234 23 88MB
English Pages 504 Year 2015
Bleda S. Düring, Claudia Glatz (Eds.) Kinetic Landscapes The Cide Archaeological Project: Surveying the Turkish Western Black Sea Region
Bleda S. Düring, Claudia Glatz (Eds.)
Kinetic Landscapes The Cide Archaeological Project: Surveying the Turkish Western Black Sea Region
Managing Editor: Katarzyna Michalak Associate Editor: Łukasz Połczyński Language Editor: Nadia Smith
Published by De Gruyter Open Ltd, Warsaw/Berlin Part of Walter de Gruyter GmbH, Berlin/Munich/Boston
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivs 3.0 license, which means that the text may be used for non-commercial purposes, provided credit is given to the author. For details go to http://creativecommons.org/licenses/by-nc-nd/3.0/.
Copyright © 2015 Bleda S. Düring, Claudia Glatz and chapters’ contributors ISBN: 978-3-11-044496-4 e-ISBN: 978-3-11-044497-1 Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.dnb.de. Managing Editor: Katarzyna Michalak Associate Editor: Łukasz Połczyński Language Editor: Nadia Smith www.degruyteropen.com Cover illustration: © Photo of Cide coastal stretch looking east from ridge at Abdulkadir
Contents Acknowledgements
XII
List of Contributors
XIV
List of Cide Archaeological Project (CAP) Team Members List of Cide Archaeological Project (CAP) Publications
XVI XVIII
Bleda S. Düring and Claudia Glatz 1 The Origins and Development of the Cide Archaeological Project Bibliography 7
1
Bleda S. Düring and Claudia Glatz 2 Remote Landscapes in Flux – The Cide and Şenpazar Region 9 2.1 The Cide-Şenpazar Landscapes 9 2.1.1 Ecology and Climate 9 2.1.2 Geology 12 2.1.3 Coastal Changes of the Black Sea 13 2.2 Humans in the Cide-Şenpazar Landscapes 15 2.2.1 Transport and Communication in a Rugged Landscape 15 2.2.2 Resources and Livelihoods in the Cide-Şenpazar Region 21 2.3 A Landscape in Flux 26 2.4 Summary 31 Bibliography 31 Murat Karasalihoğlu and Bleda S. Düring 3 Traveller’s Notes on the Cide Region through the Ages 3.1 Cide in Greek and Roman Sources 37 3.2 The Cide Region in the Ottoman Period 41 3.3 Conclusion 48 Bibliography 50
36
Claudia Glatz, Bleda S. Düring and Toby C. Wilkinson, with contributions by Bernard Gratuze, Richard Jones, Effie Photos-Jones, and Victor Klinkenberg 4 4.1 4.1.1 4.1.2
Developing an Adaptive Field Methodology for Challenging Landscapes 53 Traditions of Archaeological Survey in Turkey and Beyond 54 Extensive and Intensive Survey Traditions 54 Regional Survey in Turkey 55
4.2 4.2.1
Ambitions for CAP: A Hybrid and Adaptive Method 57 Sampling: Ecological and Topographic Zones, Visibility and Accessibility 59 4.2.2 Recording Frameworks: Data Models, Forms, Digital Database and GIS 63 4.2.3 Spatial Frameworks: Sites and Units 65 4.3 Field Methods 68 4.3.1 Remote Sensing, Mapping and Desk-Based Survey 70 4.3.2 Targeted Reconnaissance 71 4.3.3 Local Memories: Oral History and Historical Photographs 75 4.3.4 Tract-Walking and Unit-Walking 76 4.3.5 Grid-Square Collection 81 4.4 Analytical Recording Procedures and Approaches to Finds 82 4.4.1 Pottery Procedures 82 4.4.1.1 Pottery and Tile Recording and Macroscopic Analysis 82 4.4.1.2 Laboratory-based Pottery Analyses 83 4.4.2 Lithic Procedures 86 4.4.2.1 Lithic Macroscopic Inspection and Analysis 86 4.4.2.2 Sourcing Obsidian 88 4.5 Dating Procedures 89 4.5.1 Comparative Dating of Pottery, Lithics and Features 89 4.5.2 Absolute Dating 92 4.6 Geomorphological Research 92 4.7 Geochemical Research 93 4.8 Archaeometallurgy 94 4.9 Concluding Comments: Assessing the CAP Methodology 95 Bibliography 98 Bleda S. Düring and Victor Klinkenberg 5
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC 104 5.1 The Problem of the Early to Mid-Holocene Occupation of Northern Turkey 104 5.1.1 Dating Prehistoric Assemblages 106 5.2 The Palaeolithic (1,000,000-10,000 BC) 109 5.3 The Early Holocene (10,000-5500 BC) 110 5.3.1 Okçular and Abdulkadir 111 5.3.2 Aybasan Mahallesi 117 5.3.3 Çamdibi 123 5.3.4 Obsidian Exchange Networks 124 5.3.5 The Early Holocene in the Cide-Şenpazar Region: By Way of Summary 126 5.4 The Chalcolithic (5500-3000 BC) 127
5.4.1 5.4.2 5.4.3 5.4.4
Okçular Kale İni 128 Derebağ Köy Mağarası 136 Kılıçlı Mağarası 138 The Chalcolithic in the Cide-Şenpazar Region: By Way of Summary 140 5.5 Conclusion 141 Catalogue of Illustrated Artefacts 142 Bibliography 147 Claudia Glatz, with contributions by Michele Massa 6
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC 151 6.1 Chronological Considerations 151 6.2 Ceramic Assemblages from the Later Early Bronze Age in the CideŞenpazar Region 152 6.2.1 Kılıçlı Mağarası 152 6.2.2 Okçular Kale İni 165 6.2.3 Derebağ Köy Mağarası 165 6.2.4 Summary 166 6.3 Early Bronze Age Communities in the Central Turkish Black Sea Region 167 6.4 Conclusions 173 Catalogue of Illustrated Artefacts 174 Bibliography 178 Claudia Glatz 7
Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region 183 7.1 Issues and Open Questions 183 7.2 The Middle and Late Bronze Age Evidence from Cide-Şenpazar 185 7.2.1 The Middle Bronze Age (2000-1600 BC) 186 7.2.2 The Late Bronze Age (1600-1200 BC) 190 7.2.3 A Local Ceramic Tradition 192 7.3 The Turkish Central Black Sea Region in the Middle and Late Bronze Ages 194 7.4 Settlement Trends in the Turkish Central Black Sea Region 199 7.5 Middle and Late Bronze Age Cide in Context 200 7.6 Filling Empty Landscapes 202 7.7 Conclusions 204 Catalogue of Illustrated Artefacts 205 Bibliography 206
T. Emre Şerifoğlu 8 8.1
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC) 212 North and North-Central Anatolia in the Iron Age – The Historical Perspective 213 8.2 North and North-Central Anatolia in the Iron Age – The Archaeological Evidence 216 8.3 The Iron Ages in Cide and Şenpazar 221 8.3.1 Iron Age I 222 8.3.2 Iron Age II 225 8.3.3 Iron Age III 227 8.4 Locales of the Cide Iron Age 229 8.4.1 Gideros 229 8.4.2 Iron Age Caves 230 8.5 Final Remarks 232 Catalogue of Illustrated Artefacts 232 Bibliography 236 T. Emre Şerifoğlu and Caner Bakan 9 The Cide-Şenpazar Region during the Hellenistic Period (325/300 - 1 BC) 9.1 The Hellenistic Period on the Turkish Black Sea Coast 246 9.2 Hellenistic Settlement Patterns 247 9.3 Trade and Economy 248 9.4 Archaeological Evidence from Cide and Şenpazar 249 9.5 Conclusion 254 Catalogue of Illustrated Artefacts 254 Bibliography 255 Philip Bes 10 The Cide-Şenpazar Region in the Roman Period 260 10.1 Historical Background 262 10.2 The Pottery — Breakdown by Period and Functional Class 10.2.1 Table Wares 265 10.2.2 Amphorae 266 10.2.3 Cooking Wares and Other 270 10.3 Discussion 270 10.3.1 Provenance and Proportions 270 10.3.2 Spatial and Chronological Distribution 271 10.3.3 Interpretation of the Data 285 10.4 Conclusion 289 Catalogue of Illustrated Artefacts 289 Bibliography 290
263
246
Marica Cassis 11 The Cide-Şenpazar Region in the Byzantine Period 11.1 Methodological Problems 295 11.2 Historical Survey 298 11.3 Byzantine Finds and Sites in Cide and Şenpazar 11.3.1 Ceramics 303 11.3.1.1 Glazed Table Wares 305 11.3.1.2 Unglazed Table Wares 307 11.3.1.3 Kitchen Wares 309 11.3.1.4 Amphorae 309 11.3.1.5 Tiles 309 11.3.2 Byzantine Settlements 312 11.3.2.1 Okçular 312 11.3.2.2 Çamdibi 321 11.3.2.3 Abdulkadir 324 11.3.2.4 Other Settlements 325 11.3.3 Byzantine Fortifications 325 11.3.3.1 Gideros Batı Kalesi (S12) 331 11.3.3.2 Gideros Doğu Kalesi (S121) 331 11.3.3.3 Kazallı Kalesi (S91) 332 11.3.3.4 Timle Kale 1 (S110) 333 11.3.3.5 Timle Kale 2 (S112) 335 11.3.3.6 Aydos Kale (S190) 335 11.3.3.7 Hıdır Kalesi (S62) 338 11.3.3.8 Akça Kale (S135) 340 11.3.3.9 Koca Kale (S154) 340 11.3.4 Religious Spaces and Hermit Caves 343 11.3.5 Spolia and Other Remains 355 11.4 Conclusions 358 Catalogue of Illustrated Artefacts 359 Bibliography 361
294
303
Joanita Vroom 12
The Fortress of Çoban Kalesi, Late Medieval Ceramic and Small Finds 364 12.1 The Glazed Wares 366 12.2 The Unglazed Wares 368 12.3 The Small Finds 369 12.4 Some Concluding Remarks 370 Catalogue of Illustrated Artefacts 372 Bibliography 373
A.C.S. Peacock, with contributions by Claudia Glatz and T. Emre Şerifoğlu 13 Cide and its Region from Seljuk to Ottoman Times 375 13.1 Turks in Cide and the Kastamonu Region: 1071-1462 AD 376 13.2 Cide and the Kastamonu Region in the Ottoman Period 379 13.3 The Material Remains of the Ottoman Period in Cide 389 13.4 Conclusion 390 Catalogue of Illustrated Artefacts 390 Bibliography 390 Claudia Glatz and Bleda S. Düring, with contributions by Marica Cassis 14
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective 392 14.1 Okçular 394 14.1.1 The Okçular Landscape 394 14.1.2 Research at Okçular 395 14.1.3 The Anatomy of the Okçular Landscape 397 14.1.4 A Biography of the Okçular Landscape 400 14.1.4.1 Scatters in the Landscape: The Early Holocene 400 14.1.4.2 Rock with a View: The Chalcolithic and Early Bronze Age 401 14.1.4.3 Branching Out: The Second Millennium BC 403 14.1.4.4 The First Millennium BC 404 14.1.4.5 A Rural Landscape: The Roman and Byzantine Periods 406 14.1.4.6 Beylik and Ottoman Periods at Okçular 411 14.2 Abdulkadir-Gideros 412 14.2.1 The Abdulkadir-Gideros Landscape 412 14.2.2 Research in the Abdulkadir-Gideros Landscape 413 14.2.3 The Anatomy of the Abdulkadir-Gideros Landscape 414 14.2.4 A Biography of the Abdulkadir-Gideros Landscape 416 14.2.4.1 The Early Holocene and Chalcolithic 416 14.2.4.2 Elusive Millennia: The Bronze and Iron Ages 418 14.2.4.3 A Harbour and its Hinterland – Hellenistic to Byzantine Periods 418 14.2.4.4 Beylik and Ottoman Periods at Abdulkadir-Gideros 424 14.3 Çamdibi 425 14.3.1 The Çamdibi Landscape 425 14.3.2 Research at Çamdibi 426 14.3.3 Anatomy of the Çamdibi Landscape 427 14.3.4 Biography of the Çamdibi Landscape 427 14.3.4.1 Chance Finds and Buried Landscapes: The Early Holocene 427 14.3.4.2 Consuming Caves: The Later Chalcolithic and Bronze Ages 428 14.3.4.3 The Late Roman to Early Byzantine Period 431 14.3.4.4 Beylik and Ottoman to Modern Periods at Çamdibi 431
14.4 Concluding Remarks Bibliography 433
432
Claudia Glatz and Bleda S. Düring 15 Conclusions 434 15.1 Settlement, Subsistence and Production 436 15.2 12,000 Years of Solitude? - Connectivity and Particularity in the Cide Region 440 15.3 Looking Back and Moving Forward 443 Bibliography 443 Jakob Wallinga and Alice Versendaal (I); Toby C. Wilkinson (II); Richard Jones (III & IV) 16 Appendices 445 16.1 Appendix I. Luminescence Dating Report, Cide – Turkey, NCL-8111 445 16.1.1 Methods 445 16.1.2 Dating Results 446 16.1.3 Synthesis 447 Bibliography 447 16.2 Appendix II. Main Recording Forms of CAP 448 16.3 Appendix III. Chemical and Petrographic Study of Pottery from the Cide Archaeological Project 452 16.3.1 Discussion 457 16.4 Appendix IV. Chemical and Magnetic Analysis of Soils from Grid 6 (G6) at Okçular 460 16.4.1 Analyses 461 16.4.2 Results 461 16.4.3 Discussion 468 Bibliography 469 List of Figures List of Tables Index
478
470 477
Bleda S. Düring, Claudia Glatz and T. Emre Şerifoğlu
Acknowledgements
The Cide Archaeological Project was made possible through the support of a large number of people, institutions and funding bodies, and it is a pleasure to acknowledge our gratitude and debt here. First of all we are grateful to the Turkish Ministry of Culture and Tourism for granting us our research permit. We would like to thank Melik Ayaz, Emir Gökhan Bozkurtlar, and Aliye Usta for their support of our project, both at the start and throughout its development. We are equally indebted to our representatives E. Çağman Esirgemez (Çanakkale Museum), Ünver Göçen (Ereğli Museum), and Baykal Aydınbek (Eskişehir Heritage Protection Department), who were all willing to brave long hikes in challenging terrain and difficult weather conditions. We are grateful for their goodhumoured endurance of the Cide and Şenpazar mountains and their general support. We would also like to thank the staff at the Kastomonu Arkeoloji Müzesi for their support, and in particular its director, Nimet Bal. In 2011, she kindly allowed us to bring the collections of all three seasons to our project base for the final study-season, after which we temporarily stored the finds at Çanakkale Onsekiz Mart Üniversitesi. Both measures greatly facilitated the study of our finds by our team of specialists and the preparation of this book. We would also like to thank our numerous sponsors. We are grateful for the financial support of the Byvanck Fund (Leiden), the Faculty of Archaeology (Leiden), the Netherlands Organisation for Scientific Research (N.W.O., Veni programme), the British Academy, the G.A. Wainwright Fund (Oxford), the Carnegie Trust for the Universities of Scotland, the Robertson Bequest of the University of Glasgow, the UCL Graduate School, the Institute of Archaeology Awards (UCL), and Çanakkale Onsekiz Mart Üniversitesi Bilimsel Araştırma Bütçesi. We are grateful to the British Institute at Ankara and the Netherlands Institute in Turkey for assisting us in our dealings with the numerous bureaucratic issues that are connected to running an archaeological fieldwork project in Turkey. Likewise, we received institutional support (including equipment, consumables, storage facilities, and institutional backing) from the Faculty of Archaeology at Leiden University (Bleda S. Düring), the Institute of Archaeology at UCL, and Archaeology and the School of Humanities at the University of Glasgow (Claudia Glatz), as well as from Çanakkale Onsekiz Mart Üniversitesi (T. Emre Şerifoğlu). Our work would not have been possible without the help of these institutions. We are greatly indebted to our team of specialists and students, who worked extremely hard to help make this project a success (see our list of CAP team members below). We are obliged to Murat Karasalihoğlu for writing several articles about our project in the Kastamonu Posta newspaper and to Caner Bakan, our Hellenistic expert from Çanakkale Onsekiz Mart Üniversitesi, who also safeguarded our survey collection in the Çanakkale laboratories.
Acknowledgements
XIII
We received much help and information from the local residents of the villages and towns of Cide and Şenpazar, all of whom were very courteous and friendly to what must have seemed like a rather peculiar group of over-equipped strangers wandering about in their fields and backyards. In particular we would like to thank Recai Yılmaz, a teacher from Cide town, who knows the region’s landscapes and its people like the back of his hand and who brought us to many remarkable spots, some of which yielded important archaeological data. Further, we are grateful to Ali Osman, Fatma and Canan Tuğtepe for providing us with an excellent project base at the Yalı Otel. We would also like to thank the staff at De Gruyter Open for their patience and help in producing this volume. Last, but not least, we would like to thank Professor Roger Matthews for a fateful remark - ‘why not work further north?’ - in an Ankara taxi in February 2008. What an adventure it has been ever since.
List of Contributors Caner Bakan Department of Archaeology, Çanakkale Onsekiz Mart University. Chapter 9 Philip Bes Independent archaeologist, The Netherlands. Chapter 10 Marica Cassis Department of History, Memorial University of Newfoundland. Chapter 11, 14 Bleda S. Düring Faculty of Archaeology, Leiden University. Chapter 1, 2, 3, 4, 5, 14, 15 Bernard Gratuze CNRS, Institut de recherche sur les archéomatériaux - Centre Ernest Babelon, Orléans. Chapter 4 Claudia Glatz Archaeology, School of Humanities, University of Glasgow. Chapter 1, 2, 4, 6, 7, 13, 14, 15 Richard Jones Archaeology, School of Humanities, University of Glasgow. Chapter 4, Appendix 16.3 and 16.4 Murat Karasalihoğlu Kastamonu City History Museum. Chapter 3 Victor Klinkenberg Faculty of Archaeology, Leiden University. Chapter 4,5 Michele Massa Institute of Archaeology, University College London. Chapter 6 A.C.S. Peacock School of History, University of St Andrews. Chapter 13
List of Contributors
Effie Photos-Jones Scottish Analytical Services for Art and Archaeology and University of Glasgow. Chapter 4 T. Emre Şerifoğlu Department of Archaeology, Bitlis Eren University. Chapter 8, 9, 13 Alice Versendaal Department of Environmental Sciences, Wageningen University. Appendix 16.1 Joanita Vroom Faculty of Archaeology, Leiden University. Chapter 12 Jakob Wallinga Department of Soil Geography and Landscape, Wageningen University. Appendix 16.1 Toby C. Wilkinson TÜBITAK / Istanbul University. Chapter 4, Appendix 16.2
Muddy boots of CAP team members at the project base after a day in the field.
