. The Excavation, Nikos Efstratiou, Alexandra Karetsou, and Eleni Banou. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2. The Stratigraphy and Cultural Phases, Nikos Efstratiou. ...................................................25 3. Fabric Diversity in the Neolithic Ceramics of Knossos, Sarantis Dimitriadis. .......................... 47 4. Neolithic Sedimentology at Knossos, Maria-Pilar Fumanal García†. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 5. The Economy of Neolithic Knossos: The Archaeobotanical Data, Anaya Sarpaki. ....................63 6. Wood Charcoal Analysis: The Local Vegetation, Ernestina Badal and Maria Ntinou. ................95 7. Plant Economy and the Use of Space: Evidence from the Opal Phytoliths, Marco Madella. ......... 119 8. The Knossos Fauna and the Beginning of the Neolithic in the Mediterranean Islands, Manuel Pérez Ripoll. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 9. The Earliest Settlement on Crete: An Archaeozoological Perspective, Liora Kolska Horwitz. ......171

viii

THE NEOLITHIC SETTLEMENT OF KNOSSOS IN CRETE

10. R  adiocarbon Dates from the Neolithic Settlement of Knossos: An Overview, Yorgos Facorellis and Yiannis Maniatis. ................................................................. 193 11. Knossos and the Beginning of the Neolithic in Greece and the Aegean Islands, Nikos Efstratiou... 201 Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215

List of Tables in the Text

Table 4.1. Correlation of sedimentology samples with excavation levels and cultural phases. . . . . . . . . . . 55 Table 4.2. Munsell color and calculation of statistical parameters of mean size, sorting, skewness, and kurtosis for each of the analyzed sedimentology samples. ..................... 56 Table 5.1. Seed list provided to J.D. Evans by Hans Helbaek (unpublished). ............................... 67 Table 5.2. List of archaeobotanical samples from the 1997 rescue excavation, along with relative and absolute dates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Table 5.3. Aceramic Neolithic archaeobotanical sample E 97(30) from Knossos 1997 level 39, retrieved from 16 liters of water-floated soil. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Table 5.4. M  easurements of Triticum turgidum/aestivum from Aceramic and EN levels at Knossos. .............................................................................................. 71 Table 5.5. Measurements of Triticum dicoccum, Triticum monococcum, Hordeum vulgare, and Lens culinaris. ........ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Table 5.6. E  arly Neolithic I archaeobotanical (seed) samples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Table 5.7. M  easurements of Trifolium spp. and Leguminosae. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Table 5.8. M  easurements of Raphanus cf. raphanistrum and Linum cf. usitatissimum. ................ 78 Table 5.9. Early Neolithic II archaeobotanical (seed) samples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

x

THE NEOLITHIC SETTLEMENT OF KNOSSOS IN CRETE

Table 5.10. V  itis sp. measurements from EN II levels; sketch of a grape seed showing locations of dimensions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Table 5.11. M  iddle Neolithic archaeobotanical (seed) samples. ............................................... 86 Table 5.12. Late Neolithic archaeobotanical (seed) samples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Table 6.1. I nventories of plants growing in different parts of the study area. ............................. 99 Table 6.2. Absolute and relative frequencies of taxa identified in the wood charcoal assemblages from Neolithic Knossos. ............................................................................ 102 Table 6.3. P  resence of plant taxa in wood charcoal assemblages from Neolithic Knossos, along with the total number of fragments analyzed and the total number of taxa identified in each level. ............................................................................ 104 Table 7.1. Knossos 1997: south profile phytolith counts. ................................................... 122 Table 7.2. Knossos 1997: west profile phytolith counts. .....................................................125 Table 8.1. Number of identified and unidentified specimens by taxa and period. . . . . . . . . . . . . . . . . . . . . . . . 135 Table 8.2. M  easurements of bones from Bos taurus. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Table 8.3. M  easurements of bones from Ovis aries (O.a.) and Capra hircus (C.h.). . . . . . . . . . . . . . . . . . . . . . 136 Table 8.4. M  easurements of bones from Sus scrofa domesticus, Sus scrofa ferus, Capra aegagrus, Martes, and Meles meles. ....................................................138 Table 8.5. N  umber of identified specimens of Bos, Ovis/Capra, and Sus with number of marks caused by dog gnawing. ................................................................................. 140 Table 8.6. Early Neolithic I and EN II faunal remains. ...................................................... 141 Table 8.7. Early Neolithic II/MN faunal remains. ........................................................... 142 Table 8.8. M  iddle Neolithic faunal remains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .144 Table 8.9A. L  ate Neolithic faunal remains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 Table 8.9B. Late Neolithic faunal remains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Table 8.10. N  umber of identified specimens of Bos and Ovis/Capra/Sus with burn marks. . . . . . . . . . . . 151 Table 8.11. N  umber of long bone remains (the diaphysis fragments are not counted here), phalanges, and tarsi corresponding to mature and immature bones, along with the number of LN tooth remains, grouped by age, for comparison with the long bones. ......................... 156 Table 8.12. Number of mandibles (NM) for goats and sheep from the LN levels classified by age. . . . . 157 Table 8.13. Number of mandibles (NM) of Bos taurus classified by age. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 Table 8.14. N  umber of maxillary and mandibular remains of Sus scrofa domesticus classified by age. . . . 159 Table 8.15. Number of identified specimens of Bos taurus classified by sex. ............................. 160 Table 8.16. N  umber of identified specimens of Ovis aries (O.a.) and Capra hircus (C.h.) classified by sex. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

LIST OF TABLES IN THE TEXT

xi

Table 8.17. Chronological representation of the faunal species at Knossos. ................................ 161 Table 8.18. P  ercentages of identified specimens of domestic and wild species at Knossos and other sites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Table 8.19. Representation and abundance of various faunal species at Shillourokambos, Ais Yiorkis, and Khirokitia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162 Table 8.20. I ntroduction and chronological representation of wild animals at various sites in Crete.... 163 Table 9.1. Relative frequencies (percentages) of animal species from Knossos. . . . . . . . . . . . . . . . . . . . . . . . 175 Table 9.2. Schematic representation of the relative chronology (cal. b.c. dates) of sites mentioned in the text. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Table 10.1. S  ummary of the British Museum radiocarbon dates on charcoal from the excavations of J.D. Evans at Neolithic Knossos, sorted by age. . . . . . . . . . . . . . . . . . . . . . . . . . . 194 Table 10.2. Description of the samples dated in the British Museum Research Laboratory. . . . . . . . . . . . . 195 Table 10.3. S  ummary of radiocarbon dating results of carbonized samples collected in 1997 from the Neolithic settlement levels at Knossos. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

List of Figures in the Text

Frontispiece. The city, the fortifications, the harbor, and the hinterland of Khandax (Herakleion) in the first half of the 17th century. Map by unknown cartographer, 17th c., Collezione Museo Civico, Padua. Vikelaia Municipal Library, Herakleion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii Figure i. The Minoan palace and its Neolithic past. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxi Figure 1.1. Plan of the Palace of Knossos showing the Central Court and the location of the excavation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Figure 1.2. Trench II: (a) view of the Central Court of the Palace, looking northeast; (b) view of the area of the rescue dig, looking northeast; (c) view looking northeast of the stratigraphy of the upper part of the trench in the souther profile. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Figure 1.3. Plan of the excavation trenches next to the staircase. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Figure 1.4. South and west stratigraphic profiles of the trench. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Figure 1.5. View of excavation level 4, showing hearth in northwest corner of the trench. .............. 5 Figure 1.6. Plans of excavation levels 9 and 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Figure 1.7. Plan of excavation level 12, showing the round kouskouras feature (12A) in southwest corner. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Figure 1.8.

V  iew of excavation level 12, showing kouskouras deposit and feature (12A) in northwest corner. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

xiv

THE NEOLITHIC SETTLEMENT OF KNOSSOS IN CRETE

Figure 1.9. Plan of excavation level 13. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Figure 1.10. Plan of excavation level 14, showing hearth in northwest corner. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Figure 1.11. View of excavation level 14, showing hearth in northwest corner. ........................... 10 Figure 1.12. Plans of excavation levels 15 and 16, showing appearance of walls 1 and 2 running from north to south. ............. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Figure 1.13. View of excavation level 15 from above. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Figure 1.14. View of excavation level 16 from above. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Figure 1.15. View of excavation level 16 facing west section. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Figure 1.16. P  lan of excavation level 17, showing walls 1 and 2 and the first appearance of walls 3 and 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Figure 1.17. View of level 17 facing west section. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Figure 1.18. View of excavation level 18 from above. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Figure 1.19. View of excavation level 19 from above. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Figure 1.20. Plan of excavation level 21, showing walls 3, 4, 5, and 6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Figure 1.21. View of excavation level 21 from above. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Figure 1.22. View of excavation level 23. .... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Figure 1.23. View of excavation level 24, showing wall 7 and grinding stones. . . . . . . . . . . . . . . . . . . . . . . . . . 14 Figure 1.24. Plans of excavation levels 22 and 24. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Figure 1.25. Plan of excavation level 27. ..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Figure 1.26. View of excavation level 28 from above. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Figure 1.27. View of excavation level 28 facing west profile. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Figure 1.28. Plans of excavation levels 29–29a and 30–30a, showing walls and hearths. . . . . . . . . . . . . . . . 17 Figure 1.29. View of level 29A, showing hearths 1, 2, and 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Figure 1.30. View of excavation level 30, showing hearth 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Figure 1.31. View of excavation level 30, showing hearth 4 and the elliptical structure. . . . . . . . . . . . . . . . . 19 Figure 1.32. View of the elliptical stone wall from levels 24–27. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Figure 1.33. View of excavation level 31. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Figure 1.34. Plans of excavation levels 31, 32, and 34, showing hearths 5, 6, and 7. . . . . . . . . . . . . . . . . . . . . . 20 Figure 1.35. Plan of excavation level 37, showing pits 1 and 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Figure 2.1. Sedimentological samples of the middle part of the south profile. . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Figure 4.1. Fine fraction granulometry (%) of the samples from the west profile. . . . . . . . . . . . . . . . . . . . . . 57 Figure 4.2. The organic content (%) of the samples from the west profile. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Figure 4.3. The carbonate content of the samples from the west profile. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

LIST OF FIGURES IN THE TEXT

xv

Figure 4.4. Morphoscopy of sands without acid treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57 Figure 4.5.

Morphoscopy of sands after the elimination of calcareous grains (subsequent to acid treatment). ................ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57

Figure 5.1.

Drawing of Triticum turgidum L./T. aestivum from the 1997 excavations at Knossos. ...69

Figure 5.2.

Graphs of measurements and measurement ratios of Triticum turgidum/aestivum from Neolithic Knossos. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

Figure 5.3.

Measurements of Triticum turgidum/aestivum from Aceramic and Early Neolithic Knossos compared with average values for Erbaba, Ramad, and Bouqras in the Near East. ............................................................................................. 73

Figure 5.4. G  raphs of measurements and measurement ratios of Lens culinaris from Neolithic Knossos. .................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Figure 5.5. Early sites including those from mainland Greece where Triticum turgidum/aestivum is reported: 1. Tell Abu Hureyra, 2. Tell Halula, 3. Tell Aswad, 4. Tell Ghoraife, 5. Tell Sabi Abyad, 6. Servia, 7. Cafer Höyük, 8. Dhali Agridhi, 9. Otzaki, 10. Sesklo, 11. Sitagroi, 12. Haçilar, 13. Aşkli Höyük, 14. Çatal Höyük, 15. Can Hasan, 16. Cayönü, 17. El Kown, 18. Bouqras, 19. Tell Ramad. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 Figure 5.6.

S  ummary of the distribution of all categories of archaeobotanical remains at Neolithic Knossos............................................................................................. 90

Figure 6.1. Climate and topography of Knossos: (a) mean annual precipitation in Crete; (b) topographic map of the area around Knossos; (c) west–east topographic section; (d) southwest–northeast topographic section. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Figure 6.2. View of the Knossos valley from Mt. Juktas showing present-day vegetation. . . . . . . . . . . . . 98 Figure 6.3. Panoramic view of the site of Knossos showing present-day vegetation. .................... 98 Figure 6.4. Present-day phrygana vegetation on the hills in the study area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Figure 6.5. Present-day vegetation on deep soils in the study area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Figure 6.6. Anatomy of plant taxa identified in wood charcoal assemblages from Neolithic Knossos. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 Figure 6.7. Wood charcoal diagram from Neolithic Knossos showing relative frequencies of taxa in successive excavation levels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 Figure 7.1. Bar chart of phytolith per­centage frequencies from the south profile. ..................... 122 Figure 7.2. Bar chart of C3 and C4 phytolith percentage frequencies from the south profile........... 122 Figure 7.3. West profile stratigraphy and sampling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 Figure 7.4. Bar chart of phytolith percentage frequencies from the west profile. . . . . . . . . . . . . . . . . . . . . . 124 Figure 7.5. Bar chart of C3 and C4 phytolith percentage frequencies from the west profile. . . . . . . . . . 125 Figure 7.6. Trench section (southwest corner to west face) with phases identified according to phytolith composition and frequencies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

xvi

THE NEOLITHIC SETTLEMENT OF KNOSSOS IN CRETE

Figure 7.7. Silica skeleton from grass leaf (long cells and a stoma) from the EN I deposits (sample XXa, level 32). ......................................................................... 129 Figure 7.8. Wheat-type silica skeleton from the EN I deposits (sample XXIVb, level 32). ........... 129 Figure 7.9 Silica skeleton from a dicotyledonous plant from the EN I deposits (sample XIV, level 30). ...........................................................................129 Figure 7.10. Millet-type silica skeleton from the EN II deposits (sample Xa, level 16). . . . . . . . . . . . . . . . .129 Figure 8.1. Percentages of the osseous parts of cattle long bones. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 Figure 8.2. Percentages of the osseous parts belonging to the long bones of middle-sized mammals (goats, sheep, and pigs). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 Figure 8.3. Skeletal fragments of long bones of Ovis/Capra from level 14, all with dog-gnawing marks. ............................................................................................. 140 Figure 8.4. Fragments of proximal epiphyses of femur and tibia of Bos taurus with fracture marks caused by impacts from the extraction of marrow, level 24. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 Figure 8.5. Animal bones from level 3: (a) radius diaphysis; (b) radius proximal part; (c) scapula; (d) fragment of femur; (e) phalanx I of Capra aegagrus. .................................... 153 Figure 8.6. Distal metacarpus of Capra aegagrus and Ovis aries. ...................................... 153 Figure 8.7. Animal bones: (a) ulna in lateral view probably belonging to a wild boar (level 23); (b) Sus scrofa ferus: canine fragment (level 10); (c) Sus scrofa domesticus: ulna in lateral view (level 14). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154 Figure 8.8. Diaphysis width range (SD) of Sus scrofa domesticus and Sus scrofa ferus from Zambujal (Portugal), Cerro de la Virgen (Spain), Argissa-Magula, and Knossos. . . . . . . .154 Figure 8.9. Meles meles: (a) left mandible in lateral view (level 14); (b) lower canine (level 14); (c) left ulna in medial views; the proximal epiphysis is not fused (level 3). Martes: (d) distal part of humerus in cranial view (level 9). ............................................................. 155 Figure 8.10. Age classes of the mandibles of Ovis and Capra. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Figure 8.11. Age classes of the mandibles of Bos taurus. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 Figure 8.12. Distal part of metacarpus belonging to a male (possibly ox) of Bos taurus, with osseous deformations on the articular surfaces. ............................................... 158 Figure 8.13. Age classes of Sus scrofa domesticus maxillae and mandibles.. . . . . . . . . . . . . . . . . . . . . . . . . . 159 Figure 8.14. Correlation of the measurements of phalanx I belonging to Bos taurus. . . . . . . . . . . . . . . . . . . . 160 Figure 9.1. Map showing location of sites mentioned in the text: 1. Ashkelon; 2. ‘Ain Ghazal; 3. Atlit Yam; 4. Hagoshrim and Tel Ali; 5. Ras Shamra; 6. Cap Andreas Kastros; 7. Khirokitia; 8. Tenta; 9. Asikli Höyük; 10. Mersin; 11. Can Hasan III; 12. Çatalhöyük; 13. Suberde; 14. Haçilar; 15. Nea Nikomedeia; 16. Argissa-Magula; 17. Sesklo; 18. Achilleion; 19. Franchthi Cave; 20. Sidari, Corfu; 21. Cave of the Cyclops, Youra; 22. Melos; 23. Santorini; 24. Knossos, Crete; 25. Tel Aray 2; 26. Umm el Tlel; 27. Qdeir; 28. El Kowm 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Figure 9.2. Adult male agrimi (Capra aegagrus cretica) showing phenotypic resemblance to the wild bezoar goat. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177

LIST OF FIGURES IN THE TEXT

xvii

Figure 10.1. Distribution of calibrated dates sorted by stratum of the samples from the excavations of J.D. Evans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198 Figure 10.2. Distribution of calibrated dates sorted by depth of the samples from the 1997 archaeological campaign. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198 Figure 10.3. Calibrated radiocarbon dates from the 1997 excavation at Knossos plotted against the depth of the samples in order to determine the accumulation rate of the habitation deposits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198

Preface

The site of Knossos on the Kephala hill in Crete is of great archaeological and historical importance for Greece and Europe. Dating back to 7000 b.c., it is the home of one of the earliest farming societies in southeastern Europe. In later Bronze Age periods, it developed into a remarkable center of economic and social organization within the island, enjoying extensive relations with the Aegean, the Greek mainland, the Near East, and Egypt. Arthur Evans excavated the site at the beginning of the 20th century, and through his extensive and spectacular restoration and reconstruction efforts, he transformed Knossos into one of the most popular archaeological sites in the Old World (Evans 1901, 1921–1935, 1927, 1928). Knossos is now best known among both specialists and the wider public for its unique central building, conventionally called a palace, which is one of the earliest archaeological monuments to have been restored on such a scale. What was not apparent during the early archaeological research at the site was the impressive extent and depth of the earlier habitation that lies under the imposing palace, even though

the laborious work of Arthur Evans and Duncan Mackenzie in the early 20th century had revealed considerable amounts of Neolithic material (Mackenzie 1903). In 1953 Audrey Furness studied and published the Neolithic pottery from Evans’s test soundings with the aim of testing the three “Stone Age” periods discussed by Mackenzie (Furness 1953). The successful work of Furness led the British School at Athens to launch a series of systematic investigations at Knossos, directed by Sinclair Hood and John D. Evans, from 1956 to 1971 (Evans 1964, 1971, 1994; Warren et al. 1968). The well-known Trenches A to C, which were opened in the area of the Central Court of the palace, together with the peripheral soundings X and ZE, confirmed a chronological sequence of 10 strata representing at least 4,000 years of Neolithic occupation, including the still-disputed Aceramic phase. Looking back at the announcement by J.D. Evans (1971) of the first and very early radiocarbon dates for the founding of Knossos (7000 b.c.), I cannot forget the welcome surprise with which these dates were received, and I am very happy to

xx

THE NEOLITHIC SETTLEMENT OF KNOSSOS IN CRETE

see that our recent radiocarbon dates, published in this volume, confirm Evans’s early chronology that was attained without the benefit of our modern technology. Other contributions to our knowledge of the Neolithic of Crete include the work of Richard M. Dawkins at Magasa in eastern Crete in 1905 (Dawkins 1905), the investigations of Angelo Mosso and Doro Levi at Phaistos (Mosso 1908), the publication of the Phaistos material by Lucia Vagnetti (Vagnetti 1972–1973), and the pioneering research at Katsambas by Stylianos Alexiou (1953, 1954). The forthcoming publication of Katsambas by Nena Galanidou and her associates (Galanidou, ed., forthcoming) and the study of the material from older fieldwork at Gerani and Pelekita in the Zakros area, carried out by Yiannis Tzedakis and Costis Davaras, respectively (Tzedakis 1970; Davaras 1979), are expected to offer more data regarding the early occupational horizon of Crete. The recent publication by Valasia Isaakidou and Peter Tomkins of The Cretan Neolithic in Context (Isaakidou and Tomkins, eds., 2008), the latest rescue excavations carried out by the Ephorate of Central Crete in the vicinity of Katsambas, and, most importantly, the announced presence of Mesolithic material on the islands of Crete and Gavdos, show that early prehistoric research in Crete and its immediate environs is a dynamic field of investigation. A series of archaeological test soundings was opened in February 1997 in conjunction with the planning of the course of the main and secondary visitors’ routes through the palace, a process that involved widening the existing paths, establishing new ones, and examining the state of the building’s foundations. The south and east slopes of the Kephala hill were the main focus of investigation (Karetsou 2004; Ioannidou-Karetsou 2006). This research was prompted by the architect Clairy Palyvou’s suggestion to double the width of the modern narrow stone stair leading from this part of the Central Court to the first level of the Grand Staircase, where A. Evans made his last attempt to restore the Medallion Pithoi. The investigation, which lasted five weeks, was carried out under difficult weather conditions and according to a very strict timetable. We were all happily surprised that in an area often disturbed for conservation work in the 1950s

and 1960s, including the opening of rainwater channels, deep pre-Minoan deposits remained intact just a few centimeters under the visitors’ feet. I took this to be a sign of good fortune, since, after three decades of personal, systematic involvement with Minoan archaeology, the dream of my youth to look down to the “Neolithic Cretan time” was becoming a reality. A collaboration with colleagues familiar with the excavation of Neolithic sites and modern data collection and analysis methods was my next immediate concern. The chance to reexamine the succession of Neolithic occupation strata on the Kephala hilltop, some 50 years after the first such investigation at Knossos, presented me with great expectations and challenges. Professor Nikos Efstratiou of the Aristotle University of Thessaloniki contributed greatly to the success of the project, and I would like to take this opportunity to thank him. He was responsible both for the selection of the researchers who gathered at Knossos with very short notice that February and for the coordination of the project. In addition, Professor Giorgos Hourmouziadis, also of the Aristotle Uni­versity of Thessaloniki, was very helpful. Many thanks are due as well to my colleague Dr. Eleni Banou, who participated in the excavation on behalf of the Ephorate, to Nikos Daskalakis, the skilled foreman of the Knossos project, and to the late Andreas Klinis, also a Knossos foreman and a man of rare excavation experience. The general aims of the investigation in the Central Court of Knossos in 1997 were (1) to readdress questions related to the old material and conclusions reached many years ago, and (2) to obtain new data, which, considering the nature of the archaeological site, with the palace standing on top of the Neolithic tell, would have been otherwise impossible. More specific objectives included the careful study of the stratigraphy for the confirmation or revision of the already established Neolithic sequence, the determination of whether the alleged Aceramic phase was represented, the collection of new evidence for the Neolithic ceramic sequence, and the recovery of new archaeozoological and archaeobotanical data and the analysis of their stratigraphic distribution (Efstratiou et al. 2004). Most importantly, the archaeological information was to be gathered and studied using methodologies that were not available in the past—sedimentological

xxi

PREFACE

analyses, which might clarify the occupational gaps in the impressive Neolithic palimpsest, phytolith analyses, ceramic technological analyses, paleoenvironmental observations, and, most significantly, new radiocarbon analyses for the establishment of a reliable sequence of dates. The many archaeological questions relating to the long Neolithic habitation of the Knossos tell had always intrigued me, especially during my 12 years of service (1992–2004) as head of the Knossos Conservation Project. I was impressed by the extent of the Neolithic settlement and the density of the scattered material, especially that of

a

b

c

d

e

f

g

h

the Late and Final Neolithic periods (Fig. i). I was enormously pleased by the opportunity we had to investigate this early Cretan farming community, buried deep under the glorious Minoan palace, and to contribute to its understanding. There is no doubt that the Knossos Neolithic settlement— whether or not it was the first and only one in Crete—constitutes one of the earliest agricultural communities in Greece, and it is also surely the earliest in the Aegean islands. Alexandra Karetsou Honorary Ephor of Antiquities

i

j

k

l

m

n

o

p

q

r

s

A 5280

B 5270

C 5260

D 5250

E 5240

F 5230

G 5220

H 5210

I 5200

J 5190

K 5180

L 5170

M 5160

N 5150

O 5140

P 5130

17840

17830

17820

17810

17800

17790

17780

17770

17760

17750

17740

17730

17720

17710

17700

17690

17680

17670

17660

17650

Q

Figure i. The Minoan palace and its Neolithic past; areas where Neolithic deposits and ceramics are found are indicated with black dots (1997­–2004).

xxii

THE NEOLITHIC SETTLEMENT OF KNOSSOS IN CRETE

References Alexiou, S. 1953. “Ἀνασκαφὴ Κατσαμπᾶ Kρήτης,” Prakt 108 [1956], pp. 299–308.

Furness, A. 1953. “The Neolithic Pottery of Knossos,” BSA 48, pp. 94–134.

———. 1954. “Ἀνασκαφὴ Κατσαμπᾶ Kρήτης,” Prakt 109 [1957], pp. 369–374.

Galanidou, N., ed. Forthcoming. The Neolithic Settlement by the River Kairatos: The Alexiou Excavations at Katsamba.

Davaras, C. 1979. “Σπήλαιο Πελεκητών Ζάκρου,” ArchDelt 34 (B', 2 Chronika), pp. 402–404. Dawkins, R.M. 1905. “Excavations at Palaikastro. IV.2: Neolithic Settlement at Magasá,” BSA 11, pp. 260–268. Efstratiou, N., A. Karetsou, E. Banou, and D. Margomenou. 2004. “The Neolithic Settlement of Knossos: New Light on an Old Picture,” in Knossos: Palace, City, State. Proceedings of the Conference in Herakleion Organised by the British School of Athens and the 23rd Ephoreia of Prehistoric and Classical Antiquities, in November 2000, for the Centenary of Sir Arthur Evans’s Excavations at Knossos (BSA Studies 12), G. Cadogan, E. Hatzaki, and A. Vasilakis, eds., London, pp. 39–51. Evans, A.J. 1901. “The Neolithic Settlement at Knossos and Its Place in the History of Early Aegean Culture,” Man 1, pp. 184–186. ———. 1921–1935. The Palace of Minos at Knossos I– IV, London. ———. 1927. “Work of Reconstruction in the Palace of Knossos,” AntJ 7, pp. 258–266. ———. 1928. “The Palace of Knossos and Its Dependencies in the Light of Recent Discoveries and Reconstructions,” Journal of the Royal Institute of British Architects 36, pp. 90–102. Evans, J.D. 1964. “Excavations in the Neolithic Settlement of Knossos, 1957–60: Part I,” BSA 59, pp. 132–240. ———. 1971. “Neolithic Knossos: The Growth of a Settlement,” PPS 37, pp. 95–117. ———. 1994. “The Early Millennia: Continuity and Change in a Farming Settlement,” in Knossos: A Labyrinth of History. Papers in Honour of S. Hood, D. Evely, H. Hughes-Brock, and N. Momigliano, eds., Oxford, pp. 1–20.

Ioannidou-Karetsou, A. 2006 “Από την Κνωσό μέχρι τη Ζάκρο: H περιπέτεια της προστασίας των ιστορικών αρχαιολογικών χώρων στην κεντρική και ανατολική Κρήτη,” in Conservation and Preservation of the Cultural and Natural Heritage of the Large Islands of the Mediterranean, V. Karageorghis and A. Giannikouri, eds., Athens, pp. 61–76. Isaakidou, V., and P. Tomkins, eds. 2008. Escaping the Labyrinth: The Cretan Neolithic in Context (Sheffield Studies in Aegean Archaeology 8), Oxford. Karetsou, A. 2004. “Knossos after Evans: Past In­ter­ ventions, Present State and Future Solutions”, in Knossos: Palace, City, State. Proceedings of the Conference in Herakleion Organised by the British School of Athens and the 23rd Ephoreia of Prehistoric and Classical Antiquities, in November 2000, for the Centenary of Sir Arthur Evans’s Excavations at Knossos (BSA Studies 12), G. Cadogan, E. Hatzaki, and A. Vasilakis, eds., London, pp. 547–555. Mackenzie, D. 1903. “The Pottery of Knossos,” JHS 23, pp. 157–205. Mosso A. 1908 “Ceramica neolitica di Phaestos e vasi dell’epoca minoica primitiva,” MonAnt 19, pp. 142–228. Tzedakis, I. 1970. “Ἀρχαιλογική Ἐρεύvε Ἀνασκαφί Σπήλαιο Γερανίου,” ArchDelt 25 (B', 2 Chronika), pp. 474–476. Vagnetti, L. 1972–1973. “L’insediamento neolitico di Festòs,” ASAtene 34–35, pp. 7–138. Warren, P., M.R. Jarman, H.M. Jarman, N.J. Shackleton, and J.D. Evans. 1968. “Knossos Neolithic, Part II,” BSA 63, pp. 239–276.

Acknowledgments

The excavators are grateful to all the people who made this publication possible. First and foremost, we would like to express our thanks to Dr. Iordanis Dimakopoulos, former Director of the Conservation and Restoration of Monuments Service of the Greek Ministry of Culture and Tourism. He fully understood the need for the rescue excavation to be carried out, at a time when the visitors’ walkway project at the palace of Knossos was already under way with a tight deadline. Our warmest thanks are also due to the technical staff of the 23rd Ephorate of Prehistoric and Classical Antiquities and the Knossos Conservation Office, who worked through the particularly cold February of 2011. We would especially like to thank Nikos Daskalakis, Stavros Mavrakis, and Michalis Tzobanakis, who constructed a small shelter to protect both staff and trenches from the rain, since the excavation ran from sunrise to sunset. The late Andreas Klinis, a foreman of special skill and astuteness, was the person whom we entrusted with the stratigraphy; he was the only one to work at a depth of four to eight meters. We must also thank Konstantinos Ktistakis for his accurate plans, elevations, and sections of the trenches; Dr. Don Evely, former Knossos curator for the British School at Athens, for his help during the study of the material; and archaeologist Maria Kelaidi, who spent an entire summer in the courtyard of the Villa Ariadne, meticulously sieving the huge amount of soil from the excavation. Vital assistance was also provided by the head guard of the palace of Knossos, Manolis Apostolakis, and the rest of the guard staff. The Ephorate accountants Evangelia Fotaki and Litsa Kafousi also provided their services, without which the project would not have been possible.

xxiv

THE NEOLITHIC SETTLEMENT OF KNOSSOS IN CRETE

Finally, we would like to express our particular thanks for the generosity of Professor Philip Betancourt, who enthusiastically supported the publication of this volume by the INSTAP Academic Press, and to the Director of Publications, Susan Ferrence, for all her efforts in ensuring that the resulting publication was of the highest possible standard. Our warmest gratitude is also, of course, due to all the contributors to the volume.

Introduction Nikos Efstratiou

The construction of a staircase extension in the northeastern part of the Central Court of the Palace of Minos at Knossos prompted the opening of a new excavation trench in 1997. After the systematic excavation of the deep Neolithic occupation levels by J.D. Evans in the late 1950s (1964, 132) and later, more limited investigations of the Prepalatial deposits undertaken primarily during restoration work, no thorough exploration of the earliest occupation of the mound had been attempted. Although our operation was to be swift and limited in extent, we knew that the opening of a trench destined to reach the basal layers of the settlement offered us the opportunity to address many old and new research questions concerning the chronological, socioeconomic, and spatial aspects of Cretan Neolithic society (Evans 1994, 1). Since the time of Evans’s research, excavation techniques and field methods have developed rapidly, and a new, more complex picture of late Pleistocene and early Holocene developments in the Aegean and the eastern Mediterranean has emerged. The chance to reexamine the important

but inconspicuous Neolithic deposits of the Knossos tell afforded both an appealing and a demanding challenge. While the Bronze Age palace dominates the historiography of the site and its archaeological image, the Neolithic settlement at Knossos does not hold the position it deserves in discussions of the early prehistory of the eastern Mediterranean, in part because of the limited research directed toward the early prehistory of Crete and the other Aegean islands. Moreover, the publication of the Neolithic settlement has been confined to a few preliminary though excellent field reports and short studies produced by Professor J.D. Evans and his collaborators (1964, 132; 1971, 95; Warren et al. 1968, 239). When attempted, previous syntheses of this material have been either very cautious analyses of the limited data (Evans 1994, 1) or provocative interpretations containing attractive but ill-founded speculations (Broodbank 1992, 39; Whitelaw 1992, 225). Additional Neolithic material recovered from later small field investigations focusing on Bronze Age deposits has been welcome, but because such

xxvi

THE NEOLITHIC SETTLEMENT OF KNOSSOS IN CRETE

information is scarce, it cannot provide the answers to many open questions (Manteli and Evely 1995, 1). It is fortunate that certain categories of the archaeological material from Evans’s investigations have recently undergone detailed reexamination with respect to issues of spatial organization, ceramic typology and technology, lithics, and faunal remains (Isaakidou and Tomkins, eds., 2008). Despite these new and interesting studies, however, the need for a better understanding of the foundation and development of Neolithic Knossos continues. This impressive and long-lived settlement—one of the very few tells in Greece—is of paramount importance to the history of the eastern Mediterranean and the Near East (Berger and Guilaine 2009). Recent developments in the archaeology of Cyprus and the Aegean islands make the reevaluation of long-held concepts about this region and time period all the more urgent, as discussed in Chapter 11. Although a number of rigorous surface reconnaissance projects have been undertaken in Crete in the past decades, Knossos remains the only early settlement known on the island (Manning 1999, 469). The methodology employed in these allperiod surveys was not specifically designed to locate early sites, however. In the last few years field researchers have become increasingly critical of older methods used to identify traces of early habitation sites, especially in view of the geomorphological complexity of coastal and island areas (Runnels 2003, 121; Ammerman et al. 2006, 1). Until specially designed surface reconnaissance projects are carried out in various coastal areas, the presence of other early occupation sites in Crete remains an open possibility. Thus, the recently reported results of the Plakias Mesolithic Survey in Crete, in which a number of pre-Neolithic sites rich in lithic scatters were identified along the southern coast of the island, do not come as a surprise (Strasser et al. 2010). Indeed, current research in Cyprus indicates that we may encounter more new and unexpected late Pleistocene and early Holocene finds in the eastern Mediterranean (Ammerman 2011). Many older views of early habitation patterns in the Aegean islands should now be treated with skepticism (Cherry 1990, 145). The newly found Mesolithic habitation remains along the south coast of the island may ultimately support claims of a missing Early Neolithic (EN)

horizon in Crete. In the meantime, the apparent uniqueness of Knossos within the island is hard to accept in cultural terms, and as we shall see in later chapters, such a perception is undermined, albeit indirectly, by the material remains (pottery, subsistence) from Knossos, along with other evidence. The key importance of Knossos, however, for documenting the beginning of farming in the Aegean and mainland Greece, whether as a distinctive stage within a westward mobility pattern of human groups or as a well-planned colonization episode involving specific Aegean islands, remains undiminished. At present the notion of a local transition to farming in Crete undertaken by a dynamic Mesolithic population seems improbable, as is the case in continental Greece, where the archaeological evidence for the arrival of new farming groups seems overwhelming (Perlès 2001). Neolithic Knossos is also important, as suggested above, in the wider geographic context of the early island prehistory of the eastern Mediterranean. Recent discoveries on the island of Cyprus have revealed the presence of a number of pre-Neolithic inland and coastal sites, triggering an interesting debate about a possibly early date for the occupation of the largest eastern Mediterranean islands and the interpretation of this phenomenon as a historical process with its own distinctive cultural, technological, and ideological characteristics (Broodbank 2006; Ammerman 2010). Mounting archaeological evidence from the Aegean either supports or at least allows us to entertain a new picture of early island settlement (Sampson 2006). In this context, the founding of the early seventh millennium b.c. farming village of Knossos on the Kephala hill may still be viewed either as the result of a long pre-Neolithic process of development on the island or as the start of an intrusive occupation by farmers from the east. Archaeological evidence from the long stratigraphic sequence of the Knossos tell may be called upon to interpret this ambiguous cultural process. Indeed, in relation to mainland Greece, specific material evidence from Knossos, such as the EN sequence of pottery (fabric types, surface treatment), attests to idiosyncratic elements of a local island development (see Dimitriadis, this vol., Ch. 3). It is still too early to argue whether these characteristics should be interpreted as the outcome of island isolationism

INTRODUCTION

and endogenous developments in Crete or as the manifestation of a more generalized and longstanding Aegean island cultural tradition. The former would undoubtedly have resulted in a number of other distinctive material features and perhaps oddities that we may search for in the archaeological record. Both in terms of a pre-“historical” reconstruction and as far as the archaeology of the site itself is concerned, our endeavor entails a constant shift between different scales (“macro,” “micro”) and genres of field inquiry (e.g., use of space, radiocarbon dating, abandonment phases, faunal changes, pottery changes). The small size of our 1997 dig admittedly limits the overall representational validity of our findings at the site, but this does not deter us from addressing some of the broader issues mentioned above. We are particularly hopeful that the new studies presented here— sedimentology, phytoliths, anthracology, ceramic technology—together with the critical reevaluation of the other categories of material remains, such as the fauna and archaeobotany, will provide new and meaningful insights into the cultural sequence of

xxvii

the Knossos settlement. The documentation of the tell’s stratigraphic sequence, which has a depth of more than 8 m, along with its comparison to the old and well-established succession of Evans’s strata (Efstratiou, this vol., Ch. 2), also contributes to these insights, as does the the newly obtained series of radiocarbon dates from accelerator mass spectrometry (AMS), which seems to corroborate the existing chronological framework (Facorellis and Maniatis, this vol., Ch. 10). All of the categories of material remains with the exception of the pottery are analyzed and presented in the following chapters of the monograph. The detailed study of the ceramics is still in progress and will appear in a separate volume. The contributors wish to underline the contingent nature of their results and syntheses, which are constrained by the limited area of the field investigation. Nevertheless, we hope that the rigor employed in the data collection, the meticulous study of the finds, the constant cross-checking with J.D. Evans’s record, and our final synthesis will balance this unavoidable difficulty.

References Ammerman, A.J. 2010. “The First Argonauts: Towards the Study of the Earliest Seafaring in the Mediterranean,” in Global Origins (and Development) of Seafaring, A. Anderson, J. Barrett, and K. Boyle, eds., Cambridge, pp. 81–92. ———. 2011. “The Paradox of Early Voyaging in the Mediterranean and the Slowness of the Neolithic Transition between Cyprus and Italy,” in The Seascape in Aegean Prehistory (Monograph of the Danish Institute at Athens 14), G. Vavouranakis, ed., Athens. Ammerman, A.J., P. Flourentzos, C. McCartney, J. Noller, and D. Sorabji. 2006. “Two New Early Sites on Cyprus,” RDAC 2006, pp. 1–22. Berger, J-F., and J. Guilaine. 2009. “The 8200 cal bp Abrupt Environmental Change and the Neolithic Transition: A Mediterranean Perspective,” Quaternary International 200, pp. 31–49. Broodbank, C. 1992. “The Neolithic Labyrinth: Social Change at Knossos before the Bronze Age,” JMA 5, pp. 39–75.

———. 2006. “The Origins and Early Development of Mediterranean Maritime Activity,” JMA 19, pp. 199–230. Cherry, J.F. 1990. “The First Colonization of the Med­ i­terranean Islands: A Review of Recent Research,” JMA 3, pp. 145–221. Evans, J.D. 1964. “Excavations in the Neolithic Settlement of Knossos, 1957–60: Part I,” BSA 59, pp. 132–240. ———. 1971. “Neolithic Knossos: The Growth of a Settlement,” PPS 37, pp. 95–117. ———. 1994. “The Early Millennia: Continuity and Change in a Farming Settlement,” in Knossos: A Labyrinth of History. Papers in Honour of S. Hood, D. Evely, H. Hughes-Brock, and N. Momigliano, eds., Oxford, pp. 1–20. Isaakidou, V., and P. Tomkins, eds. 2008. Escaping the Labyrinth: The Cretan Neolithic in Context (Sheffield Studies in Aegean Archaeology 8), Oxford.

xxviii

THE NEOLITHIC SETTLEMENT OF KNOSSOS IN CRETE

Manning, S.W. 1999. “Knossos and the Limits of Settlement Growth,” in Meletemata: Studies in Ae­ gean Archaeology Presented to Malcolm H. Wiener on the Occasion of His 65th Birthday (Aegeum 20), P. Betancourt, V. Karageorghis, R. Laffineur, and W.-D. Niemeier, eds., Liège, pp. 469–482. Manteli, K., and D. Evely. 1995. “The Neolithic Levels from the Throne Room System, Knossos,” BSA 90, pp. 1–16. Perlès, C. 2001. The Early Neolithic in Greece: The First Farming Communities in Europe, Cambridge. Runnels, C. 2003. “The Origins of the Greek Neolithic: A Personal View,” in The Widening Harvest: The Neo­­ lithic Transition in Europe. Looking Back, Looking Forward (Colloquia and Conference Papers 6), A.J. Ammerman and P. Biagi, eds., Boston, pp. 121–133.

Sampson, A. 2006. Προϊστορία του Αιγαίου, Athens. Strasser T.F., E. Panagopoulou, C.N. Runnels, P.M. Murray, N. Thompson, P. Karkanas, F.W. McCoy, and K.W. Wegmann. 2010. “Stone Age Seafaring in the Mediterranean: Evidence from the Plakias Region for Lower Palaeolithic and Mesolithic Habitation of Crete,” Hesperia 79, pp. 145–190. Warren, P., M.R. Jarman, H.N. Jarman, N.J. Shackleton, and J.D. Evans. 1968. “Knossos Neolithic, Part II,” BSA 63, pp. 239–276. Whitelaw, T.M. 1992. “Lost in the Labyrinth? Comments on Broodbank’s Social Change at Knossos before the Bronze Age,” JMA 5, pp. 225–238.

1

The Excavation Nikos Efstratiou, Alexandra Karetsou, and Eleni Banou

The Archaeology of the Levels Two trenches were opened in the northeastern corner of the Central Court of the palace in Feb­ ruary of 1997 in a salvage operation that lasted for five weeks (Figs 1.1–1.3).* Trench I, a 2.0 x 2.0 m sloping area, was the first to be opened, but its excavation was stopped as soon as we realized that the area was heavily disturbed and had been used by previous investigators of the site as a dumping place for archaeological materials such as stones and broken pottery. We then shifted our efforts to Trench II (Fig. 1.3). The rescue character of the dig entailed a strict timetable for the completion of the work, which ultimately dictated many of the methodological decisions taken in the course of the excavation. Moreover, it was not only time that was limited but also space, since the area available for investigation *All photos by N. Efstratiou; all drawings prepared by K. Kondogiannis. M. Ntinou drew Fig. 1.8. Abbreviations used in this chapter are: Ch(s). Chapter(s) EN Early Neolithic

was extremely small in size, squeezed between the quarters of the palace (Room of Medallion Pithoi) and modern constructional remains of the Central Court staircase (Fig. 1.2:a, b). Nevertheless, in view of the importance of the excavation, all precautions were taken to safeguard the quality of the fieldwork. At the outset Trench II covered an area of 3.0 by 2.0 m, and it reached a depth of 8.0 m (Figs. 1.2:c, 1.3). Due to pressure for the conclusion of the dig, the excavation area below the depth of 4.5 m was restricted to an area of 1.5 x 1.5 m in size (Fig. 1.4). This was an undesirable situation for many reasons, including the practical difficulties of carrying out a dig in such a narrow and deep shaft where there was a lack of light, which complicated the tasks of making stratigraphic observations, FN LN m MN no(s).

Final Neolithic Late Neolithic meters Middle Neolithic number(s)

2

NIKOS EFSTRATIOU, ALEXANDRA KARETSOU, AND ELENI BANOU

m

Figure 1.1. Plan of the Palace of Knossos showing the Central Court and the location of the excavation.

recording, and sampling. These problems became even more acute when the important lower Early Neolithic (EN) I and Aceramic stratigraphic levels (30–39) at 4.5–8.0 m of depth had to be documented with the utmost accuracy. Their full reconstruction was attempted with particular emphasis on architectural remains and other minor but meaningful stratigraphic features (Fig. 1.4). Despite all the difficulties, every effort was made to prepare a comprehensive study of the data through the systematic collection of samples from almost all levels for archaeobotanical, faunal, phytolith, and anthracological (charcoal) analyses as well as for radiocarbon dating. Flotation and dry sieving were employed to retrieve archaeobotanical and anthracological remains. These efforts proved particularly useful because they provided

material for comparative studies with the Middle (MN) and Late Neolithic (LN) upper levels of the deposit (levels 1–13). Moreover, the sampling of all the levels of the western and southern profiles of Trench II for sedimentological analyses was meticulously carried out by the late MariaPilar Fumanal García and Ernestina Badal of the University of Valencia. The deposit of Trench II covers the Neolithic settlement’s entire occupation from the Aceramic to the LN period. A general trend observed within the whole of the excavated area is the paucity of finds such as tools, lithics, and other portable objects. The Final Neolithic (FN) phase, identifiable mainly by certain ceramic traits, is not represented in the trench. The deposit was excavated by 39 arbitrarily defined spits, hereafter called excavation

3

THE EXCAVATION

b

c

a

Figure 1.2. Trench II: (a) view of the Central Court of the palace, looking northeast; (b) view of the area of the rescue dig, looking northeast; (c) view looking northeast of the stratigraphy of the upper part of the trench in the southern profile; the positions of the samples taken for sedimentological analysis are also shown.

-0.00

TRENCH I -2.07

-2.29

4

TRENCH II

Figure 1.3. Plan of the excavation trenches next to the staircase.

5

NIKOS EFSTRATIOU, ALEXANDRA KARETSOU, AND ELENI BANOU

4

DEPTH 1m

SOUTH

WEST LEVEL 1 2

LN

3 4

2m

MN

9 10 12 13 14a

3m

wall 1 (EN II) 16

EN II 4m

wall 3 (EN II)

20 21 23 26 + 27 28 + 29

floor

30 floor

5m

31

floor EN I (late)

32

NOT EXCAVATED

6m

33 34

7m EN I

35 36 37

ACERAMIC

38 39

8m

BEDROCK Black soil with pieces of charcoal and pottery

Burned layer

Light brown soil

Red soil Soft yellow limestone or kouskouras

Dark brown soil

Yellow limestone (washed kouskouras)

Figure 1.4. South and west stratigraphic profiles of the trench, with indications of depths, excavation levels, soil characteristics, cultural periods, and architectural features.

THE EXCAVATION

levels, which varied in thickness depending on their various microstratigraphic features (soil, sloping surfaces, structural and spatial characteristics) and, especially in the trench’s lower section where visibility was low, the digging conditions. A general description of the different excavation levels (1–39), including their main stratigraphic, occupational, architectural, and cultural features, will follow below. For intrinsic and practical reasons the grouping of a number of levels has been attempted. Often the individual levels were not very thick, and it was difficult to excavate them in the deeper parts of the deposit (levels

5

31–36). Some levels have been subdivided in order to divide the excavation area into limited and welldefined excavation units or loci, with designations such as 9A or 10B. The allocation of all 39 levels and their features to specific chronological and cultural periods and subphases (Aceramic, EN I, EN II, MN, LN) is based on ceramic criteria and follows the site’s well-established typological sequence (Mackenzie 1903; Furness 1953; Evans 1964). A full account of the different periods will be presented in Chapter 2, where the overall sequence of Knossos is discussed.

Excavation Stratigraphy Levels 1–3 (LN) No architectural remains were found in the upper part of the deposit (Fig. 1.4). The archaeological deposits of levels 1 and 2 were not uniform across the excavated area. A thick layer of soft limestone (kouskouras), which was probably used as building material, appeared in level 3 in the central area and along the southern part of the trench. A row of stones began to appear in the north and west part of the excavated area. These surface levels were mostly mixed with debris probably deriving from leveling activities in this part of the Central Court. The deposit in the north part of the trench was clayish and sandy, with scarce finds, while the soil in its southern part was black and contained a considerable number of bones and small pieces of pottery. The Neolithic pottery increased considerably in level 3, however, along with food refuse (animal bones, seashells). The pottery was mainly burnished and is diagnostic of the LN period.

is apparently related to the formation of the upper part of the Neolithic tell. It is a characteristic that is present in all of the stratigraphic sections from the Central Court, which were published by J.D. Evans (1964, fig. 4). Levels 4–8 were marked by the appearance of the first substantial architectural remains, consisting of stones, dissolved kouskouras, and a number of clay “structures” (Fig. 1.5). Most of the architectural features appeared to continue to a considerable depth, reaching level 7. Unfortunately, the character of these apparent structures could not be determined because of their diffused outline. Nevertheless, a hearth was detected at the northwestern end of the trench. It was formed by four medium-sized stones symmetrically arranged around a small empty space, and it measured 0.40

Levels 4–8 (wMN) The general view of Trench II shows that several successive layers of light-colored building material (kouskouras) and very dark soil containing archaeological remains were encountered during the excavation (Fig. 1.2:c). This material, which was also observed in the upper part of the trench, continued through level 4 (Fig. 1.5). The west to east sloping pattern of these layers is clear-cut and

Figure 1.5. View of excavation level 4, showing hearth in northwest corner of the trench.

6

NIKOS EFSTRATIOU, ALEXANDRA KARETSOU, AND ELENI BANOU

x 0.60 m. At a slightly lower depth, ranging from 1.98 to 2.18 m, two postholes were also found. They were situated 0.20 m south of the hearth. They measured 0.50 x 0.53 m and were covered by a very reddish soil (Munsell 5YR 6/8). The posthole was coated with clay that was burned in places, probably because it came in contact with the burned pole. The area marked by the hearth and the two postholes could be considered the first habitation level of the deposit. One is tempted to suggest that these features were part of the interior space of a house with walls that extended beyond the limits of the trench. They were removed during the excavation of levels 5–8 in the middle of the trench, and they therefore are not shown in the stratigraphic section (Fig. 1.4). The pottery that belonged to levels 4–8, although not particularly diagnostic, is ascribed to the Middle Neolithic time period.

Levels 9–11 (MN) Levels 9 and 10 looked quite different from the preceding layers; they appeared denser in texture and richer in charcoal and animal bones but not seashells (Fig. 1.4). There was also a noticeable improvement in the quality of the ceramics, which had intense burnishing. Decorated pieces— mainly incised and a few rippled—became more abundant. The pottery found in these levels belongs to the MN period. A concentration of ashes was located under a small projection that was left unexcavated in the southeastern corner of the trench (Fig. 1.6). Its outline resembled that of a pit, and its contents— mainly ashes—were removed for flotation (level 9A). A deposit of kouskouras excavated in the southwestern corner of the trench was designated as level 11.

Level 12 (MN) This level covered the entire area of the trench, and it was fairly thick (Figs. 1.4, 1.7). Some of the typical characteristics of the previous levels 9 and 10, a black layer uniform in color and texture and rich in finds, remained unchanged for a considerable depth. After the complete removal of level 12 large white surfaces appeared, and they were probably the remains of structures or floors (Fig. 1.8).

The apparent shape of these surfaces or structures is misleading because it is the result of different degrees of preservation of the now decayed kouskouras that was used as building material in various places. Of course, this does not exclude the possibility that some surfaces might have been part of a structure, such as the round platform-like feature made of kouskouras that was excavated as level 12A (Fig. 1.8). A large quantity of pottery with typical MN shapes and decoration was found.

Level 13 (MN) At this level the thick layer of kouskouras, which extended over the entire extent of the trench, was removed. It was the first time that kouskouras was discovered in large lumps that did not contain any finds; it was most probably a pure dissolved building material. The deposit varied in consistency across the trench, however (Fig. 1.9). In certain places the kouskouras had a more anthropogenic character and a brownish color due to the presence of many pieces of charcoal. The interpretation of the nature and function of these surfaces is problematic given the general sloping of the layers and their limited exposure. The ceramic evidence suggests that level 13 marks the stratigraphic transition between the MN and EN II periods.

Level 14 (EN II) The thickness of Level 14 was quite substantial, and it exhibited patches of heavily burned animal bones and charcoal (Fig. 1.4). The deposit located in the western side of the trench was characterized by layers that alternated between charcoal, ashes, and kouskouras. The gradual removal of the darkcolored layer in the western area revealed a rather solid layer of kouskouras with stones aligned in a northwest direction. A hearth in the form of a round hollow space dug inside the kouskouras layer, with reddish traces of burning, was discovered close to the west profile of the trench (Figs. 1.10, 1.11). The first substantial architectural remains in the trench, which seem to continue down to level 27, were first noted in level 14. Their full description follows below. The pottery from the levels

7

THE EXCAVATION

-2.47

-2.34

-2.09

-2.40

0

LEVEL 9

0.5

1

2m

-2.29

-2.34

-2.09

-2.18

LEVEL 10

Figure 1.6. Plans of excavation levels 9 and 10.

0

0.5

1

2m

NIKOS EFSTRATIOU, ALEXANDRA KARETSOU, AND ELENI BANOU

8

-2.60

-2.95

-2.86

-2.77

-2.45

LEVEL 12

0

0.5

1

2m

Figure 1.7. Plan of excavation level 12, showing the round kouskouras feature (12A) in southwest corner.

associated with these features, including level 14, belongs to the final stages of the EN II period. The only previous noticeable architectural features—a fireplace and postholes—belonged to the MN period (levels 4–8), as discussed above.

Levels 15–16 (EN II) Level 15 comprises the soil deposit removed from the center of the trench (Fig. 1.12). The level is situated between the burned patches of kouskour­as of level 14 in the western section of the trench and a row of solid lumps of kouskouras— locally called kouskouropetres—and hard gray stones (sideropetra), which run parallel north to south (Fig. 1.13). The deposit of level 15 was very rich in charcoal remains as well as small stones that may have constituted a pavement. Level 16 refers to the removal of the fireplace already found in the previous level (14) along the western section of the trench, together with the dissolved kouskouras material from the same area (Figs. 1.14, 1.15). Finds were almost nonexistent, suggesting that the kouskouras was used solely as

Figure 1.8. View of excavation level 12, showing kouskouras deposit and feature (12A) in northwest corner.

a building material. The pottery belongs to the EN II period.

Levels 17, 18, and 18A (EN II) After the removal of level 16, the uncovering of the remains of the two almost parallel rows of stones running north to south in the western and

9

THE EXCAVATION

-3.02

-2.67

-2.85

-2.55

-2.94

0

LEVEL 13

2m

1

0.5

Figure 1.9. Plan of excavation level 13.

-3.31

hearth

-3.28

-3.03

-2.85

-2.89

-2.69 -3.25

LEVEL 14

0

0.5

Figure 1.10. Plan of excavation level 14, showing hearth in northwest corner.

1

2m

NIKOS EFSTRATIOU, ALEXANDRA KARETSOU, AND ELENI BANOU

10

hearth

Figure 1.11. View of excavation level 14, showing hearth in northwest corner (bottom left of photo).

eastern areas of the trench was completed (Fig. 1.16). These were defined as walls 1 and 2 (0.30 m and 0.20 m wide, respectively). The distance between the two walls was 0.65 m. It is difficult to know whether they were parts of one larger solid structure or if they constituted two independent stone features (Fig. 1.17). The latter seems more plausible since wall 2 proved to have been founded deeper than wall 1. The construction material with which the upper part of wall 2 was made, either pisé or mudbrick, was found dissolved all over the main central area of the trench, and it had already been removed in level 15 (see above). The architecture in the trench thus is complex. It includes a concentration of stones adjoining the west profile and shown in the stratigraphic section of the trench (Fig. 1.4), walls 1 and 2, and a long, narrow corridor-like area between the walls (Fig. 1.18). The pottery (largely nondiagnostic) from level 17 remained unchanged from level 14, but its volume was reduced significantly. Level 17 seems to have been a rather pure architectural stratum. It contained building material and was strikingly different in composition from the preceding rich black layers. Wall 1 in the western part of the trench was made of large solid lumps of kouskouras, with stones and clay (kouskourohoma) used as binding material. The much narrower wall 2 differed in construction, with more stones, less kouskouras, and the use of clay and straw pisé as a binding material. The investigation of wall 2 ended with the removal of a yellowish, sterile, loose deposit with few finds in the area of the trench’s eastern profile.

At the base of level 17 the architectural picture of the area became even more interesting, though not less enigmatic. The upper part of a massive elliptical stone structure attached to wall 2 began to appear, along with the first traces of another wall positioned south of it (Fig. 1.16, walls 3 and 4). This massive stone structure eventually became the trench’s dominant architectural feature, and it is described in detail in the next section. The clayish sterile layer from the corridor between walls 1 and 2 constitutes levels 18 and 18A. It probably represents the remains of the upper structures of both these walls. The pottery belongs to the EN period, but more precise distinctions cannot be made (EN II?).

Levels 19–21 (EN II) The removal of level 18 in the corridor-like area revealed a dark, blackish deposit (level 19) with pieces of charcoal and some hard finished surfaces, probably the remains of a floor (Fig. 1.19). In order to facilitate our work in such a restricted trench we had to remove walls 1 and 2 (level 20) after they had been recorded, drawn, and photographed. Their construction was characterized by the use of solid lumps of kouskouras (kouskouropetres) and hard gray stones (sideropetres) with pisé and pure kouskouras as binding material. The exposure of level 21 revealed a densely built area in most parts of the trench (Figs. 1.20, 1.21). A row of stones that could have been part of a roughly made wall running east–west continued underneath wall 1; this was designated wall 5 (Fig. 1.20). Similarly, below wall 2 a stone feature (wall 6) ran in a north–south orientation (Fig. 1.21). Nondiagnostic pottery occurred in small quantities. The full profile of the elliptical stone structure (wall 3) already present in level 18 (if not earlier, as the upper stones were first spotted near wall 1 in level 15) was gradually revealed. It included some fallen stones along the eastern section of the trench that were noted above (level 21; Fig. 1.21). Its total exposure involved the removal of a number of levels (18–27). The elliptical structure is considered at length in the discussion of architectural remains at the end of this chapter, but two points should be given special attention at present: its long life and

11

THE EXCAVATION

-2.86

-3.20

-3.10 -3.23 -3.26 -2.93

-3.16

LEVEL 15

0

0.5

1

2m

0

0.5

1

2m

-3.33 WALL 1

2

-3.09

-3.30

LEVEL 16

Figure 1.12. Plans of excavation levels 15 and 16, showing appearance of walls 1 and 2 running from north to south.

NIKOS EFSTRATIOU, ALEXANDRA KARETSOU, AND ELENI BANOU

12

Figure 1.13. View of excavation level 15 from above.

Figure 1.14. View of excavation level 16 from above.

careful construction. It should also be stressed that the elliptical structure does not seem to have been functionally linked to the nearby wall 2.

Levels 22–24 (EN II) It was decided that the excavation should continue along the western half of the trench, leaving the eastern section and the elliptical structure for future investigation (Figs. 1.22–1.24). Excavation levels 22, 23, and 24 were black, often burned layers that contained many pieces of charcoal and

Figure 1.15. View of excavation level 16 facing west section.

A´ -3.57 WALL 1

WALL -3.71 3 4

-3.30

2

-3.60 -3.75

A

-3.09 -3.45

-3.59

-3.59

-3.79 -3.28

LEVEL 17

0

0.5

1

Figure 1.16. Plan of excavation level 17, showing walls 1 and 2 and the first appearance of walls 3 and 4.

2m

13

THE EXCAVATION

Figure 1.18. View of excavation level 18 from above.

Figure 1.19. View of excavation level 19 from above.

Figure 1.17. View of level 17 facing west section.

WALL -3.66

3

4

WALL 6

WALL 5

-3.70

-3.78

LEVEL 21

0

Figure 1.20. Plan of excavation level 21, showing walls 3, 4, 5, and 6.

0.5

1

2m

14

NIKOS EFSTRATIOU, ALEXANDRA KARETSOU, AND ELENI BANOU

wall 6

Figure 1.21. View of excavation level 21 from above.

were rich in finds (Fig. 1.22). A flimsy wall (wall 7) was discovered in the southern part of the trench, together with two discarded grinding stones (Figs. 1.23, 1.24). The pottery was not diagnostic for the most part, except for the material from level 24, which was typical of the EN II period.

Levels 25–27 (EN II) The excavation in the western part of the trench stopped at level 24. Our efforts then shifted to the eastern section, where levels 25, 26, and 27 were removed (Fig. 1.25). The deposit was rich in charcoal remains, animal bones, and pottery. The succession of levels 25–27 along the western profile suggested that they were hard surfaces, possibly successive floors of clayish texture, with dissolved pisé and kouskouras. The pottery dates to the EN II period.

Figure 1.22. View of excavation level 23.

Levels 28–30 (EN II and EN I) Level 28 comprised the deposit from the main part of the trench, west of the massive stone wall (Figs. 1.26, 1.27). The deposit consisted of black soil, rich in charcoal, with large amounts of animal bones but little pottery. It represents a pure habitation deposit and more particularly the rich remains of a hearth made of stones (Fig. 1.28, no. 1). The contents of the hearth were carefully collected for flotation. The excavation of level 29 continued in the same area and revealed more such features (Figs. 1.28, 1.29). A second hearth (no. 2), surrounded by a thick and intensely burned deposit, lay in close proximity to the first one, and probably represented

Figure 1.23. View of excavation level 24, showing wall 7 and grinding stones.

its earlier phase. Its contents were collected as level 29A. The remains of a third hearth (no. 3) were found under the foundation of the elliptical wall. All three hearths were made of stones and covered by black burned earth. Another layer of deposit from the central area of the trench, excluding its eastern part where the

15

THE EXCAVATION

WALL

-3.92

3 WALL

4

6

-3.92

-3.95

0.5

0

LEVEL 22

1

2m

WALL

-4.10

3

4

-4.12 WALL

7

grinding stones -3.96

LEVEL 24 Figure 1.24. Plans of excavation levels 22 and 24.

0

0.5

1

2m

NIKOS EFSTRATIOU, ALEXANDRA KARETSOU, AND ELENI BANOU

16

-4.20

WALL

-4.20

3

-4.09

4

-4.18

LEVEL 27

0

0.5

1

2m

Figure 1.25. Plan of excavation level 27.

Figure 1.26. View of excavation level 28 from above.

elliptical wall is situated, was designated level 30. In this level a fourth large hearth (Figs. 1.28, no. 4, 1.30) appeared directly beneath the area where nos. 1 and 2 were situated, obviously marking an earlier habitation phase. Moreover, a group of stones in a rectangular formation appeared along the southern part of the trench (see below, Fig. 1.34; these will be discussed in conjunction with the level 31). The pottery of level 30 was of extremely good quality, with many pieces bearing incised and plastic decoration (EN I).

Figure 1.27. View of excavation level 28 facing west profile.

17

THE EXCAVATION

Hearth Hearth

1

3

WALL -4.44

29a

4

3

-4.51 Hearth 2 -4.50

LEVEL 29–29a

0

0.5

1

2m

WALL Hearth

4

4

3

30a

-4.58

LEVEL 30–30a

0

0.5

1

Figure 1.28. Plans of excavation levels 29–29a and 30–30a, showing walls and hearths.

2m

18

NIKOS EFSTRATIOU, ALEXANDRA KARETSOU, AND ELENI BANOU

Figure 1.29. View of level 29A, showing hearths 1, 2, and 3.

Figure 1.30. View of excavation level 30, showing hearth 4.

The overall functional relationship between these four adjoining and successive hearths and the elliptical structure is not clear (see below for discussion). It should be noted that by the end of levels 28 and 29 the full profile of the elliptical structure, with a total surviving height of 0.70 m, was exposed (Figs. 1.31, 1.32).

representational validity of our findings was limited even more, especially given the already fragmentary nature of the architectural and spatial evidence. Practically, we were confronted with difficulties in the actual digging, recording, and sampling of the deeper archaeological deposits with limited light and visibility; some levels ended up being quite thick (0.50 m) because of the space restrictions. Nevertheless, we are content that most of these constraints were dealt with in the most satisfactory way possible. The total thickness of levels 31–36 reached almost 3 m. A series of three hearths (nos. 5, 6, and 7) was revealed (Figs. 1.33, 1.34). Two of the fireplaces were found in level 31 aligned along the north baulk of the trench (Fig. 1.34); the third belonged to level 32. Two of them were found along the north profile, and the third (no. 7) was attached to the elliptical wall (Fig. 1.33). These levels were not uniform in color or content (Fig. 1.4). Grayish soil with burned material and scattered stones, pieces of charcoal, pottery, and the refuse of habitation floors alternated with successive thin bands of yellowish soil, probably

Levels 31–36 (EN I) After the removal of level 30, the excavation took a different course because of various difficulties: including the presence of the massive elliptical wall, which was occupying most of the eastern half of the trench; the problematic situation along the western section where the excavation had already stopped at level 18; and the pressure to complete the rescue excavation in a timely manner. It was decided that the excavation should focus on a more limited area in the western half of the trench. The new trench measured 1.50 x 1.50 m (Fig. 1.4). This unfortunate decrease in the excavation area had serious consequences. In theory, the

19

THE EXCAVATION

Figure 1.31. View of excavation level 30, showing hearth 4 and the elliptical structure.

Figure 1.32. View of the elliptical stone wall from levels 24–27.

Level 37 (EN I)

Figure 1.33. View of excavation level 31.

dissolved kouskouras from decayed walls. Some examples of sun-dried bricks with imprints of thatch were found; unfortunately the limited area exposed did not allow us to define any of the structures with which they might have been associated. It is fair to suggest, however, that although no specific architectural remains were recorded, the deposit overall probably represents the refuse of a house or structure and its contents. Certain levels were extremely rich in ceramics, most of them diagnostic of EN I. Levels 33 and 34, for example, yielded sherds with typical EN I offset rims and rippled decoration.

Level 37 is considered the earliest EN I occupation phase of Knossos, and despite its limited extent it revealed interesting new features: two pits covered by reddish soil and many stones, perhaps filling debris (Fig. 1.35). Pit 1 was found in the northwest corner of the trench, and it was completely devoid of pottery. Pit 2 was set close to the elliptical wall of the eastern part of the trench; its deposit produced some pottery similar to that of level 36, together with a few animal bones. The deposit of level 37 was scattered with pieces of mudbricks bearing imprints of straw.

Levels 38–39 (Aceramic) Levels 38 and 39 represent the basal layers of the Neolithic occupation (Fig. 1.4). The deposit was poor in finds except for some pieces of obsidian and dissolved unbaked mudbrick. It was similar in content to level 37, except that it contained no pottery. Levels 38 and 39 have thus been assigned to the Aceramic phase.

Notes on the Architectural Remains The criteria for identifying domestic structures as “houses” are the presence of well-defined ground plans, the remnants of building materials,

and habitation debris associated with inside spaces. The presence of hearths and pits or rich deposits of pottery and other finds related to domestic

NIKOS EFSTRATIOU, ALEXANDRA KARETSOU, AND ELENI BANOU

20

WALL

-4.76

-4.73 Hearth

5

3

Hearth

4

6

-4.86

-4.54

-4.74

Hearth

0.5

1

2m

0

0.5

1

2m

0

0.5

1

2m

0

LEVEL 31

WALL

7

3

4

-5.09

-4.89 -4.54

LEVEL 32

B

-8.04

-8.05

-8.14

-8.08

-8.18

LEVEL 34

B´ Figure 1.34. Plans of excavation levels 31, 32, and 34, showing hearths 5, 6, and 7.

21

THE EXCAVATION

-7.53

-7.41

PIT PIT

1

2

-7.33

-4.54

LEVEL 37

0

0.5

1

2m

Figure 1.35. Plan of excavation level 37, showing pits 1 and 2.

structures may be interpreted with less certainty, since these features and finds could easily have been associated with outdoor spaces and activities as well. Open-air spaces such as courtyards, street lanes, or corridors were most likely lively and multifunctional areas for a Neolithic community. The identification of interior and exterior spaces at Knossos was hampered by the limited scale of our excavation and the lack of contextual evidence. Nevertheless, in the long stratigraphic sequence of the Neolithic deposits investigated we were occasionally able to isolate structures and features relevant to the criteria noted above. Unfortunately, any definitive interpretation of our finds is problematic, although the architectural features documented by J.D. Evans in his campaigns were useful for comparison (1964, 132; 1994, 1). Reconstruction of the spatial development of the Neolithic settlement on the Kephala mound has been of paramount interest to archaeologists (Evans 1971, 95; Broodbank 1992, 39; Katsianis 2002). What we know so far regarding the establishment and growth of Neolithic Knossos comes primarily from the area beneath the Central Court of the palace, however (Evans 1971, fig. 1). The

trenches opened along the periphery of the Kephala mound hardly offer concrete evidence for a sound spatial reconstruction and can only inspire some very general remarks. Attempts to document the pace of growth using the evidence from specific sections of the settlement and then projecting it to the tell as a whole are undermined by the absence of solid architectural evidence (Whitelaw 1992, 225). Here we present some new spatial features that might be considered characteristic of the organizational structure of Neolithic Knossos along its eastern periphery during the site’s long and different cultural periods (Tomkins 2008, 21). Although their fragmentary nature and contextual deficiency must be acknowledged, they lead us to the important conclusion that the eastern edge of the Kephala mound was an integral part of the Neolithic tell and its occupational sequence from the very beginning of the settlement’s life (see discussion of the Aceramic in Efstratiou, this vol., Ch. 2). A brief summary of the architectural findings from Trench II is offered below. All of the architectural remains encountered in our excavation have been dated to the EN II period

22

NIKOS EFSTRATIOU, ALEXANDRA KARETSOU, AND ELENI BANOU

(levels 14–29). In levels 16–19 we encountered two parallel walls (nos. 1 and 2) that most probably formed part of a house (Fig. 1.18). The area directly below the two walls revealed a row of stones that formed a stretch of wall (no. 5) running east– west. Beneath wall 2 an earlier wall (no. 6) with a similar orientation was also found (Figs. 1.20, 1.21, 1.24). The east end of wall 5 seemed to appear again in level 22, where it was visible against the west side of wall 6 (Fig. 1.24). Whether walls 5 and 6, which were set at an angle, defined the corner of an earlier house within the EN II horizon is difficult to say. Levels 23–26 revealed the remains of a narrow wall (no. 7) running in an east–west direction; its function is unclear (Figs. 1.23, 1.24). The massive elliptical wall (no. 3) first appeared in level 18. It had a north–south orientation, with a curve inclining toward the east at its southern end (Figs. 1.16, 1.18). Measuring 0.30 m in width and preserved to an impressive height of nearly 0.70 m, it constituted the dominant architectural feature of the entire deposit, spanning levels 18–27 (Fig. 1.32). It was probably constructed in level 27 (Fig. 1.25), a yellowish-colored, mud- or clay-textured deposit. The elliptical wall was made of mediumsized stones of kouskouras bound together with mortar. A thick band of yellow soil without finds found next to the wall (see description of levels 17, 18, and 18A) may represent a superstructure made of pisé (clay and straw). Due to the small extent of the excavated area, it is impossible to estimate the total length or the course of the elliptical wall in relation to other spatial features of the settlement. Its use as a defensive or retaining feature would have been problematic since the bend of the curve faced the Central Court (west) and did not follow the natural contour of the Kephala hill so as to protect or retain it (Fig. 1.25; cf. Evans 1971, pl. VI). Moreover, another wall (no. 4) of the same width (0.30 m) was found set at a right angle to the elliptical structure and running eastward (Figs. 1.25, 1.33). It most probably served as a supporting element or a partition wall. The presence of a perpendicular adjoined wall indicates that we may be dealing with an important, large-scale, and composite building. This interpretation is suggested by the structure’s careful construction and its long life span, which extended through most of the EN II period. Elliptical

walls are rare in the Neolithic levels of Knossos, although a very flimsy elliptical wall made of clay, also of EN II date, was excavated in Stratum IV by J.D. Evans (1964, 149, fig. 14). Further attempts at interpretation are hampered by the restricted size of the excavation. No structural features were found in either the LN levels (1–3) or the MN levels (4–13) of the excavation. The MN deposits, however, contained remains of possible hearths, postholes, and habitation refuse. Similarly, no traces of architecture were found in levels 30–37, assigned to the EN I period, although four hearths (nos. 4–7) and two pits (nos. 1 and 2) were encountered in various levels. Given the limited area exposed, it is difficult to discern the spatial association of features such as the hearths and pits with settlement architecture. It is not clear whether the EN II hearths 1–3 of levels 29 and 29A (Figs. 1.28, 1.29), hearth 4 from levels 30 and 30A (Figs. 1.28, 1.30), hearths 5 and 6 from level 31, and hearth 7 from level 32 (Fig. 1.34) were situated within the interior of a house or in an open area such as a courtyard or a lane between houses of EN II date. The same applies to the two EN I pits found in level 37 (Fig. 1.35). The so-called corridor area located between walls 1 and 2 (Fig. 1.18) was kept clear from structures for over two millennia, or most of the long life of the Neolithic settlement. Whether this area belonged to an open-air space or the interior of a structure, the fact that it remained open could not have been fortuitous. It probably means that a general plan of the settlement was agreed upon and kept unchanged, at least for this part of the site. Other long spaces or corridor-like arrangements are known from the EN II house in Stratum V (Area C) of the Central Court, a few meters to the west of our trench (Evans 1964, 146, 156– 157, fig. 12). The most striking example of a corridor flanked by narrow walls, however, is the MN house at Katsambas (Alexiou 1955, fig. 2; Zois 1973, fig. 19). Alexiou (1955) regarded this long space as a courtyard with a surrounding wall (peribolos), an interpretation that probably does not apply to our corridor. Nevertheless, it appears that similar spaces, possibly of diverse function, occurred elsewhere in Crete during the Neolithic.

THE EXCAVATION

23

Conclusions Despite the small size of our trench, we can match the new data recovered with a number of suggestions and hypotheses previously advanced by J.D. Evans in his discussion of the spatial features of Neolithic Knossos (1971, 95). We can now confirm his observation regarding the common north–south orientation of the walls in all parts of Neolithic Knossos (Evans 1964, 138). Indeed, most of the walls uncovered in our sounding exemplify this orientation. This probably indicates a long duration of spatial organization patterns and social relations within the community. Moreover, the fact that the only solid architectural remains from the trench date to the EN II period corroborates Evans’s suggestion that the Neolithic settlement acquired its largest size at that time. Our sounding strongly indicates that the EN II settlement reached the northeastern fringe of the natural hill of Kephala. The interpretation of the massive elliptical structure, which is unlike anything found by Evans, remains problematic. It is possible that it served specific collective organizational purposes, perhaps defensive, although this would imply a more complex political scene in Neolithic Crete than is supported by the available archaeological evidence. Knossos appears to have been a solitary

settlement on the island throughout the seventh and sixth millennia b.c., although this picture may soon change. Finally, the habitation remains from the socalled Aceramic levels are undeniably important, not so much because they substantiate the existence of a pottery-free phase in the Aegean (a long-debated issue), but rather because they document archaeologically the arrival of the first farmers in Crete (see Efstratiou, this vol., Chs. 2 and 11). This lower part of the long Knossos sequence consists of occupation refuse such as burned material from hearths, animal bones, and pieces of sun-dried mudbricks. The latter may represent a different building technique from that associated with the pisé in use subsequently, as Evans also noticed. He concluded that the first settlement of the EN I period—“Aceramic” or not—might have had a temporary character (Evans 1964, 136). In later publications, although adhering to his first suggestion of a small-sized site and community (Evans 1994, 2), he seemed reluctant to comment on the relative permanence of the first settlement, acknowledging the existence of more than one early phase (1971, 103). Only extensive future research can elucidate the scale and characteristics of the earliest settlement of Crete.

References Alexiou, S. 1955. “Ἀνασκαφὴ Κατσαμπᾶ Kρήτης,” Prakt 110 [1960], pp. 311–320. Broodbank, C. 1992. “The Neolithic Labyrinth: Social Change at Knossos before the Bronze Age,” JMA 5, pp. 39–75. Evans, J.D. 1964. “Excavations in the Neolithic Set­ tlement of Knossos, 1957–60: Part I,” BSA 59, pp. 132–240. ———. 1971. “Neolithic Knossos: The Growth of a Set­tlement,” PPS 37, pp. 95–117. ———. 1994. “The Early Millennia: Continuity and Change in a Farming Settlement,” in Knossos: A

Labyrinth of History. Papers in Honour of S. Hood, D. Evely, H. Hughes-Brock, and N. Momigliano, eds., Oxford, pp. 1–20. Furness, A. 1953. “The Neolithic Pottery of Knossos,” BSA 48, pp. 94–134. Katsianis, M. 2002. Detecting the Growth of Neolithic and Early Bronze Age Knossos through the Modelling of the Depositional Evidence: A GIS Application, M.A. diss., University College, London. Mackenzie, D. 1903. “The Pottery of Knossos,” JHS 23, pp. 157–205.

24

NIKOS EFSTRATIOU, ALEXANDRA KARETSOU, AND ELENI BANOU

Tomkins, P.D. 2008. “Time, Space and the Reinvention of the Cretan Neolithic,” in Escaping the Labyrinth: The Cretan Neolithic in Context (Sheffield Studies in Aegean Archaeology 8), V. Isaakidou and P. Tomkins, eds., Oxford, pp. 21–48.

Whitelaw, T.M. 1992. “Lost in the Labyrinth? Comments on Broodbank’s Social Change at Knossos before the Bronze Age,” JMA 5, pp. 225–238. Zois, A. 1973. Κρήτη—Ἐποχή τοῦ Λίθου, Athens.

2

The Stratigraphy and Cultural Phases Nikos Efstratiou

The Knossos Neolithic excavation of 1997 took place within a 3 x 2 m trench (Trench II) that was excavated to a depth of 8 m (see Efstratiou, Kar­etsou, and Banou, this vol., Ch. 1).* The deposit was excavated in 39 levels that were defined in accordance with field conditions such as soil characteristics, the thickness of the natural layers, the presence of architectural features, and digging difficulties. Although the excavation was restricted in both its spatial extent and the time available for its completion, these limitations were mitigated to a considerable degree by our understanding of the stratigraphy of the site as established by the work of J.D. Evans (1964, 132; 1971, 95; 1994, 1; Warren et al. 1968, 239). Trench II was located *Abbreviations used in this chapter are: AMS accelerator mass spectrometry BM British Museum lab code cal. calibrated or calendar years DEM Laboratory of Archaeometry of N.C.S.R “Demokritos” lab code EN Early Neolithic km kilometers

very close to Evans’s excavation area AC in the Central Court (Fig. 1.1). Thus, his research served as an important point of reference for our expectations regarding the general sequence and thickness of the layers, the phases of occupation, and the cultural composition of the site. Nevertheless, our digging methodology was not strictly dictated by Evans’s results, but rather by a flexible field procedure designed to incorporate both the information (e.g., stratigraphic, ceramic) available from past campaigns and the evidence that emerged in different stages of our own digging process. We were particularly concerned with defining the natural and anthropogenic characteristics of the superimposed layers of the deposit, and LN m MN OxA sp. spp.

Late Neolithic meters Middle Neolithic Radiocarbon Accelerator Unit, Oxford University, lab code species various species

26

NIKOS EFSTRATIOU

Figure 2.1. Sedimentological samples of the middle part of the south profile. Photo N. Efstratiou.

in this we were assisted by our sedimentologist, the late Maria-Pilar Fumanal García of the University of Valencia, who was present for much of the excavation. With her help it was possible to decide on the thickness of the archaeological layers to be removed and to follow their slope along the western and southern profiles of the trench (Fig. 1.4). Our excavation methodology was supported by the analysis of the sedimentological samples collected systematically from all the levels of the deposit (Figs. 1.2:c, 2.1). Although the results usually gave no more than a gross indication of the nature of the layers, for example, open areas versus interiors, we were able to use the sedimentological analysis for grouping together a number of levels with similar features, con­tributing to our overall understanding of the stratigraphy (see Fumanal García, this vol., Ch. 4). Particular mention should be made of the awkward digging situation we faced from level 31 onward when, for practical reasons, the trench’s exposed area was limited to 1.5 x 1.5 m. The already limited visibility of level 30—by now 4.5 m below the surface—was even more reduced. This made the remainder of the excavation down to the natural bedrock an increasingly difficult and demanding process, especially when it came to identifying and recording the different layers and soil features. Considering the digging conditions, we are nonetheless satisfied that the archaeological description of the site’s deepest and perhaps most interesting levels (37–39) was adequate with regard to the recovery of architectural features and small finds. Our results from the archaeobotanical

analysis of seeds, faunal analysis, studies of phytoliths and charcoal, sedimentology, ceramic analysis, and radiocarbon dating from samples taken throughout the sequence serve to enrich the information derived from past excavations at the site. The results of the 1997 excavation are relevant to some of the most hotly debated research issues of the early Knossos cultural sequence, including the presence of an Aceramic phase, the possibility of a pre-Neolithic human presence on the island, the date of the first community on Kephala hill, and the identity of the settlers (i.e., whether they came from overseas or from elsewhere in Crete). Our findings also have an important bearing upon the question of the continuity of the Neolithic habitation of Knossos over a period of more than 1,500 years (7000–5500 cal. b.c.) and the absolute chronology of the much-discussed Early Neolithic (EN) pottery sequence of Knossos, particularly in view of the advanced ceramic technology represented (see discussion below and that of Dimitriadis, this vol., Ch. 3; Tomkins 2007a, 180). While the limited scale of the 1997 excavation area must be acknowledged, we hope that the new discoveries will provide much-needed hard evidence with which to evaluate previous speculative reconstructions of the organization of Neolithic Knossos and, in particular, its expansion along the eastern edge of the Central Court (Broodbank 1992, 39; Whitelaw 1992, 225). The 39 levels of the 1997 excavation were allocated to several cultural phases by taking into consideration a number of criteria. These included, first of all, the typology of the ceramic material and its characteristic traits, some of which have been imbued—justifiably or not—with assumptions about their cultural and relative chronological significance (Mackenzie 1903, 157; Furness 1953, 94; Evans 1964, 132; Tomkins, Day, and Kilikoglou 2004, 51). Second, we relied upon J.D. Evans’s wellknown stratigraphic sequence (Aceramic, EN I, EN II, Middle Neolithic [MN], Late Neolithic [LN]) exemplified and modified in a number of studies (1964, 1994; Warren et al. 1968; Tomkins 2007b, 9). Third, we considered the archaeology of each of the levels (e.g., architecture, lithics, small finds), and fourth, we made use of the new accelerator mass spectrometry (AMS) dates on radiocarbon samples from the 1997 excavation. Each of these criteria has its own limitations and is subject to dynamic reevaluations.