XV
List of Cide Archaeological Project (CAP) Team Members In alphabetic order and by year: 2009: Eva Boeijen Renate de Boer Bleda S. Düring E. Çağman Esirgemez Claudia Glatz
fieldwalker and database management fieldwalker director government representative director
Imke van Hagen Harmen Huigens Victor Klinkenberg Panagiotis Kontolomaios Michele Massa
fieldwalker fieldwalker fieldwalker and flint expert fieldwalker fieldwalker and GIS expert
Caterina Ruscio
fieldwalker
Toby C. Wilkinson
fieldwalker, GIS and database expert
2010: Eva Boeijen Geuch de Boer
fieldwalker and database management geoarchaeologist
Marica Cassis
Byzantine period specialist
Bleda S. Düring Claudia Glatz Ünver Göçen Imke van Hagen Sami Harju Murat Karasalihoğlu Victor Klinkenberg Michele Massa
director director government representative fieldwalker fieldwalker fieldwalker and local connoisseur fieldwalker and flint expert fieldwalker and GIS expert
T. Emre Şerifoğlu
director
Leiden University Leiden University Leiden University Çanakkale Museum University College London Leiden University Leiden University Leiden University Leiden University University College London Vatican Archaeological Service University of Sheffield
Leiden University RAAP Archaeological Consultancy Memorial University of Newfoundland Leiden University University of Glasgow Eğreli Museum Leiden University University of Glasgow Kastamonu Kent Tarihi Müzesi Leiden University University College London Çanakkale Onsekiz Mart Üniversitesi
List of Cide Archaeological Project (CAP) Team Members
Toby C. Wilkinson
XVII
fieldwalker, GIS and database expert fieldwalker
University of Sheffield
Freie Universität Berlin
Baykal Aydınbek
fieldwalker and database management government representative
Philip Bes
Roman period specialist
Geuch de Boer
geoarchaeologist
Tuğçe Çakıllık Aurora Camaño Sercan Değirmenci
fieldwalker fieldwalker and pottery processing fieldwalker
Bleda S. Düring Claudia Glatz Maarten Horn
director director fieldwalker
Gwendolynn de Groote Sami Harju Murat Karasalihoğlu Victor Klinkenberg Rita Kremer Michele Massa
fieldwalker fieldwalker fieldwalker and local connoisseur flint expert fieldwalker fieldwalker and GIS expert
Lorraine McEwan T. Emre Şerifoğlu
illustrator director
Evangelos Tourloukis Büşra Usta
Palaeolithic period specialist fieldwalker
Fiona Watson Toby C. Wilkinson
fieldwalker fieldwalker, GIS and database expert fieldwalker
Selim Yıldız 2011: Lilith Apostel
Selim Yıldız
Çanakkale Onsekiz Mart Üniversitesi
Eskişehir Heritage Protection Department Independent archaeologist, The Netherlands RAAP Archaeological Consultancy Denizli University Memorial University of Newfoundland Çanakkale Onsekiz Mart Üniversitesi Leiden University University of Glasgow IDDS Archaeological Consultancy Leiden University University of Glasgow Kastamonu Kent Tarihi Müzesi Leiden University Leiden University University College London University of Glasgow Çanakkale Onsekiz Mart Üniversitesi Leiden University Çanakkale Onsekiz Mart Üniversitesi University of Glasgow University of Sheffield Çanakkale Onsekiz Mart Üniversitesi
List of Cide Archaeological Project (CAP) Publications Düring, B. S., and Glatz, C. 2009. The Cide Archaeological Project 2009. Anatolian Archaeology 15: 15-6. —2010. 2009 Cide Arkeoloji Projesi: İlk sonuçlar. Araştırma Sonuçları Toplantısı. 28/1: 187-201. —2010. The Cide Archaeological Project 2009: First results. Anatolia Antiqua 18: 203-13. Düring, B. S., and Gratuze, B. 2013. Obsidian exchange networks in Prehistoric Anatolia: New data from the Black Sea region. Paleorient 39: 173-82. Düring, B. S., Şerifoğlu, T. E., and Glatz, C. 2011. 2010 Cide Arkeoloji Projesi: İkince sezon sonuçlar. Araştırma Sonuçları Toplantısı 29/3: 237-50. Düring, B. S., Glatz, C., and Şerifoğlu, T.E. 2011. The Cide Archaeological Project 2011. Heritage Turkey 1: 19-20. —2012. The Cide Archaeological Project 2011: Third preliminary report. Anatolia Antiqua 20: 167-75. Glatz, C., Düring, B. S., and Şerifoğlu, T. E.2010. The Cide Archaeological Project 2010. Anatolian Archaeology 16: 14-5. —2011. The Cide Archaeological Project 2010: Second preliminary report. Anatolia Antiqua 19: 279-88. Şerifoğlu, T. E., Düring, B. S., and Glatz, C. 2012. Cide ve Şenpazar 2009-2010 yılı arkeolojik araştırmaları. Belleten 277: 1-36. —2013. Cide Arkeoloji Projesi 2011: Üçüncü ve son sezon sonuçları. Araştırma Sonuçları Toplantısı 30/2: 157-68.
Newspaper and magazine articles Karasalıhoğlu, M. 2009. Cide’nin tarihi yeniden yazılacak. Kastamonu Posta: 20 August: 1 and 11. —2010. Cide Arkeoloji Projesi, Cide ve Şenpazar’da çok çarpıcı keşifler imza attı: Buluntular MÖ 20 bini gösterdi. Kastamonu Posta: 24 Temmuz: 1, 6-7. —2010. Cide Arkeoloji Projesi. Aktuel Arkeoloji Dergisi: 10-11. —2011. Antik Cide’nin kapısı aralandı. Kastamonu Posta: 28 July: 7.
Bleda S. Düring and Claudia Glatz
1 The Origins and Development of the Cide Archaeological Project The Cide Archaeological Project emerged out of a combination of different research agendas and several coincidences. We believe that the origins of the project are important for understanding the particular set of research questions we wanted to address and why specific methodological decisions were made in the course of our research. Here we will give a brief account of how this project came about. One strand of what was eventually to become the Cide Archaeological Project originates in Bleda’s post-doctoral research. During his doctoral research, which consisted of a spatial and social analysis of a number of Neolithic settlements in the southern part of central Anatolia, including sites such as Aşıklı Höyük and Çatalhöyük (Düring 2006), Bleda became increasingly aware of how much we know of the Prehistory of south-central Anatolia, and how little about the wider Neolithic of Anatolia. Northern Turkey as a whole has been very poorly investigated by archaeologists, and as a result we know next to nothing about the Neolithisation of this area, and what types of societies existed there prior to the fourth millennium BC. The problem of the ‘absent’ Neolithic of north-central Turkey was the topic of a post-doctoral research project Bleda undertook at UCL under the supervision of Professor Roger Matthews (Düring 2008). The main conclusion of this research was that intensive and targeted survey work was required to investigate the Prehistory of northern Turkey. The motive for Bleda to work specifically at UCL was Matthews’ Project Paphlagonia survey, which investigated the provinces of Çankırı and Karabük in north-central Turkey between 1996 and 2001 (Matthews and Glatz 2009a). The idea was to use the Project Paphlagonia dataset to investigate why assemblages predating the fifth millennium BC had proven so elusive. The region surveyed by Matthews was of special interest because it contains an obsidian source at Sakaeli, which is located in the Devrez valley near Orta. Obsidian from the Sakaeli source had been identified at Neolithic sites in the Marmara region, such as Ilıpınar, Pendik and Fikirtepe (Bigazzi et al. 1995, Bigazzi et al. 1998: 80-86). The artefacts from Ilıpınar come from levels X and IX, and can be dated to approximately 6000 BC, and those from Pendik in all likelihood predate Ilıpınar X and can be tentatively assigned to the second half of the seventh millennium BC (Özdoğan 1997: 21; 1999: 213). Thus, the Sakaeli obsidian sources were most probably exploited by 6500 BC, making it likely that a local population was present in the region and involved in the extraction and exchange of obsidian. The initial idea was to set up a smallscale survey in the area directly surrounding the obsidian outcrops at Sakaeli. A proposal sent to the Ministry of Culture and Tourism in Ankara at the end of 2006 was, however, unsuccessful.
2
The Origins and Development of the Cide Archaeological Project
The second strand of what was to become the Cide Archaeological Project originates in Claudia’s research on the archaeology of Hittite imperialism (Glatz 2007, 2009). Claudia, who was finishing her PhD at the time, and Bleda first met during his post-doctoral stint at UCL. Heavily involved in the analysis and publication of the Project Paphlagonia material (Matthews and Glatz 2009a), Claudia was interested in the same part of the Devrez valley that Bleda wanted to investigate. During the later second millennium BC, the Devrez valley formed part of the northern frontier of the Hittite empire and part of the stage upon which the Hittite – Kaska conflict was played out (Matthews and Glatz 2009b). Project Paphlagonia recorded a series of fortified sites along the Devrez Çay, which can be dated to the Late Bronze Age (Glatz et al. 2009). Claudia’s plan was to intensively survey the landscapes surrounding these fortified sites in order to grain a more detailed understanding of this early imperial frontier zone, its socio-economic base, and the dynamics of culture contact which would have unfolded in this region during the Late Bronze Age (Glatz and Matthews 2005). Over many coffees at the Bloomsbury Café, the idea was hatched to conduct a jointly directed intensive survey of a small section of the Devrez valley measuring about seven by three kilometres in order to investigate a set of well-defined and period-specific research questions for which we knew ample data would be available in this landscape. We applied for funding and a research permit in December 2007. At the same time, Bleda secured a Veni post-doctoral fellowship at Leiden University for a project entitled Exploring the Early Holocene Occupation of North-Central Anatolia: New Approaches for Studying Archaeological Dark Ages in which the ‘Devrez Valley Archaeological Project’ featured as a central component. The implementation of our research strategy, however, proved less straightforward than we had hoped. Despite successes in securing funding from both Dutch and British funding bodies (N.W.O., Byvanck Fund, G.A. Wainwright Fund, and UCL), we were not granted the survey permit crucial to our research. The reason for this is a systemic incompatibility between the demands of research funding bodies in the Netherlands and the United Kingdom on the one hand, and those of the Turkish Ministry of Culture and Tourism on the other. All archaeologists wishing to excavate or survey in Turkey have to apply to the Ministry of Culture and Tourism for a research permit. Provided one meets a set of standard requirements such as a PhD, a position at a major university, and adequate financial and institutional support, among other things, one can normally expect to receive a research permit with some confidence. This system also tends to protect existing permit holders from competing claims. In the case of survey permits, however, this policy, which in itself is laudable, has some undesirable consequences. Current regulations mean that most of Turkey has been carved up into a small number of exceedingly large survey areas, the majority of which have been investigated by (very) extensive strategies focused on large, conspicuous sites (Düring 2008). This territorial system does not permit
The Origins and Development of the Cide Archaeological Project
3
the more intensive investigation of selected parts of these larger survey units by other projects. In our particular case, this meant that one of our senior colleagues from Ankara University already held a permit to survey Çankırı province as well as neighbouring Çorum, which together amount to a survey area of over 20,000km² (an area larger than countries such as Slovenia or Israel). We made several attempts to convince our esteemed colleague that the work we planned in the Devrez valley was of a very different and complementary nature to his own research, namely an intensive pedestrian survey in a tiny portion (ca. 30km²) of Çankırı province. Claudia, with the support of Roger Matthews, even visited Ankara University in February 2008 to speak to our colleague. Unfortunately, however, all our efforts were of no avail, in part because we had very different understandings of archaeological survey work. In an attempt to rescue our idea of a survey in northern Turkey, we enquired with the Ministry of Culture and Tourism what other nearby regions were available for archaeological survey. It was at this point that the district of Cide in Kastamonu province was first mentioned. Cide was both archaeologically unspoken for and, from the perspective of the Ministry at least, was located near our original area of interest in Çankırı. After some hectic deliberations and a flurry of emails and phone calls, we decided to jump at the opportunity and change our research proposal from Çankırı to Cide. The Cide Archaeological Project (CAP) was born. In some ways, the same logic that was at the basis of our original research proposal also applies to the Cide region: to explore and document the archaeology of an area that had seen only limited research to date. In other respects, however, Cide differs dramatically from central Çankırı. Central Çankırı forms part of the northern extension of the central Anatolian plateau and is characterised by relatively flat, open and today intensively farmed landscapes – in other words, an ideal arena for intensive pedestrian survey. The geographical situation of Cide is completely different. Part of the Turkish western Black Sea region, Cide is dominated by a rugged coastal landscape that is separated from central Anatolia by steep mountains, which are difficult to traverse even with modern transport technologies (chapter 2) and equally challenging to survey due to dense vegetation cover and steep terrain (chapter 4). Moreover, and unlike in the case of central Çankırı, we would have no archaeological leads to guide our research agenda and survey strategy. Instead, we would have to start with a more or less blank map and a handful of mentions of Roman and Byzantine castles and chance finds from the region (Belke 1996; Marek 2003). Three broad research questions, one methodological and two socio-cultural in nature, have since guided our work in the Cide region. Faced with this extraordinarily difficult landscape, where ground visibility is limited and erosion processes have severely affected the preservation and accessibility of archaeological sites (section 2.3), the first key question that imposed itself was how best to investigate this challenging landscape. The development of a suitable field methodology that would allow us to
4
The Origins and Development of the Cide Archaeological Project
identify and document ephemeral archaeological assemblages was, therefore, central to our efforts (chapter 4). This would then allow us to tackle our social and cultural research questions. The second central research question of CAP, thus, concerns the long-term cultural development of the Cide region. More specific questions, which will be dealt with in chapters 5 to 13, include, for example, when and how farming was introduced to the region; when and in what social and economic context seafaring developed and proliferated; and when the first complex societies emerged in this marginal region. Our third question closely ties in with the second and focuses on how the history and cultural development of our research region relate to broader developments in mainland Anatolia and the Circumpontic. The topography of the region, which is dominated by the steep east-west oriented Pontic mountains, would probably have impeded communication with the Anatolian interior (section 2.2.1). By contrast, the sea could have facilitated transport along the coast as soon as seafaring was taken up. We will address more specific questions, such as whether the Early Bronze Age settlement expansion witnessed across Anatolia also affected this area; to what degree the region was involved in and affected by international trade; and whether the rise of powerful states in the second and first millennia BC had any impact on our region, in subsequent chapters. Once the decision to change our survey area from central Çankırı to Cide had been taken, the next challenge was to convince our funding bodies to do the same. Fortunately, all of them accepted our proposed changes, since our research questions and proposed methodologies remained largely the same, even if the conditions in which we wanted to answer them had become more challenging. A permit application to the Turkish Ministry of Culture was finally successful in Spring 2009. In August 2009, we started work in Cide supported by an international team of 11 experts and students and under the watchful gaze of our representative, E. Cağman Esirgemez of Çanakkale Museum. It was a relatively short season, the main purpose of which was to find our feet in this new landscape and develop a strategy for investigating its archaeology in the face of poor ground visibility, rugged mountainous terrain and difficult roads (chapter 2). Visibility in particular proved our biggest nemesis throughout the duration of the project. Much of the area is covered by dense forest and shrub that are interspersed with parcels of small open terrain, which are, however, often covered by lush meadows in turn. A decline in crop farming in recent decades meant that ploughed fields are rare and those carrying crops could be examined only partially. Our main challenge in this first season was, thus, to develop a methodology that would allow us to find any archaeology in this taxing landscape. It also soon became evident that it would be impossible to focus our research exclusively on the ‘Pre- and Protohistoric’ periods, which formed the explicit remit of our 2009 survey permit. For example, during fieldwalking we would encounter predominantly Roman and Byzantine ceramics, which were occasionally interspersed with more ephemeral
The Origins and Development of the Cide Archaeological Project
5
traces of an earlier presence. Likewise, while investigating caves we tended to find both Byzantine and earlier assemblages, often mixed to a considerable degree by illegal excavations. In both cases, it would have been impossible to investigate only the Pre- and Protohistoric components of these assemblages in a way that would have been archaeologically satisfactory. During the 2009 campaign we also discussed the possibility of collaborating with a Turkish colleague. Shortly after arriving in Cide, we soon realised that we could not always relate easily to residents and local authorities due to language and cultural barriers and that in addition to bringing complementary archaeological expertise to the project, a Turkish co-director would significantly enhance and facilitate our local relationships. Finally, we felt that the Cide district was somewhat limiting in terms of the ecological zones it comprised. Cide consists primarily of a coastal zone, adjacent foothills, and river valleys stretching inland. We wanted to expand our research area into the mountainous upland zones directly to the south of our survey area in order to gain a more detailed understanding of the variability in occupation histories of different landscape types and the region’s inland connections. In order to address these concerns we included the following changes in our 2010 permit application. We requested a multi-period permit that would allow us to investigate all periods present in the area. To this end we invited a Byzantine specialist, Dr. Marica Cassis (Memorial University of Newfoundland), to join our project. We also asked Dr. T. Emre Şerifoğlu from Çanakkale Onsekiz Mart Üniversitesi to join us as a third co-director, whose research focus would be on the Iron Age and Hellenistic periods. Finally, we requested an expansion of our survey permit to include not only Cide but also the adjacent Şenpazar district. Despite some initial uncertainty caused by an alternative claim to the Roman and Byzantine archaeology of the region, we were granted both the geographical and chronological expansion of our survey permit. Following the lessons learnt during the 2009 campaign, the structure of our project and its objectives had evolved considerably when we found ourselves in the field again in July 2010. At the start of the second season, we were in a very different position from the preceding one. We had acquired a sense of the landscape, what archaeology we could find in it and how (chapter 4). Our Turkish co-director and team members greatly enhanced our relations with local residents, facilitated the acquisition of local knowledge and made the dealings with a range of local authorities smoother than in the previous year. We had a specialist for the Byzantine period on the team and a geoarchaeologist, Geuch de Boer (RAAP consultancy) to help us understand the geological changes in the Cide and Şenpazar landscapes and how they would have affected their archaeology. A longer six-week campaign, moreover, allowed us to cover much more ground and to expand our sample of different ecological zones and landscape types. Overall, the 2010 season was a very productive one during which we were able to achieve many of our goals.
6
The Origins and Development of the Cide Archaeological Project
In 2011 little changed with regard to our research goals and overall project structure. We continued our combined intensive-targeted approach developed in 2009-10, which is to contextualise sites recorded during targeted survey in a wider landscape perspective as well as to cover a range of different geographical/ topographical zones from the coastal region to the inland valleys and higher altitude zones of southern Cide and Şenpazar. The main purpose of the 2011 campaign was to complete and solidify our existing dataset of different landscape zones and specific site types. We introduced a number of new sampling and analytical procedures, which included interviews with local residents and the collection and export of soil, pottery and obsidian samples for laboratory-based analyses and organic material for radiocarbon dating (chapter 4). In 2011, we worked with three field teams in order to cover as much terrain as possible. Two weeks of intensive survey were followed by a brief phase of targeted investigations with a smaller field team. The Cide Archaeological Project’s fieldphase was concluded by a two-week study season. Our team of experts was joined by Dr. Philip Bes, a Roman period specialist, and was able to study all three seasons worth of collected material. We also greatly benefitted from discussing our assemblages with visiting colleagues, including Dr. Alexander Bauer, Dr. Gregory McMahon, Prof. Jennifer Ross, Dr. Ulf-Dietrich Schoop, Dr. Susan Sherratt, Prof. Sharon Steadman and Dr. Ali Türkcan. Although conditions in the field in 2011 were more difficult than in previous campaigns due to a rainy spring and summer, resulting in very lush vegetation, our results for the final season were nonetheless very satisfactory. We were able to expand the range of intensively sampled ecological zones and to explore extensively areas which we had not previously worked in. We were able to increase sample sizes from key sites and to learn much about the more recent past of the region through our interviews with local residents. Overall, we were able to acquire a much more systematic understanding of the Cide-Şenpazar landscapes. Following the final field-season we were able to further expand our team of specialists: Mr. Caner Bakan for the Hellenistic period, Dr. Andrew Peacock for the Ottoman sources and Dr. Joanita Vroom for the post-Byzantine ceramics. With this augmented team we subsequently worked towards the publication in hand. We can summarise the Cide Archaeological Project as follows. In the course of three seasons, we clocked up a total of 85 field days and spent a total of € 62,514 (table 1.1). We worked with two field teams of five walkers in 2009 and 2010, and three field teams in 2011. Our overall team ranged between 12 and 22 members in size. In total this amounts to about 3½ wo/man-years spent surveying the Cide and Şenpazar landscapes. In the following chapters, we present the results of these three seasons worth of fieldwork, analysis and interpretation that form the ultimate justification for the considerable amount of resources and labour that we devoted to investigating the archaeology of Cide and Şenpazar.
Bibliography
7
Tab. 1.1: Number of field days, size of the team and budget for each of the seasons of the Cide Archaeological Project. season
field-days
team size
budget (€)
2009
20
12
13,308
2010
32
14
19,559
2011
33
22
29,647
Bibliography Belke, K. 1996. Paphlagonien und Honorias. Vienna, Verlag der Österreichische Akademie der Wissenschaften. Bigazzi, G., Oddone, M., and Yeğingil, Z. 1995. A provenance study of obsidian artifacts from Ilıpınar. In J.J. Roodenberg, ed., The Ilıpınar Excavations I. Istanbul, Nederlands Historisch Archaeologisch Istituut te Istanbul: 143-50. Bigazzi, G., Poupeau, G., Yeğingil, Z., and Bellot-Gurlet L. 1998. Provenance studies of obsidian artefacts in Anatolia using the fission-track dating method: An overview. In M.-C. Cauvin, A. Gourgaud, B. Gratuze, N. Arnaud, G. Poupeau, J.-L. Poidevin and C. Chataigner, eds., L’obsidienne au Proche et Moyen Orient, du volcan à l’outil. Oxford, Archaeopress: 69-89. Düring, B.S. 2006. Constructing Communities: Clustered Neighbourhood Settlements of the Central Anatolian Neolithic, ca. 8500-5500 Cal. BC. Leiden, Nederlands Instituut voor het Nabije Oosten. — 2008. The Early Holocene occupation of North-Central Anatolia between 10,000 and 6,000 BC cal: Investigating an archaeological terra incognita. Anatolian Studies 58: 15-46. Glatz, C. 2007. Contact, Interaction, Control. The Archaeology of Inter-Regional Relations in Late Bronze Age Anatolia. Unpublished PhD thesis, University College London. — 2009. Empire as network: Spheres of material interaction in Late Bronze Age Anatolia. Journal of Anthropological Archaeology 28: 127-41. Glatz, C. and Matthews, R. 2005. Anthropology of a frontier zone: Hittite-Kaska relations in Late Bronze Age north-central Anatolia. Bulletin of the American Schools of Oriental Research 339: 47-65. Glatz, C., Matthews, R., and Schachner, A. 2009. A landscape of conflict and control: Paphlagonia during the second millennium BC. In R. Matthews and C. Glatz, eds., At Empires’ Edge: Project Paphlagonia. Regional Survey in North-Central Turkey. London, British Institute at Ankara: 107-147. Marek, C. 2003. Pontus et Bithynia. Die römischen Provinzen im Norden Kleinasiens. Mainz am Rhein, Philipp von Zabern. Matthews, R., and Glatz, C. eds. 2009a. At Empires’ Edge: Project Paphlagonia. Regional Survey in North-Central Turkey. London, British Institute at Ankara. Matthews, R., and Glatz, C. 2009b. The historical geography of north-central Anatolia in the Hittite period: Texts and archaeology in concert. Anatolian Studies 59: 51-72.
8
The Origins and Development of the Cide Archaeological Project
Özdoğan, M. 1997. The beginning of Neolithic economies in southeastern Europe: An Anatolian perspective. Journal of European Archaeology 5: 1-3. —1999a. Northwestern Turkey: Neolithic cultures in between the Balkans and Anatolia. In M. Özdoğan and N. Başgelen, eds., The Neolithic of Turkey. Istanbul, Arkeoloji ve Sanat yayınları: 203-24.
Bleda S. Düring and Claudia Glatz
2 Remote Landscapes in Flux – The Cide and Şenpazar Region An understanding of the Cide-Şenpazar region forms the basis of any meaningful evaluation of the sampling strategies and survey methods which we developed in the course of the Cide Archaeological Project and the archaeological materials and patterns that we recorded as a result. This chapter presents an overview of key factors shaping this region in the long and shorter-term, including ecology, climate and geology, as well as the logistical parameters of transport and communication, local economy and culture.
2.1 The Cide-Şenpazar Landscapes 2.1.1 Ecology and Climate The Cide and Şenpazar districts are located in the northwest of Kastamonu province and comprise an area of approximately 930km². The districts are part of the Turkish western Black Sea region. The dominant geographic features of this region, which stretches from Ereğli to Sinop, are the Pontic mountains that rise steeply from the Black Sea, with peaks between 800 to 1200 metres above sea level only a few kilometres from the coast and that leave little space for coastal plains. The pronounced relief of this landscape has various consequences for the ecology of the region. First, along the windward slopes and valleys of the Pontic mountains and their interface with the Black Sea, precipitation is relatively high compared to other parts of Turkey due to orographic lift, in which the moist air from the north is forced upwards. Precipitation is about 1000mm per year (Fig. 2.1), part of which falls in summer (Alex 1985), and the climate is temperate with relatively cool summers and mild winters, in large part due to buffering effects of the Black Sea (Zohary 1973: 31). Second, the pronounced relief of the Turkish western Black Sea region and the effects of orography lead to a marked zoning of vegetation. General descriptions such as ‘Euro-Siberian’ (Zohary 1973: 31, 118-23) and ‘cold-deciduous forests’ (van Zeist and Bottema 1991: 27-8) encompass diverse ecological zones over close distances. There is a grading of dense maquis-type shrub forests on the lower slopes, dominated by trees such as Fagus orientalis, Carpinus orientalis, Corylus colurna, Carpinus betulus, Ostrya carpinifolia and various oak species, which give way to open forests on the higher mountains that are dominated today by pine trees, such as Abies nordmanniana, Pinus nigra and Pinus sylvestris (Zohary 1973: 112-6, 570-8; Van Zeist and Bottema 1991: 27). Without human interference the Turkish western Black Sea region would be completely forested. In the field, we observed many abandoned agricultural areas and other types of open terrain being reclaimed by surrounding forest.
10
Remote Landscapes in Flux – The Cide and Şenpazar Region
Fig. 2.1: Average monthly temperatures and precipitation for Bartın (after Alex 1985: 55).
On a general level, the landscapes of the Turkish western Black Sea region can be divided into three primary components: first, the higher mountains along the interior, which are covered by open forests and generally show few signs of human interference; second, the hills and valleys running up to the coast where many villages and hamlets can be found; and, third, the few flat areas along river courses and the littoral, the largest of which consists of the coastal plain of Cide. Further inland, the climate is dryer and landscapes are more akin to those of central Anatolia. These pronounced differences in vegetation are determined by equally diverse climate regimes, which in turn have an impact on the range of crops that can be cultivated in different landscapes. Detailed reconstructions of changes in the Holocene climate of the Turkish western Black Sea region are currently not available, but we do have various types of proxy data to work with. Across Anatolia, a large number of lakes have been sampled to reconstruct changes in climate and vegetation in the course of the Holocene (see Düring 2011: 11-17 for an overview). Of great importance is the synthetic work by Van Zeist and Bottema (1991), who reconstruct past ecologies from pollen data derived from lake cores. According to their reconstructions, the Turkish Black Sea region is the prime refugium for trees in Anatolia during the last Ice Age and remains forested up to the present. Directly south of the Turkish western Black Sea region, at Yeniçağa and Abant Gölü, forestation started in the Younger Dryas, immediately prior to the start of the Holocene (Van Zeist and Bottema 1991: 94). Subsequent investigations of lakes further east, at Kaz Gölü and Ladik Gölü have largely confirmed the idea of a forested Pontic region and a more or less open steppe ecology to the south, which was
The Cide-Şenpazar Landscapes
11
reoccupied by forests in the period between the Younger Dryas and the Mid-Holocene (Bottema et al. 1993: 50-57). An important new climatic proxy record has been obtained from the cave of Sofular Mağarası near Zonguldak (Fleitmann et al. 2009). By sampling stalactites, a sequence was obtained that dates back up to 650,000 years ago. Samples were dated through a series of Uranium-Thorium dates, and isotopic values were taken of O18 (δ18O ‰ VPDB) and C13 (δ13C ‰ VPDB). These data have been used mainly to investigate the precise timing of global climatic oscillations, the so-called Heinrich events in the Pleistocene. However, such isotope values can also be used for detailed climatological reconstructions: the O18 isotopic values are influenced by the amount and seasonality of precipitation and changes in temperature in the Black Sea surface waters, whereas C13 isotopic values are determined by the type of local vegetation (Fleitmann et al. 2009: 2-3; Zumbühl 2010: 50-4). At Soreq cave in Israel, such records have been used to reconstruct past climates (Bar-Matthews et al. 1999). The Sofular cave data has similar potential (Fig. 2.2), but we will have to await more detailed assessments of this dataset by our colleagues in geology.
Fig. 2.2: δ18O ‰ VPDB and δ13C ‰ VPDB isotopic measurements from Sofular cave from 14,000 BP (right) up to 2006 (source data: www.ncdc.noaa.gov).
Notwithstanding the absence of detailed climate and vegetation reconstructions for the Cide-Şenpazar area, it is clear that from the start of the Holocene the Turkish western Black Sea region had a temperate climate, with substantial precipitation
12
Remote Landscapes in Flux – The Cide and Şenpazar Region
distributed throughout the year. It is also clear that the area was forested throughout the Holocene. The O18 and C13 isotopic data do show considerable fluctuations, however, which suggests alternating periods of wetter and dryer climates (δ18O ‰ is lower when effective precipitation is high) as well as changes in vegetation (δ13C ‰ values). A recent study of the Turkish eastern Black Sea region on the basis of a marine core taken north of Samsun suggests that in the Early Holocene, between about 10,000 and 6000 BC, the region was characterised by a relatively open landscape similar in vegetation to modern day central Anatolia. It also indicates that it was only around 6000 BC that the current vegetation of dense mixed temperate deciduous forests developed, with a subsequent slight and most likely anthropogenic decrease in forest cover around 3000 BC (Shumilovskikh et al. 2012: 188-9). To what degree this development also occurs in the Turkish western Black Sea region is unclear, but we do know that in many parts of northwestern Turkey forests were more prominent than the new data from Samsun would suggest (Düring 2008: 29).
2.1.2 Geology The Pontic mountains, which dominate the Turkish western Black Sea region, are the result of the collision of the Eurasian and Anatolian plates in the Late Palaeocene -Early Eocene. The strata that were uplifted in the Pontic mountains have ‘European’ features exemplified by the presence of coal (Dean et al. 2000; Okay 2008). To the south of the Pontic mountains runs the North Anatolian Fault Zone, along which the Anatolian Plate moves westwards causing regular earthquakes (Okay 2008; Marsh et al. 2009). Earthquakes with a magnitude of 6 or 7 on the Richter scale occur almost every decade in the region. The geological structure of the Turkish western Black Sea region is heterogeneous (Akyol et al. 1974; Erol 1983). Moving from Kastamonu to Cide one encounters five main geological formations (Uğuz et al. 2002): first, Lower-Middle Miocene lacustrine limestone (marl, shale); second, a belt with metamorphic rocks (schist, marble, metabaste, serpentinite) dating to the Triassic-Jurasic; third, Lower Cretaceous clastic and carbonate rocks; and, fourth, Upper Senonian clastic and carbonate as well as vulcanic and sedimentary rocks. Interspersed between these geological belts, which run parallel to the Pontic mountains, are numerous smaller pockets of geological deposits. The geological composition of the Cide and Şenpazar regions has important implications for soil quality and agricultural potential, colluvial and erosion susceptibility of landscapes, and the distribution of natural resources (see below).
The Cide-Şenpazar Landscapes
13
2.1.3 Coastal Changes of the Black Sea The reconstruction of the coastal changes along the Sea of Marmara and the Black Sea in the final Pleistocene and the Holocene has been the subject of much debate. Some years ago, Ryan and Pitman claimed a dramatic flooding of the Black Sea basin in the Early Holocene, which they linked with the Biblical narrative of Noah’s flood (Ryan et al. 1997; Ryan and Pitman 1999). According to them, Black Sea levels during the Early Holocene remained about 100 metres below current sea level. A sudden highvolume influx of water from the Mediterranean through the Bosphorus, they claimed, resulted in the filling of the Black Sea basin to current surface levels at about 6000 BC. They also hypothesised that the flooding of the Black Sea displaced a large number of farming communities, which had previously lived in the basin and were forced by this fast encroaching flood to migrate to adjacent regions, including to Europe. This idea has found some resonance in the archaeological community. For example, one team of researchers found what they interpreted as a submerged Neolithic settlement off the coast between Sinop and Cide. This ‘site 82’ is presently located nearly a hundred metres below the modern sea level (Ballard et al. 2001; Hiebert et al. 2002; Doonan 2004). A plan of site 82 (Ballard et al. 2001: 617) suggests the presence of square structures of ca. five by five metres in a more or less linear arrangement. Others have modelled the migrations of people displaced from the Black Sea basin into Europe and how farming spread in the process (Turney and Brown 2007). The flood hypothesis has since been put to the test by both geologists and archaeologists. The interpretation of site 82 as a submerged Neolithic site could not be substantiated with lithic and ceramic artefacts, and wood samples collected from the site all dated to the 19th and 20th centuries AD (Ballard et al. 2001: 615). Currently, it seems most likely that site 82 consists of the remains of a sunken wooden vessel no more than a hundred years old. Other archaeologists have scrutinised the archaeological data of the Circumpontic and have not found evidence, in the form of cultural similarities along its shores, for an exodus of people from the Black Sea basin. The same scholars have also pointed out that the chronology of the supposed Black Sea flood and exodus is incompatible with archaeological datasets (Bailey 2007; Dergachev and Dolukhanov 2007; Özdoğan 2007). Finally, the hypothesis of the flooding of the Black Sea around 6000 BC has been challenged by geologists on the following grounds (Aksu et al. 2002a; 2002b; Jablonka 2003; Yanko-Hombach 2007a; 2007b). First, the assumption that Early Holocene sea levels were one hundred metres below present levels is problematic given that both precipitation and the melting of glaciers would have created a large inflow in the Black Sea during this period. Second, Ryan and Pitman have dated their flooding event by dating the youngest fresh water molluscs they were able to find on the Black Sea floor, which could not survive under the anoxic conditions that currently characterise the lower Black Sea. However, the extinction of fresh water molluscs in the Black Sea does not necessarily date a flooding event, but rather the onset of anoxic conditions
14
Remote Landscapes in Flux – The Cide and Şenpazar Region
that resulted from the exchange of salt water from the Mediterranean/Sea of Marmara on the one hand and fresh water from the Black Sea on the other. The following reconstructions for the Sea of Marmara and the Black Sea are most plausible at present. During the Last Glacial Maximum, the level of the ‘Marmara Lake’ was about 90 metres below that of the present. Around 11,000 BC water started to flow through the Dardanelles from the Aegean into the Sea of Marmara, which quickly rose to the same level. In the Black Sea, the level during the Late Glacial Maximum was about 110 metres below current sea levels. During the Younger Dryas it rose to 43 metres below current levels. At around 7500 BC water started to flow from the Black Sea into the Sea of Marmara, either through the Bosphorus or via the IzmitSapanca depression. At this point water levels were much higher in the Black Sea, and water flowed westwards only. It was only at around 6000 BC that both seas reached a balance, and at this point salt water from the Sea of Marmara started flowing into the Black Sea as an undercurrent. It was this undercurrent of salt water that created the anoxic condition that put an end to fresh water mollusc communities in the lower Black Sea rather than a flood. At the latest at about 4000 BC, the Bosphorus had become the main connection between the Black Sea and the Sea of Marmara (Özdoğan 1997; Aksu et al. 1999; Özdoğan 1999, 2003; Jablonka 2003). Dispelling the Black Sea flood hypothesis does not, however solve the question of past sea levels and their fluctuations along the coast of the Turkish western Black Sea region. Considerable efforts have been put into dating submerged shelves (old coastal terraces) in various parts of the Black Sea. However, the data obtained do not match across the sea and it has been impossible to construct Black Sea-wide changes in sea level over time (contra Shilik 1997). The most plausible explanation for these differences are neotectonic processes of uplift and subsidence of sectors of the Black Sea seafloor, and reconstructions of sea level oscillations can consequently only be undertaken for specific regions (Govedarica 2003; Yanko-Hombach 2007a: 176; 2007b: 8; Brückner et al. 2010). For the Turkish western Black Sea region the one study available, of the Sakarya delta, suggests that at the start of the Holocene the sea would have been about 40 metres below current levels, then would have stabilised at minus 20 metres for some time around 6000 BC (Fig. 2.3), and finally would have risen to more or less current sea levels at around 3000 BC (Aksu et al. 2002b: 88-91). At a general level, such reconstructions are useful. For much of the coast of the Turkish western Black Sea region sea level changes of a few metres would have had little effect on coastal geography, since mountains rising steeply from the sea dominate most of the coastline. It is possible, however, that caves along the coast have been flooded since the Early Holocene, resulting in the loss of potential sites dating to this period. There are also parts of the coast, for which sea level changes would have had more dramatic effects. For example, the bay of Gideros is the only large natural harbour in the Cide–Şenpazar region, and has been of importance from at least the period of Greek colonisation in the seventh century BC, and probably much earlier (chapters 3, 8 and 9). However, Gideros harbour today is relatively shallow and it would not have existed if
Humans in the Cide-Şenpazar Landscapes
15
Fig. 2.3: Hypothetical reconstruction of the coastline of the Cide-Şenpazar region at -40 metres (ca. 10,000 BC) and -20 metres (ca. 6000 BC) in relation to the modern coastline (from ca. 3000 BC). Based on bathymetric data. Produced by Victor Klinkenberg.