THE STRATIGRAPHY AND CULTURAL PHASES

The new evidence from the 1997 excavation, including the series of radiocarbon dates covering the entire deposit, allows for a critical reconsideration of chronology and other cultural problems. To begin with, we found a loose but persuasive correlation between the archaeology of the 39 levels and Evans’s Strata I–X (Evans 1964; Warren et al. 1968). This correlation encompasses ceramic particularities (shapes, decoration, fabric), chronological ref­erences (relative and absolute), and archaeological features (architecture, building material, small finds). Some aspects of this correspondence—e­ spec­ ially regarding the ceramics—proved very helpful in view of the small quantity of pot­tery recovered in 1997. The

27

sequence of the levels excavated in 1997 and that of the previous Knossos campaigns presented only minor variations in their stratigraphic and chronological order and in their cultural succession. In fact, the pottery analysis of the 1997 material validates the basic cultural inferences of Evans’s sequence. Any divergences will be discussed below. The different cultural phases of Neolithic Knos­sos derived from the stratigraphic sequence excavated in 1997 are presented below, making use of all the available archaeological evidence. Particular atten­tion is paid to studies of materials related to the use of space, subsistence practices, environmental conditions, radiocarbon dating, and ceramic technology.

The Aceramic Levels Levels 39 and 38 of Trench II are the basal layers of the Knossos Neolithic deposit, and they correspond to Evans’s Stratum X. Their total thickness ranges from 0.50 to 0.60 m, with their main characteristic being the complete absence of ceramics. Architectural remains are lacking except for pieces of fired mudbrick, a building material also noticed by Evans in Stratum X of nearby Area AC (1964, 142). The area excavated in 1997 does not appear similar to the much more built-up Area ZE, located to the south and considered to be the core zone of the original village (Evans 1971, 99; 1994, 2, fig. 3). Unfortunately, the scale of the excavation does not permit us to evaluate either Evans’s initial suggestion that Aceramic Knossos was a temporary camp (based on the lack of permanent structures) or his revised interpretation after the discovery of mudbricks in 1970 (Evans 1964, 142; 1971, 101). The mere fact, however, that the so-called Acer­ amic occupation extends all the way to the end of the northeastern side of the tell, beyond Area AC, is important evidence concerning the size and extent of the early settlement on the Kephala hill. This information has been fruitfully utilized in a recent GIS analysis of the long-term growth of the Neolithic settlement using data from soundings in Areas X, ZE, AC, and the 1997 Trench II (Katsianis 2002, 38).

Faunal evidence from the 1997 excavation strong­ly supports Evans’s findings concerning the presence of fully domesticated animals, key elements of the Neolithic subsistence package. In his study of the animal bones from the Aceramic and EN I levels excavated in 1997, Pérez Ripoll observes the presence of cattle, sheep, goats, pigs, and dogs (this vol., Ch. 8). As in the previous studies of faunal material from Evans’s excavations (Jarman and Jarman 1968; Isaakidou 2004), no remains of wild animals were found in any of these levels. The absence to date of any pre-Neolithic evidence for the wild ancestors of the domesticated fauna in Crete makes it more than certain that these animals were introduced from elsewhere. Similarly, there is no evidence that would support any serious argument for a local domestication process in mainland Greece (Perlès 2001). Moreover, archaeobotanical finds from the A­cer­amic levels 39 and 38 show beyond any doubt that the first settlers at Knossos were full-fledged farmers, highly acquainted with agriculture, who emphasized the cultivation of cereals over legumes and practiced some form of arboriculture (see Sarpaki, this vol., Ch. 5). Their subsistence complex included not just cultivated but also domesticated crops such as einkorn (T. monococcum), emmer (T. dicoccum), two-row and six-row

28

NIKOS EFSTRATIOU

hulled barley (Hordeum distichum and H. vulgare, respectively), and naked barley (H. vulgare var. nudum). Pulses and fruits were represented by lentil (Lens sp.), almond (Prunus amygdalus), and fig (Ficus carica). Of particular importance is the discovery of free-threshing wheat (T. turgidum L./ aestivum L.), as Helbaek identified Triticum aestivum in the archaeobotanical sample from Evans’s excavations. The presence of free-threshing wheat led Evans to propose an Anatolian origin for the Neolithic settlers (Evans 1994, 1). This hypothesis is not without problems, however, since the occurrence of naked wheat in seventh millennium b.c. Crete predates its appearance in Anatolian contexts by one thousand years. Nevertheless, early sea-going movements by human groups in the eastern Mediterranean are definitely indicated. Evans (1994, 2) described the first Aceramic set­ tlement as advanced in its economy of domesticates and cultigens but simple in its bone and stone tool technologies. The suggestion by Kozłowski that the lithics of Evans’s Stratum X (and from the entirety of the Early and Middle Neolithic) are part of a Mesolithic tradition is interesting and deserves further investigation. Unfortunately, Conolly’s recent preliminary study of the lithic as­ semblage from Evans’s Aceramic strata lacks draw­ings and contextual distribution in­formation, making it difficult to ascertain the size of the unre­ touched pieces or the characteristics of the flaking technology (Connolly 2008, 75). It is apparent, nonetheless, that the assemblage is characterized by a blade-like flakelet industry with no typical tool classes and an absence of cores. Conolly cautiously sug­gests that this material resembles the very early Franchthi as­semblage, termed Initial Neolithic by Perlès (2001), which is marked by the presence of small flakes and denticulates. Conolly also notes the absence of microliths. These characteristics might indicate that Crete and perhaps all of the Aegean islands were separated from continental developments from the very beginning of the Hol­ ocene (Kaczanowska and Kozłowski 2006, 67), but a definitive conclusion must be deferred in view of the new evidence that is accumulating from preNeolithic horizons in the Aegean (Sampson 2006). The chipped stone finds from the 1997 excavation were few in number, comprising four pieces of obsidian—small flakes and broken blades—and one flint flake. While this small obsidian sample

cannot be used to elucidate the problems of lithic technology raised above, it nonetheless attests to the prevalence of early seafaring and perhaps long-distance exchange. Although the provenance of the obsidian was not determined, a Melian origin is most probable. The obsidian samples from the following EN I level 37, which were analyzed in the Laboratory of Archaeometry of N.C.S.R. “Demokritos” using neutron activation analysis, were found to derive from Melos (A. MoundreaAgrafioti, pers. comm., 2003). With regard to the ongoing discussion of endemic animals on the island and the possible causes of their extinction (Cherry 1990; Lax and Strasser 1992; Hamilakis 1996; Jarman 1996; Lax 1996; Reese, ed., 1996), the 1997 faunal sample has only indirect evidence to offer. None of the wild animals—wild goat/agrimi, boar, badger, and marten—were found before the MN levels. Instead, they seem to have been introduced to the island at the transition between the EN II and MN periods, when the boar appeared, with the other species present in the MN and/or LN levels (Pérez Ripoll, this vol., Ch. 8). This supports the suggestion that the native fauna of “oceanic” Crete were already long extinct (Reese, Belluomini, and Ikeya 1996, 47). Lax and Strasser’s (1992, 211) suggestion that some of the last endemic fauna on the island became extinct as a result of human agricultural activities and the competing habits of wild species introduced in the Aceramic and Early Neolithic is not supported by our findings. Moreover, the overkill model or its variations for the extinction of endemic animals (Lax and Strasser 1992) cannot be disassociated from the question of a preNeolithic human presence in Crete. Visits to the island by seagoing hunter-gatherer groups of the late Pleistocene or early Holocene may have contributed to the demise of the native fauna (Sampson 2006; Ammerman 2010, 81; see also discussion by Efstratiou, this vol., Ch. 11). The more general hypothesis that intensive early farming activities at Knossos disrupted the local animal habitat long after the establishment of the Aceramic village may be supported by the ecological evidence from the 1997 deposits (this vol., Badal and Ntinou, Ch. 6; Madella, Ch. 7). In particular, analysis of the phytolith Phase A (west profile) from the Aceramic and the beginning of the EN I period shows a high frequency of phytoliths

THE STRATIGRAPHY AND CULTURAL PHASES

in the sediments and indicates a very strong use of wood and leaves from the natural vegetation around the site as fuel and/or fodder. The hypothesis that there was a wide exploitation of the available wild plant resources during the seventh millennium (for a full discussion, see Madella, this vol., Ch. 7) receives further support from the charcoal evidence (see below and Badal and Ntinou, this vol., Ch. 6). It is risky, however, to generalize the evidence from a small excavation area such as ours to the issue of habitat overexploitation on a large island like Crete. Badal and Ntinou’s study of the charcoal from the 1997 excavation was particularly informative with respect to the plant formations in the habitat surrounding Neolithic Knossos. Regarding the Aceramic level in particular, despite the scarce presence of charcoal remains in the sample, Badal and Ntinou (this vol., Ch. 6, p. 109) note that “the absolute frequency of deciduous oak . . . establishes it as the dominant taxon in this assemblage and clearly differentiates level 39 from the rest of the sequence.” This means that deciduous oak either was part of the prevailing early Holocene vegetation in the vicinity of the site at the beginning of the seventh millennium b.c. (it changed to sclerophyllous

29

woodland and evergreen oaks at the beginning of EN I), or it was present because humans selected it for use—for example, as a source of timber for building. The authors lean toward the latter explanation, although they leave open the possibility of climatic or human influences on changing vegetation profiles. The idea that we may be able to glimpse the vegetation found by the first settlers at Knossos around 7000 b.c. and the alteration that it underwent over the following centuries because of human activities from the remains of wood used in domestic fires is a particularly welcome revelation. Badal and Ntinou suggest that early farmers reduced the territory for oak growth by opening small plots of land for cultivation, thus accounting for the reduction of deciduous species after the first 1,000 years of occupation. This reconstruction is supported by the distinct changes observed in the list of flora and the frequencies of certain taxa found in the charcoal assemblages over the long life of the Neolithic settlement. These changes are evident by the end of the EN II period, possibly as a result of the intensification of human activities in the catchment area of the site.

Chronology of the Aceramic and Early Neolithic Occupations Turning now to the absolute dating of the basal layers of the tell and the Aceramic settlement, we have four 14C dates, all in calibrated years (cal.) b.c., with which to work. Three of these dates come from the 1960 excavation: BM-124, 7250– 6690 (Sample 1, Pit F, Area AC, Level 27, Stratum X); BM-278, 7030–6650 (Sample 1, Pit F, Area AC, Level 27, Stratum X); and BM-436, 6770–6430 (Sample 1, Pit F, Area AC, Level 27, Stratum X). The fourth date, OxA-9215, 7030– 6780 (Knossos-1, Trench II, level 39/1), comes from the 1997 excavation. This sample, dated using the AMS method, comes from wood charcoal of the Quercus deciduous type (see Facorellis and Maniatis, this vol., Ch. 10). All dates were calibrated using the latest international curve (Stuiver and Reimer 1993; Reimer et al. 2004), and the

consistency of their ages around 7000 b.c. is remarkable. The 1997 AMS sample, taken almost 40 years after the first excavation, confirms the early date of the Cretan Neolithic beyond any doubt. This is particularly worth noting in view of the importance of AMS dating in documenting the rate of spread of the Neolithic in Europe (Ammerman and Biagi, eds., 2003). The life span of the initial village on the Kephala mound is difficult to estimate. The only 14C date from Evans’s Stratum IX is BM-272, 6590– 6250. If accurate, this date might be taken to indicate that the beginning of the EN I period was at least 400 years later than the founding of the settlement. Should the Aceramic phase be assigned such a long time span? Or should we consider the possibility of a lengthy “come and go” period

30

NIKOS EFSTRATIOU

with many short occupational phases (Evans 1971) within the 1 m thick deposit of Stratum IX and 1997 levels 39–38? Unfortunately, the aforementioned 14C sample is from charcoal associated with a mudbrick house overlying the earliest occupation and may not be reliable due to the provenance of the sample (Facorellis and Maniatis, this vol., Ch. 10, Table 10.2; see Barker, Burleigh, and Meeks 1969; Evans 1971). The sedimentological analysis of samples XXXI and XXX from the Aceramic layers excavated in 1997 (see Fumanal García, this vol., Ch. 4) does not further illuminate the problem of possible abandonment episodes during or soon after the Aceramic period. The unequal distribution of organic matter within the 8 m thick deposit of the tell is, however, worth considering. The samples from the deepest part of the profile show a persistently low percentage of organic content, which could indicate either a temporary habitation, perhaps a camp, or an outdoor area with no evidence of human activities. Other lines of evidence make the hypothesis of periodic abandonments seem less likely. The 2.5 m thick deposit of sediments spanning the period from approximately 7000 to 5400 b.c. (the latter being one of our earliest radiocarbon dates for EN I, discussed below) implies an accumulation of habitation debris at the rate of ca. 4.5 cm every 100 years (see Facorellis and Maniatis, this vol., Ch. 10). This rough rate is subject to cultural variations, of course, and it should be used with caution when applied to different habitation areas of the settlement. Interestingly, however, no occupation gap was documented between the Aceramic and EN I periods in the 1997 excavation area; the EN I level 37 lies directly above the Aceramic levels 39 and 38. Tomkins also casts doubt on the presence of abandonment phases at the site during the Aceramic and EN I periods, suggesting that the associated dates actually come from a rather late EN I deposit associated with a spatial reorganization of the settlement, LN I according to his

revised sequence relating Knossos to other areas of the Aegean (Tomkins 2008, 22, table 3.1). The beginning and the duration of the EN I period in radiocarbon years remains a puzzle, given the uncertain duration of the preceding Aceramic phase and the absence of clear evidence for an occupational gap between the two eras. Seven radiocarbon dates (four of them AMS) were obtained from the EN I levels 30–37 of the 1997 excavation. These dates include: OxA-9216, 5210–5060 (level 37); OxA-9219, 5460–5305 (level 35); OxA9217, 5210–5030 (level 33); OxA-9220, 5210–5050 (level 32); DEM-663, 5207–5050 (level 32); DEM661, 5290–5060 (level 31); and DEM-670, 4830– 4470 (level 30). With the exception of the sample from level 30, which falls within the first half of the fifth millennium, all the dates are concentrated in the second half of the sixth millennium. The EN I period seems to end just after 5000 cal. b.c., as the beginning of the following EN II is dated by OxA-9218 to 4940–4800 (1997 level 29), by BM-577 to 4990–4540 (Evans’s Stratum IV), and by BM-279 to 4690–4370 (Evans’s Stratum IV). The 1997 date DEM-660 (6250–6100) from level 29 must be discounted. Thus, there is a wide time gap of approximately 1,500 years between the end of the Aceramic phase (as ostensibly marked by BM-272, 6590–6250) and the end of the EN I period. The lack of additional supporting dates for the span of the Aceramic is regrettable. It is important to establish better chronological control over the long duration of EN I, especially if one accepts its extraordinary time span of more than 1,500 years (Efstratiou et al. 2004, fig. 1.3). Tomkins has attempted to break the long EN I period into three subphases (EN Ia, b, and c) based on his detailed analysis of Evans’s pottery material (Tomkins 2007b, 9). The archaeological documentation of this long period cannot, however, rely solely on changes in pottery typology or technology, although the analysis of changing attributes of the ceramic technology may be part of the answer to this problem, as discussed below.

THE STRATIGRAPHY AND CULTURAL PHASES

31

The EN I Levels The beginning of the Ceramic Neolithic period at Knossos (Evans’s EN I, Strata IX–V) is first seen in stratigraphic level 37 of the 1997 excavation and continues until level 30. The depth of the deposit exceeds 3 m, similar to the thickness of deposition (3 m) reported by Evans in the nearby Central Court (1994, 10). This period is distinguished by the appearance of successive occupation features, including traces of habitation floors and remains of decayed walls (kouskouras). Pits and filling debris such as scattered stones and mudbricks were found in level 37; similarly, Evans (1964) also observed a number of pits of unspecified character cut into the Aceramic deposit in Area AC. In the following levels architectural building material, fireplaces, and the debris of habitation areas (floors, charcoal, pottery, bones) are common. We have interpreted these remains as the refuse of a house or structure (see Efstratiou, Karetsou, and Banou, this vol., Ch. 1). Although the small area exposed does not allow any conclusive interpretation of the architectural features found, Evans’s previous observations (1994) related to the employment of specific building techniques and their differentiation in the course of the deposit have been confirmed. Finds of sundried bricks with impressions of straw accord with his comments on Stratum IX, in which “the earliest traces of solid structures were found” (Evans 1964, 144). Moreover, the presence of kouskouras for the first time in the Knossos sequence—with pisé following in the upper part of the deposit—is consistent with Evans’s observations concerning changes in building methods from Stratum VII (EN I) onward (1964, 150; 1994, 8). The repetition of habitation features within the 3 m thick EN I deposit often occurred with alternating layers of dissolved kouskouras building material, indicating a long habitation sequence. Phytolith evidence also attests to the presence of successive habitation floors (see below). The end of the EN I deposit (level 30) is marked by the first evidence for an architectural structure, a group of stones. This was followed in the EN II period with the construction of the massive elliptical

wall. Whether the EN I remains, as opposed to those of the Aceramic, constitute the earliest manifestation of the first permanent Neolithic village at Knossos is difficult to determine based on the limited area of the 1997 trench. Nevertheless, we concur with the EN I expansion (LN I in the schema of Tomkins 2008) proposed by Evans and others (Broodbank 1992, 39; Whitelaw 1992, 225; Evans 1994). Due to low visibility within the trench at the great depth of the EN I levels, minor but probably significant depositional changes were difficult to detect, complicating the interpretation of the EN I radiocarbon dates. Nevertheless, the analysis of EN I sedimentological samples XXIX–XIV has aided our understanding of the overall composition of the strata. At least until sample XVII, the sediments were characterized by a low organic content, as was the case in the preceding Aceramic phase. In the upper EN I levels from which samples XXVI–XIV were taken, however, the organic content tended to increase, possibly indicating the more anthropogenic character of the deposit. Even so, a number of samples (XVII, XIX, and XXIa) from different levels of the 3 m thick EN I deposit did not contain any organic material. These samples were taken from successive thin bands of yellowish soil, probably dissolved kouskouras from decayed walls, which alternated with the refuse of habitation floors (see Efstratiou, Karetsou, and Banou, this vol., Ch. 1). Although we do not want to overemphasize the implications of these sedimentological results, they could nevertheless indicate a change in the settlement’s spatial organization (see Madella, this vol., Ch. 7). In the faunal samples from the 1997 excavation, wild animals were strikingly absent from the EN I levels, but all domestic species were present— cattle, sheep/goat, pig, and dog. Because of the small sample size, the Ovis and Capra remains were treated together, as in the previous study of Knossos fauna (Jarman and Jarman 1968). Contrary to the evidence from Evans’s sample, in which the occurrence of cattle bones doubled by Stratum VIII (1994, 10), cattle appear to have been

32

NIKOS EFSTRATIOU

relatively insignificant until the EN II/MN period. While the quantity of the remains may have been affected by the small sample size, the pronounced increase in the frequency of cattle bones in the succeeding EN II and MN periods should not pass unnoticed (Pérez Ripoll, this vol., Ch. 8, Table 8.1). The same rate of increase was observed in the case of pigs between the EN I and the LN periods. Although Isaakidou (2004, 104) notes “traction pathologies” in cattle from the end of the EN I and in EN II in Evans’s sample, such evidence only becomes apparent in the LN material from 1997. No fracture marks for the extraction of marrow on the bones of Capra, Ovis, and Sus were found in the 1997 material either; this also contrasts with the older material (Isaakidou 2006, 100). The limitations of the bone sample from the 1997 excavation should be stressed once again, however. The main cereal crops represented were naked wheat (T. turgidum/aestivum), einkorn, emmer, and barley, while legumes included the pea (Pisum sp.) and the horsebean (Vicia faba) (for a full discussion, see Sarpaki, this vol., Ch. 5). Domesticated flax (Linum usitatissimum) was introduced in the EN I period, and in view of the enigmatic absence of olives, the flax might have been used for oil extraction. Sarpaki reports the presence of wild radish (Raphanus raphanistrum) and discusses at length the interesting archaeobotanical question of whether it was collected or cultivated. The presence of possible fodder remains (Leguminosae, Trifolium spp./Medicago sp.) could attest to a change in the use of the area from living to storage. The archaeobotanical evidence from the EN I period is very important because it helps to outline the subsistence profile of the early farmers of Knossos. It complements the limited but meaningful record of the first Aceramic settlers who found their way up to the top of Kephala hill in the seventh millennium b.c. With or without abandonment intervals—a still debatable issue—the EN I occupation seems to have continued until the end of the sixth millennium b.c. and beyond. During this long period the cultivation of domesticated cereals and legumes, which constituted the basic Neolithic package, remained uninterrupted and was probably intensified in order to cover the needs of the village’s increasing population. This may have resulted in significant clearing of the native vegetation (see Badal and Ntinou, this vol., Ch. 6).

The rich phytolith assemblages of the EN I levels have been grouped into three phases: Phase A (initial EN I), Phase B (late EN I) and Phase C (the end of EN I). The remains from Phase A resemble the Aceramic deposit, with mostly wild resources present, the Phase B finds include more cultivated plants, and the Phase C remains are once again similar to those of the Aceramic. It is hypothesized that “during the seventh millennium there was a wide exploitation of the available wild plant resources” and that “the excavated area was not the focus of any cereal crop processing, or that in general the intensity of cereal processing in the settlement was minimal” (see Madella, this vol., Ch. 7, p. 126). Later in the EN I period there was evidence in the area for the processing of cultivated cereals, with the crops Triticum sp. and Hordeum sp. now replacing wild species. As mentioned above, from level 34 onward the presence of at least two successive habitation floors was documented, and it is therefore reasonable to suggest that the EN I occupation was intensive enough and/or long enough to have led to noticeable changes in the evidence for plant collection activities within the limited area exposed in 1997. The analyses of the charcoal from the EN I levels at Knossos, together with the existing pollen cores and diagrams from Hagia Galini in SouthCentral Crete (Bottema 1980), Tersana in northwestern Crete (Moody 1987; Moody, Rackham, and Rapp 1996; Rackham and Moody 1996), and Delphinos in northwestern Crete (Bottema and Sarpaki 2003), provide important information about the Neolithic vegetation of the island. These two sources of information are in general agreement regarding the presence of wooded and nonwooded areas. The authors of the charcoal report emphasize the impact of human actions on the local landscape, which may account for its compos­ite diachronic vegetation succession. The deciduous oak, which was dominant in the Aceramic period, gave way to sclerophyllous woodland (evergreen oak, Quercus sp. evergreen type) in levels 37–32 of the EN I period, possibly as the result of increasing human pressure in the plant environments (Badal and Ntinou, this vol., Ch. 6). This important information indicates that the Aceramic might have been a long occupational period—at least 1,000 years—during which intensive human activity at Knossos exerted pressures on the environment,

THE STRATIGRAPHY AND CULTURAL PHASES

causing the restriction of deciduous species to protected habitats. The suggestion that humans competed for the same environments as the deciduous oak when opening plots for cultivation should be seriously considered, especially if occupation was continuous. The vegetational changes noted could be interpreted as the earliest reported evidence for land clearance in Crete, taking place long before the changes previously detected in the MN pollen record. Nevertheless, it is also possible that

33

these changes occurred due to climatic conditions, specifically the much-discussed environmental change of 6200 b.c. (8200 cal. b.p.) (Weninger et al. 2006, 401; Berger and Guilaine 2009, 31). A noticeable climatic change such as this might justifiably be used as an additional criterion for demarcating the beginning of EN I, along with the appearance of ceramics, changes in building techniques, and other archaeological evidence.

EN I Pottery Technology Pottery recovered from the 1997 dig was the focus of a systematic study concerned not so much with typological traits and their chronological significance (both thoroughly studied long ago; see Furness 1953; Evans 1964; Tomkins 2001), but rather with the technology of manufacture (see Dimitriadis, this vol., Ch. 3). While the full typological and stylistic description of the 1997 pottery is not presented in this volume, we are optimistic that the evidence for ceramic technology discussed here will provide insights into aspects of early society of Knossos that are otherwise difficult to approach. These insights relate to questions of intrinsic social evolution at Knossos and the possible influence of external relations with other communities in Crete or abroad, since technological choices are related to both personal and collective domains of social and economic life. Evans stated that the pottery introduced to Neo­lithic Knossos in EN I was “the product of a fully developed tradition of potting” (1964, 196). This statement is strongly supported by our findings (see this vol., Ch. 3; Dimitriadis 2008, 249). Moreover, Dimitriadis (this vol., Ch. 3, p. 49) notes that “the degree of fabric diversity of the EN Knossos ceramics is significantly wider than anything suggested in earlier ceramic studies,” a conclusion also reached independently in other recent studies (Tomkins and Day 2001; Tomkins, Day, and Kilikoglou 2004, 51). What is most surprising, perhaps, is the distinctiveness of the EN I pottery in the context of the overall Knossian ceramic sequence. In particular, petrographic variation in the types of tempering material used was greater

in EN I than in EN II. Among the three groups of fabrics (with numerous subtypes) identified in the Knossos material, calcareous fabrics appear to have been dominant in EN I. The calcareous fabric 1C (see Dimitriadis, this vol., Ch. 3), which was missing from the early levels (37 and 36), suddenly appeared in level 35 and remained the most common fabric continuously thereafter. The reasons for the diversity of the Knossos fabrics and their changing frequencies over the long stratigraphic, chronological, and social time span of the site must have been complex; they were not merely functional, aesthetic, or ideological. No meaningful correlation between technological choices and variables such as shapes, surface treatment, or decoration was found at any time during the long occupation of the settlement. On the contrary, there seems to have been an arbitrary, often surprising stratigraphic, chronological, or social differentiation of fabrics. We can only speculate about what this means in terms of community organization, socioeconomic priorities, and social contacts and exchanges. Nevertheless, we will try to enrich the debate on the significance of the ceramic patterns observed with the following arguments. The mature technological characteristics seen in the forming and surface treatment of the first ceramics from level 37 point to a long tradition of pottery production, as noted by Furness (1953), Evans (1964, 196; 1994, 1), and Dimitriadis (this vol., Ch. 3). There seems to be no disagreement on the actual relative time of the appearance of the first potters and their craft at Knossos (our level

34

NIKOS EFSTRATIOU

37 and Evans’s Stratum IX). Nevertheless, there is a problem related to the absolute dating of that level. Evans’s solitary 14C date for this horizon, 6590–6250 cal. b.c. (BM-272), is 1,000 years earlier than our corresponding level 37, which has an AMS date of 5210–5060 (OxA-9216). We believe that our date is much closer to reality since it is affirmed by a number of similar dates (see discussion of chronology above), and it is also supported by the well-developed pottery style, which is of a “late” character (LN) when viewed in comparison with the typical Greek mainland pottery typology. This means that pottery was probably introduced to Knossos at a date well into the sixth millennium b.c. The apparent “lateness” of the Knossian Early Neolithic in wider Aegean terms—radiocarbon, typological, technological—has led Tomkins to propose a new chronological scheme for the settlement in order to deal with the mismatch between Crete and adjacent areas (Tomkins 2008, 22, table 3.1). The issue at stake here, however, is not so much the archaeologically constructed, ceramicbased periodization, its alleged cultural content, or its endurance within the scholarly literature, but rather what the impressive Knossian palimpsest represents in actual historical terms. In other words, what does the archaeology of the prehistoric community reveal through different lines of evidence—be it social, economic, ideological, or technological? Unfortunately this is not an easy problem to resolve. As a consequence of its burial underneath the Minoan palace complex, the exposure of the Neolithic occupation in most cases— the 1997 season included—has been limited in relation to the overall extent of the site. One important question is not so much that of the potters’ origin but rather of their relation to the preexisting Aceramic farming community. The sudden appearance of pottery in the material record of a mature Neolithic community hardly points toward an internal evolution; instead, either the arrival of newcomers at Knossos or the acceptance of the new technology by the Aceramic community must be supposed. Both possibilities are not without archaeological difficulties, however, since no substantial cultural gap (spatial, faunal, archaeobotanical) between the Aceramic and EN I village is documented, nor does there appear to have been a local “reception” period for

the newly arrived technology. Instead, the technological analysis of the pottery has shown that it was characterized by an impressive pluralism of fabrics and tempering materials, exceeding that of later periods, contrary to what one might expect. While there are cases in which specific fabrics disappeared and were replaced by new ones in the later periods of the tell’s long stratigraphic sequence, this does not appear to have been the norm. The possibility that we are dealing with a dynamic farming community in EN I Crete is high, and the ceramic evidence certainly does not suggest that Knossos was an orphan island community. Indeed, the analyses of the 1997 material indicate that the quantity of ceramics “imported” to Knossos during the Early Neolithic may have been large (Dimitriadis 2008, 260). Dimitriadis (2008, 249) justifiably hesitates to commit himself regarding the nonlocal provenance of two of the most popular EN I fabrics at Knossos (groups 2 and 3). Systematic fieldwork for locating the sources of temper used for these fabrics is needed in order to identify possible ceramic production centers or exchange patterns and their cultural trail. Almost all of the so-called nonlocal tempers may be found in a radius of 20–30 km or less of the site, for example, at Psiloritis and Talea Ori (Dimitriadis 2008, 260–261, fig. 4). Thus, what constitutes a local or nonlocal fabric could very likely be a matter of definition in social terms, particularly in view of the fact that a real nonlocal fabric (type 3F3) that was exogenous to Crete, most probably of Cycladic origin, has also been detected (see Dimitriadis, this vol., Ch. 3). Even so, we need to consider the social reasons for acquiring nonlocal finished vessels or using tempering materials obtained at some distance from the settlement. Should this phenomenon be explained by a social mechanism such as “consumption” serving status needs or the possession of symbolic capital (Tomkins 2007a, 192)? Or was it a manifestation of a technological pluralism within a community that was taking advantage of the long experience of pottery manufacture in different communities around Knossos, either within Crete or on neighboring islands? It is not at all obvious that “distance” and its symbolism or “consumption” (Tomkins 2007a, 193) had a social significance in the context of our model of fluid Neolithic households (Efstratiou 2007b, 29).

THE STRATIGRAPHY AND CULTURAL PHASES

The “pluralism” made evident by the wide variation of tempering material in the EN I pottery can be interpreted in different ways. The diversity may have been introduced as part of the tradition of the founder settlement of the “ceramic” community. Alternatively, it may have been a tradition that flourished locally, maturing and diversifying over the years. In the latter case, it may represent characteristics relating to social structure that went hand-in-hand with other communitylevel processes. For example, Tomkins (2007a) has argued that fabric diversity may reflect shifts in mobility and exchange between different communities on the island. Nevertheless, the question remains as to whether we should be talking more in terms of the movability (exchange) of ceramics or of the influx of people from abroad. Two developments of particular relevance in this context are the appearance in level 35 of the subsequently popular 1C fabric and its varieties and the fading out of the group 2 fabrics that had dominated the earlier levels 37 and 36. The fabric diversity of the Knossian EN I ceramics may be interpreted in terms of people arriving and settling in the area from places close to

35

Knossos or from farther away, bringing their pottery and technological knowledge along with them while keeping ties with their places of origin. In either case, the growth of Knossos during the EN (Evans 1971, 1994) may not have been the outcome of the natural expansion of an isolated community but rather the result of a continuous influx of new settlers. Many other innovations seen at Neolithic Knossos, including new building methods and the introduction of new plants and wild fauna, most probably arrived from outside the island. Therefore, we tend to view EN I Knossos as a “cosmopolitan” community, the dynamism of which surpassed that of the EN II settlement. Its development may not have been the outcome of internal processes as proposed by Broodbank (1992, 55). While stylistic characteristics of the pottery might support the latter interpretation (Washburn 1983, 138), it is equally or even more credible to consider technological parameters, particularly the variation in tempers seen in EN I. The striking stylistic homogeneity of the Knossos pottery tradition as a whole makes particularly dubious the use of “decoration” or “design” as criteria for cultural or social diversity (Whitelaw 1992, 229).

The EN II Levels The EN II deposit excavated in 1997 (levels 14–29) measured 2 m in depth, roughly comparable in thickness to the 1.5 m deposit excavated by Evans in area AC of the nearby Central Court. The EN II levels revealed an interesting archaeological record that contained architectural remains and a large amount of domestic refuse (seeds, animal bones, charcoal, and pottery). Despite the limited size of the area exposed, the presence of a massive elliptical structure (wall 3, Fig. 1.32), a corridor-like area defined by rows of stones, probably walls, running in an east–west direction, and large quantities of varied building material (kouskouras, kouskouropetres, and pisé) suggest that this was a lively part of the settlement with persistent habitation (see Efstratiou, Karetsou, and Banou, this vol., Ch. 1). Successive hard surfaces, possibly floors, along with fireplaces and the remains of ashes, burned material, and animal bones

were observed. A number of building phases may be represented, but their identification is difficult. The occurrence of walls and corridor-like spaces suggest a densely built-up area, but it is not easy to distinguish indoor and outdoor spaces. The massive elliptical structure founded at the beginning of the period would seem to have been part of a monumental construction effort. Our observations are consistent with Evans’s (1994, 11) proposal that there was a rapid expansion of the settlement at this time and that the EN II habitation area was reaching the extent of the later Minoan palace or beyond toward the north and west (Evans 1994, 14). The sedimentology report (samples XIII–IX) reveals alternations in the percentages of organic matter that were most probably related to fluctuating levels of anthropogenic activity. While little further elaboration is possible, certain extreme

36

NIKOS EFSTRATIOU

examples of this variation could be suggestive of specific depositional events, especially when combined with the relevant archaeological evidence. For example, sample IX, which corresponds to lev el 14A and marks the end of EN II, is characterized by a high organic content; this same level is described archaeologi­cally as a burned level and may represent an abandon­ment or reconstruction episode at the site. There is no apparent stratigraphic gap between the end of EN I and the beginning of EN II in the 1997 excavation area (levels 29–28). The absolute dates for the early part of the EN II fall around or just after 5000 cal. b.c.: OxA-9221, 4994–4856 (level 28); DEM-659, 4940–4800 (level 28); DEM658, 5199–4956 (level 24); DEM-642, 4929–4800 (level 14); and DEM-641, 5220–4910 (level 13/ early MN). These dates from the 1997 excavation fall within the range of Evans’s dates BM-577, 4990–4540 (Stratum IV), and BM-279, 4690– 4370 (Stratum IV). The period may have lasted a few hundred years, as suggested by BM-279. Un­ fortunately, the close of EN II is not well defined by the 1997 dates, but we tend to agree with Evans’s suggestion (1994, 14) that end of the period falls “within the earlier part of the 5th millennium b.c.” The pattern of animal husbandry in the EN II period resembles that of EN I, with goats and sheep dominating the archaeozoological record and relatively low frequencies of cattle (see Pérez Ripoll, this vol., Ch. 8). The presence of mostly young individuals shows that the animals were kept predominantly for meat. Fracture marks are not evident in the small sample of bone remains. The EN II archaeobotanical evidence reveals an emphasis on cereals and legumes, with an inten­ sified cultivation of the latter (see Sarpaki, this vol., Ch. 5). Sarpaki detects a shift toward legume cultivation along with a stabilization in cereal production, developments which she attributes to agricultural intensification and shortage of land, probably associated with settlement growth and population increase. All of the plants present in EN I contexts continued to be cultivated in EN II, and the quantities of fruit observed more than doubled. The practice of arboriculture is firmly attested by the increasing number of almonds and figs. Sarpaki discusses the possibility that almond trees were brought by the first inhabitants of Knossos as

part of the Neolithic package, although they might also have been part of the natural vegetation of Crete, as proposed by Badal and Ntinou (this vol., Ch. 6). The occurrence of almond trees, whether wild or domesticated, was also observed in the charcoal remains. The EN II phytolith assemblages—Phase D and samples XIII–IV—come from both the south and west profiles of the trench, and they include the remains of cereal phytoliths, especially wheat and millet. The presence of the latter, indicating the exploitation of wild grasses that might have been used as fodder, is particularly notable along the west profile. It is also possible that during its life the EN II settlement may have experienced more changes in spatial use (areas of crop processing) and fewer abandonment episodes than in the previous period (EN I), as would be consistent with the hypothesized growth of Knossos in the EN II period. The paleoenvironmental evidence for the EN II period shows two successive anthracological (charcoal) zones: that of the dominance of evergreen oak, similar to the previous EN I period, and that of Prunus/almond dominance. The former indicates the typical natural vegetation surrounding the site of Knossos in the first part of EN II—woodland interspersed with open vegetation and small cultivated plots. The latter corresponds to the end of the EN II and partly to the MN period, and it testifies to a change in the vegetation, perhaps due to the special use of almond trees for their edible fruits (see Badal and Ntinou, this vol., Ch. 6). Thus, the possibility of an early management of trees and a form of proto-arboriculture is cautiously proposed here. There is, however, another interesting and more general point to be made. By the end of EN I, at least 1,000 years after the establishment of the Aceramic village, long-lasting human activities in the vicinity of the settlement—farming, herding, and burning—seem to have taken their toll on the density of the evergreen oak woodland, favoring specific subsistence strategies such as tree management. In this context, the ongoing absence of the olive from the EN II remains at Knossos is striking. Overall, the vegetation record of Knossos at the beginning of the fifth millennium b.c. attests to an anthropogenic environment affected by a long and intense Neolithic occupation.

THE STRATIGRAPHY AND CULTURAL PHASES

The study of the EN II pottery technology may reveal important social changes within the Knossos community. The great variety in tempering material that was characteristic of the EN I ceramics was reduced in this period (see Dimitriadis, this vol., Ch. 3). As the quantity of ceramics retrieved in 1997 was small, it is not certain whether the changing selection of raw materials was affected by social factors. Technological changes noted in level 18 and continuing thereafter (EN II–LN), marked by the domination of noncalcareous and semicalcareous matrices instead of the calcareous ones typical of EN I, provoke interesting suggestions, however. One is the possibility that by the beginning of the fifth millennium b.c. the community of Knossos experienced a population increase

37

that led to agricultural intensification (cultivation of legumes, beginning of arboriculture) and spatial rearrangements involving the construction of the massive elliptical structure and settlement expansion. Internal technological innovations such as the improvement of old tempering methods or the adoption of new ones could be considered an integral part of increasing socioeconomic “complexity.” Another possibility is that a new wave of settlers arrived, either from within the island or from abroad. These people might have brought with them novel pottery techniques that were subsequently adopted by the inhabitants of Knossos. Unfortunately, in the absence of other known Cretan sites of the fifth millennium b.c., this suggestion cannot be properly evaluated at present.

The MN Levels The MN period is represented by an almost 1 m thick deposit (levels 13–4) corresponding to Stratum III of Evans’s sequence. The deposit was marked by alternating, thick, sterile layers of kouskouras building material and blackish anthropogenic layers that were full of charcoal and rich in finds. The successive building levels suggest the occurrence of multiple reconstruction episodes of unknown length. There is no doubt that this part of the site was intensively occupied at this time, as is evident from the remnants of platforms, pits, postholes, floors, and structures of considerable time depth made of stones, kouskouras, and clay. It is not easy to determine whether these features were part of an open or roofed area (see Efstratiou, Karetsou, and Banou, this vol., Ch. 1). The remains mark a distinct MN habitation phase, consistent with Evans’s discovery of substantial MN architectural remains in the nearby Central Court (1994, 15). The sedimentological study (samples VIII–IV) is not particularly suggestive as to the nature of the MN levels except that they present a high carbonate content. Level 10, however, stands out with its high percentage of organic material. This accords with the archaeological descrip­tion of the deposit as being relatively rich in charcoal and animal

bones (see Efstratiou, Karetsou, and Banou, this vol., Ch. 1). The MN period should be placed in the second half of the fifth millennium cal. b.c. on the basis of the radiocarbon dates from Evan’s 1969 campaign: BM-580, 4460–4270 (Stratum III), and BM-575, 4550–4360 (Stratum III/II). Unfortunately, the 14C dates of the three samples taken from the 1997 excavation fall closer to the beginning of the fifth millennium and are all too high: DEM-641, 5220– 4910 (level 13); DEM-640, 4932–4801 (level 12); and DEM-638, 5320–5030 (level 9). We have no reasonable explanation to offer for this discrepancy between the two groups of dates except for the unlikely possibility that levels 13–4 belong to a late part of the EN II rather than the MN period. The abundant presence of rippled sherds, described by Evans (1994, 15) as being “enormously popular” in the MN period, makes this unlikely, however. Both sets of dates suggest that the duration of the period was rather short. The Knossos community experienced noticeable changes in its subsistence economy in the MN period. Analysis of the faunal remains shows an increase in the presence of cattle, the intensified exploitation of sheep relative to goat, and the first appearance of the wild goat, boar, and badger

38

NIKOS EFSTRATIOU

(see Pérez Ripoll, this vol., Ch. 8). In earlier studies doubts were expressed regarding the explanation of the increase in cattle during the MN period (Jarman, Bailey, and Jarman, eds., 1982; Broodbank 1992). The 1997 material seems to confirm that this phenomenon was not the result of depositional factors, and that it was more probably related to the community’s subsistence require­ ments rather than to ideological expressions of prestige, as suggested by Broodbank (1992, 62). The higher demand for meat, milk, and wool must have led to the observed increase in the numbers of sheep that were kept (see Pérez Ripoll, this vol., Ch. 8). This apparent economic dynamism coincides with the hypothesized spatial expansion and population growth of Knossos in the MN period (Broodbank 1992, 39). We thus do not agree with Evans’s suggestion that Knossos did not experience great changes in the fifth millennium (1971, 95). The analysis of the 1997 faunal samples indicates that wild mammals—beginning with Sus scrofa ferus and continuing with Capra aegagrus and Meles meles—were introduced to the island by humans during the MN period (see Pérez Ripoll, this vol., Ch. 8). This can be considered a reasonable proposition in view of the developments observed in other eastern Mediterranean islands such as Cyprus (Peltenburg and Wasse, eds., 2004). The much acclaimed “isolationism” of the early seventh- and sixth-millennium Knossian communities must be reconsidered. As in the case of Cyprus (eighth millennium cal. b.c.), the introduction of wild fauna in Crete presupposes the existence of a well-established network of community contacts and exchange relations within the eastern Mediterranean (Bar-Yosef 2001, 130). At the same time it may represent a response to the subsistence needs of indigenous island communities. It seems possible that the preceding 2,000 years of Neolithic occupation at Knossos caused changes to the natural environment, seen in the paleoenvironmental and archaeobotanical evidence, which, together with population growth and the settlement’s expansion, demanded new sources of food in the fifth millennium cal. b.c. The question of the possible relationship of the wild goat and the Cretan agrimi is still open (Reese, Belluomini, and Ikeya 1996, 47). Is the agrimi goat a subspecies of the wild bezoar goat (Capra aegagrus aegagrus) and domestic animals introduced to Crete in different times in the past

(Horwitz and Kahila Bar-Gal 2006, 123), or is it a feral goat originating from the initial domestic stock brought by the first settlers that gradually acquired a “wild” phenotype (Kahila Bar-Gal et al. 2002, 376)? This is a particularly interesting problem in view of the growing evidence that many eastern Mediterranean islands were inhabited or visited by human groups in pre-Neolithic times (for further discussion, see Efstratiou, this vol., Ch. 11). The interpretation of morphological features to resolve the issue could be misleading, as Pérez Ripoll rightly points out in his study (this vol., Ch. 8). Unfortunately, all of the goat DNA samples analyzed so far have been taken either from domestic breeds (Capra hircus) from different geographical locations or from agrimi, rather than from the actual Neolithic stock of Crete. The DNA analyses by Horwitz and Kahila Bar-Gal aiming to compare goat bone samples from the 1997 MN/LN faunal material from Knossos with modern agrimi samples were unsuccessful; due to poor preservation it was possible to recover only one short sequence from one of the bones, and the results were not diagnostic for identifying wild or domestic goat (Kahila Bar-Gal, pers. comm., 2007). The archaeobotanical samples for the MN period produced limited plant remains. The olive was still absent from the MN archaeobotanical record at Knossos, as was also the case in the charcoal samples (see below). Sarpaki also reports the absence of naked wheat, barley, and the grape, perhaps indicating changes in agricultural priorities and/or in the use of the excavated space (this vol., Ch. 5). The MN phytolith data (samples VIII–IV, phytolith Phase D), together with the EN II samples, show significant variability between the south and the west profiles of Trench II, with the south profile being richer in phytoliths. According to Madella (this vol., Ch. 7), such variation may represent different uses of space, with higher rates of phytolith deposition occurring in areas used for the processing of crops. The MN phytolith record suggests the possibility that this was a period of consolidation. According to the charcoal analysis report (see Badal and Ntinou, this vol., Ch. 6), the vegetation around Knossos experienced drastic changes by the end of the EN II and the beginning of the MN period, sometime around the middle of

THE STRATIGRAPHY AND CULTURAL PHASES

the fifth millennium b.c. All the evidence shows the repercussions of over 1,000 years of human activities (e.g., farming, herding, field clearance) in the Kairatos valley, perhaps intensified in the fifth millennium. This agrees with the archaeological evidence for a peak growth of the Knossos community in the EN II and MN periods. The evergreen oak woodland became less dense, with new, sunloving taxa (e.g., almond trees) taking advantage of the open spaces. The available pollen diagrams suggest similar vegetation changes in different parts of Crete during the fifth millennium b.c. as a result of both climatic conditions and human activity

39

(Bottema 1980; Moody, Rackham, and Rapp 1996; Bottema and Sarpaki 2003). As noted above, the olive continued to be absent from MN Knossos. Lesser quantities of pottery were recovered from the relatively thin MN deposit than from those of the preceding periods. Technologically, the pottery was still marked by the selection of noncalcareous and semicalcareous matrices. This choice might be explained in purely technological terms, but it is also tempting to relate it to the emerging picture of expanded contacts between Knossos and other communities within the island in the fifth millennium b.c.

The LN Levels The LN deposit (levels 3–1, Evans’s Strata II and I) had a total thickness of less than 1 m. Architectural remains were flimsy, comprised mainly of kouskouras building material and mixed debris from leveling activities in this part of the Central Court (see Efstratiou, Karetsou, and Banou, this vol., Ch. 1). Similarly, the deposit was poor in other archaeological finds, in contrast with the more illuminating MN levels (4–13). The analysis of sediment samples recovered from the LN layers was not undertaken as these levels seemed to have been too exposed to later disturbances. The absolute dating of the LN period is uncertain. Unfortunately, no 14C dates are available from the levels excavated in 1997. Moreover, as Evans himself noted (1994, 18), the four radiocarbon dates from the 1969 excavation are puzzling, as they either fall within an earlier EN II horizon (BM-717, 4790–4500; BM-579, 4460–4330; and BM-581, 4610–4270, all cal. b.c.) or have a large deviation error (e.g., BM-716, 4040–3530). These dates most likely indicate that the LN period spanned 500 years near the end of the fifth or the beginning of the fourth millennium b.c. Relative dating on the basis of pottery styles is also precarious, as nothing strikingly new makes its appearance in this period (Evans 1994, 16). The improvement in firing techniques observed by Evans (1994) is not documented by the technological analysis of the 1997 material (see below).

Similarly, other aspects of material culture do not betray any drastic changes from the MN period. The archaeozoological evidence indicates that the transformations in the settlement’s subsistence economy first observed in the MN period, including the rising frequency of cattle and the preference for breeding sheep over goat, continued in the LN period. The possible significance of these changes has already been discussed. The intensification of both these production strategies may have been associated with the gradually changing economic and social relations of Neolithic Knossos at the end of the fifth millennium b.c. It is not easy to say whether the changes in faunal exploitation are attributable to the processes of the often-invoked “secondary products revolution” (Sherratt 1983). Important qualitative changes such as the first use of cattle for transportation and plowing are indeed documented by the evidence for bone deformations in the 1997 faunal sample (see Pérez Ripoll, this vol., Ch. 8), and they are also supported by the study of the older samples from Knossos (Isaakidou 2006, 104). It is not clear, however, why innovations like the intensification of the exploitation of cows and the use of plow technology should have been adopted from abroad, as Sherratt’s model indirectly suggests. The production of more meat, milk, and wool could easily have been a response to the growing collective subsistence needs of the fifth-millennium inhabitants

40

NIKOS EFSTRATIOU

of Knossos after a long and well-established occupation of the area. The same context of growing social demand and environmental change should be considered responsible for the introduction of wild fauna such as the wild goat, the boar, the marten, and the badger in the MN and LN periods. The case of Cyprus, where the introduction by the first farmers (eighth millennium cal. b.c.) of wild animals along with domestic ones is well documented, makes a credible precedent (see also Pérez Ripoll, this vol., Ch. 8). The different time of introduction of each of these species points to not just one but a number of multifaceted processes and phenomena—what Bailey (2007, 200) has called “time perspectivism”— that cannot be easily addressed by the static nature of just one archaeological record under study. This would mean that the introduction of new wild fauna in Crete could not be explained by the description of just one set or scale of phenom­ena but by a combination of them: the “grand narrative” of the spread of the Neolithic farmers westward (large scale), the Cretan Neolithic “politics” (medium scale), and the “history” of the Knossos community (small scale). The LN archaeobotanical evidence from Knossos is rather poor due to the limited sampling of the deposit. Nevertheless, the absence of the olive is confirmed by both the archaeobotanical and the charcoal record. The scarcity of plant remains also seems to be attested by the phytolith analysis of samples taken from the south profile (samples III– I, Phase E). The sediments were poor in phytoliths, with few cereal silica skeletons. Madella (this vol., Ch. 7) suggests that the excavation trench may have been an open area of unspecified use within the LN settlement. The anthracological evidence from level 3, the only LN stratum sampled, is similarly rather

limited. We might expect, however, that the effects (already evident in the MN period) of the previous two millennia of occupation and intense farming activities at Knossos must have been dramatic. The LN environment surrounding Knossos was most probably characterized by the decrease of oak and the increase of plants associated with farming. The absence of the olive in the wood charcoal analysis, contrary to what is known from the island’s available pollen records, is striking; it could perhaps be explained partly in terms of local peculiarities (see Badal and Ntinou, this vol., Ch. 6). The pottery of the period continued to be characterized by the presence of fabrics dominated by noncalcareous and semicalcareous matrices that first appeared in the EN I period, as discussed above. It should be emphasized again that there was no correlation between the different fabrics and pottery types, surface treatment, or decoration throughout the Knossos sequence. The lack of any new fabric types might be viewed as a decline of vitality in the Neolithic community. Thus, we cannot agree with Evans’s remark (1994, 16) that “firing techniques were clearly improved and new wares were developed.” From the perspective of the 1997 excavation, the LN remains a rather obscure period. The complete lack of any trace of the elusive Final Neolithic in this part of the tell makes its interpretation even more difficult. Therefore, we tend to see the LN settlement mainly as the closing stage of a long occupation period. Its lack of definition may relate to the changing picture of LN Crete as a whole, characterized by the expansion of communities throughout the island (Evans 1971, 99). Alternatively, it may reflect the particular history of the Central Court area, where extensive leveling work was carried out in the following period.

Conclusions We believe that the 1997 excavation reaffirms the cultural importance of the founding village of Knossos in the wider context of early developments in the eastern Mediterranean, the Aegean islands, and mainland Greece. The early chronology

of the site, long regarded as one of the earliest in the Greek Neolithic, is confirmed by a new AMS date around the beginning of the seventh millennium cal. b.c. Faunal and archaeobotanical data attest once again to the introduction by the first

THE STRATIGRAPHY AND CULTURAL PHASES

settlers of all the basic domesticated plant and animal taxa of the Neolithic subsistence package, as seen in the early excavations by Evans (1964, 1971, 1994; Warren et al. 1968). The paleoecological data derived from phytolith and anthracological analyses provide, for the first time, information of critical importance for addressing issues of vegetation composition, the locational distribution of activities, and subsistence practices, as well as the long term effects of the latter on the local environment. Furthermore, although the small area exposed cannot elucidate the full spatial extent of the Aceramic habitation, it nonethess indicates that the early occupation area extended along the northeastern edge of the hill. Certain aspects of the life of the Aceramic village, however, still remain obscure. The problems of its overall duration and its relative permanence (versus the possibility of periodic abandonments and recurring habitation) remain unresolved, as does the enigma of its apparently unique presence on the island. Our guess is that the first two questions will remain unanswered because future fieldwork exposing the deep layers of Neolithic Knossos is most unlikely (unless they are found in an unexpected depositional arrangement as in Trench ZE; Evans 1971). Moreover, documenting “permanency” or “seasonality” archaeologically is an ambiguous task (Edwards 1989). As for the solitary presence in Crete of this Aceramic village, in view of new discoveries of early Holocene sites on a number of Aegean islands (Sampson 2006), we think the perception of Knossos’s isolation is bound to change soon. The case of Cyprus, where late Pleistocene and early Holocene sites have recently been found, offers a compelling precedent, one which argues for the possibility of early seaborne travels. Indeed, the increasing evidence for sea-voyaging hunter-gatherer communities in late Pleistocene Cyprus (Ammerman et al. 2006, 1; 2007) raises a number of new questions. Should Crete and the Aegean be seen as part of the intriguing cultural scenery of the late Pleistocene in the eastern Mediterranean? How were these developments of the 11th millennium b.c. related, if at all, to the Mesolithic occupation of the Aegean islands and the beginning of the Neolithic in Greece? More importantly, where should the starting point of these early sea travelers be placed? What are the attributes of their material culture, and why has it

41

taken us so long to locate their sites? These points will be discussed further in Chapter 11. There is no doubt that the establishment of the first settlement at Knossos was just one act of the play performed on the eastern Mediterranean stage during the eighth and seventh millennia b.c. The environmental, economic, and social preferences displayed by the first agricultural communities in the Aegean must have led to the formation of a number of diversified material cultures. As I have suggested elsewhere (Efstratiou 2007a, 132), “we may still be missing the exact ‘plot’—the sequence of events that took place in the Greek mainland or the Aegean islands.” In this regard it is interesting to note, as Pérez Ripoll does in his archaeozoological study (this vol., Ch. 8), the material difference between the early seventh millennium b.c., fully domesticated fauna from Knossos and that of Neolithic Shillourokambos in Cyprus, 1,000 years earlier, where the animals appear to have been still at a “pre-domestication” stage (Vigne, Carrère, and Guilaine 2003). Is the time gap between these sites indicative of two quantitatively and chrono­logically different colonization episodes in the strikingly slow seaborne spread of the Neolithic westward (Biagi, Shennan, and Spataro 2005, 41)? Undoubtedly it is, although many intermediary historical phases and localized circumstances of this vast time span and region still seem to be missing. Regardless of the answers to all of the questions posed above, the arrival of the first Aceramic Neolithic community at Knossos may justifiably be considered a major event in the prehistory of the region, one that was part of a broader, multifaceted movement of people and domesticates (Efstratiou et al. 2004). Even if more preNeolithic sites (in addition to those recently discovered in the Plakias region; see Strasser et al. 2010) are soon discovered in Crete using new site location models (Runnels et al. 2005, 259; Ammerman et al. 2006, 1), the active involvement of these early inhabitants in the development of farming several millennia later remains to be demonstrated. Before examining the development of the Knossos settlement following the Aceramic phase, we must begin by acknowledging that the longestablished nomenclature of the Knossos cultural sequence conceals a rather uncomfortable problem: in terms of ceramic typology, some of the

42

NIKOS EFSTRATIOU

EN pottery from Knossos could easily be ascribed to a LN sequence in mainland Greece. Tomkins has recently reached the same conclusion after studying the bulk of Evans’s ceramic material (Tomkins 2007a, 180). He introduces a new chronological scheme relating the mainland and Cretan ceramic phases in a frame of cross-reference for the Aegean and adjacent areas of Anatolia. While this is a useful development, it should be noted that the Knossos pottery record of the late seventh and early sixth millennium b.c. remains unique. It cannot be compared to any other Cretan or Aegean island ceramic material of the same period because there are no comparably early finds from these regions. The earliest ceramic material from Katsambas, a site close to Knossos, falls rather late within the sixth millennium b.c. (Galanidou, pers. comm.). Although the sample of pottery retrieved from the 1997 excavation levels was small, our technological analyses of this material have afforded a number of cultural and chronological insights not available from stylistic analysis alone. In her initial study of the pottery, Furness (1953, 94) primarily made use of stylistic traits for distinguishing the EN I period, and Evans referred to the frequencies of these same traits in characterizing the homogeneity of the EN I ceramic material (1964, 194). As Evans observed (1994, 7), the earliest ceramics from Knossos were the product of a well-developed potting tradition for which there is no evident internal evolution at the site; thus the EN I pottery was almost certainly introduced from other communities beyond Knossos. Petrographic analy­sis reveals that diverse types of tempering materials were used initially, some perhaps deriving from sources within 20 to 30 km of the site, others from outside the island (e.g., from the Cyclades). Thus, from a technological perspective, the EN I pottery was not at all homogeneous, and we may speculate as to whether its diversity in turn represents the practices of different communities of potters whose products accumulated at Knossos. It is also note-worthy that new fabric groups were intro­ duced and then waxed and waned in popularity over time. Calcareous fabrics predominated in EN I, yet the most popular class of this group (1C) appeared rather late in EN I. In EN II, the overall variety of the tempers that were employed decreased, and new fabrics with noncalcareous or semicalcareous

matrices rose to prominence. These fabrics predominated in the MN and LN periods, supplanting the earlier calcareous wares. Again, one may ask whether these changes were related to the social origins of the potters. The subsistence economy of the Knossos community also apparently underwent significant transformations over the course of the Neolithic. Analysis of the faunal remains from the 1997 excavation show changing frequencies of domesticated animal species, with an increasing emphasis on cattle and sheep becoming apparent in the MN and LN periods. This development may perhaps be explained by rising needs for both meat and secondary products such as milk and wool, and the use of draft animals for plowing in a context of population growth and settlement expansion. In the MN and LN periods, wild species (including wild goat, boar, badger, and marten) not seen in the faunal samples of the Aceramic and EN periods were also introduced, presumably from outside the island. Their exploitation may further attest to the need for additional food supplies at a time when the population of the community was expanding, perhaps in part through the arrival of new settlers who introduced these very species. The archaeobotanical remains recovered from the 1997 excavation suggest a parallel expansion of the subsistence base over time, as the Aceramic complex of domesticates including emmer, einkorn, naked wheat, barley, and legumes came to be augmented by the use of the wild radish and domesticated flax (a source of oil?) and the management of trees, including the almond and the fig (but not, curiously, the olive). The increasing emphasis on legumes observed in EN II, if not an effect of sampling or the changing spatial organization of human activities, might also be interpreted as evidence for agricultural intensification, necessitated by the growth of the community and attendant local land shortages. More intensive practices of farming, herding, and land clearance, in conjunction with probable climatic change, may eventually have led to important changes in the natural vegetation of the Knossos region; already by EN I, the charcoal and phytolith evidence suggests that the original landscape dominated by deciduous oak had given way to sclerophyllous woodland, or evergreen oak, and by the MN period, this evergreen woodland in turn became less dense.

THE STRATIGRAPHY AND CULTURAL PHASES

As to the size and social structure of the Knossos community and the changes it experienced over the course of the Neolithic, the archaeological data from the small area of the 1997 excavations affords us a few intriguing insights. The EN I levels attest to a long habitation sequence during which a changing series of building materials (mudbrick, kouskouras, and pisé) was employed. From these traces of habitation it is clear that the settlement extended to the northeastern side of the hill at this time. The succeeding levels reveal that the area of the excavation trench was part of a densely built-up hab­itation zone in EN II, although it is difficult to de­termine whether the features encountered were part of building interiors or open external activity areas. A massive elliptical wall, which, we might speculate, was part of some communal or public buil­ding, was constructed in EN II. The area continued to be occupied intensively in the MN period, but the LN architectural remains were flimsy, having most likely been affected by later leveling activities in the Central Court area of the Bronze Age pal­ace. Overall, the EN II elliptical wall may be the most significant indication that social changes were un­der­way, assuming that the building of which it was a part was a focus of

43

communal activity and/or an expression of emergent social differentiation and hierarchy. Examined collectively, the various types of data recovered from the 1997 excavation greatly illum­inate our understanding of life at Neolithic Knossos. It is apparent that a process of agricultural intensification, complemented by the exploitation of a variety of wild plant and animal food sources, was underway from EN I onward, and the associated subsistence practices transformed the local landscape. We may infer that the intensification and diversification of the subsistence regime was a communal response to population increase. Whether that growth was the result purely of endogenous rates of increase or the periodic influx of new settlers from other communities within or beyond Crete cannot be determined with certainty at present, but the introduction of and long-term changes in ceramic technology, as well as the introduction of new species of wild animals, are lines of evidence that seemingly point to the ongoing migration of people from places outside of Knossos. The social consequences of communal growth under these circumstances may long remain a matter for conjecture.

References Ammerman, A.J. 2010. “The First Argonauts: Towards the Study of the Earliest Seafaring in the Mediterranean,” in Global Origins and Development of Seafaring, A. Anderson, J.H. Barrett, and K.V. Boyle, eds., Cambridge, pp. 81–92.

Bar-Yosef, O. 2001. “The World around Cyprus: From Epi-Palaeolithic Foragers to the Collapse of the PPNB Civilization,” in The Earliest Prehistory of Cyprus: From Colonization to Exploitation (Cyprus American Archaeological Research Institute Monograph Series 2), S. Swiny, ed., Boston, pp. 129–164.

Ammerman, A.J., and P. Biagi, eds. 2003. The Widening Harvest: The Neolithic Transition in Europe. Looking Back, Looking Forward (AIA Colloquia and Conference Papers 6), Boston.

Barker, H., R. Burleigh, and N. Meeks. 1969. “British Museum Natural Radiocarbon Measurements VI,” Radiocarbon 11, pp. 278–294.

Ammerman, A.J., P. Flourentzos, R. Gabrielli, P. McCartney, J. Noller, D. Peloso, and D. Sorabji. 2007. “More on the New Early Sites on Cyprus,” RDAC 2007, pp. 1–26.

Berger, J.-F., and J. Guilaine. 2009. “The 8200 cal bp Abrupt Environmental Change and the Neolithic Transition: A Mediterranean Perspective,” Quaternary International 200, pp. 31–49.

Ammerman, A.J., P. Flourentzos, C. McCartney, J. Noller, and D. Sorabji. 2006. “Two New Early Sites on Cyprus,” RDAC 2006, pp. 1–22.

Biagi, P., S. Shennan, and M. Spataro. 2005. “Rapid Rivers and Slow Seas? New Data for the Radiocarbon Chronology of the Balkan Peninsula,” in Prehistoric Archaeology and Anthropological Theory and Education (Reports of Prehistoric Research Projects

Bailey, G. 2007. “Time Perspectives, Palimpsests and the Archaeology of Time,” JAnthArch 26, pp. 198–223.

44

NIKOS EFSTRATIOU

6–7), L. Nikolova and J. Higgins, eds., Salt Lake City, pp. 41–52. Bottema, S. 1980. “Palynological Investigation on Crete,” Review of Palaeobotany and Palynology 31, pp. 193–217. Bottema, S., and A. Sarpaki. 2003. “Environmental Change in Crete: A 9000-Year Record of Holocene Vegetation History and the Effect of the Santorini Eruption,” The Holocene 13, pp. 733–749. Broodbank, C. 1992. “The Neolithic Labyrinth: Social Change at Knossos before the Bronze Age,” JMA 5, pp. 39–75. Cherry, J.F. 1990. “The First Colonization of the Mediterranean Islands: A Review of Recent Research,” JMA 3, pp. 145–221. Conolly, J. 2008. “The Knapped Stone Technology of the First Occupants at Knossos,” in Escaping the Labyrinth: The Cretan Neolithic in Context (Sheffield Studies in Aegean Prehistory 8), V. Isaakidou and P. Tomkins, eds., Oxford, pp. 73–89. Dimitriadis, S. 2008. “The Significance of Fabric Diversity in Neolithic Knossos Ceramics,” in Proceedings of the 4th Symposium of the Hellenic Society for Archaeometry, National Hellenic Research Foundation, Athens, 28–31 May 2003 (BAR-IS 1746), Y. Facorellis, N. Zacharias, and K. Polikreti, eds., Oxford, pp. 249–262. Edwards, P.C. 1989. “Problems of Recognizing Earliest Sedentism: The Natufian Example,” JMA 2, pp. 5–48. Efstratiou, N. 2007a. “The Beginning of the Neolithic in Greece—Probing the Limits of a ‘Grand’ Narrative,” in Mediterranean Crossroads, S. Antoniadou and A. Pace, eds., Athens, pp. 123–138. ———. 2007b. “Neolithic Households in Greece: The Contribution of Ethnoarchaeology,” in Building Communities: House, Settlement and Society in the Aegean and Beyond. Proceedings of a Conference Held at Cardiff University, 17–21 April 2001 (BSA Studies 15), R. Westgate, N. Fisher, and J. Whitley, eds., Cardiff, pp. 29–39. Efstratiou, N., A. Karetsou, E. Banou, and D. Margomenou. 2004. “The Neolithic Settlement of Knossos: New Light on an Old Picture,” in Knossos: Palace, City, State. Proceedings of the Conference in Herakleion Organised by the British School at Athens and the 23rd Ephoreia of Prehistoric and Classical Antiquities of Herakleion, in November 2000, for the Centenary of Sir Arthur Evans’s Excavations at Knossos (BSA Studies 12), G. Cadogan, E. Hatzaki, and A. Vasilakis, eds., London, pp. 39–49.

Evans, J.D. 1964. “Excavations in the Neolithic Settlement of Knossos, 1957–60: Part I,” BSA 59, pp. 132–240. ———. 1971. “Neolithic Knossos: The Growth of a Settlement,” PPS 37, pp. 95–117. ———. 1994. “The Early Millennia: Continuity and Change in a Farming Settlement,” in Knossos: A Labyrinth of History. Papers in Honour of S. Hood, D. Evely, H. Hughes-Brock, and N. Momigliano, eds., Oxford, pp. 1–20. Furness, A. 1953. “The Neolithic Pottery of Knossos,” BSA 48, pp. 94–134. Hamilakis, Y. 1996. “Cretan Pleistocene Fauna and Archaeological Remains: The Evidence from Sentoni Cave (Zoniana, Rethymnon),” in Reese, ed., 1996, pp. 231–239. Horwitz, L.K., and G. Kahila Bar-Gal. 2006. “The Origin and Genetic Status of Insular Caprines in the Eastern Mediterranean: A Case Study of Freeranging Goats (Capra aegagrus cretica) on Crete,” Human Evolution 21, pp. 123–138. Isaakidou, V. 2004. Bones from the Labyrinth: Faunal Evidence for the Management and Consumption of Animals at Neolithic and Bronze Age Knossos, Crete, Ph.D. diss., Univeristy College London. ———. 2006. “Ploughing with Cows: Knossos and the ‘Secondary Products Revolution,’” in Animals in the Neolithic of Britain and Europe, D. Sarjeantson and D. Field, eds., Oxford, pp. 95–112. Jarman, M.R. 1996. “Human Influence in the Development of the Cretan Mammalian Fauna,” in Reese, ed., 1996, pp. 211–229. Jarman, M.R., C.N. Bailey, and H.N. Jarman, eds. 1982. Early European Agriculture: Its Foundations and Development, Cambridge. Jarman, M.R., and H.N. Jarman. 1968. “The Fauna and Economy of Early Neolithic Knossos,” in “Knossos Neolithic, Part II,” P. Warren, M.R. Jarman, H.N. Jarman, N.J. Shackleton, and J.D. Evans, BSA 63, pp. 241–264. Kaczanowska, M., and J.K. Kozłowski. 2006. “Παλαιολιθικές παραδόσεις, Μεσολιθικές προσαρμογές και Νεολιθικές καινοτομίες στο Αιγαίο μέσα από το πρίσμα της λιθοτεχνίας,” in Προϊστορία του Αιγαίου, A. Sampson, ed., Athens, pp. 67–87. Kahila Bar-Gal, G., P. Smith, E. Tchernov, C. Greenblatt, P. Ducos, A. Gardeisen, and L.K. Horowitz. 2002. “Genetic Evidence for the Origin of the Agrimi Goat (Capra aegagrus cretica),” Journal of Zoology 256, pp. 369–377.

THE STRATIGRAPHY AND CULTURAL PHASES

Katsianis, M. 2002. Detecting the Growth of Neolithic and Early Bronze Age Knossos through the Modelling of the Depositional Evidence: A GIS Application, M.A. thesis, University College London. Lax, E.L. 1996. “A Gazetteer of Cretan Paleontological Localities,” in Reese, ed., 1996, pp. 1–32. Lax, E.L., and T. Strasser. 1992. “Early Holocene Extinctions on Crete: The Search for the Cause,” JMA 5, pp. 203–224. Mackenzie, D. 1903. “The Pottery of Knossos,” JHS 23, pp. 157–205. Moody, J. 1987. The Environmental and Cultural Prehistory of the Khania Region of West Crete, Ph.D. diss., University of Minnesota, Minneapolis. Moody, J., O. Rackham, and G. Rapp. 1996. “Environmental Archaeology of Prehistoric NW Crete,” JFA 23, pp. 273–297. Peltenburg, E., and A. Wasse, eds. 2004. Neolithic Revolution: New Perspectives on Southwest Asia in Light of Recent Discoveries on Cyprus (Levant Suppl. Series 1), Oxford. Perlès, C. 2001. The Early Neolithic in Greece: The First Farming Communities in Europe, Cambridge. Rackham, O., and J. Moody. 1996. The Making of the Cretan Landscape, Manchester. Reese, D.S., ed. 1996. Pleistocene and Holocene Fauna of Crete and Its First Settlers (Monographs in World Archaeology 28), Madison. Reese, D.S., G. Belluomini, and M. Ikeya. 1996. “Absolute Dates for the Pleistocene Fauna of Crete,” in Reese, ed., 1996, pp. 47–51. Reimer, P.J., M.G.L. Baillie, E. Bard, A. Bayliss, J.W. Beck, C. Bertrand, P.G. Blackwell, C.E. Buck, G. Burr, K.B. Cutler, P.E. Damon, R.L. Edwards, R.G. Fairbanks, M. Friedrich, T.P. Guilderson, K.A. Hughen, B. Kromer, F.G. McCormac, S. Manning, C. Bronk Ramsey, R.W. Reimer, S. Remmele, J.R. Southon, M. Stuiver, S. Talamo, F.W. Taylor, J. van der Plicht, and C.E. Weyhenmeyer. 2004. “IntCal04 Terrestrial Radiocarbon Age Calibration, 0–26 Cal Kyr bp,” Radiocarbon 46, pp. 1029–1058. Runnels, C., E. Panagopoulou, P. Murray, G. Tsartsidou, S. Allen, K. Mullen, and E. Tourloukis. 2005. “A Mesolithic Landscape in Greece: Testing a SiteLocation Model in the Argolid at Kandia,” JMA 18, pp. 259–285. Sampson, A. 2006. Προϊστορία του Αιγαίου, Athens. Sherratt, A. 1983. “The Secondary Exploitation of Animals in the Old World,” WorldArch 15, pp. 90–104.

45

Strasser, T.F., E. Panagopoulou, C.N. Runnels, P.M. Murray, N. Thompson, P. Karkanas, F.W. McCoy, and K.W. Wegmann. 2010. “Stone Age Seafaring in the Mediterranean: Evidence from the Plakias Region for Lower Palaeolithic and Mesolithic Habitation of Crete,” Hesperia 79, pp. 145–190. Stuiver, M., and P.J. Reimer. 1993. “Extended 14C Data Base and Revised Calib 3.0 14C Age Calibration Program,” Radiocarbon 35, pp. 215–230. Tomkins, P.D. 2001. The Production, Circulation and Consumption of Ceramic Vessels at Early Neolithic Knossos, Crete, Ph.D. diss., University of Sheffield. ———. 2007a. “Communality and Competition: The Social Life of Food and Social Transformation in the Aegean before the Bronze Age,” in Cooking Up the Past: Food and Culinary Practices in the Neolithic and Bronze Age Aegean, C. Mee and J. Renard, eds., Oxford, pp. 174–199. ———. 2007b. “Neolithic Strata IX–VIII, VII–VIB, VIA–V, IV, IIIA, IIIB, IIA and IC Groups,” in Knossos Pottery Handbook: Neolithic and Bronze Age (Minoan) (BSA Studies 14), N. Momigliano, ed., London, pp. 9–48. ———. 2008. “Time, Space and the Reinvention of the Cretan Neolithic,” in Escaping the Labyrinth: The Cretan Neolithic in Context (Sheffield Studies in Aegean Archaeology 8), V. Isaakidou and P. Tomkins, eds., Oxford, pp. 21–48. Tomkins, P., and P.M. Day. 2001. “Production and Exchange of the Earliest Ceramic Vessels in the Aegean: A View from Early Neolithic Knossos, Crete,” Antiquity 75, pp. 259–260. Tomkins, P.D., P.M. Day, and V. Kilikoglou. 2004. “Knossos and the Earlier Landscape of the Herakleion Basin,” in Knossos: Palace, City, State. Proceedings of the Conference in Herakleion Organised by the British School at Athens and the 23rd Ephoreia of Prehistoric and Classical Antiquities of Herakleion, in November 2000, for the Centenary of Sir Arthur Evans’s Excavations at Knossos (BSA Studies 12), G. Cadogan, E. Hatzaki, and A. Vasilakis, eds., London, pp. 51–59. Vigne, J.-D., I. Carrère, and J. Guilaine. 2003. “Unstable Status of Early Domestic Ungulates in the Near East: The Example of Shillourokambos (Cyprus, IX–VIIIth Millennia cal. B.C.),” in Le néolithique de Chypre. Actes du colloque international organisé par le Département des antiquités de Chypre et l’École française d’Athènes, Nicosie, 17–19 mai 2001 (BCH Suppl. 43), J. Guilaine and A. Le Brun, eds., Athens, pp. 239–251.

46

NIKOS EFSTRATIOU

Warren, P., M.R. Jarman, H.N. Jarman, N.J. Shackleton, and J.D. Evans. 1968. “Knossos Neolithic, Part II,” BSA 63, pp. 239–276. Washburn, D.K. 1983. “Symmetry Analysis of Ceramic Design: Two Tests of the Method on Neolithic Ma­ terial from Greece and the Aegean,” in Structure and Cognition in Art, D.K. Washburn, ed., Cambridge, pp. 138–164.

Weninger, B., E. Alram-Stern, E. Bauer, L. Clare, U. Danzeglocke, O. Joris, C. Kubatzki, G. Rollefson, H. Todorova, and T. van Andel. 2006. “Climate Forc­ ing Due to the 8200 cal yr bp Event Observed at Early Neolithic Sites in the Eastern Mediterranean,” Quaternary Research 66, pp. 401–420. Whitelaw, T.M. 1992. “Lost in the Labyrinth? Comments on Broodbank’s Social Change at Knossos before the Bronze Age,” JMA 5, pp. 225–238.

3

Fabric Diversity in the Neolithic Ceramics of Knossos Sarantis Dimitriadis

This chapter, which is a short version of an ear­lier report by the author (Dimitriadis 2008), pre­sents in succinct form the results of the petro­ graph­ic analysis of the pottery found during the 1997 exca­vation at Knossos.* It focuses on the di­ versity of fabrics in relation to the pottery types of the Early Neolithic (EN) I and II periods, paying particular at­ten­tion to issues of technology and provenance. The study is based on the petrographic analysis of 268 thin sections from ceramic sherds retrieved from the different levels of the excavation. Because

of the limited scale of the dig and the small quantity of ceramics collected, detailed references to and generalizations concerning the relation between vessel shapes, surface decora­tion, and fabric types are not presented. The stylistic analysis of the ceramic material from the 1997 ex­cavation confirms the general stylistic characteristics defined by Mackenzie (1903, 157), elaborated by Furness (1953, 94), and confirmed by Evans (1964, 132) for the Neolithic pottery of Knossos, as well as those noted in the pottery from the nearby site of Katsambas (Alexiou 1955, 311).

Methodology Samples for petrographic analyses were taken on the basis of their stratigraphic relevance, macroscopically perceptible fabric differences, wall thickness, surface treatment, and decoration. Ves­ sel shape was not a primary selection criterion,

as the fragmentary nature of the ceramic material made it very difficult or even impossible to *Abbreviations used in this chapter are: C=Celsius or Centigrade, EN=Early Neolithic, LN=Late Neolithic, and MN=Middle Neolithic

48

SARANTIS DIMITRIADIS

reconstruct vessel shapes. The primary objective was to collect samples representative of all the different types of pottery from the different levels in numbers proportionate to their relative frequencies. Thus, the EN I and II periods are better represented because of the thickness and longevity of the occupational layers, while the Middle Neolithic (MN) and Late Neolithic (LN) levels provided fewer diagnostic sherds.

With these considerations in mind, 268 sherds were selected, thin-sectioned, and examined under a polarizing microscope. In addition, a provisional geological survey of the Knossos area, which entailed the systematic collection of sands from streams in the vicinity, was conducted so that potential sources of the raw materials used in ceramic manufacture could be identified.

Ceramic Fabrics Tempering appears to have been extensively prac­ticed. Three groups of fabrics can be distinguished on the basis of the material(s) used as tem­per and the nature of the matrix. In the first group grains of various kinds of carbonates are al­ most ex­ clusively the tempering components, and the matrix is either calcareous or noncalcareous. In the se­cond group the tempering components are mainly noncarbo­nate rock fragments, and the ma­ trix is noncalcareous. In the third group carbonate and noncarbonate components have been added as temper in roughly equal amounts, and the matrix

is noncalcareous. All the ceramics examined were fired at temperatures of less than 750–800ºC. Both the forming and the surface treatment of even the earliest ceramics attest to the considerable technological knowledge of the potters, as has also been noted by Furness (1953) and Evans (1994, 1). The presence, absence, or coexistence of diverse fabrics within the various levels indicate significant differences. The greater variation in firing temperature observed by Tomkins and his associates (Tomkins, Day, and Kilikoglou 2004) might be explained by the larger size of their sample.

Fabric Types The first major group includes four fabric types (1A, 1B, 1C, and 1D), which can be further subdivided into a number of varieties (Dimitriadis 2008, 252–256, pls. I–IV). Thus, fabric 1A is tempered with fragments of various types of carbonate rocks (fine to medium sparitic, dark micritic, partly or completely dolomitized and/or recrystallized limestone). Two varieties of this fabric can be distinguished. In fabric variety 1A1 the matrix is cal­careous, whereas it is noncalcareous in fabric variety 1A2. The former is common in EN I and only scarcely present in later strata, while the latter, less common than 1A1, is present in both EN I and II but exists only sporadically in subsequent strata. Fabric 1B is tempered with fine turbid grains, most of them abraded but still clearly recognizable as partly dedolomitized rhombohedral dolomite monocrystals. Here again two varieties can be

distinguished: 1B1 with calcareous and 1B2 with noncalcareous matrices. A single repre­ sentative of 1B1 was found in level 37, and none was recovered from level 36. This fabric variety is abundant beginning in level 35 and continues throughout EN I and II. Fabric variety 1B2 is much less common than 1B1 and was only found in EN I levels. A third variety, 1B3, is tempered with angular to subrounded fragments of an equicrystalline sucrosic dolosparite plus single crystal rhombohedral dolomite, all embedded in a cal­careous matrix; it is rather common in EN I but is apparently missing from subsequent strata. Fabric 1C is tempered with cleaved, sharpedged calcite tempering grains. Variety 1C1 has a calcareous matrix. Although it is missing from levels 37 and 36, it is very common and continuously present in all the subsequent strata from EN

FABRIC DIVERSITY IN THE NEOLITHIC CERAMICS OF KNOSSOS

I until the end of the LN period. Fabric varieties 1C2 and 1C3 have semi-calcareous and noncalcareous matrices, respectively. Both are missing from levels below 18 but are common from that level upward within EN II. Fabric 1D bridges 1B and 1C fabrics as it contains both rhombohedral single crystal dolomite plus cleaved sparry-calcite. Contrary to 1B and 1C, which are the two most common fabrics among all the samples examined, 1D is uncommon and discontinuously present, first appearing in late EN I. The second major group includes one main fabric, 2E, which is further subdivided into a number of varieties and subvarieties. The first of these, fabric variety 2E1, has a noncalcareous matrix and is tempered with grains of slate phyllite, fine psammite, quartzite, and single quartz. Carbonate grains are only sporadically present and are apparently derived from carbonate veining in the phyllitic source rocks. Two subvarieties can be distinguished: 2E1a, which is characterized by the additional presence of siliceous sponge spicules and calcareous foraminifers, and 2E1b, in which such biograins are missing. Both varieties are fairly common only in EN I. Fabric variety 2E2 also has a noncalcareous ma­ trix, with tempering grains similar to 2E1 and additional components of green schist, serpentin­itic, and basic volcanic fragments. Two subvarieties can also be distinguished: 2E2a, with sponge spicules and foraminifers, and 2E2b, without such biograins. Both subvarieties are also common only in EN I.

49

Fabric variety 2E3 is characterized by abundant quartz, quartzite, chert, and carbonates, in addition to the tempering constituents of 2E2. This variety is transitional to the next group and could have been attributed to that as well. It is present primarily in EN I and only scarcely in EN II strata. The third major group, which includes one main fabric, 3F and its varieties, is nowhere prevalent in the stratigraphic sequence and has a discontinuous presence. In all its varieties the matrix of 3F is noncalcareous, and the tempering grains are fragments of various rocks, similar or identical to the ones present in the other two fabric groups, primarily group 2, with strained quartzite being the most common. Shell fragments, nearly absent from the other two groups, are occasionally present in 3F. The shape of the grains implies tempering with stream sands. Group 3 ceramics are more porous and friable than those of the other two groups. Fabric variety 3F1 is tempered with quartzites, metabasites, serpentinites, and various carbonates. Only one representative of this variety was recovered in the EN I strata; it is, however, relatively more common in the EN II levels. Fabric variety 3F2 is tempered with quartzite and various carbonates; it is apparently more common in EN I than in EN II. Fabric variety 3F3 is tempered with quartzite, feldspars, and carbonates, along with biotite and/or amphibole. Some polymineralic fragments are clearly granitic. Few representatives of this uncommon variety were found in the EN I and II levels.