the sea had been even 10 metres lower than at present. Proxy data elsewhere along the Turkish Black Sea coast suggest that during the Iron Age, sea levels were three to four metres lower than at present (Tsetskhladze 2007: 176-80). If this was the case, Gideros was a harbour that could only be used by ships with shallow keels, something which is true even today and at higher sea levels (personal communication Recai Yilmaz 4-72011). Likewise, the Cide coastal plain, which is the only large flat area in all of Cide and Şenpazar and includes a beach where ships could be hauled, repaired and built, would have been affected by relatively minor sea level changes. It was, therefore, important for our research to study local sea level oscillations and coastal formation processes in some detail and we will report on our findings later in this chapter.
2.2 Humans in the Cide-Şenpazar Landscapes 2.2.1 Transport and Communication in a Rugged Landscape The difficulties that transportation poses in the challenging terrain of the Turkish western Black Sea region can be graphically illustrated by the following personal experience of one particular travel episode in the Cide region.
16
Remote Landscapes in Flux – The Cide and Şenpazar Region
“The old road to Cide?” The question was asked by a middle-aged man in the town of Pınarbaşı, in the district adjacent to that of Cide. There was a little pause in the question to convey his feeling that perhaps he had misunderstood something and this was mixed with a hardly perceptible mirth, the source of which was unclear to me at the time. “Yes” I said stupidly, ignoring the subtle clues in his response – earlier that day we had decided we would like to try the stretch of road indicated on our map leading from the town of Pınarbaşı to that of Şenpazar, in order to explore an area we had not yet visited. The man looked at our car for a moment, a Fiat Doblo that is somewhat higher on its wheels than most cars. “Yes” he said “it is possible”, and he then pointed out the road we should take. By the time I was thanking him, I was starting to get a little bit worried whether we were doing the right thing by taking this road, and I suppose had it not been for a shared feeling in the car that we wanted to explore new territory and discover new places, I would have turned around there and then. However, the first stretch of road wasn’t all that bad and we would have many hours before dusk would set in, and the distance to be covered was not all that great, amounting to perhaps 15 kilometres. Thus, we started on the road with optimism. Our moods were lifted even further by the discovery of a large complex of rock-cut graves near the start of Valla kanyon: a gorge about 20 kilometres long and 800 metres deep considered to be the second largest canyon in the world (Tunoğlu 2008: 26). After we had forded the river at the upstream end of the canyon, however, the road steadily deteriorated. The dirt road was narrow with steep slopes on the side and frequently had large holes in inconvenient spots; it wound up the mountain at a steep angle and in a series of sharp turns. Much to our surprise there were dirt roads of similarly poor quality as the one we were trying to follow leading off in all directions. These remote mountains are home to a remarkable number of houses and many of these were apparently in use at least during parts of the year. Eventually the inevitable happened: we took a wrong turn and were lost. This was not surprising given that road signs were virtually non-existent and all roads were more or less similar, but it was a problem because I was getting tired from the difficult drive which had already lasted some three hours, and it was starting to get dark. It was clear that we could not spend the night in these mountains; there was no place where we could stay. We started to drive back the way we had come and came upon two men, who we asked for directions to Şenpazar and Cide. They briefly looked at us and our car, and they said: “Yes, it is possible”, without any mirth this time but with a touch of anxiety, and explained what to do. The good news was that the remaining stretch to Şenpazar was relatively short in terms of distance, but the bad news was that it meant descending from an impossibly steep and high mountain. In places along this trajectory the remnants of what had been once a nice stone paving could be made out, but in most places this paving had long been eroded away and had given way to gullies, holes, collapsed roadsides, and collapse from the slope above. This was clearly no longer a road in the normal sense of the word, although it had been in the past.
Humans in the Cide-Şenpazar Landscapes
17
As we began our descent, the impossible nature of our situation dawned on me. There was no mobile phone reception, the car - which was certainly not build for this type of driving - could break down at any moment, which could results either in us all falling down the slope with the car, or having to abandon the vehicle and make our way on foot. To make matters worse there was a baby in the back of the car, in need of feeding, warmth and shelter. She was miraculously sleeping throughout the shaky drive and tense atmosphere, but this too would not last indefinitely. Somehow, we made it down that mountain without as much as a scratch on the car, but the memory of this trip has remained unsettling to me ever after. (source: diary Bleda Düring 25-8-2009) This excursion along the ‘Old Cide Road’ taught us something fundamental about this landscape, which is to take seriously the degree to which even the lesser Pontic mountains are, and have been in the past, an impediment to communication between the coastal region and central Anatolia. One of the initial ideas of our survey project was to focus on natural routes from Cide across the mountains to the interior, and try and establish, for example, whether these routes were controlled by fortified settlements. Our experience in the mountains just described made it clear that there are no easy natural routes to the interior from Cide and Şenpazar, and that all communications across these mountains would have been difficult and time consuming. Thus the Cide region in the past must have been relatively isolated from nearby inland regions, at least with regard to vehicle-based transport. Today, the road trip from the town of Kastamonu to Cide, a distance of about 130km, takes about two hours, but this has been made possible only by the recent construction of a good road. In the relatively recent past, that is before 2005, the trip to Kastamonu – the provincial capital where people have to go for all sorts of official and judicial matters – took a solid eight hours of driving. Before the advent of cars, when this distance was covered with horses and wagons, it could take up to three to four days to reach Kastamonu from Cide (personal communication Recai Yilmaz 4-72011). A fitting local saying puts it this way: “Askerlik zor değil, savaşmak da zor değil. Esas zor olan kış aylarında Kastamonu yolunu kullanarak Ankara’ya ulaşmak.’’ (Army service and war are not difficult. What is difficult is to travel the Kastamonu roads to Ankara in winter.) It is plausible that the preferred route in the past was along the coast eastwards to the port of Sinop, where a major road led to Kastamonu and further south from at least the Roman times (Marek 2003: 183). In this regard, one may wonder to what degree the modern provincial boundaries, in which Cide has become part of Kastamonu province, reflect pre-modern social realities. For example, in Roman times the Turkish western Black Sea coast was unified in the province of Pontus and distinguished from interior provinces such as Paphlagonia (Madsen 2009: 12). In modern Cide and Şenpazar, a large number of people told us that they did not feel connected with Kastamonu. Instead we were often told that we should regard the region as culturally part of Istanbul. Although at first we were somewhat wary of
18
Remote Landscapes in Flux – The Cide and Şenpazar Region
these statements, which sounded very much like the sort of thing one would say to impress liberal and secular foreigners, we gradually began to appreciate the fact that this was not an idle boast. For one thing, we came to realise that the region’s inhabitants are indeed more liberal and secular than in most rural regions of Turkey, which is partly related to the importance of tourism for the local economy. We also increasingly became aware of a much more direct connection with the metropolis in the form of a large-scale migration in recent years from Cide-Şenpazar. We encountered people with links to European countries such as the Netherlands and Germany, but the vast majority had migrated within Turkey and the destination was almost exclusively Istanbul. The reasons for this are of course the scale of the economy in the Istanbul region with its many employment opportunities for migrants from the countryside. In the case of Cide, however, there are also historic reasons for a preferential migration to Istanbul, which we will discuss in more detail in the following section. Today, the main transport route from Cide to the outside world, both in terms of volume of goods and the number of people travelling, leads not to Kastamonu, but follows the coast to Amasra, where the road turns south to join the main AnkaraIstanbul highway. The Cide-Amasra coastal route and its eastward extension to Sinop consist of a long series of sharp turns at steep angles on a very narrow road, which allows for an average speed of about 30 kilometres per hour, or much less when stuck behind one of the many lorries travelling this route. The importance of this route as the main transport artery of this coastal region is, therefore, all the more remarkable. However, the coastal road is also a relatively recent achievement. The Cide-Amasra stretch was built in 1967 with NATO support (personal communication Ali Osman Tuğtepe 6-7-2011). Prior to the construction of this road, there was virtually no land-based traffic to and from the Cide-Şenpazar region (also chapter 3). The senior inhabitants of the Cide area in particular have vivid memories of how transport was arranged prior to the construction of metalled roads. Before the construction of the artificial harbours at Cide and İlyasbey, large ships coming from Istanbul or the eastern Black Sea would anchor in the open waters outside the natural harbour of Gideros or off Cide beach. People and goods were loaded onto smaller boats and transported to and from the larger ships (Fig. 2.4). Such transfers could only take place when the sea was relatively calm and often the larger ships would simply by-pass the region. Nonetheless, this type of transport seems to have been of great importance to the Cide-Şenpazar region prior to the 1970s. By contrast, today Turkish ferry companies have completely abandoned Black Sea routes, probably because they are now outcompeted by coach companies which provide cheaper, faster and more flexible services. The above reflection on transport to and from the Cide-Şenpazar area and the relatively recent increase in their accessibility, may create the impression that this region was rather isolated throughout much of its history. This is not the case, however.
Humans in the Cide-Şenpazar Landscapes
19
Fig. 2.4: Photograph from the 1950s showing pier and water taxis to the steamboat at Cide. From the album of Fikriye Doğan. Provided by Nesrin Şahin and Murat Karasalihoğlu.
While it appears that the transport of staple goods in bulk, which is so common today, was less important before the recent construction of metalled roads, exchange of less voluminous goods has a long tradition in the region. The archaeological evidence collected during our survey attests to the region’s involvement in the long-distance exchange of goods from the earliest periods. This includes obsidian imported from central Anatolia in the Early Holocene (Düring and Gratuze 2013; chapter 5, this volume) and evidence for the participation of the Cide-Şenpazar region in cultural traditions centred elsewhere in Turkey and beyond, indicating sustained interaction across a wide area from the third millennium BC onwards (chapters 6 to 13). One key technology that significantly impacted on the connectivity of the CideŞenpazar region is seafaring. Broodbank (2006) has argued that in the Mediterranean, seafaring took off during the Younger Dryas (ca. 10,800-9600 BC), immediately before the onset of the Holocene. More systematic seafaring using canoes started around 6200 BC and was followed by voyaging in long boats commencing in the Aegean around 2600 BC. Sailing boats are first documented in the eastern Mediterranean and Egypt in the early second millennium BC, culminating in the well-attested counterclockwise seaborne trade which is exemplified by the Gelidonya and Uluburun shipwrecks found off the Lycian coast (Wachsman 1988; Bachhuber 2006; Broodbank 2006; 2008), as well as harbours dating to this period (Coleman 1986; Watrous 2012).
20
Remote Landscapes in Flux – The Cide and Şenpazar Region
In contrast to the Mediterranean, the earliest well-attested seafaring in the Black Sea dates to the seventh century BC, when the Ionian Greeks colonised the Black Sea shores (Tsetskhladze 1994; chapter 8, this volume). For example, Mycenaean pottery, which is so ubiquitous in the eastern Mediterranean and an index of inter-regional trade, is absent from the shores of the Black Sea, as are other types of evidence for maritime trade dating to the Late Bronze Age (Kolb 2004: 591-3). Likewise, no convincing shipwrecks or harbour installations predating the first millennium BC have been discovered on or off the shores of the Black Sea. The differences between the developments of seafaring in the Mediterranean and the Black Sea are remarkable. One element that might in part explain these divergent histories of maritime interaction is the geographical and climatic differences of the two seas. In the Mediterranean, early seafaring is documented especially in favourable geographical configurations – or ‘sea nurseries’ – of intervisible islands and coasts, such as the Aegean, the Gulf of İskenderun and Cyprus (Broodbank 2006). Such ideal geographical settings for the emergence of seafaring are almost entirely absent in the Black Sea. Except for the north-central area, for instance, there are no islands in the Black Sea. Furthermore, on the steep and rocky shores of its southern coast there are very few natural harbours where ships could seek shelter. Finally, the Black Sea is infamous for its bad weather conditions, which include the sudden onset of heavy storms, rainfall and fog, even in mid-summer (King 2004: 12). As a result, there is no season with predictably good weather conditions, such as the summer in the Mediterranean. This together with the fact that there are few places one can find refuge when the weather changes, make the Black Sea a much more hazardous sea to navigate than the Mediterranean. Travellers’ reports on their troubles at sea even in the relatively advanced ships of the early modern period speak volumes (chapter 3). Seafaring on the Black Sea must have been an even greater challenge in the boats and ships that were available in antiquity. Indeed, the Black Sea was known as Póntos Áxeinos to the Greeks, or ‘the inhospitable sea’ precisely because it is difficult to navigate and has few secure anchorages. It became known as Póntos Eúxeinos, or ‘the hospitable sea’, only after Greek colonisation (Strab. 1. 2. 10). Some have argued that this relatively late colonisation of the Black Sea in the seventh century BC may have been facilitated by advances in shipbuilding resulting in increased oar power (West 2003: 153). Once seafaring did take off in the Black Sea, the potential for trade along its shores and beyond was enormous. The diverse landscapes and climates of the regions surrounding the Black Sea held many desirable resources. A series of often competing trade networks encompassed the sea, including those of Greek, Genoese, and Venetian traders which flourished in periods when states surrounding the sea encouraged trade and managed to keep pirates at bay, but collapsed when they did not (Kortepeter 1966; King 2004; Bresson et al. 2007; Di Cosmo 2010). The location of the Cide region was of particular importance for seafaring due to the manner in which sea currents, or gyres, flow in the Black Sea basin (King 2004: 16). In the eastern part
Humans in the Cide-Şenpazar Landscapes
21
of the Black Sea there is a counter-clock wise current which facilitates crossings from the Crimea to Sinop (but not in reverse) and along the eastern shores to Georgia and back to the Crimea. In the western Black Sea there is also a counter-clock wise current which facilitates a crossing from the Cide region to the Crimea (but not in reverse) and from there to Romania, Bulgaria and the Bosphorus (Fig. 2.5). While trade across the Black Sea has been limited in recent history due to political circumstances, maritime trade to the west has been of great significance up until recent times. Within living memory, the Cide region was involved in largescale shipping of goods towards the Istanbul metropolitan area. This involved large wooden ships with gross tonnages of up to 450 tons (personal communication Recai Yilmaz 4-7-2011), which were able to transport their goods in the opposite direction of the prevailing currents.
2.2.2 Resources and Livelihoods in the Cide-Şenpazar Region The Cide-Şenpazar region has witnessed enormous changes in the past half-century, which make it difficult to conceive of its pre-modern landscapes and land-uses. Today the cultural landscapes of Cide-Şenpazar are in decline. Historic photographs show a countryside that was very different from the one we encounter today (Fig. 2.6). In the
Fig. 2.5: Sea currents in the Black Sea. Produced by Victor Klinkenberg.
22
Remote Landscapes in Flux – The Cide and Şenpazar Region
1950s, for instance, it was a much more open and intensively cultivated one. Large areas that are now forested were in use for agricultural production. Traditional villages, moreover, almost completely avoided the coast (Ilgaz 1998: 15). The tourist industry has resulted in a recent sprawl of apartment buildings along the Cide coastal plain. Maps of the Cide-Şenpazar region show a very dense network of traditional villages in the coastal hinterlands, many of which are almost completely abandoned today. Many village houses are now used as holiday homes by family members who live in Istanbul or elsewhere, and during the summer months, cars with Istanbul plates outnumber local ones. For the remainder of the year, only a few elderly permanent residents remain in most villages. They no longer practice any form of substantial agriculture. It is only a matter of time before these villages will be abandoned entirely, with the more remote places affected more severely than those that can be reached from metalled roads. In the upland village of Altıntaş in the Aydos valley, for instance, we encountered an old man with a bad leg living alone in what had been a small village in his youth, without a car or any other means of transport. He pointed to the forested mountain slopes below, on which the fields of the village had been located and which had been completely overgrown after their abandonment. This process of reforestation and abandonment of (remote) villages is occurring everywhere in the Cide-Şenpazar region and makes archaeological survey work much more challenging than it would have been only a few decades ago.
Fig. 2.6: Photograph from the 1950s: Cide town and surrounding landscape. Provided by Nesrin Şahin and Murat Karasalihoğlu.
Humans in the Cide-Şenpazar Landscapes
23
What did all the villagers in the Cide-Şenpazar region subsist on in the premodern era? Clearly one important component in the sub-recent economy was agriculture. Prior to the construction of metalled roads, most of the food consumed in the region must have been produced locally. Today, most of it is produced elsewhere – where production is cheaper - and agriculture has dwindled in importance in the local economy. Important subsistence crops that were grown in the region included wheat, barley, rice and maize, supplemented by a range of vegetables. These crops were grown on a substantial scale. Cuinet reports that no less than 121 mills existed in the Cide district in the late 19th century AD (chapters 3, 13 and 14), and these were presumably used for grinding cereals. Indeed we have found a substantial cluster of watermills near the village of Kalafat (S164), and additional watermills were encountered at Okçular (S105) and Öveçler Köy (not recorded). Subsistence agriculture would also have included animal husbandry for meat and dairy products. Today, cattle and water buffalo, rather than sheep and goats, are kept on a limited scale and we came across numerous cattle tracks used to drive animals to pastures away from the valley floors. Finally, flax seems to have been grown on some scale, mainly to produce linen textiles and clothes for local consumption. In antiquity, however, the area around Amasra, located due west of the Cide-Şenpazar region, was also known for its high quality wool (Madsen 2009: 22). Agricultural products exported from the Cide-Şenpazar region included eggs, apples and pears, all of which were shipped in large quantities to Zonguldak and Istanbul within living memory (personal communication Recai Yilmaz 4-7-2011). The main agricultural export of the area today is hazelnuts. Around 75% of global hazelnut production comes from northern Turkey.1 Hazelnut orchards are ubiquitous in the Cide-Şenpazar region, although most are located in the Turkish eastern Black Sea region. Hazelnuts were exported from the area as far as Egypt from at least the Hellenistic era onwards (Reger 2007: 274-5). In general, the Turkish Black Sea region provides an excellent climate for various cultivated trees such as apple, pear, cherry, walnut, chestnut and hazel (Zohary 1973: 629-33). Some of these trees, such as walnut and chestnut, were almost exclusively found in the Pontic refugium of northern Turkey at the end of the last Ice Age, and it has been suggested that the subsequent domestication and spread of these two species might have originated in this region (Zohary and Hopf 1994: 178-9). Cherries might also have been first domesticated in the Black Sea region. The word ‘cherry’ is related to that of the ancient town of Kerasus, today Giresun, a town on the eastern Turkish Black Sea coast from where the Romans adopted their cherries in the first century BC (Zohary and Hopf 1994, 172; Ascherson 1996: 177; Manoledakis 2010: 147; Mariette et al. 2010).
1 See: http://www.fas.usda.gov/htp/Hort_Circular/2004/3-05-04%20Web%20Art/03-04%20Hazelnut%20Web%20Article.pdf
24
Remote Landscapes in Flux – The Cide and Şenpazar Region
The agricultural potential of particular landscapes in the Cide-Şenpazar region is determined to a large degree by the distribution of soils and their characteristics. A map of the geological makeup of the Cide-Şenpazar region in relation to modern villages clearly shows the preference for some areas and the avoidance of others (Fig. 2.7). Notwithstanding the fact that the geological map lacks detail, this suggests that some soils were better suited for agriculture than others.
Fig. 2.7: Geological makeup of the Cide and Şenpazar districts (from Uğuz et al. 2002 and Aksay et al. 2002) and the distribution of villages and hamlets (stars) in the research area.
The distribution of villages also shows that there was no clustering of settlement along the coast, which suggests that fishing was not important as a full-time specialisation. According to local residents, fishing was done by individual households and fish was not traded (personal communication Recai Yilmaz 4-7-2011; personal communication Ali Osman Tuğtepe 6-7-2011) as it was abundant and easy to catch (also chapter 3). The Black Sea in the pre-modern era was famous for its rich fish resources, including schools of bonito (palamut), bluefish (lüfer) and anchovy (hamsı) (King 2004: 18), which have been decimated severely in recent decades due to overfishing (Ascherson 1996: 6, 248). Classical sources describe fishing on a large scale off the southern Pontic coast and the export of presumably smoked or dried fish, tuna in particular, to the Mediterranean (Lund and Gabrielsen 2005; Madsen 2009: 22), although the
Humans in the Cide-Şenpazar Landscapes
25
best archaeological evidence derives from the north Pontic coast (Opait 2007). In 14th century AD Byzantium, Black Sea caviar, probably from the north coast, was considered food for the poor because it was abundant and cheap (Ascherson 1996: 5). Another important activity related to the littoral up until recent decades was shipbuilding (Fig. 2.8). The long Cide beach and the adjacent coastal plain is ideal for hauling, repairing and building ships and the forested mountain slopes in the vicinity provide ample timber.
Fig. 2.8: Photograph from 1939 showing a newly constructed boat from Cide town being hauled to the sea by a group of men and water buffalos. Provided by Nesrin Şahin and Murat Karasalihoğlu.
In the 1980s, a large shipyard (tersane), which had been a major employer in the Cide area, was shut down and transferred to Istanbul. Many of the workers were offered employment in the Istanbul factory and migrated at that point. The migrant Cide community that formed in this manner supported others who followed the same route. Although the transfer of the shipyard to Istanbul was a significant event for the relationship between Cide-Şenpazar and Istanbul, this connection already existed much earlier. A model of a fast rowing boat of 30 metres length, called Sultan kayıkları, is displayed in the Cide kaymakamlık (district headquarters). In this barge, which was manufactured in Cide, the Ottoman sultans were transported across the Bosphorus. As summarised by Hannestad (2007: 86-7, 92-3; chapter 3), ancient authors such as Xenophon, Theophrastus and Strabo all remarked on the quality of shipbuilding timber and ships from the Turkish western Black Sea region. A related industry is the production and export of timber used for the construction of buildings or ships (personal communication Recai Yilmaz 4-7-2011). The timber (kereste) industry is still alive today, but it is difficult to estimate its importance in the
26
Remote Landscapes in Flux – The Cide and Şenpazar Region
past. Another relatively minor industry is quarrying. There are several marble quarries in the area, and we found one old stone quarry near Gideros harbour (S182), which is difficult to date however. None of the Classical or Ottoman sources mention stone quarries in this region (chapter 3) suggesting that these were not of great importance in the past. Other types of mining seem unimportant in the Cide-Şenpazar region, although in recent centuries there has been coal mining at Nanepınar (chapter 3). Copper was mined in the nearby Küre mountains from the fourth millennium BC onwards (De Jesus 1980; Yener et al. 1994) but not as far as we know in the CideŞenpazar region. A concentration of iron oxides or oxy-hydroxides (iron ore) at Okçular (chapters 4 and 14), however, points to small scale metallurgical activities probably dating to the Roman or Byzantine period. Summing up, the economy of the Cide-Şenpazar region in the early 20th century AD was to a large extent a local economy in which agriculture served local subsistence needs and could support a remarkably dense population of craftspeople, farmers and a middle class. Some export also took place of agricultural produce, such as eggs, apples and pears to the urbanised regions of Zonguldak and Istanbul. It is difficult to assess whether this agricultural export constitutes a specific development best understood within the context of the industrialisation of Turkey and the Ottoman empire, or whether the trade in fresh agricultural products also existed in earlier periods. Hazelnuts, which are transported much more easily than apples and eggs, were exported from the Hellenistic period onwards. Thus, we should be careful in conceptualising the regional economy as isolated in the past. Fishing, hunting and gathering no doubt supplemented agricultural production in the more recent past, but it is possible that fishing was of greater importance in some periods, such as the Roman era, than in others. The main industries in the region were timber export and shipbuilding, for which the Cide coastal plain is ideally suited. What is difficult to assess is how changing maritime trade networks impacted on the economy of the region. While in the last century, trading across the Black Sea almost completely ceased, in the past the Cide-Şenpazar region would have been well placed to play a role in Black Sea exchange networks.
2.3 A Landscape in Flux Anthropologists have deconstructed the tendency to view societies under study as static and frozen in time (Fabian 1983). This critique applies equally to archaeologists working in the Near East. In the Cide-Şenpazar region it quickly became apparent that people are enmeshed in economic and social relations extending across Turkey and beyond, and that both local society and economy changed considerably over the past decades. There is no reason to think that similarly large-scale socio-economic transformations would not have occurred regularly in the occupation history of the Cide-Şenpazar region. In this section the focus will be on the taphonomic changes in the area, which, as will be shown here, are both substantial and varied.
A Landscape in Flux
27
On the basis of a geoarchaeological survey, led by Geuch de Boer in 2010 and 2011, the Cide-Şenpazar region can be divided into a series of distinct landscape zones and processes: 1. The Cide coastal plain, which seems to consist mainly of seaborne deposits. 2. The rivers and river valleys. Rivers in the region are without exception highgradient and thus fast flowing and carry large volumes of water. They do not aggrade or deposit sediments, but cut into existing sediments. Some of the river valleys provide natural routes into the interior of Cide and Şenpazar, and remains of old roads, bridges and many villages are located along these corridors. 3. The slopes, minor valleys and hills that are located adjacent to the coast, which are among the most densely occupied areas of the region. 4. Cultivated slopes in higher altitudes and hilltops are found in the interior. 5. High-altitude mountain landscapes that are too rocky or steep to be suitable for agriculture apart from pasturing animals. Our geoarchaeological investigations focused on the coastal plain, the river valleys leading up to the coast, the slopes, minor valleys and hills located adjacent to the coast. These are all areas in which landscape transformations appear to have been substantial (Fig. 2.9). By contrast, the cultivated uplands of the interior appear to have been more stable and were not investigated in as much detail.
Fig. 2.9: Locations of geoarchaeological investigations in the Cide-Şenpazar region. Produced by Toby C. Wilkinson.
28
Remote Landscapes in Flux – The Cide and Şenpazar Region
We investigated the Cide coastal plain and the adjacent smaller Kalafat coastal plain through a series of profiles exposed by gravel pits, building foundations, and in river sections. This method was based on two practical considerations: first, investigating already exposed sections was the most efficient use of the limited amount of time available; and, second, coring, the traditional method used for geoarchaeological reconstructions, is not normally permitted in the context of a standard survey permit in Turkey. Fortunately, because of low ground water levels and the many gravel and construction pits in Cide and the Kalafat area, we nevertheless were able to gather information about the formation of these coastal plains. Due west of Cide town, a cluster of gravel extraction pits yielded much information about the geology of the coastal plain. We were able to recognise several phases of sea regression and transgression, indexed by an alteration of fine, off shore deposits and gravelly near shore sediments. In one profile (S67) there was a dark band of clay with heavy silt and a moderate amount of humus about a metre below the current surface which yielded Roman ceramics. This suggests that substantial seaborne deposits have accumulated in the Cide coastal plain since the Roman period, a hypothesis confirmed by other archaeological observations. Roman period spolia and ceramics are found in specific places in the coastal plain and appear to be almost always connected to substantial ground disturbances such as large pits (also chapter 10). In an adjacent profile (S71) there were Byzantine period sherds in the section ca. 40cm below the topsoil, perhaps suggesting that no substantial sedimentation had occurred since that period. Various proxy data suggest a transgression of the sea starting in the Late Byzantine period. At Gideros (S12), a rock-cut mooring surface in the harbour is visible which can be reached via a rock-cut stairway from the Byzantine castle, and is now submerged under about 70cm of water, attesting to a sea level rise in the area. At Kalafat, a Byzantine church and associated cemetery (S116) is currently being eroded by the sea. Originally this location must have been at some distance from the sea. Finally, at Gökçekale we found various structures, including what appear to be a wine or olive oil press, a building and a church apse (S174/176/178) on the present beach. Again, we may assume that at the time of their construction the sea was at a lower level. Combining the data from the two Cide profiles (S67 and S71) we have the following broad sequence (table 2.1). A fairly recent but undated transgression phase at minus 30-38cm; a long regression phase prior to this, with evidence for Roman and Byzantine period occupation; and, finally, an earlier transgression phase at minus 175-200cm. Below this is a clean sand beach deposit, and given that the profiles are located close to the foothills, this can be taken to represent the start of the formation of the Cide coastal plain. Two optically stimulated luminescence (OSL) samples from this deposit date it to 1850 BC (section 16.1). This means that the entire Cide coastal plain postdates 1850 BC and that it might not have existed in earlier periods. However, it is also possible that the sea was a few meters lower,
A Landscape in Flux
29
and that there was an equivalent coastal plain at a lower elevation. The available bathymetric data are not precise enough to identify possible older terraces or shelves at Cide. Tab. 2.1: Combined description of two profiles in the Cide coastal plain (S67 and S71). Regression phases indicated in italics. depth (cm)
description
ceramics / date
0-30
silty clay with medium-fine subangular blocky structure, brownish grey rounded pebbles up to 10cm in sand matrix silty clay with coarse prismatic structure and a little humus, light brownish grey silty clay deposit with some humus and medium prismatric structure, dark grey silty clay deposit with some humus, brownish grey rounded beach pebbles and sand, brownish grey clean sand, dark grey
-
30-38 38-75 75-110 110-175 175-200 200-275
Byzantine Roman OSL date 1850 BC
The Kalafat coastal plain, located directly west of that of Cide, has a similar sequence. However, colluvial processes are more apparent in the smaller Kalafat plain, which is surrounded by hills. At the mouth of the Kalafat river (S72), the river section shows an old beach at the bottom and seaborne clay deposits above. About 100 metres inland, a ca. 2m deep foundation trench (S73) yielded Early Byzantine ceramics at the bottom. In this section there were seaborne clayey deposits, which were overlain by colluvial material that had eroded from the surrounding hills. In Kalafat, we seem to have an old (undated beach) about 130cm below the current surface, overlain by silty clay deposits, presumably deposited behind a coastal beach barrier, and followed by colluvial sediments close to the hills and erosion caused the sea transgression in the last century. This is documented by the erosion of sub-recent graves at Kalafat cemetery (S116). We also investigated several locales in the immediate coastal hinterland, which is today among the most densely occupied areas of the region. Much of our archaeological work focused on the landscapes of Okçular and Abdulkadir (chapter 14). In both landscapes there is a vertical zoning with, first, rock outcrops surrounded by macquis on top; second, relatively flat shoulders on the south-facing slopes, where modern houses are located and where we found archaeological clusters; third, hill slopes with substantial gully and sheet erosion; and fourth, the flat valley bottom in which sediments accumulate. Okçular and Abdulkadir differ from many other local landscapes in the degree of erosion that has occurred in them. However, we could not detect differences in the land-use of the Okçular and Abdulkadir valleys and of those landscapes that were less
30
Remote Landscapes in Flux – The Cide and Şenpazar Region
affected by erosion. Therefore, these differences are most likely related to the local geology. Of note are folded and oblique occurrences of different rocks and sediments. It is likely that erosion is more pronounced where the softer shale deposits surface, for example in the eastern area of Okçular, whereas other areas with the same land-use are less affected. At the same time, it is plausible that better maintenance of the land, by means of wooded fences and the filling up of emergent erosion gullies, might have prevented part of this erosion in the past. On the eroded slopes of Okçular and Abdulkadir we encountered considerable quantities of sherds, tile fragments and lithics. It is plausible that much of this material derives from the eroding shoulders of the hills above, where houses are located today and where we found clusters of artefacts in our survey (chapter 14). Typically, on the lower hill slopes and in the valley bottoms, tile fragments and sherds dating to the Byzantine period occur at considerable depths. At one cleaned gully profile at Abdulkadir (S76) there were tile fragments up to a depth of 130cm below surface. Also, a profile in the Adulkadir valley bottom (S78) showed several horizontal layers, some cleaner, others with surfaces with a lot of anthropogenic material (tile fragments and charcoal). There was a lot of anthropogenic material until about 130cm below surface, where more or less clean clay was encountered, which contained a Roman sherd. This particular profile suggests multiple erosion episodes occurring during the Byzantine period. However, unlike in the coastal plain, these processes would not necessarily have sealed off pre-Byzantine materials, given that the erosion of the hill shoulders also could have released older artefacts. In the river valleys leading up to the interior, several geological profiles were investigated. One aim of the geoarchaeological survey was to find ancient river terraces, but none were identified in the regions investigated. Instead, the river valleys of the Aydos, Güble, and Devrekani have all cut into steep v-shaped valleys with a lot of colluvial erosion (most clearly visible at sites S200 and S202). Although the few ploughed fields which we encountered obviously suffered from erosion, no precautions were taken against this in the form of terracing, for instance. It appears, also from talking to farmers, that large deposits of sediment are available for agriculture and that loss of soil was not a concern. When a field becomes depleted, there is more than enough land to relocate fields and houses if necessary. Achieving an ecological balance does not seem a high priority among the farmers in the Cide-Şenpazar region. The landscapes of the cultivated higher slopes and hilltops located in the interior of the Cide-Şenpazar region appear more stable than other areas in the region. These landscapes show fewer signs of erosion, presumably because geologically stable landscapes were selected for farming. It is perhaps no coincidence that our best prehistoric site on the surface, consisting of a large cluster of chipped stone, was found in one of these landscapes at Aybasan (chapter 5). The various types of landscapes in the Cide-Şenpazar region thus have markedly different properties in terms of the geological processes that occur in them and the ways in which they affect what type of archaeology we can expect
Summary
31
to find in them (Tab. 2.2). These differences will be illustrated in the subsequent chapters of this book. Tab. 2.2: Landscape types in the Cide-Şenpazar region, the geological processes that occur in them, and how this may affect archaeology. landscape type
geological processes
archaeology
coastal plain
sediments deposited/eroded by the sea colluvial processes on steep slopes erosion and colluvial processes of shoulders and slopes
only recent archaeology accessible on surface only recent archaeology accessible on surface erosion of hill shoulders also exposes pre-Byzantine assemblages partial preservation of Pre- and Protohistoric surfaces unsuitable for settlements, but some inhabited caves, etc.
river valleys coastal hinterland/lower hills
cultivated inland slopes and hilltops high mountains
erosion less pronounced relatively stable
2.4 Summary An assessment of the archaeology of the Cide-Şenpazar region can only be undertaken by taking its ecological context seriously. The ecology of a region affects what types of foodstuffs and materials are available in the area and what livelihoods are possible. The geological properties of the region determine which areas are most suitable for agriculture, and a range of geological processes profoundly affect the preservation, accessibility and visibility of archaeological assemblages. Transport technologies to transverse both the region’s rugged mountains and the varyingly hospitable Black Sea would have had a great impact on how people in the region interacted economically and socially with other societies. Despite transport difficulties there are strong arguments for cultural links with inner Anatolia and the Circumpontic in our assemblages, but it is also clear that the intensity of interactions in both directions fluctuated considerably in the course of history.