Discussion The degree of fabric diversity of the EN Knossos ceramics is significantly wider than anything suggested in earlier ceramic studies (Furness 1953; Evans 1971; Noll 1982). The authors of a recent study and reexamination of the Neolithic ceramics from J.D. Evans’s excavations at Knossos appear to have reached similar conclusions (Tomkins and Day 2001). Despite the limited area exposed in the 1997 dig, all important fabric groups mentioned by Tomkins and Day are present in the 1997

sample, with the exception of the fabric tempered with blueschist fragments (Dimitriadis 2008, 260). This confirms the representative nature of the 1997 material and the validity of the conclusions drawn. A number of interesting general and specific points regarding the wider cultural implications of the ceramic data are raised by Efstratiou (this vol., Ch. 2). First, fabrics belonging to group 2 are only common in EN I. They dominate the bottom of the

50

SARANTIS DIMITRIADIS

Knossos stratigraphic sequence (levels 37 and 36) and become less common in level 35, after which they decline. They are gradually replaced by the appearance of group 1 fabrics, especially fabric 1C. Does this signify something more than a change in ceramic technology? It should also be noted that 1C remained the most common fabric subsequently (EN I–LN), perhaps signifying either the arrival of new Neolithic groups at Knossos after the earliest colonization episode or the simple adoption of a new tempering material (see discussion by Efstratiou, this vol., Ch. 2). Second, from very early on the EN I ceramics are characterized by a wider variation of tempering material than those of EN II. Fabrics 1B3, 2E, and 3F are typical of EN I; fabric 1A is characteristic of EN I, but it is also found in EN II. Fabric 1B1 is typical of both EN I and II, and it can be found sporadically in later phases. Fabric 1B3 is typical of EN I. It is not without significance that the raw material used in Fabric 1B1, persistent and common in later phases, is a naturally or technologically more refined cognate of the raw material used in fabric 1B3, which was more common in early phases. Third, an enigmatic technological change with pos­sible cultural implications seems to oc­cur from level 18 upward (EN II–LN): the domi­na­tion of noncalcareous and semicalcareous ma­trices, which surpass the calcareous varieties of EN I (see Efstratiou this vol., Ch. 2). Fourth, no relation can be seen between the different fabrics, vessel shapes, surface treatment, and decoration of the Knossos ceramics (Efstratiou et al. 2004). It is therefore reasonable to assume that choices of clay and temper were not affected by the shape, surface treatment, or decoration of any one period. Fifth, the diversity of the ceramic fabrics of Knossos is impressive. The Knossian potters’

access to multiple raw material sources should be considered a necessary precondition for the production of these wares. All of these sources can be found within Crete, although their distances from the settlement are difficult to determine. A provisional field survey conducted within a radius of 10 km around Knossos failed to locate potential sources for a significant number of the components identified in the EN ceramics, especially the components of group 2 and to some extent those of the group 3 fabrics. Therefore, on the basis of the evidence available so far it is possible to argue that some Neolithic ceramics from Knos­sos may not have been of strictly local origin. The question arises, however, regarding what makes a component “local” or “nonlocal.” If the catchment area around Knossos is expanded to 20 or 30 km, nearly all the components present in the ceramics studied can be traced. Should this distance be considered local or nonlocal? Nevertheless, it is proposed here that EN pottery production and exchange networks in Crete must have been complex, as the recent work by Tomkins confirms (Tomkins and Day 2001; Tom­ kins, Day, and Kilikoglou 2004, 51; Tomkins 2007, 174; 2008). Our study suggests that the quantity of the not strictly Knossian ceramics present at the settlement during the EN period may have been significantly larger than initially thought (Dimitriadis 2008, 249). It is also possible that the rare 3F3 fabric, as well as the fabric with blueschists mentioned in Tom­ kins and Day (2001), may not be Cretan at all; a Cycladic origin could be the most likely alternative. This is not as wild a hypothesis as it sounds if one considers the presence in the very early (EN I) levels of Knossos of other material, such as Melian obsidian, that came from far away places.

References Alexiou, S. 1955. “Ἀνασκαφὴ Κατσαμπᾶ Kρήτης,” Prakt 110 [1960], pp. 311–320. Dimitriadis, S. 2008. “The Significance of Fabric Diversity in Neolithic Knossos Ceramics,” in Proceed­ ­ ­ ings of the 4th Symposium of the Hellenic Society

for Archaeometry, National Hellenic Research Foundation, Athens, 28–31 May 2003 (BAR-IS 1746), Y. Facorellis, N. Zacharias, and K. Polikreti, eds., Oxford, pp. 249–262.

FABRIC DIVERSITY IN THE NEOLITHIC CERAMICS OF KNOSSOS

Efstratiou, N., A. Karetsou, E. Banou, and D. Margomenou. 2004. “The Neolithic Settlement of Knossos: New Light on an Old Picture,” in Knossos: Palace, City, State. Proceedings of the Conference in Herakleion Organised by the British School at Athens and the 23rd Ephoreia of Prehistoric and Classical Antiquities of Herakleion, in November 2000, for the Centenary of Sir Arthur Evans’s Excavations at Knossos (BSA Studies 12), G. Cadogan, E. Hatzaki, and A. Vasilakis, eds., London, pp. 39–49. Evans, J.D. 1964. “Excavations in the Neolithic Settlement of Knossos, 1957–60: Part I,” BSA 59, pp. 132–240. ———. 1971. “Neolithic Knossos: The Growth of a Settlement,” PPS 37, pp. 95–117. ———. 1994. “The Early Millennia: Continuity and Change in a Farming Settlement,” in Knossos: A Labyrinth of History. Papers in Honour of S. Hood, D. Evely, H. Hughes-Brock, and N. Momigliano, eds., Oxford, pp. 1–20. Furness, A. 1953. “The Neolithic Pottery of Knossos,” BSA 48, pp. 94–134. Mackenzie, D. 1903. “The Pottery of Knossos,” JHS 23, pp. 157–205. Noll, W. 1982. “Mineralogie und Technik der Keramiken ­ Altkretas,” Neues Jahrbuch für Mineralogie Ab­ handlungen 143, pp. 150–199.

51

Tomkins, P.D. 2007. “Communality and Competition: The Social Life of Food and Social Transformation in the Aegean before the Bronze Age,” in Cooking Up the Past: Food and Culinary Practices in the Neolithic and Bronze Age Aegean, C. Mee and J. Renard, eds., Oxford, pp. 174–199. ———. 2008. “Time, Space and the Reinvention of the Cretan Neolithic,” in Escaping the Labyrinth: The Cretan Neolithic in Context (Sheffield Studies in Aegean Archaeology 8), V. Isaakidou and P. Tomkins, eds., Oxford, pp. 21–48. Tomkins, P., and P.M. Day. 2001. “Production and Exchange of the Earliest Ceramic Vessels in the Aegean: A View from Early Neolithic Knossos, Crete,” Antiquity 75, pp. 259–260. Tomkins, P.D., P.M. Day, and V. Kilikoglou. 2004. “Knossos and the Earlier Landscape of the Herakleion Basin,” in Knossos: Palace, City, State. Proceedings of the Conference in Herakleion Organised by the British School at Athens and the 23rd Ephoreia of Prehistoric and Classical Antiquities of Herakleion, in November 2000, for the Centenary of Sir Arthur Evans’s Excavations at Knossos (BSA Studies 12), G. Cadogan, E. Hatzaki. and A. Vasilakis, eds., London, pp. 51–61.

4

Neolithic Sedimentology at Knossos Maria-Pilar Fumanal García†

This study was initially undertaken by the late Maria-Pilar Fumanal García, who took part in the fieldwork at Knossos in the winter of 1997 and carried out the sampling of the sections.* Un­for­ tunately, the analysis of the samples, which was already in an advanced stage at the Laboratorio de Geomorfología, Departamento de Geografía, Universidad de Valencia, was interrupted by her sudden death. Her colleague P. Carmona González kindly agreed to resume the study of the material

and to summarize the principal findings, taking into account the preliminary remarks made by Fumanal García in the field. This was a very difficult task, and we appreciate the efforts of Carmona González to make the most of the data collected by someone else. The following report may look incomplete in certain respects, but we hope nevertheless that it will contribute to the better understanding of the formation of the Knossos tell.

Geomorphological Background The archaeological site of Knossos is located 5.2 km from the north coast of the island approximately 90 m asl on a ridged surface of Middle Pliocene marls (kouskouras). The area to the east and west of the site is surrounded by Miocene bioclastic limestone platforms, 150–250 m high. The hill where Knossos is situated is bordered by the Kairatos

River to the east and its tributary Vlychia to the south; a gentle slope of approximately 20 m leads to the Kairatos riverbed. Holocene terraces can be *All figures and tables prepared by the author. Abbreviations used here are: EN=Early Neolithic, km=kilometers, m=meters, m asl=meters above sea level, MN=Middle Neolithic.

54

MARIA-PILAR FUMANAL GARCÍA

seen in the valleys of both rivers, their water flow being presently reduced due to irrigation needs. The shaping of the Pliocene marls is gently undulating with glacis surfaces sloping toward the north and the valley bottoms. The Miocene limestone of

the eastern platform presents an inclination of 15º to 20º, with talus morphology toward the Kairatos valley. The Pliocene glacis seen at the bottom of this slope are affected by gullies.

The Sedimentological Study During her fieldwork at Knossos in 1997, Fumanal García collected sediment samples from the western stratigraphic profile while the excavation was in progress. Her aim was to carry out sedimentological analyses at the geomorphology laboratory of the University of Valencia. The stratigraphy of the west profile is over 8 m deep. Altogether, 31 samples were collected, starting with sample I from the topmost layer and ending with sample XXXI at the base. The study commenced with sample IV (Table 4.1). The following analyses were carried out: 1. Granulometry of fine fraction (sand, silt, clay) based on phi scale (Fig. 4.1); for each sample, data were represented with histograms and accumulation curves

2. Calculation of statistical parameters of mean size, sorting, skewness, and kurtosis (Table 4.2) according to formulas of Folk and Ward (1957) 3. Organic content (Fig. 4.2) 4. Carbonate content (Fig. 4.3) 5. Morphoscopy of sands in samples that included calcareous grains (Fig. 4.4) 6. Morphoscopy of sands after elimination of calcareous grains (Fig. 4.5) 7. Munsell color (Table 4.2)

Sedimentological Description of the Samples Sample IV (Level 4, MN) Color 10YR 6/3 pale brown. Clayish silt sample with bones, small pieces of charcoal, and pottery. It is presented in aggregates and nodules. Texture analyses indicate very poor grain size sorting. In the analysis of siliceous grain morphoscopy the predominance of subangular and angular grains is observed, while calcareous grains are rounded and subrounded. The organic content is less than 1%. The carbonate content is 61%.

Sample V (Level 9, MN) Color 10YR 7/1 light gray. Clayish silt sample with charcoal, pottery, and bones. It is presented in aggregates and slightly hardened nodules. Texture

analyses indicate very poor grain size sorting. In the analysis of siliceous grain morphoscopy the predominance of subangular and angular grains is observed, while calcareous grains are subrounded and rounded. The organic content is 0.56%. The carbonate content is 64.4%.

Sample VI (Level 10, MN) Color 10YR 6/2 light brownish gray. Silty clay sample with bones, pottery, and abundant yellowishwhite clay nodules. The sediment is poorly sorted. The analyses of sand morphoscopy show similar results with the samples in the upper part of the profile. The carbonate content is 62.4%. The organic content reaches 1.03%.

55

NEOLITHIC SEDIMENTOLOGY AT KNOSSOS

Sample VII (Level 12, MN) Color 10YR 7/1 light gray. Silty clay sample with bones, pottery, and some charcoal. There are abundant whitish and gray clay nodules (with microcharcoals). Texture analyses indicate very poor grain size sorting. The form of sand grains is similar to the previously described samples. The organic content is 0.77%. The carbonate content is 65.6%.

Sample VIII (Level 13, MN) Color 10YR 7/2 light gray. Silty clay sample with aggregates. Microfragments of bones, shell, and charcoal are less abundant than in previous samples. The sample is poorly sorted. Morphoscopy analyses indicate a considerable increase of angular and subangular sands. The form of the sand grains is similar to the previously described samples. The carbonate content also increases. The organic content continues at 0.77%.

Sample IX (Level 14a, EN II) Color 10YR 5/2 grayish brown. Silty clay sample with abundant charcoal, clay nodules, bones, and pottery. Fragments of plaster are incorporated. The sample is poorly sorted. The sand morphoscopy shows a percentage of angular and subangular grains similar to samples IV and VII. The carbonate content is 62.6%. The content in organic matter increases compared to the previous samples and reaches the value of 2.25%.

Sample Xa (Level 16, EN II) Color 10YR 8/2 white. Clay-rich sample (57%) with silt and sand. Aggregates and nodules are incorporated. Bones or other archaeological material are not observed. Given the high clay content, the sorting is better than in other samples. Angular and subangular sands increase. The organic content decreases to 0.31%, and the carbonate content is 71.2%.

Sample Xb (Level 20, EN II) Color 10YR 7/3 very pale brown. Silty clay sample with aggregates and without anthropogenic remains. The degree of sorting is lower than in

Sedimentology Sample

Level

Cultural Phase

IV

4

MN

V

9

MN

VI

10

MN

VII

12

MN

VIII

13

MN

IX

14a

EN II

Xa

16

EN II

Xb

20

EN II

Xc

21

EN II

XI

23

EN II

XII

26+27

EN II

XIII

28+29

EN II

XIV

30

EN I

XV

30

EN I

XVI

30

EN I

XVII

30

EN I

XVIII

31

EN I

XIX

31

EN I

XX

32

EN I

XXI

32

EN I

XXII

32

EN I

XXIII

32

EN I

XXIV

32

EN I

XXV

33

EN I

XXVI

34

EN I

XXVII

35

EN I

XXVIII

36

EN I

XXIX

37

EN I

XXX

38

Aceramic

XXXI

39

Aceramic

Table 4.1. Correlation of sedimentology samples with excavation levels and cultural phases.

previous samples. The organic and carbonate content have values similar to the previous sample.

Sample Xc (Level 21, EN II) Color 2.5Y 8/2 white. Clayish silt sample with nodules and hardened aggregates. It incorporates charcoal and pottery bone fragments. The sample is very poorly sorted. The carbonate content is 56.8%.

MARIA-PILAR FUMANAL GARCÍA

56

Sample

Color

Mean Size (u)

Sorting (u)

Skewness (u)

Kurtosis (u)

IV

10YR 6/3 pale brown

6.44

2.21

-0.19

0.65

V

10YR 7/1 light gray

6.56

2.22

-0.21

0.69

VI

10YR 6/2 light brownish gray

6.82

2.27

-0.36

0.72

VII

10YR 7/1 light gray

7.05

2.28

-0.18

0.66

VIII

10YR 7/2 light gray

7.23

2.50

-0.27

0.72

IX

10YR 5/2 grayish brown

7.44

2.55

-0.39

0.75

Xa

10YR 8/2 white

6.96

1.67

-1.00

1.02

Xb

10YR 7/3 very pale brown

7.00

2.37

-0.36

0.65

Xc

2.5Y 8/2 white

6.85

2.02

-0.11

0.56

XI

10YR 6/2 light brownish gray

5.36

1.53

-0.04

1.21

XII

10YR 7/1 light gray

6.82

2.30

-0.17

0.62

XIII

10YR 6/3 pale brown

6.16

2.48

0.14

0.97

XIV

10YR 8/2 white

5.81

2.35

-0.04

0.67

XV

2.5Y 6/2 light brownish gray

6.43

3.03

-0.23

0.76

XVI

7.5YR 7/2 pinkish gray

6.04

1.81

-1.00

0.95

XVII

10YR 7/3 very pale brown

7.50

2.52

-0.35

0.65

XVIII

10YR 7/1 light gray

5.48

1.63

-1.00

0.98

XIX

2.5Y 8/2 white

7.42

2.34

-0.29

0.66

XXa

10YR 7/1 light gray

6.91

2.74

-0.21

0.69

XXb

10YR 6/1 gray

6.05

1.79

-1.00

0.95

XXc

7.5YR 7/2 pinkish gray

6.76

2.61

-0.26

0.67

XXIa

2.5Y 7/2 light gray

6.79

2.48

-0.14

0.58

XXIb

10YR 6/2 light brownish gray

7.09

2.99

-0.17

0.65

XXII

10YR 6/1 gray

6.30

2.84

-0.36

0.68

XXIII

10YR 7/3 very pale brown

6.66

2.85

-0.20

0.63

XXIVa

10YR 7/1 light gray

6.12

2.57

-0.14

0.70

XXIVb

10YR 6/1 gray

7.47

2.95

-0.18

0.69

XXV(1)

—

6.77

1.60

-1.00

0.98

XXV (2)

2.5Y 7/2 light gray

7.02

2.36

-0.28

0.66

XXVI

10YR 6/3 pale brown

7.23

2.71

-0.41

0.80

XXVII

7.5YR 5/2 brown

6.77

2.62

-0.35

0.64

XXVIII

10YR 5/6 yellowish brown

5.65

1.63

-1.00

0.92

XXIX

10YR 7/3 very pale brown

6.73

2.33

-0.20

0.65

XXX

7.5YR 5/4 brown

6.57

2.56

-0.27

0.69

XXXI

5YR 4/4 reddish brown

7.86

2.96

-0.30

0.62

Table 4.2. Munsell color and calculation of statistical parameters of mean size, sorting, skewness, and kurtosis for each of the analyzed sedimentology samples.

100 100 100

75 75 75

% 50

50 %% 50

25

25 25

0

00

IV V VI VII VIII IX Xa Xb Xc XI XII XIII XIV XV XVI XVII XVIII XIX XXa XXb XXc XXIa XXIb XXII XXIII XXIVa XXIVb XXV1 XXV2 XXVI XXVII XXVIII XXIX XXX XXXI

25

0 0

clay

rounded

silt sand

Figure 4.1. Fine fraction granulometry (%) of the samples from the west profile.

%%

subrounded

subangular

angular

Figure 4.4. Morphoscopy of sands without acid treatment. IV V VI VII VIII IX Xa Xb Xc XI XII XIII XIV XV XVI XVII XVIII XIX XXa XXb XXc XXIa XXIb XXII XXIII XXIVa XXIVb XXV1 XXV2 XXVI XXVII XXVIII XXIX XXX XXXI

% 50

2525 2525

00

00

IV IV V V VI VI VII VII VIII VIII IX IX Xa Xa Xb Xb Xc Xc XI XI XII XII XIII XIII XIV XIV XV XV XVI XVI XVII XVII XVIII XVIII XIX XIX XXa XXa XXb XXb XXc XXc XXIa XXIa XXIb XXIb XXII XXII XXIII XXIII XXIVa XXIVa XXIVb XXIVb XXV1 XXV1 XXV2 XXV2 XXVI XXVI XXVII XXVII XXVIII XXVIII XXIX XXIX XXX XXX XXXI XXXI

IV V VI VII VIII IX Xa Xb Xc XI XII XIII XIV XV XVI XVII XVIII XIX XXa XXb XXc XXIa XXIb XXII XXIII XXIVa XXIVb XXV1 XXV2 XXVI XXVII XXVIII XXIX XXX XXXI

100

IV V VI VII VIII IX Xa Xb Xc XI XII XIII XIV XV XVI XVII XVIII XIX XXa XXb XXc XXIa XXIb XXII XXIII XXIVa XXIVb XXV1 XXV2 XXVI XXVII XXVIII XXIX XXX XXXI

IV IV V V VI VI VII VII VIII VIII IX IX Xa Xa Xb Xb Xc Xc XI XI XII XII XIII XIII XIV XIV XV XV XVI XVI XVII XVII XVIII XVIII XIX XIX XXa XXa XXb XXb XXc XXc XXIa XXIa XXIb XXIb XXII XXII XXIII XXIII XXIVa XXIVa XXIVb XXIVb XXV1 XXV1 XXV2 XXV2 XXVI XXVI XXVII XXVII XXVIII XXVIII XXIX XXIX XXX XXX XXXI XXXI

NEOLITHIC SEDIMENTOLOGY AT KNOSSOS

7575 7575

5050 5050

rounded

subrounded

57

2.5

75 2

% 1.5

1

0.5

Figure 4.2. The organic content (%) of the samples from the west profile.

%%

Figure 4.3. The carbonate content of the samples from the west profile: percentage of general calcimetry (left) and percentage of Bernard calcimetry (right).

subangular

angular

Figure 4.5. Morphoscopy of sands after the elimination of calcareous grains (subsequent to acid treatment).

58

MARIA-PILAR FUMANAL GARCÍA

Sample XI (Level 23, EN II)

Sample XVI (Level 30, EN I)

Color 10YR 6/2 light brownish gray. Sandy silt sample with abundant and quite altered clay nodules and charcoal. It contains bone and pottery remains. In this sample the sorting increases slightly. The sands are angular and subangular. The carbonate content is 55.4%.

Color 7.5YR 7/2 pinkish gray. Silty clay sample with charcoal, bone, and pottery fragments as well as pieces of broken pebbles. The sample is very poorly sorted. The carbonate content reaches the value of 60.7%, and the organic content is 0.41%. The sand grains present the same characteristics as in previous samples.

Sample XII (Levels 26 and 27, EN II) Color 10YR 7/1 light gray. Sandy silt sample with abundant pottery and bone remains. Clays are present in aggregates and nodules of variable size. The sample is very poorly sorted. The carbonate content is 62.1%, and the organic content is 0.74%.

Sample XIII (Levels 28 and 29, EN II) Color 10YR 6/3 pale brown. Clayish silt sample with bone and pottery remains, charcoal and clay nodules. It is presented in the form of aggregates. The sample is slightly sorted. The carbonate content is 28.8%. The sand morphoscopy, with the calcareous material included, indicates a predominance of rounded and subrounded grains. The organic content is 0.92%.

Sample XIV (Level 30, EN I)

Sample XVII (Level 30, EN I) Color 10YR 7/3 very pale brown. Silty clay sample in the form of aggregates that include some rounded pebbles, angular stones, and charcoal. The sample is very poorly sorted. The carbonate content reaches 50%. Organic matter is not present. The proportion of angular sand grains increases slightly.

Sample XVIII (Level 31, EN I) Color 10YR 7/1 light gray. Silty clay sample in the form of aggregates and nodules with charcoal and bones. The sample is poorly sorted. The carbonate content is 59%, and the organic content is 0.92%. The proportion of angular and subangular sand grains is similar to the previous samples.

Sample XIX (Level 31, EN I)

Color 10YR 8/2 white. Clayish silt sample with fragments of charcoal, bone, and pottery. It is presented in the form of aggregates and nodules. The sample is well sorted. The carbonate content is 59%, and the organic content is 0.31%. The sands, including the calcareous grains, are rounded and subrounded.

Color 2.5Y 8/2 white. Silty clay sample in the form of aggregates with slightly hardened nodules. The sample is very poorly sorted. The carbonate content is 62.4%, and the organic content is null. The proportion of angular and subangular sand grains is similar to the previous samples.

Sample XV (Level 30, EN I)

Sample XXa (Level 32, EN I)

Color 2.5Y 6/2 light brownish gray. Silty clay sample with fragments of charcoal, bone, and pottery, along with some small rounded pebbles. The sample is very poorly sorted. The carbonate content reaches the value of 62.3%, and the organic content is 1.25%. As in previous samples, the calcareous sands are rounded and subrounded.

Color 10YR 7/1 gray. Silty clay sample. It is presented in the form of nodules that include charcoal and bone fragments. The sample is very poorly sorted. The carbonate content is 60.1%, and the organic content is 0.92%. The proportion of rounded and subrounded grains increases slightly.

NEOLITHIC SEDIMENTOLOGY AT KNOSSOS

59

Sample XXb (Level 32, EN I)

Sample XXIII (Level 32, EN I)

Color 10YR 6/1 gray. Silty clay sample. It is presented in the form of aggregates and nodules. The sample includes charcoal, bone remains, and lithic material. It is very poorly sorted. The organic content is 1.25%. The proportion of rounded and subrounded sand grains is similar to the previous sample.

Color 10YR 7/3 very pale brown. Silty clay sample. It is presented in the form of aggregates that include ash, charcoal, bone, and pottery fragments. The sample is very poorly sorted. The carbonate content is 52.9%, and the organic content is 0.52%. The carbonated sand grains are rounded and subrounded.

Sample XXc (Level 32, EN I)

Sample XXIVa (Level 32, EN I)

Color 7.5YR 7/2 pinkish gray. Silty clay sample. It is presented in the form of aggregates and nodules. The sample includes charcoal, bone remains, and lithic material. It is very poorly sorted. The carbonate content is 59.1%. The organic content is 1.42%. The proportion of rounded and subrounded grains is similar to the previous sample.

Color 10YR 7/1 light gray. Clayish silt sample. It is presented in the form of aggregates that include charcoal, a few sherds of pottery, and bone remains. The carbonate content is 59.8%, and the organic content is 1.34%. The sample is poorly sorted. The carbonate grains are rounded and subrounded.

Sample XXIa (Level 32, EN I)

Sample XXIVb (Level 32, EN I)

Color 2.5YR 7/2 light gray. Clayish silt sample. It is presented in the form of nodules with rare small pottery and bone fragments. The sample is poorly sorted. The carbonate content is 60.6%. It does not contain organic matter. The carbonated sand grains are rounded and subrounded.

Color 10YR 6/1 gray. Silty clay sample. It is presented in the form of aggregates with small charcoals. The sample is poorly sorted. The carbonate content is 58.5%, and the organic content is 1.08%. The carbonate grains are rounded and subrounded.

Sample XXV 1 (Level 33, EN I) Sample XXIb (Level 32, EN I) Color 10YR 6/2 light brownish gray. Silty clay sample. It is presented in the form of aggregates that include charcoal and some pebbles. The sample is poorly sorted. The carbonate content is 59.5%, and the organic content is 1.28%. The carbonated sand grains are rounded and subrounded.

Silty clay sample. It is presented in the form of aggregates and small nodules. No organic remains are observed. The sample is poorly sorted. The carbonate content is 59.3%, and the organic content is 0.61%. The carbonated sand grains are subangular and subrounded.

Sample XXV 2 (Level 33, EN I) Sample XXII (Level 32, EN I) Color 10YR 6/1 gray. Silty clay sample. It is presented in the form of aggregates that include charcoal, bones, and pottery fragments. The sample is very poorly sorted. The carbonate content is 60.2%, and the organic content is 1.66%. The carbonated sand grains are rounded and subrounded.

Color 2.5Y 7/2 light gray. Silty clay sample. It is presented in the form of nodules and aggregates. No organic remains are observed except for a few charcoal fragments. The sample is poorly sorted. The carbonate content is 57.2%, and the organic content is 0.64%. The percentage of subangular

60

MARIA-PILAR FUMANAL GARCÍA

and subrounded carbonated sands corresponds to almost the whole sample.

Sample XXVI (Level 34, EN I) Color 10YR 6/3 pale brown. Silty clay sample. It is presented in the form of aggregates including charcoal and bone fragments. The sample is very poorly sorted. The carbonate content is 26.8%, and the organic content is 1.15%. The sample includes carbonated material and is dominated by sands of subangular and subrounded form.

Sample XXVII (Level 35, EN I) Color 7.5YR 5/2. Silty clay sample in the form of nodules with small charcoal fragments. The sample is very poorly sorted. The carbonate content is 55.4%, and the organic content is 0.24%. The sample includes carbonated material and is dominated by sands of subangular and subrounded form.

Sample XXVIII (Level 36, EN I) Color 10YR 5/6 yellowish brown. Silty clay sample. It is presented in the form of aggregates or nodules with a few bone remains and charcoal fragments. The sample is very poorly sorted. The carbonate content is 38.7%, and the organic content is 0.55%. The sample includes carbonated material and is dominated by sands of subangular and subrounded form.

Sample XXIX (Level 37, EN I) Color 10YR 7/3 very pale brown. Clayish silt sample presented in the form of aggregates including charcoal. The sample is poorly sorted. The carbonate content is 52.6%, and the organic content is 0.11%. The sample includes carbonated material and is dominated by sands of subangular and subrounded form.

Sample XXX (Level 38, Aceramic Neolithic) Color 10YR 5/4 brown. Silty clay sample. It is presented in the form of aggregates that include charcoal and construction remains. The carbonate content is 46.8%, and the organic content is 1.15%. The sample includes carbonated material and is dominated by sands of subangular and subrounded form.

Sample XXXI (Level 39, Aceramic Neolithic) Color 5YR 4/4 reddish brown. Silty clay sample. It is presented in the form of aggregates with small white nodules. The carbonate content is 26.3%, and the organic content is 0.45%. The sample includes carbonated material and is dominated by sands of subangular and subrounded form.

Interpretation and Discussion of the Results Sediment analyses indicate that the texture of all levels is silty clay or clayish silt. In very few samples sand reaches values over 20% of the total, while in many of them the sands content is 10%– 15%. These results, together with the light coloration of the sediment (pale brown, gray, or white), indicate that the bedrock out of which this material formed is the Pliocene marl of the surroundings and of the site’s substrate.

The content in organic matter is extremely variable along the profile. Samples IX (described as a burned layer) and XXII (described as red earth in the profile and gray soil in the sedimentology sample) stand out for their high organic content. Samples XVII, XIX, and XXIa do not contain any organic material. The latter two probably correspond to natural marls with minimal contamination (in line with the description of the profile),

NEOLITHIC SEDIMENTOLOGY AT KNOSSOS

while the former one is described in the profile as reddish soil. The basal part of the profile, from which samples XXVII, XXVIII, XXIX, and XXXI were derived, shows a very low percentage of organic content. The analyses of the carbonate contents show a homogeneous distribution of percentages along the profile, although the values of carbonate content decrease notably from sample XXVII to sample XXXI. The red and brown colors of those basal

samples, along with their low content in organic matter, might indicate either a different provenance of the sediment or a mixture with other materials, probably clays and silt from the nearby riverbed. It should be noted, however, that we do not know the exact location and altitude of the profile in relation to the river valley. There is a possibility that the bedrock under the site was a river terrace adjacent to the marls.

References Folk, R.L., and W. Ward. 1957. “Brazos River Bar: A Study in the Significance of Grain Size Parameters,” Journal of Sedimentary Petrology 27, pp. 3–26.

61

5

The Economy of Neolithic Knossos: The Archaeobotanical Data Anaya Sarpaki

The archaeobotanical data presented in this paper came principally from the 1997 rescue excavation in the southeastern area of the Central Court of the Palace of Knossos.* The discussion is enriched, however, with information from the unpublished report of Hans Helbaek on the Neolithic botanical material excavated by J.D. Evans (Evans 1964, 1971, 1994; Renfrew 1979). I have also examined some seed finds from Evans’s excavation *All figures and tables prepared by the author. Abbreviations used in this chapter are: B breadth BM British Museum lab code cal. calibrated or calendar years cf. compare Ch(s). Chapter(s) cotyl. cotyledon EN Early Neolithic fr. fragment(s) I Teledyne Isotopes, Westwood, NJ, lab code km kilometers L length LN Late Neolithic lvs. leaves

stored in the Stratigraphical Museum at Knossos, but more arch­aeobotanical material recovered by Evans remains to be studied. Unlike Evans’s samples, the archaeobotanical finds from the 1997 excavations de­rive from a long diachronic sequence extending from the Aceramic through the Late Neolithic (LN) periods. The study of this material may provide an insight into the cultural origins of the Neolithic of m min. mm MN No. OxA PPN s.l. sp. spp. subsp. T

meters mineralized millimeters Middle Neolithic number Radiocarbon Accelerator Unit, Oxford University, lab code Pre-pottery Neolithic (with phases PPNA, PPNB, PPNC) sensu lato, meaning “loose sense” species various species subspecies thickness

64

ANAYA SARPAKI

Knossos and a better understanding of the agricultural preferences, practices, and food procurement strategies of Knossos in comparison with other Neolithic sites of mainland Greece and the eastern Mediterranean in general. A more complete picture of Neolithic subsistence should emerge after additional material from Knossos and other sites in Crete has been studied (Sarpaki 2009). Knossos and Phaistos are the earliest Neolithic sites in Crete that have been studied to date. Earlier sites have been identified at Trypiti and Rouses near Herakleion and at Asfendou near Sphakia in western Crete, however, and more examples may be found as research progresses (Zois 1973, 58– 66; Kopaka and Matzanas 2009; see also van Andel and Runnels 1995). Strasser (1996, 327–328) has noted that the first farmers were quite aware of their environment, as attested by their choice of settlement locations and of cultivars. At present the archaeobotanical evidence suggests that the Aceramic inhabitants of Knossos were either immigrants who established themselves on Crete and introduced a “Neolithic package” of economic plants and animals, or local huntergatherers who most likely imported agricultural produce. While “migration” and “diffusion” are taboo words for some archaeologists, it is time to revisit these concepts with fresh data, a step that may help us reformulate our explanations for developments in Crete (Sherratt 1996, 140). The imported species observed in the early agricultural complex include a wide range of plants and animals such as cattle (see Pérez Ripoll, this vol., Ch. 8; Cherry 1990, 161), sheep, goats, pigs, dogs, cereals, and perhaps legumes. Evans (1994, 5) hypothesized a link with Asia Minor, since, although the other plants and animals present have been attested at contemporary sites in mainland Greece as well as in western Asia, bread wheat is known to have been cultivated specifically at Çatalhöyük. The practice of building with mudbricks (see Efstratiou, Karetsou, and Banou, this vol., Ch. 1) also points to sites such as Aceramic Haçilar and Aşikli Höyük in southwestern Asia Minor as one possible area of dissemination. The closest parallels for the coarse pottery occur in western Anatolia and the eastern Aegean islands (Warren et al. 1968, 273; Sakellarakis 1973). Therefore, it appears that the Neolithic cultivars

could have originated in the area extending from western Turkey to Palestine because the region of the Syro-Levant and Turkey is where agriculture first originated. The evidence from Franchthi points to the acquisition of obsidian, albeit in tiny amounts, in Upper Paleolithic levels dated to 10,880 ± 160 b.p. in radiocarbon years (I-6129; Perlès 1979). This is an indication of what must have been a more widespread pattern of movement and exploration, further attested by the three pieces of obsidian found at Knossos (two of Early Neolithic [EN] I and one of EN II date) originating from the island of Giali and suggesting a probable contact with southwestern Asia Minor (Evans 1994, 5 n. 10). Whether the introduction of farming involved immigration or importation, the necessary sailing might have been accomplished with the level of technology available if islands were used as stepping stones for voyages. One possible route might have passed by Rhodes, Karpathos, and Kasos. Another might have skirted the islands of the Dodecanese (such as Kos) and the southern Cyclades (Anaphi, Thera). Sakellarakis (1973, 134) favors the immigration to Crete from the Cyclades or the Dodecanese. A third passage might have been undertaken via the Peloponnese, Kythera, and Antikythera, or even through Attica via some of the Cycladic islands (see Broodbank 1999, 34, fig. 1.9; 2000, 135, fig. 38; see also Cherry 1985 for possible colonization sequences). Lambeck (1996) discusses sea level changes, which must have been an important factor affecting colonization in the early part of the Neolithic. Crete is visible at certain times of the year from some of the islands along each of these routes (e.g., Karpathos, Thera, and Kythera). There is no evidence so far to confirm or refute any of these hypothetical journeys, and it is very probable that there were two or more such events, whether simultaneous or successive. For Greece proper, Perlès (2001) has argued that colonization was a maritime phenomenon that started with small groups of people from different points of origin. She later adds that these migrations could have occurred in different periods (Perlès 2005, 280), and one might suggest that parallel processes were underway in Crete.

THE ECONOMY OF NEOLITHIC KNOSSOS: THE ARCHAEOBOTANICAL DATA

65

Background to Cultivation Modern Crete has a Mediterranean-type climate characterized by hot, dry summers and cool, wet winters. The climate of Neolithic Crete seems to have been less arid than that of the present, possibly comparable to that of modern Epirus, as suggested by the evidence from pollen cores taken from Tersana and Limnes (Moody, Rackham, and Rapp 1996; Rackham and Moody 1996, 39). Knossos is located in one of the driest areas of the island. It presently receives an average annual rainfall of under 600 mm, considerably less than the levels observed in areas to the west, which may receive up to 1,800 mm. It is likely that in the past, as is the case today, arable agriculture was based upon autumn- and winter-sown crops because crops planted in the spring would have received insufficient rainfall to survive spring and summer droughts. The soil around the site is light, fine-grained, friable marl and redzina soils—which are characterized by a horizon of high organic content and which are highly arable—that are very workable and fairly fertile (Jarman, Bailey, and Jarman, eds., 1982, 147). Vines and olive trees are grown there today. For people who cultivated the soil with the hoe or digging stick rather than the plow, the choice of a site with a friable, easily cultivable soil was of paramount importance (Strasser 1996, 327–328). Local soils may have changed somewhat since the Neolithic, however, due to the effects of erosion and other anthropogenic and climatic events. The Kairatos River, located immediately below the site to the east, along with its tributary, the Vlychia, would have been a major source of water for Knossos. The Kairatos rises from springs near Archanes, and before the use of pump irrigation, it was a perennial river (Roberts 1979, 231). Furthermore, before recent seismic activity, local springs and a high water table adjacent to the site justified the digging of wells, although we cannot say whether wells were present in prehistoric times. The wheat found at ancient Knossos (especially Triticum turgidum/aestivum L.) certainly indicates that water was available in some form (Sherratt 1980).

Thus, Neolithic Knossos occupied an area of prime agricultural land. Nearby there were highquality redzina soils suitable for cultivation. Some of the steeper areas south of Knossos on Mt. Juktas could have been used for grazing, along with the surrounding lowlands, which might have sustained the cattle whose bones were recovered from the site. As the whole of Crete is characterized by pockets of varied microenvironments, a type of small-scale, mixed farming may have been favored, as Halstead (1996a; 1996b, 35) has suggested. Contrary to some present views, a form of “transhumance” may also have been practiced, as was the case in several areas of Crete until recently (for a thorough history of transhumance, see Arnold and Greenfield 2006). So far, the evidence indicates that the residents of Neolithic Knossos were full-fledged mixed far­ mers. No crops were found at the transitional stage between “wild” and “domesticated,” although we cannot exclude the stage of “cultivation” for some plants, such as the radish, the fig, and the almond. If the first settlers were colonists, we need to understand why they left their homes and what the “pressures” were that pushed them to migrate from their native cultural, social, and natural environments. If they were not immigrants, we need to understand what made the local people change from hunting and foraging to farming for a livelihood. The answers to such questions are undoubtedly complex, and they must be addressed by integrating interdisciplinary research. The available climatic data suggest that following the Late Glacial global warming, a cold spell known as the Younger Dryas took place from ca. 10,800 to 9800 b.c. in northern Europe and globally. This triggered extremely dry conditions in the Near East, and it seems to have been followed by the onset of agriculture (Blumler 1996, 41). It has been demonstrated by ample evidence that temperature change is not as crucial a factor as moisture for plant species. The pollen figures, according to Bottema (1992, 104), demonstrate that precipitation declined at the same pace as temperature

66

ANAYA SARPAKI

change, resulting in greater aridity. This is believed to have happened at the end of the Pre-Pottery Neolithic A (PPNA) and the beginning of the PPNB in northern Syria (Helmer et al. 1998, 30) and the Levant (Peltenburg et al. 2000). Although the effects were archaeologically noted in those periods, the trend would, nevertheless, have started earlier without leaving any evidence that has been detectable to date. The lowering of temperatures and the diminishing of precipitation probably led to an exodus and a bottleneck situation, as Sherratt (1996, 136) infers for the Near East, as communities searched for improved conditions for mixed agriculture. Cyprus seems to have received a first wave of settlers some 500–1,000 years later in the early PPNB period (Peltenburg et al. 2000). Farther afield, Crete may have experienced a related demic influx approximately 1,000 years after Cyprus. The bioarchaeological remains found to date at Knossos represent an agro-pastoral farming

regime that was established at the first habitation of the site. It was difficult to understand why cattle were preferred to sheep and goats at the earliest periods of the site’s existence under relatively unfavorable environmental conditions. However, according to the recent study by Pérez Ripoll (this vol., Ch. 8), sheep/goat predominated in the EN, and cattle did not rise significantly until the Middle Neolithic (MN). The Delphinos core, sampled west of Rethymnon (ca. 80 km to the west of Knossos) points, however, to a landscape of open vegetation with low tree cover around 7375–6310 cal. b.c. (Bottema and Sarpaki 2003). The reasons behind agricultural and pastoral choices may well have been triggered not only by economic constraints and environmental conditions but also by cultural choices. Those choices are often “veiled” by their “biographies,” or social histories, as well as by their taphonomic histories.

The Archaeobotany The archaeobotanical material from Knossos comes from three sources. One sample was a cache of seeds found by J.D. Evans (Evans 1994) in Level IX in a single specific context—the socalled Aceramic level—referred to as stratum X in earlier excavations (Warren et al. 1968, 269, 272). This was the sample that Helbaek studied but never properly published (Table 5.1). Some other material was retrieved by Michael and Heather Jarman (1968), but the archaeobotanical study was never completed. The third group of samples came from the 1997 rescue excavation, from which 33 samples of soil were water-floated (Table 5.2). A majority of the levels from this investigation were sampled for bioenvironmental data, the exceptions being levels 1, 5, 6, 11, 15, 16, 18, 19, 21, 22, 25– 27, 30, 36, and 38. The soil processed amounted to ca. 600 liters. The nature of the information from each data source is different but complementary. Evans believed that Helbaek’s material was the threshing product of a crop containing contaminants and

weeds. He mentioned that in area AC, which measured 11 X 5 m, no buildings were found, but there was evidence of other activities including threshing and corn grinding (Evans 1994, 2). He also stated (Evans 1994, 4) that “in one area grain from a field of breadwheat had apparently been threshed.” One side of this area had a row of stake holes from which a radiocarbon date was obtained. This proved similar to a date obtained from level 39 of the 1997 excavations (see Facorellis and Maniatis, this vol., Table 10.1: BM-436, Area AC, Stratum IX, 7740 ± 140 b.p. or 7050–6370 cal. b.c. on carbonized grain; cf. Table 10.3: OxA-9215, level 39, dated to 7050–6690 cal. b.c.). Unfortunately, the water-floated material from the 1997 rescue dig cannot be assigned to interior or exterior spatial contexts (see Efstratiou, Karetsou, and Banou, this vol., Ch. 1), but it represents a long diachronic sequence. Whenever there is any structural or featural evidence pertaining to context, however, it is discussed below.

67

THE ECONOMY OF NEOLITHIC KNOSSOS: THE ARCHAEOBOTANICAL DATA Seed Type Wheat grains

Description

Sorted in Copenhagen

bread wheat

2,900

emmer einkorn

Wheat spikelet fragment

10 9

4

hulled, twisted

2

2

naked, straight

12

1

naked, twisted

2

bread wheat

3

emmer

191*

einkorn hulled, two-row Barley spike fragment**

Lentil

Weeds

2,500 10

hulled, straight Barley grains

Sorted in England

1

9

hulled, six-row

3

naked, six-row

none

seeds

210

wild oats (awn)

1

ryegrass (grain)

1

mallow (nutlets)

52

plantain (seed)

1

140

Table 5.1. Seed list provided to J.D. Evans by Hans Helbaek (unpublished). *Helbaek did not separate einkorn from emmer grains. **Here he must have meant rachis fragment. 1997 Level Number

Water Flotation Number

Relative Date

J.D. Evans Strata and Dates* (uncal. b.p.)

2

E 97(1)

LN

3

E 97(2a) E 97(2b)

LN

4

E 97(4) E 97(5a)

MN

—

7

E 97(6a)

MN

—

8

E 97(5b)

MN

—

9

E 97(6b) E 97(7)

MN

Calendar Age (b.c.) —

Stratum II

Stratum III

—

5470–4850

10

E 97(8)

MN

—

10b

E 97(9)

MN

—

12

E 97(10) E 97(11)

MN

4990–4731

14

E 97(12) E 97(13a) E 97(13b)

EN II

4982–4774

17

E 97(14)

EN II

20

E 97(15)

EN II

23

E 97(16)

EN II

—

24

E 97(17)

EN II

5208–4936

28

E 97(18)

EN II

5042–4779

29

E 97(19) E 97(20)

EN II

— Stratum IV

Stratum IV

—

5000–4730

Table 5.2. List of archaeobotanical samples from the 1997 rescue excavation, along with relative and absolute dates. *Only short-lived samples (i.e., seeds/grains) are included here. For all the others, see Facorellis and Maniatis (this vol., Ch. 10).

ANAYA SARPAKI

68

1997 Level Number

Water Flotation Number

Relative Date

31

E 97(21)

EN I

5310–5000

32

E 97(22)

EN I

5211–5016

E 97(23)

33

E 97(24) E 97(25)

34

E 97(26a) E 97(26b)

35

E 97(27)1

37

E 97(28)2

39

E 97(30)

3

EN I

J.D. Evans Strata and Dates* (uncal. b.p.)

Calendar Age (b.c.)

5220–4950

Strata VI–V

EN I

—

EN I

5468–5228

EN I

Strata IX–VI

5300–5000

Aceramic

Stratum IX: 7740 ± 130 b.p. (BM-436)** Stratum X: no seeds dated

7050–6690 OxA-9215

Table 5.2, cont. List of archaeobotanical samples from the 1997 rescue excavation, along with relative and absolute dates. *Only short-lived samples (i.e., seeds/grains) are included here. For all the others, see Facorellis and Maniatis (this vol., Ch. 10). **The dated grains of naked wheat originally assigned to Stratum X (Warren et al. 1968, 272) were later attributed to Stratum IX (Evans 1994, 20). 1 T. spp 5297–5055 b.c. (calibrated) (2 sigma) (OxA-21420). 2 T. turgidum/aestivum 5206–4843 b.c. (calibrated) (2 sigma) (OxA-21419). 3 Triticum spp. 6639–6480 b.c. (calibrated) (2 sigma) (OxA-21418).

Plant Species

Common Name

Count

Amygdalus communis

almond

(1)

Ficus cf. carica fragment

fig

(4)*

Lens culinaris

lentils

2

cf. Lens sp.

(4)

Legume fragments

(6)

Leguminosae (medium)

1

cf. Trifolium sp./cf. Astragalus sp.

clover

1

Triticum sp.

wheat

5 + (3)

Triticum turgidum L./aestivum L.

1

Triticum sp. glume base

1**

Hordeum sp. hulled

barley

(3)

Cerealia sp. (Triticum or Hordeum)

(2)

cf. Cerealia fragments

(26)

Total

11 + (49)

Table 5.3. Aceramic Neolithic archaeobotanical sample E 97(30) from Knossos 1997 level 39, retrieved from 16 liters of water-floated soil. Values shown in parentheses represent fragments. *All fragments charred. **More like T. dicoccum but too fragmented.

THE ECONOMY OF NEOLITHIC KNOSSOS: THE ARCHAEOBOTANICAL DATA

69

The Aceramic Sample Material from the Aceramic period includes the sample from what is now known as Stratum IX (not X) of Evans’s excavation, examined by Helbaek (Table 5.1), and sample E 97(30) from 1997 level 39, examined by the author (Table 5.3). Various species of cereals and pulses are represented, along with fruits such as the almond (Prunus amygdalus) and the fig (Ficus carica). All of the crops present should be considered fully domesticated, except for the almond and the fig, which could have been wild, as discussed at greater length in conjunction with the EN II finds. The most striking archaeobotanical find of the Aceramic was the presence of free-threshing bread wheat, identified by Helbaek as Triticum aestivum, which, he stated, was in the “overwhelming majority” (Helbaek 1968, 5). The 1997 rescue dig produced, regrettably, only grains that are preferably to be referred to as T. turgidum L./aestivum L. (Fig. 5.1). No rachis was preserved in the waterfloated sample. There is no secure way of separating tetraploid from hexaploid wheats on the basis of grain shape and size; only rachis morphology can be used to distinguish their ploidy level. Helbaek found rachis fragments in his archaeobotanical sample, and he observed that the spike must have been dense rather than compact. On compact spikes, he wrote, “[t]he narrow tops of the internodes suggest that the spikelet did not yet normally develop several grains as in later times, but was two-grained like emmer, one of its direct parents” (Helback, unpublished report, Knossos). He noted that a compact spikelet would have been no more than 1 mm long, whereas in the Orient the internodes of bread wheat (T. aestivum) varied from 2 to 4 mm in length. One complete internode from Helbaek’s sample measured 2.20 mm long and 1.48 mm at its widest point. Unfortunately, it was not illustrated (Helbaek unpublished report, Knossos). The measurements reported by Helbaek are comparable to T. parvicoccum as identified by Kislev (1979–1980, 99–100). DNA analysis is needed to verify the presence of T. parvicoccum as some archaeobotanists believe it is an aberration (see Kislev’s 2009 publication of T. turgidum sp. parvicoccum). The present author plans to restudy the archaeobotanical material from Evans’s

excavations and will check for any unpublished notes relating to this material. It is important to elucidate the problem of whether the Knossos wheat is a free-threshing tetraploid (4X, T. turgidum L.) or a hexaploid (6X, T. aestivum), for it would add to our knowledge about the spread of these wheats in the eastern Mediterranean. It is well known that T. aestivum evolved under cultivation from the already cultivated T. turgidum L. stock (Helbaek 1970, 211; Maier 1996, 47; Zohary and Hopf 2000, 51). It is a product of hybridization between a tetraploid T. turgidum L. wheat and a diploid wild grass, Aegilops squarrosa L. This wild grass does not occur in the Mediterranean-bordering regions of the Near East, although it is present in temperate parts of central Asia, such as northern Iran, Transcaucasia, and Afghanistan. Thus, T. aestivum could have developed only after the domestication of emmer (T. dicoccum) and the spread of cultivated tetraploid (T. turgidum L.) wheat to northern Iran and adjacent Transcaucasia. This expansion is believed to have occurred sometime between 6000 and 5000 b.c. (Maier 1996, table 2, figs. 8, 9; Zohary and Hopf 2000, 54), postdating the postulated occurrence of hexaploids in Crete some 1,000–1,500 years earlier. Experimental evidence has indicated that the first hexaploid wheats were spelt-like, or hulled (Zohary and Hopf 2000, 52), although genetic analysis shows that naked hexaploid wheats could have evolved rather quickly from their more primitive, hulled relatives (Zohary and Hopf 2000, 57). As noted above, the material presented in this study consists only of the grains of free-threshing wheat; no chaff of this cereal was retrieved from water flotation. The overall morphological character

0

5 mm

Figure 5.1. Drawing of Triticum turgidum L./T. aestivum from the 1997 excavations at Knossos. Drawing A. Kontonis.

70

ANAYA SARPAKI

of the grain is short and blunt. The dorsal view of the grains is rather rounded in comparison to the more truncated and acute dorsal view of tetraploids (Kislev 1984, 143), and the outline of the cheeks (ventral view) is also rounded instead of angular as tetraploids tend to be. Due to the effects of charring deformation on the grains, however, definitive identification of species cannot be made without the rachis internodes. Jacomet has made several charring experiments on Cerealia, and she specifically mentions that “the most varied shapes are produced” (1987, 40; see similar observations by Hopf 1955; Helbaek 1970; van Zeist 1972, 49). Measurements of the grains of T. aestivum s.l. and T. aestivo-compactum from Knossos (Table 5.4; Fig. 5.2) have been presented for comparison in Figure 5.3 with the data from the sites of Erbaba, Ramad, and Bouqras studied by Jacomet (1987, 58), van Zeist and Buitenhuis (1983), and van Zeist and Waterbolk-van Rooijen (1985). The Knossos finds fall largely within these measurements and are consistent with those of the denseeared form of bread wheat, T. aestivo-compactum Schiem, or club wheat (van Zeist 1972, 53–54; Jacomet 1987, 38–40). According to Jacomet, the ratios of the length and breadth measurements are crucial for distinguishing between the T. aestivum group (lax-eared bread wheat) and the T. aestivo-compactum type. She claims that the length/breadth ratio of T. aestivum is 50 fragments.

77

THE ECONOMY OF NEOLITHIC KNOSSOS: THE ARCHAEOBOTANICAL DATA E 97 (28)

E 97 (27)

E 97 (25)

E 97 (24)

E 97 (23)

E 97 (22)

E 97 (21)

1997 Level/ J.D. Evans Stratum

37/VI

35/VI

35/VI

34/VI

34/VI

33/VI

33/VI

32/VI

31/VI

 

Liters

14

10

14

14

16.5

15

10

15

15

123.5

Malva cf. sylvestris

—

—

—

1

2

—

—

—

1

4

Malva (type B)

—

—

—

—

—

—

—

—

1

1

Galium aparine

—

—

—

—

1

—

—

—

—

1

Boraginaceae fr.

—

—

—

1

—

—

—

—

—

1

Verbena officinalis

—

—

—

—

—

1

—

—

—

1

Rubiaceae (cf. Valerianella sp.)

2

—

—

—

—

—

—

—

—

2

Total

8

3

2

7

19

2

1

0

15

57

Ignota (identifiable?)

—

7

3

3

9

1

—

—

—

23

Ignota fr. (very damaged)

—

1

7

11

23

13

5 (+++)

20

15

95

Ignota (featureless)

4

—

—

—

—

—

—

—

—

4

Ignota (shell) fr.

—

—

—

—

—

—

—

—

—

1

cf. spores

—

—

—

—

3

—

—

—

—

3

Total

4

8

10

14

35

15

5

20

15

126

Sample Number

E 97 (26b)

E 97 (26a)

Total 

Weeds, cont.

Ignota

Table 5.6, cont. Early Neolithic I archaeobotanical (seed) samples. *In addition to counted specimens, (+) = up to 10 fragments that cannot be counted; (++) = 11–50 fragments; (+++) > 50 fragments.

material that sinks is subjected to greater mechanical damage than the material that is prone to float. Another problem is that pulses often explode when they come in contact with water. Nonetheless, although we used a Siraf-type water flotation machine, which made its appearance in the 1970s (and is still recommended; see de Moulins 1996), and not a water-sieving or froth flotation machine, the legumes were fairly well represented. Among the Leguminosae observed were not only Lens sp. but also possibly the pea, Pisum sp., the horsebean, Vicia faba, and the clovers/medicks, Trifolium sp./Medicago sp. (Table 5.7). The lastmentioned could either have been remnants of fodder, suggesting that animals were stalled, or it could be indicative of the fuel used, that is, animal dung

(Miller 1984). Its presence is most noticeable in levels 33–35. The main cereal crop seems to have been the naked wheat T. turgidum/aestivum, but einkorn and emmer also appear to have been present, together with barley. An interesting increase in fruit was observed, along with the appearance of two plants whose presence needs further study (Table 5.8). One is the wild radish (Raphanus raphanistrum), which is encountered as a weed of cultivation, yet its unusually high presence in some of the Knossos samples suggests that its use should be further investigated. Some form of exploitation of this plant may be attested, especially as the sample was collected from what has been described as a floor, on which hearths 5 and 6 were excavated. The

ANAYA SARPAKI

78 Sample Number

Phase

Length

Breadth

Thickness

L:B*

L:T**

B:T***

E 97(6b)

MN

E 97(6b)

MN

0.75

0.5

—

1.50

—

—

0.8

0.55

—

1.45

—

—

Trifolium spp.

E 97(9)

MN

0.8

0.5

—

1.60

—

—

E 97(13b)

EN II

0.7

0.6

0.4

1.17

1.75

1.50

E 97(13b)

EN II

1.1

0.7

0.6

1.57

1.83

1.17

E 97(14)

EN II

0.9

0.6

0.5

1.50

1.80

1.20

E 97(14)

EN II

1.1

0.9

0.7

1.22

1.57

1.29

E 97(19)

EN II

2.0

1.5

—

1.33

—

—

E 97(19)

EN II

0.9

0.6

0.55

1.50

1.64

1.09

E 97(20)

EN II

0.8

0.5

0.3

1.60

2.67

1.67

E 97(24)

EN I

0.75

0.5

0.4

1.50

1.88

1.25

E 97(25)

EN I

1.2

0.8

0.7

1.50

1.71

1.14

E 97(26b)

EN I

0.9

0.6

0.5

1.50

1.80

1.20

Leguminosae E 97(10)

MN

1

0.8

—

1.25

—

—

E 97(12)

EN II

0.9

0.6

—

1.50

—

—

E 97(27)

EN I

0.7

0.5

0.5

1.40

1.40

1.00

Table 5.7. Measurements of Trifolium spp. and Leguminosae. *L:B = ratio of length (L) to breadth (B). **L:T = ratio of length to thickness (T). ***B:T = ratio of breadth to thickness.

Sample Number

Phase

Length

Breadth

Thickness

L:B*

L:T**

B:T***

5.6

3.1

3.0

1.81

1.87

1.03

Raphanus cf. raphanistrum pod segment E 97(13b)

EN II

Linum cf. usitatissimum E 97(21)

EN I

3.5

1.95

1.1

1.79

3.18

1.77

E 97(21)

EN I

3.7

2.1

1.0

1.76

3.70

2.10

Table 5.8. Measurements of Raphanus cf. raphanistrum and Linum cf. usitatissimum. *L:B = ratio of length (L) to breadth (B). **L:T = ratio of length to thickness (T). ***B:T = ratio of breadth to thickness.

human-plant relationship is impossible to define at present, but some would argue that the ancestor of Raphanus sativus L., the cultivated radish, is R. raphanistrum. Körber-Grohne (1987, 200– 202) proposes that the Mediterranean, among other areas, could have seen the beginnings of its cultivation, and she also believes that the origins of radish cultivation are very old. The radish is rarely found in archaeobotanical assemblages. The root system has not yet been identified, as the chance of roots being charred is far less lower for seeds. The microscopic work by

Hather (1993, 46) has proven that if the R. sativus tuber is charred, it is identifiable if submitted for analysis. The chances for the seeds to be preserved are also minimal for various reasons. One would not expect them to be stored except for use in planting, most likely in rather small quantities. Also, the seeds have a very high oil content, and as with all oleaginous seeds, such as flax, several cruciferous crops, and sesame, they char more than other species and are often so badly damaged that their identification is problematic. This is the case at Knossos, where it seems that R. raphanistrum

THE ECONOMY OF NEOLITHIC KNOSSOS: THE ARCHAEOBOTANICAL DATA

seeds are preserved but their identification is made difficult for the same reasons. Moreover, if the crop was used for making oil, which is unlikely in the Aceramic as the olive was not found at Knossos at that time, it probably would not have been stored for any length of time, having been pressed after reaping. One should, however, note the olive pollen from the Delphinos pollen cores (Bottema and Sarpaki 2003) and the MN–LN (6200 b.p.) cores from Tersana, as well as the evidence for oil imbibed in MN and LN pottery from Gerani Cave in the Rethymnon area (see Beck’s analysis in Tzedakis and Martlew, eds., 1999, 82–83). Finally, the plant’s greens could also have been collected and consumed, but the archaeobotanical visibility of leaves is nil. Here again Beck’s analysis (Tzedakis and Martlew, eds., 1999, 82–83) leaves open the possibility of interpretation. Although the radish is a common weed of cultivation and also one that favors acidic soils, its high representation in certain samples at Knossos demands another explanation. Obviously the residents were collecting this plant, perhaps for any or all of its edible parts. Even the thin, hard root could be used as a condiment, as it smells and tastes like radish. Therefore, we are faced with the question of whether the wild radish was an intensively collected plant that existed in the wild in Crete (thus exemplifying the exploitation and adaptation of native plants in a new environment) or whether it was a cultivated plant brought together in the “Neolithic package.” In the latter case, we would expect to find, if not R. sativus, at least evidence of a proto–R. sativus, since intensive cultivation should be discernible. At present, as the evidence is poor, this question remains unanswered. The other important new species observed is Linum usitatissimum, or flax (Table 5.8). Flax tends to have low archaeobotanical visibility, in part because of the high oil content of the seeds, which explode when in contact with fire and are damaged beyond recognition. It was possible, however, to measure two seeds from the EN I remains and to determine that they belonged not to L. usitatissimum subsp. bienne, wild flax, but to the cultivated form. Wild flax from the Near East does not exceed 3.0 mm in length (van Zeist, de Roller, and

79

Bottema 2000, 141), while cultivated flax measures 3.0 mm or longer (Zohary and Hopf 2000, 130). The Knossos specimens measure 3.5 and 3.7 mm. It has been shown (Helbaek 1959; van Zeist and Bakker-Heeres 1975; van Zeist and de Roller 1991–1992) that flax seeds shrink in length and breadth when charred. Van Zeist mentions a decrease in length of 12%–15%, whereas de Roller increases this to 13%–21% (van Zeist, de Roller, and Bottema 2000, 141). Flax could have been introduced with the other domesticates of the “Neolithic package” found at Knossos, or it could have been indigenously cultivated and/or domesticated, as the distribution of wild flax includes Crete and Greece (Zohary and Hopf 2000, 129, map 12). So far no wild flax has been identified at Knossos, however. All the specimens found are within the range of domesticated flax, and there are no indications of protocultivation, that is, of seeds close to L. bienne in size but in the fringe of L. usitatissimum as well. The presence of cultivated flax at Sabi Abyad II (PPNB), where the specialized production of flax may have been practiced, demonstrates that the species was brought under cultivation in the Near East at an early date (van Zeist, de Roller, and Bottema 2000, 141). As olives did not exist at Knossos, flax seeds (and perhaps also the seeds of R. raphanistrum) might have served as a source of oil, but, alternatively, the stems could have been used to make fiber. If the seeds had been more numerous, it might have been possible to determine the plant’s use, as flax tends to be reaped at different times according to the product needed. It is harvested early, before the seeds are fully ripe, when used for fiber, but the seeds are usually allowed to ripen to their maximum if they are needed for oil. Other archaeological evidence that might allow us to deduce the processing of fibers is also lacking. Spindle whorls only made their appearance in the transitional phase between the EN II and MN periods in Evans’s excavations (Evans 1994, 14), and no such objects are reported from the 1997 dig. It is possible, however, that earlier examples were made out of perishable material.

80

ANAYA SARPAKI

The EN II Samples The EN II samples (Table 5.9) show the continued cultivation of cereals and legumes, with an apparent increase in emphasis on the latter (34 legumes, 68 cerealia) relative to the preceding period. The significance of this trend cannot be determined. The other archaeobotanical material attests to the ongoing cultivation of flax, the collection or cultivation of wild radish, and the collection of several other aromatic plants (Labiatae). The trends that made their appearance in the previous (EN I) period (Table 5.6) seem to have persisted and strengthened in the subsequent era. A much stronger emphasis on arboriculture, with a high presence of almond fragments (Amygdalus communis) and figs (Ficus carica), is evident in the EN II period, however, and the grape (Vitis sp.) appeared as well. Early exploitation of the almond tree in the Aegean is attested at Franchthi Cave in the Argolid, where two whole almonds dated to the Lower Mes­olithic were found (Hansen 1991, 66–68, fig. 30a, pl. 11a; 1992). Although we do not know the exact contexts of these finds, Hansen (1991, 124) mentions that the seed assemblages were associated with several hearths. According to Hansen, the almonds are closer to P. webbii, a species that exists in the Aegean, than the domesticated almond. Their presence as whole (uncrushed) almonds may indicate the use of wild trees as fuel. The use of wild almonds to make oil was definite­ly not possible as the prussic acid content would have been fatal. Moreover, it is reasonable to believe that the extraction of oil was a technological feat achieved after cultivation or domestication. Only fragmented almonds have been recovered at Knossos. Their broken condition indicates that they were most likely a product for consumption rather than for fuel or storage. These fruits were surely not from totally wild trees, which produce intensely bitter seeds. This bitterness is a defensive mechanism of the tree against predators, and it is caused by the presence of the glycoside amygdalin, which becomes the deadly prussic acid (hydrogen cyanide) after crushing, chewing, or any other injury to the seed (Zohary and Hopf 2000, 186). The consumption of several seeds would prove fatal for human beings. Whether or not the almonds used at

EN Knossos were fully domesticated, the inhabitants must have eliminated trees that produced bitter fruits. Unlike some other fruit trees, the almond can be planted from seed (Zohary and Hopf 2000, 185), and, therefore, it could have been imported as seed by the first settlers and cultivated within a few years. It has been noted that 75% or more of the trees grown from the seeds of nonbitter specimens produce sweet fruit. By the stage of cultivation, most, if not all, trees would have been producing sweet almonds. The almond tree also existed in the wild in Greece and Crete, however (Browicz and Zohary 1996, 232). Zohary and Hopf (2000, 186) mention that “A. webbii Sprach is native to the Aegean basin and south Italy,” and they continue (2000, 187), saying, that “the local wild species could have facilitated the development of locally adapted A. Communis cultivars.” Browicz and Zohary (1996, 229) claim that wild almond species (including A. webbii, the local Greek species) constituted the primary gene pool of the cultivated crop and that they saw (Browicz and Zohary 1996, 244) intermediate forms (A. webbii and A. communis L.) at the edges of almond cultivation in Crete. Wild forms of A. communis L. (not feral) abound in the Levant and southern Turkey (Browicz and Zohary 1996, 236, map 245), and prior to the advent of agriculture, this species was confined to the Levant. Perhaps as more data are recovered from Mesolithic and Early Neolithic sites in Greece we will need to challenge the view that cultivation began in the Near East and to include Crete and other parts of Greece among the areas of first cultivation. Cultivation of the almond would have led to domestication, a process that should be visible in the archaeobotanical remains. The changes expected with cultivation include a shift from bitter, poisonous seeds, an increase in the size of the drupes, and the appearance of softer, thinner shells (Browicz and Zohary 1996, 246). Unfortunately, none of the drupes were preserved whole at Knossos and very few have been found elsewhere, so it is not possible to undertake meaningful metric analyses, and the same problem applies to the comparison of the shells.

81

THE ECONOMY OF NEOLITHIC KNOSSOS: THE ARCHAEOBOTANICAL DATA

Sample Number

E 97 (20)

E 97 (19)

E 97 (18)

E 97 (17)

E 97 (16)

E 97 (15)

E 97 (14)

E 97 (13)

E 97 (12)

Total

1997 Level/ J.D. Evans Stratum

29/IV

29/IV

28/IV

24/IV

23/IV

20/IV

17/IV

14/IV

14/IV

 

Liters

21.5

19

17

29

19.5

18

16

48.5

20.5

209

12

15

2

4

4

—

—

3

2

42

(+++)*

(+++)

—

—

(+)

—

—

(+++)

(++)

0

Ficus carica fr. (mineralized)

—

—

—

(+)

—

—

—

—

—

0

Ficus carica (mineralized)

—

1

—

—

—

—

—

—

—

1

Ficus carica charred

63

18

—

6

1

—

—

62

17

167

Vitis vinifera (cf. sylvestris)

—

1

—

—

—

—

—

—

—

1

V. vinifera sp. vinifera

—

1

—

—

—

—

—

—

—

1

cf. Vitis sp. fr.

2

—

—

—

—

—

—

—

—

2

Vitis sp. fr.

3

2

—

—

—

—

—

—

—

5

80

38

2

10

5

0

0

65

19

219

Legume sp. (medium)

—

—

—

—

1

—

—

1

1

3

Legume fr.

2

2

—

—

—

—

—

—

—

4

Lens culinaris

1

1

—

—

—

—

—

—

1

3

Trifolium sp. (small)

—

1

1

1

2

—

2

2

1

10

Trifolium sp. (medium)

1

1

—

—

—

—

—

1

—

3

cf. Trifolium sp.

2

—

—

—

—

—

—

—

—

2

cf. Trigonella sp.

—

—

—

—

—

—

—

1

1

2

Leguminosae (pod)

1

—

—

—

—

—

—

—

—

1

Leguminosae (small)

—

—

—

—

—

—

—

2

—

2

Leguminosae (medium)

—

—

—

—

—

—

—

3

—

3

cf. Onobrychis sp.

—

1

—

—

—

—

—

—

—

1

Total

7

6

1

1

3

0

2

10

4

34

Triticum sp.

12

9

—

1

—

—

—

—

—

22

Triticum sp. (cf. dicoccum)

1

—

—

—

—

—

—

—

—

1

Triticum turgidum L./ aestivum L.

6

4

—

4

1

—

—

3

1

19

Triticum sp. glume base

1

—

—

—

—

—

—

—

—

1

cf. Triticum sp. awns

4

—

—

—

—

—

—

1

—

5

Hordeum sp. (hulled)

—

1

—

—

—

—

—

—

—

1

cf. Hordeum sp.

—

2

1

—

—

—

—

—

—

3

Fruit Amygdalus communis fr. Ficus cf. carica fr. (charred)

Total

Legumes

Cerealia

Table 5.9. Early Neolithic II archaeobotanical (seed) samples. *In addition to counted specimens, (+) = up to 10 fragments that cannot be counted; (++) = 11–50 fragments; (+++) > 50 fragments.

ANAYA SARPAKI

82

Sample Number

E 97 (20)

E 97 (19)

E 97 (18)

E 97 (17)

E 97 (16)

E 97 (15)

E 97 (14)

E 97 (13)

E 97 (12)

Total

1997 Level/ J.D. Evans Stratum

29/IV

29/IV

28/IV

24/IV

23/IV

20/IV

17/IV

14/IV

14/IV

 

Liters

21.5

19

17

29

19.5

18

16

48.5

20.5

209

Hordeum sp. (hulled) fr.

—

1

—

1

—

—

—

—

—

2

Hordeum sp. (cf. naked)

—

—

1

—

—

—

—

—

—

1

Hordeum sp. rachis (damaged)

1

—

—

—

—

—

—

—

—

1

cf. Hordeum sp. awn

1

—

—

—

—

—

—

—

—

1

Cerealia sp. (Triticum/ Hordeum sp.)

—

7

—

1

—

—

—

—

—

8

Cerealia fr.

—

—

—

—

—

—

—

(+)

—

0

cf. Cerealia fr.

—

—

—

—

—

—

—

1

—

1

Cerealia rachis

—

—

—

—

—

—

—

—

1

1

Cerealia, cont.

—

1

—

—

—

—

—

—

—

1

26

25

2

7

1

0

0

5

2

68

Gramineae fr.

—

—

—

1

1

—

—

—

—

2

Gramineae (very small; cf. Arundo sp.)

2

—

—

—

—

1

—

—

2

5

Gramineae (cf. Poa)

—

—

—

—

—

—

—

1

—

1

Gramineae (medium)

—

—

2

—

—

—

—

—

—

2

Gramineae (type 1; cf. Cynodon)

—

1

1

—

1

—

—

2

—

5

Gramineae (type 2; medium)

—

3

—

—

—

—

—

—

—

3

Gramineae (type 3; small)

—

1

—

—

—

—

—

—

—

1

Gramineae rachis

1

—

—

—

—

—

—

—

1

2

Phalaris sp.

—

—

—

—

—

—

—

1

—

1

cf. Bromus sp.

—

—

—

—

—

—

—

1

—

1

cf. Lolium sp. (small)

1

—

—

—

—

—

—

—

—

1

cf. Lolium sp. (medium)

1

—

—

—

—

—

—

—

—

1

Rumex sp.

1

—

—

—

—

—

—

—

—

1

Rumex sp. (R. sanguineus–type)

—

22

—

—

—

—

—

—

—

22

Rumex sp. fr.

—

2

—

—

—

—

—

—

—

2

Cruciferae (cf. Moricandia arvensis)

—

—

—

—

—

—

—

1

—

1

cf. Cruciferae

—

—

—

—

—

—

—

1

—

1

Silene sp.

—

—

—

—

—

—

—

—

1

1

cf. Avena sp. fr.

Total

Weeds

Table 5.9, cont. Early Neolithic II archaeobotanical (seed) samples. *In addition to counted specimens, (+) = up to 10 fragments that cannot be counted; (++) = 11–50 fragments; (+++) > 50 fragments.

83

THE ECONOMY OF NEOLITHIC KNOSSOS: THE ARCHAEOBOTANICAL DATA

Sample Number

E 97 (20)

E 97 (19)

E 97 (18)

E 97 (17)

E 97 (16)

E 97 (15)

E 97 (14)

E 97 (13)

E 97 (12)

Total

1997 Level/ J.D. Evans Stratum

29/IV

29/IV

28/IV

24/IV

23/IV

20/IV

17/IV

14/IV

14/IV

 

Liters

21.5

19

17

29

19.5

18

16

48.5

20.5

209

Malva sp.

—

1

—

—

—

—

—

—

—

1

Sherardia arvensis

—

1

—

—

—

—

—

—

—

1

Galium rivale

—

—

—

2

—

—

—

—

—

2

Buglossoides arvensis

—

—

—

—

—

—

—

5

—

5

Valerianella sp. (cf. microcarpa)

—

—

—

1

—

—

—

—

—

1

6

31

3

4

2

1

0

12

4

63

Weeds, cont.

Total

Condiment-aromatic-industrial(?) cf. Raphanus raphanistrum seeds

1

1

—

—

—

—

—

—

—

2

Raphanus raphanistrum pod fr.

72

5

1

9

—

—

1

1

—

89

Raphanus raphanistrum pod segment

4

—

—

—

—

—

—

2

—

6

Thymelaea (cf. hirsuta) (lvs.)

—

—

—

—

—

—

—

—

—

0

Thymelaea hirsuta seed

1

—

—

2

1

—

—

2

—

6

cf. Linum sp.

—

2

—

—

—

—

—

—

—

2

Satureja thymbra L.

—

1

—

—

—

—

—

—

1

2

Satureja thymbra L. sp.

—

1

—

—

—

—

—

—

—

1

cf. Thymus sp.

—

—

—

—

—

—

—

—

—

0

Labiatae (type B)

—

1

—

—

—

—

—

—

—

1

Total

78

11

1

11

1

0

1

5

1

109

Ignota (identifiable?)

1

—

5

1

2

—

—

—

—

9

Ignota (type A)

—

—

—

—

—

—

—

90

—

90

Ignota drupe fr.

—

—

1

—

—

—

—

—

—

1

Ignota fr. (very damaged)

47 (+++)

11 (+++)

5

33 (++)

48

2

37

87 (+++)

21

291

9

—

—

2

—

—

—

9

9

29

Ignota (shell) fr.

1

1

—

—

—

—

—

—

—

2

stem fr.

—

1

—

—

—

—

—

—

—

1

cf. spores

—

—

—

—

—

2

—

—

3

5

58

13

11

36

50

4

37

186

33

428

Ignota

Ignota (featureless)

Total

Table 5.9, cont. Early Neolithic II archaeobotanical (seed) samples. *In addition to counted specimens, (+) = up to 10 fragments that cannot be counted; (++) = 11–50 fragments; (+++) > 50 fragments.

84

ANAYA SARPAKI

The fig tree (Ficus carica) presumably existed since very early times in Crete and elsewhere in Greece, as it is easily transported in the stomachs of migrating birds. The distribution map of the wild fig drawn by Browicz (1986, map V; see also Zohary and Hopf 2000, 162, map 16) indicates its wide presence in the Mediterranean. The only areas where it has not been found must have been ones that presented unfriendly environmental and climatic conditions. The earliest sites that have produced fig remains so far are located in the Near East, however, and they are of PPNA date, that is, earlier than the Knossos finds (Zohary and Hopf 2000, 163). Although the fig was not found in the archaeobotanical remains from Franchthi, its absence may possibly be explained by the recovery techniques employed; the size of the mesh used for sieving was 1.5 mm, and the fig falls well under this size. The grape (Vitis sp.) appeared in Knossos EN II level 29. In order to distinguish the wild from the cultivated grape, measurements and formulas published by Mangafa and Kotsakis (1996) were used. Their measurements were obtained from two modern varieties, Limnio and Asyrtiko, both from Macedonia, and from wild populations of grapes found in three different localities in western and eastern Macedonia (Mangafa and Kotsakis 1996, 410). On the basis of their formulas, both the archaeobotanical grape pips found at Knossos appear more likely to have been wild than cultivated (Table 5.10). This conclusion is surprising to the present

author as the relative lengths of the stalks suggest that one pip was wild and the other one, which had a longer stalk, was cultivated. Application of the formulas of Mangafa and Kotsakis also categorized as wild the grape pips found at the Late Minoan IB site of Mochlos (Sarpaki and Bending 2004). Although the wild grape was widely distributed in the Mediterranean (Zohary and Hopf 2000, 154; for Crete, see Bottema and Sarpaki 2003), its discovery in the sample of archaeobotanical remains from within the site of Mochlos was totally unexpected. Interestingly, the use of another method of Mangafa and Kotsakis by Jacquat and Martinoli (1999) in the study of pips from Petra in Jordan (150 b.c. to 400 a.d.) also resulted in an identification of wild grapes, even though the morphology of the remains, which displayed a long stalk, was more consistent with that of the cultivated species. The presence of wild grapes in a private Roman dwelling at Petra also seems rather unlikely. Therefore, while the methods of Mangafa and Kotsakis (1996) seem to have a great deal of potential, they probably need to be refined with measurements from more populations of wild grape and present day “old varieties” of cultivars. Thymelaea, a maquis plant, was also represented in the EN II samples and occurred subsequently in the MN period as well. It may have been used for fuel, and the stems, which are very tough, are known to be used for making rope (Polunin and Huxley 1972).

The Middle Neolithic Samples In the Middle Neolithic samples (Table 5.11), the agricultural remains appear to demonstrate a continuation of previous trends, but the grape, naked wheat, and barley are not visible. It is uncertain whether their absence is a reflection of a

change in agricultural priorities and practices or the changing function of the excavated area within the settlement. Whatever the reason, the number and density of archaeobotanical finds waned visibly in comparison to the preceding period.

85

THE ECONOMY OF NEOLITHIC KNOSSOS: THE ARCHAEOBOTANICAL DATA

Vitis sp. Sample Number

Phase

Length

Breadth

Thickness

BF

LF

LS

LCH

BCH

PCH

TS

E 97(19)

EN II

4.7

4.1

3.3

—

—

0.6

1

—

2.2

1.6

E 97(19)

EN II

5.1

3.9

3.3

0.3

2

1

2.3

1.5

1.9

1.1

Formula 2 = 0.2951 + (-12.64PCH/L - 1.6416L + 4.5131PCH + 9.63LS/L) F2 < - 0.2

→

Wild seed

-0.2 < F2 < 0.4

→

90% wild seed

0.4 < F2 < 0.9

→

63% cultivated

F2 > 0.9

→

Cultivated seed

E 97(19): F2 = -2.18

→

Wild seed

E 97(19): F2 = -2.32

→

Wild seed

F3 < 0

→

Wild seed

0 < F3 < 0.5

→

93.5% wild seed

0.5 < F3 < 0.9

→

63.3% cultivated

F3 > 0.9

→

Cultivated seed

E 97(19): F3 = -1.91

→

Wild seed

E 97(19): F3 = -1.88

→

Wild seed

Formula 3 = -7.491 + (1.7715PCH + PCH/L + 9.56LS/L)

BF

TS

LS PCH L LF

LCH

BCH

Ventral side Ventral Side

Dorsal side Dorsal Side

Lateral view Lateral View

Table 5.10. Vitis sp. measurements from EN II levels, analyzed with the formulas of Mangafa and Kotsakis (1996). Sketch of a grape seed showing locations of dimensions: BF = breadth of fossete; LF = length of fossete; LS = length of stalk; L = total length; LCH = length of chalaza; BCH = breadth of chalaza; PCH = placement of chalaza; TS = thickness of stalk.

ANAYA SARPAKI

86

Sample Number

E 97(11)

E 97(10)

E 97(9)

E 97(8)

E 97(7)

E E 97(6b) 97(5b)

E 97(6a)

E 97(5a)

E 97(4)

1997 Level/ J.D. Evans Stratum

12/III

12/III

10/?

10/?

9/?

9/?

8/?

7/?

4/?

4/?

Liters

21

24

29

17

18

22.5

27

22

22

27

229.5 

Amygdalus communis fr.

—

14

4

9

6

23

11

12

16

—

95

cf. Amygdalus sp.

—

—

—

Ficus cf. carica fr. (charred)

—

—

1

—

—

—

—

1

(+)*

11

—

—

—

—

—

—

2

8

21

Ficus carica (mineralized)

4

1

—

1

4

—

—

—

12

2

27

Ficus carica (charred)

3

1

1

2 (++)

1

1

4

2

2

6

23

7

27

5

12

11

25

15

17

32

16

167

2

4

—

1

1

3

—

—

—

— 

12

Total

Fruit

Total

Legumes Legume fr. Lens culinaris

 

1

—

—

—

—

—

—

—

—

2

cf. Lens sp. (cotyl.)

—

—

—

1

1

1

—

—

—

—

3

cf. Pisum sp.

—

—

—

1

—

—

—

—

—

—

1

Trifolium sp. (small)

2

—

1

—

—

—

—

—

—

—

5

Trifolium sp. (cf. arvense)

—

—

—

—

—

2

—

—

—

—

2

cf. Trigonella sp.

—

1

—

—

—

—

—

—

—

—

1

4

6

1

3

2

8

0

0

2

0

26

Triticum sp.

—

—

—

1

—

—

—

1

—

—

—

Cerealia sp. (Triticum/ Hordeum sp.)

—

—

—

—

—

—

—

—

—

1

1

Cerealia fr.

—

2

6

—

—

—

—

—

6

 3

17

0

2

6

1

0

0

0

1

6

4

20

Total

Cerealia

Total

Condiment-aromatic-industrial(?) cf. Capparis sp. mineralized fr.

—

—

—

—

—

—

—

—

1

—

1

Raphanus raphanistrum pod fr.

—

1

—

1

1

1

1

—

2

—

7

Thymelaea hirsuta seed

—

—

—

—

—

1

—

—

—

—

1

Satureja thymbra L.

1

—

—

—

—

—

—

—

1

—

2

1

1

0

1

1

2

1

0

4

0

11

Gramineae fr.

—

—

—

—

—

—

—

—

1

—

4

Gramineae (very small, cf. Arundo sp.)

—

1

—

—

—

—

—

—

—

—

1

Gramineae (type 1, cf. Cynodon)

—

—

—

—

—

—

—

—

—

2

Total

Weeds

2

Table 5.11. Middle Neolithic archaeobotanical (seed) samples. *In addition to counted specimens, (+) = up to 10 fragments that cannot be counted; (++) = 11–50 fragments; (+++) > 50 fragments.

87

THE ECONOMY OF NEOLITHIC KNOSSOS: THE ARCHAEOBOTANICAL DATA

Sample Number

E 97(11)

E 97(10)

E 97(9)

E 97(8)

E 97(7)

E E 97(6b) 97(5b)

E 97(6a)

E 97(5a)

E 97(4)

1997 Level/ J.D. Evans Stratum

12/III

12/III

10/?

10/?

9/?

9/?

8/?

7/?

4/?

4/?

Liters

21

24

29

17

18

22.5

27

22

22

27

229.5 

Gramineae (small, very damaged)

—

—

—

—

—

—

—

—

—

1

1

Euphorbia helioscopia

—

—

—

Malva sp. (mineralized)

—

—

—

2

—

—

 —

2

—

1

—

—

—

—

—

—

—

—

1

0

4

0

0

0

0

2

3

1

1

11

Ignota (identifiable?)

—

1

—

—

—

—

1

—

—

—

2

Ignota (type A)

—

—

—

—

—

2 (++)

5 (+++)

1 (+)

—

—

8

Ignota fr. (very damaged)

6

6

7

8

5

1

10

5

5

1

54

Ignota (featureless)

—

—

—

—

—

—

3

—

—

—

3

stem fr.

—

—

—

—

—

—

—

—

1

— 

1

cf. spores

—

—

—

—

—

—

2

1

—

—

3

6

7

7

8

5

3

21

7

6

1

71

Total

Weeds, cont.

Total

Ignota

Total

Table 5.11, cont. Middle Neolithic archaeobotanical (seed) samples. *In addition to counted specimens, (+) = up to 10 fragments that cannot be counted; (++) = 11–50 fragments; (+++) > 50 fragments.

The Late Neolithic Samples The Late Neolithic archaeobotanical material (Table 5.12) is still poorer than that of the Middle Neolithic, even taking into account the fact that

fewer soil samples were subjected to flotation. Bread wheat, barley, and grape were absent once again.

Discussion The rescue excavations of 1997 have paved the way for a reconsideration of the beginning of the Neolithic in Crete, and the archaeobotanical material provides a tool for reevaluation of ideas about the onset of agriculture on the island. This process, which at first proved to have been somewhat different than developments on mainland Greece, including the Peloponnese, where the Neolithic was established in a landscape of prior Mesolithic

habitation, now seems in the light of the new findings at Gavdos (Kopaka and Mantzanas 2009) and Plakias (Strasser et al. 2010) to possibly have parallel development. What we can detect beyond any doubt is that the first settlement at Knossos was established by people who were full-fledged farmers, highly acquainted with agriculture. The wide range of crops such as almond, fig, lentil, pea, horse bean, possibly

ANAYA SARPAKI

88

Sample Number

E 97(2a)

E 97(2b)

E 97(1)

Total

1997 Level/J.D. Evans Stratum

3/?

3/?

2/?

Liters 

5

6

17

28

Amygdalus communis fr.

1

1

3

5

Ficus cf. carica fr. (charred)

3

—

—

3

Total

4

1

3

8

—

1

— 

1

0

1

0

1

1

—

—

1

1

0

0

1

1

—

—

1

1

0

0

1

Fruit

Legumes Legume fragments

Total

Cerealia Cerealia fr.

Total

Condiment-aromatic-industrial(?)  Raphanus raphanistrum pod fr.

Total

Weeds Gramineae fr.

1

—

—

1

Fumaria sp.

1

—

—

1

Total

2

0

0

2

—

1

—

1

Ignota Ignota (identifiable?) Ignota (type A)

1 (+)*

1

—

2

Ignota fr. (very damaged)

5

2

—

7

cf. spores

—

2

—

2

6

6

0

12

Total

Table 5.12. Late Neolithic archaeobotanical (seed) samples. *In addition to counted specimens, (+) = up to 10 fragments that cannot be counted; (++) = 11–50 fragments; (+++) > 50 fragments.

clovers/medicks, einkorn, emmer, naked wheat, hulled two-row and six-row barley, naked barley, flax, and wild radish(?) is similar to the intensive horticultural regime already described for mainland Greece (Halstead 1996a). If we had a larger number of samples and weed seeds, it might have been possible to identify crop husbandry practices through the study of weed phytosociology and the use of discriminant analysis (Jones 1992). Such analyses may be possible in the future when

the study of the material collected by Helbaek is completed. The presence of T. turgidum/aestivum, naked wheat, is evident from the earliest Aceramic habitation and continues in the EN I and II periods. We cannot say whether its absence in the MN and LN periods represents a change in the cultivated crops or is merely a trend within this area of the site. The important point is that this cultivar is most probably an import from the east (Turkey or the Levant).

89

THE ECONOMY OF NEOLITHIC KNOSSOS: THE ARCHAEOBOTANICAL DATA

Black Sea 11 6 9 10 Aegean Sea

13

cyclades

12

15

•

14

7

16

2 5

Melos Kythera Thera Antikythera

Rhodes

knossos

Karpathos Kassos

17

8

18 te ra ph r Eu ive R

3, 4

s

Mediterranean Sea

1

19

N

0 50

250

Hexaploid Tetraploid

500 km

Figure 5.5. Early sites including those from mainland Greece where Triticum turgidum/aestivum is reported: 1. Tell Abu Hureyra, 2. Tell Halula, 3. Tell Aswad, 4. Tell Ghoraife, 5. Tell Sabi Abyad, 6. Servia, 7. Cafer Höyük, 8. Dhali Agridhi, 9. Otzaki, 10. Sesklo, 11. Sitagroi, 12. Haçilar, 13. Aşikli Höyük, 14. Çatal Höyük, 15. Can Hasan, 16. Cayönü, 17. El Kown, 18. Bouqras, 19. Tell Ramad. Drawing A. Sarpaki.

All the early sites where it has been found are noted in Figure 5.5. The sites in mainland Greece are all later in date than Knossos. Due to the low presence of naked wheat in the north of Greece (Valamoti 2004), one would assume that the immigrants did not come from this region, but rather from the east, and that they most probably used islands of the Dodecanese, Karpathos, Kasos, and perhaps the Cyclades as stepping stones. Extensive studies of other aspects of material culture, such as chipped stone and bone technology, are needed in order to assess the cultural similarities and differences between regions and to elucidate the origins of the the Greek Neolithic agricultural complex (Perlès 2001, 2005). Even if it is claimed that there was at least one very early immigration from the east, we still cannot be sure of how large the “crop package” was and whether some plants, including the almond, fig, wild radish, and others, were cultivated from local stock. The early inhabitants of Knossos were surely masters of the technological skills of agriculture.

The changing frequencies of all categories of archaeobotanical remains from the 1997 Knossos rescue excavation are summarized in Figure 5.6. The main conclusions that emerge from the study of the seed material are as follows: 1. In the Aceramic there was an emphasis on cereal cultivation compared to legumes, and some form of arboriculture was already present, as indicated by the occurrence of figs. Figs might have been part of the natural vegetation and not deliberately tended initially, however. If figs were indigenous, then arboriculture may have started as early the EN I period. 2. There seems to have been a steady increase in the use of fruits from the EN I to the EN II period. Although the number of samples is comparable (9 for EN I and 10 for EN II), their volumes are 123.5 and 209.0 liters, respectively—a difference that could have affected the results and

ANAYA SARPAKI

90 2008

Aceramic

EN I

EN II

MN

LN

Fruit

5

156

434

230

8

Legumes

14

38

34

26

1

Cerealia

38

473

73

20

1

Condiment-aromatic-industrial(?)

—

74

109

12

1

Weeds

—

58

63

11

2

Ignota

—

181

490

187

16

Total Number of Seeds Total Number of Seeds

1200 1200

1000 1000

490 490

181 181

800 800

Ignota Ignota Weeds Weeds Condiment-aromaticCondiment-aromaticindustrial (?) industrial (?) Cerealia Cerealia Legumes Legumes Fruit Fruit

58 58 74 74

600 600

109 109 73 73

473 473

400 400

434 434

200 200 38 38

14 0 5 14 0 5 Aceramic Aceramic

156 156 EN I EN I

38 38

63 63 34 34 11 1112 12

187 187 12 12 20 2026 26 230 230

EN II EN II

MN MN

8

8

2 1 8

2 1 8

16 16

LN LN

Figure 5.6. Summary of the distribution of all categories of archaeobotanical remains at Neolithic Knossos.

the comparability of the samples. Still, their proportions (per liter of soil) should be fairly representative of their overall frequency, and from this perspective the use of fruits is more than double in the EN II period. It is to be expected that the inhabitants would begin with trees, such as almonds and figs, that did not need grafting but could propagate vegetatively. The EN II occurrence of Vitis is problematic, for according to the formulas of Mangafa and Kotsakis (1996) the specimens are both grouped as wild, whereas the morphology of one seems to be that of the cultivated species, as discussed above. The secure presence of domesticated grape would imply the beginning of grafting and/or intentional vegetative propagation. 3. There was a great emphasis on cereals, especially naked wheat, in the Aceramic and

EN I periods, but there seems to have been a shift toward greater legume cultivation in the EN II period. Production of cereals and legumes seems to have stabilized, which might perhaps be an indication of agricultural intensification and/or a shortage of land (Sarpaki 1992). This trend continued up to the LN period. Even at an early date, Knossos must have been a huge settlement by the standards of the time. 4. The importation of naked wheat was a conscious choice, as it is much easier to thresh than glume wheats (einkorn and emmer). This would have been an important consideration for a people on the move. 5. It might be speculated that the appearance of flax in the EN I period is indicative of a new wave of immigration. Once the initial

THE ECONOMY OF NEOLITHIC KNOSSOS: THE ARCHAEOBOTANICAL DATA

group of settlers arrived, all kinds of contacts and revisits were possible. 6. The wild radish appeared in EN I and remained part of the agricultural scene until the Late Neolithic. The foraging or cultivation of this plant might be a local development. At present it is impossible to tell whether it was cultivated as a crop since no cache or storage of these seeds has been detected at Knossos or elsewhere. 7. The presence of aromatic plants indicates that garrigue vegetation was available and that the samples retrieved were not strictly from storage contexts but were of mixed provenance. Weeds of cultivation were also present.

91

8. There is a total absence of olives at Knos­ sos from the Aceramic to the Late Neo­ lithic, as also observed by Badal and Ntinou (this vol., Ch. 6). While the archaeobotanical results may have been affected, to some extent, by the sampling strategy for the collection of soil samples, one would like to believe that they are representative both of the excavated area and of the agricultural tendencies of the site overall. The conclusions presented here will be evaluated further once the material from the rest of the Neolithic excavations is studied in more detail. It will then be possible to assess whether the archaeobotanical finds from the 1997 excavation speak only for a part of the site or for more general developments at Knossos.

Dedication and Acknowledgments This contribution is dedicated to the memory of Hans Helbaek. I thank Alexandra Karetsou and Eleni Banou of the 23rd Ephoreia of Herakleion and Nikos Efstratiou of the University of Thessaloniki for entrusting me with the study of the seed remains from the 1997 excavation. I am deeply indebted to Maria Balanou, who did the water flotation, as well as to Angeliki Kossyva, Niki Spanou, and Lena Mandalara, who sorted the

flots with a stereoscope microscope. I would also like to thank Sue Colledge and Tania Valamoti for information on naked wheats. I am especially grateful to the British School at Athens and J.D. Evans for granting me permission to study and photograph the botanical material from Evans’s excavations and to the latter for entrusting me with Helbaek’s unpublished report.

References Arnold, E.R., and H.J. Greenfield. 2006. The Origins of Transhumant Pastoralism in Temperate South Eastern Europe: A Zooarchaeological Perspective from the Cental Balkans (BAR-IS 1538), Oxford.

Bottema, S., and A. Sarpaki. 2003. “Environmental Change in Crete: A 9000-Year Record of Holocene Vegetation History and the Effect of the Santorini Eruption,” The Holocene 13, pp. 733–749.

Blumler, M.A. 1996. “Ecology, Evolutionary Theory and Agricultural Origins,” in Harris, ed., 1996, pp. 25–50.

Broodbank, C. 1999. “Colonization and Configuration in the Insular Neolithic of the Aegean,” in Neolithic Society in Greece (Sheffield Studies in Aegean Archaeology 2), P. Halstead, ed., Sheffield, pp. 15–41.

Bottema, S. 1992. “Cereal-Type Pollen in the Near East as Indicators of Wild or Domestic Crops,” in Préhistoire de l’agriculture: Nouvelles approches ex­ périmentales et ethnographiques (Monographies du CRA 6), P.C. Anderson-Gerfaud, ed., Paris, pp. 95–106.

———. 2000. An Island Archaeology of the Early Cyclades, Cambridge.

92

ANAYA SARPAKI

Browicz, K. 1986. Chorology of Trees and Shrubs in South-West Asia and Adjacent Regions 5, Kornik.

Harris, D.R., ed. 1996. The Origins and Spread of Agriculture and Pastoralism in Eurasia, London.

Browicz, K., and D. Zohary. 1996. “The Genus Amygdalus L. (Rosaceae): Species Relationships, Distribution and Evolution under Domestication,” Genetic Resources and Crop Evolution 43, pp. 229–247.

Helbaek, H. 1959. “Notes on the Evolution and History of Linum,” Kuml 1959, pp. 103–120.

Cherry, J.F. 1985. “Islands out of the Stream: Isolation and Interaction in Early East Mediterranean Insular Prehistory,” in Prehistoric Production and Exchange: The Aegean and Eastern Mediterranean (UCLAMon 25), A.B. Knapp and T. Stech, eds., Los Angeles, pp. 12–29.

———. 1970. “The Plant Husbandry of Haçilar,” in Excavations at Haçilar I, J. Mellaart, ed., Edinburgh, pp. 189–244.

———. 1990. “The First Colonization of the Mediter­ ranean Islands: A Review of Recent Re­search,” JMA 3, pp. 145–221. de Moulins, D. 1996. “Sieving Experiment: The Con­ trolled Recovery of Charred Remains from Modern and Archaeological Samples,” in Early Farm­­ing in the Old World: Recent Advances in Archaeobotanical Research (Vegetation History and Archaeobotany Special Vol.), K.-E. Behre and K. Oeggl, eds., Heidelberg, pp. 153–156. Evans, J.D. 1964. “Excavations in the Neolithic Set­ tlement of Knossos, 1957–60: Part I,” BSA 59, pp. 132–240.

———. 1968. “Knossos Wheat.” Unpublished manuscript, second draft.

Helmer, D., V. Roitel, M. Saña, and G. Wilcox. 1998. “Interprétations environnementales des données archéozoologiques et archéobotaniques en Syrie du nord de 16,000 bp à 7000 bp, et les débuts de la domestication des plantes et des animaux,” in Espace naturel, espace habité en Syrie du nord (10e–2e millénaires av. J.-C.) (Travaux de la Maison de l’Orient 28), M. Fortin and O. Aurenche, eds., Lyon, pp. 9–33. Hopf, M. 1955. “Formveränderunggen von Ge­ trei­ dekörnern beim Verkohlen,” Berichte der Deutschen Botanischer Gesellschaft 68, pp. 191–193. Jacomet, S. 1987. Prehistoric Cereal Finds: A Guide to the Identification of Prehistoric Barley and Wheat Finds, J. Greig, trans., Basel.

———. 1971. “Neolithic Knossos: The Growth of a Settlement,” PPS 37, pp. 95–117.

Jacquat, C., and D. Martinoli. 1999. “Vitis vinifera L.: Wild or Cultivated? Study of the Grape Pips Found at Petra, Jordan; 150 b.c.–a.d. 40,” Vegetation History and Archaeobotany 8, pp. 25–30.

———. 1994. “The Early Millennia: Continuity and Change in a Farming Settlement,” in Knossos: A Labyrinth of History. Papers in Honour of S. Hood, D. Evely, H. Hughes-Brock, and N. Momigliano, eds., Oxford, pp. 1–20.

Jarman, M.R., C.N. Bailey, and H.N. Jarman, eds. 1982. Early European Agriculture: Its Foundations and Development, Cambridge.

Green, F.J. 1975. Large Seeded Legumes of the Old World, B.A. diss., University of Sheffield. Halstead, P. 1996a. “The Development of Agriculture and Pastoralism in Greece: When, How, Who and What?” in Harris, ed., 1996, pp. 296–309. ———. 1996b. “Pastoralism or Household Herding? Problems of Scale and Specialization in Early Greek Animal Husbandry,” WorldArch 28, pp. 20–42. Hansen, J.M. 1991. The Palaeoethnobotany of Franchthi Cave (Franchthi 7), Bloomington. ———. 1992. “Franchthi Cave and the Beginnings of Agriculture in Greece and the Aegean,” in Préhistoire de l’agriculture: Nouvelles approches expérimentales et ethnographiques (Monographies du CRA 6), P.C. Anderson-Gerfaud, ed., Paris, pp. 231–247. Hather, J.G. 1993. An Archaeological Guide to Root and Tuber Identification 1: Europe and South West Asia (Oxbow Monograph 28), Oxford.

Jarman, M.R., and H.N. Jarman. 1968. “The Fauna and Economy of Early Neolithic Knossos,” in “Knossos Neolithic, Part II,” P. Warren, M.R. Jarman, H.N. Jarman, N.J. Shackleton, and J.D. Evans, BSA 63, pp. 241–264. Jones, G. 1992. “Weed Phytosociology and Crop Husbandry: Identifying a Contrast Between Ancient and Modern Practice,” Review of Palaeobotany and Palynology 73, pp. 133–143. Kislev, M. 1979–1980. “Triticum parvicoccum sp. Nov.: The Oldest Naked Wheat,” Israel Journal of Botany 28, pp. 95–107. ———. 1984. “Botanical Evidence for Ancient Naked Wheats in the Near East,” in Plants and Ancient Man: Studies in Palaeoethnobotany, W. van Zeist and W.A. Casparie, eds., Rotterdam, pp. 141–152. ———. 2009. “Reconstructing the Ear Morphology of Ancient Small-Grain Wheat (Triticum turgidum ssp. parvicoccum),” in From Foragers to Farmers. Papers

THE ECONOMY OF NEOLITHIC KNOSSOS: THE ARCHAEOBOTANICAL DATA

93

in Honor of Gordon Hillman, A. Fairburn and E. Weiss, eds., Oxford, pp. 235–238.

Rackham, O., and J. Moody. 1996. The Making of the Cretan Landscape, Manchester.

Kopaka, K., and Ch. Mantzanas. 2009. “Palaeolithic Industries from the Island of Gavdos, near Neighbour to Crete in Greece,” Antiquity 83 (321), http://antiquity.ac.uk/projgall/Kopaka321/

Renfrew, J. 1979. “The First Farmers in South East Europe,” in Festschrift Maria Hopf zum 65. Geburtstag am 14. September 1979 (ArchaeoPhysika 8), U. Körber-Grohne, ed., Cologne, pp. 243–265.

Körber-Grohne, U. 1987. Nutzpflanzen in Deutschland, Stuttgart. Ladizinsky, G. 1989. “Origin and Domestication of the Southwest Asian Grain Legumes,” in Foraging and Farming: The Evolution of Plant Exploitation, D.R. Harris and G.C. Hillman, eds., London, pp. 374–389. Lambeck, K. 1996. “Sea-Level Change and Shore-Line Evolution in Aegean Greece since Upper Palaeolithic Time,” Antiquity 70, pp. 588–611. Maier, U. 1996. “Morphological Studies of Freethreshing Wheat Ears from the Neolithic Site in Southwest Germany, and the History of the Naked Wheats,” in Early Farming in the Old World: Recent Advances in Archaeobotanical Research (Vegetation History and Archaeobotany Special Vol.), K-E. Behre and K. Oeggl, eds., Heidelberg, pp. 39–55. Mangafa, M., and K. Kotsakis. 1996. “A New Method for the Identification of Wild and Cultivated Charred Grape Seeds,” JAS 23, pp. 400–418. Miller, N. 1984. “Intentional Burning of Dung as Fuel: A Mechanism for the Incorporation of Charred Seeds into the Archaeological Record,” Journal of Ethnobiology 4, pp. 15–77. Moody, J., O. Rackham, and G. Rapp. 1996. “Environmental Archaeology of Prehistoric NW Crete,” JFA 23, pp. 273–297. Peltenburg, E., S. Colledge, P. Croft, A. Jackson, C. McCartney, and M.A. Murray. 2000. “Agro-Pastoralist Colonization of Cyprus in the 10th Millennium bp: Initial Assessments,” Antiquity 74, pp. 844–853. Perlès, C. 1979. “Des navigateurs méditerranéens il y a 10,000 ans,” La Recherche 96, pp. 82–83. ———. 2001. The Early Neolithic in Greece: The First Farming Communities in Europe, Cambridge. ———. 2005. “From the Near East to Greece: Let’s Reverse the Focus. Cultural Elements That Didn’t Transfer,” ­ in How Did Farming Reach Europe? Anatolian-European Relations from the Second Half of the 7th through the First Half of the 6th Millennium cal bc. Proceedings of the International Workshop, Istanbul, 20–22 May 2004 (Byzas 2), G. Lichter, ed., Istanbul, pp. 275–290. Polunin, O., and A. Huxley. 1972. Flowers of the Mediterranean, London.

Roberts, N. 1979. “The Location and Environment of Knossos,” BSA 74, pp. 231–240. Sakellarakis, Y. 1973. “Neolithic Crete,” in Neolithic Greece, D.R. Theocharis, Athens, pp. 131–146. Sarpaki, A. 1992. “The Palaeoethnobotanical Approach: The Mediterranean Triad or Is It a Quartet?” in Agriculture in Ancient Greece (SkrAth 4°, 42), B. Wells, ed., Stockholm, pp. 61–76. ———. 2009. “Knossos, Crete: Invaders, ‘Sea-goers,’ or Previously ‘Invisible,’ the Neolithic Economy Appears Fully-Fledged in 9000 b.p.,” in From Foragers to Farmers. Papers in Honour of Gordon Hillman, A. Fairbairn and E. Weiss, eds., Oxford, pp. 220–234. Sarpaki, A., and J. Bending. 2004. “Archaeobotanical Assemblages,” in Mochlos IC: Period III. Neopalatial Settlement on the Coast: The Artisans’ Quarter and the Farmhouse at Chalinomouri. The Small Finds (Prehistory Monographs 9), J.S. Soles, C. Davaras, J. Bending, T. Carter, D. Kondopoulou, D. Mylona, M. Ntinou, A.M. Nicgorski, D.S. Reese, A. Sarpaki, W.H. Schoch, M.E. Soles, V. Spatharas, Z.A. StosGale, D.H. Tarling, and C. Witmore, Philadelphia, pp. 126–131. Sherratt, A. 1980. “Water, Soil, and Seasonality in Early Cereal Cultivation,” WorldArch 11, pp. 313–330. ———. 1996. “Plate Tectonics and Imaginary Prehistories: Structure and Contingency in Agricultural Origins,” in Harris, ed., 1996, pp. 130–140. Stavropoulos, N., A. Zamanis, P. Efthimiadis, S. Samaras, and A. Matthaiou. 1992. “Phenotypic Differences in Greek Populations of Wild Wheat Triticum monococcum L. subsp. boeoticum (Boiss) and of Cultivated Diploid Wheat T. monococcum subsp. Monococcum” (Greek with English summary), in Proceedings of the Panhellenic Conference on Agricultural Research, vol. 2, Thessaloniki. Strasser, T. 1996. “Soils and Settlements on Neolithic Crete,” in Pleistocene and Holocene Fauna of Crete and Its First Settlers (Monographs in World Archaeology 28), D.S. Reese, ed., Madison, pp. 317–336. Strasser, T., N. Thompson, E. Panagopoulou, P. Karkanas, C. Runnels, F. McCoy, P. Murray, and K. Wegmann. 2010. “Stone Age Seafaring in the Mediterranean:

94

ANAYA SARPAKI

Evidence from the Plakias Region for the Paleolithic and Mesolithic Habitation of Crete,” Hesperia 79, pp. 145–190. Tzedakis, Y., and H. Martlew, eds. 1999. Minoans and Mycenaeans: Flavours of Their Time, Athens.

van Zeist, W., G.J. de Roller, and S. Bottema. 2000. “The Plant Remains” in Tell Sabi Abyad II: The Prepottery Neolithic B Settlement (Uitgaven van het Nederlands Historisch-Archaeologisch Instituut te Istanbul 90), M. Verhoeven and P.M.M.G. Akkermans, eds., Istanbul, pp. 137–147.

Valamoti, S.M. 2004. Plants and People in Late Neo­lithic and Early Bronze Age Northern Greece: An Archaeo­ botanical Investigation (BAR-IS 1258), Oxford.

van Zeist, W., and W. Waterbolk-van Rooijen. 1985. “The Palaeobotany of Tell Bouqras, Eastern Syria,” Paléorient 11, pp. 131–147.

van Andel, T.H., and C.N. Runnels. 1995. “The Earliest Farmers in Europe,” Antiquity 69, pp. 481–500.

Warren, P., M.R. Jarman, H.N. Jarman, N.J. Shackleton, and J.D. Evans. 1968. “Knossos Neolithic, Part II,” BSA 63, pp. 239–276.

van Zeist, W. 1972. “Prehistoric and Early Historic Food Plants in the Netherlands,” Palaeohistoria 14, pp. 41–173. van Zeist, W., and J.A.H. Bakker-Heeres. 1975. “Evidence for Linseed Cultivation before 6000 b.c.,” JAS 2, pp. 215–219. van Zeist, W., and H. Buitenhuis. 1983. “A Pa­ laeo­ botanical Study of Neolithic Erbaba, Turkey,” Anatolica 10, pp. 47–89. van Zeist, W., and G.J. de Roller. 1991–1992. “The Plant Husbandry of Aceramic Cayönü, SE Turkey,” Palaeohistoria 33/34, pp. 65–96.

Wilson, D.G. 1984. “The Carbonization of Weed Seeds and Their Representation in Macrofossil Assemblages,” in Plants and Ancient Man: Studies in Palaeoethnobotany, W. van Zeist and W.A. Casparie, eds., Rotterdam, pp. 201–206. Zohary, D., and M. Hopf. 2000. Domestication of Plants in the Old World: The Origin and Spread of Cultivated Plants on West Asia, Europe, and the Nile Valley, 3rd ed., Oxford. Zois, A. 1973. Κρήτη—Ἠ Ἐποχή τοῦ Λὶθου, Athens.

6

Wood Charcoal Analysis: The Local Vegetation Ernestina Badal and Maria Ntinou

The excavation of a small trench in the Central Court area at Knossos in 1997 explored the Neo­ lithic sequence of the site. Previous excavations of the Neolithic deposits (Evans 1964) had already established the prominent role of Knossos in the early spread of agriculture in the Mediterranean. New excavations at the site, although much more limited in scale, presented us with the opportunity to add complementary information on various aspects of the process of neolithization. One of the aims of the 1997 excavation was to undertake paleoenvironmental research at the

*Abbreviations used in this chapter are: AN Aceramic Neolithic C Celsius, Centigrade cal. calibrated or calendar years cf. compare Ch(s). Chapter(s) EM Early Minoan EN Early Neolithic km kilometers

site.* Charcoal analysis, or anthracology, has proven to be a valuable tool for paleoecological and paleoethnographic reconnaissance at many arch­ aeological sites (Vernet, ed., 1992; Chabal et al. 1999; Thiébault, ed., 2002; Dufraisse, ed., 2006; Fiorentino and Magri, eds., 2008). The study of the charcoal from the 1997 Knossos excavation has helped us to reconstruct both the plant formations that prevailed in the Knossos area during the Neolithic and the ways in which they were used and modified by the first Neolithic settlers. The results of our analyses are presented in this chapter.

LN m m asl mm MN Mt. no. sp.

Late Neolithic meters meters above sea level millimeters Middle Neolithic Mount number species

ERNESTINA BADAL AND MARIA NTINOU

96

Physical Background Location and Geology The site of Knossos in Crete lies in the valley of the Kairatos River, which runs along the east side of the site. The Vlychia, a tributary of the Kairatos, borders the site to the south (Fig. 6.1). The river flows south from Mt. Juktas to the Aegean in the Nea Alikarnassos area, approximately 5 km to the north of Knossos. It carries a low volume of water and occasionally dries up in the summer. The Kairatos valley is narrow and gorge-like in its upper part, south of Knossos, and becomes slightly wider downstream. The terrain is higher to the

1400 mm

south. The area around Knossos is surrounded by hills lower than 300 m, including a limestone ridge to the east and the “Acropolis” hill to the west (Roberts 1979). The geology of the area is characterized by Cretaceous limestone overlain in places by kou­ skouras, a soft white marl of Pliocene age. Related gypsum deposits form the low hills of Gypsades south of Knossos. Debris fans of Plio-Pleistocene age surround the hills and the higher elevations to the south of the site (Roberts 1979).

a

b

Cretan Sea

N

atos Kair r

Rive

Herakleion

10

0

10

0

c

400 300

0

200

m asl

10

0

10

400

knossos

B

300

3

B´ 20

200

Kairatos River

100 m asl

20

200

2

0

100 B

Mt. Juktas

800

0

knossos

W–E

B´

d

30

0

800

5

700

700

600

600

30

0

40

40

0

0

0

60

0

50

500

500

0

0

400 KNOSSOS

200

1

300

Kairatos River

200

100

100 A

A´ SW–NE

700

400

300

30

Mt. Juktas

40

4

N

Topographic curves every 50 m River City 0 0

1

2 1

3

4 2

5

6 km

3 miles

Figure 6.1. Climate and topography of Knossos: (a) mean annual precipitation in Crete (after Rackham and Moody 1996); ( b) topographic map of the area around Knossos; (c) west–east topographic section; (d) southwest–northeast topographic section. The numbers on topographic sections correspond to plant inventories (see Table 6.1).

WOOD CHARCOAL ANALYSIS: THE LOCAL VEGETATION

97

Climate The climate in the Knossos area is typically Mediterranean, and the bioclimatic conditions are of the thermomediterranean type. The area is warm throughout the year with the average annual temperature estimated between 17°C and 19°C. January is the coldest month, with an average temperature of 11°C. Lower temperatures are experienced inland and at higher altitudes. Frosts, though rare, can occur from December to February. Maximum summer temperatures often

exceed 35°C. Winter is the wettest season, and summers are dry and hot. Mean annual precipitation at the station of Herakleion has been estimated to be 476.5 mm by Roberts (1979). According to a map of Cretan precipitation in 1970–1982 drawn by Rackham and Moody (1996), the area of Knossos receives approximately 500 mm annually (Fig. 6.1:a). As in other Mediterranean areas, however, annual totals may vary considerably in consecutive years.

Flora and Vegetation The vascular flora of Crete comprises 1,706 species. Many of these are widespread Mediterranean and Euro-Siberian plants, but some special elements associated with the historical geography of the area also exist. Certain habitats, particularly calcareous cliffs and naturally treeless mountain summits, are very rich in relict endemic species (Turland, Chilton, and Press 1993). The island’s natural vegetation is typically Mediterranean. Zohary and Orshan (1966, cited in Roberts 1979) offered the following scheme of the altitudinal zoning of climax communities: a. Evergreen maquis, which includes wild olive, pistachio, carob, juniper, and oak (Quercus coccifera), prevailing in all lowland areas up to 300 m b. Evergreen oak forest extending between 300 m and 800 m of altitude, with pine forests partly substituting on rendzina soils c. Cupresso-Aceretum orientale covering the zone above 800–1000 m, which corresponds to the oromediterranean zone in Crete The Cretan landscape presents a variety of vegetation types, and occasionally altitudinal zoning is difficult to distinguish, in part due to intense use of the vegetation over the millennia, which has caused different plant formations such as maquis, phrygana, and pastureland to intermingle and fade

into one another. Wooded areas exist in the great mountain ranges as well as at lower elevations. During the last decades and especially since the number of livestock has decreased and agriculture has retreated to the best land, the Cretan vegetation is recovering fast. Agriculture is practiced extensively in the Kairatos valley around Knossos (Figs. 6.2, 6.3). The major crops are olive trees and vines, and there are also irrigated areas where vegetable orchards and orange plantations are located. Natural vegetation is restricted to small patches of the landscape. Thermomediterranean communities can be found from sea level to 600 m asl. They are very degraded by intense grazing. The elevations to the east of the site are covered with phrygana formations. Spiny shrubs such as Sarcopoterium spinosum and species of the Leguminosae, Labiatae, Cistaceae, and Compositae families are common components of these communities (Table 6.1; Figs. 6.1, 6.4). At higher elevations on the sides of Mt. Juktas (highest elevation 811 m) phrygana communities also dominate, but evergreen oak woodland in a degraded state is found in places. Conifers such as Cupressus sempervirens and Pinus brutia are rare. Scattered deciduous oaks grow on the western side of Mt. Juktas. Rich plant formations, which combine evergreen and deciduous species, are found on deeper soils and along watercourses (Figs. 6.1:d, 6.5). More detailed lists (inventories) of the plants growing in different parts of the study area are presented in Table 6.1.

98

ERNESTINA BADAL AND MARIA NTINOU

Figure 6.2. View of the Knossos valley from Mt. Juktas showing present-day vegetation. Photo E. Badal.

Figure 6.3. Panoramic view of the site of Knossos showing present-day vegetation. Photo E. Badal.

Figure 6.4. Present-day phrygana vegetation on the hills in the study area. Photo E. Badal.

Figure 6.5. Present-day vegetation on deep soils in the study area. Photo E. Badal.

99

WOOD CHARCOAL ANALYSIS: THE LOCAL VEGETATION Inventory No. 1 River Kairatos 50–100 m asl

Inventory No. 2 Eastern Hill 100–200 m asl

Inventory No. 3 Eastern Hill 200 m asl and up

Inventory No. 4 Juktas 400–425 m asl

Inventory No. 5 Juktas 700 m asl

Platanus orientalis

Pyrus sp.

Phlomis fruticosa

Quercus coccifera

Quercus coccifera

Phragmites australis

Salvia fruticosa

Cistus creticus

Cistus salvifolius

Quercus calliprinos

Arundo donax

Capparis cf. ovata

Calycotome villosa

Osyris alba

Cupressus sempervirens (horizontalis)

Vinca sp.

Sambucus ebulus

Asparagus sp.

Salvia fruticosa

Calycotome villosa

Parietaria sp.

cf. Asphodelus

Salvia fruticosa

Phlomis fruticosa

Cistus cf. crispus

Ecbalium elaterium

Thymus capitatus

Crocus sp.

Genista acanthoclada

Phlomis fruticosa

Corylus avellana

Satureja sp.

Lavatera cf. arborea

Calycotome villosa

Euphorbia sp.

Morus sp.

Lavatera cf. arborea

Thymus capitatus

Thymus capitatus

Sarcopoterium spinosum

Ficus carica

Calycotome villosa

Ebenus cretica

Hypericum empetrifolium

Phlomis lanata

Hedera helix

Ebenus cretica

Olea europaea

Pistacia lentiscus

Osyris alba

—

Asparagus sp.

Prunus amygdalus

Sarcopoterium spinosum

Spartium junceum

—

Pistacia terebinthus

cf. Asphodelus

Ebenus cretica

Pistacia terebinthus

—

Helichrysum sp.

Osyris alba

Erica multiflora

Hypericum empetrifolium

—

Crocus sp.

Ficus carica

Asphodelus sp.

Olea europaea

—

Phlomis fruticosa

Hypericum empetrifolium

Euphorbia sp.

Ranunculus sp.

—

Euphorbia sp.

Euphorbia sp.

Oreganum microphyllum

Asparagus sp.

—

Sarcopoterium spinosum

Capparis cf. ovata

Lavatera cf. arborea

Rhamnus oleoides

—

Rhamnus alaternus

Thymelea hirsuta

Asparagus sp.

Pinus brutia

—

Genista acanthoclada

Salicornia sp.

Phlomis lanata

—

—

—

Sarcopoterium spinosum

Ruscus aculeatus

—

—

—

Genista acanthoclada

Rubia peregrina

—

—

—

—

Fumana sp.

—

—

—

—

Oreganum vulgare

—

—

—

—

Quercus calliprinos

—

—

—

Styrax officianalis

—

—

—

Rhamnus oleoides

—

—

—

—

Cistus creticus

—

—

—

Pyrus amygdaliformis

—

—

—

—

Crataegus sp.

—

—

—

—

Spartium junceum

—

—

—

—

Ficus carica

—

—

—

—

Thymelea hirsuta

—

—

—

—

Ephedra cf. fragilis

—

—

—

—

Pistacia terebinthus

—

—

—

—

Capparis spinosa

—

Table 6.1. Inventories of plants growing in different parts of the study area. Plants are listed in the order of occurence as observed in each area.

100

ERNESTINA BADAL AND MARIA NTINOU

Wood Charcoal Analysis: Methodology and Fieldwork Daily human activities are reflected in the cultural remains at archaeological sites. Wood charcoal is a category of such remains. It originates either from firewood or from timber and plants used for various purposes and burned deliberately or accidentally at some point during a site’s history. Although wood charcoal has been employed mainly for 14C dating, it can also be used for the identification of plant taxa, thus providing ethnobotanical and paleoenvironmental information. Wood charcoal analysis requires detailed sampling of all the excavated deposits and precise information concerning the state of deposition and origin of the remains. The woody plant species can be identified by using a metallurgical microscope. The results provide information on the types of vegetation that existed in an area in the past, the climate regime under which they grew, and, most importantly, how they were used by the human groups visiting or settling in an area (Chabal 1988; 1997; Badal 1990; Vernet, ed., 1992; Chabal et al. 1999; Thiébault, ed., 2002; Dufraisse, ed., 2006; Fiorentino and Magri, eds., 2008). Excavation in the Central Court of the palace of Knossos was carried out in 1997 jointly by the 23rd Ephorate of Classical and Prehistoric Antiquities of the Ministry of Culture and the Department of History and Archaeology of the University of Thessaloniki. A 3 x 2 m trench was opened in the eastern part of the Central Court, and 39 archaeological levels were excavated. The study of the material culture and pottery typology,

together with radiocarbon dating, have established the sequence of archaeological levels and phases of the Neolithic as follows: Aceramic Neolithic: levels 39–38 Early Neolithic I (EN I): levels 37–30 Early Neolithic II (EN II): levels 29–14 Middle Neolithic (MN): levels 13–4 Late Neolithic (LN): levels 3–1 Because the paleoenvironmental study of the site was one of the main aims of the 1997 excavation campaign, sediment samples were taken from every archaeological level and subsequently drysieved and floated. Wood charcoal was extracted from the samples, and this material was analyzed to obtain information on the vegetation surrounding the site and its use by the Neolithic settlers of the area. In general, the wood charcoal was dispersed but not abundant in the sediment. Its relatively low frequency may be explained by the characteristics of the physical site, which is an open-air settlement exposed to wind, rain, and other erosive agents that may have caused the displacement and loss of remains not concentrated in closed features. Some levels did not provide any wood charcoal at all. Material was recovered from 25 archaeological levels altogether, and these fortunately correspond to all of the Neolithic phases listed above (Table 6.2). The wood charcoal assemblage from each archaeological level comprises the material recovered from both dry-sieving and flotation.

Results The Plant List A total of 29 taxa have been identified. These include evergreen broad-leaved species, deciduous species, and conifers (Table 6.3). The identified

taxa are Acer sp. (maple), Anacardiaceae, Arbutus sp. (strawberry tree), Cistus sp. (rockrose), Conifer, Cupressus sempervirens (cypress; Fig. 6.6:a, b),

WOOD CHARCOAL ANALYSIS: THE LOCAL VEGETATION

Daphne sp. (garland flower), Erica sp. (heather), Fraxinus sp. (ash), Ficus carica (fig tree), Juniperus sp. (juniper), cf. Laurus nobilis (laurel), Leguminosae (the pea family), Monocotyledons, cf. Oreganum (marjoram), Phillyrea/Rhamnus (mock privet/buckthorn), Pinus brutia (the Cretan pine; Fig. 6.6:c, d), Pinus sp., Pistacia lentiscus (lentisk; Fig. 6.6:e), Pistacia sp., Pistacia terebinthus (turpentine), Platanus orientalis (oriental plane; Fig. 6.6:f, g), Prunus amygdalus (almond tree; Fig. 6.6:h, i), Prunus sp., Quercus sp. evergreen type (prickly and/or Holm oak), Quercus sp., Quercus sp. deciduous type (deciduous oak), Rosaceae/Maloideae (the apple tree family), and Tamarix sp. (tamarisk). Five fragments were not identified because their small size did not permit observation of all three anatomical sections. Overall, both the plant list and individual assemblages are rich in taxa. The taxon Quercus sp. evergreen type includes the species Quercus coccifera (prickly oak) and Quercus ilex (Holm oak), which are undifferentiated on the basis of their anatomy. Prickly oak is the most common tree in Crete and can grow at sea level as well as on the highest Cretan mountains (up to 1,780 m in the White Mountains). It grows on any type of substrate, although preferably on limestone (Turland, Chilton, and Press 1993, 5; Rackham and Moody 1996, 64). Holm oak is rather uncommon in Crete and is usually restricted to crevices of calcareous cliffs where it is protected from browsing, wood cutting, and burning. It usually forms thickets with the strawberry tree (Arbutus unedo) and turpentine (Pistacia terebinthus) on noncalcareous soils. It also occurs in open woodland on calcareous soils and rocky ground together with Acer sempervirens, Crataegus, Phillyrea, Pistacia terebinthus, Prunus spinosa, and Quercus pubescens (Turland, Chilton, and Press 1993, 5). In the assemblages from Neolithic Knossos both species, namely prickly oak and Holm oak, might be represented as components of various plant formations since all other participants of present-day communities are also included. The taxon Quercus sp. deciduous type may include different deciduous oak species that cannot be differentiated by their xylem anatomy. For Crete the species Q. pubescens and Q. brachyphylla are mentioned as native (Turland, Chilton, and Press 1993, 6; Rackham and Moody 1996, 65). As

101

discussed by Bottema and Sarpaki (2003), there are different opinions concerning the taxonomic status of deciduous oak in Crete. Some experts accept the presence of only one of the two species mentioned above or consider them synonymous; some even describe Q. brachyphylla as a subspecies of Q. pubescens. The taxon Phillyrea/Rhamnus includes two different genera, Phillyrea and Rhamnus, belonging to different families. Due to their anatomical similarity they cannot be differentiated. Both are quite common in the Cretan landscape and participate in various sclerophyllous communities. Arbutus sp. as a single taxon includes two species both native to Crete: Arbutus unedo (strawberry tree) is the more common, while Arbutus andrachne (andrachne) has a relatively limited distribution, especially on limestone substrate (Rackham and Moody 1996, 70). Anatomical distinction between the two species is impossible, however, and therefore we present the taxon at the genus level. The taxon Erica sp. may include the tree heather E. arborea and other smaller species. The anatomical distinction is based on the width of the rays in the tangential section (Schweingruber 1990, 367, 369). The preservation of anatomical characteristics of wood charcoal fragments is not always optimal for the anatomical distinction of different species. For this reason we use a single taxon, but we can confirm the existence of fragments with ray widths of five cells or more, and these could be attributed to E. arborea. Acer sp. (maple) is a taxon that includes many species, which, as far as their xylem anatomical characteristics are concerned, cannot be differentiated except in a few cases (Acer platanoides/ Acer pseudoplatanus; Schweingruber 1990, 175, 177). In the Cretan landscape the most common Acer species is A. sempervirens, the Cretan maple, a small deciduous tree or shrub that forms thickets in the mountains and may descend gorges almost to sea level (Rackham and Moody 1996, 70). In this presentation, we use the taxon Acer sp. due to the limitations of xylem anatomy. A few wood charcoal fragments were identified as Platanus orientalis (plane). In general the taxon is rare in pollen cores from pre-Neolithic mainland Greece, but it appears early in pollen cores from Crete (Bottema and Sarpaki 2003). The species

ERNESTINA BADAL AND MARIA NTINOU

102



Level

39

37

35

34

33

32

31

30

29, 28

24

Taxa

No. (%)

No. (%)

No. (%)

No. (%)

No. (%)

No. (%)

No. (%)

No. (%)

No. (%)

No. (%)

Acer sp.

—

—

—

—

—

—

—

—

1 (0.5)

—

Anacardiaceae

—

—

—

—

—

1 (0.4)

—

—

—

—

Arbutus sp.

—

—

1 (1.3)

4 (2.5)

2 (1.1)

9 (3.2)

3 (16.7)

—

5 (2.5)

3 (2.8)

Cistus sp.

—

5 (5.2)

—

6 (3.7)

—

1 (0.4)

—

—

2 (1.0)

—

Conifer

—

—

—

1 (0.6)

—

3 (1.1)

—

—

2 (1.0)

3 (2.8)

Cupressus sempervirens

—

2 (2.1)

—

6 (3.7)

2 (1.1)

1 (0.4)

—

—

7 (3.4)

—

Daphne sp.

—

—

—

—

—

—

—

—

—

1 (0.9)

Erica sp.

—

16 (16.7)

5 (6.4)

19 (11.7)

31 (16.6)

22 (7.8)

—

—

11 (5.4)

4 (3.7)

Ficus carica

—

4 (4.2)

—

—

1 (0.5)

—

—

—

1 (0.5)

2 (1.8)

Fraxinus sp.

—

1 (1.0)

—

—

—

—

—

—

—

—

Juniperus sp.

—

—

—

5 (3.1)

—

2 (0.7)

—

—

1 (0.5)

2 (1.8)

Laurus nobilis

—

—

—

—

—

—

—

—

1 (0.5)

—

Leguminosae

—

1 (1.0)

—

6 (3.7)

2 (1.1)

—

—

—

—

1 (0.9)

Monocotyledons

—

—

1 (1.3)

—

—

—

—

—

—

—

cf. Oreganum sp.

—

1 (1.0)

—

—

—

—

—

—

—

—

Phillyrea/ Rhamnus

—

13 (13.5)

12 (15.4)

12 (7.4)

5 (2.7)

33 (11.7)

—

2 (16.7)

14 (6.9)

6 (5.5)

Pinus sp.

—

—

—

—

—

—

—

—

—

—

Pinus brutia

—

1 (1.0)

2 (2.6)

4 (2.5)

7 (3.7)

1 (0.4)

—

—

10 (4.9)

1 (0.9)

Pistacia lentiscus

—

3 (3.1)

—

4 (2.5)

5 (2.7)

5 (1.8)

—

—

—

2 (1.8)

Pistacia terebinthus

—

1 (1.0)

—

5 (3.1)

4 (2.1)

3 (1.1)

1 (5.6)

—

17 (8.4)

1 (0.9)

Pistacia sp.

—

8 (8.3)

—

10 (6.2)

40 (21.4)

57 (20.2)

—

4 (33.3)

4 (2.0)

3 (2.8)

Platanus orientalis

—

—

—

—

—

—

—

—

—

—

Prunus amygdalus

—

—

—

1 (0.6)

5 (2.7)

12 (4.3)

—

—

3 (1.5)

9 (8.3)

Prunus sp.

—

4 (4.2)

3 (3.8)

1 (0.6)

7 (3.7)

11 (3.9)

—

—

29 (14.3)

10 (9.2)

Quercus sp. deciduous type

20 (66.7)

2 (2.1)

—

—

7 (3.7)

1 (0.4)

—

—

4 (2,0)

—

Quercus sp. evergreen type

7 (23.3)

24 (25.0)

32 (41.0)

65 (40.1)

46 (24.6)

116 (41.1)

10 (55.6)

5 (41.7)

73 (36.0)

52 (47.7)

Quercus sp.

3 (10.0)

8 (8.3)

22 (28.2)

5 (3.1)

10 (5.3)

4 (1.4)

4 (22.2)

—

13 (6.4)

3 (2.8)

Rosaceae

—

2 (2.1)

—

6 (3.7)

13 (7.0)

—

—

—

2 (1.0)

1 (0.9)

Tamarix sp.

—

—

—

1 (0.6)

—

—

—

—

2 (1.0)

1 (0.9)

Indeterminate

—

—

—

1 (0.6)

—

—

—

1 (8.3)

—

1 (0.9)

Nutshell fragment

—

—

—

—

—

—

—

—

1 (0.5)

—

Parenchymatous tissue

—

—

—

—   

—

—

—

—

—

3 (2.8)

30 (100)

96 (100)

78 (100)

162 (100)

187 (100)

282 (100)

18 (100)

12 (100)

202 (100)

106 (97)

1 (3.2)

15 (13.5)

15 (16.1)

46 (22.1)

31 (14.2)

34 (10.8)

—

2 (14.3)

27 (11.7)

29 (21.0)

31 (100)

111 (100)

93 (100)

208 (100)

218 (100)

316 (100)

18 (100)

14 (100)

229 (100)

138 (100)

Subtotal Unidentifiable Total

Table 6.2. Absolute and relative frequencies of taxa identified in the wood charcoal assemblages from Neolithic Knossos. Relative frequency of taxa has not been calculated for levels 20 and 7 due to the scarcity of wood charcoal.

103

WOOD CHARCOAL ANALYSIS: THE LOCAL VEGETATION



23

21

No. (%)

No. (%)

—

—

20

18

17

14

No. No. (%) No. (%) No. (%) — 

—

—

2 (1.1)

12

10

9

No. (%)

No. (%)

No. (%)

—

—

—

8

7

4

3

No. (%) No. No. (%) No. (%) —

—

—

—

—

—

— 

—

—

—

—

—

—

—

—

—

—

3 (5.4)

3 (13.6)

 —

1 (2.9)

4 (5.4)

8 (4.4)

8 (8.6)

16 (18.0)

6 (7.8)

2 (3.8)

—

5 (16.1)

—

—

—

 —

—

3 (4.1)

—

1 (1.1)

4 (4.5)

—

—

—

—

—

1 (1.8)

—

 —

—

—

—

—

—

2 (2.6)

—

—

—

—

—

—

 —

2 (5.7)

—

—

—

—

—

—

—

—

—

—

—

 —

—

—

—

—

—

2 (2.6)

—

—

—

—

2 (3.6)

—

 —

2 (5.7)

3 (4.1)

1 (0.6)

4 (4.3)

5 (5.6)

2 (2.6)

—

—

—

—

—

—

—

1 (2.9)

1 (0.6)

1 (1.1)

—

—

—

 —

—

—

—

—

 —

—

—

—

—

—

—

—

—

—

—

1 (1.8)

1 (4.5)

— 

—

—

—

2 (2.2)

1 (1.1)

—

2 (3.8)

—

—

—

—

—

 —

—

—

—

—

—

—

—

—

—

—

—

—

— 

1 (2.9)

—

—

—

—

—

—

—

—

—

—

—

 —

—

—

—

—

—

—

—

—

—

—

—

—

 —

—

—

—

—

—

—

—

 —

—

—

—

4 (18.2)

 —

—

3 (4.1)

3 (1.7)

2 (2.2)

5 (5.6)

5 (6.5)

1 (1.9)

—

1 (3.2)

4 (14.3)

—

—

 —

—

2 (2.7)

—

3 (3.2)

2 (2.2)

—

—

2

—

—

—

1 (4.5)

1

1 (2.9)

—

1 (0.6)

4 (4.3)

—

1 (1.3)

—

—

1 (3.2)

—

—

—

— 

—

1 (1.4)

—

—

1 (1.1)

—

—

—

—

—

—

—

 —

—

—

—

2 (2.2)

—

1 (1.3)

—

 —

—

—

—

1 (4.5)

 —

—

—

1 (0.6)

5 (5.4)

—

8 (10.4)

—

 —

—

—

—

—

 —

—

—

3 (1.7)

—

—

—

—

—

—

—

1 (1.8)

—

 —

1 (2.9)

3 (4.1)

79 (43.9)

—

8 (9.0)

5 (6.5)

11 (21.2)

—

9 (29.0)

6 (21.4)

6 (10.7)

—

 —

1 (2.9)

25 (26.9)

28 (31.5)

15 (19.5)

10 (19.2)

3

6 (19.4)

8 (28.6)

—

—

— 

—

1 (1.1)

—

—

2 (3.8)

—

—

—

31 (55.4)

8 (36.4)

1

18 (19.4)

17 (19.1)

19 (24.7)

22 (42.3)

—

6 (19.4)

7 (25.0)

6 (10.7)

4 (18.2)

 —

2 (5.7)

1 (1.4)

3 (1.7)

2 (2.2)

2 (2.2)

6 (7.8)

2 (3.8)

—

3 (9.7)

2 (7.1)

1 (1.8)

—

 —

—

1 (1.4)

—

3 (3.2)

—

3 (3.9)

—

—

—

1 (3.6)

—

—

 —

—

—

—

—

—

—

—

—

—

—

—

—

— 

—

—

1 (0.6)

1 (1.1)

—

—

—

—

—

—

4 (7.1)

—

— 

—

—

—

4 (4.3)

—

2 (2.6)

—

—

—

—

—

—

— 

3 (8.6)

4 (5.4)

2 (1.1)

7 (7.5)

—

—

—

—

—

—

52 (93)

22 (100)

2

32 (91)

70 (95)

178 (99)

82 (88)

89 (100)

75 (97)

52 (100)

5

31 (100)

28 (100)

13 (18.8)

3 (12.0)

1

7 (16.7)

16 (17.8) 26 (12.6)

24 (20.5)

22 (19.8)

12 (13.5)

13 (20.0)

— 

9 (22.5)

2 (6.7)

69 (100)

25 (100)

3

42 (100)

90 (100) 206 (100) 117 (100)

111 (100)

89 (100)

65 (100)

5

40 (100)

30 (100)

26 (35.1) 54 (30.0) —

—

20 (57.1) 23 (31.1) 21 (11.7)

ERNESTINA BADAL AND MARIA NTINOU

104 Cultural period Calendar age years b.c. (95.4% probability)

AN

EN I

7050– 6690

5300– 5000

5468– 5228

Archaeological level

39

37

35

Number of charcoal fragments

31

111

93

EN II

5220– 4950

5290– 4960

5310– 5000

5010– 4350

5000– 4730

5208– 4936

34

33

32

31

30

29–28

24

208

218

316

18

14

230

138

Acer sp.

•

Anacardiaceae

•

Arbutus sp.

•

Cistus sp.

•

Conifer Cupressus sempervirens

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

Daphne sp.

•

Erica sp.

•

Ficus carica

•

Fraxinus sp.

•

•

•

Juniperus sp.

•

•

cf. Laurus nobilis

•

Leguminosae

•

Monocotyledons

•

•

•

•

cf. Oreganum

•

Phillyrea/Rhamnus

•

•

•

•

•

Pinus brutia

•

•

•

•

•

•

•

•

•

•

Pinus sp. Pistacia lentiscus

•

•

•

•

Pistacia sp.

•

•

•

•

Pistacia terebinthus

•

•

•

•

• • •

•

•

•

•

Platanus orientalis Prunus amygdalus Prunus sp.

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

Quercus sp.

•

•

Quercus sp. deciduous type

•

•

Quercus sp. evergreen type

•

•

Rosaceae/ Maloideae

•

•

Tamarix sp.

•

• •

•

•

Nutshell fragment

•

•

•

•

•

•

•

Panenchymatous tissue

•

Indeterminate Number of taxa

• 3

17

8

19

• 16

17

4

4

• 21

20

Table 6.3. Presence of plant taxa in wood charcoal assemblages from Neolithic Knossos, along with the total number of fragments analyzed and the total number of taxa identified in each level.

105

WOOD CHARCOAL ANALYSIS: THE LOCAL VEGETATION EN II, cont.

MN 4982– 4774

4990– 4731

LN

23

21

20

18

17

14

12

10

9

8

7

4

3

69

25

3

42

90

206

117

111

89

65

5

40

30

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

• •

• • •

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

• •

• •

• •

• •

•

• •

•

•

•

• •

•

•

•

•

•

•

•

•

•

• •

•

•

•

•

• •

•

• •

•

•

• •

•

•

•

• •

•

•

•

•

• •

• •

•

•

•

•

•

•

•

•

• •

•

•

•

•

•

•

•

• •

10

•

8

2

11

•

12

•

•

•

•

14

18

9

14

8

2

7

6

ERNESTINA BADAL AND MARIA NTINOU

106

25.0 μm

100 μm

a. Cupressus sempervirens, tangential longitudinal section, x150.

b. Cupressus sempervirens, radial longitudinal section, x1100.

25.0 μm

d. Pinus brutia, radial longitudinal section, x350.

100 μm

e. Pistacia lentiscus, transverse section, x130.

10.0 μm

250 μm

250 μm

c. Pinus brutia, transverse section, x60.

250 μm

f. Platanus orientalis, transverse section, x100.

100 μm

g. Platanus orientalis, tangential longitudinal h. Prunus amygdalus, transverse section, x80. i. Prunus amygdalus, tangential longitudinal section, x1800. section, x250. Figure 6.6. Anatomy of plant taxa identified in wood charcoal assemblages from Neolithic Knossos. Photos M. Ntinou.

was probably native in Crete, and this corroborates its presence in the charcoal assemblages from Knossos. The species Pinus brutia is anatomically similar to P. halepensis, and according to Schweingruber (1990, 121) the two are undifferentiated. In discussions of Cretan flora and vegetation P. brutia is considered the only native pine on the island

(Turland, Chilton, and Press 1993, 34; Rackham and Moody 1996, 63), however, and we believe that this is the species represented in the wood charcoal assemblages. Finally we should mention the presence of few small fragments of burned nutshell. Their presence in the assemblages may indicate the discard of food residues in domestic fires.

WOOD CHARCOAL ANALYSIS: THE LOCAL VEGETATION

A general characteristic of the assemblages is their homogeneity in composition (Table 6.3); some of the taxa are constantly present. Evergreen oak (Quercus sp. evergreen type) is found in all assemblages, while mock privet/buckthorn (Phillyrea/Rhamnus), Prunus sp., the almond (Prunus amygdalus), strawberry tree (Arbutus sp.), Pistacia sp., the Cretan pine (Pinus brutia),

107

and heather (Erica sp.) are present in most of them. Deciduous oak (Quercus deciduous type), members of the Rosaceae family, juniper (Juniperus sp.), cypress (Cupressus sempervirens), rockrose (Cistus sp.), and the pea family (Leguminosae) appear in many assemblages. The remaining taxa are rather infrequent.

The Charcoal Diagram Combined qualitative and quantitative results (Table 6.2) for all the assemblages in chronological sequence are presented in a diagram that shows the representation of the taxa diachronically (Fig. 6.7). Comparison between successive assemblages allows us to distinguish vegetation types, their characteristics, and possible changes through time. The frequency of the taxa is shown for those assemblages that included enough fragments (over 70) for a coherent qualitative and quantitative study (i.e., the assemblages from levels 37, 35, 34, 33, 32, 29, 28, 24, 17, 14, 12, 10, and 9). For the remaining, charcoal-poor assemblages (those from levels 39, 31, 30, 21, 20, 18, 8, 7, 4, and 3), only the presence of taxa is indicated with a square black symbol. The best represented taxon in the entire sequence is evergreen oak (Quercus sp. evergreen type). In assemblages 37–32, corresponding to the EN I period, the taxon’s relative frequency increases from 25% to 41%. This frequency, with some fluctuations, is maintained or slightly higher in assemblages 29–24, ascribed to the EN II periods. The taxon’s frequency decreases in assemblage 17 and reaches the lowest point in the sequence (11.7%) at the end of EN II in assemblage 14. The following assemblages 12, 10, and 9, corresponding to the MN period, document a slight increase in the taxon’s frequency, which is maintained at approximately 20%. Mock privet/buckthorn (Phillyrea/Rhamnus) shows a fluctuating tendency in assemblages 37– 24, ascribed to EN I and part of EN II. The taxon’s frequency clearly decreases in the following levels 17–9, that is, at the end of the EN II and in the MN period. The strawberry tree (Arbutus sp.) shows the opposite tendency, starting with low values in the older assemblages and increasing in

frequency in assemblages 14–9. The frequencies of lentisk (Pistacia lentiscus) remain constantly below 5%. The lentisk might be partly represented under the taxon Pistacia sp., as is also the case with turpentine (Pistacia terebinthus). The frequency of Pistacia sp. fluctuates between 10% and 20% in the oldest part of the sequence. The taxon is very well represented in the EN I period (levels 37–32). If all three taxa—Pistacia sp., P. lentiscus, and P. terebinthus—are viewed together, they present similar trends as other evergreen taxa, namely Quercus evergreen and Phillyrea/Rhamnus. Pistacia appears to have been widely used for firewood in EN I and to some extent in EN II, but it was markedly abandoned as a source by the end of the EN II period. Some increase in the Pistacia sp. frequency is observed in level 9, where it reaches 10%. A quite constant and frequent taxon is heather (Erica sp.). The taxon has its highest frequencies (16.7%) of occurrence in the first assemblage in which it appears, level 37, and in level 33, both of the EN I period, and its presence is also notable at the beginning of the EN II period. Its value decreases by the end of the EN II, but it is maintained at approximately 4%–5% thereafter. The frequency of rockrose (Cistus sp.) fluctuates between approximately 1% and 5% in some assemblages, and the leguminosae have a somewhat lower frequency (1%–3%). Other taxa such as Daphne sp., Monocotyledons, and cf. Oreganum are scarce. Deciduous taxa are represented in the Neolithic Knossos diagram by deciduous oak (Quercus sp. deciduous type), almond (Prunus amygdalus), other Prunus species, maple (Acer sp.), ash (Fraxinus sp.), the fig tree (Ficus carica), turpentine (Pistacia terebinthus), and members of the

Archaeological level

Calendar age years b.c. (95.4% probability)

MN

EN II

EN I

Cultural period

10

20

30

40% 10

20%

10%

5%

10%

10% 20

5% 2%

* *

10

20

30

40%

*

10

20

*

*

30% 2 2% 2%

*

*

5%

*

*

5% 2%2% 2%

* *

*

*

10%

*

2% 2 2%2% 2 2% 1 1%

*

*

C U

A C

C N C U P N C U

P U P P

N P U

C N

Figure 6.7. Wood charcoal diagram from Neolithic Knossos showing relative frequencies of taxa in successive excavation levels. Relative frequencies of taxa are calculated on the basis of the fragments identified. Black squares indicate presence of taxa in charcoal-poor assemblages; asterisks indicate taxa with frequencies lower than 0.8%.

No. of charcoal fragments

AN 7050–6690 39 31

5300–5000 37 111

5468–5228 35 93

34 208

5220–4950 33 218

5290–4960 32 316

Quercus evergreen type

5010–4350 30 14 5310–5000 31 18

Quercus sp.

23 69

Phillyrea/Rhamnus

5208–4936 24 138 5000–4730 29– 28 230

Arbutus sp.

3

Pistacia lentiscus

21 25

Pistacia sp.

20

Pistacia terebinthus

18 42

Quercus deciduous type

17 90

*

Prunus amygdalus

4990–4731 12 117 4982–4774 14 206

Prunus sp.

89

Acer sp.

9

Fraxinus sp. Ficus carica

10 111

Rosaceae/Maloideae

5

Pinus brutia

65

Pinus sp.

8

Juniperus sp.

40

Cupressus sempervirens

7

Erica sp.

30

Cistus sp.

3

Leguminosae Daphne sp. cf. Oreganum Monocotyledons Tamarix sp. Platanus orientalis cf. Laurus nobilis

4

A Anacardiaceae C Conifer N Nutshell fragment P Parenchymatous tissue I Indeterminate

LN

1

2

Anthracological zone

108 ERNESTINA BADAL AND MARIA NTINOU

WOOD CHARCOAL ANALYSIS: THE LOCAL VEGETATION

Rosaceae/Maloideae family. With the exception of Prunus sp., the almond, and the turpentine, all the remaining deciduous taxa have a modest representation. Deciduous oak is present in some of the assemblages with low values overall. The presence of the taxon in the earliest occupation level of the sequence, the Aceramic Neolithic level 39, is interesting, however. Wood charcoal was very scarce in this level, and consequently qualitative and quantitative data are incomplete. Nevertheless, the absolute frequency of deciduous oak (Table 6.2) establishes it as the dominant taxon in this assemblage and clearly differentiates level 39 from the rest of the sequence (see discussion below). Prunus sp. and the almond are regularly represented in EN I (assemblages 37–32) with values that approach 4% on average. The taxa show a stable increase during EN II (assemblages 29– 23). The end of EN II, when the almond tree reaches 43.9% and Prunus sp. reaches 30%, is the

109

culminating point for these taxa (especially assemblage 14 for both). High frequency of Prunus sp. is maintained in the MN period. Other deciduous taxa are occasionally present with low values in the assemblages. Their mean relative frequency is around 2%, and only Rosaceae surpass this value once in level 33. Taxa associated with riversides and humid environments, such as the oriental plane (Platanus orientalis) and laurel (Laurus nobilis), are present in only one assemblage each. Tamarisk (Tamarix sp.) is present with low frequencies in three assemblages. The conifers are represented by the Cretan pine (Pinus brutia), the cypress (Cupressus sempervirens), and juniper (Juniperus sp.). The Cretan pine is more constant than the other conifers, although its frequency hardly exceeds 5%. Only in level 29 does it reach 8%. The values of the cypress fluctuate between 2% and 5% and those of juniper between 2% and 4%.

Interpretation The basal assemblage of the diagram and the sequence derives from level 39, the Aceramic Neolithic. Unfortunately, wood charcoal was scarce, and thus it does not allow for a good understanding of the vegetation characteristics of that period. Furthermore, the large chronological distance between this assemblage and the next one represented in the diagram (assemblage 37) makes the comparison between them difficult. Nevertheless, the main characteristic of assemblage 39, the abundance of deciduous oak—the dominant taxon—clearly distinguishes the earliest phase of habitation from the rest of the sequence. The other taxon represented in the assemblage is evergreen oak, and Quercus sp. might be either deciduous or evergreen oak. The abundance of the deciduous oak in this assemblage could be the result of environmental conditions that favored the growth of those trees in the surroundings of the settlement, conditions that changed sometime before EN I, and/or the selective use of these trees for purposes other than firewood. According to the first hypothesis, during the Aceramic (dated to 7050–6690 b.c.), deciduous oaks prevailed in the vegetation around the site.

A change took place between the first occupation and the beginning of EN I (level 37), dated to the late sixth millennium. Sclerophyllous woodland and evergreen oaks dominated the environment thereafter. It is difficult to say if this change occurred due to climatic reasons or because of human activities. There is a long period of time separating the AN occupation from what is considered to be the EN I occupation (almost 1,500 years). However, if human presence was constant in the area during this period of time, it might have affected the natural vegetation and caused the restriction of certain species to protected habitats. Deciduous oaks and humans compete for the same environments, valley bottoms, and in Crete’s present-day vegetation these trees are frequently found growing in abandoned fields. It is possible that deciduous oak groves existed in the Kairatos valley, close to the first Neolithic settlement. If Neo­­ lithic farmers opened small plots for cultivation in the valley, the constant practice of mixed farming activities would eventually have caused the territory of deciduous oaks to shrink. Such a hypothesis seems to be supported by Isaakidou’s (2008) modeling of the subsistence requirements of Neolithic

110

ERNESTINA BADAL AND MARIA NTINOU

Knossos, which, according to the author, could have been easily accommodated in the valley alluvium throughout this period. The early human presence at Knossos and the constant use of the valley floor, possibly a favorable location for deciduous oaks, might have accounted for the reduction of the trees in the vegetation sometime after the first 1,000 years of occupation, especially if that occupation was continuous. According to the second hypothesis, wood charcoals in assemblage 39 were not firewood remains. The presence of only two species in the assemblage might indicate that the remains originated from burned timber or wooden material associated with a structure not revealed in the limited excavation exposure. In general, assemblages with poor plant lists are typical of selective plant use or incidental, instantaneous use (Badal 1992; Chabal 1997; Chabal et al. 1999). The poor plant list of assemblage 39 contrasts with the rest of the sequence, in which a wide array of plants has been identified in individual levels (see Table 6.3). The dominance of deciduous oak in assemblage 39, therefore, would not reflect the characteristics of the vegetation but the selection of these trees for their timber. The discovery of post and stake holes in Evans’s Aceramic Stratum X, along with the end of a burned stake identified as oak, probably of the deciduous type, supports this hypothesis (Evans 1964; Western 1964). The abundance of deciduous oak in the earliest occupation level would thus be a reflection of timber use rather than of the locally dominant vegetation. Even so, these trees were certainly growing in the vicinity of the settlement. Following the Aceramic Neolithic assemblage 39, the diagram shows two anthracological zones that can be distinguished on the basis of the relative representation of the evergreen oak and the almond, Prunus sp. In the first zone, which includes the EN I and most of the EN II period, evergreen oak is dominant. In the second zone, corresponding to the end of the EN II and part of the MN period, the almond is the most abundant taxon. The first anthracological zone is characterized by the dominance of evergreen oak accompanied by other sclerophyllous taxa, as well as deciduous species and conifers. This zone provides a good picture of the EN vegetation of the area and its different environments. The characteristics of the vegetation are typically Mediterranean.

Prickly oak and the other evergreen species would have participated in Mediterranean sclerophyllous formations, which at present are found mostly in a shrub-like state, but may reach an arboreal state and form a dense canopy when left undisturbed by the pressure of coppicing or browsing by animals. Lentisk and juniper would have extended to the areas closer to the coast, giving way farther inland to evergreen oak woodland, in which mock privet/buckthorn would have played an important role. Evergreen woodland would have grown around the settlement, probably forming a noncontinuous mantle interrupted by more open space. Strawberry trees with tree heather might either have grown in separate formations, resembling their modern analogue on the phyllite areas of western Crete, or, most probably, they would have formed part of the understory of evergreen oak thickets. Deciduous species of the Rosaceae/ Maloideae family-subfamily such as Pyrus amygdaliformis and species of the Prunus genus, which are sun-loving and resistant to drought and poor soils, might have occupied rocky areas with an open canopy or areas barren of other arboreal vegetation that were covered by phrygana, rockrose, Daphne, and members of the Leguminosae and Labiatae (such as cf. Oreganum) family. Deciduous oaks would not have been abundant. They probably grew in more humid areas and in the deeper soils of the evergreen woodland. Holm oak might have been found in such places as well, as it demands more humid environments than the other evergreen species, the prickly oak. Cretan pine and cypress are not abundant in the assemblages from Neolithic Knossos. The examples present might represent scattered individuals or, since both are gregarious species, they might have constituted limited inland groves in areas bordering the lower elevations of the mountains. Laurel, plane tree, and probably monocotyledons and ash would have grown on riverbanks or slopes in the upper part of the Kairatos valley. Tamarisk might have grown either in sandy and coastal areas or along riverbanks. The low frequency of these taxa is indicative of the limited extension of riverine formations along the river course. The first zone presents all the characteristics of an area ascribed to the thermomediterranean bioclimatic level, consistent with the lowland,

WOOD CHARCOAL ANALYSIS: THE LOCAL VEGETATION

almost coastal location of the site. The identified taxa could grow under dry (precipitation of 300–500 mm/year) or subhumid (precipitation of 500–700 mm/year) conditions. Neither precipitation nor temperatures would have been much different from the present. The natural plant cover surrounding the site of Knossos was probably a mosaic of formations, with dense woodland interspersed with open vegetation and small cultivated plots. The evergreen formation components of the first anthracological zone persisted in the second zone observed at the end of EN II and into the MN period. The factor that defines the new anthracological zone is the high frequency of Prunus sp. (almond) in some assemblages, surpassing that of the evergreen oak, which is dominant in the rest of the sequence. The abundance of Prunus sp. could be due to a special use of these trees for their edible fruits and/or to a change in the vegetation. The identification of the almond is well documented. The anatomy of this species (ring porous distribution and wide rays) clearly differentiates it from other Prunus species and from P. webbii, a wild almond with bitter and poisonous fruit. Many other small wood charcoal pieces, although they were identified as Prunus sp., could also correspond to the almond, but the identification was limited to the genus because, despite the large and wide rays, they did not preserve an entire growth ring. Therefore, at Neolithic Knossos people were using the almond for firewood and probably for the consumption of the edible fruits (see Sarpaki, this vol., Ch. 5) as early as EN I. These trees were probably part of the natural vegetation, as they occur in Greek mainland areas from the end of the Pleistocene (Ntinou 2002a). By the end of EN II the almond and Prunus in general became the most abundant taxa. This proliferation may reflect the managing of these trees. By this we mean planting with seeds and pruning or even grafting the wild trees, activities that could indicate proto-arboriculture. In the archaeobotanical record only the end products of tree tending (wood from pruning used for fuel) and fruit consumption (nutshells) are preserved to testify to such activities. Consequently, it is difficult to support the

111

idea of either intensive fruit gathering or more sophisticated techniques of fruit tree management. Nevertheless, the end of the EN II period is characterized by a selective use of Prunus species, probably focused on the fruit of these trees. Moreover, there is further evidence in the second zone to support the idea that a change in the vegetation was taking place. Evergreen oaks were still present, but their frequency was lower. In addition, all the other taxa composing previous assemblages also decreased in frequency, with the exception of strawberry tree (Arbutus sp.) and Prunus sp. In well-developed evergreen oak formations Arbutus forms part of the understory, but it tends to spread when the evergreen oak woodland is set on fire, for it benefits to a certain extent from the opening of the canopy; if the degradation is continuous and repetitive, however, it is also adversely affected, and the population diminishes (Braun-Blanquet 1936). The above succession has been observed at prehistoric sites in the western Mediterranean and is attributed to increasing human intervention from the Neolithic onward (Badal, Bernabeu, and Vernet 1994). In the case of Knossos the opposing tendencies of evergreen oak and Arbutus might be related to changes in the density of the evergreen formations caused by human activities. The increase in Prunus sp. (almond) would be in line with this change, especially given that these taxa thrive in open formations and are favored by solar radiation. Changes in the composition and density of the sclerophyllous woodland might have occurred after more than 1,000 years of human presence in the area. Farming, herding, and burning would have affected the fragile equilibrium of evergreen Mediterranean formations, and the evergreen oak woodland surrounding the site would have become less dense. As reported elsewhere (Efstratiou et al. 2004), after a long period (more than 1,000 years) the occupation of the site became more solid with substantial architecture by the end of EN II. In the light of this information, we could explain the increase of Prunus sp. in conjunction with changes in the vegetation and adaptations of subsistence strategies to local resources.

112

ERNESTINA BADAL AND MARIA NTINOU

The Olive The wood charcoal results from Neolithic Knos­ sos show that typically Mediterranean formations were growing around the site. These would ordinarily include components of thermomediterranean vegetation characteristic of the coastal areas of Crete such as evergreen oaks, lentisk, Cretan pine, laurel, and wild olive (Quézel and Barbéro 1985). The olive in its wild state is the indicator of the thermomediterranean bioclimatic level (Ozenda 1982), and at present the olive is a major crop in Crete. It is, however, remarkably absent from the entire charcoal sequence of Knossos. The absence of the olive from the charcoal assemblages of Knossos might indicate that: (a) the species was not native to the island, or at least it did not grow along this part of the northern coast; (b) the species was very rare in the landscape, and it was not gathered for firewood or other purposes; or (c) the species grew around the site, but for cultural reasons it was not used for fuel. The third hypothesis is difficult to evaluate given that the selection or avoidance of plants varies considerably in relation to ideologies and taboos. Therefore, it remains a possible explanation for the absence of the olive from the Neolithic sequence of Knossos. The other two hypotheses can be checked in relation to relevant information from Crete and adjacent areas. The olive appears in the pollen record of Crete late in the Holocene. During the early part of the Holocene the taxon is absent from the Hagia Galini core in South-Central Crete, and Bottema suggests that “the wild olive must have been either rare or even absent from the island” (Bottema 1980, 214). The olive is also absent in the lower spectra of the pollen diagrams from Delphinos (Bottema and Sarpaki 2003) and Tersana (Moody, Rackham, and Rapp 1996), northwestern Crete. It appears for the first time at 6200 b.p. (around 5000 cal. b.c.) in the Delphinos diagram and presents a continuous closed curve after 5700 b.p. (ca. 4700 cal. b.c.). In the Tersana diagram the olive appears at 6000 b.p. According to Bottema and Sarpaki (2003), olives were introduced to the island through overseas contacts, and they were certainly grown before Early Minoan (EM) I, although cultivation and oil production on a larger scale is not documented until the Middle Minoan I period. Moody,

Rackham, and Rapp (1996) have argued, however, that the olive was a natural element of the Pleistocene vegetation of Crete. During that period it survived in refugia somewhere on the island and spread with the onset of the Holocene. The authors attribute its presence during the MN (4750 b.c.) to a native origin, being “a natural part of the oak woodland” (Moody, Rackham, and Rapp 1996, 286). The abundance of the taxon thereafter is an indication of the manipulation of the vegetation by humans and of local cultivation, which could have been an imported practice or a local development (Moody, Rackham, and Rapp 1996; Rackham and Moody 1996, 20). The pollen record is inconclusive regarding the Holocene presence and natural growth of Olea in Crete. Moreover, the oldest olive archaeobotanical remnant, namely an olive stone from the site of Myrtos, dates to the Early Bronze Age (Rackham 1972; Renfrew 1972), millennia after the first Neolithic settlements were established on the island. The only Neolithic charcoal evidence comes from Knossos, and the EN I charcoal results are in agreement with the pollen record regarding the absence of the olive. Furthermore, although the species appears in the pollen record after 5000 b.c., it continues to be absent from the charcoal sequence. According to Moody, Rackham, and Rapp (1996), the olive might have grown naturally in small numbers in the oak woodland of Crete. In the case of Knossos, if the wild olive was a rare element of the natural vegetation, it might have escaped being collected for firewood. The plant list from the site is very rich in taxa, however, and although some of them are scarcely represented, they appear nonetheless, reflecting the use of a variety of environments. Therefore, we would expect the olive to appear at least once, as is the case with other rare taxa like the plane, the tamarisk, and the laurel. The early presence of the wild olive in Neolithic contexts in other parts of the Mediterranean contrasts with the wood charcoal results from the site of Knossos. If the olive grew naturally in the environment, it would probably have been used. In Cyprus the olive is well documented both in the form of wood charcoal and archaeobotanical

WOOD CHARCOAL ANALYSIS: THE LOCAL VEGETATION

remains in the earliest Neolithic settlements of Shillourokambos, Khirokitia, Cape Andreas– Kastros, and Ayios Epiktitos–Vrysi (Thiébault 2003). The analysis of wood charcoal from the site of Shillourokambos shows that the olive became very abundant already by 7500 b.c. Similarly, in the southern latitudes and coastal areas (the thermomediterranean bioclimatic level) of the eastern, central, and western Mediterranean, the olive was already present in Epipaleolithic, Mesolithic, and Early Neolithic contexts, mainly in the form of charcoal remains rather than olive stones (Galili, Weinstein-Evron, and Zohary 1989; Liphschitz et al. 1991; Bernabeu and Badal 1992; Galili et al. 1993; Badal, Bernabeu, and Vernet 1994; Liphschitz 1997; Colledge 2001; Badal 2002; Aura et al. 2005; Rodríguez-Ariza and Montes Moya 2005). The presence of an olive stone (dated to 6415–6089 b.c.) from a Mesolithic context at El Abric de la Falguera, Spain (García Puchol and Aura Tortosa, eds., 2006, 115) corroborates the idea that the species was growing spontaneously in certain places and was used by pre-Neolithic populations. The olive was a native element of the vegetation in these areas, and while the fruit might not have been extensively used, at least during the Neolithic, the wild plants were used for fuel and fodder (Badal 1999, 2002). In mainland Greece the olive does not appear in wood charcoal samples from Neolithic sites, although most of these are located in the northern part of the country, which probably did not offer the optimum conditions for the natural growth of

113

the species. The olive is absent from the Neolithic levels of the site of Limenaria on Thassos (Ntinou 2012), and it only appears in small numbers in LN assemblages from the Cave of the Cyclops on the island of Youra (Ntinou 2002b, 2011). In view of the information about the taxon in the archaeobotanical remains from Neolithic sites in Greece, questions concerning the native origin and use of the species remain open. The early presence of the wild olive in Neolithic contexts in other parts of the Mediterranean contrasts with the wood charcoal results from the site of Knossos. Thus, it seems probable that the olive either was not native to Crete or did not grow in the wider area of Knossos. Moreover, even if the plant was grown locally in western Crete before EM I, as the pollen results indicate (Moody, Rackham, and Rapp 1996; Bottema and Sarpaki 2003), this activity probably did not take place at Neolithic Knossos. If it had been grown there, we would expect to find olive wood charcoal remains among the other archaeobotanical material. The pruning of olive trees would provide wood that could be used for fuel and would eventually be represented, even in small numbers, in the charcoal assemblages. Whether the species was introduced to Crete relatively late in the Neolithic as Bottema and Sarpaki (2003) postulate or was present on the island during the EN period but did not grow/was not grown in the wider area of Knossos will remain an open question until more charcoal results are available from early contexts and from different locations on the island.

Discussion The wood charcoal data from Neolithic Knossos offer information regarding the history of Mediterranean plant formations and their use by the first settlers of the island from the Aceramic to the Late Neolithic. The vegetation around Knossos was typically Mediterranean, presenting a mosaic of evergreen oak woodland and open xerophytic formations. The conifers associated with the vegetation of Crete, namely Pinus brutia and Cupressus sempervirens, probably grew at some distance from the

site at higher altitudes. Riverine and estuarine environments were only used sporadically by the settlers of Knossos, but the existence of such habitats is documented in the charcoal results and points to the biodiversity of the area. The wood charcoal data from Neolithic Knossos can be compared to the Holocene pollen record. The pollen record for Crete starts during the early Holocene, in pre-Neolithic times, with the pollen core from Hagia Galini (Bottema 1980) on the southern coast. This shows a relatively high pine

114

ERNESTINA BADAL AND MARIA NTINOU

frequency at first, after which the pine decreases rapidly and is replaced by oak as the main arboreal species. During pre-Neolithic times woodland was relatively abundant. Plants characteristic of dry open conditions such as Leguminosae, Compositae, Umbelliferae, and Asphodelus existed as well. The absence of wild olive pollen suggests that the olive might not have been present on Crete in early Holocene times. According to the Hagia Galini diagram (Bottema 1980), throughout the Neolithic the southern coast continued to be a mosaic of open formations and woodland dominated by deciduous oak. Other pollen types included evergreen oak and pine. These characteristics prevailed until 4300 b.c., when a decline in the woodland vegetation is documented. Two more pollen diagrams from northwestern Crete, Delphinos and Kournas, add information concerning the Holocene vegetation of the island (Bottema and Sarpaki 2003). The percentages of nonarboreal pollen are higher than arboreal pollen until 6300 b.c., the beginning of EN I, indicating that climatic conditions were dryer and that the forest near the coast was sparser than today. Deciduous and evergreen oaks could have grown in a few places near the coast, or, alternatively, the observed arboreal pollen might represent the vegetation at higher elevations. According to Bottema and Sarpaki, the presence of central European taxa (lime, hazel, hornbeam) in these two cores, as well as in others from northwestern Crete (see below for Tersana), is due to long-distance transport. After 6300 b.c. and during the EN I period, there is an increase in both evergreen and deciduous oaks. Pistacia and Phillyrea start forming continuous curves. The landscape included all the typical Mediterranean components. The main cause for the spread of oak forest appears to have been an increase in winter precipitation that might have brought about a shift from the previous dry conditions to the modern situation. Changes in the vegetation are observed around 5000 b.c., that is, at the end of EN I and the beginning of EN II. Quercus decreases slightly, Ericaceae increase, the olive appears for the first time, and indicators of crop cultivation and animal husbandry appear or increase. The indicators of human activity increase considerably after 4870 b.c., the end of EN II, and the anthropogenic impact becomes apparent in the fifth millennium.