Bibliography websites http://www.fas.usda.gov/htp/Hort_Circular/2004/3-05-04%20Web%20Art/03-04%20 Hazelnut%20Web%20Article.pdf (last accessed 3 December 2014) www.ncdc.noaa.gov (last accessed 3 December 2014).
32
Remote Landscapes in Flux – The Cide and Şenpazar Region
Greco-Roman source Strabo, Geographica. [Translated by H.C. Hamilton and W. Falconer]. London, G. Bell & Sons, 1903. published sources Aksay, A., Pehlivan, S., Gedik, I., Bilginer, E., Duru, M., Akbaş, B., and Altun, I. 2002. Geological Map of Turkey - Zonguldak (1:500:000). Ankara, Maden Tekikve Arama Genel Müdürlüğü. Aksu, A.E., Hiscott, R.N., and Yaşar, D. 1999. Oscillating Quaternary water levels of the Marmara Sea and vigorous outflow into the Aegean Sea from the Marmara Sea–Black Sea drainage corridor. Marine Geology 153: 275-302. Aksu, A.E., Hiscott, R.N., Mudie, P.J. Rochon, A., Kaminski, M.A., Abrajano, T., and Yaşar, D. 2002a. Persistent Holocene outflow from the Black Sea to the Eastern Mediterranean contradicts Noah’s flood hypothesis. Geological Society of America 12(5): 4-10. Aksu, A.E., Hiscott, R.N., Yaşar, D., Işler, F.I., and Marsh, S. 2002b. Seismic stratigraphy of Late Quarternay deposits from the southwestern Black Sea shelf: Evidence for non-catastrophic variations in sea level during the last 10 000 yr. Marine Geology 190: 61 - 94. Akyol, Z., Arpat, E., Erdoğan, B., Göğer, E., Güney, Y., Şaroğlu, F., Şentürk, İ., Tütüncü, K. and Uysal, Ş. 1974. Geological Map of the Cide-Kurucaşile Region, Scale 1:50,000. Ankara, Maden Tetkikve Arama Enstitüsü. Alex, M. 1985. Klimadaten ausgewählter Stationen des Vorderen Orients. Wiesbaden, Dr. Ludwig Reichert Verlag. Ascherson, N. 1996. Black Sea: The Birthplace of Civilisation and Barbarism. London, Hill and Wang. Bachhuber, C. 2006. Aegean interest on the Uluburun ship. American Journal of Archaeology 110: 345-63. Bailey, D.W. 2007. Holocene changes in the level of the Black Sea: Consequences at human scale. In V. Yanko-Hombach, A.S. Gilbert, N. Panin, and P.M. Dolukhanov, eds., The Black Sea Flood Question: Changes in Coastline, Climate and Human Settlement. Dordrecht, Springer: 515-36. Ballard, R.D., Hiebert, F.T., Coleman, D.F., Ward, C., Smith, J.S., Willis, K., Foley, B., Croff, K., Major, C., and Torbe, F. 2001. Deepwater archaeology of the Black Sea: The 2000 season at Sinop, Turkey. American Journal of Archaeology 105: 607-23. Bar-Matthews, M., Ayalon, A., Laufman, A., and Wasserburg, G.J. 1999. The eastern Mediterranean paleoclimate as a reflection of regional events: Soreq cave, Israel. Earth and Planetary Science Letters 166: 85-95. Bottema, S., Woldring, H., and Aytuğ, B. 1993/1994. Late Quarternary vegetation history of northern Turkey. Palaeohistoria 35/36: 13-73. Bresson, A., Ivantchik, A., and Ferray J.-L., eds. 2007. Une koinè pontique: cités greques, societies indigenes et empires mondiaux sur le littoral nord de la mer noire. Bordeaux, Ausonius. Broodbank, C. 2006. The origins and early development of Mediterranean maritime activity. Journal of Mediterranean Archaeology 19: 199-230. — 2008. The Early Bronze Age in the Cyclades. In C.W. Shelmerdine, ed. The Cambridge Companion to the Aegean Bronze Age. Cambridge, Cambridge University Press: 47-76. Brückner, H., Kelterbaum, D., Marunchak, O., Porotov, A., and Vött, A. 2010. The Holocene sea level story since 7500 BP - Lessons from the Eastern Mediterranean, the Black and Azov Seas. Quarternary International 225: 160-79. Coleman, J.E. 1986. Excavations at Pylos in Elis. Princeton, American School of Classical Studies at Athens. De Jesus, P.S. 1980. The Development of Prehistoric Mining and Metallurgy in Anatolia. Oxford, British Archaeological Reports.
Bibliography
33
Dean, W.T., Monod, O., Rickards, R.B., Demir, O., and Bultynck, P. 2000. Lower Palaeozoic stratigraphy and palaeontology, Karadere–Zirze area, Pontus Mountains, northern Turkey. Geological Magazine 137(5): 555-82. Dergachev, V.A., and Dolukhanov, P.M. 2007. The Neolithization of the north Pontic area and the Balkans in the contexts of the Black Sea floods. In V. Yanko-Hombach, A.S. Gilbert, N. Panin, and P.M. Dolukhanov, eds., The Black Sea Flood Question: Changes in Coastline, Climate and Human Settlement. Dordrecht, Springer: 489-514. Di Cosmo, N. 2010. Black Sea emporia and the Mongol empire: A reassessment of the pax mongolica. Journal of the Economic and Social History of the Orient 53: 83-108. Doonan, O. 2004. Sinop Landscapes, Exploring Connectivities in a Black Sea Hinterland. Philadelphia, University of Pennsylvania Museum of Archaeology and Anthropology. Düring, B.S. 2008. The Early Holocene occupation of North-Central Anatolia between 10,000 and 6,000 BC cal: Investigating an archaeological terra incognita. Anatolian Studies 58: 15-46. — 2011. The Prehistory of Asia Minor: From Complex Hunter-Gatherers to Early Urban Societies. Cambridge, Cambridge University Press. Düring, B.S., and Gratuze, B. 2013. Obsidian exchange networks in Prehistoric Anatolia: New data from the Black Sea region. Paléorient 39(2): 173-82. Erol, O. 1983. Die naturräumliche Gliederung der Türkei. Wiesbaden, Dr. Ludwig Reichert Verlag. Fabian, J. 1983. Time and the Other, How Anthropology Makes its Object. New York, Columbia University Press. Fleitmann, D., Cheng, H., Badertscher, S., Edwards, R.L., Mudelsee, M., Göktürk, O.M., Fankhauser, A., Pickering, R., Raible, C.C., Matter, A., Kramers, J., and O. Tüysüz 2009. Timing and climatic impact of Greenland interstadials recorded in stalagmites from northern Turkey. Geophysical Research Letters 36: L19707. Govedarica, B. 2003. On the oscillations of the Black Sea level in the Holocene period from an archaeological viewpoint. In G.A. Wagner, E. Pernicka, and H.-P. Uerpmann, eds., Troia and the Troad, Scientific Approaches. Berlin, Springer: 95-104. Hannestad, L. 2007. Timber as a trade resource of the Black Sea. In V. Gabrielsen and J. Lund, eds., The Black Sea in Antiquity, Regional and Interregional Economic Exchanges. Aarhus, Aarhus University Press: 85-99. Hiebert, F.T., Ballard, R.D. Coleman, F.D. Ward, C. Torre, F. Miller, N. and Woods W. 2002. Deepwater archaeology of the Black Sea. TÜBA-AR 5: 95-117. Ilgaz, R. 1998 [1976]. Sarı Yazma. Istanbul, Cınar. Jablonka, P. 2003. The link between the Black Sea and the Mediterranean since the end of the last Ice Age, archaeology and geology. In G.A. Wagner, E. Pernicka, and H.-P. Uerpmann, eds., Troia and the Troad, Scientific Approaches. Berlin, Springer: 77-94. King, C. 2004. The Black Sea: A History. Oxford, Oxford University Press. Kolb, F. 2004. Troy VI: A trading center and commercial city? American Journal of Archaeology 108(4): 577-614. Kortepeter, C.M. 1966. Ottoman imperial policy and the economy of the Black Sea region in the sixteenth century. Journal of the American Oriental Society 86: 86-113. Lund, J., and Gabrielsen, V., 2005. A fishy business. Transport amphorae of the Black Sea region as a source for the trade in fish and fish products in the Classical and Hellenistic periods. In T. Bekker-Nielsen, ed., Ancient Fishing and Fish Processing in the Black Sea Region. Aarhus, Aarhus University Press: 161-9. Madsen, J.M. 2009. Eager to be Roman: Greek Responses to Roman Rule in Pontus and Bithynia. London, Duckworth.
34
Remote Landscapes in Flux – The Cide and Şenpazar Region
Manoledakis, M. 2010. Choirades, Kerasous, Pharnakaie. Observations on three ancient placenames in the southern Black Sea. Ancient West and East 9: 137-55. Marek, C. 2003. Pontus et Bithynia. Die römischen Provinzen im Norden Kleinasiens. Mainz am Rhein, Philipp von Zabern. Mariette, S., Tavaud, M., Arunyawat, U, Capdeville, G., Millan, M., and Salin, F. 2010. Population structure and genetic bottleneck in sweet cherry estimated with SSRs and the gametophytic self-incompatibility locus. BMC Genetics 11(77): 1-13. Marsh, B., Roberts, N., Toprak, V., Matthews, R., Eastwood, W., Carolan, J., Arcasoy, A. and Süzen, M.L. 2009. Contexts of human interaction: Geology, geography and environment. In R. Matthews and C. Glatz, eds., At Empires’ Edge: Project Paphlagonia. Regional Survey in NorthCentral Turkey. Ankara, British Institute at Ankara: 27-74. Okay, A.I. 2008. Geology of Turkey: A synopsis. Anschnitt 21: 19-42. Opait, A. 2007. A weighty matter: Pontic fish amphorae. In V. Gabrielsen and J. Lund, eds., The Black Sea in Antiquity, Regional and Interregional Economic Exchanges. Aarhus, Aarhus University Press: 101-121. Özdoğan, M. 1997. The beginning of Neolithic economies in Southeastern Europe: An Anatolian perspective. Journal of European Archaeology 5(2): 1-33. — 1999. Northwestern Turkey: Neolithic cultures in between the Balkans and Anatolia. In M. Özdoğan and N. Başgelen, eds., The Neolithic of Turkey. Istanbul, Arkeoloji ve Sanat Yayınları: 203-224. — 2003. The Black Sea, the Sea of Marmara and Bronze Age archaeology: An archaeological predicament. In G.A. Wagner, E. Pernicka, and H.-P. Uerpmann, eds., Troia and the Troad, Scientific Approaches. Berlin, Springer: 105-120. — 2007. Coastal changes of the Black Sea and Sea of Marmara in archaeological perspective. In V. Yanko-Hombach, A.S. Gilbert, N. Panin, and P.M. Dolukhanov, eds., The Black Sea Flood Question: Changes in Coastline, Climate and Human Settlement. Dordrecht, Springer: 651-69. Reger, G. 2007. Traders and travellers in the Black and Aegean Seas. In V. Gabrielsen and J. Lund, eds., The Black Sea in Antiquity, Regional and Interregional Economic Exchanges. Aarhus, Aarhus University Press: 273-85. Ryan, W.B.F., Pitman, W.C., Major, O.C., Shimkus, K., Moskalenko, V., Jones, G.A., Dimitrov, P. Goriir, N., Saking, M., and Yiice, H. 1997. An abrupt drowning of the Black Sea shelf. Marine Geology 138: 119-26. Ryan, W., and Pitman, W. 1999. Noah's Flood. The New Scientific Discoveries about the Event that Changed History. New York, Simon & Schuster. Shilik, K.K. 1997. Oscillations of the Black Sea and ancient landscapes. In J. Chapman and P. Dolukhanov, eds., Landscapes in Flux. Central and Eastern Europe in Antiquity. Oxford, Oxbow Books: 115-29. Shumilovskikh, L.S., Tarasov, P., Arz, H.W., Fleitmann, D., Marret, F., Nowaczyk, N., Plessen, B., Schlütz, F., and Behling, H. 2012. Vegetation and environmental dynamics in the southern Black Sea region since 18 kyr BP derived from the marine core 22-GC3. Palaeogeography, Palaeoclimatology, Palaeoecology 337/8: 177-93. Tsetskhladze, G.R. 1994. Greek penetration of the Black Sea. In G.R. Tsetskhladze and F. De Angelis, eds., The Archaeology of Greek Colonisation, Essays Dedicated to Sir John Boardman. Oxford, Oxford University Committee for Archaeology: 111-35. — 2007. Greeks and locals in the southern Black Sea littoral: A re-examination. In G. Herman and I. Shatzman, eds., Greeks between East and West. Essays in Greek Literature and History in Memory of David Asheri. Jerusalem, The Israel Academy of Sciences and Humanities: 160-95.
Bibliography
35
Tunoğlu, C. 2008. Kastamonu’nun Jeolojikve Doğal Özellikleri. In L. Seymen, ed., Üsküdar’a kadar Kastamonu. Istanbul, Yapı Kredi Yayınları: 23-34. Turney, C. S. M., and H. Brown 2007. Catastrophic early Holocene sea level rise, human migration and the Neolithic transition in Europe. Quarternary Science Reviews 26: 2036-41. Uğuz, M.F., Sevin, M., and Duru, M. 2002. Geological Map of Turkey - Sinop (1:500,000). Ankara, Maden Tekikve Arama Genel Müdürlüğü. Wachsmann, S., ed., 1988. Seagoing Ships and Seamanship in the Bronze Age Levant. London, Texas A & M University Press. Watrous, L.V. 2012. The harbour complex of the Minoan town of Gournia. American Journal of Archaeology 116: 521-41. West, S. 2003. The most marvellous of all seas: The Greek encounter with the Euxine. Greece & Rome 50(2): 151-67. Yanko-Hombach, V. 2007a. Controversy over Noah’s flood in the Black Sea: Geological and foraminifera evidence from the shelf. In V. Yanko-Hombach, A.S. Gilbert, N. Panin, and P.M. Dolukhanov, eds., The Black Sea Flood Question: Changes in Coastline, Climate and Human Settlement. Dordrecht, Springer: 149-203. — 2007b. Late Quaternary history of the Black Sea: An overview with respect to the Noah’s Flood Hypothesis. In G. Erupt and S. Mitchell, eds., The Black Sea: Past, Present and Future. London, British Institute at Ankara: 5-20. Yener, K.A., Geçkinli, E., and Özbal, H. 1994. A brief survey of Anatolian metallurgy prior to 500 BC. In S. Demirçi, A.M. Özer, and G.D. Summers, eds., Archaeometry 94, The Proceedings of the 29th International Symposium on Archaeometry. Ankara, Tübitak: 375-90. Zeist, W. van, and Bottema, S. 1991. Late Quarternary Vegetation of the Near East. Wiesbaden, Dr. Ludwig Reichert Verlag. Zohary, D. 1973. Geobotanical Foundations of the Middle East. Stuttgart, Gustav Fischer Verlag. Zohary, D., and Hopf, M. 1994 [second edition]. Domestication of Plants in the Old World: The Origins and Spread of Cultivated Plants in West Asia, Europe, and the Nile Valley. Oxford, Clarendon Press. Zumbühl, A. 2010. History of the Black Sea recorded in Stalagmites from Northern Turkey. MA Thesis, Bern University.
Murat Karasalihoğlu and Bleda S. Düring
3 Traveller’s Notes on the Cide Region through the Ages The Cide region1 features in a remarkable variety of travel documents written over the course of more than two millennia. The earliest written attestation of the region is in the Iliad, in which troops from the area take part in the Trojan War. In later periods, the Cide region is often mentioned and described as an important part of the southern Black Sea coast. This is the case for Greek, Roman, medieval and early modern sources. In this chapter we aim to discuss the information on the Cide region that can be gleaned from these sources. At the outset the limitations of our approach in this discussion should be stated clearly. First, our coverage does not claim to include all written comments on the Cide region: without doubt there are many sources we have not been able to track down. In particular, we expect that for the early modern period there are many sources in languages other than English, French and Turkish (for example Russian and Greek) that we have not been able to include, for the simple reason that such sources are more difficult to locate and require additional language skills. Further, it is plausible that many Italian reports on the region exist. After all, the Venetian and Genoese both had active trading interests in the Black Sea (Di Cosmo 2010) and controlled some of the key ports such as Amasra, and probably also the stronghold of Çoban Kalesi (S32) in the Cide region (chapter 12). They also produced detailed maps of the Black Sea coasts (Crow and Hill 1995; Belke 1996: 185). In this chapter, we focus on easily accessible sources. Much more data could be gathered, if more resources and time were available. Second, the temporal coverage and quality of the available sources dealing with the Cide region vary enormously, creating a rather uneven picture. Greek and Roman sources often describe distances between coastal settlements along the Black Sea, harbour facilities, and natural resources, and in some cases mythology. After their reports of the region cease around AD 400, there is a full millennium for which we have been unable to find historical sources. Subsequently, the earliest reports, dating to the 15th and 16th centuries AD, are travel itineraries focusing on the description of towns, resources, local customs, and industries of the Turkish western Black Sea region. Many of the later sources also refer to Greek and Roman authors describing the same topography. In early modern travel reports, an increasing amount of information
1 All available sources either deal with the coastal part of the Cide-Şenpazar region or with the aggregate Cide region before Şenpazar became a separate district. Thus, no information dealing specifically with Şenpazar was encountered and we will therefore simply discuss the Cide region here.
Cide in Greek and Roman Sources
37
is included, especially in those dating to the 19th century AD, although the accuracy of this information is perhaps not always beyond doubt. Despite the limitations of these accounts of the Cide region, much information that can be gleaned from them. This includes the location of towns and villages in antiquity, the political and cultural affiliations of the region, how people travelled and traded along the Black Sea coast, and what the local economy was like.