Complementary information for the Neolithic period comes from the Tersana core in northwestern Crete, which documents the existence of mosaic vegetation of phrygana and woodland at the beginning of the Neolithic (Moody 1987; Moody, Rackham, and Rapp 1996). The woodland included Mediterranean and central European taxa, namely evergreen and deciduous oaks, lime, hazel, and hornbeam. The Central European taxa would indicate that the climate was moister than today. During the MN, ca. 4750 b.c., olive pollen that was not present earlier begins to appear in small quantities and indicates human manipulation of the local vegetation. By the LN it is abundant enough to indicate local cultivation (Rackham and Moody 1996). From this time onward the decrease in oak woodland and the increase in phrygana and steppe taxa suggest a modification of the natural plant environment due to human activities, especially land clearance for agriculture. The pollen cores and the wood charcoal diagram from Knossos show similarities both in the components of the vegetation of successive periods and in the timing of the changes that took place in the plant formations. The different locations of the study areas may account for the discrepancies between them. The pollen cores are from western Crete, which is considerably moister and presents more microenvironments than Central Crete, where Knossos is located. The vegetation that the first settlers of Knossos encountered around 7000 b.c. is difficult to describe in detail because of the scarcity of charcoal from the earliest occupation level. Even so, deciduous and evergreen oaks grew in proximity to the site. The arboreal pollen of the same period is composed mainly of the same species. The extraordinary presence of central European taxa is interpreted by Bottema and Sarpaki (2003) as an effect of long-distance transport during drier conditions in the first three millennia of the Holocene, while others (Moody 1987; Moody, Rackham, and Rapp 1996) interpret these taxa as evidence of moister conditions. Although the charcoal evidence cannot resolve this question, since only oaks are represented in the earliest level, it shows clearly that deciduous species grew in lowland areas and near the coast. Later on, during the sixth millennium, the pollen cores show that the vegetation in western Crete

WOOD CHARCOAL ANALYSIS: THE LOCAL VEGETATION

was characterized by an expansion of woodland in which deciduous oaks played an important role. Typical Mediterranean elements, namely evergreen oaks, mock privet, and Pistacia expanded as well. According to the wood charcoal results, approximately at the same time in EN I and II, a typically Mediterranean woodland of evergreen oaks grew around Knossos. Woodland or more open formations are documented in the wood charcoal assemblages, and their components are the same as those found in the pollen cores. Deciduous oaks are not so important in the vegetation around Knossos, unlike the situation in western Crete. This is probably due to differences in the precipitation and topography of these two regions. Western Crete is rainier, and within a few kilometers of the coast mountain peaks rise to more than 1,000 m of altitude, a situation that favors the existence of different microclimates and microenvironments with diverse vegetation and plant formations. These conditions probably contributed to the proliferation of deciduous oak in the western parts of the island. As one moves to the east the conditions become relatively drier, perhaps accounting for the lower representation of deciduous oak in the wood charcoal diagram from Knossos. Moreover, early and continuous human presence at Knossos might also have restricted the growth of deciduous oak, especially if Neolithic farmers competed with these trees for the deeper soils of the valley bottoms. Drier conditions around Knossos probably favored the growth of evergreen oaks and xerophytic formations. Cretan pines, cypresses, and maples might have grown in the nearby low mountains. In general terms, both the pollen cores and the charcoal diagram attest to the existence of a mosaic of woodland and open vegetation areas. Evidence for changes in the vegetation appears and increases during the fifth millennium. In the pollen cores, the decrease of oak, the increase of Ericaceae and plants associated with farming activities, and, most importantly, the first appearance of the olive are prominent indicators of human activities that would have caused changes in the vegetation. By the end of EN II the charcoal diagram from Knossos shows similar characteristics. Evergreen oaks decrease, while strawberry trees (a member of the Ericaceae) and Prunus increase, probably as a result of the opening of the

115

woodland caused by human activities. Contrary to what is observed in the pollen cores, the olive is absent from the whole Neolithic sequence of Knossos. Other indications of tree management and early arboriculture, however, may be seen in the abundance of Prunus sp. (almond) from the end of EN II onward, consistent with the proposed use of the olive based on the pollen evidence. Members of Rosaceae such as Prunus are usually absent from pollen cores because they are pollinated by insects. Their presence and abundance in the charcoal diagram might be an effect of clearance of the woodland caused by human activities and/or special treatment of these trees with an emphasis on fruit collection. Changes in the vegetation and evidence for the intensification of farming practices at Knossos occurred after more than 1,000 years of Neolithic presence at the site, when, according to all the archaeological information, the consolidation of the settlement took place. Concerning the olive tree, the pollen record from Crete diverges from the wood charcoal results from Knossos. The earliest appearance of the olive is in the Delphinos pollen diagram, around 5000 b.c. (Bottema and Sarpaki 2003), and after approximately 4500 b.c. it shows continuous curves in all pollen cores. The Tersana pollen diagram documents the appearance and increase of olive pollen grains from the MN onward. Rackham and Moody (1996) argue that the wild olive is native to Crete, having survived glaciations in warm gorges and expanding later to the coastal areas. At Neolithic Knossos the olive is completely lacking, probably because it did not grow spontaneously in the area, nor was it deliberately grown by humans. Wood charcoal analysis results from other Neolithic sites in Crete are not available, and therefore the data from Knossos can only be compared to relevant information from a few other coastal areas and islands in the Aegean. Information for the vegetation during the second half of the sixth millennium comes from two sites in the northern Aegean, the coastal site of Makri, Thrace, and the settlement of Limenaria on the island of Thassos (Ntinou 2002a, 2012). Both these areas are located at a long distance from Crete and at a higher latitude, which may explain the importance of deciduous oaks in their natural vegetation, contrasting with the dominance of evergreen oak woodland

116

ERNESTINA BADAL AND MARIA NTINOU

and xerophytic formations at Knossos. At Makri, evergreen oaks, lentisk, strawberry tree, and other thermophilous plants are absent. On the island of Thassos all of the above-mentioned species are present, along with deciduous oaks. Farther to the south on the island of Youra, the Mesolithic huntergatherers used the evergreen formations with Phillyrea/Rhamnus and evergreen oaks early in the Holocene, and the same vegetation thrived during the Neolithic (Ntinou 2011). These vegetational characteristics are quite similar to the ones from Knossos and contrast with those from the northern latitudes. Thus, we can see a north–south transect along which Mediterranean deciduous formations

give way to evergreen formations in response to latitude, temperature, and moisture. The presentday evergreen maquis and the shiblyak formations (associations of deciduous scrubs and short trees as a result of the degradation of oak forests) of the northern areas may be interpreted as the result of human activities that caused the substitution or modification of the Holocene deciduous woodland through the millennia. At sites in the southern latitudes such as those in the Sporades and Crete, a mosaic of evergreen oak woodland and xerophytic formations formed the natural vegetation used by the human groups, who gradually transformed the landscape through their agricultural practices.

Conclusions Wood charcoal analysis of the Neolithic deposits at Knossos was undertaken in order to describe the local vegetation and the way it was used by the first settlers of the area. Although the small size of the excavation placed limitations on sampling and recovery of detailed paleoenvironmental information, we believe that the charcoal results for the Neolithic sequence are coherent and in agreement with other lines of paleoenvironmental evidence. The area around the Neolithic settlement presented a variety of environments that are reflec­ted in the identified plant taxa. A mosaic of evergreen oak woodland and open plant formations was the most common plant cover in the area and the most extensively used by Neolithic people. Deciduous oaks were a rare component, probably associated with mature evergreen woodland and growing in favorable places with deeper soils also used by Neolithic farmers. These deciduous trees were widely used in the first Aceramic settlement. Cretan pines and cypresses, characteristic species of the Cretan flora, would have grown in the nearby mountains. The riverside and saline environments were seldom used for the gathering of

firewood, although the valley of the river Kairatos would have been the main area of farming activity. Throughout the Neolithic, the Knossos settlers made use of the local vegetation for firewood. Changes relating to the density of the plant formations, especially the oak woodland, become evident by the end of the EN II period, and they should be interpreted in conjunction with the consolidation of the settlement and human activities in the area. Among other farming activities, tree management or arboriculture of the almond/ Prunus sp. is reflected in the abundant remains of these taxa. Such activities would have been the result of a longer process involving the adoption of local resources in the diet. It is remarkable that the olive, a typical component of the Mediterranean sclerophyllous forma­ tions, is absent from the wood charcoal of Neolithic Knossos. Without other charcoal or archaeobotanical data from relevant chronological contexts, and given the late appearance of the olive in the pollen cores from Crete, we are inclined to believe that the species did not grow and/or was not purposely grown in the area.

WOOD CHARCOAL ANALYSIS: THE LOCAL VEGETATION

117

References Aura, J.E., Y. Carrión, E. Estelles, and G. Pérez. 2005. “Plant Economy of Hunter-Gatherer Groups at the End of the Last Ice Age: Plant Macroremains from the Cave Santa Maira (Alacant, Spain) ca. 12,000– 9000 b.p.,” Vegetation History and Archaeobotany 14, pp. 542–550. Badal, E. 1990. Aportaciones de la antracología al estudio del paisaje vegetal y su evolución en el cuaternario reciente, en la costa mediterránea del País Valenciano y Andalucía (18.000–3000 b.p.), Ph.D. diss., Universitat de València. ———. 1992. “L’anthracologie préhistorique: À propos de certains problèmes mèthodologiques,” in Les charbons de bois, les anciens écosystèmes et le rôle de l’homme (Bulletin de la Société Botanique de France 139, Actualités botaniques 1992 [2/3/4]), Paris, pp. 167–189. ———. 1999. “El potencial pecuario de la vegetación mediterránea: Las cuevas redil,” in Actes del II Congrés del Neolític a la Península Ibérica (Saguntum-PLAV Extra 2), J. Bernabeu and T. Orozco, eds., Valencia, pp. 69–75. ———. 2002. “Bosques, campos y pastos: El potencial económico de la vegetación mediterránea,” in Neolithic Landscapes of the Mediterranean (Saguntum-PLAV Extra 5), E. Badal, J. Bernabeu, and B. Martí, eds., Valencia, pp. 129–146. Badal, E., J. Bernabeu, and J.-L. Vernet. 1994. “Vegetation Changes and Human Action from the Neolithic to the Bronze Age (7.000–4.000 b.p.) in Alicante, Based on Charcoal Analysis,” Vegetation History and Arcaeobotany 3, pp. 155–166. Bernabeu, J., and E. Badal. 1992. “A View of the Vegetation and Economic Explotation of the Forest in the Late Neolithic Sites of Les Jovades and Niuet (Alicante, Spain),” in Les charbons de bois, les anciens écosystèmes et le rôle de l’homme. Colloque organisé à Montpellier du 10 au 13 septembre 1991 (Bulletin de la Société Botanique de France 139; Actualités botaniques 1992 [2/3/4]), Paris, pp. 697–714. Bottema, S. 1980. “Palynological Investigation on Crete,” Review of Palaeobotany and Palynology 31, pp. 193–217. Bottema, S., and A. Sarpaki. 2003. “Environmental Change in Crete: A 9000-Year Record of Holocene Vegetation History and the Effect of the Santorini Eruption,” The Holocene 13, pp. 733–749.

Braun-Blanquet, J. 1936. “La forêt d’yeuse languedocienne,” Memoire de la Société de Sciences Naturelles de Nîmes 5, pp. 1–47. Chabal, L. 1988. “Pourquoi et comment prélever les charbons de bois pour la période antique: Les méthodes utilisées sur le site de Lattes (Hérault),” Lattara 1, pp. 187–222. ———. 1997. Forêts et sociétés en Languedoc (Néolithique final, Antiquité tardive): L’anthracologie, méthode et paléoécologie (Documents d’Archéologie Française 63), Paris. Chabal, L., L. Fabre, J.-F. Terral, and I. Théry-Parisot. 1999. “L’Anthracologie,” in La Botanique, C. Bourquin-Mignot, J.-É. Brochier, L. Chabal, S. Crozat, L. Fabre, F. Guibal, P. Marinval, H. Richard, J.-F. Terral, and I. Rhéry, Paris, pp. 43–104. Colledge, S. 2001. Plant Exploitation on Epipalaeolithic and Early Neolithic Sites in the Levant (BAR-IS 986), Oxford. Dufraisse, A., ed., 2006. Charcoal Analysis: New Analytical Tools and Methods for Archaeology. Papers from the Table-Ronde Held in Basel 2004 (BAR-IS S1483), Oxford. Efstratiou, N., A. Karetsou, E. Banou, and D. Margomenou. 2004. “The Neolithic Settlement of Knossos: New Light on an Old Picture,” in Knossos: Palace, City, State. Proceedings of the Conference in Herakleion Organised by the British School at Athens and the 23rd Ephoreia of Prehistoric and Classical Antiquities of Heraklion, in November 2000, for the Centenary of Sir Arthur Evans’s Excavations at Knossos (BSA Studies 12), G. Cadogan, E. Hatzaki, and A. Vasilakis, eds., London, pp. 39–51. Evans, J.D. 1964. “Excavations in the Neolithic Settlement of Knossos, 1957–60. Part I,” BSA 59, pp. 132–240. Fiorentino, G., and D. Magri, eds. 2008. Charcoals from the Past: Cultural and Palaeoenvironmental Implications. Proceedings of the Third International Meeting of Anthracology, Cavallino, Lecce (Italy), June 28th—July 1st, 2004 (BAR-IS 1807), Oxford. Galili, E., M. Weinstein-Evron, and D. Zohary. 1989. “Appearance of Olives in Submerged Neolithic Sites along the Carmel Coast,” Mitekufat Haeven 22, pp. 95–97. Galili, E., A. Hershkowitz, A. Gopher, M. WeinsteinEvron, O. Lernau, M. Kislev, and L. Horwitz. 1993.

118

ERNESTINA BADAL AND MARIA NTINOU

“Atlit-Yam: A Prehistoric Site on the Sea Floor of the Israeli Coast,” JFA 20, pp. 133–157. García Puchol, O., and E. Aura Tortosa, eds. 2006. El abric de la Falguera (Alcoi, Alacant): 8000 años de ocupación humana en la cabecera del río de Alcoi, Alcoi. Isaakidou, V. 2008. “The Fauna and Economy of Neolithic Knossos Revisited,” in Escaping the Labyrinth: The Cretan Neolithic in Context (Sheffield Studies in Aegean Archaeology 8), V. Isaakidou and P. Tomkins, eds., Oxford, pp. 9–114. Liphschitz, N. 1997. “Wood Remains from Two PPNB Sites: Horvat Galil and Nahal Beset,” Tel Aviv 24, pp. 237–239. Liphschitz, N., R. Gophna, M. Hartman, and G. Biger. 1991. “The Beginning of Olive (Olea europaea) Cultivation in the Old World: A Reassessment,” JAS 18, pp. 441–453. Moody, J. 1987. The Environmental and Cultural Prehistory of the Khania Region of West Crete, Ph.D. diss., University of Minnesota, Minneapolis. Moody, J., O. Rackham, and G. Rapp. 1996. “Environmental Archaeology of Prehistoric NW Crete,” JFA 23, pp. 273–297. Ntinou, M. 2002a. El paisaje en el norte de Grecia desde el Tardiglaciar al Atlantico: Formaciones vegetales, recursos y usos (BAR-IS 1038), Oxford. ———. 2002b. “Vegetation and Human Communities in Prehistoric Greece,” in Neolithic Landscapes of the Mediterranean (Saguntum-PLAV Extra 5), E. Badal, J. Bernabeu, and B. Martí, eds., Valencia, pp. 91–103. ———. 2011. “Charcoal Analysis at the Cave of the Cyclops, Youra, Northern Sporades,” in The Cave of the Cyclops: Mesolithic and Neolithic Networks in the Northern Aegean, Greece. Vol. II: Bone Tool Industries, Dietery Resources, and the Paleoenvironment and Archaeometrical Studies (Prehistory Monographs 31), A. Sampson, ed., Philadelphia, pp. 297–314. ———. 2012. “Anthracological Analysis at the Neolithic Settlement of Limenaria, Thassos,” in Δέκα Χρόνια Ανασκαφικής Έρευνας στον Προϊστορικό Οικισμό Λιμεναρίων Θάσου, S. Papadopoulos and D. Malamidou, eds., Thessaloniki, pp. 77–93. Ozenda, P. 1982. Les végétaux dans la biosphère, Paris. Quézel, P., and M. Barbéro. 1985. Carte de la végétation potentielle de la région méditerranéenne, Paris.

Rackham, O. 1972. “Appendix III: Charcoal and Plaster Impressions,” in Myrtos: An Early Bronze Age Settlement in Crete (BSA Suppl. 7), P. Warren, London, pp. 299–304. Rackham, O., and J. Moody. 1996. The Making of the Cretan Landscape, Manchester. Renfrew, J. 1972. “Appendix V: The Plant Remains,” in Myrtos: An Early Bronze Age Settlement in Crete (BSA Suppl. 7), P. Warren, London, pp. 315–317. Roberts, N. 1979. “The Location and Environment of Knossos,” BSA 74, pp. 231–240. Rodríguez-Ariza, M.O., and E. Montes Moya. 2005. “On the Origin and Domestication of Olea europaea L. (Olive) in Andalucia, Spain, Based on the Biogeographical Distribution of Its Finds,” Vegetation History and Archaeobotany 14, pp. 551–561. Schweingruber, F.H. 1990. Anatomy of European Woods, Bern. Thiébault, S. 2003. “Les paysages végétaux de Chypre au néolithique: Premières données anthracologiques,” in Le Néolithique de Chypre. Actes du colloque international organisé par le Département des Antiquités de Chypre et l’École Française d’Athènes, Nicosie 17–19 mai 2001 (BCH Suppl. 43), J. Guilaine and A. Le Brun, eds., Athens, pp. 221–230. Thiébault, S., ed. 2002. Charcoal Analysis: Method­ ­ ological Approaches, Palaeoecological Results and Wood Uses. Proceedings of the Second International Meeting of Anthracology, Paris, September 2000 (BAR-IS 1063), Oxford. Turland, N.J., L. Chilton, and J.R. Press. 1993. Flora of the Cretan Area: Annotated Checklist and Atlas, London. Vernet, J-L., ed. 1992. Les charbons de bois: Les anciens écosystèmes et le rôle de l’homme. Colloque International, Montpellier du 10 au 13 septembre 1991 (Bulletin de la Société Botanique de France 139; Actualité botanique 1992-2/3/4), Paris. Western, A.C. 1964. “Appendix 2: Timber from Neolithic Knossos, Pit F, Stratum X, Area AC, 1960,” in “Excavations in the Neolithic Settlement of Knossos. Part I,” J.D. Evans, BSA 59, pp. 240–241. Zohary, D., and G. Orshan. 1966. An Outline of the Geobotany of Crete (Israel Journal of Botany 14, Suppl. 5), Jerusalem.

7

Plant Economy and the Use of Space: Evidence from the Opal Phytoliths Marco Madella

Knossos is a key settlement for understanding the spread of farming societies in southeast Europe and the Mediterranean islands.* Research on the introduction of agricultural practices and cultigens in the islands has regained momentum with the appearance of new evidence for the spread of farming populations to Cyprus as early as the eighth millennium b.c. (Peltenburg et al. 2000). With the possibility that island Mesolithic populations, wherever they existed, were capable of exploiting local plants (e.g., einkorn, barley, lentils) still archaeologically undocumented, the conclusion that domesticated crops were brought from the east by people who were either the first colonists and the first farmers or just the latter seems inescapable.

*Abbreviations used in this chapter are: cc cubic centimeters Ch(s). Chapter(s) cm centimeters EN Early Neolithic g grams

While maritime voyaging in the Aegean was postulated for early prehistory (Cherry 1990; van Andel and Runnels 1995; Broodbank 1999), the full archaeological picture of Mesolithic habitation in the islands has only recently started to appear (Broodbank 2006; Sampson 2006, 2008). At the same time, the beginning of the Neolithic period in the islands, although better known, still requires the gathering of more data related to many important issues that are presently open to speculation, such as eustatic changes, dates and patterns of spread, and the effect of research biases (Cherry, Bennet, and Wilson 1991; Hansen 1992). Aceramic Knossos is the earliest Neolithic site found in Crete so far (Zois 1973; Manning 1999)

LN m mm MN sp.

Late Neolithic meters millimeters Middle Neolithic species

120

MARCO MADELLA

and one of the earliest in Greece—perhaps the earliest. Knossos is thus extremely important and interesting for a number of reasons, not only because of its insular character but also with regard to the geographical direction of the westward spread of farming from the eighth millennium b.c. onward. It is significant that although the overall colonization process in Crete remains obscure due to the lack of other sites, Neolithic Knossos was clearly founded at least 1,000 years after the first farming settlements in Cyprus (Peltenburg et al. 2000). Rescue excavations carried out in the Central Court of the palace of Knossos in 1997 have given us the opportunity to investigate the earliest occupation levels of the site, including the Aceramic Neolithic, and to devote special attention to the plant economy. For this purpose both plant macroremains, that is, charred plant remains and wood charcoal (see Sarpaki, this vol., Ch. 5; Badal and Ntinou, this vol., Ch. 6) and phytoliths were analyzed. Phytoliths are silica bodies produced in the cell lumen and in the intercellular space of many vegetal tissues. A plant family primarily composed of silica-accumulating species is the Gramineae (Poaceae), to which belong several plants of economic interest, including the cereals. Phytoliths are deposited in sediments after the decay of the tissues’ organic matter, and they can easily survive

as opaline silica, which is very resistant to decay. Phytoliths may be released in the sediments as single elements—the single phytolith formed in a plant cell—or as sheets of articulated elements. These articulated elements, called silica skeletons, are produced when most or all the cells of the epidermis (and sometimes of the mesophyll) of a plant are filled with opal silica. Complete silicification of the epidermal cell occurs when water availability and evapotranspiration are high. When the organic matter of the tissues decays, the phytoliths maintain the same anatomical position as in the original tissue, therefore preserving the original tissue’s information. This occurs very often in the case of irrigated crops (Rosen and Weiner 1994). Due to the retained anatomical information, silica skeletons are often important for the identification of the plant taxa. Phytoliths have been successfully utilized for the identification of agricultural practices in archaeological sites of the Old World dating from the Early Neolithic (EN) to the Bronze Age (e.g., Rosen 1993; Madella 2001, 2003; Harvey and Fuller 2005). The study of phytoliths from the Neolithic levels at Knossos was conducted especially to examine the plant economy of the earliest phases of occupation and to investigate the environment of this early Aegean settlement.

Methodology Forty-eight sediment samples representing the Aceramic–Late Neolithic cultural sequence were selected for the present study from the 1997 excavation at the southeast edge of the Central Court of the palace of Knossos (see Fig. 1.2). Of these, 16 samples were collected from the south profile and 32 from the west profile of Trench II. The samples originated from noncontextual sediments, which means that they came from general anthropic layers but not from specific structures (see Fig. 1.4). Approximately 10 cc of sediment from each level were collected and sealed in airtight bags in order to avoid contamination. A subsample of 3 g of sediment from each sample was carefully weighed on a three-decimal balance and processed according to the extraction technique of Madella, Powers-Jones,

and Jones (1998). This extraction protocol allows the concentration of the silica component of the sediment while eliminating carbonates, organic matter, and clays. The silica fraction was then isolated by gravimetric separation by means of a heavy liquid solution, and it was subsequently dried. A small amount (0.001 g) of the final residue of each sample was mounted in Styrolite™ with a 22 x 40 mm cover slip. A count of phytoliths from a fixed five-field catchment area was undertaken. The count of a fixed number of fields per slide with a known amount of final residue allows for an approximate estimate of the abundance of phytoliths in the sediments, even if it is not as precise an estimate as the calculation of the absolute number of phytoliths per gram of sediment. The identification

PLANT ECONOMY AND THE USE OF SPACE: EVIDENCE FROM THE OPAL PHYTOLITHS

of phytolith typologies was carried out at 200x, 400x, and 1000x magnification in polarized light on a Nikon Labophot optical microscope. Since the aim of the present study was to elucidate the early plant economy and environment of the Neolithic settlement of Knossos, the phytolith scanning was designed accordingly and targeted to the anatomical identification of plant material (e.g., culm vs. leaves) and the possible identification of plant taxa. The morphological categories used were: Single morphotypes Grass long cells, inflorescence Grass long cells, culm/leaf Articulated morphotypes (silica skeletons) Wheat type (Triticum sp.), inflorescence Barley type (Hordeum sp.), inflorescence Millet type (Panicum sp./Setaria sp.), inflorescence Grass silica skeleton, inflorescence Grass silica skeleton, culm/leaf Dicotyledons

121

Since anatomical characteristics in plant tissues may reflect adaptation to the climate, short cell phytoliths from grass epidermis have also been analyzed. Short cells are characteristic at the subfamily level, and three main taxonomic groups can be identified on the basis of these phytoliths: Pooideae, Panicoideae, and Chloridoideae (Twiss, Suess, and Smith 1969). The three grass subfamilies are broadly adapted to distinct climates, and they show anatomical differences in the leaf cells because they have different photosynthetic pathways. The phytoliths that characterize these grass subfamilies are: Pooidoid, mainly C3, e.g., short trapezoids typologies Panicoid, mainly C4, e.g., dumbbells typologies Chloridoid, both C3 and C4, e.g., saddles typo­logies. The C3 photosynthetic pathway is typical of grasses adapted to temperate, cool climates while the C4 pathway is typically found in grasses of warm environments.

Results South Profile Sixteen samples were analyzed from the south profile (see Fig. 1.4). The four top phytolith samples (I, II, IIIos, IIIam), attributed to the Late Neolithic (LN) stratigraphic levels 1–3, have low phytolith frequencies (Table 7.1). Samples I to IIIam are mainly composed of single, nonarticulated phytoliths from grass culm and leaves, which constitute up to 70% of the assemblages, and of grass inflorescence phytoliths (Fig. 7.1). Silica skeletons from grasses and dicotyledons are almost completely absent in these samples, with the exception of a few millet-type examples in samples II and IIIos. In samples IV–VIII (Middle Neolithic [MN] levels 4–13), single, nonarticulated phytoliths are still important. There is also, however, a

substantial presence of silica skeletons from undetermined cereal inflorescence (20% or more), as well as wheat-type, millet-type, and some barleytype silica skeletons (Fig. 7.1). Dicotyledon phytoliths account for up to 5% of the assemblages. Samples IX–XIII from the EN II levels 14a– 28+29 have similar frequencies of phytoliths from the culm and inflorescence of grasses—both as single cell and silica skeletons—as well as wheat type, barley type, millet type, and dicotyledon silica skeletons. Samples from the Aceramic deposit were not available from this profile. Grass short cells in the south profile have a frequency of C3 types varying between 59% and 78% in a 200 short cell count (Fig. 7.2).

MARCO MADELLA

122 South Profile

LN

MN

Level

1

2

Sample Number

I

II

IIIos

IIIam

Grass long cells, spiny edges

36

63

105

40

Grass long cells, smooth or slightly wavy edges

78

92

84

3

EN II

4

9

10

12

13

14a

16

20

21

23

26+27

28+29

IV

V

VI

VII

VIII

IX

Xa

Xb

Xc

XI

XII

XIII

129 118

74

85

76

60

126 147 152 132

155

93

31

152

38

92

79

59

85

93

132

112

Grass single morphotypes

139 148 125

Grass articulated morphotypes (silica skeletons) Wheat type (Triticum sp.), inflorescence

0

0

0

0

0

2

4

1

4

13

3

8

6

2

0

12

Barley type (Hordeum sp.), inflorescence

0

0

0

0

2

0

0

0

1

9

2

0

1

0

0

2

Millet type (Panicum sp., Setaria sp.), inflorescence

0

4

4

0

0

0

1

2

3

6

24

3

5

3

6

0

Grass inflorescence

0

0

0

0

6

62

47

56

92

85

35

46

52

41

36

30

Grass culm/leaf

0

0

0

0

1

0

18

14

43

66

18

49

50

43

44

47

0

0

2

7

4

1

5

18

3

2

5

1

Dicotyledons Total

114 159

0

0

193

71

292 227 240 238 283 342 304 394 419 347

2

11

375

307

e G sk ras in ele s s flo to il re ns ica sc , en G ce r s as cu kele s s lm to ilic /le ns a af , D ic ot yl ed on s

yp tt

M

ille

rle

y

ty Ba

he

at

Grass long cells, culm/leaf

W

Grass long cells, inflorescence

ty

pe

pe

Table 7.1. Knossos 1997: south profile phytolith counts.

Phase E

0

200

Phase D

Depth in cm

100

300

Figure 7.1. Bar chart of phytolith percentage frequencies from the south profile.

400 % 0 10 20 30 40 50

80

0 10 20 30 40 50 60

0 10

0 10

0

10

0 10 20

0

10

0 10

C3 C4

70 Percentage (%)

60 50 40 30 20 10 0

I

II

IIIos

IIIam

IV

V

VI

VIII VII Sample Numbers

IX

Xa

Xb

Figure 7.2. Bar chart of C3 and C4 phytolith percentage frequencies from the south profile.

Xc

XI

XII

XIII

PLANT ECONOMY AND THE USE OF SPACE: EVIDENCE FROM THE OPAL PHYTOLITHS

123

West Profile A total of 32 samples were analyzed from the stratigraphic section of the west profile (Fig. 7.3). Samples from stratigraphic levels 1–3, corresponding to the LN (samples I–III in the south profile), were not collected. Therefore, sampling started from level 4 (sample IV). Samples IV–VIII, attributed to the MN levels 4–13, are very poor in phytoliths, and most should be considered sterile (see Table 7.2). The samples from the EN II layers (stratigraphic levels 14a–28+29 corresponding to samples IX–XIII) have good frequencies of phytoliths, with assemblages composed of phytoliths from grass culm, leaves, and inflorescence. Silica skeletons identified as wheat type, barley type, and millet type are common (e.g., samples IX and Xa). Silica skeletons from dicotyledonous plants are constantly present in these samples (Table 7.2; Fig. 7.4). The deposits from the late EN I period (samples XIV–XXVI) show some variability in the concentration of phytoliths, with five field-counts varying from 23 in sample XXVI to 663 in sample XXII (see Table 7.2). The phytolith assemblages from this period may be arranged in three groups (Fig. 7.4): one group in which assemblages are composed mainly (more than 65% of phytoliths) of dicotyledon silica skeletons (samples XIV–XVII, level 30); a second group in which assemblages still have dicotyledons, but their frequency is 5% or less, and frequencies of grass silica skeletons are often more than 10% (samples XVIII–XXIVb, levels 31 and 32); and finally, a third group in which grass phytoliths are still a majority, although dicotyledonous plant phytoliths are more than 5% but less than 35% of the total composition (samples XXV and XXVI, levels 33 and 34). The samples from the deepest EN I levels (sample XXVII–XXIX, levels 35–37) and the

Aceramic levels (samples XXX and XXXI, levels 38 and 39) show characteristics similar to the third group of the late EN I, but they do not have any silica skeletons. Also, they are generally very poor in phytoliths. Grass short cells in the west profile have a frequency of C3 types varying between 55% and 77% in a 200 short cell count (Fig. 7.5).

Figure 7.3. West profile stratigraphy and sampling. Photo N. Efstratiou.

MARCO MADELLA

124

West Profile

MN

Level Sample Grass single morphotypes

EN I Late 

EN II 

4

9

10

12

13

14a

16

20

21

23

IV

V

VI

VII

VIII

IX

Xa

Xb

Xc

XI

Grass long cells, spiny edges

8

3

9

7

7

59

Grass long cells, smooth or slightly wavy edges

2

6

4

9

3

26+27 28+29

30

XII

XIII

XIV

XV

XVI

XVII

119 139 147 128

121

63

12

18

23

25

72

84

87

74

17

21

16

20

134 141 123

Grass articulated morphotypes (silica skeletons) Wheat type (Triticum sp.), inflorescence

0

0

0

0

0

15

2

7

6

2

2

4

0

1

2

1

Barley type (Hordeum sp.), inflorescence

0

0

0

0

0

11

3

0

3

0

2

4

0

0

0

1

Millet type (e.g., Panicum sp., Setaria sp.), inflorescence

0

0

0

0

0

7

22

2

4

4

12

7

0

0

0

0

Grass inflorescence

0

0

3

2

0

89

34

48

61

52

31

85

3

1

4

2

Grass culm/leaf

0

0

3

3

0

71

16

51

60

49

14

81

8

2

3

2

0

0

0

0

0

14

5

3

5

2

5

6

123

98

102

76

10

9

19

21

10

338 285 384 427 360

274

324

163

141

150

127

Dicotyledons Total

e

0 10

Dicotyledons

Phase E

0 10

G s ras in kele s s flo to ili re ns ca sc , en ce G sk ras cu ele s s lm to ilic /le ns a af ,

yp ille

tt

ty M

rle y

Grass long cells, culm/leaf

Ba

Grass long cells, inflorescence

W

0

he

at

ty

pe

pe

Table 7.2. Knossos 1997: west profile phytolith counts.

Wall

Phase C

300

400

Phase B

Depth in cm

200

Phase D

100

500

Phase A

600

700 % 0 10 20 30 40 50 60 70

0 10 20 30 40 50 60 0 10

0 10 20 0 10 20 0 10 20 30 40 50 60 70

Percentages

Figure 7.4. Bar chart of phytolith percentage frequencies from the west profile.

125

PLANT ECONOMY AND THE USE OF SPACE: EVIDENCE FROM THE OPAL PHYTOLITHS

EN I Late, cont. 31

EN I

32

Aceramic

33

34

35

36

37

38

39

XVIII

XIX

XXa

XXb

XXc

XXIa

XXII

XXIII

XXIVb

XXV

XXVI

XXVII

XXVIII

XXIX

XXX

XXXI

36

53

150

148

139

129

298

168

142

13

6

7

5

4

2

2

20

46

72

71

62

72

235

144

62

24

9

9

6

3

5

9

0

2

5

13

8

11

5

9

14

0

0

0

0

0

0

0

2

0

2

12

8

5

5

11

18

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

34

55

106

57

67

53

44

108

3

0

0

0

0

0

0

2

29

58

72

52

71

67

39

68

4

0

0

0

0

0

0

1

9

12

11

9

6

0

5

10

5

8

3

4

7

4

6

61

173

354

433

335

361

663

420

422

49

23

19

15

14

11

17

100

90

C3 C4 80

70

Percentage (%)

60

50

40

30

20

10

XXX

XXXI

XXIX

XXVII

XXVIII

XXV

XXVI

XXIVb

XXII

Sample Numbers

Figure 7.5. Bar chart of C3 and C4 phytolith percentage frequencies from the west profile.

XXIII

XXc

XXIa

XXb

XIX

XXa

XVIII

XVI

XVII

XV

XIV

XIII

XI

XII

Xc

Xb

IX

Xa

VIII

VI

VII

V

IV

0

126

MARCO MADELLA

Interpreting the Phytolith Evidence from Neolithic Knossos The analysis of the phytoliths from the Neolithic layers under the Central Court raises quite a few interesting points of discussion in terms of plant economy and the use of space at the settlement. The assemblages of phytoliths deposited in the sediments of the more than 7 m long sequence may be arranged in five distinct phytolith phases (A–E, from bottom to top; Fig. 7.6). Phases A–C are

composed of samples deriving only from the west profile, phase D is based on samples from both south and west profiles, and phase E is based only on samples from the south profile (see Tables 7.1, 7.2). These groups tend to be internally consistent in terms of phytolith typology and composition, reflecting similar plant input.

Phase A Phase A includes phytolith samples XXXI– XXV (stratigraphic levels 39–33), which are attributed to the Aceramic and the earlier part of the EN I period. Although the sediments of this phase are very poor in phytoliths (almost sterile), some interpretations may be put forward for this early material. The assemblages seem to have originated from wild grasses, nonherbaceous plants, and, in much smaller quantity, cereals. The nonherbaceous plants component comes from the wood and leaves of dicotyledonous taxa (Figs. 7.4, 7.6). Dicotyledons represent between 10% and 50% of the total assemblages of this phytolith phase. Considering that the production of silica in woody plants is in general rather low (Albert and Weiner 2001; Carnelli, Madella, and Theurillat 2001; Tsartsidou et al. 2007), these phytolith frequencies in the sediments are high and reflect a very strong use of resources like wood and leaves from the natural vegetation surrounding the site. Ash from wood and dicotyledon leaves is indeed characteristic for its component in phytolith morphologies that cannot be attributed to plant genus or species (Albert et al. 1999; Carnelli, Madella, and Theurillat 2001; Tsartsidou et al. 2007), like the ones observed in these sediments. The presence of wild grass phytoliths also supports the hypothesis that during the seventh millennium there was a wide exploitation of the available wild plant resources. The exploitation of trees is also attested in the archaeobotanical data from the same period (see Sarpaki, this vol., Ch. 5). In the case of

phytoliths, remains of wood and leaves of trees and bushes may indicate the use of these resources as fuel and/or fodder. None of the sediments from this phase yielded any silica skeletons from cereals. The lack of cereal silica skeletons and the rather scanty presence of phytoliths (Table 7.2) in the sediments dating from the Aceramic to the EN I seem to suggest that these assemblages had a strong input from noncrop plants. The absence of silica skeletons may be related to preservation and/or taphon­ omy, however. As the archaeobotanical analysis shows that the people at Neolithic Knossos were cultivating a full “agricultural package” since the founding of the settlement, it is likely that during the Aceramic and initial EN I periods the excavated area was not the focus of any cereal crop processing, or that in general the intensity of cereal processing in the settlement was minimal. This hypothesis seems also to be supported by the low number of grass phytoliths from inflorescence and the extremely low concentration of phytoliths (Table 7.2). Most of the phytoliths are from grass culm and dicotyledonous plants, which might indicate the ephemeral use of vegetal material only for fires or other domestic activities (e.g., ash scatters from hearth cleaning or fodder). The low frequency of phytoliths might also attest to a scanty occupation of the excavated part of the settlement during the Aceramic period. The small excavated area, however, does not allow for an easy generalization. Additional samples from other areas of the

PLANT ECONOMY AND THE USE OF SPACE: EVIDENCE FROM THE OPAL PHYTOLITHS

127

0

0.5

Section of southwest corner of trench WEST FACE

1 1

LN

2

Phytolith Phase E • No dicotyledons • Grass single cells • Few grass silica skeletons (millet type)

3 MN

4

2

9 10 12 13 14a wall 1 (EN II)

3 EN II

16

Phytolith Phase D • Few dicotyledons • Grass single cells • Grass silica skeletons • Millet- and wheat-type phytoliths

20 21 23 27+26

4

29+28 30

Phytolith Phase C • Many dicotyledons • Few grass single cells and silica skeletons • Wheat-type phytoliths

31

5 NOT EXCAVATED

EN I (late)

32

Phytolith Phase B • High frequency of phytoliths • Few dicotyledons • Grass single cells • Grass silica skeletons • Wheat-type and barley-type phytoliths

6

33 34

7 EN I

35

ACERAMIC

36 37

Phytolith Phase A • Low frequency of phytoliths • Some dicotyledons • Grass single cells • No grass silica skeletons

38 8

39 meters

BEDROCK

Figure 7.6. Trench section (southwest corner to west face) with phases identified according to phytolith composition and frequencies.

128

MARCO MADELLA

settlement would be needed to determine whether the assemblages were the result of deposition

during a period of low density occupation or if they were related to intrasite space utilization.

Phase B Phase B includes samples XXIVb–XXa (level 32) and sample XIX (level 31), which are attributed to the late EN I period. Phase B is characterized by sediments with high concentrations of phytoliths and phytolith assemblages dominated by grass typologies (Figs. 7.4, 7.6, 7.7). Silica skeletons from wheat (Fig. 7.8) and barley were recovered for the first time in level 32 (sample XXIVb). Silica skeletons from wheat and barley husk attain the highest frequency in this phase, with a maximum of about 5% of the total assemblage in sample XXIII. In general the assemblages show high frequencies of both grass long cells and silica skeletons from inflorescence, mainly dendritic forms from cereals. It seems that these phytolith assemblages represent a deposition originating from the

continued and significant processing of cultivated cereals in the compound. According to the phytolith evidence, both the early stages of grain processing (in the form of hay) and the late stages of pounding, winnowing, and sieving (in the form of husk remains) seem to be represented in these samples (see Hillman 1981 for a related discussion). The differences in the phytolith assemblages from Phase B could be interpreted as the result of a change in the use of space in the settlement or of changes in plant utilization strategies. The two principal crops identified from the anatomical characteristics of the silica skeletons (articulated phytoliths) are Triticum sp. and Hordeum sp., confirming the results of the archaeobotanical analysis (see Sarpaki, this vol., Ch. 5).

Phase C Phase C includes phytolith samples XVII–XIV from stratigraphic level 30, assigned to the end of the EN I period. The samples are grouped in a different phytolith phase because dicotyledon phytoliths gain importance once again and account for up to 75% of the assemblages (Figs. 7.4, 7.6, 7.9). The frequency of phytoliths in these sediments, however, is lower than in the previous phase (Table 7.2). The Phase C assemblages show some similarities in composition with those of Phase A (samples XXV–XXXI, stratigraphic levels 33–39, Aceramic and beginning of EN I). In comparison to the Phase A samples, however, the Phase C assemblages also show the presence of silica skeletons from wheat, unidentified cereal crops (grass silica skeletons—inflorescence), and grass culm/ leaf. The phytolith composition of these samples seems to be related partially to activities of crop processing and partially to the use of plants for fire

or fodder. Only a small number of silica skeletons from wheat were identified in this phase. The composition of Phase C phytolith assemblages may be interpreted in terms of either a fluctuation in plant exploitation or a change in fuel choice. To the wood component are here associated the by-products of crop processing. Changes in fuel choice marked by the introduction of nonwood material may sometimes occur when a depletion of the wood reserves in the vicinity of the settlement forces people to look for alternative sources of fuel. The wood charcoal analysis, however, does not seem to indicate a depletion of forest resources (see Badal and Ntinou, this vol., Ch. 6). Nevertheless, it is possible that a diversification in the use of the available plant resources for fuel took place during this period with the employment of both wild resources and the by-products of crop processing.

129

PLANT ECONOMY AND THE USE OF SPACE: EVIDENCE FROM THE OPAL PHYTOLITHS

50 microns

50 microns

Figure 7.7. Silica skeleton from grass leaf (long cells and a stoma) from the EN I deposits (sample XXa, level 32). Photo M. Madella.

Figure 7.8. Wheat-type silica skeleton from the EN I deposits (sample XXIVb, level 32). Photo M. Madella.

50 microns

50 microns

Figure 7.9 Silica skeleton from a dicotyledonous plant from the EN I deposits (sample XIV, level 30). Photo M. Madella.

Figure 7.10. Millet-type silica skeleton from the EN II deposits (sample Xa, level 16). Photo M. Madella.

Phase D The Phase D samples come from both the south and west profiles. They comprise samples from the EN II stratigraphic levels 28+29–14a (samples XIII–IX) and the MN stratigraphic levels 13–4 (samples VIII–IV). They are characterized by cereal phytoliths, especially millet and wheat (Figs. 7.1, 7.4, 7.6). Most of the assemblages in this phase have a significant amount of the millet-type phytoliths, specifically in the west profile samples (e.g., samples Xa and XIII; Fig. 7.10). Under the category

of millet type, two genera are considered here, Panicum and Setaria; these cannot always be separated on the basis of the morphological characteristics of glume phytoliths. Panicum sp. glume phytoliths may have an interlocking pattern between long cells with very square edges or very irregularly wavy edges. The latter morphology is also characteristic of the Setaria sp. husk epidermis, and it is the only pattern observed in the millet (sensu lato) silica skeletons from Knossos. It is

130

MARCO MADELLA

possible to consider the millet-type silica skeletons from the settlement as belonging to Setaria sp. The cultivated species is Setaria italica (L.) P. Beauv., and the wild progenitor is S. viridis (L.) P. Beauv., a common summer weed widely spread across Eurasia (De Wet, Oestry-Stidd, and Cubero 1979; Zohary and Hopf 2000, 86). In Europe, S. italica first appears in the Bronze Age settlements of central Europe (Netolitzky 1914) and France (Hopf 1985) during the second millennium. In Greece it is reported at the Late Bronze Age site of Kastanas in Macedonia (Kroll 1983). The earliest evidence from the Near East for the cultivation of S. italica comes from the site of Tille Höyük, an Iron Age settlement in southeast Turkey (Nesbitt and Summers 1988). The available archaeological evidence indicates that S. italica is a relatively old domesticate, one which does not, however, belong to

the Neolithic Near Eastern agricultural package. The presence of Setaria silica skeletons at Knossos is probably attributable to the exploitation of wild grasses, of which S. viridis was one species, perhaps used as fodder. Millet-type phytoliths are constantly present in the EN II sediments, sometimes with frequencies accounting for more than 5% (e.g., sample Xa in the south profile or Xa in the west profile). Phase D samples VIII–IV from the south profile and samples VIII–IV from the west profile have phytolith assemblages with different compositions and different overall frequencies. The south profile samples are richer in phytoliths than the samples from the west profile. This variability might be due to a different use of the space, resulting in areas that experienced higher rates of phytolith deposition (e.g., areas dedicated to the processing of crops), while others received a lesser input.

Phase E Phase E samples (IIIam–I) came from deposits in the south profile only, from the LN stratigraphic levels 3–1 (Figs. 7.1, 7.6). These sediments are poor in phytoliths and have very few cereal silica skeletons (some millet types in sample II and IIIam, levels 2 and 3). The assemblages from this phase are dominated by single elements from grass inflorescence and culm/leaf. None of these phytoliths could be identified as definitely originating from cereals because the anatomical attributes of nonarticulated typologies are not always sufficient to identify the taxa. Some of the grass inflorescence phytoliths, however, belong to the

dendritic type, which should be produced mainly in cereals. This phase, characterized by grass single phytoliths, might represent the moment when the excavation area, which later constituted the Central Court of the palace, came to be used for specific functions no longer linked to daily activities of crop processing and fuel utilization. The almost total absence of silica skeletons (Table 7.1; Fig. 7.1) could support the hypothesis that the area became an open space devoted to social activities. Such a use of the area might have affected the preservation of any articulated phytoliths, as in an area of frequent trampling.

Environmental Remarks The environmental evidence from the phytolith assemblages from Neolithic Knossos is biased because of the predominantly anthropic origin (e.g., from cultivated crops) of the samples. The short cells from C3 grasses dominate the deposit, and they probably originate from the cereal input (wheat and barley are C3 plants). Plants with a C4 photosynthetic pathway (like millets) contributed to the phytolith assemblage, but the ratio between the

two groups is rather constant. Even when the origin of phytoliths is more related to wild plant input, the proportion of C3 versus C4 phytoliths does not change. Because human activities in the settlement mask the possible natural vegetation phytolith input, future phytolith studies should focus on the pedological sequences from around the site in order to clarify the issue of environmental change during the earliest phases of the Knossos settlement.

PLANT ECONOMY AND THE USE OF SPACE: EVIDENCE FROM THE OPAL PHYTOLITHS

131

Conclusion The analysis of the phytolith assemblages from the Neolithic deposits of Knossos has revealed two major points relevant to the interpretation of the settlement’s history and plant economy. First, the differences in the composition of the phytolith assemblages might be explained as changes in the importance of different crops during the time span from the Aceramic to the LN period, when changes in crop exploitation and in the spectrum of exploitation (cultivated versus wild resources) occurred. The rather constant proportion of wheat and barley phytoliths in most of the deposits (apart from the Aceramic, the initial EN I, and the LN), however, coupled with the strong variability of the wild plant remains (e.g., the Setaria and dicotyledon phytoliths), seems to suggest that the excavation area experienced varying amplitudes of activities in different periods, with emphasis shifting from cultivated to wild resources and vice versa, superimposed on a rather

regular influx from the major cereal crops (wheat and barley). Second, the variability of phytolith composition in time and sometimes in space may also be explained in terms of the changes in the use of space of the early Knossos settlement. Periods of partial abandonment and shifting activities (e.g., the construction of the wall during the late EN II) indicate continuous rearrangements that are mirrored in the phytolith input (e.g., the early EN II levels with mainly dicotyledon phytoliths). Phytolith assemblages thus represent phases of qualitative and quantitative differences in the exploitation of subsistence resources. The varying inputs may have been related to phases of adaptation and consolidation within the new environment, in which the development of a local complex of plant exploitation (i.e., noncereal plants) was superimposed on the agricultural package brought from outside the island.

References Albert, R.M., A. Tsatskin, A. Ronen, O. Lavi, L. Estroff, S. Lev-Yadun, and S. Weiner. 1999. “Mode of Occupation of Tabun Cave, Mt. Carmel, Israel, during the Mousterian Period: A Study of the Sediments and Phytoliths,” JAS 26, pp. 1249–1260. Albert, R.M., and S. Weiner. 2001. “Study of Phytoliths in Prehistoric Ash Layers Using a Quantitative Approach,” in Phytoliths: Applications in Earth Sciences and Human History, J.D. Meunier and F. Coline, eds., Lisse, Netherlands, pp. 251–266. Broodbank, C. 1999. “Colonization and Configuration in the Insular Neolithic of the Aegean,” in Neolithic Society in Greece (Sheffield Studies in Aegean Archaeology 2), P. Halstead, ed., Sheffield, pp. 15–41. ———. 2006. “The Origins and Early Development of Mediterranean Maritime Activity,” JMA 19, pp. 199–230. Carnelli, A.L., M. Madella, and J.P. Theurillat. 2001. “Biogenic Silica Production in Selected Alpine Plant Species and Plant Communities,” Annals of Botany 87, pp. 425–434.

Cherry, J.F. 1990. “The First Colonization of the Mediterranean Islands: A Review of Recent Research,” JMA 3, pp. 145–221. Cherry, J.F., D.J.L. Bennet, and A. N. Wilson. 1991. “A Gazetteer of Neolithic and Bronze Age Civilization in the Aegean,” unpublished manuscript, Göteborg. De Wet, J.M.J., L.L. Oestry-Stidd, and J.I. Cubero. 1979. “Origins and Evolution of Foxtail Millets Setaria italica,” Journal d’Agriculture Tropicale et de Botanique Appliquée 26, pp. 53–64. Hansen, J.M. 1992. “Franchthi Cave and the Beginnings of Agriculture in Greece and the Aegean,” in Préhistoire de l’agriculture: Nouvelles approches expérimentales et ethnographiques (Monographie du CRA 6), P.C. Anderson-Gerfaud, ed., Paris, pp. 231–247. Harvey, E.L., and D.Q. Fuller. 2005. “Investigating Crop Processing Using Phytolith Analysis: The Example of Rice and Millets,” JAS 32, pp. 739–752. Hillman, G. 1981. “Reconstructing Crop Husbandry Practices from Charred Remains of Crops,” in

132

MARCO MADELLA

Farming Practice in British Pehistory, R. Mercer, ed., Edinburgh, pp. 123–162. Hopf, M. 1985. “Bronzezeitliche Sämereien aus Ouroux-Marnay, Dép. Saône-et-Loire,” JRGZM 32, pp. 255–264. Kroll, H. 1983. Kastanas: Ausgrabungen in einem Siedlungshügel der Bronze- und Eisenzeit Makedoniens 1975–79 (Prähistorische Archäologie in Südosteuropa 2–7), Berlin. Madella, M. 2001. “Understanding Archaeological Structures by Means of Phytolith Analysis: A Test from the Iron Age Site of Kilise Tepe, Turkey,” in Phytoliths: Applications in Earth Sciences and Hu man History, J.D. Meunier and F. Colin, eds., Lisse, pp. 173–182. ———. 2003. “Investigating Agriculture and Environment in South Asia: Present and Future Contributions from Opal Phytoliths,” in Indus Ethnobiology: New Perspectives from the Field, S.A. Weber and W.R. Belcher, eds., Lanham, Md., pp. 199–249. Madella, M., A.H. Powers-Jones, and M.K. Jones. 1998. “A Simple Method of Extraction of Opal Phytoliths from Sediments Using a Non-Toxic Heavy Liquid,” JAS 25, pp. 801–803. Manning, S.W. 1999. “Knossos and the Limits of Settlement Growth,” in meletemata: Studies in Aegean Archaeology Presented to Malcolm H. Wiener on the Occasion of His 65th Birthday (Aegeum 20), P. Betancourt, V. Karageorghis, R. Laffineur, and W.-D. Niemeier, eds., Liège, pp. 469–482. Nesbitt, M., and G.D. Summers. 1988. “Some Recent Discoveries of Millets (Panicum miliaceum and Setaria italica) at Excavations in Turkey and Iran,” AnatSt 38, pp. 85–97. Netolitzky, F. 1914. “Die Hirse aus alten Funden,” Sitzungsberichte Kaiserliche Akademie der Wissenschaft (Wien) mathematische-naturwissenschaftliche Klasse 73 (B), pp. 1–35.

Peltenburg, E., S. Colledge, P. Croft, A. Jackson, C. McCartney, and M.A. Murray. 2000. “Agro-Pastoralist Colonization of Cyprus in the 10th Millennium bp: Initial Assessments,” Antiquity 74, pp. 844–853. Rosen, A.M. 1993. “Phytolith Evidence for Early Cereal Exploitation in the Levant,” in Current Research in Phytolith Analysis: Applications in Archaeology and Paleoecology (MASCAP 10), D.M. Pearsall and D.R. Piperno, eds., Philadelphia, pp. 160–171. Rosen, A.M., and S. Weiner. 1994. “Identifying Ancient Irrigation: A New Method Using Opaline Phytoliths from Emmer Wheat,” JAS 21, pp. 132–135. Sampson, A. 2006. Προϊστορία του Αιγαίου, Athens. ———. 2008. “Conclusions,” in The Cave of the Cyclops: Mesolithic and Neolithic Networks in the Northern Aegean, Greece. Vol. I: Intra-Site Analysis, Local Industries, and Regional Site Distribution (Prehistory Monographs 21), A. Sampson, ed., Philadelphia, pp. 199–227. Tsartsidou, G., S. Lev-Yadun, R.M. Albert, A. MillerRosen, N. Efstratiou, and S. Weiner. 2007. “The Phytolith Archaeological Record: Strengths and Weaknesses Evaluated Based on a Quantitative Modern Reference Collection from Greece,” JAS 34, pp. 1262–1275. Twiss, P., E. Suess, and R.M. Smith. 1969. “Morphological Classification of Grass Phytoliths,” Proceedings of the Soil Science Society of America 33, pp. 105–109. van Andel, T.H., and C.N. Runnels. 1995. “The Earliest Farmers in Europe,” Antiquity 69, pp. 481–500. Zohary, D., and M. Hopf. 2000. Domestication of Plants in the Old World: The Origin and Spread of Cultivated Plants in West Asia, Europe, and the Nile Valley, 3rd ed., Oxford. Zois, A. 1973. Κρήτη—Ἐποχή τοῦ Λίθου, Athens.

8

The Knossos Fauna and the Beginning of the Neolithic in the Mediterranean Islands Manuel Pérez Ripoll

The island of Crete is situated in a key geograph­ ical zone for the understanding of the spread of the Neolithic from east to west. The fauna associated with the Neolithic deposits of the settlement of Knossos reveals that the domestication process was at an advanced stage when it was introduced on the island, as demonstrated both by the size of the bones and by the sex and age selection process. The economy of the site was mixed, and there was a close link between agriculture and animal husbandry. The homogeneity of the new subsistence regime present from the beginning of the Neolithic both on the island of Crete and in continental Greece is attested particularly by the similar composition in species and proportions of livestock. Cattle had little numerical importance initially. Sheep were more important than goats, and together they constituted the basis of livestock exploitation. Pigs had small numerical importance; their real value lay in their capacity to transform domestic waste and agricultural by-products into meat proteins. The dog was always present, guarding

the livestock and assisting the shepherd, and this is why it is part of the domestic animal “package” from the very beginning of the Neolithic. The economy of the Neolithic communities was characterized not only by the use of domestic fauna but also by the exploitation of wild animals. Con­ tinental sites have a bone record of wild species comprising deer, roe deer, wild goats, boars, and a series of medium- and small-sized carnivorous animals such as foxes, badgers, martens, and wild cats. These species did not exist on the islands, and they had to be imported from a continent in order to populate the mountains and unproductive zones. This seems to have been a common feature of all Mediterranean islands, from Cyprus to Crete, Corsica, Sardinia, and the Balearic Islands. The islanders aimed to exploit these geographical zones to produce meat and skins without foregoing the satisfactions of hunting and sport. The limited excavation at Knossos in 1997 offered a unique chance to reexamine the Neolithic fauna of the site. Although only a small area was

134

MANUEL PÉREZ RIPOLL

exposed, it was possible to study the faunal remains and to compare them with the material from the older excavations (Jarman and Jarman 1968; see also the recent work of Isaakidou 2004 on J.D. Evans’s material). The following abbreviations are used here: Ad AN ant Ap b Bcrown Bd BD BDd Beta BG BmD Bp BPC BT c cal. C.h. cm comp. d Dd Dl Dm DLS DPA dt. dt.fg. EN

minimum distal width Aceramic Neolithic anterior minimum proximal width burned bone breadth at the base of the crown breadth of the distal end greatest breadth of the diaphysis breadth of the distal diaphysis Beta Analytic lab code breadth of the glenoid cavity breadth in the middle of the diaphysis breadth of the proximal end breadth of the coronoid processus breadth of the trochlea lithic cut calendar years Capra hircus centimeters complete digested by dogs depth of the distal end depth of the lateral half medial depth of the distal end diagonal length of the sole shortest distance from the processus an­co-­ naeus to the causal border of the ulna distal part distal fragment Early Neolithic

f fr GBp GLl GLm GLP Glpe gn LA Lbcrown Ld LG LM LN LO m mand. mm MM MN NISP NM No. O.a. post px. px.fg. qy SD SDO sh. sh.fg. SLC vy y

female anthropic fractured bone greatest breadth proximal greatest length of the lateral half greatest length of the medial half greatest length of the processus articularis maximum peripheral longitude gnawed length of the acetabulum Antero-posterior length at the base of the crown length of the dorsal surface length of the glenoid cavity Late Minoan Late Neolithic length of the olecranon male mandible millimeter Middle Minoan Middle Neolithic number of individual specimens present number of mandibles number Ovis aries posterior proximal part proximal fragment quite young minimum width of diaphysis smallest depth of the olecranon complete shaft diameters shaft fragment smallest length of the Collum scapulae very young young

Methodology and General Results The total number of bone remains studied from the 1997 excavation of Neolithic deposits under the Central Court of the Knossos palace was 3,955, of which 2,414 bones were identified and 1,541 were not. The high number of identified specimens is due to the good preservation of the bones, especially those from the Middle Neolithic (MN) period and later (Table 8.1). The determination of sheep versus goat was made following the guidance of Boessneck, Müller, and Teichert (1964) and the use of the reference collection of the Laboratory of the Department of Archaeology and Prehistory, University of Valencia. The measurements (Tables

8.2–8.4) were taken in accordance with the directions of von den Driesch (1976). Age has been calculated as indicated in the sections below. Bones have been classified as articular parts, fragments of articular parts, fragments of diaphyses, diaphyses, and whole bones. For each bone, the approximate age is indicated according to the fusion state of the epiphyses and the growth development of the bone (Silver 1969; Zeder 2005). The different marks on each bone have been studied in order to establish the taphonomy of the bone assemblage. The domesticated fauna consists of cattle, sheep, goats, pigs, and dogs. These taxa are present in the

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 135

Identified Specimens

EN I

EN II

EN/MN

MN

LN

TOTAL

NISP

%

NISP

%

NISP

%

NISP

%

NISP

%

Bos taurus

10

11.76

12

11

105

50.9

141

33.2

578

36.41

846

Ovis-Capra

61

71.76

80

73.39

65

31.55

228

53.77

673

42.32

1,107

Ovis aries

6

7.05

2

1.83

6

2.91

6

1.41

98

6.16

118

Capra hircus

—

—

—

—

2

0.97

2

0.47

14

0.88

18

Ovis–Capra hircus– Capra aegagrus

—

—

—

—

—

—

3

0.7

29

1.82

32

Capra aegagrus

—

—

—

—

—

—

1

0.23

11

0.69

12

Sus scrofa domesticus

7

8.23

15

13.76

26

12.62

40

9.43

179

11.25

267

Sus scrofa ferus

—

—

—

—

1

0.48

—

—

1

0.06

2

Sus scrofa ferus(?)/ Sus scrofa domesticus

—

—

—

—

1

0.48

1

0.23

1

0.06

3

Canis familiaris

1

1.17

—

—

—

—

—

—

4

0.24

5

Martes foina

—

—

—

—

—

—

—

—

1

0.06

1

Meles meles

TOTAL

—

—

—

—

—

—

2

0.47

1

0.06

3

85

100

109

100

206

100

424

100

1,590

100

2,414

Unidentified Specimens Bos Ovis-Capra-Sus

TOTAL

5

17

28

29

183

262

114

63

34

49

1,019

1,279

119

80

62

78

1,202

1,541

Table 8.1. Number of identified and unidentified specimens by taxa and period.

Bos taurus M3 mandible Level Length

Scapula

Radius

6

8

9

Level

4

21

Level

36.2

38

37

SLC

46.8

—

Bp

GLP

60.5

65.2

LG

54.5

—

BG

45.5

—

Metacarpus

9 71.2

Metatarsus

Level

2

3

8

8

12

14

Level

Bp

—

—

—

—

—

—

Bd

Bd

52.5

58.3

54.8

59.5

62.5

54.8

Pelvis

14

15

23

24

50.3

53.2

62.5

73

3

3 post

Tibia

Level LA

9

14

15

16

19

24

64.3

67

59

62

73.5

61.2

Bd

2

2

2

3

3

ant

post

ant

post

post

post

post

post

62

61.8

58

63.5

66

59.2

60.6

60.3

—

30.5

34.5

36.5

30.5

27

—

28.2

27.2

Astragalus

Level

4

14

66.8

66.1

3

3

Phalanx I

Level

19

22

Level

GLl

65

70.2

GLm

—

62.5

Glpe

Dl

36.2

38.5

Bp

32.5

Bd

—

48.2

SD

26

25

—

31

25.5

24

22.3

23

—

Bd

—

30.5

—

33

32

26.2

26.3

26

—

Table 8.2. Measurements of bones from Bos taurus according to the methodology of von den Driesch (1976).

MANUEL PÉREZ RIPOLL

136 Bos taurus Phalanx I Level

Glpe

Phalanx II 4

4

6

8

9

9

10

ant 58

16

post

post

ant

post

ant

ant

post

58

58.5

54

54

—

56

66.6

Level

2

Glpe

3

3

ant

ant

post

43

39.8

44.5

Bp

—

26.5

26.2

26

28

—

30

32.6

Bp

33

30

31.5

SD

—

21.5

21.2

22.2

22.5

—

26.2

26.8

SD

27.8

24.5

26.5

Bd

—

24.8

24.8

—

26.5

28.2

29

28

Bd

29.2

25

26.1

Phalanx II, cont. Level

3

3

3

3

3

3

4

4

4

7

9

9

10

post

post

post

post

post

post

ant

post

post

ant

ant

post

post

Glpe

43.1

40.3

39

37.5

35

42

41.6

40.5

39

40

43.5

40.5

39.5

Bp

29.8

27.5

26

25.5

23.5

31

30

26.8

27.2

27.5

34.5

—

25.6

SD

25

22.7

21.2

20.5

19

24.5

25

22.6

22

22.8

28.8

—

—

Bd

25.6

22.8

22.5

21.5

20.5

26.1

25

22.5

22.5

—

30.6

—

—

15

15

16

23

24

Level

3

3

3

6

ant

DLS

71.6

63.5

62

76.5

Ld

52.1

51

—

58.2

25

23

—

24

10

10

10

14

14

Phalanx II, cont.

Phalanx III

Level

14

14

ant

post

ant

post

post

post

Glpe

45.6

37

40.2

44

43.5

40.6

Bp

35.3

25

27.5

30.8

33.6

27.5

30.7

SD

29.8

20.5

22

25.2

28

22.8

25.5

Bd

33.5

22

25

26.8

27.8

23.2

26.5

10

10

10

10

10

10

GBp

Phalanx III, cont. Level

6

9

DLS

73.5

75

65

69.5

74.1

68.7

69

70.1

80

88

93.3

68.2

62.2

Ld

53.2

—

52.8

53

56.2

50.5

55

53

67.2

—

67

55

48.8

24

24.2

24.3

25

22.7

22.2

25

26

30.2

30.7

31

24

24.1

GBp

Phalanx III, cont.  Level

14

15

22

22

22

22

23

24

DLS

64

—

—

69

—

—

—

82

Ld

53.2

—

—

55.2

—

—

—

 —

GBp

21.8

30.4

24

25

24.5

27.2

24

31.1

Table 8.2, cont. Measurements of bones from Bos taurus according to the methodology of von den Driesch (1976).

Ovis aries, Capra hircus Scapula Species

C.h.

O.a.

O.a.

O.a.

O.a.

O.a.

O.a.

O.a.

O.a.

O.a.

O.a.

O.a.

C.h.

Level

2

3

3

3

3

4

4

9

9

9

9

16

9

SLC

—

17

17.2

18

14

16

—

16.8

17

18.5

17.8

16.8

19.3

GLP

33.5

27.5

26.1

29

—

—

26.5

—

27.5

26

28

—

28.2

LG

23

—

—

—

—

—

—

—

—

—

—

21.8

26

BG

17.2

18.2

19.2

21

—

17.5

17

17

17.2

17.5

—

18

20.4

Table 8.3. Measurements of bones from Ovis aries (O.a.) and Capra hircus (C.h. ) according to the methodology of von den Driesch (1976).

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 137 Ovis aries, Capra hircus Humerus O.a.

O.a.

O.a.

O.a.

O.a.

O.a.

O.a.

C.h.

O.a.

O.a.

C.h.

O.a.

3

3

4

4

4

9

10

10

15

15

19

24

Bd

25.5

27.3

26

26

25.5

28.3

26

31

24

25

25.5

24

Dld

22.5

24.3

23.5

23.3

21.3

24.7

22.5

—

22.3

21.5

—

21

BT

25.2

26

24.4

24.6

24.3

27.5

24.5

29

23.5

24.5

24.8

23.2

Species Level

Radius

Ulna O.a.

O.a.

C.h.

O.a.

O.a.

C.h.

O.a.

Level

3

3

3

8

8

8

14

Bp

—

23

—

27.5

—

—

28

BmD

—

—

—

15

14.2

18.5

—

 

Bd

23

—

30

—

—

—

—

 

Species

O.a.

O.a.

Level

9

9

DPA

24.4

23

Species

Metacarpus

Femur  C.h.

C.h.

O.a.

O.a.

C.h.

O.a.

O.a.

C.h.

O.a.

Level

2

3

3

3

8

8

10

12

21

Level

Bp

—

22.5

—

—

22.8

—

19

—

—

Bd

BmD

—

—

—

—

16

—

—

—

—

Bd

25

—

21.3

20.3

—

22.2

—

25

24.5

O.a.

O.a.

O.a.

O.a.

O.a.

C.hi.

O.a.

O.a.

O.a.

O.a.

O.a.

O.a.

O.a.

3

3

3

4

4

4

6

8

9

9

9

9

10

Bp

35.5

—

—

—

—

—

—

—

—

—

—

—

—

Bd

—

23

23

23.8

20.5

23

22

23.2

20.8

24

22

21.5

21

BDd

—

16.7

—

18.5

—

19

17.2

16.5

16.8

18.5

15.5

16.6

—

Species

O.a.

Species 

19 32.8

Tibia Species Level

Tibia, cont.

Metatarsus O.a.

O.a.

O.a.

C.h.

C.h.

Level

10

10

10

10

19

Bd

21

21.5

23.5

22.5

16.6

16.5

17.8

18

Species

BDd

Astragalus C.h.

O.a.

O.a.

Level

3

10

12

19

GLl

24

27

26

22.5

25

Dl

14

Bd

15

GLm

C.h.

C.h.

Level

7

10

10

21.8

Bp

16

18.1

—

18

BD

—

11.6

—

Bd

—

—

27.2

Phalanx I O.a.

Species

O.a.

Species

Phalanx II  O.a.

O.a.

O.a.

Level

2

9

19

25.2

Glpe

31.5

34

24.9

24.1

Bp

10.5

14.8

14.5

14.2

SD

17.1

16.5

16.3

Bd

Species

O.a.

O.a.

Level

4

9

31

Glpe

20

18

12.2

10

Bp

11.2

10.5

8

8.8

8

SD

8.5

7.5

10

—

9.9

Bd

9.2

8

Species 

Table 8.3, cont. Measurements of bones from Ovis aries ( O.a. ) and Capra hircus ( C.h. ) according to the methodology of von den Driesch (1976).

MANUEL PÉREZ RIPOLL

138 Sus scrofa domesticus Scapula

Humerus

Level

3

6

9

10

10

14

Level

2

10

SLC

24.3

21.8

—

—

22.2

—

Bd

35.5

39

GLP

—

32.5

31.2

35.8

—

37.5

Dld

38

39

LG

—

29

—

—

—

—

BT

26

28.2

BG

—

21.5

21.5

24.8

—

25.2

Level

4

9

14

14

DPA

33.3

37.5

33

36.5

Bd

29.8

SDO

—

—

25.6

—

Dd

26.7

LO

—

—

—

—

BPC

—

—

18.3

—

Ulna

Tibia Level

14

Sus scrofa ferus Radius

Ulna (Sus scrofa ferus?)

Level

20

SD

25.8

Level

Phalanx I (Sus scrofa ferus?) 23

Level

10

LO

61.8

Glpe

36.2

DPA

39.8

Bp

20.7

SDO

31.9

SD

16.3

Bd

18

Capra aegagrus Radius Level

Phalanx I

Phalanx II

3

3

9

Level

3

8

Level

9

Bp

35.5

—

34

Glpe

43

46.5

Glpe

29.5

SD

—

26.3

—

Bp

15.4

16

Bp

SD

13

13.2

SD

Bd

16

16.2

Bd

Martes

Meles meles

Humerus

Canine

15

Ulna

Level

9

Level

14

Level

Bd

16

Lbcrown

7.5

DPA

15.1

Dd

8.2

Abcrown

5.1

BPC

10.5

Table 8.4. Measurements of bones from Sus scrofa domesticus, Sus scrofa ferus, Capra aegagrus, Martes, and Meles meles according to the methodology of von den Driesch (1976).

Neolithic sites of continental Greece and characterize the whole Mediterranean Neolithic. The domestic character of the fauna has been established using the criteria of bone size, assessed by the measurements of the articular parts of long bones (humerus, radius, metacarpus, femur, tibia, and metatarsus) and small bones (e.g., phalanges, tarsi, molars), along with the age patterns obtained from the teeth and bones. The observed sizes for each species are the same as those studied at other Mediterranean Neolithic sites,

both in continental Greece and in the western Mediterranean. They are fully comparable with those of Argissa-Magula (Boessneck 1962), Cova de 1’Or, Cova de Cendres, and the Nerja Cave (Pérez Ripoll 1980, 1999). Therefore, the state of domestication is fully verified by the decrease in the bone measurements, which occurs due to the process of selection of the desired characteristics (Saña 1999; Pérez Ripoll 2001b). In contrast, metric measurements of cattle, goat, and sheep bones from Aceramic sites such as Shillourokambos and

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 139

Mylouthkia on the island of Cyprus are similar to those of the wild taxa in the continental zones of Anatolia and the Near East. Nevertheless, the fact that they were imported to Cyprus, the evidence of their mortality profiles, and the presence of all parts of the skeleton are all considerations that seem to be more consistent with husbandry than with hunting, leading to the conclusion that the animals were already domesticated (Peltenburg et al. 2000; Vigne 2000). Thus, while the animal bones from Shillourokambos and Mylouthkia reflect an

initial state of the domestication process, the bones from Knossos clearly belong to an advanced stage of the same process. The presence of wild animals at the site of Knos­ sos is interesting and intriguing. Species such as the marten and the badger were imported, as was the case in Cyprus, while the presence of others such as the wild goat (agrimi) and the boar may be explained in two ways: either they were imported in a wild state, or they became feral from the domesticated stock.

Taphonomic Studies Taphonomic studies are of great importance for explaining the state of the bone assemblages. Considerable insight may be gained from the

analysis of fragmentation patterns, gnawing by animals, and bone marks produced by humans.

Bone Fragmentation The bone assemblage of Knossos is quite fragmented, as we can observe from the long bones (humerus, radius, metacarpus, femur, tibia, and metatarsus). The only complete long bone of Bos is a metacarpus from the MN period (Fig. 8.1), and complete long bones of medium-sized mammals (goats, sheep, and pigs) are sparse (Fig. 8.2). Diaphysis fragments are predominant in both diagrams. The articular proximal parts of both Bos

and the medium-sized taxa are scarce compared to the articular distal parts, owing to their different bone densities and different resistance to destruction; distal parts are more resistant than proximal ones. Diaphyses (shafts) are relatively abundant in medium-sized mammals (Fig. 8.2). A large number of the articular part fragments and dia­ physis fragments were produced by the action of dogs; this is especially evident from the cylindrical

100

100

EN MN LN

80

60

60

40

40

20

20

0

0

Complete

px.

px.fg.

sh.

sh.fg.

dt.

dt fg.

Figure 8.1. Percentages of the osseous parts of cattle long bones.

EN MN LN

80

Complete

px.

px.fg.

sh.

sh.fg.

dt.

dt.fg.

Figure 8.2. Percentages of the osseous parts belonging to the long bones of middle-sized mammals (goats, sheep, and pigs).

MANUEL PÉREZ RIPOLL

140

shape of the diaphyses (Fig. 8.3). Therefore, this carnivorous animal played an important role in the final formation of the bone assemblage under study. Another action that causes fragmentation is trampling. Certain fragments could be joined together, and the morphology of their fractures differed from that associated with other activities of dogs or humans. These fractures must have been the result of people walking on top of the bones or of sediment pressure. Figure 8.3. Skeletal fragments of long bones of Ovis/Capra from level 14, all with dog-gnawing marks: two fragments of radius diaphysis, a fragment of tibia diaphysis, a metacarpus diaphysis, and a metatarsus diaphysis. Scale in cm. Photo M.P. Ripoll.

Gnawing of Bones by Dogs Dog bone remains are rare, but their presence is detected due to the alterations caused by gnawing and other forces in certain bones (Fig. 8.3). Such alteration is very common at Neolithic sites (Pérez Ripoll 1992; Bernabeu, Pérez Ripoll, and Martínez Valle 1999; Bernabeu, Barton, and Pérez Ripoll 2001). Dogs were most probably used for controlling livestock or protecting it from predators and possible thieves. Gnawed bones appear already at the beginning of the Early Neolithic (EN) period, indicating the transformations in bone exploitation strategies that took place in the Neolithic and signaling a break from the practices of hunter-gatherer societies. The number of bones with gnawing marks varies in the different taxa (Table 8.5). It is low in the case of Bos bones (between 8% and 13% of the total) and higher in Capra, Ovis, and Sus bones (between 18% and 55%). Moreover, significant variations have been observed in the stratigraphic sequence of Knossos; bones with gnawing marks

Species

EN

are more abundant in the EN levels (e.g., 55% of the bones of goats and sheep from the EN compared to 28% from the Late Neolithic [LN]). This disparity can be explained by the different conditions of preservation in the upper and lower layers of the tell; in the former, bones were altered more due to the effects of plant roots and other environmental factors, making the study of their marks difficult, while in the latter the bones are well preserved with the gnawing marks clearly defined. In addition, there is a noticeable occurrence of bone remains altered by dog gastric fluids. These are small fragments of various skeletal elements (skull, vertebra, humerus, metacarpus, talus, metatarsus, phalanx) that were gnawed, bitten, and then swallowed (Tables 8.6–8.9B). There are also some bones with rodent gnaw marks. This type of mark usually appears at sites with a farming economy; they are made by small rodents that are common in such agricultural communities. MN

LN

Total NISP

Gnawed

%

Total NISP

Gnawed

%

Total NISP

Gnawed

%

Bos

22

3

13.4

246

38

15.4

578

48

8.5

Ovis/Capra

149

83

55.7

313

64

20.4

825

234

28.3

Sus

22

11

50

69

17

24.6

181

34

18.7

Table 8.5. Number of identified specimens of Bos, Ovis/Capra, and Sus with number of marks caused by dog gnawing.

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 141

EN I Bone

Bos taurus NISP gn

Horn Skull Maxilla Mandible Teeth Hyoid Vertebrae Ribs Scapula Humerus comp. px. px.fg. sh. sh.fg. dt. dt.fg. Radius comp. px. px.fg. sh. sh.fg. dt. dt.fg. Ulna Carpal bones Metacarpus comp. px. px.fg. sh. sh.fg. dt. dt.fg. Pelvis Femur comp. px. px.fg. sh. sh.fg. dt. dt.fg. Patella Tibia comp. px. px.fg. sh. sh.fg. dt. dt.fg.

1 1 2 1 1 1

1 1

EN II

Ovis- Ovis Capra aries NISP

NISP

Sus scrofa Bos Canis domesticus taurus c b

gn

7 1 3 15

5

3 6 1

3(1d) 3 1

2

1 5

1 4

NISP

gn

NISP

NISP b 2

1 1 3

1

1

1

OvisCapra

Ovis aries

NISP

NISP

Sus scrofa domesticus b

4 6

gn

b

gn

2(1d)

3

1 10

1 1

14 4

8 3

1

5

4

2 1 7

1

NISP

3 3

2

1

1

1

1

1

1

1

1 1 7

1 1

1

1

1

1

1

1

2

2

1

1

1

2 1 1 1

2 1 1

1

1 8

4

1

6

3

2

3

3

2

1 11 1

1 7

1

1

1

Table 8.6. Early Neolithic I and EN II faunal remains: list of number of identified specimens, with remarks on the bone parts, taphonomic marks, and bone age.

MANUEL PÉREZ RIPOLL

142

EN I Bone

Bos taurus NISP gn

Fibula Talus Calcaneus Centrotarsale Os malleolar Other tarsals Metatarsus comp. px. px.fg. sh. sh.fg. dt. dt.fg. Phalanx I comp. px./px. fg. dt./dt.fg. sagittal part lateral part Phalanx II comp. px./px.fg. dt./dt.fg. Phalanx III Sesamoid TOTAL

EN II

Ovis- Ovis Capra aries NISP

NISP 1

Sus scrofa Bos Canis domesticus taurus c b 1

1

gn

NISP

gn

NISP

OvisCapra

Ovis aries

NISP

NISP

NISP b

1 1(d)

Sus scrofa domesticus b

gn

1

1

NISP

b

gn

1

1 2 1

1

2 1

1

1 3

10

3

61

6

3(3d)

1 1

37

7

2

1

1 12

1

80

2

1

46

1

1(d)

1

1(d)

14

9

Table 8.6, cont. Early Neolithic I and EN II faunal remains: list of number of identified specimens, with remarks on the bone parts, taphonomic marks, and bone age.

EN II/MN Bone

Bos taurus NISP

Horn Skull Maxilla Mandible Teeth Hyoid Vertebrae Ribs Scapula

b

gn

1 9

NISP

NISP

2 1 7 3

7 5 7 20 4

Sus scrofa domesticus

Ovis-Capra Ovis aries Capra hircus

1 10

3 16 1

1

NISP

b

gn

NISP

b

4

1

1

1 2

2 1

1 2 2

Sus scrofa ferus bn

NISP

1 1

Table 8.7. Early Neolithic II/MN faunal remains: list of number of identified specimens, with remarks on the bone parts, taphonomic marks, and bone age.

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 143

EN II/MN Bone

Bos taurus NISP

Humerus comp. px. px.fg. sh. sh.fg. dt. dt.fg. Radius comp. px. px.fg. sh. sh.fg. dt. dt.fg. Ulna Carpal bones Metacarpus comp. px. px.fg. sh. sh.fg. dt. dt.fg. Pelvis

b

3(3y)

Sus scrofa domesticus

Ovis-Capra Ovis aries Capra hircus gn

NISP

NISP

NISP

b

NISP

5 1 1

1

b

bn

NISP

1 2 5

1(y)

1

1 1

2

gn

Sus scrofa ferus

1 1

2

1

1 1

1 1

1 2 3

1

1(y)

2

1

1

1 2 2 8(3m-2f)

1 2

1 3 1 1(y) 2(1qy)

1 1 1

1

1

1 2(1y)

Femur comp. px. px.fg.

1

sh. sh.fg.

2

1

3

3

dt.

1

dt.fg.

2

1(y) 1

Patella Tibia comp.

1(y)

px. px.fg. sh. sh.fg.

3

1

1

2

1

1

dt. dt.fg.

1(y)

Fibula Talus

2(1y)

Calcaneus

1(y)

1 1

Table 8.7, cont. Early Neolithic II/MN faunal remains: list of number of identified specimens, with remarks on the bone parts, taphonomic marks, and bone age.