3.1 Cide in Greek and Roman Sources The Cide region is mentioned in one way or another in a remarkably large number of Greco-Roman sources. Many of these mention Cide only in passing, while a few provide detailed information on travel distances and towns in the area. We should bear in mind, however, that not all authors actually travelled in the region. Other sources are concerned with reconstructing global geography and mention the Turkish western Black Sea region only in this context. One of the earliest mentions of the Cide region appears in the Iliad. The genesis of the Iliad and the chronology of different aspects of the narrative have been the subject of an extended debate (e.g. Heiden 2008) that need not concern us here. In the description of troops fighting for Troy, Homer lists several contingents from Paphlagonia, a large region in north-central Turkey which includes Cide: — “And the Paphlagonians did Pylaemenes of the shaggy heart lead from the land of the Eneti, whence is the race of wild she-mules. These were they that held Cytorus and dwelt about Sesamon, and had their famed dwellings around the river Parthenius and Cromna and Aegialus and lofty Erythini.” (Hom. Il.2. 851-855 [Murray 1924]).
Here, Homer mentions various localities in the Turkish western Black Sea region, including Cytorus: probably to be equated with the small natural harbour of Gideros. Subsequently the name of Gideros may have been transferred to the modern town of Cide. Sesamon, which was later renamed Amastris, is today’s Amasra, and Cromna can be equated with Kuruçaşile. Aegialus is the modern town of Cide and Erythini modern Çakraz (Belke 1996). In the Periplus of the Pseudo-Skylax, which should probably be dated in the fourth century BC, several Hellenic towns in the Turkish western Black Sea region are mentioned as follows: — “Paphlagonia. And after Assyria is the nation of Paphlagonia. And there is in it Stephane, a harbour; Koloussa, a Hellenic city; Kinolis, a Hellenic city; Karambis, a Hellenic city; Kytoris, a Hellenic city; Sesamos, a Hellenic city, with Parthenios, a river; Tieion, a Hellenic city with a harbour; Psylla with a river, Kallichoros.” (Ps.-Skylax 90 [Shipley 2002]).
38
Traveller’s Notes on the Cide Region through the Ages
Here various localities are listed moving from east to west, including Karambis: located by present Kerempe Burnu near the modern town of İlyasbey; Kytoris: Gideros/Cide; and Sesamos: Amasra (Avram et al. 2004). In the third century BC Apollonius of Rhodes, writing about the Argonauts’ quest for the Golden Fleece, mentions the Turkish western Black Sea region. In his version, the Argonauts travelled along the southern Black Sea coast: — “Now there is a headland opposite Helice the Bear, steep on all sides, and they call it Carambis, about whose crests the blasts of the north wind are sundered. So high in the air does it rise turned towards the sea. And when ye have rounded it broad Aegialus...” (Apollon. 2. 360-365 [Seaton 1919]).
— “Then they sped onward in the night without ceasing, and passed Sesamus and lofty Erythini, Crobialus, Cromna and woody Cytorus. Next they swept round Carambis at the rising of the sun, and plied the oars past endless Aegialus, all day on through the night.” (Apollon. 2. 940-950 [Seaton 1919]).
In a later rendition of the quest for the Golden Fleece, Valerius Flaccus, who died around AD 90, depicted the Cide region as follows: — “Why should I tell thee of Carambis that rises upon her cloud-encircled cliff, why of the whirling waters of Iris or of Ancon?” (V. Fl. 4. 499 [Mozley 1934]).
— “And next, as the vessel speeds along, they send deep beneath the horizon Cromne’s ridge and pale Cytorus and thee, Erythia. And now heaven was bringing back the night; closely skim they high Carambis, and vast upon the sea trembles the shadow of Sinope.” (V. Fl. 4. 101-110 [Mozley 1934]).
By far the most detailed ancient source dealing with the geography of the Turkish western Black Sea region is Strabo who lived from ca. 64 BC to AD 21. In his Geography, he provides much information about Cide, such as the origins of the name and who founded it: — “After the river Parthenius is Amastris, bearing the same name as the princess by whom it was founded. It is situated upon a peninsula, with harbours on each side of the isthmus. Amastris was the wife of Dionysius, the tyrant of Heraclea, and daughter of Oxyathres, the brother of the Darius who fought against Alexander. She formed the settlement out of four cities, Sesamus, Cytorum, Cromna, (mentioned by Homer in his recital of the Paphlagonian forces) and Tieium, which city however soon separated from the others, but the rest continued united. Of these, Sesamus is called the citadel of Amastris. Cytorum was formerly a part of the people of Sinop. It had its name from Cytorus, the son of Phrixus, according to Ephorus. Box-wood of the best quality grows in great abundance in the territory of Amastris, and particularly about Cytorum. Ægialus is a line of sea-coast, in length more than 100 stadia. On it is a village of the same name, which the poet mentions in these lines, “Cromna, and Ægialus, and the lofty Erythini.” (Strab. 12. 3. 10 [Hamilton 1903]).
Cide in Greek and Roman Sources
39
Thus, according to Strabo, Cytorus was named after the son of Phrixus, a person known from the story of the Argonauts and the Golden Fleece as the main adversary of the Argonauts. Here, we can question the accuracy of Strabo’s account, given that the Argonauts were not historical personae, but mythical figures. While Strabo’s account of the Black Sea littoral was mainly descriptive, other authors report precise distances between localities and provide their locations within a coordinate system. Gyres, winds and how they impact navigation are also being discussed. One of the most eminent geographers dealing with the region is Pliny, who lived between AD 23 and 79. In his Natural History he recounts many specific details about the Cide region, for example that the promontory of Criumetopon, the southernmost point of the Crimea is directly opposite Carambis (ibid.: 6. 1. 26 [Bostock 1855]), and: — “[t]he city also of Sesamon, now called Amastris, Mount Cyotorus, distant sixty-three miles from Tium, the towns of Cimolis and Stephane, and the river Parthenius. The promontory of Carambis, which extends a great distance into the sea, is distant from the mouth of the Euxine three hundred and twenty-five miles, or, according to some writers, three hundred and fifty, being the same distance from the Cimmerian Bosphorus, or, as some persons think, only three hundred and twelve miles. There was also a town of the same name, and another near it called Armene; we now find there the colony of Sinop, distant from Mount Cyotorus one hundred and sixty-four miles.” (Plin. Nat. 6.2.(2) [Bostock 1855]).
Pomponius Mela, who wrote around AD 43, also notes that the Crimean peninsula is directly opposite cape Carambis (Pomp. Mel. 2. 3 [Romer 1998]; see also Amm. Mar. 22. 8. 20 [Rolfe 1935-1940]), and describes the Turkish western Black Sea region as follows: — “After that comes the town of Tios, in fact a colony of the Milesians, but now belonging to the land and people of Paphlagonia. More or less in the middle of their littoral is Point Carambis. On its nearer side is the Parthenius River; the cities of Sesamus, Cromnos, and Cytorus (founded by Cytisorus, the son of Phrixus); then Cinolus, Collyris, and Armene, which marks the end of Paphlagonia.” (Pomp. Mel. 1. 104 [Romer 1998]).
Arrian, who lived some time between AD 86 and 180, provides precise distances between various places in his Periplus Ponti Euxini and discusses whether a suitable harbour is present: — “From here onwards is Paphlagonia. From the Parthenios to the Greek city of Amastris is 90 stades; a mooring for the ships. From there to Erythinoi, 60. And from Erythinoi to Kromna, another 60. From there to Kyotoros, 90; there is a mooring for ships in Kyotoros. And from Kyotoros to Aigialoi, 60. To Thymena, 90. And to Karambis, 120. From there to Zephrion, 60. From Zephyrion to Abonouteichos, a little town, 150.” (Arr. Peripl. M. Eux. 14 [Liddle 2003]).
40
Traveller’s Notes on the Cide Region through the Ages
Finally, Claudius Ptolemy who lived between ca. AD 90 and 168, proposed coordinates for places such as Cromna, Cytorum, and Carambis (Ptol. 5. 6.). Another type of reference to the Cide region revolves around the apparently much appreciated quality of the boxwood from the area, a very hard and durable material fashioned into spoons, boxes, waxboards, and combs. For example, the Greek author Theophrastus (371-287 BC), who was interested in biology, physics, ethics, metaphysics, and botany, noted that great quantities of Buxus grew in the Cide region (Thphr. Hist. Pl. 1. 10). Pliny notes the same (Pl. Nat. 16. 3. 8.), as does Virgil (Verg. G. 2. 426-436). Ovid (43 BC –AD 17) in his description of the nymph Salmacis characterises her as decadent, infatuated with her own beauty and combing her hair with a comb made of boxwood – apparently a much sought-after item (Ov. Met. 4. 305-312). He also mentions a spool made from boxwood from Cytorus with which Arachne is struck (Ov. Met. 4. 127-133). Finally, another aspect of the Cide region that is mentioned frequently relates to its role in shipbuilding. Apparently ships from the area were highly coveted, as can be deduced from the following lyrical description by Catullus, who lived from 84 to 54 BC: — “My friends, the yacht you see before you claims to have been the swiftest of ships, and that she could outstrip in speed any bark, whether flying by oar or sail. This she defies the threatening Adriatic’s shores to deny, or the Cyclad islands, or far-famed Rhodes, or inhospitable Thrace or Propontic and Euxine’s storm-swept bay, where what is now my yacht grew up a leafy tree: for on the Cytorian heights the winds oft whistled through her rustling leaves. Pontic Amastris and box-clad Cyotorus, to you my yacht declares these truths to be known, that her birth was on your summit, that in your waters she first dipped her oars; thence over the wild seas she in safety bore her master, whether the wind drove on the port or starboard tack, or whether a favouring gale tautened both her sheets at the same time…” (Catu. 4. 10-13).
Summarising this brief overview of Greek and Roman sources mentioning the Cide region, the following types of relevant information can be extracted. First, the geographical descriptions inform us about the names and locations of ancient settlements – most of which can be located in the modern landscape – and the distances between them. They also provide information of nautical interest that would have been useful for navigators. Second, the sources provide us with information on the types of landscapes that ancient voyagers encountered, for example that the Cide region was densely forested and that it was famous for its high quality boxwood. Although this is not mentioned in the texts, we may assume that the production of boxwood artefacts would have been a local industry, one which persisted until the sub-recent period in the Cide region (chapter 2). Third, we learn from the texts that Cide was home to a shipbuilding industry, which is again something that is well documented in later periods in the same region (see below, also chapter 2).
The Cide Region in the Ottoman Period
41
3.2 The Cide Region in the Ottoman Period Following the last mentions of the Cide region in AD 400, there is a gap of a full millennium before we can pick up the thread of travel and other reports. This is not to say that there are no textual mentions of places in the Cide region (see e.g. Bryer and Winfield 1985; Belke 1996; Drakoulis 2013), but the information they provide is relatively limited. They allow the identification of a few of the main place names such as Cytorus and Carambis, but little else (Kalinka 1933; Belke 1996). As we have mentioned in the introduction, for reasons of expediency we have focused this discussion on easily accessible sources and do not claim comprehensiveness. Most likely additional information on the Cide region could be found in Byzantine, Genoese, and Venetian sources, which would fill much of our gap between AD 400 and 1400. However, this would require the inspection of various archives and considerable language skills, and therefore falls outside the scope of this chapter (but see chapters 11 and 12). The earliest post-Byzantine report at our disposal dates to the early 15th century AD, when the Cide region was under the control of the Candaroğlu principality (also known as İsfendiyaroğlu), before the Ottoman empire took control in AD 1461 and annexed this Black Sea state (Kortepeter 1966: 90; Yücel 1980; see also chapter 13). Subsequent sources date to the 17th, 18th and 19th centuries AD. Over time, an increase in both the number and the degree of detail of travel reports written predominantly by English and French travellers can be observed, mirroring the rise of these states as the paramount world empires in the industrial age. Some might view the travel reports as part of a broader quest for encyclopedic knowledge typical of the age (Westenholz 2006), but as Edward Saïd (2003 [1978]) famously argued, this process of gathering information also served imperial aspirations of control and exploitation. While his thesis has been challenged from various sides, some of the travellers’ reports discussed here certainly support Saïd’s argument. For example, many of the travels were undertaken under the auspices and instruction of the governments of Britain and France. Likewise, the detailed descriptions of geology, geography, (military) industries, and populations do not always make for interesting reading, but would certainly have been useful if these regions had come under direct or indirect colonial control, as was the case for other parts of the Ottoman empire after 1917. Ruy Gonzale De Clavijo travelled by order of the Spanish king to the court of Timurlane in Samarkand between AD 1403 and 1406 and passed by Cide on the way: — “They remained that day, and Thursday, and on Good Friday they got under way, and reached a port called ‘the two castles”, at the hour of vespers. On Saturday they went on, and there was a dense fog. Afterwards there was a fresh breeze, and the sea rose, and made great waves: and they knew not whether they were near or far from land.” (De Clavijo 2010: 57).
42
Traveller’s Notes on the Cide Region through the Ages
While De Clavijo does not mention Cide here by name, from the itinerary – he came from Amasra - it is clear that the harbour described here is that of Gideros, which has Byzantine fortifications on both sides of the entrance (chapter 11). The Ottoman traveller Evliya Çelebi (AD 1611-1682) published extensively on his many travels in the Ottoman lands, and in his Seyahatname (book of travels) there is the following on the Cide region: — “From there we crossed the river Kıyu, which forms the boundary between the provinces of Bolu and Kastamonu. From there it is 40 miles to the harbour of Geduz. From there we went to the cape of Kerçe through the Kerence harbour and on, which is a narrow cape like Sinop. It is 7 miles long. Inscriptions are carved on its rocks which are worth mentioning.” (Çelebi 1970: 74, translated from the Turkish by authors).
Here Geduz is almost certainly to be equated with Gideros and Kerçe with Carambis. At the start of the 18th century, the Frenchmen Joseph Pitton de Tournefort (AD 1656-1708) travelled in the same region. He arrived in Istanbul in 1701 and travelled across the Black Sea in 1702 and on to the Caucasus. He describes the Turkish western Black Sea region in some detail: — “We will leave the city of Amastro, in the hands of the Turks, and pursue our journey. The 4th of May we left the river Sita, which I find neither in Maps nor Books: We went but 30 Miles further, and the North-wind obliged us to encamp on a wretched flat Shore,2 where we had much ado to defend ourselves from the Wind. The 5th of May we doubled Cape Pisello, which the Ancients knew under the name of Carambis, and which they make to be just opposite to the Ram’s Head of the Chersonesus Taurica, now called the Little or Crim Tartary. The Ancients, as Strabo observes, compared the Black Sea to a bow bent, the String being represented by the South Coast, which would be almost in a straight line, but for Cape Pisello.” (de Tournefort 2003: 36).
Here, de Tournefort explicitly links his observation with those of Greek and Roman authors. Also of interest is that cape Carambis is given an otherwise undocumented name, which sounds Italian and would be a testimony to the centuries of Genoese and Venetian maritime trade in and domination of the Black Sea (also chapter 12). de Tournefort then goes on to describe the town of İnebolu and the shipbuilding industry of the region and its importance to the Ottoman navy: — “That Day, the 5th, we made but 50 Miles, and encamp’d on the Banks of the Sea at Abono, where are nothing but wretched Cazerns for a great number of Workmen that are employed in making Cordage for the Grand Signior’s Ships and Gallies. I forgot to mention that the Coasts of the Black Sea furnish in abundance every thing necessary for stocking the Arsenals, Magazines, and Ports of that Emperor. As they are cover’d with Forests and Villages, the Inhabitants are oblig’d to cut and saw the Wood for the Navy. Some make the Nails, others the Sails, the Cables, and all the necessary Rigging. There are Janizaries to oversee these Workmen, and Commissioners to press Seaman. From hence the Sultans have had their strongest Fleets in the time of their
2 This is probably the coastal plain of Cide.
The Cide Region in the Ottoman Period
43
Conquests, and nothing would be easier than to restore their Navy. The Country is excellent; it abounds in Provisions, as Corn, Rice, Meat, Butter, Cheese, and the people live very soberly.” (de Tournefort 2003: 36-7).
More than a century later, James Renell also notes the importance of shipbuilding in the area: “Cytorus, now Kadarôs or Ghidrus, lies to the east of Amâsarah. This is a great place of export of ship timber; and at Kara-Agatch, in its neighbourhood, ships of war are built, and cordage and sails are made.” (Rennell 1831: 118).
Like de Tournefort, his fellow countryman Jean Baptiste Bourguignon d’Anville (AD 1697-1782) was interested in bringing together the ancient and modern topography of the Cide region: — “Cytorus is recognised in the modern name of Kudros; beyond which position the most important object is the promontory of Carambis, whose name is perpetuated in that of Kerembi: and, in describing the Tauric Chersonese, we have said that this is directly opposite the Criu-metopon of that head. Abonitichos, which was also called Ionopolis, retains this last-mentioned name in that of Ainehboli.” (D’Anville 1814: 298).
The English traveller William Martin Leake (AD 1777–1860), who worked for the Ottoman army, arrived in the Cide region in 1824 and also linked ancient toponyms with the places he encountered: — “On the sea-coast all the most important sites of antiquity are determined by the actual names. These sites are […] Bartan, in Greek Parthéni; Amastris, Amásera; Cytorus, Kidros; Thymena, Temena; Carambis, Kerempe […]. With these data it will not be difficult for the future traveller to fix the intermediate names of the three Periplus: especially as existing vestiges of antiquity, and the rivers which form a large proportion of the ancient names, will greatly facilitate the task.” (Leake 1824: 307).
Similarly, the Frenchman P. Amaedee Jaupert (AD 1779-1847), who passed by the Cide region in 1806, combined modern and ancient geography, noting the properties of Cape Carambis and Gideros harbour in some detail: — “Nous voguâmes jusqu’au soir, et alors nous primes terre entre les roches qui forment le redoutable cap Kerempéh que nous nous disposâmes à doubler le lendemain matin de bonne heure, ce que nous fimes. A partir de ce cap jusqu’à Kidros, l’ancienne Kithoros de Paphlogonie, c’està-dire sur un espace de douze à treize lieues, la calme nous permit de faire usage de la rame, et nous passâmes, sans nous arrêter, devant Temenéh, l’ancienne Thymène, et devant Cara-Agadj (l’arbre noir), position qui semble convenir à celle d’Ægiali, dont il est fait mention dans le Périple d’Arrien. Le 18 au soir nous jetâmes l’ancre dans le port de Kidros, que de hautes montagnes couronnées de forêts ferment presque de toutes part, et dans lequel on est à l’abri de tous les orages. Il est surprenant qu’on ne trouve dans ce lieu paisible aucune autre habitation que quelques masures qui suffisent à peine pour mettre à couvert de la pluie.” (Jaupert 1821: 401-402).