MANUEL PÉREZ RIPOLL

144

EN II/MN Bone

Bos taurus NISP

Centrotarsale

b

2

Sus scrofa domesticus

Ovis-Capra Ovis aries Capra hircus gn

NISP

NISP

NISP

b

gn

NISP

Sus scrofa ferus

b

bn

NISP

1

5

2

1

Os malleolar Other tarsals

2

Metatarsus comp. px. px.fg.

1

1

1

sh.

1

sh.fg.

1

dt.

2

1

1

1

1

dt.fg.

1

1

1

1

1(y)

22

26

Phalanx I comp.

2

px./px. fg. dt./dt.fg. sagittal part lateral part Phalanx II comp.

4

px./px.fg. dt./dt.fg. Phalanx III

6

Sesamoid TOTAL

105

4

18

65

6

2

2

Table 8.7, cont. Early Neolithic II/MN faunal remains: list of number of identified specimens, with remarks on the bone parts, taphonomic marks, and bone age.

MN Bone

NISP Horn Skull Maxilla Mandible Teeth Hyoid Vertebrae Ribs Scapula

OvisCapra

Bos taurus b

fr gn

1 5 7 9 1 5 42 3

NISP

Ovis Capra Ovis–Capra hircus– Capra aries hircus Capra aegagrus aegagrus NISP

NISP

NISP

NISP

Sus scrofa domesticus b

gn

6 1 4 11 2 9

3 26 5(1y)

2

1(d) 4 1

NISP

Gn

10 2 2 2

1

2 5 4(1qy)

1 1 1

Meles NISP

1 1

Table 8.8. Middle Neolithic faunal remains: list of number of identified specimens, with remarks on the bone parts, taphonomic marks, and bone age.

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 145

MN Bone

OvisCapra

Bos taurus NISP

b

Humerus comp. px. px.fg. 1 sh. sh.fg. 3 1 dt. 1(y) dt.fg. 1 Radius comp. px. px.fg. 2 sh. sh.fg. 3 dt. dt.fg. Ulna 2(y) Carpal bones 1 Metacarpus comp. 1(y) px. 1 px.fg. sh. sh.fg. 4 dt. 2 dt.fg. Pelvis 7(2m,1vy) Femur comp. px. px.fg. 2(1y) sh. sh.fg. 3 dt. dt.fg. 1(y) Patella 1 Tibia comp. px. px.fg. sh. 1(y) sh.fg. 8 dt. 2(1y) dt.fg. Fibula Talus Calcaneus 4(3y) Centrotarsale 1 Os malleolar Other tarsals Metatarsus

fr gn

NISP

Ovis Capra Ovis–Capra hircus– Capra aries hircus Capra aegagrus aegagrus NISP

NISP

NISP

NISP

Sus scrofa domesticus b

gn

1

4

NISP

Gn

3

3

Meles NISP

1(vy) 1 9 1

2

1

1

1 12 1(y) 2(y)

3

1 1 1

1

2(1y)

2(1vy) 2

1 1

1 1 1

1 2

11 1 1

4(2v)

1

5

1 4

1

3

1(y) 1(y) 2 5

1

2

1(vy) 1(y) 1 2 1

1

5 11 2(2y) 1

3

1 1

1

1 2

1

1

Table 8.8, cont. Middle Neolithic faunal remains: list of number of identified specimens, with remarks on the bone parts, taphonomic marks, and bone age.

MANUEL PÉREZ RIPOLL

146

MN Bone

NISP comp. px. px.fg. sh. sh.fg. dt. dt.fg. Phalanx I comp. px./px.fg. dt./dt.fg.

OvisCapra

Bos taurus b

fr gn

NISP

1(y)

Ovis Capra Ovis–Capra hircus– Capra aries hircus Capra aegagrus aegagrus NISP

NISP

NISP

NISP

Sus scrofa domesticus b

1 1 3

3 2

gn

Meles

NISP

Gn

1

1

41

12

NISP

1 1 2

1(y) 1(qy) 2

2 1

1(d)

sagittal part lateral part Phalanx II comp. px./px.fg. dt./dt.fg. Phalanx III Sesamoid TOTAL

4

4

1

141

1

9

20

228

6

2

3

1

5

42

2

Table 8.8, cont. Middle Neolithic faunal remains: list of number of identified specimens, with remarks on the bone parts, taphonomic marks, and bone age.

LN Bone

Bos taurus NISP

fr

c

gn

Ovis-Capra

Ovis aries

Capra hircus

Ovis– Capra hircus– Capra aegagrus

Capra aegagrus

NISP

NISP

NISP

NISP

NISP

Horn

2

1

4(1f)

Skull

28

16

1

Maxilla

b

fr

gn 2(1d)

7

Mandible

63

1

1

Teeth

56

55

Hyoid

4

1

Vertebrae

35

2

31

13

5 2

2

4(2d)

Table 8.9A. Late Neolithic faunal remains: list of number of identified specimens, with remarks on the bone parts, taphon­ omic marks, and bone age.

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 147

LN Bone

Bos taurus NISP

Ribs

95

Scapula

16

fr

c

gn 23

Ovis-Capra

Ovis aries

Capra hircus

Ovis– Capra hircus– Capra aegagrus

Capra aegagrus

NISP

NISP

NISP

NISP

NISP

11(3m,1f)

2

1

195 24(3f,1m)

b

fr

3

gn 1

1

4

Humerus comp. px. px.fg.

1(y) 2

1

23(2y)

8

sh. sh.fg.

7 1

dt.

7

30 6(1vy,1qy,1y)

4 8

26

2

9(1d)

dt.fg. Radius comp. px.

2(1f)

px.fg.

2(1m)

1

1

27

14

1

2

5

1

11

1

1

32

1

3

dt.

5

1

dt.fg.

2

sh. sh.fg.

Ulna

3

Carpal bones

6

2

12(5y)

1

1(y)

6 1

10 1

1(y)

23 2 1

12

1

1

6

1

Metacarpus comp.

3

3

2

2

4

px.

2

1

px.fg.

3

2

2 4

1

9

4

14

7

dt.

4(1y)

2

2(2y)

4(2f)

dt.fg.

2(1y)

1

2(1m)

2

23

10(5m,4f)

12

sh. sh.fg.

Pelvis

33(7f)

11

2(1m)

1 1

1

1

1 3(2d)

1

Femur comp. px. px.fg.

1(y)

3(1y)

sh. sh.fg.

3(1vy) 24

10

1

46

dt. dt.fg.

1 2

28

2(1y) 1(y)

3

3

Table 8.9A, cont. Late Neolithic faunal remains: list of number of identified specimens, with remarks on the bone parts, taphonomic marks, and bone age.

MANUEL PÉREZ RIPOLL

148

LN Bone

Bos taurus NISP

fr

c

gn

Patella

Ovis-Capra

Ovis aries

Capra hircus

Ovis– Capra hircus– Capra aegagrus

Capra aegagrus

NISP

NISP

NISP

NISP

NISP

b

fr

gn

1

17

1

Tibia comp. px. px.fg.

1(y)

1

2

1

36

10

2

sh. sh.fg. dt. dt.fg.

1 1(1vy) 18(4y)

1

14

42

3(y)

2(1y)

3

1

8 14

2

1 1

Fibula Talus

2

1

1

Calcaneus

4(3y)

1

Centrotarsale

1(y)

1

Os malleolar

1

Other tarsals

1

1

2(2d)

1

1(d)

Metatarsus comp.

1(vy)

px. px.fg.

6

4

sh.

1

2

2

16

1

2 1

6(1qy)

sh.fg.

10

2

19

dt.

3

2

2

dt.fg.

2

2

1

6

1

5

1

1

1

1

2

1

4(2y)

1

1

10 1 3(2d)

Phalanx I comp.

14(1y)

px./px.fg.

3

1

dt./dt.fg.

3

1

sagittal part

6

2

1(d)

1(d)

6

lateral part Phalanx II comp.

15

2

19

1

1

1

11

5

px./px.fg. dt./dt.fg. Phalanx III Sesamoid TOTAL

578

69

1

48

673

98

14

29

4

234

Table 8.9A, cont. Late Neolithic faunal remains: list of number of identified specimens, with remarks on the bone parts, taphonomic marks, and bone age.

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 149

LN Bone

Sus scrofa domesticus

Sus scrofa ferus

NISP

NISP

b

gn

Canis

Martes

Meles

NISP

NISP

NISP

Horn Skull

14

Maxilla

8

Mandible

27

Teeth

8

1

1 1

1

Hyoid Vertebrae

13

4

Ribs

19

1

Scapula

9

4

Humerus comp.

1(qy)

px. px.fg.

1(y)

sh. sh.fg.

6

dt.

1

1

5 1

dt.fg. Radius comp. px. px.fg.

1

1

5

1

3(1y)

2

sh. sh.fg. dt. dt.fg. Ulna

1(y) 5(1vy,2y)

3

1

1

Carpal bones Metacarpus comp.

3(3y)

px. px.fg. sh. sh.fg. dt.

1(y)

dt.fg. Pelvis

8

2

Femur comp. px.

3(2y)

px.fg.

Table 8.9B. Late Neolithic faunal remains: list of number of identified specimens, with remarks on the bone parts, taphonomic marks, and bone age.

MANUEL PÉREZ RIPOLL

150

LN Bone

Sus scrofa domesticus

Sus scrofa ferus

NISP

NISP

b

gn

Canis

Martes

Meles

NISP

NISP

NISP

1

1

Femur, cont. sh. sh.fg.

6

4

dt. dt.fg.

1(y)

Patella Tibia comp. px.

2(2vy) 1(y)

px.fg. sh. sh.fg.

1 16

4

dt. dt.fg. Fibula

1

Talus

4(2y)

Calcaneus Centrotarsale Os malleolar Other tarsals

1

Metatarsus comp. px.

3(2y) 1

px.fg. sh. sh.fg. dt. dt.fg. Phalanx I comp.

3(1y)

px./px.fg.

1(y)

1

dt./dt.fg. sagittal part lateral part Phalanx II comp.

2

2(d)

px./px.fg. dt./dt.fg. Phalanx III Sesamoid TOTAL

179

2

1

34

4

Table 8.9B, cont. Late Neolithic faunal remains: list of number of identified specimens, with remarks on the bone parts, taphonomic marks, and bone age.

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 151

Marks Caused by Humans The marks caused by humans include bone fractures for extracting marrow, lithic cuts produced during the butchering process, and marks caused by fire. The extraction of marrow depends on the quantity present in the bones of the different taxa. In general it is present in the bones of Bos, but it is almost nonexistent in the bones of Capra, Ovis, and Sus (Pérez Ripoll 1992; Bernabeu, Barton, and Pérez Ripoll 2001). At Knossos the absence of Bos bones with fracture marks in the EN levels may be due to the scarcity of the faunal remains. There are only four bones with fracture marks from the transitional levels (EN/MN), representing 18.1% of the total bones of Bos. There are nine bones from the MN levels, representing 6.3% of the total bones of Bos, and four bones from the LN levels, representing 11.9% (Fig. 8.4). Four LN bones from the medium-sized taxa (Capra, Ovis, and Sus) with fracture marks caused by the extraction of marrow may have served as dog food. These findings are similar to those recorded at other EN sites in the western Mediterranean (Pérez Ripoll 1992; Bernabeu, Pérez Ripoll, and Martínez Valle 1999; Bernabeu, Barton, and Pérez Ripoll 2001), where important alterations of bones caused mainly by dogs (and to a lesser extent by the human use of the marrow) are observed. Only bones of Bos were broken. This pattern of exploitation is very different from that employed by hunter-gatherers, who fractured all bones from large, medium, and small mammals, including the bones of rabbits (Pérez Ripoll 1992, 2001a, 2004, 2005–2006; Pérez Ripoll and Martínez Valle 2001), birds, and porcupines (Martínez Valle 2001). Indeed, the study of bone marks caused both by gnawing and by fractures for the extraction of marrow has helped to construct a reliable model for determining the domestic or wild status of a bone assemblage. In the case of Knossos, the existing marks indicate that there was little interest in the marrow, except in the case of Bos, which is consistent with the needs of a Neolithic community rather than of hunter-gatherers. Marks caused by lithic tools are almost nonexistent. They are seen only on a tibia diaphysis fragment of Bos belonging to the LN period

and a talus of goat/sheep from the EN. Generally speaking, marks from lithic cuts on bones of the Neolithic period are usually few in comparison with those seen on bones of the Paleolithic and Epipaleolithic. Usually the Neolithic diaphyses do not have defleshing marks, which are typical of the Paleolithic/Epipaleolithic bones. The most common marks in a Neolithic context are those of disarticulation (Pérez Ripoll 1992; Aura et al. 2002). The lack of lithic marks at Knossos may be explained by the high frequency of bone alteration in the middle and upper levels of the deposit, as discussed above; in the majority of the cases, their recognition and study is impossible. Marks caused by fire are also rare at Neolithic Knossos (Table 8.10). The low frequency of such marks contrasts with the high frequencies observed among the bones from Paleolithic and Epipaleolithic contexts, owing to the direct importance of a fire and the use of coals for the preparation of meat in earlier periods. Neolithic societies in general made use of pottery for cooking, and as a result bones did not come into direct contact with fire or coals. Fire marks on bone may also have resulted from the scattering of the content of hearths.

Figure 8.4. Fragments of proximal epiphyses of femur and tibia of Bos taurus with fracture marks caused by impacts from the extraction of marrow, level 24. Scale in cm. Photo M.P. Ripoll. EN

EN/MN

MN

LN

Bos

1

0

1

0

Ovis/Capra/Sus

3

3

5

5

Species

Table 8.10. Number of identified specimens of Bos and Ovis/ Capra/Sus with burn marks.

152

MANUEL PÉREZ RIPOLL

The Domesticated Species In the EN I and II periods at Knossos all domestic species are present (cattle, sheep, pig, and dog), although goat bones cannot be separated with certainty from sheep in these levels and are grouped as Ovis-Capra (Table 8.1). This difficulty may be explained by the small size of the faunal sample from the EN levels. There are 85 bones from the EN I levels, of which 66 belong to goat/sheep, while there are 109 bones from EN II levels, of which 82 come from goat/sheep (Table 8.1). It was also the case with the samples from Evans’s excavations that few bones of domestic goat were securely identified in the early levels (Jarman and Jarman 1968). In her recent work on Evans’s material, Isaakidou (2004) tried to separate sheep and goat bones, reaching the conclusion that goats are more or less steadily represented at ca. 10–20% of the total (Isaakidou 2004, 205, 224, table 6:16, fig. 6:27). The overall percentage occurrence of sheep/goat observed in the

present study is very similar to those found in the continental Neolithic sites of Greece and other places in the Mediterranean, as we shall see later on. Important changes are observed from the MN period onward, when the frequency of cattle rose from 11% to 33%–50% (Table 8.1). This increase might represent economic changes related especially to the use of cattle as a means of transport and a pulling force, in addition to the consumption of their meat and milk (see also Isaakidou 2004, 2006). Moreover, in the MN period the percentage of sheep became three times higher than that of goats, and by the LN the proportion of sheep was seven times higher than that of goats (Table 8.1). Thus, the sheep flocks formed the basis of the animal economy and were exploited for their meat and wool.

The Wild Species Four wild species—the wild goat (agrimi), the boar, the badger, and the marten—have been identified. The wild goat is found in the MN and LN levels, the boar in the transitional EN/MN as well as in the LN levels, the badger in the MN and LN, and the marten in the LN levels (Table 8.1). The taxa of the wild goat (Capra ibex or Capra aegagrus) are usually determined by the presence of the horns (Hecker 1982). In the material studied no horns were found, but the identification of Capra aegagrus at Phaistos (Wilkens 1996a, 2003), Smari (Tsoukala 1996), and Kavousi-Kastro (Klippel and Snyder 1991), as well as the depiction of the agrimi in Minoan iconography (Porter 1996), make it probable that the wild goat of Knossos corresponds to this taxon. At Knossos 12 bone remains are attributed to the agrimi. One (an ulna) derives from the MN period, and 11 bones come from the LN (one scapula, two proximal parts of radius, one diaphysis of radius, one ulna, one distal fragment of metacarpus, one proximal fragment of metatarsus, two phalanx

I, and one phalanx II; see Tables 8.8, 8.9A, 8.9B). There are another 32 articular fragments, and judging from their size and robustness they could also belong to the agrimi, but this could not be confirmed with certainty. Thus, they have been listed as Ovis–Capra hircus–Capra aegagrus (see Tables 8.1, 8.8, 8.9A, 8.9B), although their size, the protruding muscle insertion crests, and the robustness of the bones indicate that they belong to the agrimi (Fig. 8.5). With respect to the size of the bones, the minimum distal width (Ad) measurements of the proximal radius (Ad = 35.5 and 34.0 mm; the latter has a minimum diaphysis width of 20.5 mm) are within the measurement range of the bones from Phaistos (Wilkens 1996a) and within the minimum range of those from Asiab in the Kermanshah valley of Iran (Bökönyi 1977). There is some doubt regarding the radius with a proximal measurement of 34.0 mm, but if we take into account the minimum width of the diaphysis, that is, 20.5 mm, we can perceive the robustness of this bone part. The

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 153

diaphysis width exceeds the measurements seen at Argissa-Magula, where the Ad range of the radius of Ovis/Capra is 13.5–19.5 mm (with wild goats possibly included) and the average is 14.4 mm (Boessneck 1962, 90–91). In addition, the morphological descriptions are also the same (Wilkens 1996a). To these two bones, we must add a radius diaphysis, the minimum width of which is 26.3 mm, which is comparable to the maximum measurements from Asiab and corresponds to a male. In addition to these three radii, there are three phalanges that correspond to a wild goat: two phalanx I and one phalanx II (Tables 8.8, 8.9A, 8.9B). The measurements of phalanx I are higher than those of Capra hircus at Phaistos. One phalanx exceeds the maximum measurements observed at Tell Halula in the Euphrates valley of Syria, and the other falls within the maximum range of the phalanges at Halula (Saña 1999). The values of phalanx II are within the range of maximum values at Tell Halula. The remaining bones, out of a total of 12 Capra aegagrus remains, could not be measured, but judging from their size and robustness, they may be included in this taxon (Figs. 8.5, 8.6). The boar has been identified by consideration of the bone size. Only two remains can be assigned to this species with certainty: a large canine from the LN period (Fig. 8.7) and a radius diaphysis from the EN/MN period. There are another three remains that might belong to a boar, but this is uncertain. Figure 8.8 clearly shows that the diaphysis measurements are much larger than the respective measurements of pigs. In a sample of nine specimens from the site of Argissa-Magula, the maximum width of the pig radius diaphysis ranges between 13.5 and 17.5 mm (Boessneck 1962, 97), very similar to the measurements from Zambujal in Portugal (von den Driesch and Boessneck 1976, 63) and Cerro de la Virgen in Spain (von den Driesch 1972, 240–241). In contrast, the radius measurement from Knossos is 25.8 mm and perfectly matches the boar measurements from Zambujal and Cerro de la Virgen. The three doubtful bones comprise two ulnae, one from an EN/MN context and the other from the MN period (Fig. 8.7), and one phalanx I from the LN period. It was possible to measure only one of the two ulnae; the shortest distance from the processus anconaeus to the caudal border of the

a

c b

d

e

Figure 8.5. Animal bones from level 3: (a) radius diaphysis; (b) radius proximal part; (c) scapula; (d) fragment of femur; and (e) phalanx I of Capra aegagrus. Scale in cm. Photo M.P. Ripoll.

Figure 8.6. Distal metacarpus of Capra aegagrus and Ovis aries. Scale in cm. Photo M.P. Ripoll.

ulna (DPA) is 39.8 mm. This value falls within the minimum range of values for a sample of four boar specimens from Cerro de la Virgen, the range of which is 40–49 mm. The measurement range for pigs from a sample of 88 specimens from the same site is 31–39 mm. The ulnae measurements for a

MANUEL PÉREZ RIPOLL

154

a

c

b

Figure 8.7. Animal bones: (a) ulna in lateral view probably belonging to a wild boar (level 23); (b) Sus scrofa ferus canine fragment (level 10); (c) Sus scrofa domesticus ulna in lateral view (level 14). Scale in cm. Photo M.P. Ripoll.

sample of eight specimens from Argissa-Magula are between 30 mm and 33.7 mm. The Knossos measurement falls between the minimum measurements of boar and the maximum of pig, and thus it could belong to either a boar or a pig. The LN phalanx I has a maximum peripheral longitude (GLpe) of 36.2 mm, a value close to the maximum measurements of 65 specimens from Cerro de la Virgen (29–40 mm). Its proximal width (Ap), which is 20.7 mm, is bigger than those measured at Cerro de la Virgen (having a range of 13–17 mm in the same sample of 65 specimens). The minimum width of the diaphysis (SD), which is 16.3 mm, is bigger than that observed at Cerro de la Virgen (11– 14 mm; same sample), and the distal width (Ad), which is 18 mm, is also much bigger than the measurements of Cerro de la Virgen (11.5–15.3 mm for this site; same sample). Therefore, it is probable that this phalanx belongs to a boar. The marten is represented by a left distal humerus (Fig. 8.9) assigned to the LN period. The badger is represented by a left mandible without teeth, a lower canine, and an immature left ulna; the first two belong to the MN and the latter to the LN period.

(No. =7) 4 3 (No. =35) 2 (No. =9)

1 0

10

12

14

16

18

20

22

24

26

Zambujal and Cerro de la Virgen

Zambujal and Cerro de la Virgen

Knossos

Argissa-Magula

28

30

SD

Figure 8.8. Diaphysis width range (SD) of Sus scrofa domesticus and Sus scrofa ferus from Zambujal (Portugal), Cerro de la Virgen (Spain), Argissa-Magula, and Knossos.

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 155

The wild and domestic fauna were introduced to the island by the Neolithic inhabitants. During the Pleistocene the mammals of Crete included pygmy hippos, pygmy and large elephants, cervids, and murids. The pygmy hippo (Hippotamus creutzburgi) was associated with a pygmy elephant (Elephas creticus) and a large murid (genus Kritimys) that belong to the biostratigraphic zone of Kritimys. Its chronological limits are placed within the Middle Pleistocene. The next biostratigraphic zone (zone of Mus) consisted of a small murid, a large elephant (Elephas antiquus), and Cretan cervids (genus Candiacervus), including a small-sized cervid with a withers height of about 40 cm, a medium-sized one with a height of about 65 cm, and a large one with a height of about 90– 120 cm (De Vos 1996; Lax 1996; Mol et al. 1996; Spaan 1996). Dated bones place the cervids within a range from 150,000 to 22,000 years b.p. (Reese, Belluomini, and Ikeya 1996). During the Holocene the only mammal that may be regarded as a survivor of the endemic species is Crocidura zimmermanni, the Cretan white-toothed shrew (Masseti 2003). Therefore, the endemic fauna was already extinguished when Neolithic people arrived on the island. The mammals from Neolithic Knossos

a

d

b

c

Figure 8.9. Meles meles: (a) left mandible in lateral view (level 14); (b) lower canine (level 14); (c) left ulna in medial views; note that the proximal epiphysis is not fused (level 3). Martes : (d) distal part of humerus in cranial view (level 9). Scale in cm. Photo M.P. Ripoll.

must have been introduced during the colonization of Crete and within the context of the expansion of the Neolithic communities in the Mediterranean islands. Their appearance should not be interpreted as a separate insular phenomenon, but as a development taking place in a more general context common to other Mediterranean islands as well.

Age and Sex Selection Neolithic communities employed selective husbandry practices aimed at ensuring the reproduction of the species through the control of age and sex and selection of the most appropriate genetic characteristics. This selection is recognizable in the reduction of horn size and the decrease in the body size (Helmer 1992; Saña 1999; Pérez Ripoll 2001b). The domestication of the animals at Knossos was at an advanced stage, both in terms of the reduction of bone sizes and in the selective practices relating to age and sex. The age at death has been determined by analysis of the teeth and bones. The studies based on tooth eruption and wear are more precise than the

studies based on bones. Moreover, teeth are better preserved because they were not subjected to alteration or destruction by dogs. The bones do not present fixed patterns, and the proportion of immature bones varies extensively from one level to another, as the comparison of the number of immature bone remains and young teeth of the LN period indicates (Table 8.11). The discrepancy is strong between Ovis/Capra due to the alteration of long bones caused by dogs. The young age of the slaughtered animals is reflected, however, in the high quantity of immature bones found in the assemblages from Knossos and other Neolithic sites.

MANUEL PÉREZ RIPOLL

156

Bones mature/immature

Species

Tooth aging adult/young

EN II/MN

MN

LN

LN

NISP mature

43

45

135

NISP adult: 48

NISP immature

6

12

16

NISP immature: 9

% mature

84.1

73.4

88.2

% adult: 81.3

% immature

13.9

26.6

11.8

% immature: 18.7

26

38

250

66

Bos taurus

Ovis-Capra NISP mature NISP immature

5

16

22

29

% mature

80.8

57.9

90.4

% adult: 56.1

% immature

19.2

42.1

9.6

% immature: 43.9

9

16

54

8

Sus scrofa domesticus NISP mature NISP immature

3

2

18

2

% mature

66.7

87.5

66.7

% adult: 75

% immature

33.3

12.5

33.3

% immature: 25

Table 8.11. Number of long bone remains (the diaphysis fragments are not counted here), phalanges, and tarsi corresponding to mature and immature bones, along with the number of LN tooth remains, grouped by age, for comparison with the long bones.

Slaughter Age in Sheep and Goats Age has been determined on the basis of the teeth, taking into account the studies of Deniz and Payne (Payne 1973; Deniz and Payne 1982). The data included in these studies have been compared and adjusted to the data obtained from the study of the collections coming from the Mediterranean Neolithic sites of Valencia, Spain. The ages have been divided into classes, with each age class corresponding to the eruption of the permanent molars and their wear (Table 8.12). Moreover, the heights of the molars have been measured in order to differentiate between individuals with similar wear, something that was not taken into account in the system of Deniz and Payne (1982). The advantage of this measurement is that it enables us to compare the age classes of different species and to compare the selective patterns. The age profile (Fig. 8.10) was completed on the basis of the teeth and mandibles corresponding to the LN period, since the sample taken from the EN and MN levels was not sufficiently representative. A concentration in the class of young animals is observed, which is consistent with expectations for a livestock farming system that aims

to produce meat. This profile has been compared with that of the current production methods used in the Pomak community of Thrace in northern Greece, which still employs traditional exploitation systems transmitted from one generation to another with an apparently high degree of purity. This husbandry system aims toward the production of both goat and sheep meat. The slaughtering pattern is based on the reproduction cycle, which is fully adjusted to the annual seasonal cycles. The majority of the male animals and some of the females not destined for reproduction are slaughtered while they are young. Slaughtering ends with the beginning of the fertility cycle, at the age of one year for the males and one-and-a-half years for the females. For this reason, young animals will be slaughtered progressively in classes Ia, Ib, IIa, and IIb according to the data obtained from the study of mandibles from Neolithic Knossos (see Table 8.12). The maximum frequency occurs in class IIb. The progressive slaughter of adult animals takes place on the basis of the individual’s reproduction capacity, the use of adult animals in the celebration of festivals, and the occurrence of lesions

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 157 Class

Months

Years

Comment

NM

0

0–2

—

Milk dentition

0

Ia

2–6

—

M1 eruption

9

Ib

6–9

—

—

8

IIa

9–12

1

M2 eruption

1

IIb

12–24

2

—

11

IIIa

24–27

—

M3 eruption

8

IIIb

27–36

3

M3 initial wearing

6

IVa

—

4

—

5

IVb

—

5

—

3

IVc

—

6

—

9

Va

—

7

—

4

Vb

—

8 and +

—

2

NISP 20

15

10

5 0 0

I

II

III

IV

V

Age Class

Table 8.12. Number of mandibles (NM) for goats and sheep from the LN levels classified by age.

Figure 8.10. Age classes of the mandibles of Ovis and Capra.

or illnesses among the animals. The case of animals growing old (more than seven years of age) is rather rare. The slaughter patterns are necessary to maintain the balance between the size of the livestock and the food potential of pastures. Moreover, this procedure tries to keep the numerical component of the livestock to ensure food for humans. Although this slaughter profile is basically adjusted to the production of meat, milk is also consumed. Pomaks slaughter their animals after they have been weaned. The Bedouins of Jordan use a system in which milk is shared, with the output of one nipple kept for feeding the animals and the other for human consumption. The animals are milked in spring and summer. Both systems aim at a production of milk limited only to family use. Some of the milk is drunk fresh, but most of it is turned into yogurt, butter, and cheese. It is highly

likely that the Neolithic communities of Knossos consumed milk, using the methods employed both by the Pomaks and the Bedouins, within a slaughter profile that is characteristic of meat production (Chatty 1986; Weir 1990; Pérez Ripoll 2003). Livestock plays another important role in the economic system of traditional communities, however; it serves as a food stock in the event of an agricultural crisis. This has been verified by studies now taking place among traditional herding and farming communities in the Spanish Mediter­ ranean. In seasonal dry cycles the livestock turns into a food alterative by castrating the males or slaughtering females in order to confront the critical seasons and the lack of food. This slaughtering pattern involves age classes II and especially III as seen at Knossos (Table 8.12).

Slaughter Age for Cattle The age table for cattle (Table 8.13) was created with the same methodology used for the classification of goats and sheep. The only difference is in the months of molar eruption (for the data, see Habermehl 1975; Grant 1982). The slaughter profile of cattle shows a pattern related to the production of meat, but unlike the case of sheep and goats, slaughtering mainly focuses on class III,

between ca. 27 and 30 months. The greater delay in slaughter maximizes meat production. Although the model of selection is typical of meat production (Fig. 8.11), it is highly likely that cow’s milk was used for household consumption as in the case of the Pomaks, where each family keeps one to three cows for meat and milk (Pérez Ripoll 2003).

MANUEL PÉREZ RIPOLL

158 Class

Months

Comment

NM

0

5

Milk dentition

0

I

6

M1 eruption

1

II

18

M2 eruption

8

IIIa

27–30

M3 eruption

10

10

IIIb

30–48

M3 initial wearing

9

5

IVa

48 and +

—

9

IVb

adult

—

6

V

old

—

5

NISP 20

Table 8.13. Number of mandibles (NM) of Bos taurus classified by age.

15

0 0

I

II

III

IV

V

Age Class

Figure 8.11. Age classes of the mandibles of Bos taurus.

In addition to meat and milk, cattle may also be used for their force. The Pomaks use cows to cultivate their land. This practice produces malformations in the joints of the limbs, visible in the bone remains scattered around permanent and seasonal houses in the outskirts of the village of Sarakini. The same deformations have been identified in some cattle bones from Knossos, indicating that during the LN period cattle were also used as traction animals. Figure 8.12 shows important alterations in the articular surface of a distal metacarpus caused by the use of the animal in demanding duties. Therefore, the increasing exploitation of the different qualities of cattle may explain the noticeable increase of cattle bones at Knossos from the MN period onward.

Figure 8.12. Distal part of metacarpus belonging to a male (possibly ox) of Bos taurus, with osseous deformations on the articular surfaces. Scale in cm. Photo M.P. Ripoll.

Slaughter Age for Pigs Although the sample of pig bones is not very large, it offers some insight about age selection (Table 8.14). Age classes are based on molar eruption and wear (Habermehl 1975; Grant 1982). There is a concentration of individuals in classes

II and III (between 12 and 24 months) and especially in class IV (adults). The slaughter diagram illustrates this death profile (Fig. 8.13). Such selection practices are aimed at providing more meat.

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 159

Class

Months

Comment

Maxillae

Mandibles

0

5

Milk dentition

0

0

I

6

M1 eruption

1

0

II

12

M2 eruption

3

2

III

20–24

M3 eruption

1

2

IV

adult

—

4

4

V

old

—

0

0

NISP

Table 8.14. Number of maxillary and mandibular remains of Sus scrofa domesticus classified by age.

4

2

0 0

I

II

III

IV

V

Age Class

Figure 8.13. Age classes of Sus scrofa domesticus maxillae and mandibles.

Sex Distribution of Cattle Sex can be determined either by examining the morphological features of certain bones, such as the horns and the pelvis, or through consideration of their respective measurements. Unfortunately, horns of all bovines are rare and fragmented, and therefore they do not provide enough information. Pelvises are more numerous and contribute some information. Most of the data on sex, however, come from the analysis of bone measurements. Sexual dimorphism has been established as indicated in Table 8.15, using pelvic morphology and bone measurements from the metacarpus, metatarsus, phalanx I, and phalanx III. Measurements of phalanx II were not used because of their unreliable sex classification. There were no remains amenable to sex classification from the EN levels; the most numerous bones came from the LN deposits.

In addition to the dispersion and classification of the measurements shown in the corresponding diagrams (Fig. 8.14), we have also taken into account the limits of values that mark sex difference based on measurements from faunal samples at various sites in the Iberian peninsula (Pérez Ripoll 1999). The metric values from these sites are very similar to those found at Knossos. The majority of the cattle bones belong to females. Taking into account that the bones belong to adult animals, we can suggest that the scarcity of males is due to the fact that they were slaughtered at a young age, i.e., when they produce more meat. This practice is attested by the values for age classes II and III in the age chart (Table 8.13). It is also possible that females were kept longer for the production of milk.

Sex Distribution of Sheep and Goats Morphological indicators of sex in sheep and goats are associated with the horns and pelvis. Unfortunately, these bones were rare. Measure­ ments pertaining to sexual dimorphism were possible only on four bones: the scapula, humerus, pelvis, and tibia. The results of this analysis are illustrated in Table 8.16. All bones apart from the

pelvis indicate that females were in the majority, an observation that is consistent with the data concerning the slaughtering of other types of animals. The males were culled before adulthood, as were the females not destined for reproduction; females were also killed when they were ill or wounded.

MANUEL PÉREZ RIPOLL

160 Bos taurus

EN/MN

MN

LN 40

Pelvis Male

3

2

0

Female

2

0

7

35 Bp 30 25

Metacarpus Bd Male

—

0

0

Female

—

2

4

Male

1

—

—

Female

3

—

—

Male

1

—

3

Female

0

—

13

Male

1

1

3

Female

5

3

12

Metatarsus Bd

Phalanx I

Phalanx III

20 50

The data for age and sex indicate that sheep were used for the production of meat, with the exploitation of milk being feasible at a family level in the period between weaning and gestation. Although the data for goats are very limited, the numbers would seem to indicate that the production of meat as well as milk was important. Males, including adults, may have been used for meat. In traditional pastoral societies of Spain, male goats were castrated so that they would gain more weight and become docile, thus avoiding problems related to the rut.

60

65

70

GLpe

Figure 8.14. Correlation of the measurements of phalanx I belonging to Bos taurus : proximal breadth (Bp) and maximum peripheral longitude (great long peripheric; GLpe). The circles indicate the distribution of the values by sex (males have larger measurements).

Ovis-Capra

Table 8.15. Number of identified specimens of Bos taurus classified by sex on the basis of morphological features and bone measurements.

55

EN II/MN

MN

LN

Male

—

—

0

Female

—

—

1

Male

—

—

3 (O.a.) 1 (C.h.)

Female

—

—

8 (O.a.) 1 (C.h.)

Horn

Scapula

Humerus Male Female

—

—

2 (O.a.) 1 (C.h.)

1 (O.a.) 1 (C.h.)

2 (O.a.)

5 (O.a.)

—

—

2 (O.a.)

1 (C.h.)

—

11 (O.a.) 2 (C.h.)

Tibia Male Female Pelvis Male

—

—

5 (O.a.)

Female

—

—

4 (O.a.)

Table 8.16. Number of identified specimens of Ovis ariews ( O.a. ) and Capra hircus ( C.h. ) classified by sex on the basis of morphological features and bone measurements.

Evaluation of Domesticated Fauna The composition of the livestock (Tables 8.1, 8.17), with a high proportion of sheep and goats and lesser numbers of cows and pigs, was very closely connected to the overall regime of agricultural practices. Sheep must have been perfectly integrated in the farming cycle, since they consume stubble and straw from the fields as well as the waste from the grinding of grain. Thus, sheep can be considered a complement to agriculture

because they transform by-products of no apparent use to rural communities into meat proteins. Cattle played a rather different role. There is no doubt that their milk was consumed by the community, but the utilization of their force for rural tasks must have been equally important. Wear observed in certain bones from the limbs seems to indicate exactly this practice (Fig. 8.12). The exploitation of this capability probably became most

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 161

important in the MN and LN periods, as a significant increase in the numbers of cows and the above-mentioned bone deformations were observed in these levels. This new use of animals may have arisen due to increased needs for rural productivity through plowing and transportation. Isaakidou (2004, 2006) comes to similar conclusions from the analyses of female cattle bones in Evans’s faunal sample, which show pathological conditions with a selective anatomical distribution that suggests work-related stress. According to her study, the earliest indications of these conditions come from EN Ic material, and she proposes that adult cows were being used beginning in this period for traction as well as breeding (Isaakidou 2004, 247–250, table 7:4, fig. 7:6). Pigs were also very important in the diet of traditional Mediterranean households. They were relegated to the role of transforming rural by-products, both cereal and horticultural, into proteins for human consumption. There was no need for a large number of pigs per family; one male and one female would have been sufficient to fulfill the dietary needs of one family for a whole year. The proportions of the Knossos livestock are very similar to those observed at other continental Greek and Cretan Neolithic sites (Table 8.18). According to the data from Argissa-Magula

% NISP

Knossos

AN

EN I

EN II

MN

LN

Bos

X

X

X

X

X

Ovis

X

X

X

X

X

Capra hircus

X

X

X

X

X

Sus

X

X

X

X

X

Canis

X

X

X

X

X

Cervus

X

Meles

X

X

Martes

X

X

X

X

Table 8.17. Chronological representation of the faunal species at Knossos according to Jarman (1996).

(Boessneck 1962), Nea Nikomedeia (Higgs 1962), Achilleion (Bökönyi 1989), and Franchthi (Payne 1975), cattle had little importance in the EN; sheep and goats were more important, while the numerical relevance of pigs was limited. There was a tendency at these sites, depending on their locations, for either cattle or pigs to increase from the MN period onward. This observation coincides with the data from Knossos, Phaistos (Wilkens 1996a, 2003), and Hagia Triada (Wilkens 1996a, 1996b, 2003). Therefore, there are common patterns among the different sites of Greece in terms of livestock exploitation.

Wild

Domestic

Cattle

Sheep/goat

Pig

Dog

Knossos EN I–II (Pérez Ripoll)

0

100

11.3

76.8

11.3

0.5

Knossos Camp (Jarman and Jarman 1968)

0

100

6.5

74.7

18.4

0.2

Knossos Ia (Jarman and Jarman 1968)

0

100

13.6

64.6

20.9

0.9

Knossos Ib (Jarman and Jarman 1968)

0

100

22.7

61.5

13.9

1.9

0.9

99

4.7

84.1

10

0.1

7

93

13.6

65.9

13.8

0.2

Achilleion (Bökönyi 1989)

5.3

94.7

5.1

74.9

13.9

0.9

Knossos MN–LN (Pérez Ripoll)

0.8

99.2

37.6

50.1

11.2

0.1

Phaistos LN (Wilkens 1996a, 2003)

4.9

95.1

23.9

48.1

22.4

0.3

Hagia Triada EM (Wilkens 1996b, 2003)

0.5

99.5

17.9

68.1

13.3

0

Hagia Triada MM (Wilkens 1996b, 2003)

0.4

99.5

5.9

66.9

26.6

0

Hagia Triada MM III/LM IA (Wilkens 1996b, 2003)

1.4

98.6

13.3

54.1

31.1

0

Hagia Triada LM (Wilkens 1996a, 1996b, 2003)

3.4

96.6

6.1

71.7

18.5

0

0

100

11.8

61.3

23.8

0

Argissa-Magula (Boessneck 1962) Nea Nikomedeia (Higgs 1962)

Knossos LM (Bedwin 1984)

Table 8.18. Percentages of identified specimens of domestic and wild species at Knossos and other sites.

162

MANUEL PÉREZ RIPOLL

Evaluation of the Wild Fauna in the Context of the Mediterranean Islands In contrast to the emphasis placed on plant cultivation and animal husbandry, the importance of wild fauna in the farming economies of the Mediterranean islands has not been sufficiently evaluated. The first agriculturalists who settled in Cyprus brought with them both wild animals such as Dama mesopotamica and domesticated goat, sheep, cow, and pig (Table 8.19), as seen in the faunal remains from the Aceramic sites of Shillourokambos (Guilaine et al. 1996; Vigne 2000, 2001; Vigne, Carrère, and Guilaine 2003; Vigne et al. 2004), Ais Yiorkis (Reese 1996, 1999), Cape Andreas Kastros, and Khirokitia (Davis 1987, 1994). Moreover, remains of Dama and boar have also been identified at older sites such as Akrotiri-Aetokremnos (Reese and Roler 1999). Fallow deer were probably introduced in a wild state (Davis 1987) for the purpose of repopulating the island, since the big endemic fauna had long been extinct. They may also have been used to exploit the island’s uncultivated zones. In addition to these herbivores, small carnivores were introduced to Cyprus, possibly for their skins. Genetta was noted at Akrotiri-Aetokremnos (Steensma and Reese 1999); Vulpes vulpes at Shillourokambos (Vigne 2001), Ais Yiorkis, and Khirokitia (Davis 1987, 1994; Simmons 1998); a cat at Shillourokambos (Vigne et al. 2004); and another small feline at Ais Yiorkis and Khirokitia (Davis 1987, 1994; Simmons 1998). Similarly, the large island of Crete, with plateau zones and high mountains devoid of endemic mammals (Cherry 1990), was probably repopulated during the Neolithic by the wild goat (agrimi) and small carnivores such as the marten and the badger, as indicated by the finds from Knossos and Phaistos (Table 8.20). As in Cyprus, the aim of those who introduced these animals may have been to exploit extensive areas not suitable for cultivation. The presence of the boar may be interpreted from the same perspective. Although it is still an open possibility that some of the animals such as the goat and boar were introduced to the island in a domesticated state and later became feral (Groves

1989; Broodbank and Strasser 1991), there is no doubt that the marten and the badger were imported (Steensma and Reese 1999). The badger seems to have been present from a very early date, as it can be seen in the Aceramic Neolithic (AN) levels of Evans’s excavations (Jarman 1996). Both carnivores have also been identified in the later Minoan levels at the site of Hagia Triada (Wilkens 1996a, 1996b). The same process of repopulation continued in Crete in later periods when new species were brought in from the continent. Thus, during the Minoan period different bones appear in the archaeological catalogues of Knossos (Bedwin 1984; Jarman 1996) and Hagia Triada (Wilkens 1996a, 1996b). Among them were new species such as the horse, donkey, fallow deer, red deer (noted by Jarman already in the LN), and hare (Table 8.20). Mustelids and the wild cat were found at the site of Kastro in the levels belonging to the orientalizing period (Snyder and Klippel 1996). In the central Mediterranean islands the largeand medium-sized endemic mammals had become extinct prior to the arrival of the Neolithic population (Vigne 1995a, 1995b). On the island of Corsica, for instance, at the sites of AraguinaSennola XVIII, Curacchiaghiu 7, Strete II XXIV– XXII, Monte Leone, and Pietracorbara 7 and 8, the pre-Neolithic fauna basically consisted of a small lagomorph, Prolagus sardus, rodents, fish, marine malacofauna, birds, and, in Araguina, the remains

Species Bos Ovis/ Capra Sus

Shillourokambos

Ais Yiorkis

X

X

X (++)

X (+)

Khirokitia

X (++)

X

X (++)

X (+)

Dama

X (+)

X (++)

X (++)

Vulpes

X

X

X

Felis

X

X

X

Table 8.19. Representation and abundance of various faunal species at Shillourokambos (Guilaine et al. 1996; Vigne 2000), Ais Yiorkis (Reese 1996, 1999), and Khirokitia (Davis 1987, 1994) (+ = abundant, ++ = very abundant).

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 163

Sites in Crete

Equus Equus Cervus Dama Capra caballus asinus elaphus dama aegagrus

Felis silvestris

Mustela nivalis

Knossos EN (Jarman 1996)

Meles meles

Martes Lepus foina

X

X

?

Knossos MN (Pérez Ripoll)

X

X

Knossos LN (Pérez Ripoll)

X

X

X

X

X

X

X

X

X

X

X

X

X

X

Knossos LN (Jarman 1996)

?

X

Phaistos LN (Wilkens 1996a, 1996b)

X

Knossos Minoan (Jarman 1996)

X

Knossos LM (Bedwin 1984)

X

X

X

X

X

Hagia Triada EM (Wilkens 1996a, 1996b)

X

Hagia Triada MM (Wilkens 1996a, 1996b)

X

Hagia Triada MM/LM (Wilkens 1996b)

X

X

Hagia Triada LM (Wilkens 1996b)

X

Kastro LM–Early Orientalizing (Snyder and Klippel 1996)

X

Gortys Byzantine (Wilkens 1996a)

Oryctolagus cuniculus

X

X

X

X

X

X

Table 8.20. Introduction and chronological representation of wild animals at various sites in Crete.

X

X

164

MANUEL PÉREZ RIPOLL

of a monk seal (Vigne 1995a, 1995b; Vigne and Desse-Berset 1995). On the island of Sardinia at its key pre-Neolithic site Corbeddu, the endemic fauna consisted of Megaceros cazioti (a cervid), Cynotherium sardus (a carnivore similar to a fox), Prolagus sardus, insectivores, birds, and malacofauna (Klein Holfmeijer 1998). In the case of Corsica, both Megaceros and Cynotherium had become extinct before the arrival of the pre-Neolithic hunters (Vigne 1990). The sites on these two large islands with occupations dating to the EN are Filiestru in Sardinia and Basi 7, Strete XIV, Strete XIII (final EN), AraguinaSennola XVII (final EN), and Pietracorbara 6 (final EN) in Corsica. The domestic fauna is predominant at all these sites, with the exception of Araguina XVII and Strette XIII–XIV, where it represents only 1.4% and 9.6%, respectively, of the total remains. The domesticated animals included cattle, sheep, goat, pig, and dog. The wild fauna consisted of Prolagus, malacofauna, and some remains of Vulpes vulpes in Basi 7 and Filiestru D 6–7 (Levine 1983; Vigne and Alcover 1985; Vigne 1987a, 1987b, 1987c, 1990, 1995a, 1995b). The presence of Vulpes vulpes is significant, indicating that it was introduced together with the domestic fauna, following the same pattern observed in Cyprus and Crete. On the Balearic Islands, the continental domestic fauna arrived late. Unfortunately, the first cases of the introduction of domesticated animals have

not been adequately studied; only the animals from levels 24–28 of Son Matge and strata 5 and 6 of Sector X of Moleta are known. Present were goat, sheep, and pig, all associated with the first pottery (Waldren 1982). The same species were found at the site of Cova Simó (Coll 2000). Radiocarbon dating has not been particularly helpful because of stratigraphic problems at the first two sites (Alcover et al. 2001). According to pottery typologies the animals must belong to the Final Neolithic (Guerrero 2000) or Early Chalcolithic periods (Coll 2000). The most direct dates come from a goat mandible from Cova des Moro (Beta-155645: 3750±40 b.p., 2290–2030 cal. b.c., with a probability of 95%; see Alcover et al. 2001) and a molar of Caprinae from Cova Simó, level 35 (Beta-154196: 3760±40 b.p., 2300–2030 cal. b.c., with a probability of 95%; see Coll 2001). The introduced wild fauna dates to the Talayotic period (1200–123 b.c.) and consists of deer, fallow deer, boar, wild cat, hare, and rabbit (Uerpmann 1971; Estévez 1984; Ramis 2002–2003; Morales Pérez 2005). It is apparent that the introduction of wild animals in Crete was not an isolated case. Faunal developments on the island fit well with those ob­ served in the broader westward expansion of Neo­ lith­ ic communities, one that began in Cyprus, con­ tinued to Crete and the Aegean islands, and moved on to the islands of central and the western Mediterranean.

Conclusions The fauna of Neolithic Knossos is very important both due to its early date in the island’s prehistory and with respect to the ongoing discussion of the spread of the farming communities to the west. The arrival of agriculturalists in Crete at the end of the eighth millennium b.c., as verified by the new set of 14C dates from Knossos, seems to be the second most important episode of island colonization in the eastern Mediterranean after Cyprus. The latter became the home of farmers who by the ninth millennium b.c. (early Cypriot Pre-Pottery Neolithic B) had already crossed the sea and settled along the island’s southern coast at places like Kalavasos-Tenta,

Shillourokambos, and Mylouthkia (Peltenburg et al. 2000, 2001; Guilaine and Briois 2001; Todd 2001). Shillourokambos typifies the difficulties of interpreting the faunal record of an early island habitation. The small size of the bones of pig and dog is indicative of domestication. Metric measurements of the bones of cattle, goat, and sheep, however, are identical to those of their wild ancestors on the Near Eastern mainland. Nevertheless, the distribution of age classes, the presence of all parts of the skeleton, and the selection by sex all suggest the employment of husbandry practices (Guilaine et al. 1996; Vigne 2000, 2001; Vigne, Carrère,

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 165

and Guilaine 2003). Similar considerations of size are also relevant to the remains of a cat associated with a human burial; the size of these bones is similar to those of the wild cat, but the strong social association reinforces the domestic status of the remains (Vigne et al. 2004). This may be seen as part of a local “pre-domestication” stage that is tentatively documented in Cyprus and corresponds to an early stage of the domestication process (Vigne and Buitenhuis 1999). The composition of the livestock at Knossos is the same as that of Shillourokambos, although the domestication process was much more advanced, as indicated by the reduction of the bone size. The measurements of the bones from Knossos are very similar to those of the other Neolithic sites of continental Greece as well as those of the central and western Mediterranean. The composition and proportion of domestic taxa are also strikingly homogeneous at the Mediterranean Neolithic sites, with

a predominance of Ovis/Capra (Ovis being more abundant than Capra) and a minor representation of Bos during the EN, followed by the steady increase of Bos during the MN and LN periods because of milking and traction practices. The bones of Sus were of little importance. The wild fauna from island contexts, although not important numerically, nonetheless has a qualitative significance. In the Aceramic levels of Cyp­ riot sites, Dama, Genetta, Vulpes, and a small Felis were documented. In Crete, Capra aegagrus (MN and LN), Cervus elaphus (LN), perhaps Sus scrofa ferus (EN/MN), Meles meles (EN, MN, LN, and Evans’s Aceramic), and Martes foina (EN and LN) have been identified. Farther west in Corsica Vulpes vulpes appeared in the EN period, while in the Balearic Islands the wild fauna arrived in the Talayotic period. The introduction of wild mammals to Crete was thus not an isolated phenomenon but part of a general island process.

References Alcover, J.A., D. Ramis, J. Coll, and M. Trias. 2001. Bases per al coneiximent del contacte entre els primers colonitzadors humans i la naturalesa de Les Balears (Endins 24), Palma.

Neolithikum sowie die Tier- und Pflandzenreste (Beitrage zur ur-und frühgeschichtlichen Archäologie des Mittelmeer-Kulturraumes 23), V. Milojčić, J. Boessneck, and M. Hopf, Bonn, pp. 27–99.

Aura, J.E., V. Villaverde, M. Pérez Ripoll, R. Martínez Valle, and P. Guillem. 2002. “Big Game and Small Prey,” Journal of Archaeological Method and Theory 9, pp. 215–268.

Boessneck, J., H.H. Müller, and M. Teichert. 1964. “Osteologische Unterscheidungsmerkmale zwischen Schaf (Ovis aries, Linné) und Ziege (Capra hircus, Linné),” Kühn-Archiv 78, pp. 1–129.

Bedwin, O. 1984. “The Animal Bones,” in The Minoan Unexplored Mansion at Knossos (BSA Suppl. 17), M.R. Popham, ed., Oxford, pp. 307–308.

Bökönyi, S. 1977. The Animal Remains from Four Sites in the Kermanshah Valley, Iran: Asiab, Sarab, Dehsavar, and Siahbid. The Faunal Evolution, Environmental Changes and Development of Animal Husbandry, VIII–III Millennia b.c. (BAR Suppl. 34), Oxford.

Bernabeu, J., C.M. Barton, and M. Pérez Ripoll. 2001. “A Taphonomic Perspective on Neolithic Beginnings: Theory, Interpretation, and Empirical Data in the Western Mediterranean,” JAS 28, pp. 597–612. Bernabeu, J., M. Pérez Ripoll, and R. Martínez Valle. 1999. “Huesos, neolitización y contextos arqueológicos aparentes,” in Actes del II Congrés del Neolític a la Península Ibérica (Saguntum-PLAV Extra 2), J. Bernabeu and T. Orozco, eds., Valencia, pp. 589–596. Boessneck, J. 1962. “Die Tierreste aus der Argissa-Magula vom präkeramischen Neolithikum bis zur mittleren Bronzezeit,” in Die deutschen Ausgrabungen auf der Argissa-Magula in Thessalien I: Das präkeramische

———. 1989. “Animal Remains,” in Achilleion: A Neolithic Settlement in Thessaly, Greece, 6400–5600 bc (Monumenta Archaeologica 14), M. Gimbutas, S. Win, and D. Shimabuku, eds., Los Angeles, pp. 315–332. Broodbank, C., and T.F. Strasser. 1991. “Migrant Farmers and the Neolithic Colonization of Crete,” Antiquity 65, pp. 233–245. Chatty, D. 1986. From Camel to Truck: The Bedouin in the Modern World, New York.

166

MANUEL PÉREZ RIPOLL

Cherry, J.F. 1990. “The First Colonization of the Mediterranean Islands: A Review of Recent Research,” JMA 3, pp. 145–221. Coll, J. 2000. “Excavación arqueológica del coval Simó: Estado actual y perspectivas de futuro,” in Colontización humana en ambients insulars: Interacció amb el medi i adaptació cultural, V. Guerrero and S. Gornés, eds., Palma, pp. 371–381. ———. 2001. “Primeres datacions absolutes del jaciment de covall Simó,” Endins 24, pp. 161–168.

Guilaine, J., F. Briois, J. Coularou, J.-D. Vigne, and I. Carrère. 1996. “Shillourokambos et les débuts du néolithique à Chypre,” Espacio, Tiempo y Forma 1°, Prehistoria y Arqueología 9, pp. 159–171. Habermehl, K.-H. 1975. Die Alterbestimmung bei Haus-und Labortieren 2, Berlin. Hecker, H.M. 1982. “Domestication Revisited: Its Implications for Faunal Analysis,” JFA 9, pp. 217–236. Helmer, D. 1992. La domestication des animaux par les hommes préhistoriques, Paris.

Davis, S.J.M. 1987. “La faune,” in Le néolithique précéramique de Chypre (L´Anthropologie 91 [1]), A. Le Brun, S. Cluzan, S.J.M. Davis, J. Hansen, and J. Renault-Miskovsky, Paris, pp. 305–309.

Higgs, E.S. 1962. “Fauna,” in Excavations at the Early Neolithic Site at Nea Nikomedeia, Greek Macedonia, 1961 Season (PPS 28), R.J. Rodden, ed., pp. 271–274.

———. 1994. “Even More Bones from Khirokitia: The 1988–1991 Excavations,” in Fouilles récentes à Khirokitia (Chypre), 1988–1991, A. Le Brun, ed., Paris, pp. 305–333.

Isaakidou, V. 2004. Bones from the Labyrinth: Faunal Evidence for the Management and Consumption of Animals at Neolithic and Bronze Age Knossos, Crete, Ph.D. diss., Univeristy College London.

Deniz, E., and S. Payne. 1982. “Eruption and Wear in the Mandibular Dentition as a Guide to Ageing Turkish Angora Goats,” in Ageing and Sexing Animal Bones from Archaeological Sites (BAR-BS 109), B. Wilson, C. Grigson, and S. Payne, eds., Oxford, pp. 155–205.

———. 2006. “Ploughing with Cows: Knossos and the ‘Secondary Products Revolution,’” in Animals in the Neolithic of Britain and Europe (Neolithic Studies Group Seminar Papers 7), D. Serjeantson and D. Field, eds., Oxford, pp. 95–112.

De Vos, J. 1996. “Taxonomy, Ancestry and Speciation of the Endemic Pleistocene Deer of Crete Compared with the Taxonomy, Ancestry and Speciation of Darwin’s Finches,” in Reese, ed., 1996, pp. 111–124.

Jarman, M.R. 1996. “Human Influence in the Development of the Cretan Mammalian Fauna,” in Reese, ed., 1996, pp. 211–229.

Estévez, J. 1984. “Restos alimentarios e industria ósea de Son Fornés (Mallorca),” in Son Fornés I: La Fase Talayótica. Ensayo de reconstrucción socioeconómica de una comunidad prehistórica de la isla de Mallorca (BAR-IS 209), P. Gasull, V. Lull, and M.E. Sanahuja, Oxford, pp. 138–178. Grant, A. 1982. “The Use of Tooth Wear as a Guide to the Age of Domestic Ungulates,” in Ageing and Sexing Animal Bones from Archaeological Sites (BARBS 109), B. Wilson, C. Grigson, and S. Payne, eds., Oxford, pp. 91–108. Groves, C.P. 1989. “Feral Animals of the Mediterranean Islands: Documents of Early Domestication,” in The Walking Larder: Patterns of Domestication, Pastoralism and Predation (One World Archaeology 2), J. Clutton-Brock, ed., London, pp. 46–58. Guerrero, V.M. 2000. “La colonización humana de Mallorca en el contexto de las islas occidentales del mediterráneo: Una revisión crítica,” in Colonizació humana en ambients insulars: Interacció amb el medi i adaptació cultural, V. Guerrero and S. Gornés, eds., Palma, pp. 99–190. Guilaine, J., and F. Briois. 2001. “ParekklishaShillourokambos: An Early Neolithic Site in Cyprus,” in Swiny, ed., 2001, pp. 37–54.

Jarman, M.R., and H.N. Jarman. 1968. “The Fauna and Economy of Early Neolithic Knossos,” in “Knossos Neolithic, Part II,” P. Warren, M.R. Jarman, H.N. Jarman, N.J. Shackleton, and J.D. Evans, BSA 63, pp. 239–276. Klein Hofmeijer, G. 1998. Late Pleistocene Deer Fossils from Corbeddu Cave: Implications for Human Colonization of the Island of Sardinia (BAR-IS 663), Oxford. Klippel, W.E., and L.M. Snyder. 1991. “Dark-Age Fauna from Kavousi, Crete: The Vertebrates from the 1987 and 1988 Excavations,” Hesperia 60, pp. 179–186. Lax, E.L. 1996. “A Gazetteer of Cretan Paleontological Localities,” in Reese, ed., 1996, pp. 1–32. Levine, M. 1983. “La fauna de Filiestru (Trincea D),” in La grotta de Filiestru a Bonu Ighinu, Mara (Sassari), D.H. Trump, Sassari, pp. 1–32. Martínez Valle, R. 2001. “Cazadores de pequeñas presas,” in Villaverde, ed., 2001, pp. 129–130. Masseti, M. 2003. “Holocene Endemic and non Endemic Mammals of the Aegean Islands,” in Zooarchaeology in Greece: Recent Advances (BSA Studies 9), E. Katjabopoulou, Y. Hamilakis, P. Halstead, C. Gamble, and P. Elefanti, eds., London, pp. 53–63.

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 167

Mol, D., J. De Vos, G.D. van den Bergh, and P.Y. Sondaar. 1996. “The Taxonomy and Ancestry of the Fossil Elephants of Crete: Faunal Turnover and a Comparison with Proboscidean Faunas of Indonesian Islands,” in Reese, ed., 1996, pp. 81–98. Morales Pérez, J. V. 2005. “Estudi zooarqueològic del jaciment de Talatí de Dalt: Economia pecuaria i explotació dels mamífers,” in Talatí de Dalt 1997–2001 (Treballs del Museu de Menorca 29), G.J. Benejam and J. Pons, Palma, pp. 189–216. Payne, S. 1973. “Kill-Off Patterns in Sheep and Goats: The Mandibles from Asvan Kale,” AnatSt 23, pp. 281–303. ———. 1975. “Faunal Change at Franchthi Cave from 20,000 bc to 3,000 bc,” in Archaeozoological Studies, A.T. Clason, ed., Amsterdam, pp. 120–131. Peltenburg, E., S. Colledge, P. Croft, A. Jackson, C. McCartney, and M.A. Murray. 2000. “AgroPastoralist Colonization of Cyprus in the 10th Millennium bp: Initial Assessments,” Antiquity 74, pp. 844–853. Peltenburg, E., P. Croft, A. Jackson, C. McCartney, and M.A. Murray. 2001. “Well-Established Colonists: Mylouthkia 1 and the Cypro-Pre-Pottery Neolithic B,” in Swiny, ed., 2001, pp. 61–95. Pérez Ripoll, M. 1980. “La fauna de vertebrados,” in Cova de l´Or (Beniarrés-Alicante) (Serie de Trabajos Varios 65), B. Martí, V. Pascual, M.D. Gallart, P. López, M. Pérez, J.D. Acuña, and P. Robles, Valencia, pp. 193–255. ———. 1992. Marcas de carnicería, fracturas intencionadas y mordeduras de carnívoros en huesos prehistóricos del Mediterráneo español (Colección Patrimonio 15), Alicante. ———. 1999. “La explotación ganadera durante el III milenio a.C. en la Península Ibérica,” in Actes del II Congrés del Neolític a la Península Ibérica (Saguntum-PLAV Extra 2), J. Bernabeu and T. Orozco, eds., Valencia, pp. 95–103.

internationales d´archéologie et d´histoire d´Antibes, J.P. Brugal and J. Desse, eds., Antibes, pp. 191–206. ———. 2005–2006. “Caracterización de las fracturas antrópicas y sus tipologías en huesos de conejo procedentes de los niveles gravetienses de la Cova de les Cendres (Alicante),” Munibe 57, pp. 239–254. Pérez Ripoll, M., and R. Martínez Valle. 2001. “La caza, el aprovechamiento de las presas y el comportamiento de las comunidades cazadoras prehistóricas,” in Villaverde, ed., 2001, pp. 73–98. Porter, R. 1996. “The Cretan Wild Goat (Capra aegagrus cretica) and the Theran Antelopes,” in Reese, ed., 1996, pp. 295–315. Ramis, D. 2002–2003. La fauna del asentamiento de navetes de Canyamel, Mallorca (II milenio cal bc), Ph.D. diss., Universidad Nacional a Distancia, Palma de Mallorca. Reese, D.S. 1996. “Cypriot Hippo Hunters No Myth,” JMA 9, pp. 107–112. ———. 1999. “The Faunal Assemblages. Pygmy Elephant. Pig. Deer,” in Simmons, ed., 1999, pp. 161–167. Reese, D.S., ed. 1996. Pleistocene and Holocene Fauna of Crete and Its First Settlers (Monographs in World Archaeology 28), Madison. Reese, D.S., G. Belluomini, and M. Ikeya. 1996. “Absolute Dates for the Pleistocene Fauna of Crete,” in Reese, ed., 1996, pp. 47–51. Reese, D.S., and K. Roler. 1999. “The Faunal Assemblages: Pygmy Hippopotamus,” in Simmons, ed., 1999, pp. 156–160. Saña, M. 1999. Arqueología de la domesticación animal: La gestión de los recursos en Tell Halula (valle del Éufrates-Siria) del 8.800 al 7.000 bp (Treballs d´Arqueologia del Pròxim Orient 1), Barcelona. Silver, I.A. 1969. “The Ageing of Domestic Animals,” in Science in Archaeology, D. Brothwell and E.S. Higgs, eds., London, pp. 283–302.

———. 2001a. “Marcas antrópicas en los huesos de conejo,” in Villaverde, ed., 2001, pp. 119–124.

Simmons, A.H. 1998. “Of Tiny Hippos, Large Cows and Early Colonists in Cyprus,” JMA 11, pp. 232–241.

———. 2001b. “El proceso de domesticación animal en el Próximo Oriente: Planteamiento y evolución,” Archivo de Prehistoria Levantina 24, pp. 65–96.

Simmons, A.H., ed. 1999. Faunal Extinction in an Island Society: Pygmy Hippopotamus Hunters of Cyprus, New York.

———. 2003. “Preliminary Study of Pastoral Activities among the Pomaks in Greek Thrace,” in Kotjabopoulou et al., eds., 2003, pp. 291–295.

Snyder, L.M., and W.E. Klippel. 1996. “The Cretan Badger (Meles meles) as a Food Resource at Late Bronze/ Early Iron Age Kavousi-Kastro,” in Reese, ed., 1996, pp. 283–293.

———. 2004. “La consumation des lapins pendant le Paléolithique dans la région de Valencia (Espagne) et l’étude des niveaux gravétiens de la Cova de les Cendres (Alicante),” in XXIV rencontres

Spaan, A. 1996. “Hippopotamus creutzburgi: The Case of the Cretan Hippopotamus,” in Reese, ed., 1996, pp. 99–110.

168

MANUEL PÉREZ RIPOLL

Steensma, K., and D. Reese. 1999. “Genet,” in Simmons, ed., 1999, pp. 167–169. Swiny, S., ed. 2001. The Earliest Prehistory of Cyprus: From Colonization to Exploitation (Cyprus American Archaeological Research Institute Monograph Series 2), Boston. Todd, I.A. 2001. “Kalavasos-Tenta Revisited,” in Swiny, ed., 2001, pp. 95–108. Tsoukala, E. 1996. “The Animal Bones from Smari: A Preliminary Report,” in Reese, ed., 1996, pp. 273–275. Uerpmann, H.-P. 1971. Die Tierknochenfunde aus der Talayot-Siedlung von S’illot (San Lorenzo, Mallorca) (Studien über frühe Tierknochenfunde von der Iberischen Halbinsel 2), Munich. Vigne, J.-D. 1987a. “L’exploitation des ressources alimentaires carnées en Corse du VII au IV millénaire,” in Premières communautés paysannes en Méditeranée occidentale. Colloque International du C.N.R.S., Montpellier, 1983, J. Guilaine, J. Courtin, J.-L. Roudil, and J.-L. Vernet, eds., Montpellier, pp. 193–198. ———. 1987b. “L’extinction holocène du fond de peuplement mammalien indigène des îles de Méditerranée occidentale,” Mémoires Société Géologique de France 150, pp. 167–177. ———. 1987c. “L’origine du peuplement mammalien de la Corse: Quelques réflexions biogéographiques,” Bulletin de la Societé Zoologique de France 111, pp. 165–178. ———. 1990. “Biogeographical History of the Mammals on Corsica (and Sardinia) since the Final Pleistocene,” in International Symposium on Biogeographical Aspects of Insularity, Rome, pp. 369–392. ———. 1995a. “Aproximacions arqueozoológiques de la relació home-animal en els territoris insulars: L’exemple mediterrani,” Cota Zero 11, pp. 61–70. ———.1995b. “Faunes sauvages et sociétés humaines sur les îles méditerranéennes: Exemples corses,” Anthropozoologica 21, pp. 41–54. ———. 2000. “Les débuts néolithiques de l’élevage des ongulés au Proche Orient et en Méditerranée: Acquis récents et questions,” in Premiers paysans du monde: Naissances des agricultures, J. Guilaine, ed., Paris, pp. 143–167. ———. 2001. “Large Mammals of Early Aceramic Neolithic Cyprus: Preliminary Results from Parekklisha-Shillourokambos,” in Swiny, ed., 2001, pp. 55–60. Vigne, J.-D., and J.A. Alcover. 1985. “Incidence des relations historiques entre l’homme et l’animal dans la

composition actuelle du peuplement amphibien, reptilien et mammalien des îles de Méditerranée occidentale,” in Actes du 110 Congrès National des Sociétés Savantes, Montpellier, 1985 (Sciences 2), Montpellier, pp. 79–91. Vigne, J.-D., and H. Buitenhuis. 1999. “Les premiers pas de la domestication animale à l’ouest de l’Euphrate: Chypre et l’Anatolie central,” in Les débuts de l’élevage au Proche-Orient: Données nouvelles et réflexions (Paléorient 25 [2]), J.-D. Vigne and M. Mashkour, eds., Paris, pp. 49–62. Vigne, J.-D., I. Carrère, and J. Guilaine. 2003. “Unstable Status of Early Domestic Ungulates in the Near East: The Example of Shillourokambos (Cyprus, IX–VIIIth Millennia cal. b.c.),” in Le néolithique de Chypre (BCH Suppl. 43), J. Guilaine and A. Le Brun eds., Athens, pp. 239–251. Vigne, J.-D., and N. Desse-Berset. 1995. “The Exploitation of Animal Resources in the Mediterranean Islands during the Pre-Neolithic: The Example of Corsica,” in Man and Sea in the Mesolithic: Coastal Settlement above and below Present Sea Level (Oxbow Monograph 53), A. Fischer, ed., Oxford, pp. ­ 309–318. Vigne, J.-D., J. Guilaine, K. Debue, L. Hayne, and P. Gérard. 2004. “Early Taming of the Cat in Cyprus,” Science 304, p. 259. Villaverde, V., ed. 2001. De neandertales a cromañones: El inicio del poblamiento humano en las tierras valencianas, Valencia. von den Driesch, A. 1972. Osteoarchäologische Untersuchungen auf der iberichen Halbinsel (Studien ubre frühe Tierknochenfunde von der iberischen Halbinsel 3), Munich. ———. 1976. A Guide to the Measurement of Animal Bones from Archaeological Sites (Peabody Museum Bulletin 1), Cambridge, MA. von den Driesch, A., and J. Boessneck. 1976. Castro do Zambujal: Die Fauna (Studien ubre frühe Tierknochenfunde von der iberischen Halbinsel 5), Munich. Waldren, W. 1982. Balearic Prehistoric Ecology and Culture: The Excavation and Study of Certain Caves, Rock Shelters and Settlements (BAR-IS 149), Oxford. Weir, S. 1990. The Bedouin, London. Wilkens, B. 1996a. “Faunal Remains from Italian Excavations on Crete,” in Reese, ed., 1996, pp. 241–261. ———. 1996b. “I resti faunistici di Haghia Triada (Creta) in età neo e postpalaziale: Nota preliminaire,” in Tai e memorie del secondo congresso

THE KNOSSOS FAUNA AND THE BEGINNING OF THE NEOLITHIC IN THE MEDITERRANEAN ISLANDS 169

internazionale di micenologia, Roma-Napoli, 14–20 ottobre 1991, E. De Miro, L. Godart, and A. Sacconi, eds., Rome, pp. 1511–1520. ———. 2003. “Hunting and Breeding in Ancient Crete,” in Kotjabopoulou et al., eds., 2003, pp. 85–90.

Zeder, M.A. 2005. “Reconciling Rates of Long Bone Fusion and Tooth Eruption and Wear in Sheep (Ovis) and Goat (Capra),” in Recent Advances in Ageing and Sexing Animal Bone, 9th ICAZ Conference, Durham, 2002, D. Ruscillo, ed., Oxford, pp. 87–118.

9

The Earliest Settlement on Crete: An Archaeozoological Perspective Liora Kolska Horwitz

I found also that the Island I was in was barren and, as I saw good reason to believe, uninhabited, except by wild beasts, of whom, however, I saw none. —Daniel Defoe, 1719, The Life and Strange Surprising Adventures of Robinson Crusoe, of York, Mariner

The maritime capabilities of pre-Neolithic human populations in the Mediterranean are attested by the presence of obsidian from the island of Melos at continental sites and by ephemeral occupations, best defined as landfalls, on insular sites such as AkrotiriAetokremnos on Cyprus, Grava on Corfu, Maroula on Kythnos, the Cave of the Cyclops on Youra, and several sites on Corsica and Sardinia (Cherry

1990; Vigne and Desse-Berset 1995; Simmons 1999; Costa et al. 2003; Mavridis 2003; Broodbank 2006; Masseti 2007).* Permanent settlement on a Mediterranean island is, however, a relatively late phenomenon, with Cyprus providing the earliest evidence. This event, dating to the second half of the ninth millennium cal. b.c., signals the first phase in the assimilation of the Mediterranean seascape

* I would like to express my warm thanks to Nikos Efstratiou, University of Thessaloniki, for inviting me to participate in this publication and for his assistance in obtaining materials from Knossos and information relating to the excavation. I am also grateful to Gila Kahila Bar-Gal, the Hebrew University of Jerusalem, for sharing her research on the DNA of the modern and ancient agrimi of Crete and to Marco Masseti, University

of Florence, for access to the unpublished dates of the pigs from Youra. Part of the research for this paper was supported by the Arts and Humanities Research Council (AHRC) project number AH/D503434/1, awarded to Stephen Shennan, University College London, and Keith Dobney, University of Durham.

LIORA KOLSKA HORWITZ

172

and its islands into the mainland interaction sphere. It was characterized by the introduction of a wide range of techno-cultural items as well as new plant and animal species (Guilaine et al. 2000; Peltenburg et al. 2000, 2001; Peltenburg 2003; Vigne, Carrére, and Guilaine 2003). Crete, the second Mediterranean island to be settled, was first occupied a millennium after Cyprus, at the end of the eighth millennium cal. b.c. (Evans 1964; Warren et al. 1968; Broodbank and Strasser 1991; Efstratiou et al. 2004). This event is generally considered part of the westward demic diffusion of the European Neolithic package, comprising domestic cultivars, herd animals, dogs, and other features (Broodbank and Strasser 1991; van Andel and Runnels 1995). The settlement of Cyprus and Crete differed, as discussed by Isaakidou (2004). The geographic origin of the first Cretans is still unclear. Perlès (2001, 2003) has discussed the similarities in material culture of the Levant and Anatolia with

Aceramic and Early Neolithic (EN) Greece. The authors of a recent genetic study of Y-chromosome sequences concluded that modern Cretan populations resemble those from Anatolia more closely than populations from mainland Greece (King et al. 2008). Consideration of the archaeozoological record may shed additional light on this issue, since the first settlers on Crete brought with them a range of anthropochorous (introduced) animals not previously found on the island (Masseti 2006). Comparison of the species spectrum, the domestic/wild status of the animals in the founder assemblage, and their mode of husbandry and/or hunting with archaeozoological assemblages from mainland Greece, Anatolia, and the Levant may aid in pinpointing the source of the introduced stock and hence the geographic origin of the settlers. The underlying assumption is that, despite cultural filtering, the anthropochorous stock must have originated from a location with a fauna analogous to that introduced onto the island.

Earliest Neolithic Occupation on Crete Crete is the fifth largest island in the Medi­ terranean, covering an area of ca. 8,260 km2. It has a diverse natural environment that includes high mountains in the central and western part, gorges such as Samaria in the southwest, the fertile plateaus of Lasithi, Omalos, and Nidha, a natural lake, and numerous small perennial streams. Its flora is also diverse, with ca. 1,700 species, of which about 10% are endemic (Yale Peabody Museum 2008). It is on the main migration route of birds. There are many endemic invertebrates, including over 40% of land snails and darkling beetles, and two species of endemic vertebrate: the Cretan shrew, Crocidura zimmermanni; and the Cretan spiny mouse, Acomys minous (Sfikas 1989; Masseti 2003; Yale Peabody Museum 2008).

Abbreviations used in this chapter are: aDNA ancient DNA cal. calibrated or calendar years Ch(s). Chapter(s) EN Early Neolithic km kilometers LN Late Neolithic

While occurrences of both Lower Paleolithic and Mesolithic hunter-gatherer utilizations of Crete have now been identified in the Plakias region of southwestern Crete (Strasser et al. 2010), Knossos (Fig. 9.1) remains the earliest site to have yielded evidence for permanent human occupation on the island (Evans 1964, 1971, 1994; Warren et al. 1968; Efstratiou et al. 2004). Stratum X, described by Evans as “the Camp” (Evans 1971), was an Aceramic archaeological horizon found at the base of three small soundings in the Central Court (Trenches X, ZE, and AC) of the Minoan palace. Although the existence and character of an Aceramic or Initial Neolithic period in Greece has been discussed by various researchers (see review in Perlès 2001), most researchers define it

MN mtDNA NISP PPN uncal.

Middle Neolithic mitochondrial DNA number of individual specimens Pre-Pottery Neolithic (with phases PPNA, PPNB, PPNC) uncalibrated

173

THE EARLIEST SETTLEMENT ON CRETE: AN ARCHAEOZOOLOGICAL PERSPECTIVE

15 20

GREECE 16 17 21

18

A N AT O L I A 9

12

11

14

19

10 13 22

25 23 5

6

27 28

24 8

CRETE

7

26

CYPRUS L E V A N T

Mediterranean Sea 4 3 2

N 0

300 km 1

Figure 9.1. Map showing location of sites mentioned in the text: 1. Ashkelon; 2. ‘Ain Ghazal; 3. Atlit Yam; 4. Hagoshrim and Tel Ali; 5. Ras Shamra; 6. Cap Andreas Kastros; 7. Khirokitia; 8. Tenta; 9. Asikli Höyük; 10. Mersin; 11. Can Hasan III; 12. Çatalhöyük; 13. Suberde; 14. Haçilar; 15. Nea Nikomedeia; 16. Argissa-Magula; 17. Sesklo; 18. Achilleion; 19. Franchthi Cave; 20. Sidari, Corfu; 21. Cave of the Cyclops, Youra; 22. Melos; 23. Santorini; 24. Knossos, Crete; 25. Tel Aray 2; 26. Umm el Tlel; 27. Qdeir; 28. El Kowm 2.

as an era with few or no baked clay sherds. At Knossos this occupation was overlain by stratified deposits assigned to EN I and II and later periods. Three radiocarbon dates from Evans’s excavation (8050±180 uncal. b.p., 7480–6540 cal. b.c.; 7910±140 uncal. b.p., 7170–6470 cal. b.c.; 7740±140 uncal. b.p., 7050–6370 cal. b.c.; Perlès 2001, table 5.3; see also Facorellis and Maniatis, this vol., Ch. 10), and a recent radiocarbon date (7050–6690 cal. b.c.) from level 39 of the 1997 excavation (see Facorellis and Manaitis, this vol., Ch. 10) place the Stratum X occupation at the end of the eighth millennium cal. b.c. Architectural features of Stratum X are scanty and include pits and postholes dug into the bedrock. A threshing area lined with a row of wooden stakes and the remains of domesticated cereals (wheat,

barley) and legumes (lentils) attest to the agricultural basis of this pioneer community. Knossos is located 5 km from the coast today, but it was probably farther inland in the Neolithic, given fluctuating sea levels. Its positioning may have been affected by the scarcity of suitable agricultural land on the coast due to the presence of marshes and the contamination of fresh water supplies. The site is located near perennial fresh water sources (the Kairatos stream and springs) and agricultural land; the latter, although not extensive, would have offered suitable alluvial soils for cultivation (Roberts 1979). An interesting techno-cultural feature of the lithic assemblage from the so-called camp is the absence of arrowheads, artifacts generally associated with hunting. Instead, the lithic assemblage contains unretouched flakes, a few retouched or shaped

174

LIORA KOLSKA HORWITZ

pieces, and rare stone axe-heads manufactured of local chert, quartz, and obsidian from Melos,

along with groundstone querns and grinders (Evans 1994; Perlès 2001).

The Cretan Faunal Record The last Pleistocene Cretan fauna (of the Mus biozone) comprised two species of elephant (Elephas creutzburgi and Elephas antiquus), several endemic deer species, murid rodents, an insectivore (Crocidura zimmermanni), an otter species, and birds (Mayhew 1977; Weesie 1988; Lax and Strasser 1992; Jarman 1996; Reese, ed., 1996; Isaakidou 2004). Radiometric dates indicate that there was a hiatus between the earliest human settlers and the Pleistocene fauna (aside from one insectivore). Indeed, it is generally accepted that Crete was devoid of endemic mammals at the time that it was settled in the Aceramic Neolithic (see Isaakidou 2004 for an up-to-date review; cf. Lax and Strasser 1992). This situation may explain why the early settlers brought with them to the island a variety of animals and plants, assuming they had a priori knowledge of the locality, that is, pre-Neolithic visits. Faunal remains recovered from Evans’s excavation of the Camp comprise 445 bones (Table 9.1) identified by Jarman and Jarman (1968; Jarman 1996). This assemblage was composed exclusively of the remains of anthropochorous taxa. As has been discussed at length by numerous researchers (Cherry 1990; Reese, ed., 1996; Masseti 2003; Mavridis 2003), these taxa were new to the island and did not form part of the insular Pleistocene fauna. Indeed, Crete lies outside the natural biogeographical range of all of the anthropochorous species found. The most common animals are caprines, which constituted 75% of the assemblage, with sheep (Ovis) significantly outnumbering goats (Capra). Next in prominence were pigs (Sus), making up 18.3% of the total, and cattle (Bos), representing 6.5% of the total (Table 9.1). Isolated remains of a canid, probably a dog (Canis familiaris), were also identified, along with finds of badger (Meles meles) and hare (Lepus) (Jarman and Jarman 1968; Jarman 1996). Aside from the latter two taxa, all animals were described by Jarman as domestic. A close reading of the original faunal reports (Jarman

and Jarman 1968; Jarman 1996), however, as well as the more recent studies by Pérez Ripoll (this vol., Ch. 8) and Isaakidou (2004), indicates that the domestic versus wild status of these early anthropochorous taxa may have been complex. The animals introduced into Crete have been variously described as domestic, feral, wild, and protodomestic. Domestic animals are defined by changes in their morphology and biometry as specified in the archaeozoological criteria given by Meadow (1989). Ferals are domestic stock that have become freeranging and have reverted in many (but not necessarily all) features to their wild phenotypes, though they still retain a domestic genotype (Kahila BarGal et al. 2002; Isaakidou 2004, 229). Another important category is that of “proto-domestic” forms, defined here as animals that are indistinguishable from wild forms in their phenotype (biometry and morphometry) but which are under some form of cultural management that is manifest in selective culling of age and sex classes (Ducos 1996, 2000; Zohary, Tchernov, and Horwitz 1998). In contrast to proto-domestic forms, wild animals will exhibit mortality profiles and sex ratios characteristic of natural populations. The domestic status of the sheep remains from the Camp was inferred by Jarman and Jarman (1968) from their small size, which was comparable to that of domestic animals from several sites from the Greek mainland (EN Argissa-Magula and Nea Nikomedeia; Fig. 9.1), Britain, and Europe. It must be emphasized that the sample of sheep bones that could be measured from Knossos and the other sites was extremely small and cannot be considered representative of the full size range in a wild or domestic population of mixed sexes. Moreover, the geographic range of animals used for comparison included different climatic zones and chronological periods. Despite these limitations, the sheep appear to be of comparable small size to mainland domesticates and as such are assumed to represent domestic animals.

THE EARLIEST SETTLEMENT ON CRETE: AN ARCHAEOZOOLOGICAL PERSPECTIVE

Jarman (1996, 212) described the Aceramic Neolithic goats from Knossos as domestic, but the original biometric and morphological evidence presented in his earlier paper (Jarman and Jarman 1968, 242) and the conclusions he reached then are less conclusive. Only five bones were positively identified as those of goat. Of these, only two were measured, and these were found to be larger than those of domestic animals from the sixth millennium uncal. b.c. site of Choga Mami in Iraq (Jarman and Jarman 1968, fig. 7). Furthermore, there were no horncores preserved from this layer, remains that may offer conclusive evidence as to domestic/ wild status of the animals. Given the small size of the sample examined, the status of the goats is best left undefined. As noted by Jarman (1996), since only a few goats are present in the assemblage, they would have had little impact on the age structure of the combined caprine sample, which thus reflects primarily that of the sheep. The age profile for the caprine sample indicates an economy geared to meat exploitation, with a peak cull of immature animals, of which few survived into adulthood (Jarman and Jarman 1968, 245). This finding is surprising since it might be expected that the settlers would have safeguarded the founder stock, especially reproductively active females, to ensure herd survival and growth. As discussed by Redding (1981, 185),

175

a strategy aimed at herd security would have as its goal the maximization of the physical survival of the herd and not the production of a marketable surplus. Instead, it appears that the vast majority of animals from Knossos Stratum X were slaughtered for meat while young. The biometric study of the pig remains from the Camp was limited by the small number of teeth. The measurements were compared with Holocene wild boar from European sites and Pleistocene boar from the Lebanese site of Ksar Akil, a disparate group of sites both geographically and temporally. Despite these problems, Jarman concluded that in addition to small pigs there were some extremely large specimens, especially when they were compared to animals from the early Neolithic site of Argissa-Magula on the Greek mainland. As suggested by Jarman and Jarman (1968), this may indicate that the Aceramic Neolithic specimens from Knossos were in an early phase of domestication, or proto-domestic (Table 9.1). Some 73% of the pigs were culled by two years of age. While the mortality patterns demonstrate selective culling of immature pigs, this pattern may fit both hunting and herding strategies for pigs, as their double birth peak produces a population with more immature animals than those of caprines or cattle (Russell and Martin 2005). As for caprines, the emphasis on an immature cull for meat production does not

Animal Species

Aceramic Neolithic “The Camp” (Evans 1971)

EN Ia (Jarman)

EN Ia (Pérez Ripoll)

Suggested Status

Sheep/goat

67

61.2

71.7

domestic + proto-domestic/wild

Sheep

6.5

2.7

7

domestic

Goat

1.1

0.4

Pig

18.3

20.9

8.2

proto-domestic

Cattle

6.5

13.6

11.7

proto-domestic

Dog

0.2

0.9

1.1

domestic

Hare

0.2

—

—

wild/intrusive?

Badger

0.2

—

—

wild/intrusive?

NISP

446

645

85

proto-domestic/wild

Table 9.1. Relative frequencies (percentages) of animal species from Knossos based on NISP counts and the status of the animals as proposed here.

176

LIORA KOLSKA HORWITZ

reflect a management strategy aimed at herd security, since few adult animals were kept for breeding. For Bos, the situation is also complex. Again, the observations were based on a limited number of specimens from Knossos, while the comparative material spanned a wide temporal and geographic range (Jarman and Jarman 1968). As for pigs, Jarman (1996) noted that small-sized “domestic” animals occurred along with several large specimens, the latter comparable in size to aurochs (Bos primigenius). He interpreted this as reflecting “a single, relatively stable population of cattle which included some large individuals” (Jarman 1996, 213). It should be noted that remains of largesized cattle identified as aurochs have been reported from later sites on the island (Jarman 1996, n. 6), although these claims have been dismissed as unlikely by most researchers (Jarman 1996, 215; Isaakidou 2004, 250). Only a few cattle remains from the Camp could be aged, but at least two young animals (less than two years of age) were represented, while three were older than two years (Jarman and Jarman 1968). Despite the small numbers of cattle remains, given their far larger size and meat weight, the importance of cattle in the diet would have been far higher than is suggested by their bone counts. In the overlying EN phases Ia and Ib, Jarman and Jarman (1968; Jarman 1996) reported a spectrum of anthropochorous species similar to that found in the Aceramic (Table 9.1). In these two phases, they reported the presence of badger, canid (possibly dog, 0.9% and 1.9%, respectively, in EN Ia and Ib), domestic cattle (13.6%, 22.7%), pig (20.9%, 13.9%) and caprines (63.9%, 61.5%), with sheep being the most common species in both phases. The goat sample was small but included large animals, the same size as those found in the preceding Aceramic period. As such, their domestic/wild status was not determined. Badger, identified by Jarman (1996) in both Aceramic Neolithic and EN I levels at Knossos, represents an introduced wild species, as does the isolated tooth of a hare from the Aceramic deposits (Table 9.1). Both may represent intrusions from overlying strata; the badger is a burrowing animal, while additional remains of the hare are not found again at the site until the Early Bronze Age (Jarman 1996; Steensma and Reese 1999; Masseti 2003; Pérez Ripoll, this vol., Ch. 8). Remains of badger

occur in all levels at Knossos, however, suggesting that it is in situ. If so, its presence in the earliest levels at Knossos increases the possibility that other wild animals were introduced at this time. The new 1997 excavations at Knossos (Efstratiou et al. 2004; Efstratiou, Karetsou, and Banou, this vol., Ch. 1) did not yield an Aceramic faunal assemblage, but a small sample from early EN I was recovered, comprising 85 identified bones, in addition to remains from later Neolithic deposits. Pérez Ripoll (this vol., Ch. 8), who studied these remains, corroborated many of the initial observations of Jarman and Jarman (1968) concerning EN I. Despite the fact that bones of caprines dominate this assemblage, only remains of sheep were positively identified (Table 9.1). This pattern does not seem to have been related to sample size since, even in the larger faunal samples from Knossos studied by Jarman and Jarman (1968), goats represented a minor faunal component. Both domestic cattle and pig were identified, with cattle having been of greater importance. In the mixed EN/Middle Neolithic (MN), MN, and Late Neolithic (LN) contexts, Pérez Ripoll identified remains of wild boar as well as wild goat on the basis of the large size of the remains. Remains of wild boar have been reported from eighth millennium cal. b.c. deposits at the Cave of the Cyclops on the island of Youra, but a nonanthropogenic colonization with animals swimming to the island has not been ruled out by Masseti (2007, 160). The presence of wild boar at later sites on Crete has not been corroborated (Jarman 1996, n. 18). A reexamination of Evans’s faunal collection from the Camp and the EN Ia stratum was recently undertaken by Isaakidou (2004, 2006). She confirmed Jarman’s conclusion that the sheep and cattle represent domestic animals. For pigs, she concluded that the bimodal size distribution reflects the presence of small-sized domestic animals, while she attributed the large-sized remains to ferals. She reached no conclusion as to the status of the goats based on their biometry, although the mortality profile of the combined sheep/goat sample in the pooled Aceramic–EN Ia periods indicated a typical domestic management strategy, that is, selective culling of young animals (males). She suggested that the surviving adults (ewes) were exploited for their milk (Isaakidou 2004, 268; 2006, 103), but since this sample is composed predominantly of sheep, it does not help to clarify the status

THE EARLIEST SETTLEMENT ON CRETE: AN ARCHAEOZOOLOGICAL PERSPECTIVE

of the goats. Moreover, the fact that it is a pooled sample may obscure the Aceramic Neolithic management strategy. Isaakidou’s analysis of the cattle in a pooled Aceramic–EN Ia sample corroborates the results of both Jarman and Pérez Ripoll. She noted that 30%– 40% of cattle were slaughtered as calves, that is, by two-and-a-half years, with another 10%–25% culled by five years. Thus, few cattle survived into adulthood (Isaakidou 2004, 244). Such a mortality profile demonstrates “considerable potential for production of meat” (Isaakidou 2006, 104). It is unclear, however, given the limited number of mature adult cattle in this sample, why Isaakidou then claims that this mortality profile “is equally compatible with smallscale exploitation for milk and power” (2006, 104). As concluded by all the archaeozoologists who have worked on the Knossos assemblage, in the Aceramic and EN levels, sheep/goat, cattle, and pigs were characterized by a high immature cull, reflecting an exploitation system aimed at meat production. As discussed above, in order to ensure herd security, sufficient numbers of animals—predominantly females—need to be kept for breeding. The expected mortality pattern, as explained by Redding (1981), is based on changes in the reproductive value of caprines; for sheep this value is highest at two years and for goats at three years, and then it drops off steadily. Moreover, the survival rate of ewes falls off after five or six years (Redding 1981). Consequently, the retention of only a few animals beyond this age, as attested to at Knossos, may signal both natural attrition of the adult herd as well as culling of older, less fecund animals. Such practices would, however, affect the size of the breeding pool. Remains of an extremely large-sized goat were found in association with those of small-sized domestic sheep and goats in mixed EN/MN deposits at Knossos and in later assemblages from the island (Pérez Ripoll, this vol., Ch. 8; Jarman and Jarman 1968, 268; Jarman 1996, 213; Wilkens 1996, 241). These remains have been attributed to free-ranging goats called agrimi (Capra aegagrus cretica) that are still found on Crete (Masseti 1998, 2003). Agrimi (Fig. 9.2) are slightly smaller in body size, and hence also in brain size, than the mainland wild bezoar goat (Capra aegagrus), but they closely resemble them in pelage (coat color and patterning), horn form, and behavior (Schultze-Westrum 1963; Papageorgiou 1974; Husband and Davis

177

Figure 9.2. Adult male agrimi ( Capra aegagrus cretica ) showing phenotypic resemblance to the wild bezoar goat.

1984; Groves 1989; Nicholson and Husband 1992). Researchers have alternately assigned them a wild status (e.g., Schultze-Westrum 1963; Mason 1984; Harrison and Bates 1991) or perceived them as feral descendants of the Neolithic domesticates brought to the island (Groves 1989; Masseti 1998, 2003; Kahila Bar-Gal et al. 2002). With respect to the latter possibility, it has been suggested that selective pressures, high inbreeding, and decreased genetic variation acted on escaped domestic Capra to reconstruct a “wild” phenotype—the agrimi (Kahila Bar-Gal et al. 2002). Indeed, in a genetic study of mtDNA from modern agrimi, Kahila Bar-Gal et al. (2002) illustrated that they are closer to domestic than wild goats. This finding may also imply, however, that the original wild goats introduced onto the island subsequently interbred with imported domestic animals (Horwitz and Kahila Bar-Gal 2006). This is the case with the free-ranging boar on Sardinia and Corsica that are thought to have their origins in pigs and wild boar introduced in the seventh millennium b.c. (Albarella et al. 2006). Today, cross-breeding between boar and domestic pigs takes place regularly, and it probably occurred in the past (Albarella et al. 2007). The presence of goats of varying size in the mixed EN/MN and MN assemblages at Knossos has important implications for the domestic status of the earliest goats from the Aceramic and EN I

178

LIORA KOLSKA HORWITZ

levels at the site, as the archaeozoologists concur that the goats belong to large animals of uncertain domestic status (see above). Indeed, in contrast to sheep, the Knossos goats show little diminution in size throughout the EN, an observation that shows it is premature to assume that only fully domestic animals were introduced into Crete. Several attempts were made by Kahila Bar-Gal to extract aDNA from EN goat bones and a tooth from the new excavations at Knossos. Only one bone produced a sequence that proved this specimen was a goat, but its genetic affinity to either wild or domestic goats was not ascertained since the sequence could not be replicated (Kahila BarGal, pers. comm., 2007). In conclusion, although past and present archaeozoological studies of the Aceramic Neolithic and EN Ia assemblages from Knossos have been limited by sample size constraints, especially by the scarcity of well-preserved bones that could be aged, sexed, or measured, there is a broad consensus on most of the results. Based on the work of Jarmon and Jarmon (1968, 257) and Isaakidou (2004) some of the salient points concerning the pioneer Neolithic fauna of Crete may be summarized as follows: 1. With the exception of one insectivore, Crete was devoid of endemic mammals at the time of its settlement in the Aceramic Neolithic. 2. There is no overlap between the earliest human settlers and the last Pleistocene island fauna, so there is no corroboration for supposed anthropogenic involvement in the extinction of Pleistocene fauna on this island. 3. The faunal package found in the Aceramic Neolithic and EN I at Knossos is exotic and contains only new, anthropochorous taxa. 4. Remains of fish and marine mollusks are notably absent from the Aceramic Neolithic and EN Ia levels. This is unlikely to be the result of retrieval methods during excavation (e.g., the lack of fine sieving), since small mammalian bones and even a tooth of a hare were recovered (although, as noted above, the latter may be intrusive from overlying levels).

5. The Aceramic Neolithic and EN I fauna comprised animals of different statuses, including domestic, wild, and feral/wild/ proto-domestic taxa (Table 9.1). That sheep were introduced onto the island as already domesticated animals is firmly established based on biometric comparisons with other continental sites and is further substantiated by cull profiles. As will be discussed below, the preferential cull of young animals raises the question of longterm herd viability and how the supply of surplus animals was maintained. There were two wild taxa in the Aceramic Neolithic deposit—the badger and the hare. Both, however, may be intrusive. Regarding the feral/wild/proto-domestic taxa, Isaakidou (2004, 230) proposed that “in the earlier phases of the ‘feralization’ process, feral animals should be expected to fall within the same size range as the population of domesticates from which they derive.” Only subsequently will they deviate from the founder herd as a result of having experienced different selective pressures under different living conditions. As attested at Knossos, however, from the very inception of the settlement in the Aceramic Neolithic pigs and cattle exhibited a wide size range. Given Isaakidou’s definition of feralization, insufficient time would have elapsed since the importation of these animals onto the island for a distinct phenotypic population (with large body size) to be established, unless pre-Neolithic animal release—an option that is not substantiated—is assumed. Indeed, her results are compatible with the large range of variation expected in proto-domestic animals following the relaxation of selective pressures that operated in the wild (Zohary, Tchernov, and Horwitz 1998). It is also unlikely that people would have transported both wild and domestic forms of the same species simultaneously. The data for goats are ambivalent, and it is likely that they were introduced in a proto-domestic state or as wild animals. It is clear that some of these goats either escaped or were released and subsequently established free-ranging populations that are phenotypically indistinguishable from wild forms, that is, agrimi. It is noteworthy that sheep and cattle at Knossos show a decrease in size over time, while pigs and goats retained a more stable body size. Isaakidou

THE EARLIEST SETTLEMENT ON CRETE: AN ARCHAEOZOOLOGICAL PERSPECTIVE

(2004, 262) attributes this to their less restricted feeding and/or breeding activities. It is suggested here that this may relate to differences in the status of the animals, with pigs and goats, introduced while still proto-domestic, undergoing slower changes than sheep, imported as fully domesticated. Despite their proto-domestic status, cattle would have undergone diminution due to their larger body size, which probably would have made them more susceptible to even minor changes in selective pressures. The subsequent biometric

179

and morphological changes that took place in these proto-domestic animals may be associated with interbreeding with fully domestic forms that were subsequently introduced from the mainland (Ducos 2000). Consequently, in the search for the source of the founder stock on Crete, prime areas to be considered are those where, in the late eighth millennium cal. b.c., sheep were fully domesticated, while goats, pigs, and possibly cattle were still in a proto-domestic state.

Managing the Anthropochorous Fauna The presence, from the very outset, of a complex, well-developed, exotic Neolithic package that included domestic plants as well as wild, domestic, and proto-domestic fauna has led many researchers to conclude that the settlement of Crete was not a casual dispersion resulting from accidental landfall by mariners. Rather, as Perlès (2001, 45) has remarked, it “demonstrates purposeful and planned displacements of populations” by seafaring pioneers (Broodbank and Strasser 1991; Broodbank 2000; Cucchi and Vigne 2006; but see Cherry 1981, 1990, for a different view). An event of such magnitude and ambition as that outlined above would have required a high degree of planning since it entailed not only the transport of people (human population size for the initial settlement of a large island is estimated to have been at least 40 by Cucchi and Vigne 2006) and their material possessions, but also water, food supplies, seeds, and live animals needed to establish an agropastoral economy. Both latter elements would require special care or else they would, respectively, spoil or die. One may even argue that since the island was barren (i.e., depleted of Pleistocene terrestrial fauna), such supplies were essential to make the settlers self-sufficient. Such provisioning would imply prior knowledge, that is, pre-Neolithic landfalls, and it negates the idea that Crete was settled accidentally by mariners who drifted off their course due to winds or currents. A sufficient number of animals in the founder herd to sustain both population growth and slaughter for food from the outset would have been

critical for the success of colonization. In order to assess the numbers involved, Ducos (2000) modeled the growth rate and survivorship of a founder population composed of a few pairs of Capra (ca. 10 animals) introduced into a predator-free island. By setting the sex ratio at birth at 1:1, fecundity rate at 0.9 (based on demography of modern agrimi), and a random cull rate of 70% of young males, he calculated that in order to produce a minimum of 50 slaughtered animals per annum (i.e., less than one animal slaughtered per week), a newly established settlement would have had to wait 25 years or even more before the animals could make a substantial contribution to the diet and still retain a viable breeding population. While a larger initial herd size would have augmented the number of surplus animals that could be consumed, higher mortality rates than those calculated by Ducos (e.g., due to sickness, accidental death, newborn mortality) would have resulted in an even longer time span before the herd could be regularly culled and still remain viable. Even higher mortality rates could have resulted in a total population crash. Another simulation based on data for Middle Eastern caprines was run by Redding (1981), who reached a similar conclusion. He demonstrated that for a founder population of 20 ewes aged two to three years, the population first doubled itself after eight years and thereafter doubled every 11 years, thus increasing only 5.46 times over 24 years (to total 109.2 animals). For goats, a higher growth potential was obtained for the same size founder herd, with the number of animals doubling after five years

180

LIORA KOLSKA HORWITZ

and increasing 15.46 times over the full 24-year period (i.e., to 309.2 animals). The expected annual meat yield from the founder flock of 100 sheep ewes was 497.9 kg (from 48.9 individuals). For the goat does, the yield would be 506.6 kg (from 67.7 individuals). Although both these simulations rely on estimated parameters for the founder population (herd size, sex ratios, cull rate, and age at cull), they are based on documented demographic data for modernday caprine populations. The results demonstrate the potentially low meat yield, and hence the low animal protein, available from small pioneer herds. This is especially pertinent for Crete, where current archaeozoological data point to an island depleted of terrestrial fauna at the time of its settlement (Isaakidou 2004). Given the low meat yield of the first herds, initially the diet of settlers on Crete would have been based primarily on crops and locally available wild foods, while meat would have been a minor component. The lack of any evidence for large-scale exploitation of marine resources (fish or shellfish) or wild birds in the Camp assemblage suggests that, at least at Knossos, they were not consumed. Obviously the inland location of the site may have been a factor in the absence of these taxa. Nevertheless, since the settlement of Crete required maritime skills, it seems unlikely that the early inhabitants lacked the technical know-how for fishing in deep water or otherwise. Indeed, given the pioneer nature of the occupation, it seems strange that a combined terrestrial-marine mode of subsistence was not practiced, since this would have been far more stable and less risky than one focusing only on imported agro-pastoralism (Galili et al. 2002). A similar trend has been reported, however, from the Aeolian Islands north of Sicily and on Sardinia, where the first Neolithic settlers relied mainly on agriculture, with agro-pastoral fishing communities appearing only much later (Castagnino Berrlinghieri 2000–2001). Similarly, on Cyprus the scale of maritime resource exploitation appears to have been of greater intensity in the later Khirokitian phase than in the early Pre-Pottery Neolithic (PPN) B settlements (Desse and Desse-Berset 1989, 1994; Cerron-Cerrasco 2003). Perhaps the arduous and dangerous nature of fishing, as well as its seasonal nature, may have made it a low-priority activity when alternative subsistence forms were available (Galili et al. 2002).

Given our current knowledge of Cretan prehistory, it is most likely that a combination of factors helped to provide sustenance for the newcomers and contributed to the survival and subsequent development of the early settlements. A few possible solutions relating to the low meat yield of the founder herds are discussed briefly below. First of all, it may be suggested that a stock of wild/proto-domestic animals was released on the island and that these animals served as supplementary sources of food. As an analog for Crete, Cyprus offers documentation of the simultaneous introduction of a clearly domestic species (sheep), proto-domestic taxa (goat, pig, and cattle), and wild taxa (fallow deer, cat, and fox). The introduction of fox and cat in Cyprus echoes the introduction of badger in Crete. These are atypical dietary elements and perhaps indicate the portage of animals for other reasons (e.g., cultural symbolism or the keeping of pets). Alternately, these introductions may have been accidental. Fallow deer, although they served as the major protein source on Cyprus throughout the Neolithic and especially in the Late Neolithic and Chalcolithic periods (Croft 2003), appear not to have been domesticated on either Cyprus or the mainland. Vigne (1993) has suggested that such appropriation of wild species by past populations may have been an integral component of the domestication process. The historic use of Aegean islands as natural enclosures for free-ranging domestic and wild animals has been explored by Masseti (1998, 2003). Furthermore, there are many ethnographic and archaeological examples of the deliberate introduction of wild taxa onto islands to serve as “living larders” (Horwitz, Tchernov, and Hongo 2004). These free-ranging animals ensured the availability of fresh food for people while freeing them of management responsibilities. Thus, it is possible that in Crete during the Aceramic and EN Ia periods, domestic sheep and proto-domestic pigs and cattle were herded and raised, while wild goats that had been intentionally released were hunted simultaneously. As discussed by Isaakidou (2004), the absence of large- or medium-sized endemic ungulates on Crete during the Neolithic meant that there was no competition for the introduced animals. This, together with the availability of more browse (arboreal deciduous and evergreen trees) than is found on the island today (Isaakidou 2004, 280), would

THE EARLIEST SETTLEMENT ON CRETE: AN ARCHAEOZOOLOGICAL PERSPECTIVE

have greatly contributed to the initial success of a wild goat population. The scarcity of wild goat remains in the Aceramic and EN I levels at Knossos may then relate to the fact that they could only be obtained through hunting. In addition, it is probable that at least initially wild goats were present in low numbers, having only recently been introduced onto the island, and as such they were rare. This would also account for the absence of lithic projectile points in the early Cretan toolkit (Evans 1964; Warren et al. 1968; Perlès 2001). The settlers probably used other technologies such as nets and traps for hunting. Over time, the size of the agrimi herds increased, making them more readily available, while the introduction of metal arrowheads in the Bronze Age would have increased hunting success. Indeed, the agrimi is frequently portrayed as a hunted animal in Minoan artistic depictions (Porter 1996). A second possibility with regard to the founder herds is that the number of animals imported initially was sufficient to sustain culling. The number of animals introduced at the outset would have been contingent upon parameters that we know little about, such as the size of Neolithic boats and the number of boats and people in the colonizing flotilla. There is an interesting association between the relative importance of the animals in the Aceramic assemblage and their status, perhaps relating to the ease with which animals were captured and transported. Sheep, the only unequivocal domesticate in the assemblage, are the most common animals. The second most common taxa, namely cattle and pigs, were probably proto-domestic and relatively easy to handle for transport, while the least-represented species are the wild animals—goats and perhaps also badger and hare. Certainly the first two of these wild taxa would have been the most difficult animals to control. The most likely age class of animals to have been transported, despite their being hobbled or trussed, would have been that of young individuals due to their smaller size, greater docility, superior reproductive potential, and lower food and water demands. We can conclude that the preferred age of founder animals was less than one year of age (when ca. 90% of adult weight is attained) and immediately after weaning, that is, ca. two to three months for sheep and goats (Redding 1981), ca.

181

four to five months for cattle (Balasse and Tresset 2002), and ca. two months for pigs (de Fredrick and Osborne 1977). There would have been an obvious bias in the selection of the sex ratio, since females, as well as being critical for reproduction, can serve as sources of meat, milk, and wool or hair. A third possibility with regard to the perpetuation of the founder herds is that there was continuous human population movement between Crete and the source locations—continental or insular—of those herds, bringing in more animals on a regular basis. This scenario was proposed by Cherry (1981), who envisaged small-scale settlement of the islands, often accompanied by reciprocal movements. Corroboration of this model, perhaps the most viable for herd maintenance, may be found in the material cultural record of Crete. Ceramics first appear in EN I in an advanced technological state, suggesting that they did not develop in situ during the Aceramic Neolithic (e.g., Manteli 1996; Perlès 2001), while the introduction of new ceramic forms in EN II has been interpreted as attesting to increased contact with the Aegean (Evans 1971). With respect to animals, there is evidence for new faunal arrivals in Crete during the Neolithic, with the marten appearing in the EN II period and the donkey and red deer in LN. The house mouse occurs from ca. 5700 to 4700 b.c. (Jarman and Jarman 1968; Jarman 1996; Cucchi and Vigne 2006). The previous claim for the house mouse at Knossos between 6600 and 5800 b.c. is now considered doubtful, while the 5700–4700 b.c. date is accepted as more probable (Cucchi, Vigne, and Auffray 2005). The introduction of new domestic and wild taxa from various sources continued late into the Minoan period (Masseti 2003; Pérez Ripoll, this vol., Ch. 8). Hence, as outlined by Horwitz, Tchernov, and Hongo (2004), the maintenance of viable populations of domestic animals on an island, at least in the first phase following their arrival, would have been facilitated by the replenishment of those populations from outside sources. This would have been especially important since the age profiles on Crete point to meat production, with the majority of animals slaughtered young and few kept into adulthood. The ongoing influx of animals thus raises the problem of the geographic origin of the Aceramic Neolithic Cretan faunal package.

182

LIORA KOLSKA HORWITZ

Faunal Origins In the search for the geographic origins of the anthropochorous animals of Crete, there are two key considerations in addition to geographical proximity and the obvious chronological prerequisite (namely, that the source sites are coeval with Stratum

X at Knossos). The first criterion pertains to the faunal spectrum; all the animals found on Crete must occur at the sites of proposed origin. The second criterion is that the animals should have the same domestic or wild status as those found on Crete.

Island Stepping-Stones Researchers favoring an Anatolian origin for the settlement of Crete (e.g., Broodbank and Strasser 1991; Evans 1994, 5) have raised the possibility of islands acting as stepping-stones for settlers coming from the east. This is due partly to the large distance between Crete and the Anatolian mainland (ca. 200 km) and partly to the fact that the mountain chain of Crete is not visible from the Greek mainland, but only from two islands— Melos and Santorini (Fig. 9.1). The presence of obsidian from Melos in the Camp assemblage provides tangible evidence for a connection between that island and Crete. There is however, no evidence for Neolithic occupation on any of the Mediterranean islands between Anatolia and Greece predating or contemporaneous with the Aceramic occupation of Knossos (Broodbank 1999; Perlès 2001). For example, a date of 7530–7100 cal. b.c. was obtained for pig/wild boar remains (Masseti 2007) from the Cave of the Cyclops on Youra (Table 9.2; Fig. 9.1), while a goat bone from the same site has been dated to 6327– 6137 cal. b.c. (Masseti 2003), making them both slightly younger than the Aceramic at Knossos. Likewise, the earliest Neolithic (level C base) at the site of Sidari, Corfu (Table 9.2; Fig. 9.1), has yielded incised pottery, sheep, and goat remains dating to 6643–6433 cal b.c. (Perlès 2001, 49). Thus, chronology effectively rules these sites out as candidates for the earliest Cretan stock at present.

But what of Cyprus? This is the first occupied island, with Neolithic habitation dated to 8200/8300– 7800 cal. b.c., or the early PPNB (Guilaine et al. 2000; Vigne et al. 2000). Cyprus has a long record of human settlement, making it an excellent jumping-off point for Cretan colonization. Examination of the archaeozoological record for Cyprus, however, demonstrates that this is improbable. The Khirokitian fauna—those that deriving from the sites of Khirokitia, Cap Andreas Kastros, and the upper levels of Kalavasos-Tenta (Table 9.2; Fig. 9.1)—comprises elements that are not found in the Cretan assemblage at the time of its first settlement—fallow deer, house mouse, fox, cat, and marine fish (Ducos 1981; Davis 1984, 1989, 1994, 2003; Desse 1984; Desse and DesseBerset 1989, 1994; Todd 1998; see Weninger et al. 2006 for radiocarbon dates). More importantly, cattle, a species present on Cyprus in the early PPNB, is not found in the Khirokitian phase, indicating that cattle had become extinct (Croft 2003). Yet, cattle are present in the contemporaneous Cretan founder fauna. Without examining any of the other archaeozoological features, it may be concluded that the available data exclude Cyprus as the source of the Cretan settlers. Nor, at the present time, is there evidence to support an island stepping-stone model for the colonization of Crete.

The Greek Mainland The archaeozoological evidence from the Greek mainland and its relevance to the colonization

of Crete has not previously been discussed in depth. Published data on fauna for the Aceramic

183

THE EARLIEST SETTLEMENT ON CRETE: AN ARCHAEOZOOLOGICAL PERSPECTIVE

Neolithic period on the Greek mainland are currently available for only three sites (Table 9.2; Fig. 9.1): Franchthi Cave (Payne 1975), ArgissaMagula (Boessneck 1962), and Sesklo (Schwartz 1981). They demonstrate the establishment of a well-developed animal economy based on domestic herds, along with the presence of dog. Domestic sheep, as attested by NISP counts, are numerically the most common taxon, followed by domestic cattle and pigs. Remains of domestic goat are, however, scarce. Von den Driesch (1987) attributed this to the small size of the identified caprine assemblages and the inadequate methods used to separate sheep and goats at the time the faunal assemblages were studied. Several researchers (Broodbank and Strasser 1991; Perlès 2001, table 8.2; Isaakidou 2004; Pérez Ripoll, this vol., Ch. 8) have noted the

Cal. b.c.

Crete

Eastern Mediterranean Islands

similarities in the range of taxa and relative frequencies seen at Knossos with those of EN sites such as Nea Nikomedeia (Higgs 1962), Argissa-Magula (Boessneck 1962), Achilleion (Bökönyi 1989), and Franchthi Cave (Payne 1975), all located on the Greek mainland. In sharp contrast to the situation in Crete, however, all herd animals from Aceramic or EN sites on the Greek mainland have unequivocally been identified as fully domesticated. In itself, this observation excludes the region as a source area. The most critical consideration, however, is that the presently available calibrated radiocarbon dates for the Aceramic Neolithic occupations on the Greek mainland (Perlès 2001, table 5.3) are slightly more recent than those obtained for Knossos: the dates for Franchthi range from 7044– 6479 cal. b.c. to 7034–6367 cal. b.c., those for Argissa-Magula range from 7422–6708 cal b.c. to

Mainland Greece

Anatolia

7000–6000 Sesklo Argissa Franchthi

Mersin 6785–6620 to 6463–6192

Levant

6000 Cypriot Khirokitian 7000–5000 Tenta, Khirokitia Cap Andreas Kastros Cyclops Cave 6327–6137 (goat) Knossos “The Camp” 7483–6465 to 7006–6230

Sidari 6643–6433 (sheep/goat)

6500

7000

Central Anatolian sites 7600–6200 Suberde, Çatal Höyük, Hacilar V–VII, Can Hasan III

PPNC sites 7000–6300 ‘Ain Ghazal, Ashkelon, Tel Ali, Atlit Yam, Hagoshrim 6, Qdeir, Umm el Tlel, El Kowm 2

Cyclops Cave 7530–7100 (pig bone)

7500 Cyprus EPPNB 8200–7800

Table 9.2. Schematic representation of the relative chronology (cal. b.c. dates) of sites mentioned in the text.

184

LIORA KOLSKA HORWITZ

6477–6070 cal b.c., and the dates for Sesklo span 6992–6387 cal b.c. to 6352–5959 cal b.c. (Table 9.2). The radiocarbon data, then, demonstrate that

Crete could not have been settled by populations from the Greek mainland.

The Levant The important role played by the Levant in the neolithization of Greece has been discussed by Perlès (2003, 54), who notes the similarities in the archaeological record (i.e., human figurines, stone vessels, and green stone axes). The archaeozoological record may offer further insight on the connections. In the Levant, there are only a limited number of faunal assemblages that are coeval with the earliest occupation on Crete (Table 9.2). They fall within the terminal or Final PPN period, also known as the PPNC, dating to 7000–6300 cal. b.c. (Kuijt and Goring-Morris 2002). During the PPNC, many previously occupied sites contracted or were abandoned, with people dispersing into new localities. The observed shifts in settlement pattern have alternately been attributed to deteriorating ecological conditions or the collapse of the social system (e.g., Rollefson and Köhler-Rollefson 1993; Rollefson 1996; Kuijt 2000). These developments might have created a motive for Levantine PPNC communities to disperse, although the distance to Crete would have placed an important constraint on immigration. Thus, the Levant would have been a viable point of origin only if an “island stepping-stone” model, one which is currently unsupported by the archaeological evidence, is adopted. A consideration of the archaeozoological data from sites in the northern and southern Levant may help in the further evaluation of the Levant as a point of origin. During the PPNC, animal domestication was already well advanced in the northern Levant. At the sites of Umm el Tlel, Qdeir, and El Kowm 2 in the El Kowm Basin (Fig. 9.1), domestic sheep, goat, and cattle (but no pigs) have been reported. At El Kowm 2, the domesticated animals are found together with remains of wild sheep and wild goat (Helmer 1992, 2000; Helmer and Saña 1993). Ras Shamra (Syria) is the only PPN settlement in the northern Levant that is located on the coast. Remains of all four domestic herd animals and dog

were already present in the late PPN level Vc, together with a rich assemblage of fish bones and the tentatively identified remains of aurochs and wild boar (Helmer 1989, 1992). In the sixth millennium b.c. Ceramic Neolithic levels at Tel Aray 2, located farther inland, wild caprines were identified together with remains of aurochs (Hongo 1996), indicating the continued existence of these taxa in the region at least a millennium after the first occupation of Crete. Thus, intersite variation associated with local environmental conditions, rather than chronology, appears to have been responsible for the spectrum of species found at each of the northern Levantine sites. The absence of proto-domestic forms and presence of only fully domestic herd animals during the PPNC eliminates the northern Levant as a candidate for the source of the first Cretan stock. The PPNC settlements in the southern Levant include Atlit Yam and Ashkelon on the coast, Tel Ali, Hagoshrim level 6, and ‘Ain Ghazal in the interior. According to Galili et al. (2002), the PPNC coastal sites represent the first Mediterranean fishing villages since they were engaged in a new mode of subsistence that merged agro-pastoralism with marine resource exploitation. Fish remains were scarce or absent, nonetheless, in inland PPNC sites, demonstrating that the role of marine resources as a dietary element was related to local site catchments. From an archaeozoological perspective, the southern Levantine PPNC sites are only partly suitable as points of origin. They are similar to Knossos in that domestic sheep are the most common species at many sites; however, in contrast to Knossos domestic goats predominate at others (Rollefson and Köhler-Rollefson 1993; Horwitz et al. 1999; Lev-Tov 2000; Galili et al. 2002; Haber, Dayan, and Getzov 2003; Garfinkel et al. 2005). Another point of divergence with Aceramic Knossos is that the PPNC pigs are currently considered wild, while on Crete they appear to represent proto-domestic animals. The status of cattle in most PPNC sites

THE EARLIEST SETTLEMENT ON CRETE: AN ARCHAEOZOOLOGICAL PERSPECTIVE

is unclear, although at some sites small-sized animals have tentatively been identified as domestic (von den Driesch and Wodtke 1997; Horwitz et al. 1999; Haber, Dayan, and Getzov 2003; Garfinkel et al. 2005; Horwitz and Ducos 2005). In conclusion, the coastal southern Levantine PPNC settlements do not fully conform to the

185

criteria expected of the parent population of Crete. Another consideration is the distance between this region and Crete, which probably would have been too great for an uninterrupted journey and would have necessitated landfall. Currently, the archaeological record for eastern Mediterranean islands does not support such a model.

Central Anatolia There are no sites in western or northern Anatolia coeval with Knossos Stratum X, and only one EN site, Mersin-Yumuktepe, is located on the coast (Table 9.2; Fig. 9.1). This site is slightly younger than the earliest occupation on Crete, with dates ranging from 6785–6620 cal. b.c. to 6463–6192 cal. b.c. (Caneva 1999). Mersin had a well-developed agro-pastoral economy based on domestic sheep and goats, followed by cattle and then pigs in importance (Buitenhuis and Caneva 1998). The assemblage comprises marine mollusks and fish as well as low frequencies of small mammals, reptiles, and birds. There are several sites in central Anatolia that roughly span the occupation of Aceramic Knossos and may serve as suitable candidates for places of origin. They include Hacilar V–VII, Can Hasan III, Suberde, and the upper levels of Çatalhöyük (Martin, Russell, and Carruthers 2002; Russell and Martin 2005; see individual site dates in Weninger et al. 2006). It should be cautioned that with the exception of the recent investigations at Çatalhöyük, the archaeozoology of these sites presents an unbalanced picture, as most assemblages were studied over 30 years ago when limited comparative material was available. Moreover, the reports often lack data on biometry, morphology, or age profiles. The Aceramic level at Can Hasan III has commonly been associated with Crete in the archaeological literature due to three common factors: the use of unfired mudbricks as building material, the presence of bread wheat (disputed by Isaakidou 2004, 62), and the supposed evidence for domestic cattle in both assemblages (Evans 1994). The last claim cannot be corroborated since the preliminary publication on the Can Hasan III fauna (Payne 1972) provides neither quantitative data nor other information on the domestic status of

the main animals. The report states that the most common taxon was cattle, followed by sheep/goat (not distinguished by species). Also present were pig/wild boar, equids (onager), deer, and canids. In contrast, at both Hacilar and Suberde all taxa aside from the dog were identified as wild (Perkins and Daly 1968; Westley 1970). Martin, Russell, and Carruthers (2002) have recently suggested that the sheep at Suberde were proto-domestic, while the status of the other herd animals is uncertain. The larger, more recently analyzed faunal samples from Aşikli Höyük and Çatalhöyük are presently the key assemblages for central Anatolia, although the former is a mid PPNB site and hence far earlier than the earliest settlement at Knossos. Buitenhuis (1997) concluded that the caprines at this site were not domestic animals, but they were in a proto-domestic stage. The domestic status of pigs and cattle at Aşikli Höyük remains uncertain, but they are morphologically and biometrically indistinguishable from wild animals. The assemblage most comparable to Knossos is that from phase 3 at the late PPN site of Çatalhöyük (Russell and Martin 2000, 2005; Martin, Russell, and Carruthers 2002; Russell, Martin, and Buitenhuis 2005). Caprines dominate the Çatalhöyük assemblage, with sheep consistently outnumbering goats. They are smaller than the proto-domestic sheep from Aşikli Höyük, indicating their domestic status. The biometric analysis of goats was less conclusive as it was based on a relatively small bone sample (n=25). Few wild sheep but numerous wild goat horncores were found at the site. This disparity may demonstrate differences in their status. Russell and Martin (2005, 78) note that the extent of genetic isolation of goats cannot be determined. It is thus possible that goats at Çatalhöyük were still in an early stage of domestication (proto-domestic).

186

LIORA KOLSKA HORWITZ

In phases 2 and 3 at Çatalhöyük there is a shift in the caprine sex ratio to one dominated by females, and there is a heavy cull of juveniles through subadult age classes, with few old adults represented (Russell and Martin 2005). This pattern is interpreted to represent a strategy geared toward meat production. Given the predominance of sheep in the sample, this may more specifically reflect the management of sheep rather than of goats. Based on their biometry, cull patterns, and isotopic values (Pearson et al. 2007), the cattle from Çatalhöyük have been identified as wild aurochs. The site of Çatalhöyük is unique in that it contains a large array of animal remains used predominantly for symbolic or ritual purposes (Russell and Martin 2005). Consequently, this may have biased the composition of the faunal assemblage with selection of special size and/or age classes. It is unclear what physical differences are to be expected in proto-domestic cattle as opposed to aurochs, unless foddering is assumed. Bearing in mind, however, that distinguishing between proto-domestic and wild animals relies not on biometry but on the interpretation of cull patterns and sex ratios, it is proposed that the Çatalhöyük cattle data correspond well to what would be expected in proto-domestication. Moreover, the clear increase in the proportion of females and subadult aurochs in phase 3 at the site may be taken as an indication

of their proto-domestic status rather than as a shift in the social structure and timing of hunting and salt lick availability, as proposed by Russell and colleagues (Russell and Martin 2005; Russell, Martin, and Buitenhuis 2005). Only remains of wild boar have been reported from Çatalhöyük, and this identification is based on biometry. Since sample sizes of measured pig remains for the site are extremely small, however, this identification should probably be treated with some caution. In general, aside from sheep, the Anatolian record lacks evidence for fully domestic herd animals in the period contemporaneous with Aceramic Knossos. The earliest examples of other domesticates are found at the later site of Mersin. Currently, Çatalhöyük offers the closest parallels for Knossos in terms of the faunal elements represented, that is, the combination of domestic sheep, wild sheep and goats, and probably proto-domestic cattle, pig(?), and goat(?). Unfortunately, there is no detailed information available for the sites of Hacilar, Suberde, and Can Hasan III, but it is likely that animals at these sites were also in a proto-domestic stage (Martin, Russell, and Carruthers 2002). If central Anatolian sites (or ones like them nearer the coast) furnished the pioneer stock for Crete, one would expect to find a mixture of domestic, wild, and proto-domestic taxa in Crete.

Conclusion Examination of the evidence from the four potential source regions (other Mediterranean islands, the Greek mainland, the Levant, and Anatolia) illustrates the value of using faunal data to identify the starting point for the animals found in Knossos Stratum X. Given the present status of our knowledge, Anatolia offers the best match, since it is the only region where fully domesticated sheep are found together with proto-domestic and/or wild cattle, pig, and goats. This corroborates known affinities in the material culture (Perlès 2001, 2003), as well as recent genetic data (King et al. 2008) that points to Anatolia as the area of origin of the Cretan settlers. It is possible that the impetus for migrating out of this region came from deteriorating

climatic conditions that culminated in 8200 cal. b.p. (Weninger et al. 2006). At this time, occupation at many Anatolian sites was disrupted and sites were abandoned. Loss of the subsistence base due to ecological stress and/or increasing competition for access to reduced resources (whether associated or not with climate change) may have led people to move to islands in order to increase or even to maintain their food supplies. From this perspective, the occupation of Crete echoes many features of the earlier settlement of Cyprus (Bar-Yosef 2003; Davis 2003; Peltenburg 2003, 2004). As noted by Isaakidou (2004) and Pérez Ripoll (this vol., Ch. 8), the Cypriot Neolithic record is an expression of the initial development and spread of

THE EARLIEST SETTLEMENT ON CRETE: AN ARCHAEOZOOLOGICAL PERSPECTIVE

the Near Eastern Neolithic package, while the subsequent movement to Crete is part of the westward demic diffusion of the Neolithic complex from the Near East into western Europe (van Andel and Runnels 1995). In both events, anthropochorous fauna played an integral role. Due to the absence of large- and medium-sized Pleistocene mammals, the food supply on the islands would have been tenuous without anthropochorous dietary resources, and certainly the rate and scale of development, as well as duration of human settlement, would have been jeopardized. Together with a spectrum of cultivated plants, the transportation of wild, domestic, and proto-domestic animals allowed the pioneering population to recreate a familiar environment, mold the new one, and establish

187

a viable and independent subsistence base. Once established, the mixed agro-pastoral economy would have ensured a varied diet as well as the production of surpluses—crops that could be stored and animals that could serve as living larders. These resources would have augmented the economic security of the settlers. In the same manner as the introduction of the horse changed the history of the North American continent, so the anthropochorous faunal and floral “package” of Crete molded its history. The study of the nature and composition of this package offers key information that has the potential to expand and deepen our understanding of the colonization process and its origin.

References Albarella, U., F. Manconi, P. Rowley-Conwy, and J.-D. Vigne. 2006. “Pigs of Corsica and Sardinia: A Biometrical Re-evaluation of Their Status and History,” in Studi di archeozoologia in onore di Alfredo Riedel, U. Tecchiati and B. Sala, eds., Bolzano, pp. 285–302. Albarella, U., F. Manconi, J-D. Vigne, and P. RowleyConwy. 2007. “Ethnoarchaeology of Pig Husbandry in Sardinia and Corsica,” in Pigs and Humans: 10,000 Years of Interaction, U. Albarella, K. Dobney, A. Ervynck, and P. Rowley-Conwy, eds., Oxford, pp. 285–307. Balasse, M., and A. Tressset. 2002. “Early Weaning of Neolithic Domestic Cattle (Bercy, France) Revealed by Intra-tooth Variation in Nitrogen Isotope Ratios,” JAS 29, pp. 853–859. Bar-Yosef, O. 2003. “Away from Home: Prehistoric Colonizations, Exchanges and Diffusions in the ­ Mediterranean Basin,” in Vandermeersch, ed., 2003, pp. 71–79. Boessneck, J. 1962. “Die Tierreste aus der ArgissaMagula vom präkeramischen Neolithikum bis zur mittleren Bronzezeit,” in Die deutschen Ausgrabungen ­ auf der Argissa-Magula in Thessalien I: Das präkeramische Neolithikum sowie die Tier- und Pflandzenreste (Beitrage zur ur-und frühgeschichtlichen Archäologie des Mittelmeer-Kulturraumes 23), V. Milojčić, J. Boessneck, and M. Hopf, Bonn, pp. 27–99. Bökönyi, S. 1989. “Animal Remains,” in Achilleion: A Neolithic Settlement in Thessaly, Greece, 6400–5600

bc (Monumenta Archaeologica 14), M. Gimbutas, S. Win, and D. Shimabuku, eds., Los Angeles, pp. 315–332.

Broodbank, C. 1999. “Colonization and Configuration in the Insular Neolithic of the Aegean,” in Neolithic Society in Greece (Sheffield Studies in Aegean Archaeology 2), P. Halstead, ed., Sheffield, pp. 15–41. ———. 2000. An Island Archaeology of the Early Cyclades, Cambridge. ———. 2006. “The Origins and Early Development of Mediterranean Maritime Activity,” JMA 19, pp. 199–230. Broodbank, C., and T.F. Strasser. 1991. “Migrant Farmers and the Neolithic Colonization of Crete,” Antiquity 65, pp. 233–245. Buitenhuis, H. 1997. “Aşikli Höyük: A ‘ProtoDomestication Site,’ Anthropozoologica 25–26, pp. 655–662. Buitenhuis, H., and I.V. Caneva. 1998. “Early Animal Breeding in South-Eastern Anatolia: MersinYumuktepe,” in Man and the Animal World. Studies in Archaeozoology, Archaeology, Anthropology and Palaeolinguistics in Memoriam of Sándor Bökönyi, P. Anreiter, L. Bartosiewicz, E. Jerem, and W. Meid, eds., Budapest, pp. 121–130. Caneva, I. 1999. “Early Farmers on the Cilician Coast: Yumuktepe in the Seventh Millennium bc,” in Neolithic in Turkey: The Cradle of Civilization. New

188

LIORA KOLSKA HORWITZ

Discoveries, M. Ozdogan and N. Basgeden, eds., Istanbul, pp. 105–114. Castagnino Berrlinghieri, E.F. 2000–2001. The Aeolian Islands: Crossroads of the Mediterranean Maritime Routes of the South Tyrrhenian, Ph.D. diss., University of Bristol. Cerron-Cerrasco, R. 2003. “Fish Remains,” in The Colonisation and Settlement of Cyprus: Investigations at Kissonerga-Mylouthkia, 1976–1999 (SIMA 70 [4]), E. Peltenburg, ed., Sävedalen, pp. 81–82. Cherry, J.F. 1981. “Pattern and Process in the Earliest Colonization of the Mediterranean Islands,” PPS 47, pp. 42–68. ———. 1990. “The First Colonization of the Mediterranean Islands: A Review of Recent Research,” JMA 3, pp. 145–221. Costa, L.-J., J.-D. Vigne, H. Bocherens, N. DesseBerset, F. Heins, F. de Lanfranchi, J. Magdeleine, M.-P. Ruas, S. Thiebault, and C.A. Tosí. 2003. “Early Settlement on Tyrrehnian Islands (8th Millennium cal. bc): Mesolithic Adaptation to Local Resources in Corsica and Northern Sardinia,” in Mesolithic on the Move. Papers Presented at the Sixth International Conference on the Mesolithic in Europe, Stockholm, 2000, L. Larsson, H. Krindgen, K. Knutsson, D. Loeffler, and A. Åkerlund, eds., Oxford, pp. 3–10. Croft, P. 2003. “The Animal Bones,” in The Colonisation and Settlement of Cyprus: Investigations at Kissonerga-Mylouthkia, 1976–1996 (SIMA 70 [4]), E. Peltenburg, ed., Sävedalen, pp. 225–237. Cucchi, T., and J.-D. Vigne. 2006. “Origin and Diffusion of the House Mouse in the Mediterranean,” Human Evolution 21, pp. 95–106. Cucchi, T., J.-D. Vigne, and J.-C. Auffray. 2005. “First Occurrence of the House Mouse (Mus musculus domesticus Schwartz & Schwartz, 1943) in Western Mediterranean: A Zooarchaeological Revision of the Sub-Fossil House Mouse Occurrences Using a Zooarchaeological Critical Grid,” Biological Journal of the Linnean Society 84, pp. 429–445. Davis, S.J.M. 1984. “Khirokitia and Its Mammal Remains: A Neolithic Noah’s Ark,” in Fouilles récentes à Khirokitia (Chypre) 1977–1981 (Études néolithiques 41), A. Le Brun, ed., Paris, pp. 147–162. ———. 1989. “Some More Animal Remains from the Aceramic Neolithic of Cyprus,” in Fouilles récentes à Khirokitia (Chypre) 1983–1986 (Études néolithiques 81), A. Le Brun, ed., Paris, pp. 189–221. ———. 1994. “Even More Bones from Khirokitia: The 1988–1991 Excavations,” in Fouilles récentes à Khirokitia (Chypre) 1988–1991 (Études néolithiques), A. Le Brun, ed., Paris, pp. 305–333.

———. 2003. “The Zooarchaeology of Khirokitia (Neolithic-Cyprus), Including a View from the Mainland,” in Le néolithique de Chypre (BCH Suppl. 43), J. Guilaine and A. Le Brun, eds., Athens, pp. 253–268. de Fredrick, D.F., and H.G. Osborne. 1977. “Pig Production in the Solomon Islands III: The Influence of Breed, Diet and Housing on Reproduction and Growth,” Tropical Animal Health and Production 9, pp. 203–210. Desse, J. 1984. “L’ichthyofaune du site néolithique de Khirokitia (Chypre),” in Fouilles récentes à Khirokitia (Chypre) 1977–1981 (Études neolithiques 41), A. Le Brun, ed., Paris, pp. 167–168. Desse, J., and N. Desse-Berset. 1989. “Les poissons de Khirokitia (campages 1983, 1984 et 1986),” in Fouilles récentes à Khirokitia (Chypre) 1983–1986 (Études neolithiques 81), A. Le Brun, ed., Paris, pp. 223–233. ———. 1994. “Osteometry and Fishing Strategies at Cap Andreas Kastros (Cyprus, 8th millennium bp),” in Fish Exploitation in the Past (Annales du Musée Royal de l’Afrique Centrale, Sciences Zoologiques 274), W. Van Neer, ed., Tervuren, pp. 69–79. Ducos, P. 1981. “Notes sur les grands mammifères du cap Andreas-Kastros,” in Un Site néolithique précéramique en Chypre: Cap Andreas–Kastros (Recherche sur les grandes civilisations 5), A. Le Brun, ed., Paris, p. 89. ———. 1996. “La faune de Khirokitia (Chypre) et la diffusion des prédomestications proche-orientales vers la méditerranée occidentale,” in Vandermeersch, ed., 2003, pp. 41–50. ———. 2000. “The Introduction of Animals by Man in Cyprus: An Alternative to the Noah’s Ark Model,” in Mashkour et al., eds., 2000, pp. 74–82. Efstratiou, N., A. Karetsou, E. Banou, and D. Margomenou. 2004. “The Neolithic Settlement of Knossos: New Light on an Old Picture,” in Knossos: Palace, City, State. Proceedings of the Conference in Herakleion Organised by the British School at Athens and the 23rd Ephoreia of Prehistoric and Classical Antiquities of Heraklion, in November 2000, for the Centenary of Sir Arthur Evans’s Excavations at Knossos (BSA Studies 12), G. Cadogan, E. Hatzaki, and A. Vasilakis, eds., London, pp. 39–51. Evans, J.D. 1964. “Excavations in the Neolithic Settlement of Knossos, 1957–60. Part I,” BSA 59, pp. 132–240. ———. 1971. “Neolithic Knossos: The Growth of a Settlement,” PPS 37, pp. 95–117.

THE EARLIEST SETTLEMENT ON CRETE: AN ARCHAEOZOOLOGICAL PERSPECTIVE

———. 1994. “The Early Millennia: Continuity and Change in a Farming Settlement,” in Knossos: A Labyrinth of History. Papers in Honour of S. Hood, D. Evely, H. Hughes-Brock, and N. Momigliano, eds., Oxford, pp. 1–20. Galili, U., B. Rosen, A. Gopher, and L.K. Horwitz. 2002. “The Origin and Dispersion of the Eastern Mediterranean Fishing Village: Evidence from Submerged Neolithic Settlements off the Carmel Coast, Israel,” JMA 15, pp. 167–198. Garfinkel, Y., D. Dag, B. Hesse, P. Wapnish, D. Rookis, G. Hartman, O. Bar-Yosef, D.E. Mayer, and O. Lernau. 2005. “Neolithic Ashkelon: Meat Processing and Early Pastoralism on the Mediterranean Coast,” Eurasian Prehistory 3, pp. 43–72. Groves, C.P. 1989. “Feral Animals of the Mediterranean Islands: Documents of Early Domestication,” in The Walking Larder: Patterns of Domestication, Pastoralism and Predation (One World Archaeology 2), J. Clutton-Brock, ed., London, pp. 46–58. Guilaine, J., F. Briois, J.-D. Vigne, and I. Carrère. 2000. “Découverte d’un Néolithique précéramique ancien chypriote (fin 9e, début 8e millénaires cal. bc), apparenté au PPNB ancien/moyen du Levant nord,” Comptes Rendus d l’Académie des Sciences, series IIA, Sciences de la terre et des planètes 330, pp. 75–82. Haber, A., T. Dayan, and N. Getzov. 2003. “Pig Exploitation at Hagoshrim: A Prehistoric Site in the Southern Levant,” in First Steps of Animal Domestication, J.-D. Vigne, J. Peters, and D. Helmer, eds., Oxford, pp. 80–85. Harrison, D.L., and P.J. Bates. 1991. Mammals of Arabia, 2nd ed., Seven Oaks, Kent. Helmer, D. 1989. “Le développement de la domestication au Proche-Orient de 9500 à 7500 b.p.: Les nouvelles données d’El Kowm et de Ras Shamra,” in Préhistoire du Levant: Processus des changements culturels (Paléorient 15 [1]), O. Aurenche, M.-C. Cauvin, and P. Sanlaville, eds., Paris, pp. 111–121. ———. 1992. La domestication des animaux par les hommes préhistoriques, Paris. ———. 2000. “Étude de la faune mammalienne d’El Kowm 2 (Syrie),” in El Kowm 2: Une île dans le désert. La fin du néolithique préceramique dans la steppe syrienne, D. Stordeur, ed., Paris. Helmer, D., and M. Saña. 1993. “Etude de la faune PPNB d’Umm el Tlel (El Kowm, Syrie), Fouilles M. Molist,” Cahiers de l’Euphrate 7, pp. 93–105. Higgs, E.S. 1962. “Fauna,” in Excavations at the Early Neolithic Site at Nea Nikomedeia, Greek

189

Macedonia, 1961 Season (PPS 28), R.J. Rodden, ed., pp. 271–274. Hongo, H. 1996. “Faunal Remains from Tell Aray 2, Northern Syria,” Paléorient 22, pp. 125–144. Horwitz, L.K., and P. Ducos. 2005. “Counting Cattle: Trends in Neolithic Bos Frequencies from the Southern Levant,” Revue de Paléobiologie 10, pp. 209–224. Horwitz, L.K., and G. Kahila Bar-Gal. 2006. “The Origin and Genetic Status of Insular Caprines in the Eastern Mediterranean: A Case Study of Free-Ranging Goats (Capra aegagrus cretica) on Crete,” Human Evolution 21, pp. 123–138. Horwitz, L.K., E. Tchernov, P. Ducos, C. Becker, A. von den Driesch, L. Martin, and A. Garrard. 1999. “Animal Domestication in the Southern Levant,” Paléorient 25, pp. 63–80. Horwitz, L.K., E. Tchernov, and H. Hongo. 2004. “The Domestic Status of the Early Neolithic Fauna of Cyprus: A View from the Mainland,” in Neolithic Revolution: New Perspectives on Southwest Asia in Light of Recent Discoveries on Cyprus (Levant Suppl. 1), E. Peltenburg and A. Wasse, eds., Oxford, pp. 35–48. Husband, T.P., and P.B. Davis. 1984. “Ecology and Behaviour of the Cretan Agrimi,” Canadian Journal of Zoology 62, pp. 411–420. Isaakidou, V. 2004. Bones from the Labyrinth: Faunal Evidence for the Management and Consumption of Animals at Neolithic and Bronze Age Knossos, Crete, Ph.D. diss., Univeristy College London. ———. 2006. “Ploughing with Cows: Knossos and the ‘Secondary Products Revolution,’” in Animals in the Neolithic of Britain and Europe, D. Sarjeantson and D. Field, eds., Oxford, pp. 95–112. Jarman, M.R. 1996. “Human Influence in the Development of the Cretan Mammalian Fauna,” in Reese, ed., 1996, pp. 211–229. Jarman, M.R., and H.N. Jarman. 1968. “The Fauna and Economy of Early Neolithic Knossos,” in “Knossos Neolithic, Part II,” P. Warren, M.R. Jarman, N.J. Shackleton, and J.D. Evans, BSA 63, pp. 241–264. Kahila Bar-Gal, G., P. Smith, E. Tchernov, C. Greenblatt, P. Ducos, A. Gardeisen, and L.K. Horowitz. 2002. “Genetic Evidence for the Origin of the Agrimi Goat (Capra aegegrus cretica),” Journal of Zoology 256, pp. 369–377. King, R.J., S.S. Özcan, T. Carter, E. Kalfoglou, C. Triandaphyllidis, A. Kouvasi, A.A. Lin, C.-E.T. Chow, L.A. Zhivotovsky, M. Michalodimitrakis, and P.A.

190

LIORA KOLSKA HORWITZ

Underhill. 2008. “Differential Y-Chromosome Anatolian Influences on the Greek and Cretan Neolithic,” Annals of Human Genetics 72, pp. 205–214.

(BSA Studies 9) E. Kotjabopoulou, Y. Hamilakis, P. Halstead, C. Gamble, and P. Elefanti, eds., London, pp. 65–74.

Kuijt, I. 2000. “People and Space in Early Agricultural Villages: Exploring Daily Lives, Community Size and Architecture in the Late Pre-Pottery Neolithic,” JAnthArch 19, pp. 75–102.

Mayhew, D. 1977. “The Endemic Pleistocene Murids of Crete, I and II,” Koninklijke Nederlandse Akademie van Wetenschappen B 80, pp. 182–214.

Kuijt, I., and A.N. Goring-Morris. 2002. “Foraging, Farming and Social Complexity in the Pre-Pottery Neolithic of the Southern Levant: A Review and Synthesis,” Journal of World Prehistory 16, pp. 361–440. Lax, E.L., and N. Strasser. 1992. “Early Holocene Extinctions on Crete: The Search for the Cause,” JMA 5, pp. 203–224. Lev-Tov, J. 2000. “Late Prehistoric Faunal Remains from New Excavations at Tel Ali (Northern Israel),” in Mashkour et al., eds., 2000, pp. 208–217. Manteli, K. 1996. “Pottery. Crete, ” in Neolithic Culture in Greece, G.A. Papathanassopoulos, ed., Athens, pp. 132–134. Martin, L., N. Russell, and D. Carruthers. 2002. “Animal Remains from the Central Anatolian Neolithic,” in The Neolithic of Central Anatolia: Internal Developments and External Relations during the 9th–6th Millennia cal. bc, F. Gerard and L. Thissen, eds., Istanbul, pp. 193–206. Mashkour, M., A.M. Choyke, H. Buitenhuis, and F. Poplin, eds. 2000. Archaeozoology of the Near East IV (ARC-Publicatie 32), Groningen. Mason, I.L. 1984. Evolution of Domesticated Animals, London. Masseti, M. 1998. “Holocene Endemic and Anthropochorous Wild Mammals of the Mediterranean Islands,” Anthropozoologia 28, pp. 3–20. ———. 2003. “Holocene Endemic and non Endemic Mammals of the Aegean Islands,” in Zooarchaeology in Greece: Recent Advances (BSA Studies 9), E. Kotjabopoulou., Y. Hamilakis, P. Halstead, C. Gamble, and P. Elefanti, eds., London, pp. 53–63. ———. 2006. “Domestic Fauna and Anthropochorous Fauna,” Human Evolution 21, pp. 85–93. ———. 2007. “The Economic Role of Sus in Early Human Fishing Communities,” in Pigs and Humans: 10,000 Years of Interaction, U. Albarella, K. Dobney, A. Ervynck, and P. Rowley-Conwy, eds., Oxford, pp. 156–170. Mavridis, F. 2003. “Early Island Archaeology and the Extinction of Endemic Fauna in the Eastern Mediterranean: Problems of Interpretation and Methodology,” in Zooarchaeology in Greece: Recent Advances

Meadow, R.H. 1989. “Osteological Evidence for the Process of Animal Domestication,” in The Walking Larder: Patterns of Domestication, Pastoralism and Predation (One World Archaeology 2), J. CluttonBrock, ed., London, pp. 80–90. Nicholson, M.C., and T.P. Husband. 1992. “Diurnal Behaviour of the Agrimi, Capra aegagrus,” Journal of Mammology 73, pp. 135–142. Papageorgiou, M. 1974. Population Energy Relationships of the Agrimi (Capra aegagrus cretica) on Theodorou Island, Greece, Ph.D. diss., University of Michigan, Ann Arbor. Payne, S. 1972. “Can Hasan III, the Anatolian Aceramic and the Greek Neolithic,” in Papers in Economic Prehistory I, E.S. Higgs, ed., Cambridge, pp. 191–194. ———. 1975. “Faunal Change at Franchthi Cave from 20,000 bc to 3,000 bc,” in Archaezoological Studies, A.T. Clason, ed., Amsterdam, pp. 120–131. Pearson, J.A., H. Buitenhuis, R.E.M. Hedges, L. Martin, N. Russell, and K.C. Twiss. 2007. “New Light on Early Caprine Herding Strategies from Isotope Analysis: A Case Study from Neolithic Anatolia,” JAS 34, pp. 2170–2179. Peltenburg, E. 2003. The Colonisation and Settlement of Cyprus: Investigations at Kissonerga-Mylouthkia, 1976–1996 (SIMA 70 [4]), Sävedalen. ———. 2004. “Cyprus: A Regional Component of the Levantine PPN,” Neo-Lithics: The Newsletter of Southwest Asian Neolithic Research 1 (04), pp. 3–7. Peltenburg, E., S. Colledge, P. Croft, A. Jackson, C. McCartney, and M.A. Murray. 2000. “Agro-Pastoralist Colonization of Cyprus in the 10th Millennium bp: Initial Assessments,” Antiquity 74, pp. 844–853. ———. 2001. “Neolithic Dispersals from the Levantine Corridor: A Mediterranean Perspective,” Levant 33, pp. 36–64. Perkins, D., and P. Daly. 1968. “A Hunters’ Village in Neolithic Turkey,” Scientific American 219, pp. 96–106. Perlès, C. 2001. The Early Neolithic in Greece: The First Farming Communities in Europe, Cambridge. ———. 2003. “Le rôle du Proche-Orient dans la néolithisation de la Grèce,” in Vandermeersch, ed., 2003, pp. 91–104.

THE EARLIEST SETTLEMENT ON CRETE: AN ARCHAEOZOOLOGICAL PERSPECTIVE

Porter, R. 1996. “The Cretan Wild Goat (Capra aegagrus cretica) and the Theran Antelopes,” in Reese, ed., 1996, pp. 295–315. Redding, R.W. 1981. Decision Making in Subsistence Herding of Sheep and Goats in the Middle East, Ph.D. diss., University of Michigan, Ann Arbor. Reese, D.S., ed. 1996. Pleistocene and Holocene Fauna of Crete and Its First Settlers (Monographs in World Archaeology 28), Madison. Roberts, N. 1979. “The Location and Environment of Knossos,” BSA 74, pp. 231–240. Rollefson, G.O. 1996. “The Neolithic Devolution: Ecological Impact and Cultural Compensation at ‘Ain Ghazal, Jordan,” in Retrieving the Past. Essays on Archaeological Research and Methodology in Honor of Gus W. Van Beek, J.D. Seger, ed., Winona Lake, Ind., pp. 219–229. Rollefson, G.O., and I. Köhler-Rollefson. 1993. “PPNC Adaptations in the First Half of the 6th Millennium b.c.,” Paléorient 19, pp. 33–42. Russell, N., and L. Martin. 2000. “Neolithic Çatalhöyük: Preliminary Zooarchaeological Results from the Renewed Excavations,” in Mashkour et al., eds., 2000, pp. 164–170. ———. 2005. “Çatalhöyük Mammal Remains,” in Inhabiting Çatalhöyük: Reports from the 1995–1999 Seasons (British Institute of Archaeology at Ankara Monograph 38), I. Hodder, ed., pp. 33–98. Russell, N., L. Martin, and H. Buitenhuis. 2005. “Cattle Domestication at Çatalhöyük Revisited,” CurrAnthr 46, pp. 101–108. Schultze-Westrum, T. 1963. “Die Wildziegen der ägäischen Inseln,” Saugetierkundliche Mitteilungen 11, pp. 145–182. Schwartz, C.A. 1981. “The Fauna from Early Neolithic Sesklo,” in The Early Neolithic I Settlement at Sesklo: An Early Farming Community in Thessaly, Greece (Analecta Praehistorica Leidensia 14), M. Wijnen, ed., Leiden, p. 112. Sfikas, G. 1989. Birds and Mammals of Crete, Athens. Simmons, A.H. 1999. Faunal Extinction in an Island Society: Pygmy Hippopotamus Hunters of Cyprus, New York. Steensma, K. and D. Reese. 1999. “Genet,” in Faunal Extinction in an Island Society: Pygmy Hippoptamus Hunters of Cyprus, A.H. Simmons, ed., New York, pp. 167–169. Strasser T.F., E. Panagopoulou, C.N. Runnels, P.M. Murray, N. Thompson, P. Karkanas, F.W. McCoy, and

191

K.W. Wegmann. 2010. “Stone Age Seafaring in the Mediterranean: Evidence from the Plakias Region for Lower Palaeolithic and Mesolithic Habitation of Crete,” Hesperia 79, pp. 145–190. Todd, I.A. 1998. Kalavasos-Tenta, Nicosia. van Andel, T.H., and C.N. Runnels. 1995. “The Earliest Farmers in Europe,” Antiquity 69, pp. 481–500. Vandermeersch, B., ed. 2003. Échanges et diffusion dans la préhistoire méditerranéenne. Actes des congrès nationaux des sociétés historiques et scientifiques, Section de pré- et protohistoire, 121e, Nice, 1996, Paris. Vigne, J.-D. 1993. “Domestication ou appropriation pour la chasse: Histoire d’un choix socio-culturel depuis le néolithique. L’example des cerfs (Cervus),” in Exploitation des animaux sauvages à travers le temps. XIIIe Rencontres internationales d’archéologie et d’histoire d’Antibes, 15–16–17 octobre 1992, J. Desse and F. Audoin-Rouzeau, eds., Juan Les Pins, pp. 201–220. Vigne, J.-D., I. Carrère, and J. Guilaine. 2003. “Unstable Status of Early Domestic Ungulates in the Near East: The Example of Shillourokambos (Cyprus, IX–VIIIth Millennia cal. b.c.),” in Le néolithique de Chypre (BCH Suppl. 43), J. Guilaine and A. Le Brun eds., Athens, pp. 239–251. Vigne, J-D., I. Carrère, J.-F. Saliége, A. Person, H. Bocherens, J. Guilaine, and F. Briois. 2000. “Predomestic Cattle, Sheep, Goat and Pig during the Late 9th and 8th Millennium cal bc on Cyprus: Preliminary Results of Shillourokambos (Parekklisha, Limassol),” in Mashkour et al., eds., 2000, pp. 83–106. Vigne, J.-D., and N. Desse-Berset. 1995. “The Exploitation of Animal Resources in the Mediterranean Islands during the Pre-Neolithic: The Example of Corsica,” in Man and Sea in the Mesolithic: Coastal Settlement above and below Present Sea Level, (Oxbow Monograph 53), A. Fischer, ed., Oxford, pp. 309–318. von den Driesch, A. 1987. “Haus und Jagdtiere im vorgeschichtlichen Thessalien,” PZ 62, pp. 1–21. von den Driesch, A., and U. Wodtke. 1997. “The Fauna of ‘Ain Ghazal, a Major PPN and Early PN Settlement in Central Jordan,” in The Prehistory of Jordan II: Perspectives from 1997 (Studies in Early Near Eastern Production, Subsistence and Environment 4), H.G.K. Gebel, Z. Kafafi, and G.O. Rollefson, eds., Berlin, pp. 511–556. Warren, P., M.R. Jarman, H.N. Jarman, N.J. Shackleton, and J.D. Evans. 1968. “Knossos Neolithic, Part II,” BSA 63, pp. 239–276

192

LIORA KOLSKA HORWITZ

Weesie, P.D.M. 1988. “Quaternary Avifauna of Crete, Greece,” Palaeovertebrata 18, pp. 1–94.

Wilkens, B. 1996. “Faunal Remains from Italian Excavations on Crete,” in Reese, ed., 1996, pp. 241–261.

Weninger, B., E. Alram-Stern, E. Bauer, L. Clare, U. Danzeglocke, O. Joris, C. Kubatzki, G. Rollefson, H. Todorova, and T. van Andel. 2006. “Climate Forcing Due to the 8200 cal yr bp Event Observed at Early Neolithic Sites in the Eastern Mediterranean,” Quaternary Research 66, pp. 401–420.

Yale Peabody Museum. 2008. “The Biodiversity of Crete at the Yale Peabody Museum,” http://www. peabody.yale.edu/explore/cretebiodiv.html

Westely, B. 1970. “The Mammalian Fauna,” in Excavations at Hacilar I, J. Mellaart, ed., Edinburg, pp. 245–247.

Zohary, D., E. Tchernov, and L.K. Horwitz. 1998. “The Role of Unconscious Selection in the Domestication of Sheep and Goats,” Journal of Zoology 245, pp. 129–135.

10

Radiocarbon Dates from the Neolithic Settlement of Knossos: An Overview Yorgos Facorellis and Yiannis Maniatis

Recent archaeological excavations of the Neo­ lithic settlement levels at Knossos, Crete (35°31' N, 25°10' E) yielded carbonized material suitable for radiocarbon dating.* Ten of the samples collected were dated in the Laboratory of Archaeometry, N.C.S.R. “Demokritos,” using the CO2 conventional radiocarbon technique, while seven more were dated by the accelerator mass spectrometry (AMS) method at the Radiocarbon Accelerator Unit of the Research Laboratory for Archaeology and the History of Art, Oxford University. The radiocarbon dates obtained previously in the Research Laboratory of the British Museum from samples

collected by J.D. Evans during his 1957–1960 and 1969–1970 archaeological campaigns are also considered here. In this study we discuss the absolute dating of the Neolithic phases of the settlement based on the samples from the 1997 excavation, and we compare these dates with those collected by Evans after converting the latter to calendar dates using the latest issue INTCAL04 of the international calibration curve. Our results, along with the archaeological data, help to establish an absolute chronology for the beginning of the Neolithic in Crete and the Aegean islands, which is now more accurately placed in the range of 7030–6780 b.c.

*The authors would like to thank Nikos Efstratiou of the Uni­ versity of Thessaloniki for helpful comments and discussion. Many thanks are also due to the technical staff of the Laboratory of Archaeometry, N.C.S.R. “Demokritos,” Kiki Gogidou, Melina Korozi, and Marigo E. Kyriazi, for their assistance in the preparation of the samples for radiocarbon dating. Abbreviations used in this chapter are: AMS accelerator mass spectrometry BM British Museum, London, lab code cal. calibrated or calendar years

DEM EN FN LN m MN OxA yr(s)

Laboratory of Archaeometryof N.C.S.R. “Demokritos,” Athens, lab code Early Neolithic Final Neolithic Late Neolithic meters Middle Neolithic Radiocarbon Accelerator Unit, Oxford University, lab code year(s)

YORGOS FACORELLIS AND YIANNIS MANIATIS

194

Archaeological Samples and Analytical Methods Samples collected by J.D. Evans were submitted for dating to the Research Laboratory of the British Museum (registered laboratory code BM). The measurements of these samples were carried out using the conventional liquid scintillation and gas counting techniques (CO2 and C2H2) as described elsewhere (Barker and Mackey 1959, 1968). The radiocarbon ages of the 19 samples in this group were published in the journal Radiocarbon in four different British Museum Date Lists (Barker and Mackey 1963; Barker, Burleigh, and Meeks 1969; Evans 1971; Burleigh, Hewson, and Meeks 1977; Burleigh and Matthews 1982). The dates came from charcoal samples found in Neolithic levels under the West and Central Courts of the Minoan palace (Table 10.1; for a brief archaeological description of the samples, see Table 10.2).

Sample Number

Context

Collection Year

In 1998, 17 samples collected from the 1997 excavation were submitted for dating to the Laboratory of Archaeometry of N.C.S.R. “Demokritos.” The samples originated from 14 different levels with depths from 2.0 to 7.8 m (Table 10.3). Ten of these samples were measured using the CO2 conventional radiocarbon technique (see description in Facorellis, Maniatis, and Kromer 1997). The other seven samples, which were undersized, were sent to the Oxford Radiocarbon Accelerator Unit of the Research Laboratory for Archaeology and the History of Art, Oxford University. These samples have been identified by Maria Ntinou as wood charcoal of the Quercus evergreen type, except for the sample from level 39, which is Quercus deciduous type. Their ages have been calculated by the methods described by Hedges and colleagues (1992).

d13C (‰)

Age (yrs b.p.)

Calendar Age (yrs b.c.)

Calibrated Age 14 C (yrs b.p.)

Probability

West Court BM-716*

Sounding FF, Level 38, Sample 3

1970

-25.0

5003 ± 213

4040–3530 4330–3360

5480–5990 5310–6280

(68.3%) (95.4%)

BM-579**

Stratum II

1969

not reported

5534 ± 76

4460–4330 4530–4240

6280–6410 6190–6480

(68.3%) (95.4%)

BM-575**

Stratum III/II

1969

not reported

5636 ± 94

4550–4360 4700–4330

6310–6500 6280–6650

(68.3%) (95.4%)

BM-580**

Stratum III

1969

not reported

5522 ± 88

4460–4270 4550–4070

6220–6410 6020–6500

(68.3%) (95.4%)

BM-717*

Sounding EE, Level 18, Sample 19

1970

-25.8

5806 ± 124

4790–4500 4950–4370

6450–6740 6320–6890

(68.3%) (95.4%)

BM-718*

Sounding EE, Level 27, Sample 23, Level 34, Samples 27–29

1970

-24.5

5892 ± 91

4900–4620 4990–4540

6570–6850 6490–6940

(68.3%) (95.4%)

BM-719*

Sounding AA/BB, Level 164 Sample AR (IA), Level 174, Sample AY (IA), Level 181 Sample BA (IA), Level 183, Sample BI (IA)

1970

-24.4

5967 ± 41

4907–4793 4948–4728

6742–6856 6677–6897

(68.3%) (95.4%)

BM-1371*

Sounding AA/BB, Level 272, Sample CW (II) Level 277, Sample CY (II)

1970

-24.7

6201 ± 252

5460–4850 5620–4550

6790–7410 6500–7570

(68.3%) (95.4%)

BM-1372*

Sounding AA/BB, Level 279, Samples CM, DF, DG (II), Level 286, Sample CL (II)

1970

-24.3

6482 ± 161

5610–5310 5710–5060

7260–7560 7010–7660

(68.3%) (95.4%)

Table 10.1. Summary of the British Museum radiocarbon dates on charcoal from the excavations of J.D. Evans at Neolithic Knossos, sorted by age. * Burleigh and Matthews 1982; ** Burleigh, Hewson, and Meeks 1977.

195

RADIOCARBON DATES FROM THE NEOLITHIC SETTLEMENT OF KNOSSOS: AN OVERVIEW

Sample Number

Collection Year

Context

d13C (‰)

Age (yrs b.p.)

Calendar Age (yrs b.c.)

Calibrated Age 14 C (yrs b.p.)

Probability

Central Court BM-581**

Stratum II

1969

not reported

5588 ± 145

4610–4270 4770–4050

6220–6560 6000–6720

(68.3%) (95.4%)

BM-279***

Sample 6, Stratum IV

1960

not reported

5680 ± 150

4690–4370 4900–4240

6320–6640 6180–6850

(68.3%) (95.4%)

BM-577**

Stratum IV

1969

not reported

5884 ± 188

4990–4540 5210–4360

6490–6940 6310–7160

(68.3%) (95.4%)

BM-274***

Sample 4, Area A, Level 15, Stratum V

1960

not reported

6140 ± 150

5290–4860 5460–4720

6800–7240 6670–7410

(68.3%) (95.4%)

BM-126****

Sample 5, Pit F, Area A, Level 16A, Stratum V

1960

not reported

7000 ± 180

6050–5720 6230–5560

7670–7990 7510–8180

(68.3%) (95.4%)

BM-273***

Sample 3, Area AC, Level 17, Stratum VI

1960

not reported

6210 ± 150

5320–4960 5470–4800

6910–7270 6750–7420

(68.3%) (95.4%)

BM-272***

Sample 2, Area AC, Level 24, Stratum IX

1960

not reported

7570 ± 150

6590–6250 6750–6070

8200–8540 8020–8700

(68.3%) (95.4%)

BM-436***

Sample 1, Pit F, Area AC, Level 27, Stratum X

1960

not reported

7740 ± 140

6770–6430 7050–6370

8380–8720 8320–8990

(68.3%) (95.4%)

BM-278***

Sample 1, Pit F, Area AC, Level 27, Stratum X

1960

not reported

7910 ± 140

7030–6650 7170–6470

8600–8980 8420–9120

(68.3%) (95.4%)

BM-124****

Sample 1, Pit F, Area AC, Level 27, Stratum X

1960

not reported

8050 ± 180

7250–6690 7480–6540

8640–9200 8480–9430

(68.3%) (95.4%)

Table 10.1, cont. Summary of the British Museum radiocarbon dates on charcoal from the excavations of J.D. Evans at Neolithic Knossos, sorted by age. ** Burleigh, Hewson, and Meeks 1977; *** Barker, Burleigh, and Meeks 1969; **** Barker and Mackey 1963.

Sample Number

Description

Publication

BM-716

charcoal belonging to the FN

Burleigh and Matthews 1982

BM-579

charcoal found in Stratum II belonging to the LN

Evans 1971; Burleigh and Matthews 1982

BM-575

charcoal found in Stratum III/II belonging to the LN/MN transition

Burleigh and Matthews 1982

BM-580

charcoal found in Stratum III belonging to the LN–MN

Evans 1971; Burleigh and Matthews 1982

BM-717

charcoal belonging to the LN

Burleigh and Matthews 1982

West Court

BM-718

charcoal belonging to the MN

Burleigh and Matthews 1982

BM-719

charcoal belonging to the EN II

Burleigh and Matthews 1982

BM-1371

charcoal belonging to the EN I

Burleigh and Matthews 1982

BM-1372

charcoal belonging to the EN I

Burleigh and Matthews 1982

BM-581

charcoal found in Stratum II belonging to the LN

Evans 1971; Burleigh and Matthews 1982

BM-279

charcoal from upper Neolithic level in Stratum IV

Barker, Burleigh, and Meeks 1969; Evans 1971

BM-577

charcoal found in Stratum IV belonging to the late EN II

Evans 1971; Burleigh and Matthews 1982

BM-274

charcoal from habitation level in Stratum V

Barker, Burleigh, and Meeks 1969; Evans 1971

BM-126

charcoal from the same excavation region (above the bedrock) as BM-124 near the top of the EN levels (Stratum V)

Barker and Mackey 1963; Evans 1971

BM-273

charcoal from habitation deposit in Stratum VI

Barker, Burleigh, and Meeks 1969; Evans 1971

Central Court

Table 10.2. Description of the samples dated in the British Museum Research Laboratory.

196 Sample Number

Description

Publication

Central Court, cont. BM-272

charcoal from occupation layer associated with first mudbrick houses and immediately overlying earliest camp occupation above bedrock (Stratum IX)

Barker, Burleigh, and Meeks 1969; Evans 1971

BM-436

carbonized grain associated with the carbonized stake of the sample BM-278

Barker, Burleigh, and Meeks 1969

BM-278

charcoal, remains of carbonized wooden stake found in Stratum X associated with sample BM-124

Barker and Mackey 1963; Barker, Burleigh, and Meeks 1969

BM-124

charcoal, remains of carbonized wooden stake originating from the lowest level above the bedrock (Stratum X)

Barker and Mackey 1963; Evans 1971

Table 10.2., cont. Description of the samples dated in the British Museum Research Laboratory.

Sample Number

Context

Type

d13C (‰)

Age (yrs b.p.)

Calendar Age (yrs b.c.)

Calibrated Age 14 C (yrs b.p.)

Probability

DEM-638

Knossos-6, Trench II, level 9, depth 2.00 m

charcoal

-25.00

6223 ± 120

5320–5030 5470–4850

6980–7260 6800–7420

(68.3%) (95.4%)

DEM-640

Knossos-10, Trench II, level 12, depth 2.20 m

charcoal

-25.00

5980 ± 43

4932–4801 4990–4731

6750–6881 6680–6939

(68.3%) (95.4%)

DEM-641

Knossos-12, Trench II, level 13, depth 2.40 m

charcoal

-25.00

6134 ± 116

5220–4910 5330–4780

6860–7170 6730–7280

(68.3%) (95.4%)

DEM-642

Knossos-13, Trench II, level 14, depth 2.50 m

charcoal

-25.00

5977 ± 36

4929–4800 4982–4774

6749–6878 6723–6931

(68.3%) (95.4%)

DEM-658

Knossos-20, Trench II, level 24, depth 3.95 m

charcoal

-25.00

6106 ± 40

5199–4956 5208–4936

6905–7148 6885–7157

(68.3%) (95.4%)

DEM-659

Knossos-22, Trench II, level 28, depth 4.15 m

charcoal

-25.00

5991 ± 55

4940–4800 5010–4730

6750–6890 6680–6950

(68.3%) (95.4%)

OxA-9221

Knossos-7, Trench II, level 28/2, depth 4.15 m

wood charcoal (Quercus evergreen type)

-25.0

6042 ± 34

4994–4856 5030–4843

6805–6943 6792–6979

(68.3%) (95.4%)

Mean value of level 28

—

—

6028 ± 29

4989–4847 5042–4779

6797–6939 6729–6992

(68.3%) (95.4%)

DEM-660

Knossos-23, Trench II, level 29, depth 4.25 m

charcoal

-25.00

7339 ± 57

6250–6100 6360–6070

8040–8190 8020–8310

(68.3%) (95.4%)

OxA-9218

Knossos-4, Trench II, level 29/4, depth 4.25 m

wood charcoal (Quercus evergreen type)

-23.3

5990 ± 50

4940–4800 5000–4730

6750–6890 6680–6950

(68.3%) (95.4%)

DEM-670

Knossos-24, Trench II, level 30, depth 4.50 m

charcoal

-25.00

5801 ± 151

4830–4470 5010–4350

6410–6780 6290–6960

(68.3%) (95.4%)

DEM-661

Knossos-25, Trench II, level 31, depth 4.90 m

charcoal

-25.00

6213 ± 65

5290–5060 5310–5000

7010–7240 6950–7260

(68.3%) (95.4%)

DEM-663

Knossos-26, Trench II, level 32, depth 5.80 m

charcoal

-25.00

6154 ± 30

5207–5050 5211–5016

6999–7156 6965–7160

(68.3%) (95.4%)

OxA-9220

Knossos-6, Trench II, level 32/2, depth 5.80 m

wood charcoal (Quercus evergreen type)

-22.5

6160 ± 50

5210–5050 5290–4960

7000–7160 6910–7230

(68.3%) (95.4%)

Mean value of level 32

—

—

6156 ± 26

5207–5050 5211–5016

7000–7156 6938–7169

(68.3%) (95.4%)

Table 10.3. Summary of radiocarbon dating results of carbonized samples collected in 1997 from the Neolithic settlement levels at Knossos.

197

Sample Number

Context

Type

d13C (‰)

Age (yrs b.p.)

Calendar Age (yrs b.c.)

Calibrated Age 14 C (yrs b.p.)

Probability

OxA-9217

Knossos-3, Trench II, level 33/7, depth 6.50 m

wood charcoal (Quercus evergreen type)

-24.0

6145 ± 50

5210–5030 5220–4950

6980–7160 6900–7170

(68.3%) (95.4%)

OxA-9219

Knossos-5, Trench II, level 35/2, depth 7.10 m

wood charcoal (Quercus evergreen type)

-22.7

6361 ± 37

5460–5305 5468–5228

7254–7409 7177–7417

(68.3%) (95.4%)

OxA-9216

Knossos-2, Trench II, level 37/16, depth 7.40 m

wood charcoal (Quercus evergreen type)

-23.2

6185 ± 50

5210–5060 5300–5000

7010–7160 6950–7240

(68.3%) (95.4%)

OxA-9215

Knossos-1, Trench II, level 39/1, depth 7.80 m

wood charcoal (Quercus deciduous type)

-24.3

7965 ± 60

7030–6780 7050–6690

8730–8980 8640–9000

(68.3%) (95.4%)

Table 10.3, cont. Summary of radiocarbon dating results of carbonized samples collected in 1997 from the Neolithic settlement levels at Knossos.

Results Table 10.1 presents all the necessary information on the dated samples of the British Museum Knossos series (laboratory code number, location, collection year, and δ13C value if reported). The dates are sorted by level. The conventional radiocarbon ages, the corresponding calibrated calendar dates in years b.c., and the calibrated radiocarbon ages in years b.p. within one and two standard deviations (probability 68.3% and 95.4%, respectively) are also given. The calibration of the conventional radiocarbon ages was performed with the latest version of the international calibration curve INTCAL04 using the calibration program Calib rev.5.0 (Pearson, Becker, and Qua 1993; Stuiver and Pearson 1993; Stuiver and Reimer 1993; Reimer et al. 2004). Until recently, the Neolithic chronology in Crete was based on conventional radiocarbon ages, which were converted to calendar years simply by deducting 1,950 years (Evans 1971). This practice was prevalent prior to the development of an international calibration curve, the use of which has resulted in significant shifts in the dates. Figure 10.1 shows the distribution of the calibrated dates of the BM samples, which are sorted by stratum. The samples from the West and Central Courts of the Minoan palace are plotted separately. The length of the bars represents the age range; the height represents the percent probability that

the sample lies in the specific range (Maniatis and Kromer 1990). In this plot one can see that the age of the samples increases systematically with the stratum except for BM-126, which has an earlier age. This might be explained by an anthropogenic or other natural disturbance of the habitation deposits in the past. Table 10.3 presents all the necessary information on the dated samples collected in 1997 (laboratory code number, level, depth, type, and δ13C value). The mean depth of each level has been estimated from the diagram of the section of the southwestern corner of the trench (Efstratiou et al. 2004). The internationally registered codes DEM and OxA are used to distinguish the samples measured in the Laboratory of Archaeometry, N.C.S.R. “Demokritos” and the Oxford Radiocarbon Accelerator Unit, respectively. The radiocarbon dates are shown sorted by level, and they are converted to calendar dates and calibrated radiocarbon ages using INTCAL04. In the case of the samples measured in the Laboratory of Archaeometry, the usual value for the age correction of charcoal samples due to the isotopic fractionation (δ13C = -25.00‰) was used (Polach 1976; Stuiver and Polach 1977). Three pairs of heterogeneous samples from levels 28, 29, and 32 (level 28: DEM-659 and OxA9221; level 29: DEM-660 and OxA-9218; level 32: DEM-663 and OxA-9220) were dated in both

198 BM-716

West Court

BM-579 DEM-638

BM-575

DEM-640

BM-580 BM-717

DEM-641

BM-718

DEM-642

BM-719

DEM-658

BM-1371

OxA-9221

BM-1372

DEM-659 DEM-660

BM-581

OxA-9218

BM-279

Central Court

DEM-670

BM-577

DEM-661

BM-274

OxA-9220

BM-126

DEM-663

BM-273

7000

6500

6000

5500

5000

4500

4000

BM-272

OxA-9217

BM-436

OxA-9219

BM-278

OxA-9216

BM-124

OxA-9215

3500

7000

6500

6000

5500

5000

4500

4000

3500

cal. B.C.

cal. B.C.

Figure 10.1. Distribution of calibrated dates sorted by stratum of the samples from the excavations of J.D. Evans, dated at the radiocarbon unit of the Research Laboratory of the British Museum (BM). The length of the bars represents the age range; the height represents the percent probability of the sample within the specific range.

1

Figure 10.2. Distribution of calibrated dates sorted by depth of the samples from the 1997 archaeological campaign, dated at the radiocarbon unit of the Laboratory of Archaeometry, N.C.S.R. "Demokritos" (DEM), and at the Radiocarbon Accelerator Unit of the Research Laboratory for Archaeology and the History of Art, Oxford University (OxA). The length of the bars represents the age range; the height represents the percent probability of the sample within the specific range.

Accumulation rate 110 cm/100 yr

2

DEM-640 DEM-642

Depth from surface (m)

DEM-641

3 DEM-650

4

DEM-659 OxA-9218 DEM-661

5

DEM-663

6

Accumulation rate

OxA-9217

4.5 cm/100 yr

7 OxA-9215

OxA-9219

8 7000

6500

6000

5500

5000

4500

Calendar age (years B.C.)

Figure 10.3. Calibrated radiocarbon dates from the 1997 excavation at Knossos plotted against the depth of the samples in order to determine the accumulation rate of the habitation deposits.

199

laboratories. In the case of the first and the third pairs, both samples gave the same age within one standard deviation (1σ, probability 68.3%). The mean value of each pair (Table 10.3) has been calculated taking into account the calibration curve standard deviation (Ward and Wilson 1978; Stuiver and Reimer 1993). The conventional radiocarbon ages of the two samples from level 29 differ by ca. 1,350 years, perhaps due to some form of disturbance. Figure 10.2 shows the distribution, sorted by depth, of the calibrated dates of the samples from the 1997 archaeological campaign. In this plot one can see that the ages increase with the depth, but a few inversions occur (e.g., the samples DEM-660 and DEM-670 give higher and lower ages, respectively, than expected). This may also be attributed to the disturbance of the habitation deposits by the action of anthropogenic and/or other natural factors. Nevertheless, the high precision dating of these samples compared to those of Evans’s group is obvious, thus reflecting the improvement of the radiocarbon dating technique through the time. The time gaps seen in Figure 10.2 can be explained by

the fact that among the 39 levels of the trench, only 14 yielded material suitable for dating. An attempt was made to calculate the accumulation rate of the habitation deposits using the time interval versus the mean depth at consecutive levels dated by radiocarbon. The samples presenting age inversions were disregarded. Figure 10.3 shows the relationship between the calibrated radiocarbon ages of samples from 11 different levels, dated between ca. 7000 and 4800 b.c., and their depths from the surface. The slope of the line of the best linear fit changes after ca. 5400 b.c. The accumulation rate was ca. 4.5 cm/100 years from 7000 to 5400 b.c., increasing to ca. 110 cm/100 years from 5400 to 4800 b.c. This could be explained by more intensive human activity during the late Early Neolithic (EN) and Middle Neolithic (MN) periods in comparison with the Aceramic and the beginning of the EN period. These results are tentative, as many factors cannot be controlled, including sediment erosion and human intervention. The curves, however, show a great difference in the accumulation rate between the two different periods of time.

Conclusions The systematic radiocarbon dating of samples from the Neolithic settlement levels at Knossos forms the basis for the interpretation of the archaeological finds and for the reconstruction of the paleoenvironment. Use of the international calibration curve INTCAL04 allowed the conversion to calendar years of all the conventional radiocarbon ages from three different radiocarbon laboratories, making it possible to compare them directly. Radiocarbon dating has established

absolute dates for the beginning of the Neolithic in Crete at 7030–6780 b.c. It has also permitted the detection of possible stratigraphic disturbances. Furthermore, the comparison of accumulation rates of the habitation deposits from ca. 7000 to 5400 b.c. and 5400 to 4800 b.c. indicates more intensive human activity during the late EN and MN phases relative to the Aceramic and the beginning of the EN period.

References Barker, H., R. Burleigh, and N. Meeks. 1969. “British Museum Natural Radiocarbon Measurements VI,” Radiocarbon 11, pp. 278–294.

Barker, H., and J. Mackey. 1959. “British Museum Natural Radiocarbon Measurements I,” American Journal of Science Radiocarbon Supplement 1, pp. 81–86.

200

———. 1963. “British Museum Natural Radiocarbon Measurements IV,” Radiocarbon 5, pp. 104–108. ———. 1968. “British Museum Natural Radiocarbon Measurements V,” Radiocarbon 10, pp. 1–7. Burleigh, R., A. Hewson, and N. Meeks. 1977. “British Museum Natural Radiocarbon Measurements IX,” Radiocarbon 19, pp. 143–160. Burleigh, R., and K. Matthews. 1982. “British Museum Natural Radiocarbon Measurements XIII,” Radiocarbon 24, pp. 151–170. Efstratiou, N., A. Karetsou, E. Banou, and D. Margomenou. 2004. “The Neolithic Settlement of Knossos: New Light on an Old Picture,” in Knossos: Palace, City, State. Proceedings of the Conference in Herakleion Organised by the British School at Athens and the 23rd Ephoreia of Prehistoric and Classical Antiquities of Heraklion, in November 2000, for the Centenary of Sir Arthur Evans’s Excavations at Knossos (BSA Studies 12), G. Cadogan, E. Hatzaki, and A. Vasilakis, eds., London, pp. 39–51. Evans, J.D. 1971. “Neolithic Knossos: The Growth of a Settlement,” PPS 37, pp. 95–117. Facorellis, Y., Y. Maniatis, and B. Kromer. 1997. “Study of the Parameters Affecting the Correlation of Background versus Cosmic Radiation in CO2 Counters: Reliability of Dating Results,” Radiocarbon 39, pp. 225–238. Hedges, R.E.M., M.J. Humm, J. Foreman, J. van Klinken, and C.R. Bronk. 1992. “Developments in Sample Combustion to Carbon Dioxide, and in the Oxford AMS Carbon Dioxide Ion Source System,” Radiocarbon 34, pp. 306–311. Maniatis, Y., and B. Kromer. 1990. “Radiocarbon Dating of the Late Neolithic–EBA Site of Mandalo, Macedonia, Greece,” Radiocarbon 32, pp. 149–153.

Pearson, G.W., B. Becker, and F. Qua. 1993. “HighPrecision 14C Measurements of German and Irish Oaks to Show the Natural 14C Variations from 7890 to 5000 bc,” Radiocarbon 35, pp. 93–104. Polach, H.A. 1976. “Radiocarbon Dating as a Research Tool in Archaeology: Hopes and Limitations,” in Proceedings of the Symposium on Scientific Methods in the Study of Ancient Chinese Bronzes and South East Asian Metal and Other Archaeological Artefacts, October 6–10, 1975, National Gallery of Victoria, Melbourne, N. Barnard, ed., Melbourne, pp. 255–298. Reimer, P.J., M.G.L. Baillie, E. Bard, A. Bayliss, J.W. Beck, C. Bertrand, P.G. Blackwell, C.E. Buck, G. Burr, K.B. Cutler, P.E. Damon, R.L. Edwards, R.G. Fairbanks, M. Friedrich, T.P. Guilderson, K.A. Hughen, B. Kromer, F.G. McCormac, S. Manning, C. Bronk Ramsey, R.W. Reimer, S. Remmele, J.R. Southon, M. Stuiver, S. Talamo, F.W. Taylor, J. van der Plicht, and C.E. Weyhenmeyer. 2004. “INTCAL04 Terrestrial Radiocarbon Age Calibration, 0–26 Cal Kyr bp,” Radiocarbon 46, pp. 1029–1058. Stuiver, M., and G.W. Pearson. 1993. “High-Precision Bidecadal Calibration of the Radiocarbon Time Scale, ad 1950–500 bc and 2500–6000 bc,” Radiocarbon 35, pp. 1–23. Stuiver, M., and H.A. Polach. 1977. “Discussion: Reporting of 14C Data,” Radiocarbon 19, pp. 355–363. Stuiver, M., and P.J. Reimer. 1993. “Extended 14C Data Base and Revised Calib 3.0 14C Age Calibration Program,” Radiocarbon 35, pp. 215–230. Ward, G.K., and S.R. Wilson. 1978. “Procedures for Comparing and Combining Radiocarbon Age Determinations: A Critique,” Archaeometry 20, pp. 19–31.

11

Knossos and the Beginning of the Neolithic in Greece and the Aegean Islands Nikos Efstratiou

Perceptions of Neolithic Knossos in Past Research Until very recently the Aegean archipelago seemed to be a desolate place for Neolithic communities and their archaeology (Cherry 1981, 1990; Davis 1992).* A number of articles based on limited fieldwork and armchair valuations of the area gave the impression that the islands of the Aegean experienced little or no cultural stimulus until well into the fourth millennium b.c. The very few exceptions of early island communities such as Hagios Petros in the Northern Sporades (Efstratiou 1985) and Saliagos in the Cyclades (Evans and Renfrew 1968), dated to the preceding sixth and fifth millennia, respectively, have material records that are impressive and surely suggestive of an intrinsic, albeit hidden, island dynamism, yet they were treated as isolated occurrences. These perceptions have changed in the last decade or so, as we shall see below. Looking back, two recurring and monotonous explanatory themes seem to have dominated archaeological attempts to reconstruct Aegean island

societies that existed before the Bronze Age. The first theme is the difficulty of finding and archaeologically documenting “early” sites due to visibility issues relating to rises in sea level and geomorphology (Lambeck 1996). The second theme is the attempted application of rigid archaeological explanatory models of colonization based either on a number of preconceived diachronic motivations or on ethnographic examples. The latter, deriving mainly from the Pacific, were considered by archaeologists to be suitable case studies for explaining the admittedly very different historical “locales” of the maritime eastern Mediterranean world from the early Holocene onward (see the three scenarios suggested by Broodbank 2000, 126). By “locale” I mean a historical scene and circumstances where *Abbreviations used in this chapter are: cal.=calibrated or calendar years, Ch(s).=Chapter(s), EN=Early Neolithic, PPN= Pre-Pottery Neolithic (with phases PPNA, PPNB, PPNC).

202

NIKOS EFSTRATIOU

the performance of individual and collective actions, expressed by learned behavior, practices, or symbols, recurs in multiple and distinctive episodes. These locales or historical scenes owe their decisive and modulatory power to the deep socioeconomic and political structures they create in time. Their archaeological interpretation, in turn, relies on contextual explanations of high resolution taking the form of different kinds of palimpsests (Bailey 2007, 199). It is important to note that maritime locales, in the above sense, are not bounded exclusively by insular contexts or regulated by the cumulative meaning of notions like island social identity or movement. Rather, they “are affected by broader, often dense and entangled interaction spheres” (Knapp 2007, 46). I consider these “macro-scale” island-mainland relations to be essential for understanding the early prehistory of the eastern Mediterranean maritime world and the complex individuality of specific “microscale” case studies. Returning to the previous axiomatic positions concerning the location of early sites, it is worth noting that their validity was never tested on the ground. Instead, they remained models based on the statistical processing of randomly available generic data, mostly biogeographical factors of size, location, and distance (Cherry 1981, 1990). Moreover, the historical reconstructions put forward could not navigate creatively through different scales of phenomena, processes, and, ultimately, narratives so as to arrive at a pluralism in the questions asked with regard to the archaeological record. One may claim that this was the result of the limited archaeological data available—the solitariness of Knossos is deafening—but it may equally have been the unavoidable outcome of attempting to accommodate the rich debate about islands worldwide into a weak Aegean archaeological example (Broodbank 2006, 199). I must admit that the historical and anthropological debate focusing on Mediterranean practices and motives is provocatively rich, and it is perhaps too hard for archaeologists to keep a critical position vis-à-vis that literature (Herzfeld 1991; Horden and Purcell 2000). Any environmental, geographic, or cultural concept borrowed by archaeologists, however,

must be archaeologically particularized and historically “time-bound,” a requirement that is especially demanding when dealing with prehistory. It is an open secret that most of the known Aegean island Neolithic sites were either found by chance or reported in the context of extensive surveys, never as part of single-period surveys. The latter might have allowed the acquisition of comparative material from archaeological palimpsests within an island or between islands of the same cultural and chronological horizon, therefore increasing the resolution of the specific period (Bailey 2007, 198). This potential has been demonstrated clearly by the discovery of new pre-Neolithic sites in Cyprus datable to the Akrotiri-Aetokremnos period (Ammerman et al. 2006, 2008). The specific site location model employed was, as in similar cases elsewhere, responsible for their successful discovery (Runnels et al. 2005, 259). Concepts like insularity, the founder effect, and island identity, the referential components of the early island historical context upon which archaeologists have long relied (MacArthur and Wilson 1967; Knapp 2007), excessively in my opinion, were linear in their perception and use. Although occasionally modified by specific sociogeographical parameters such as the emphasis on social factors and relations (Patton 1996), the period’s historical context continued to be defined in a descriptive and atemporal manner. This view has lately been counterbalanced by the postmodern concept of “islandscapes” and constructs such as agency, experience, memory, and visualization, through which social life is regarded more as a cultural and symbolic category and less as a specific historical unity (Gosden and Pavlides 1994, 162). The question of how to conceptualize in historical terms the specific causal forces that shape what is often uncritically termed an “island way of life” still remain the object of archaeological research. It has proven unproductive to approach the “island way of life” of the eastern Mediterranean from the early Holocene period onward in terms of atemporal processes involving idiosyncratic cases and explanations based on motivations detection and logistical analysis. Instead we need to consider the different maritime historical locales where

KNOSSOS AND THE BEGINNING OF THE NEOLITHIC IN GREECE AND THE AEGEAN ISLANDS

deep socioeconomic and political structures were in force, exercising an unfailingly decisive and modulatory—but not deterministic—impact on prehistoric behavior, and therefore on cultural developments (Braudel 1972). These structures may not be limited to identity-driven, particularized, island cultural choices but are shaped rather by social life. The archaeology of this “historicity” could only be the offspring of both “micro” and “macro” levels of archaeological analysis (Efstratiou 2007, 129). I consider that the case of the tell mound of Knossos, with its high-resolution palimpsest, requires a combination of both of these levels of analysis in order to interpret the typical but at the same time complex components of the early farming “maritime world” of the Aegean that are reflected in the materiality of the Knossos example. The way in which the discovery of Neolithic Knossos was incorporated into the archaeological historiography of the time is interesting. The integration of the impressive deposits of Knossos, revealed during the 1960s, into a rather unimaginative Aegean island Neolithic scene occurred silently. The excellent archaeological work of J.D. Evans at the site and his even more impressive finds failed to surface noticeably on the Kephala mound or to disturb the public admiration of the Bronze Age Palace and its Minoan inhabitants. In a metaphorical sense, this was the first, the last, and the most vociferous attempt of the Neolithic farmer pastoralists to acquire some of the historicity and archaeological credibility that the palace’s residents enjoyed. The very few archaeological studies dealing with the life and history of the early farming community of Knossos range from being contained and descriptive (Whitelaw 1992; Evans 1994) to being speculative. While the former was predictable, the latter was characterized by the tendency to overstretch the meaning and the implications in social life of specific categories of archaeological data such as pottery style (Broodbank 1992; Tomkins 2007, 2008). It is still a mystery as to why, despite the early date of its basal strata (7000 b.c.) and their undeniable richness—and what this might have implied for Aegean island archaeology—Neolithic Knossos as a whole kept such a low profile in

203

Aegean historiography. Surprisingly enough, even the extreme paradox of Knossos being “the one and only” Neolithic settlement on the island was not enough to prompt fieldwork attempts to locate the remains of other early Neolithic communities in Crete (Broodbank 2006). Instead, Knossos was treated and is still seen today as a particularly adventurous event of targeted Neolithic island colonization (Broodbank and Strasser 1991; Broodbank 2006, 214), even though the comprehensive archaeological data, both older and more recent, point toward a classic case of a broad farming front spreading westward (Efstratiou et al. 2004; Isaakidou and Tomkins, eds., 2008). It seems as though the extraordinary size and oddness of the Knossos Neolithic tell was designated to match in magnitude and uniqueness the sovereignty of the superimposed palace community. The negative results of surveys in Crete aiming toward the identification of other Neolithic sites of the seventh and sixth millennia b.c.—mostly multi-period surveys carried out by Minoan specialists (Manning 1999)—were reinforced by the persistent, widespread lack of similar archaeological evidence from other Aegean islands (Cherry 1990). This void, of course, does not explain either the solitariness of Neolithic Knossos or its long occupation sequence—two contradictory features. Nor does it explain Knossos’s distinctive pottery typology, with its seemingly late character, which also passed unquestioned until recently (Tomkins 2008, 21). It is as though these Knossian material eccentricities were precisely what deprived the Neolithic society of the Kairatos valley of its virtual historicity. This observation is more than an interesting matter of archaeological historiography. In my opinion, it is clearly the outcome of the disassociation of the archaeological data and their intrinsic cultural metaphors from history and its causal forces of social life. The material individuality of Knossos is perhaps the reason for the continued emphasis on the “earliness” of the settlement, despite the strong evidence, mainly from pottery, for the opposite, as well as for the poor documentation of the 1,500 succeeding years of occupation, which have received very little investigation.

204

NIKOS EFSTRATIOU

Late Pleistocene and Early Holocene Seafarers The recently announced evidence for Mesolithic habitation in Crete will undoubtedly change the present picture in ways that are not easy to envision at this stage (Strasser et al. 2010). The significance of the new discovery will depend on the dynamism of the Mesolithic communities as reflected in the number of sites, their geographic distribution across the island, their lithic assemblages, and other material data, along with their chronology. The latter in particular will determine whether Final Mesolithic groups in Crete were actively involved in the introduction of cultigens to the Aegean in the eighth millennium b.c. or if they played a marginal role and kept a low profile in the agricultural transformations taking place at the beginning of the seventh millennium b.c. In the latter case the Mesolithic inhabitants of Crete should be considered irrelevant to the neolithization of the island. From this perspective it is unfortunate that the lithic assemblage of the Aceramic layer of Knossos has not produced an unequivocal Final Mesolithic element (Conolly 2008, 88). Nevertheless, a possible pre-Neolithic/ early Holocene date for the newly discovered Cretan Mesolithic assemblage (ninth millennium b.c.) opens up a new chapter in the early prehistory of the islands, one that may in turn have followed an Upper Paleolithic habitation in the Aegean. The recent recovery of the rich hunting-foraging Epigravettian coastal site of Ouriakos on the island of Lemnos, with a lithic assemblage unparalleled in the Aegean area and a possible date around 14,000 b.p., illustrates a hitherto well-hidden and dynamic pre-Neolithic scene (Paolo Biagi, pers. comm., 2011). While the Aegean island pre-Neolithic archaeological picture remains obscure, it is worth considering the cultural developments in the adjacent regions at the end of the Paleolithic. In Cyprus new data and, most importantly, new methods for retrieving that data have greatly changed the way in which we view early prehistoric island developments (Swiny, ed., 2001; Ammerman et al. 2006; Simmons and Mandel 2007, 475; Ammerman et al. 2008). Indeed, the new evidence from Cyprus upsets many wellestablished views on island habitation patterns, early navigation, and colonization mechanisms, and the emerging interpretations in turn have farreaching effects on the way we see early Holocene

developments in the eastern Mediterranean and the Near East (Bar-Yosef 2001; Peltenburg and Wasse, eds., 2004; Broodbank 2006). Fieldwork at the Aceramic Neolithic sites of Kissonerga-Mylouthkia and Shillourokambos, situated along the southern coast of Cyprus and dated to the second half of the ninth–eighth millennium cal. b.c., have demonstrated a close Cypriot-Levant late Pre-Pottery Neolithic A (PPNA) and early PPNB cultural connection (Peltenburg et al. 2001; Guilaine 2003). Moreover, the recent discovery of a number of new preNeolithic sites most probably of the “Akrotiri phase” in the Cypriot sequence (10th millennium b.c.) at Nissi Beach and Aspros along the southern coast indicates an even earlier cultural dynamism, taking the form of short visits by seagoing forager groups from the opposite coast of the Levant (McCartney 2005, 10; Ammerman et al. 2006, 1; 2007, 2008). Though the available absolute dates of these new pre-Neolithic sites (Ammerman et al. 2006, 18; 2008) need further confirmation, their lithic material, in terms of typology and technology, seems to pre-date the Kissonerga-Mylouthkia, Shillourokambos, and Kalavasos-Tenta occupation phases of ca. 8200 b.c. (McCartney 2004, 103). The more general issues raised by these new finds—visibility, early island colonization, the circulation of obsidian, and late Pleistocene navigation incentives, though not novel per se, have significant effects on the way we perceive Near Eastern societies at the turn of the Holocene when farming began (Mithen 2003). The mainland forager communities of that period seem to have enjoyed a new dynamism in their collective social and material lives, experiences, perceptions, and politics as they engaged in short-term seasonal visits from the Levantine mainland to Cyprus (Ammerman 2011). The association of these developments with the contemporary climatic phenomenon of the Younger Dryas (early 11th–mid 10th millennium cal. b.c.), with its apparent impact on early Holocene developments in the eastern Mediterranean and the start of farming, is nowadays inevitable (Rosen 2007). The discovery of pre-Neolithic sites in Cyprus was made possible by a change in methodological

KNOSSOS AND THE BEGINNING OF THE NEOLITHIC IN GREECE AND THE AEGEAN ISLANDS

approaches to the investigation of islands. Ammerman and his associates have persuasively argued that pre-Neolithic camp sites were previously visited rarely by archaeologists looking for early sites because of their low farming potentials (Ammerman et al. 2006), even though sites like Nissi Beach and Aspros, which rest on exposed formations of aeolianites where tiny lithics are easily spotted, offer high geomorphological visibility. Ammerman’s point could be a crucial one and, pending additional confirmation, it may be expected to change the way in which we set our surveying priorities in island and coastal geomorphological environments. Does this state of events create a precedent for the Aegean island area as well? Oddly enough, research was moving in a similar direction, although less overtly, long before the recent finds in Cyprus (Sampson 1998, 1). Archaeologists entertained the idea of an early habitation horizon associated with seagoing human groups in light of the discovery of a few pieces of Melian obsidian at Franchthi Cave (lithic phase VI) dated to the ninth millennium b.c. or earlier (Torrence 1986; Perlès 1987, table 29.1), although the occurrence of the obsidian was considered unusual or even extraordinary. Honea’s dramatic proposal of the existence of a Mesolithic site on Kythnos was treated even more skeptically despite the supporting radiocarbon evidence (Honea 1975). It was not until Sampson began to uncover the early deposits of the Cave of the Cyclops on Youra in the Northern Sporades that the idea of a preNeolithic presence in some of the Aegean islands became a strong possibility (Sampson, Kozłowski, and Kaczanowska 2003; Sampson 2008a). It is suggested here that the use of the term “Mesolithic,” which is extensively used for the recent finds, is rather constraining in view of the atypical character of the Greek Mesolithic (Galanidou and Perlès, eds., 2003). The reservations initially expressed on taphonomic grounds about the cave stratigraphic sequence of Youra, especially regarding the “Mesolithic” microlithic finds, are still valid (Kaczanowska and Kozłowski 2008, 172, 178). Nevertheless, the new preNeolithic finds from a second site, that of Maroulas on Kythnos (Sampson et al. 2002), and more recently from other sites on the island of Ikaria in the eastern Aegean (Sampson 2006; Sampson, Kaczanowska, and Kozłowski 2009, 321), seem to confirm the presence of a remarkable pre-Neolithic

205

island horizon of uncertain origin and unknown relationship to the next Neolithic period. In terms of both relative and absolute chronology, these sites open up a whole new chapter in Greek prehistory and eastern Mediterranean island archaeology (Efstratiou 2005; Ammerman 2010). They attest to a very early seafaring tradition in the Aegean and to the probable presence of late Pleistocene (Epipaleolithic) maritime-oriented groups in the same area, as confirmed by the discovery of an Epigravettian site on Lemnos. This picture becomes even more interesting if one considers the possible involvement of these early hunter-gatherer and fishing groups in the events of the beginning of the Neolithic in Greece. The chronological horizons of the Mesolithic levels of both Maroulas and Youra are radiocarbondated to the middle of the ninth millennium b.c., with the occupation of the former perhaps starting a few centuries earlier (Sampson 2006). These dates are a close match for the late Paleolithic levels (phase VI) of Franchthi Cave (Perlès 2001). The occupation span of these most probably seasonal preNeolithic settlements is not easy to determine. As Sampson (2008b, 200) reports in the final publication of the site, the Cave of the Cyclops on Youra has produced only scarce Early Neolithic (EN) habitation evidence, along with a few radiocarbon dates, although the occupation of Maroulas seems to have continued well into the seventh millennium b.c., conceivably (pending its full publication) extending into the EN period (Sampson 2006). The maritime nature of these early island hunter-gatherer communities, which seem to have exhibited different degrees of permanency, is unequivocal, however. The occupants of both Maroulas and Youra settled in remote and small island environments and very effectively made use of a large variety of fish with a seemingly efficient fishing technology employing bone hooks of different sizes. They were also involved in the circulation of raw materials such as obsidian from Melos. Moreover, they hunted or exploited “wild” animals like pigs and goats that were either endemic or, most probably, introduced and perhaps constantly replenished by these traveling groups (see Horwitz, this vol., Ch. 9). Many aspects of the material culture of these archaeologically ill-defined pre-Neolithic or Mesolithic fishing communities in the Aegean are intriguing. Both the lithic material—with its emphasis

206

NIKOS EFSTRATIOU

on the production of flakes—and the early ninth millennium b.c. radiocarbon dates of Youra and Kythnos raise the possibility of an Epipaleolithic/ Epigravettian provenance (Kozłowski 1999; Sampson and Kozłowski 1999; Kaczanowska and Kozłowski 2006, 67; Sampson 2006, 46). Unfortunately, the limited number of Mesolithic chipped stone artifacts from Youra (179 pieces in total) and their stratigraphic provenance do not allow for a detailed study of the technology (Kaczanowska and Kozłowski 2008, 170, 172, 178). The lithic typological and technological affinities of these assemblages to the Cypriot preNeolithic sites and the Levant industries have been proposed, however (Sampson and Katsarou 2004). One wonders if this similarity suggests the presence of seagoing forager and hunter-gatherer groups who visited the Aegean islands as early as the 11th millennium b.c. (Broodbank 2006; Ammerman 2010). If so, what were the motives of these apparently episodic visits, and what was the actual range of these early travelers? What were the social and historical circumstances of their mobility practices? Were these sea travels related only to specific fishing activities, or to exchange practices such as the obsidian trade, or to both (Ammerman 2010)? In this context the early presence of obsidian on coastal and island Aegean sites (Franchthi, Youra, Kythnos)—often voluminous as in the case of Kythnos—should not pass unnoticed (Perlès 1979; Sampson 2001, 2008b; Sampson et al. 2002). Does this in the long run constitute a maritime tradition? How far back in time can the maritime way of life be pushed, and when did it end? One may also ask if the evidence for early maritime activity represents a full-fledged, seagoing forager lifestyle, or if it reflects the performance of a more restricted set of practices carried out by forager task groups, as suggested respectively by Ammerman (2011) and Broodbank (2006, 211). The yet-to-be-determined closing date of the pre-Neolithic or Mesolithic stage in the Aegean will have important implications for our understanding of the transition to agriculture, for in theory it signals the beginning of the Neolithic in the Aegean islands and Greece (Perlès 2001). It has been suggested, however, that Mesolithic forager groups may have kept their distinct way of life even after the establishment of the first Neolithic communities on the islands by the beginning of the seventh millennium b.c. (Ammerman 2010,

2011). Is it reasonable to expect that two distinct ways of life existed side-by-side for a long period of time and well into the Neolithic period? It is interesting that Ammerman’s argument goes one step further in this respect. In his attempt to explain the belated arrival of the Neolithic in Crete, the Aegean, and places further west compared to Cyprus (also at a remarkably slow pace with regard to the alternative Balkan river routes; see Biagi, Shennan, and Spataro 2005), Ammerman proposes that these foraging groups, which were still occupying parts of the seascape well into the eighth millennium b.c., may have obstructed the arrival of the seaborne farming communities in the Aegean (Ammerman 2011). Only future archaeological research on the islands will permit the evaluation of this appealing idea. The founding of Neolithic Knossos in Crete a­round 7000 b.c. took place within the historical maritime locale of the Aegean that had already been established for some time in the eastern Mediterranean. But what are its main historical trajectories, and how can they be approached archaeologically? The case of Cyprus may offer a possible paradigm. In Cyprus seagoing forager groups seem to have begun their seasonal visits from the Levantine mainland from at least the 11th millennium b.c. These maritime visits apparently continued uninterrupted until the second half of the ninth millennium b.c., when the first Neolithic farmers from the northern Levant found their way to the southern shores and valleys of the island during PPNB (Peltenburg and Wasse, eds., 2004; Vigne and Cucchi 2005). The archaeological relationship between these two early horizons in Cyprus is far from clear, considering that AkrotiriAetokremnos may not have been the only site of the period. Rather, there were probably more sites all over the island, the northern coast included, and a complex and widespread pattern of interaction must have taken place (Ammerman and Noller 2005; Simmons 2008). Unfortunately, many aspects of the material record of the 10th millennium b.c. forager groups in Cyprus other than their tool typology and technology (i.e., their subsistence strategies, locational preferences, origins, and date) are still missing. The recently detected archaeological deposits in the underwater bay

KNOSSOS AND THE BEGINNING OF THE NEOLITHIC IN GREECE AND THE AEGEAN ISLANDS

of Aspros, situated in front of the pre-Neolithic site, give a glimpse of the prospects of further archaeological research (Ammerman et al. 2008). Based on the archaeological evidence available so far, however, these campsites hardly describe a well established pre-Neolithic colonization pattern of the island; they seem instead to represent episodes of short seasonal visits of mainland foragers to the island (Ammerman et al. 2006, 17). In 2008 a team of archaeologists from the Uni­ versity of Thessaloniki discovered the inland site of Rhoudias in the upper valley of the Xeros River. Rhoudias is provisionally dated to the preNeolithic horizon on the basis of its lithic assemblage. Its presence in the foothills of the Troodos Mountains indicates that forager groups were moving to the interior looking for raw material sources and hunting grounds. By the last quarter of the ninth millennium early farmers were permanently settled in Cyprus, having brought with them cultivated cereals and do­ mes­ ticated, proto-domesticated, and wild animals (Guilaine et al. 2000, 75). This pattern of introduction of wild and domesticated animals by human groups, which occurred on all the large Mediterranean islands including Crete, constituted a dynamic and integral practice within what we have described in historical terms as the maritime locale of the period (Horwitz, this vol., Ch. 9). At around 8200 b.c. these farming communities were well established in Cyprus, and an internal island Neolithic (Aceramic) evolutionary process was underway, leading by the seventh millennium to the development of the well-known Khirokitia culture. It took more than 1,000 years before a fully developed farming community appeared on another Mediterranean island, that of Crete. This is not to say that Cyprus was the origin of the Cretan farmers, although that is not a far-fetched possibility. As Evans (1994) suggested on the basis of archaeobotanical and architectural evidence, the other possible place of departure for the settlers of Neolithic Knossos is Anatolia. An Anatolian provenance of the early Knossian farmers is also entertained by Horwitz in her discussion of the origin of the site’s wild and domestic fauna (see Horwitz, this vol., Ch. 9). This hypothesis is reinforced by the findings of a recent comparative genetic study that shows a close Y-chromosomal affinity between

207

modern Cretan and Anatolian populations (King et al. 2008). Whether the findings of this study can be extrapolated to Neolithic human groups is open to question; it is nevertheless a good example of the potential supporting evidence that may be derived from contemporary genetic studies. It is a challenge to pinpoint the origin and actual maritime route of the first settlers since all of the eastern Aegean islands—northeast or southeast of Crete—are presently void of early Neolithic sites (Sampson 2006). There is no doubt, however, that the maritime route—long or short in distance— was of paramount importance for the spread of the Neolithic in the eastern Mediterranean. An Anatolian origin of the Knossos farmers would certainly help to explain the striking time gap between the Neolithic colonization of Cyprus and that of Crete, implying the presence of an overland route for the spread of the farming groups westward. Nevertheless, considering the multiple and dynamic pre-Neolithic maritime locales that were active in the Aegean and Cyprus in the 10th and ninth millennia (or even earlier as the Epigravettian site on Lemnos shows), it is difficult to discount a possible open sea route for Neolithic groups pursuing a Cyprus to Crete itinerary. Whatever the origin of the early Cretan farmers, the importance of Neolithic Knossos for early Aegean developments is incontestable. The precedent of 10th-millennium sites in Cyprus and the growing pre-Neolithic evidence from the Aegean islands point to two common and in many ways parallel historical trajectories: (1) at least 2,000 years of maritime “action” in which huntergatherer and fishing groups were apparently traveling between Cyprus and the opposite mainland, camping at selected beaches and probably moving into the interior, and (2) the visiting of the Greek archipelago by similar foraging groups from the ninth millennium b.c. onward. In both cases the long historical duration of their maritime activities seems adequate for the generation of specific social experiences within each of the two locales—similar but not identical—forming the subject of a fascinating archaeological inquiry that has just begun. In the case of the Aegean, we know nothing yet of the frequency of the foragers’ visits to the islands or the duration of every stop or habitation episode, the degree of permanence of their camps, or

208

NIKOS EFSTRATIOU

even the sources of the raw material they used for making tools—the kind of archaeological evidence that may be used for documenting both material provenance and human movement. One may argue that although the raw materials used seem to have been mostly local—quartz from Kythnos and flint from the Sporades—these traveling groups must also have searched for and consistently used good flint sources spotted during their long wandering. This practice remains to be documented in future fieldwork, however. Even so, we know something of their fishing practices—such as the preference for local catch—indicating adaptation to local conditions and probably long-term exploitation of the area. Pre-Neolithic evidence for the circulation of Melian obsidian from Youra, Kythnos, and Franchthi Cave further attests to the development of long-lived social networks in the region. Unfortunately, the pre-Neolithic archaeological palimpsests excavated and studied so far in the Aegean provide only a glimpse of the long historical time and locally particular circumstances that they represent. The available material is often enigmatic or at least complex, especially when viewed within the context of the wider eastern Mediterranean “neolithization” process. One example of this complexity is the claim that the foragers of Youra kept pigs and goats in a proto-domesticated or even a domesticated state in the mid ninth millennium b.c. (Trantalidou 2003, 165; Sampson 2005; 2008b, 202, 224). A similar presence of domesticated pigs has also been reported from contemporary levels at Maroulas in Kythnos (Sampson 2006, 59). This evidence would seem to argue, directly or indirectly, for a domestication process in the Aegean islands involving both Mesolithic groups and Neolithic farmers through a cultural diffusion or even a local, independent domestication process outside the Near East (Sampson 2006, 59; 2008b, 210, 225). Such a scenario presupposes either the presence of mature Mesolithic social groups who were confident enough to embark on this process by themselves or an osmosis between two different sets of historical actors—the much older forager groups and the newly emergent farmers—that occurred within a very active and competitive maritime Aegean social and political scene. The dates of the two different historical locales, characterized by alternative ways of life, are important and need to be further documented archaeologically.

It remains open to speculation as to whether we are facing a sequential transformation or two contemporary and parallel “ways of life,” one forager/ fishing and another farming. The latter possibility is gathering momentum from the latest finds in both Greece and Cyprus. There no evidence so far to suggest that pre-Neolithic fishing groups were involved in any meaningful way with the incoming farming communities, a point that is illustrated by the case of Knossos in the seventh millennium b.c. The suggestion by Kaczanowska and Kozłowski (2006, 82) that the lithic characteristics of the Aceramic and early EN I levels point to an Aegean Mesolithic tradition (eighth millennium b.c.) rather than to a PPN Anatolian tool technology is premature and cannot be endorsed without more evidence; it is unfortunate that the recent study of the early Knossos lithics by Conolly (2008, 88) is not particularly illuminating in this respect. The material evidence from the new Mesolithic discoveries in Crete (Strasser et al. 2010) may clarify some of these intriguing problems. Nevertheless, the mere presence of Mesolithic material now confirms that wandering pre-Neolithic forager and fishing groups did not leave the impressive mass of the Cretan island unexploited. Indeed, it is reasonable to expect that substantial early campsites will soon be discovered. Whether the early Holocene material record of Crete and the Aegean islands will yield evidence for farming features prior to the beginning of the seventh millennium b.c. (i.e., before Aceramic Knossos) is doubtful, however, considering the generally slow pace of the spread of the Neolithic westward (Rowley-Conwy 2004; Pinhasi, Fort, and Ammerman 2005; Guilaine 2007, 91, fig. 1). The earliest Neolithic radiocarbon dates from the Cave of the Cyclops on Youra are all grouped in the second half of the seventh millennium b.c., with evidence for a possible gap from the preceding Final Mesolithic, which is dated surprisingly late, after 7000 b.c. (Sampson 2008b, 210, table 12.1). Does this chronology exclude the coexistence of both a forager-fishing and a farming way of life in the Aegean? Certainly not, as Ammerman (2011, 31) rightly notes. Such a coexistence would point, nonetheless, toward a much more complex and less rigid historical, social, and cultural island scene than that which we archaeologists are prone to accept—a historical time characterized

KNOSSOS AND THE BEGINNING OF THE NEOLITHIC IN GREECE AND THE AEGEAN ISLANDS

by a plurality of collective, often contradictory experiences and representations, practices, choices, codes, and material lives. Cyprus presents a persuasive and welldocumented case study for the Aegean area (Swiny, ed., 2001). The present status of Knossos as a mature farming community with no obvious local antecedents is reminiscent of the supposed long-lasting loneliness of Khirokitia as it was perceived fifteen years ago (Guilaine et al. 2000; Guilaine 2005). The similarities are striking, and the course of events that led to the development of both communities may have been comparable.

209

The archaeological record of Neolithic Knossos from the deep Aceramic levels amply attests to the presence of a mature farming community with well-developed social characteristics and a subsistence complex based on fully domesticated plants and animals. The Knossos evidence, however, contrasts with the subsistence components of the much earlier (mid ninth millennium b.c.) Cypriot farming village of Shillourokambos in Cyprus. Nevertheless, on both islands domesticated and wild animals were introduced by sea, although this occurred later in the case of Knossos.

Local Developments and Large-Scale Processes A consideration of specific issues related to the beginning of farming on the islands reveals the subject’s complexity. Addressing and ultimately solving “small-scale” archaeological research issues is only part of the strategy that needs to be employed in order to understand such phenomena; it remains important not to lose track of the modulatory efficacy of the wider historical picture that prevailed in the eastern Mediterranean during the early Holocene period. To give an example, and leaving aside the archaeological difficulties of visibility, dating, and taphonomy related to early sites, I would like to stress the problems that surface when one attempts to distinguish “wild,” “managed,” or “domestic” fauna—more specifically goats and pigs—based on phenotypic criteria or even genetic findings and to incorporate them into the more general discussion of the first farmers outside the core area of the Near East. The relevant bibliography is extensive. The results of mitochondrial DNA analysis suggest that the pig was involved in a complex set of interactions between local and introduced animal populations (Larson et al. 2007, 15276). It is possible that in the case of the Aegean islands the mixing of local wild boar with introduced domesticated pigs (pig was already part of the Neolithic package) might have led to local interbreeding episodes taking place after the arrival of the first farmers. This, however, does not seem to have occurred at Youra and Kythnos, or

at least in their lower Mesolithic levels, which are radiocarbon-dated to the mid ninth millennium (Sampson 2008b, 209), unless one argues that the same interaction processes between humans and pigs documented in Anatolia and Cyprus (Vigne, Carrère, and Guilaine 2003) in the second half of the ninth millennium b.c. also took place in the Aegean islands, as suggested by Sampson (2006). This far-fetched suggestion needs to be supported by new pig genetic data as well as by archaeologically well-documented and stratigraphically secure early farming evidence from the same area. The genetic evidence is still pending, while the archaeological evidence for farming is still poor outside of Aceramic and EN I Knossos, which was founded 1,500 years later. On a more practical note, it should be remembered that wild pigs are thought to have reached the Aegean islands by boat in pre-Neolithic times, as they were not part of the endemic wild fauna. The claim for the presence of domesticated goat or its wild and feral forms in the pre-Neolithic levels of the Aegean islands is an even more complex subject, and more genetic data are needed before a plausible scenario can be proposed (Horwitz and Kahila Bar-Gal 2006). The island of Youra, with its present-day wild goat (Capra aegagrus) population that is morphologically similar to that of Crete (Capra aegagrus cretica), along with the archaeological data from the cave, may afford an

210

NIKOS EFSTRATIOU

informative comparative study involving genetic material from the two islands (Kahila Bar-Gal et al. 2002). Horwitz offers another possible scenario for the presence of fauna in the pre-Neolithic Aegean island sites and Neolithic Knossos, one that has many interesting implications for early colonization and seafaring motives and mechanisms (Horwitz, this vol., Ch. 9). She entertains the idea that not all animals that were introduced to the islands—Crete included—were domesticated, as is often suggested. Goat is a possible candidate for a wild introduced species, given the size and number of the animals present in the early material record of Knossos. This would mean that the first farmers of Crete brought with them, once or repeatedly, a blend of fully domesticated (sheep) and wild (goat) animals. The arguments laid out by Horwitz in her contribution are persuasive though heretic, and I have nothing to add at this stage apart from my suggestion that these well-orchestrated moves were historically and politically conditioned. They entailed a complex process in which human groups chose between a number of alternatives and exercised different forms of control. Once again (see also Efstratiou et al. 2004), I would like to stress that under these circumstances the suggestion that the founding of Knossos in Crete was an idiosyncratic episode of a Neolithic island colonization process involving groups with an exploring mentality and adventurous logistics (Broodbank 2006) need to be revised. Instead I suggest that the spread of farming communities across the Aegean islands, a movement that included the founding of Knossos,

was broad-fronted and did not bear the characteristics of a venture. This brings us back, in turn, to the “modulatory efficacy of the wider historical picture” mentioned above. The reconstruction of this wider historical scene is is as important archaeologically as the recovery of local data. The reasons for the slow spread of seagoing farming communities remain unknown. The presence of pre-Neolithic forager and fishing groups over a large geographic area, including some of the Aegean islands and groups with locally specialized practices of fishing, mobility, and raw materials acquisition, speaks to the long-term development of maritime experiences and their incorporation in “material life,” at least from the early Holocene period. Such island “circumstances”—environmental, social, ideological, and technological—must have generated in the long run their own structural socioeconomic and political characteristics. The historical actors of the period acquired autonomy but not isolationism, and that autonomy ultimately placed certain limits on their motives, practices, and opportunities. Although the hunter-gatherer groups and farming communities in the eastern Mediterranean had different, often antithetical social experiences and ideological codes, they were both the protagonists in the same historical scene. The archaeological palimpsest of the Knossos tell would seem to reflect a rich and dynamic cultural florescence taking place on the Cretan stage of the seventh millennium b.c., a process that appears full of apparent contradictions and is as yet open to multiple interpretations.

Conclusion The pre-Neolithic horizon of the Aegean islands reveals aspects of a number of long and dynamic maritime locales—an ensemble of well-founded and not static structural social, political, environmental, and technological features—that were active at least by the beginning of the Holocene. It is only a matter of time before specific archaeological case studies in the Aegean will be able to particularize these different locales in time and space, bringing to the foreground the socioeconomic processes,

choices, and ideational representations of the communities involved. It is premature to say whether their development was causally related to the environmental deterioration of the Younger Dryas (Broodbank 2006, 210) and its alleged impact on the social organization and mobility of population groups in the Levant, the adjacent areas of Anatolia and Cyprus, or more distant areas such as the Aegean (Bar-Yosef 2001). We would expect that the resolution of this question will ultimately depend

KNOSSOS AND THE BEGINNING OF THE NEOLITHIC IN GREECE AND THE AEGEAN ISLANDS

greatly on the material evidence (radiocarbon dating, stratigraphy, subsistence) from the new preNeolithic sites discovered in Cyprus (Nissi Beach, Aspros, Hagia Varvara-Paliokamina, Rhoudias) and the Aegean (Youra, Kythnos, Ikaria). The affinities of the lithic assemblages from these sites with Epipaleolithic societies either of the eastern Mediterranean or the Balkans is another interesting line of evidence to be considered. The new Epigravettian site found on Lemnos is important in this respect, as it may bridge the cultural gap between the late Pleistocene and early Holocene

211

periods in the Aegean, exemplifying the characteristics of a long maritime continuum and way of life. Similarly, the relationship between the preNeolithic sites and their Neolithic successors, both in Cyprus and the Aegean, will remain the principal focus of early prehistoric studies. Archaeological research in both of these regions will help to define better the different locales involved, their dynamism, and their interaction, thus giving back to these early groups of foragers, fishermen, and farmers their history.

References Ammerman, A.J. 2010. “The First Argonauts: Towards the Study of the Earliest Seafaring in the Mediterranean,” in Global Origins and Development ­ ­ ­ of Seafaring, A. Anderson, J. Barrett and K. Boyle, eds., Cambridge, pp. 81–92. ———. 2011. “The Paradox of Early Voyaging in the Mediterranean and the Slowness of the Neolithic Transition between Cyprus and Italy,” in The Seascape in Aegean Prehistory (Monographs of the Danish Institute at Athens 14), G. Vavouranakis, ed., Athens, pp. 31–49. Ammerman, A.J., P. Flourentzos, R. Gabrielli, T. Highan, C. McCartney, and T. Turnbull. 2008. “Third Report on Early Sites on Cyprus,” RDAC 2008, pp. 1–32. Ammerman, A.J., P. Flourentzos, R. Gabrielli, P. McCartney, J. Noller, D. Peloso, and D. Sorabji. 2007. “More on the New Early Sites on Cyprus,” RDAC 2007, pp. 1–26. Ammerman, A.J., P. Flourentzos, C. McCartney, J. Noller, and D. Sorabji. 2006. “Two New Early Sites on Cyprus,” RDAC 2006, pp. 1–22. Ammerman, A.J., and J.S. Noller. 2005. “New Light on Aetokremnos,” WorldArch 37, pp. 533–543. Bailey, G. 2007. “Time Perspectives, Palimpsests and the Archaeology of Time,” JAnthArch 26, pp. 198–223. Bar-Yosef, O. 2001. “The World around Cyprus: From Epi-Palaeolithic Foragers to the Collapse of the PPNB Civilization,” in The Earliest Prehistory of Cyprus: From Colonization to Exploitation (Cyprus American Archaeological Research Institute Monograph Series 2), S. Swiny, ed., Boston, pp. 129–164.

Biagi, P., S. Shennan, and M. Spataro. 2005. “Rapid Rivers and Slow Seas? New Data for the Radiocarbon Chronology of the Balkan Peninsula,” in Prehistoric Archaeology and Anthropological Theory and Education (Reports of Prehistoric Research Projects 6–7), L. Nikolova and J. Higgins, eds., Salt Lake City, pp. 41–52. Braudel, F. 1972. The Mediterranean and the Mediterranean World in the Age of Philip II, Vol. 1, New York. Broodbank, C. 1992. “The Neolithic Labyrinth: Social Change at Knossos before the Bronze Age,” JMA 5, pp. 39–75. ———. 2000. An Island Archaeology of the Early Cyclades, Cambridge. ———. 2006. “The Origins and Early Development of Mediterranean Maritime Activity,” JMA 19, pp. 199–230. Broodbank, C., and T.F. Strasser. 1991. “Migrant Farmers and the Neolithic Colonization of Crete,” Antiquity 65, pp. 233–245. Cherry, J.F. 1981. “Pattern and Process in the Earliest Colonization of the Mediterranean Islands,” PPS 47, pp. 42–68. ———. 1990. “The First Colonization of the Mediterranean Islands: A Review of Recent Research,” JMA 3, pp. 145–221. Conolly, J. 2008. “The Knapped Stone Technology of the First Occupants at Knossos,” in Isaakidou and Tomkins, eds., 2008, pp. 73–89. Davis, J.L. 1992. “Review of Aegean Prehistory I: The Islands of the Aegean,” AJA 96, pp. 699–756.

212

NIKOS EFSTRATIOU

Efstratiou, N. 1985. Agios Petros: A Neolithic Site in the Northern Sporades (BAR-IS 241), Oxford.

Herzfeld, M. 1991. A Place in History: Social and Monumental Time in a Cretan Town, Princeton.

———. 2005. “Tracing the Story of the First Farmers in Greece—A Long and Winding Road,” in How Did Farming Reach Europe? Anatolian-European Relations from the Second Half of the 7th through the First Half of the 6th Millennium cal bc. Proceedings of the International Workshop, Istanbul, 20–22 May 2004 (Byzas 2), C. Lichter, ed., Istanbul, pp. 143–153.

Honea, K. 1975. “Prehistoric Remains on the Island of Kythnos,” AJA 79, p. 277.

———. 2007. “The Beginning of the Neolithic in Greece—Probing the Limits of a ‘Grand’ Narrative,” in Mediterranean Crossroads, S. Antoniadou and A. Pace, eds., Athens, pp. 123–138.

Horden, P., and N. Purcell. 2000. The Corrupting Sea: A Study of Mediterranean History, Oxford. Horwitz, L.K., and G. Kahila Bar-Gal. 2006. “The Origin and Genetic Status of Insular Caprines in the Eastern Mediterranean: A Case Study of Freeranging Goats (Capra aegagrus cretica) on Crete,” Human Evolution 21, pp. 123–138. Isaakidou, V., and P. Tomkins, eds. 2008. Escaping the Labyrinth: The Cretan Neolithic in Context (Sheffield Studies in Aegean Archaeology 8), Oxford.

Efstratiou, N., A. Karetsou, E. Banou, and D. Margomenou. 2004. “The Neolithic Settlement of Knossos: New Light on an Old Picture,” in Knossos: Palace, City, State. Proceedings of the Conference in Herakleion Organised by the British School at Athens and the 23rd Ephoreia of Prehistoric and Classical Antiquities of Heraklion, in November 2000, for the Centenary of Sir Arthur Evans’s Excavations at Knossos (BSA Studies 12), G. Cadogan, E. Hatzaki, and A. Vasilakis, eds., London, pp. 39–51.

———. 2008. “Chipped Stone Artifacts,” in Sampson 2008a, pp. 169–178.

Evans, J.D. 1994. “The Early Millennia: Continuity and Change in a Farming Settlement,” in Knossos: A Labyrinth of History. Papers in Honour of S. Hood, D. Evely, H. Hughes-Brock, and N. Momigliano, eds., Oxford, pp. 1–20.

Kahila Bar-Gal, G., P. Smith, E. Tchernov, C. Greenblatt, P. Ducos, A. Gardeisen, and L.K. Horowitz. 2002. “Genetic Evidence for the Origin of the Agrimi Goat (Capra aegegrus cretica),” Journal of Zoology 256, pp. 369–377.

Evans, J.D., and C. Renfrew. 1968. Excavations at Saliagos near Antiparos (BSA Suppl. 5), London. Galanidou, N., and C. Perlès, eds. 2003. The Greek Mesolithic: Problems and Perspectives (BSA Studies 10), London.

King, R.J., S.S. Özcan, T. Carter, E. Kalfoglou, C. ­ Triandaphyllidis, A. Kouvasi, A.A. Lin, C.-E.T. Chow, L.A. Zhivotovsky, M. Michalodimitrakis, and P.A. Underhill. 2008. “Differential Y-Chromosome Anatolian Influences on the Greek and Cretan Neolithic,” Annals of Human Genetics 72, pp. 205–214.

Gosden, C., and C. Pavlides. 1994. “Are Islands Insular? Landscape vs. Seascape in the Coast of the Arawe Islands, Papua New Guinea,” Archaeology in Oceania 29, pp. 162–171.

Knapp, A.B. 2007. “Insularity and Island Identity in the Prehistoric Mediterranean,” in Mediterranean Crossroads, S. Antoniadou and A. Pace, eds., Athens, pp. 37–63.

Guilaine, J. 2003. “Parekkleisha-Shillourokambos: périodisation et aménagements domestiques,” in Le Néolithique de Chypre (BCH Suppl. 43), J. Guilaine and A. Le Brun, eds., Athens, pp. 3–14.

Kozłowski, J.K. 1999. “Gravettian/Epigravettian Sequence in the Balkans: Environment, Technologies, Hunting and Raw Material Procurement,” in The Palaeolithic Archaeology of Greece and Adjacent Areas. Proceedings of the ICOPAG Conference, Ioannina, 1994 (BSA Studies 3), G. Bailey, E. Adam, E. Panagopoulou, C. Perlès, and K. Zachos, eds., London, pp. 319–320.

———. 2005. Aux marges des grands foyers du néolithique: Périphéries débitrices ou créatrices? (Collections des Hespérides), Paris. ———. 2007. “Les enjeux de Sidari,” in Prehistoric Corfu and Its Adjacent Areas: Problems-Perspectives, G. Metallinou, ed., Athens, pp. 91–97. Guilaine, J., F. Briois, J.-D. Vigne, and I. Carrère. 2000. “Découverte d’un néolithique précéramique ancien chypriote (fin 9e, début 8e millénaires cal. bc), apparenté au PPNB ancien/moyen du Levant nord,” Sciences de la terre et des planètes 330, pp. 75–82.

Kaczanowska, M., and J.K. Kozłowski. 2006. “Παλαιολιθικές παραδόσεις, Μεσολιθικές προσαρμογές και Νεολιθικές καινοτομίες στο Αιγαίο μέσα από το πρίσμα της λιθοτεχνίας,” in Sampson 2006, pp. 67–87.

Lambeck, K. 1996. “Sea-Level Change and Shore-Line Evolution in Aegean Greece since Upper Palaeolithic Time,” Antiquity 70, pp. 588–611. Larson, G., U. Albarella, K. Dobney, P. RowleyConwy, J. Schibler, A. Tresse, J.-D. Vigne, C. Edwards, A. Schlumbaum, A. Dinu, A. Balacsescu, G. Dolman, A. Tagliacozzo, N. Manaseryan, P.

KNOSSOS AND THE BEGINNING OF THE NEOLITHIC IN GREECE AND THE AEGEAN ISLANDS

Miracle, L. Van-Wijngaarden-Bakker, M. Masseti, D. Bradley, and A. Cooper. 2007. “Ancient DNA, Pig Domestication, and the Spread of the Neolithic into Europe,” Proceedings of the National Academy of Sciences of the United States of America 104 (39), pp. 15276–15281. MacArthur, R.H., and E.O. Wilson. 1967. The Theory of Island Biogeography, Princeton. Manning, S.W. 1999. “Knossos and the Limits of Settlement Growth,” in Meletemata: Studies in Aegean Archaeology Presented to Malcolm H. Wiener on the Occasion of His 65th Birthday (Aegeum 20), P. Betancourt, V. Karageorghis, R. Laffineur, and W.-D. Niemeier, eds., Liège, pp. 469–482. McCartney, C. 2004. “Cypriot Neolithic Chipped Stone Industries and the Progress of Regionalization,” in Neolithic Revolution: New Perspectives on Southwest Asia in Light of Recent Discoveries on Cyprus (Levant Suppl. 1), E. Peltenburg and A. Wasse, eds., Oxford, pp. 103–122. ———. 2005. “Preliminary Report on the Re-survey of Three Early Neolithic Sites in Cyprus,” RDAC 2005, pp. 10–21. Mithen, S. 2003. After the Ice: A Global Human History, 20,000–5,000 bc, London. Patton, M. 1996. Islands in Time: Island Sociogeography and Mediterranean Prehistory, London. Peltenburg, E., S. Colledge, P. Croft, A. Jackson, C. McCartney, and M.A. Murray. 2001. “Neolithic Dispersals from the Levantine Corridor: A Mediterranean Perspective,” Levant 33, pp. 36–64. Peltenburg, E., and A. Wasse, eds. 2004. Neolithic Revolution: New Perspectives on Southwest Asia in Light of Recent Discoveries on Cyprus (Levant Suppl. 1), Oxford. Perlès, C. 1979. “Des navigateurs méditerranéens il y a 10,000 ans,” La Recherche 96, pp. 82–83. ———. 1987. Les industries du mésolithique et du néolithique initial (Franchthi 2), Bloomington. ———. 2001. The Early Neolithic in Greece: The First Farming Communities in Europe, Cambridge. Pinhasi, R., J. Fort, and A.J. Ammerman. 2005. “Tracing the Origin and Spread of Agriculture in Europe,” Public Library of Science Biology 3, pp. 2220–2228. Rosen, A.M. 2007. Civilizing Climate: Social Response to Climate Change in the Ancient Near East, Lanham, MD. Rowley-Conwy, P. 2004. “How the West Was Lost,” CurrAnthr 45, pp. 83–113.

213

Runnels, C., E. Panagopoulou, P. Murray, G. Tsartsidou, S. Allen, K. Mullen, and E. Tourloukis. 2005. “A Mesolithic Landscape in Greece: Testing a Sitelocation Model in the Argolid at Kandia,” JMA 18, pp. 259–285. Sampson, A. 1998. “The Neolithic and Mesolithic Occupation of the Cave of Cyclope, Youra, Alonnessos, Greece,” BSA 93, pp. 1–22. ———. 2001. Αρχαιολογική έρευνα στις Βόρειες Σποράδες, Alonnessos. ———. 2005. “Early Productive Stages in the Aegean Basin from the 9th to 7th Mill bc,” in How Did Farming Reach Europe? Anatolian-European Relations from the Second Half of the 7th through the First Half of the 6th Millennium cal bc. Proceedings of the International Workshop, Istanbul, 20–22 May 2004 (Byzas 2), C. Lichter, ed., Istanbul, pp. 131–141. ———. 2006. Προϊστορία του Αιγαίου, Athens. ———. 2008a. The Cave of the Cyclops: Mesolithic and Neolithic Networks in the Northern Aegean, Greece Vol. I: Intra-Site Analysis, Local Industries, and Regional Site Distribution (Prehistory Monographs 21), Philadelphia. ———. 2008b. “Conclusions,” in Sampson 2008a, pp. 199–227. Sampson, A., and S. Katsarou. 2004. “Cyprus, Aegean, and Near East during the PPN,” Neo-Lithics: The Newsletter of Southwest Asian Neolithic Research 1/04, pp. 13–15. Sampson, A, M. Kaczanowska, and J.K. Kozłowski. 2009. “The First Mesolithic Site in the Eastern Part of the Aegean Basin: Excavations into the Site Kerame I on the Island of Ikaria in 2008,” Rocznik Polskiej Akademii Umiejtnosci 1958, pp. 321–329. Sampson, A., and J.K. Kozłowski. 1999. “The Cave of Cyclope in the Northern Aegean: A Specialized Fishing Shelter of the Mesolithic and Neolithic Periods,” Neo-Lithics: The Newsletter of Southwest Asian Neolithic Research 3/99, pp. 5–7 Sampson, A., J.K. Kozłowski, and M. Kaczanowska. 2003. “Mesolithic Chipped Stone Industries from the Cave of Cyclope on the Island of Youra (Northern Sporades),” in The Greek Mesolithic: Problems and Perspectives (BSA Studies 10), N. Galanidou and C. Perlès, eds., London, pp. 123–130. Sampson, A., J. K. Kozłowski, M. Kaczanowska, and B. Giannouli. 2002. “The Mesolithic Settlement at Maroulas, Kythnos,” Mediterranean Archaeology and Archaeometry 2, pp. 45–67. Simmons, A.H. 2008. “American Researchers and the Earliest Cypriots,” NEA 71 (1–2), pp. 21–29.

214

NIKOS EFSTRATIOU

Simmons, A.H., and R. Mandel. 2007. “Not Such a New Light: A Response to Ammerman and Noller,” WorldArch 39, pp. 475–482. Strasser T.F., E. Panagopoulou, C.N. Runnels, P.M. Murray, N. Thompson, P. Karkanas, F.W. McCoy, and K.W. Wegmann. 2010. “Stone Age Seafaring in the Mediterranean: Evidence from the Plakias Region for Lower Palaeolithic and Mesolithic Habitation of Crete,” Hesperia 79, pp. 145–190. Swiny, S., ed. 2001. The Earliest Prehistory of Cyprus: From Colonization to Exploitation (Cyprus American Archaeological Research Institute Monograph Series 2), Boston. Tomkins, P.D. 2007. “Communality and Competition: The Social Life of Food and Social Transformation in the Aegean before the Bronze Age,” in Cooking up the Past: Food and Culinary Practices in the Neo­ lithic and Bronze Age Aegean, C. Mee and J. Renard, eds., Oxford, pp. 174–199. ———. 2008. “Time, Space and the Reinvention of the Cretan Neolithic,” in Escaping the Labyrinth: The Cretan Neolithic in Context (Sheffield Studies in

Aegean Archaeology 8), V. Isaakidou and P. Tomkins, eds., Oxford, pp. 21–48. Torrence, R. 1986. Production and Exchange of Stone Tools: Prehistoric Obsidian in the Aegean, Cambridge. Trantalidou, C. 2003. “Faunal Remains from the Earliest Strata of the Cave of Cyclops, Youra,” in The Greek Mesolithic: Problems and Perspectives (BSA Studies 10), London, pp. 143–172. Vigne, J.-D., I. Carrère, and J. Guilaine. 2003. “Unstable Status of Early Domestic Ungulates in the Near East: The Example of Shillourokambos (Cyprus, IX–VIIIth Millennia cal. b.c.),” in Le néolithique de Chypre (BCH Suppl. 43), J. Guilaine and A. Le Brun, eds., Athens, pp. 239–251. Vigne, J.-D., and T. Cucchi. 2005. “Premières navigations au Proche-Orient: Les informations indirectes de Chypre,” Paléorient 31, pp. 186–194. Whitelaw, T.M. 1992. “Lost in the Labyrinth? Comments on Broodbank’s Social Change at Knossos before the Bronze Age,” JMA 5, pp. 225–238.

Index

abandonment, xxvii, 30, 32, 36, 41, 131 Aceramic, ix, xv, xix, xx, 2, 4, 5, 19, 21, 23, 26–32, 34, 36, 41, 42, 55, 60, 63, 64, 66, 68–74, 79, 88–91, 94, 95, 100, 109, 110, 113, 116, 119–121, 123, 125–128, 131, 134, 138, 162, 165, 168, 172, 174–178, 180–186, 188, 190, 199, 204, 207–209 agriculture, 27, 44, 65–66, 80, 87, 89, 91–93, 95, 97, 114, 131–133, 160, 168, 180, 206, 213 Anatolia, 28, 42, 64, 70, 139, 172, 182, 183, 185–187, 190, 207–210, 212 archaeobotany, 66, 92, 93, 117, 118 archaeozoology, 185, 187, 188, 190, 191

barley, 28, 32, 38, 42, 67, 68, 70, 73, 74, 77, 84, 87, 88, 92, 119, 121–124, 127, 128, 130, 131, 173

cattle, xvi, 27, 31, 32, 36, 37, 39, 42, 64–66, 133, 134, 138, 139, 152, 157, 158–161, 164, 174–187, 189, 191 Central Court, xiii, xix, xx, xxv, 1–3, 5, 21, 22, 25, 26, 31, 35, 37, 39, 40, 43, 63, 95, 100, 120, 126, 130, 134, 171, 194–198 ceramic, vii, xiii, xx, xxi, xxvi, xxvii, 2, 5, 6, 19, 25–27, 30, 31, 33–35, 37, 41–45, 47–51, 181, 184

cereal crops, 32, 70, 128, 131 charcoal, vii, x, xi, xv, 2, 4, 6, 8, 10, 12, 14, 18, 26, 29– 32, 35–38, 40, 42, 54, 55, 58–60, 95, 97, 99–102, 104– 118, 120, 128, 194–197 chronology, xi, xx, 26, 27, 29, 34, 40, 43, 182–184, 193, 197, 204, 205, 208, 211 colonization, xxvi, xxvii, 41, 43, 44, 50, 64, 91–93, 120, 131, 132, 155, 164–168, 176, 179, 182, 187, 188, 190, 201, 203, 204, 207, 210, 211, 213 community, xxi, 21, 23, 26, 33–35, 37–43, 151, 156, 160, 173, 190, 191, 203, 207, 209 crops, 27, 32, 38, 65, 69, 70, 78, 87, 88, 91, 97, 119, 120, 128, 130, 131, 180, 187 cultivation, 27, 29, 32, 33, 36, 37, 65, 69, 70, 77–80, 89– 91, 93, 94, 109, 112, 114, 118, 130, 162, 173 Cyprus, xxvi, xxvii, 38, 40, 41, 43, 45, 66, 93, 112, 119, 120, 132, 133, 139, 162, 164–168, 171, 172, 180, 182, 183, 186, 188–191, 202, 204–211, 213, 214

deer, 133, 162, 164, 166, 167, 174, 180–182, 185 diffusion, 64, 172, 187, 188, 190, 208 dog, x, xvi, 27, 31, 43–45, 51, 64, 117, 133, 134, 139, 140, 151, 152, 155, 161, 164, 172, 174–176, 183–185, 188, 200, 212

216

THE NEOLITHIC SETTLEMENT OF KNOSSOS IN CRETE

domestication, 27, 41, 69, 80, 92–94, 118, 132, 133, 138, 155, 164–166, 168, 175, 180, 184–192, 208, 212

Early Neolithic, xv, xxvi, xxvii, 1, 2, 25, 28, 29, 33, 34, 42, 44–47, 51, 53, 63, 73, 80, 92, 93, 95, 100, 113, 117, 119, 120, 166, 172, 175, 189–191, 193, 201, 203, 207, 213 economic, xix, 33, 34, 38, 39, 41, 64, 66, 117, 120, 152, 157, 187, 190 economy, vii, 28, 37, 42, 44, 63, 92, 93, 117, 119–121, 126, 131, 133, 140, 152, 166, 175, 179, 183, 185, 187, 189 einkorn, 27, 32, 42, 67, 70, 74, 77, 88, 90, 119 emmer, 27, 32, 42, 67, 69, 70, 74, 77, 88, 90, 132 environment, 32, 36, 38, 40, 41, 64, 65, 79, 93, 109, 110, 112–114, 116, 118, 120, 121, 131, 132, 172, 187, 191, 205, 212 Epipaleolithic, 113, 151, 205, 206, 211 exchange, 28, 33–35, 38, 45, 50, 51, 92, 187, 206, 214

fabric, vii, xxvi, 27, 33–35, 40, 42, 44, 47–50 farmers, xxvi, 23, 27, 29, 32, 40, 64, 65, 87, 93, 94, 109, 115, 116, 119, 132, 164, 165, 187, 188, 191, 203, 206–212 farming, xix, xxi, xxii, xxvi–xxviii, 23, 28, 34, 36, 39– 42, 44, 45, 51, 64–66, 92, 93, 109, 111, 115, 116, 119, 120, 132, 140, 156, 157, 160, 162, 164, 189–191, 203– 210, 212, 213 fauna, x, xi, xxvi, xxvii, 2, 26–28, 31, 34, 35, 37–42, 44, 45, 92, 93, 117, 133, 134, 138, 141–152, 155, 159– 162, 164–169, 174, 176, 178, 181, 182, 184–187, 189– 192, 207, 209, 210, 214 feral, 38, 80, 139, 162, 166, 174, 176–178, 189, 209 Final Neolithic, xxi, 1, 2, 40, 164, 193 fireplace, 8, 17, 18, 20, 31, 35 foraging, 65, 91, 93, 190, 204, 206, 207 forest, 97, 114, 116, 117, 128 fruits, 28, 36, 69, 80, 89, 90, 111

geomorphology, 54, 201 goat, x, xvi, xvii, 27, 28, 31, 36–40, 42, 44, 64, 66, 133, 134, 138–140, 151–153, 156, 157, 159–162, 164, 166, 167, 169, 174–186, 189–192, 205, 208–210, 212

habitat, 28, 29, 33, 97, 109, 113 habitation, xvii, xix, xxi, xxvi, xxviii, 6, 14, 16, 18, 19, 22, 23, 26, 30–32, 35, 37, 41, 43, 45, 66, 87, 88, 94, 119, 164, 182, 191, 195, 197–199, 204, 205, 207, 213 hearth, xiii, xiv, 5, 6, 9, 10, 14, 16–20, 22, 23 77, 80, 126, 151

herding, 36, 39, 42, 92, 111, 157, 175, 190, 191 holocene, xxv, xxvi, 28, 29, 41, 44, 45, 53, 91, 93, 112– 114, 116, 117, 155, 166–169, 175, 189, 190, 192, 201, 202, 204, 208–210 house, 6, 19, 22, 30, 31, 44, 93, 158, 181, 182, 188, 196 hunters-gatherers, 64, 151 husbandry, 36, 88, 92, 94, 114, 131, 133, 139, 155, 156, 162, 164, 172, 187

ideological, xxvi, 33, 34, 38, 210 innovation, 35, 37, 39 isolationism, xxvii, 38, 210

Katsambas, xx, 22, 42, 47 Kephala, xix, xx, xxvi, 21–23, 26, 27, 29, 32, 203 kouskouras, xiii, 4–6, 8, 10, 14, 19, 22, 31, 35, 37, 39, 43, 53, 96

landscape, 32, 42, 43, 45, 51, 66, 87, 93, 97, 101, 112, 114, 116–118, 212 Late Neolithic, 1, 2, 25, 34, 40, 42, 47, 63, 87, 91, 94, 95, 100, 113, 117, 119, 120, 172, 180, 193, 200 layers, xxv, 5, 6, 10, 12, 19, 25–27, 29–31, 37, 39, 41, 48, 120, 123, 126, 131, 140 lithic assemblage, 28, 173, 204, 207, 211 locale, 201, 202, 206–208, 210, 211

management, 36, 42, 44, 111, 115, 116, 166, 174, 176, 180, 186, 189 Melos, xvi, 28, 89, 171, 173, 174, 182, 205 Mesolithic, xx, xxvi, xxviii, 28, 41, 45, 80, 87, 94, 113, 116, 118, 119, 132, 168, 172, 188, 191, 204–206, 208, 209, 212–214 microliths, 28 Middle Neolithic, 1, 6, 25, 28, 47, 53, 63, 84, 87, 95, 100, 119, 134, 171, 193 migration, 43, 64, 172 Minoan, xiii, xx, xxi, 34, 35, 45, 63, 94, 95, 119, 134, 152, 162, 163, 165, 172, 181, 194, 197, 203 mudbrick, 10, 19, 23, 27, 30, 31, 43, 64, 185, 196

Near East, xv, xix, xxvi, 45, 65, 66, 69, 70, 73, 79, 80, 84, 91–93, 130, 139, 164, 168, 187, 188, 190, 191, 204, 208, 209, 213, 214

obsidian, 19, 28, 50, 64, 171, 174, 182, 204–206, 208, 214 olive, 32, 36, 38–40, 42, 65, 79, 91, 97, 112–118

INDEX

package, 27, 32, 36, 41, 64, 79, 89, 126, 130, 131, 133, 172, 178, 179, 181, 187, 209 Paleolithic, xxviii, 43, 45, 93, 117, 191, 211–213 phytoliths, vii, xv, x, xxi, xxvii, 26, 28, 36, 38, 40, 119– 121, 123, 126–132 pig, xvi, 27, 31, 32, 64, 133, 134, 139, 152–154, 158, 160–162, 164, 167, 171, 174–191, 204–209, 211, 212 pisé, 10, 14, 22, 23, 31, 35, 43 pit, xiv, 6, 19, 21, 22, 29, 31, 37, 173, 195 plant, vii, x, xv, xvi, 29, 32, 35, 36, 38, 40–43, 64, 65, 67, 68, 77–80, 84, 89, 91–101, 104, 106, 110–121, 123, 125–132, 140, 162, 167, 172, 174, 179, 187, 209 Pleistocene, xxv, xxvi, 28, 41, 44, 45, 93, 96, 111, 112, 155, 166–169, 174, 175, 178, 179, 187, 189, 190, 192, 204, 205, 211 pottery, xix, xxii, xxvi, xxvii, 1, 4–6, 8, 10, 14, 16, 18, 19, 23, 26, 27, 30, 31, 33–35, 37, 39, 40, 42, 44, 45, 47, 48, 50, 51, 54, 55, 58, 59, 63, 64, 79, 94, 100, 151, 164, 167, 172, 182, 190, 201, 203 pre-Neolithic, xxvi, 26–28, 38, 41, 101, 113, 162, 164, 168, 171, 174, 178, 179, 191, 202, 204–211 proto-domestic, 174, 175, 178–181, 184–187, 207, 208

radiocarbon dates, viii, xi, xvii, xix, xx, xxi, xxvii, 27, 30, 31, 37, 39, 173, 182, 183, 193–195, 197, 198, 205, 206, 208

seafaring, xxvii, xxviii, 28, 43, 45, 93, 179, 191, 205, 210, 211, 213 sedimentology, vii, ix, xxvii, 26, 35, 53, 55, 56, 60 sheep, x, xvi, 27, 31, 36–39, 42, 64, 66, 133, 134, 138– 140, 151, 152, 156, 157, 159–162, 164, 167, 169, 174– 186, 191, 192, 210

217

social, xix, xxvii, xxviii, 23, 24, 33–35, 37, 39–46, 51, 65, 66, 130, 165, 184, 202–204, 206–214 social relations, 23, 39 stratigraphic sequence, xxvi, xxvii, 21, 26, 27, 34, 49, 50, 140, 205 structures, 5, 6, 10, 19, 21, 22, 27, 31, 37, 120, 132, 202, 203 subsistence, xxvii, 27, 32, 36–39, 41–43, 64, 109, 111, 131, 133, 180, 184, 186, 187, 191, 206, 209, 211

taphonomy, 126, 134, 209 technology, xx, xxvi, xxvii, 26–28, 30, 33, 34, 37, 39, 43, 44, 47, 50, 64, 89, 204–206, 208, 211

vegetation, vii, xv, 29, 32, 33, 36, 38, 39, 41, 42, 44, 66, 89, 91–93, 95, 97, 98, 100, 106, 107, 109–118, 126, 130 village, xxvi, 27–29, 31, 32, 34, 36, 40, 41, 158, 184, 189, 190, 209

walls, xiv, 6, 10–17, 19, 22, 23, 31, 35 wheat, xvi, 28, 32, 36, 38, 42, 64–70, 73, 77, 84, 87–93, 121–124, 127–132, 173, 185 wild, xi, xvi, xvii, 27–29, 31, 32, 35–38, 40, 42, 43, 50, 65, 67, 69, 70, 77, 79, 80, 84, 85, 88–93, 97, 111–115, 126, 128, 130, 131, 133, 139, 151–155, 161–165, 167, 171, 172, 174–182, 184–187, 190, 191, 205, 207, 209, 210 woodland, 29, 32, 36, 39, 42, 97, 101, 109–116