44
Traveller’s Notes on the Cide Region through the Ages
Yet another traveller who took up the same theme is the Frenchman Charles Texier (AD 1802-1871), who notes: — “Nous retrouvons sur la côte le cap Carambis, qui porte aujourd’hui le nom Kérembé. Les trois petites villes, Sesame, Cytorus et Cromna, concoururent à former la population d’Amastris….” (Texier 1862: 622).
In contrast to the above authors who were mainly interested in historical geography, the Ottoman author P. Minas Bijişkyan (AD ?-1851) travelled along the Turkish Black Sea coast in 1817-19 and made many interesting observations about the places and people he encountered: — “Kidros, or Kitoron, is a town 18 miles from Tekkeönü. There is a small islet in front of the harbour which is suitable for ships. Much timber is harvested in the surroundings and ships are built here. On the coast there are two old castles on its western and eastern sides. Wood for rafts is brought here from Salderesi, which is 4 miles away. Karaağaç is 18 miles away from Kidros. On the coast there are shops and there is a stream, but the town is located on higher grounds and is not an important one. This area is densely forested and all the mountains are green. Arrian mentions the existence of a town called Timena here which was inhabited by an old people called the Aegiali. Kerembe (Karamvis Akra in Greek, meaning Cape Karambis) in 36 miles from Karaağaç, a tall and extended cape located opposite the Crimea. Because there is no harbour, this is a very dangerous place for ships at the time of strong storms. The cape starts from Kidros stretching into the sea and ships can only halt on the western side with some difficulty. In the past this was an inhabited place. Pliny mentions that storks crossed to the other side from this cape because it was close. Indeed, all species of birds cross to Crimea from here and often fall into boats as they do so.” (Bijişkyan 1969: 25-6, translated from the Turkish by authors).
Bijişkyan then continues his description of the Black Sea coast further east, noting for example that İnebolu had a Christian quarter, two churches and an old monastery. Given that he does not make similar observations for the Cide region, this may imply that it had a more limited Christian population. Bijişkyan, like de Tournefort, also remarks on the importance of the shipbuilding industry in the region. A very similar description of the region is given by Eduard Taitbout de Marigny who travelled in the region in AD 1830 and notes the following: — “At 30 miles E. By N. of Amastris, to the east of a high mountain, called Tepehfoulla, or Mount Segra, and at the foot of another, in the form of a sugar-loaf, is the anchorage of Kitross or Kidross. It is small, and the entrance is obstructed by a rock, which defends a few vessels from the swell of the sea; but it is exposed to the north winds. To avoid the rock going in, vessels should range along the east shore. Kara Agatch lies 5 miles to the eastward of Kidross. The village is small; but vessels of considerable size are built there, and at different small villages along the coast. The anchorage is open, but used by coasting vessels. Cape Kerempeh: this cape is one of the highest in the Black Sea, and therefore, the most easily made by vessels coming from the Crimea. From
The Cide Region in the Ottoman Period
45
Cape Saritche, the southernmost point of that peninsula, the distance is 130 miles; and it is said, in clear weather, they have both been seen at the same time. These two promontories divide the Black Sea into two parts, the east and the west; often very distinct by the different winds blowing at the same time in each. Cape Kerempeh deserves the name of Spartivento (separator of winds) which has been given by the Italians to several capes in the Mediterranean; for a strife between the winds is very often observed abreast of it, forming a line of separation, very distinct.” (de Marigny 1847: 33).
The French traveller Xavier Hommaire de Hell (AD 1812-1848) undertook a trip to Turkey and Persia for the French government. His account was published in three volumes, including detailed images and descriptions. On the Cide region he notes: — “19 juillet. Nous reprenons la mer devenue houleuse sous un fort vent de N.O. et dont la violence, au bout d’une heure, nous force à entrer dans le petit port de Kidros, où se refugient d’ordinaire les barques surprises par le mauvais temps. Au sortir de Kidros (Cide) en passant près de Kara Agati Bournou (à la sortie de Cide) force est de nous réfugier derrière un petit cap où il y a juste assez de place pour nous abriter. [..] Le vent s’étant calmé, nous essayons quelques bordées. [..] mais ces efforts ont peu de succès et nous devons choisir notre campement au milieu de rochers éboulés. Après une nuit orageuse, le calme renaît sur le matin, et nous en profitons pour reprendre la mer. Une bonne brise nous conduit à Aidos-Skelessi, située à l’entrée d’une vallée parfaitement cultivée. De hautes montagnes la dominent. Cette échelle possède un chantier de construction. Forcés par le vent contraire de rétrograder, nous allons prentre terre à Koumé-Sélessi, où nous trouvons toute la population en émoi, assistant à la mise à flot d’une grande barque chargée de bois et d’ouvriers destinés à l’arsenal de Constantinople. Au moment où nous arrivons, une centaine d’hommes entourent la barque qui se trouve à terre, assez loin du rivage; mais grâce à un gros cabestan, et à la quantité de bras qui agissent, la barque est bientôt hissée sur un plan incliné, à deux pas de la mer, et, dès que l’équipage se trouve à bord, le capitaine coupe avec la hache le cable qui la retient…Au bout d’une demi-heure, elle disparaît derrière le cap. [..] Bientôt nous sommes en face du célèbre cap de Kerembeh, signalé par M. Taitbout de Marigny comme un promontoire gigantesque, mais qui n’offre en réalité qu’un méchant petit cap, d’une trentaine de mètres au plus d’élévation. D’autres le suivent, parmi lesquels celui de FakasBournou. Tous ces caps se rattachent à une chaîne de médiocre hauteur qui borne cette côte… Nous passons devant Fakas Skelessi, village d’une soixantaine de maisons où l’on construit jusqu’à des bricks. Puis viennent le cap et l’échelle de Mezet, Kaïran Skelessi, Zarpana Skelessi, avec une grande plage, une belle vallée et un fond de hautes montagnes; et enfin Inéboli, où nous débarquons.” (Hommaire de Hell 1854: 334-8).
Hommaire de Hell also provides extensive notes on the geology of the Cide region based on his limited explorations of the area, and one cannot help but wonder what purpose this geological data was intended to serve. Hommaire de Hell also provides us with a vivid account of the vagaries of navigation in the Black Sea in the 19th century AD and mentions several places not noted by other travellers: Aydos, Koumé
46
Traveller’s Notes on the Cide Region through the Ages
(probably present Timle/Uğurlu), and Fakas (probably İlyasbey). Finally, Hommaire de Hell vividly describes how a boat with timber and other supplies for the Ottoman navy in Istanbul is being loaded and departs, again attesting to the importance of ship industries in the area. Another detailed report on the Turkish western Black Sea region was provided by Eugene Bore (AD 1809-1878). Unlike any of the other reports, Bore’s prose is that of novelist in the sense that he describes landscapes, buildings, and people in articulate detail. Like other western travellers he was interested in the region’s past, but he is the only one who actually reports on the exploration of the material remains of this past: — “Levés avec le soleil nous allâmes suivis d’un serviteur du bey examiner ce qui reste de Cytore appelée Kydros par les Turcs. Nous trouvâmes les vestiges de la tour récemment démolie et à travers l’épaisse lisière d’un taillis bordant la colline élevée au dessus de la mer nous reconnûmes le prolongement d’un mur, qui devait être la muraille d’enceinte. De l’autre côté du port, étroit et mal défendu contre les ouragans du Nord, se montrent les ruines d’une forteresse. Une ou deux barques à l’ancre représentaient l’état d’abaissement auquel est tombé le commerce du lieu. Anciennement, des montagnes environnantes qu’ombrageaient des forêts de pins, de chênes et de hêtres, la marine ottomane tirait les bois de ses flottes; mais la mine s’est épuisée; et l’incurie de l’administration qui n’a percé aucune route, rend inutiles les autres forêts, plus éloignées dans l’intérieur des terres. Kydros n’a pas profité de la position maritime de Cytore. Cette bourgade est à une lieue environ du rivage. Derrière la maison du bey apparaît une montagne, dont le flanc est percé de cavernes. Ces excavations étaient un genre de monument religieux et funéraire, qui se multipliera à nos yeux sous, des formes variées et quelquefois d’une hardiesse prodigieuse, à mesure que nous pénétrerons dans l’Asie. Peut être était-ce le pic qui portait autrefois le même nom que la ville, suivant la remarque de Pline.” (Bore 1840: 243).
Bore then proceeds to a description of his attempts to muster new sailors, which proved difficult because the villagers were engaged in harvesting their fields, some bad Black Sea weather, and their subsequent arrival in the picturesque bay of ‘Delikalsali’, a location which we cannot pinpoint. Bore describes Gideros in some detail, including the nature of the fortifications, a tower no longer extant today, the former importance of the place to naval industry, and an associated mine. He also mentions that the real town of Cide lay a mile inland, possibly at the current location of Abdulkadir village. He also mentions many caves in the mountains framing the bay that were used for religious and funerary purposes. We were able to locate some of these caves mentioned by Bore in the course of our survey (chapters 11 and 14). The final report on the Cide region which we will discuss here is that by Vital Cuinet who undertook a census of the Ottoman empire towards the end of the 19th century AD in order to assess whether it would be able to pay its debts. He provides astonishingly detailed information on the region, although the accuracy of his figures is difficult to assess. It is possible that he copied Ottoman records on this occasion (see chapter 13). He also offers lengthy observations on the local economy and its population. Regarding hunting he notes:
The Cide Region in the Ottoman Period
47
— “Les animaux sauvages qui fréquentent ces forêts [de Kastamonu] sont principalement les loups, les renards et les lièvres, qui donnent lieu à un commerce assez important de peaux pour fourrures.” (Cuinet 1894: 426). — “On doit cependant mentionner un fait particulier assez intéressant. Les tempêtes et les pluies d’automne jettent sur les côtes de Djiddé et d’Inéboli une multitude innombrable de cailles venant de Crimée. La population de ces deux cazas va les rammaser la nuit à la lueur des torches; on en tire un bon bénéfice en les vendant de 30 à 40 paras Tune (17 à 23 centimes), et l’on exporte un grande quantité. — Ces contrées sont d’ailleures extrêmement giboyeuses, comme le savent bien les chasseurs européens en résidence à Constantinople, qui, fort souvent, viennent, dans les parages de Djiddé et d’Inéboli, oragniser des battues à l’ours et au sanglier.” (Cuinet 1894: 452).
Thus according to Cuinet, hunting was of considerable importance to the local economy in the late 19th century AD, which is remarkable given that it is not very common at present. Cuinet (1894: 490-2) also notes that the district of Cide, which was then larger, encompassed 36 villages in which 40,306 people lived in 8,060 houses, 31 of which were Christians. The town of Cide had 2,510 inhabitants, of which 31 were Christians (thus all Christians lived in the main town). He also notes that there were 42 primary schools with 960 pupils, and one secondary school with 121 students. Further, the district had 55 mosques, two dervish tekke (monasteries), public baths, and no less than 121 mills. The region also produced a large amount of agricultural produce. Cuinet mentions the importance of lumber for the local economy and provides details on a major mine presently no longer in operation: “Mines et minières. – Kidros, village maritime situé à 10 kil. ouest de Djiddè, dont il est séparé par le Dévrikan-Irmaghi, est bâti sur les ruines de l’antique vilel de Cytorus. Entre la montagne et le village se trouve un mine de houille d’une grande richese et qui produit un charbon comparable en quantité aux meilleurs charbons de houille de l’Angleterre. La concession de cette mine à été accordée par décret impérial au géneral de division Vitaluis Pacha, ancien officier supérieur français de grand mérite.” (Cuinet 1894: 492).
Cuinet, thus, provides us with much detail on the local economy at the end of the 19th century AD. Interestingly, he does not mention the shipbuilding industry focused on by other authors. At the same time, the coal mining near Gideros that Bore reports as having ceased some decades earlier, is described as an important economic activity by Cuinet. Was the mine reopened subsequently? Or is Cuinet’s information not completely reliable? Also of interest are the 121 mills in the area that attest to a scale of agricultural production no longer evident today (chapter 2) but that is corroborated by the remains of mills we encountered throughout the region during fieldwork (chapter 14). In summary, we can distinguish between several different types of Ottoman period accounts of Cide. The earliest reports by De Clavijo and Evliya Çelebi list landmarks and towns that were encountered on their travels. Later travellers, including de
48
Traveller’s Notes on the Cide Region through the Ages
Tournefort, d’Anville, Leake, Jaupert, and Texier also discuss these landmarks and towns and often try to link them with the toponyms mentioned in Greek and Roman sources. Finally, many 18th and 19th century travellers provide increasing amounts of detail about the industries, resources and people they encountered in the Cide region. De Tournefort discusses the importance of shipbuilding in the area and the region’s agricultural economy. Shipbuilding in particular is mentioned also by Bijişkan, de Marigny, and Hommaire de Hell. By contrast, Cuinet focuses on the value of natural resources, such as game, lumber and coal, but does not mention local industries (table 3.1). As stated at the start of this section, these later works, which contain very detailed assessments of the local economy and its resources, are not simply to be understood as travel reports but also reflect the growing interests of European empires in the region. In particular the work of Cuinet is clearly motivated by imperial interests. Tab. 3.1: Economic activities in the Cide region mentioned in travellers’ accounts. agriculture de Tournefort
lumber
x
Bijişkyan
hunting
mining
x
x
x x
de Marigny Hommaire De Hell Cuinet
shipbuilding x x
x
x x
x
3.3 Conclusion The information provided by the Greek, Roman and Ottoman period sources on the Cide region cover a period of about 2500 years. The earliest mention of the region is in the Iliad, which has to be dated to the later part of the Iron Age. The most informative sources, however, provide important information on the society and economy of the region in the periods between ca. 400 BC and AD 400 and ca. AD 1400 and 1900. In the Classical era, Kytoros was a relatively minor town when compared to others in northern Turkey, such as Heraclea, Amasra and Sinop. At the same time, the region receives frequent mentions, owing mainly to its importance to maritime travel in the Black Sea. Greek and Roman sources, thus, mention primarily landmarks and towns of the region and distances between them and the region’s most widely appreciated resources. The Cide region also features in earlier Greek stories, such as the Trojan War and the quest for the Golden Fleece by the Argonauts, but given that the first Greeks arrived in the area only in the seventh century BC (West 2003; also chapter 8), these
Conclusion
49
sources should be considered as relatively late elaborations with limited value for understanding the region in earlier periods. Of particular interest with regards to both sets of sources is the remarkable degree of continuity in place names and the locations of towns that they attest to (Fig. 3.1). We encounter the harbour Gideros described as two castles and under the names of Geduz, Kytoros, and Kydros, the major cape of Kerempe Burnu as Carambis, Keremhi, Kerempeh, the town of Thymena as Time and Temeneh, which is called Denizkonak today. Aegialos was located where Cide town is today, which was known in the Ottoman period as Karaağaç (‘elm tree’ in Turkish) and changed to Cide, probably a derivate of Gideros, sometime in the late 19th century AD. In the sources from the Ottoman period we also find many toponyms which continue to be used. Some of these have Greek rather than Turkish origins. Examples include the Aydos river, Timle which we find as Gümle, Koumi Pier which is Uğurlu today, Fakaz, probably at İlyasbey. Finally, while working in the Cide and Şenpazar districts we often came across places with seemingly Greek names, such as Malyas, Terme, and Muna. In the 1960s many village names were altered. For example, Timle became Uğurlu, Loç became Çamdibi, Güble became Çayyakka, Terme became Gökceören, Malyas became Gökçeler, and Gürcü became Bağyurdu (Anonymous 1970). Interestingly, in many cases, local residents continue to use the old names and the official names on maps are often unknown to them. It would also appear that for some time, Italian toponyms were in use for local towns and landmarks, most likely a reflection of the Geneose and Venetian presence in the Black Sea (Di Cosmo 2010). In particular, de Tournefort reports that Cape Carambis was also called Cape Pisello and İnebolu, Classical Ionopolis, was called Abono. In later periods, however, Italian toponyms disappear completely from written sources. Although there is a remarkable degree of continuity in the locations of towns and place names, there are also remarkable absences. In particular the locations and names of many Byzantine period towns and castles are not attested in earlier sources. These include prominent sites and landmarks, often easily spotted from the sea, such as Okçular Kale (S24), Hıdır Kale (S62), Kazallı Kale (S91), Aydos Kale (S190), and Çoban Kalesi (S32), which indicates that these sites, by and large, postdate AD 400. Local industries and resource exploitation too shows much continuity. One widely sought after commodity is boxwood (Turkish şimşir), which was fashioned into objects such as combs, spoons, spools, writing boards, and boxes and exported far beyond the Cide region. This industry was present in antiquity and persisted into sub-recent periods before it was prohibited for preservation purposes. Another local industry featuring in Greek, Roman and Ottoman sources is shipbuilding. In antiquity, high-quality ships were constructed here, and in Ottoman times the region was of great significance as a production centre for the imperial navy. This tradition of shipbuilding persisted until recently, when major shipyards were relocated from the region to Istanbul (chapter 2). Further, Ottoman period sources in particular show that the Cide region, unlike today, had a burgeoning agricultural economy, something borne out also by the oral accounts of senior residents in the region (also chapter 2).
50
Traveller’s Notes on the Cide Region through the Ages
Fig. 3.1: Map of the Cide and Şenpazar region showing the main locations of towns and landmarks discussed in this chapter. Produced by Toby C. Wilkinson.
Equally interesting are the activities not mentioned by any of these sources. Fishing and smoking fish for export which is attested elsewhere along the Black Sea (Madsen 2009), for instance, does not seem to have been a mainstay of the local economy. Likewise, hunting for fur, a major industry in many parts of the world in the early modern period (Lightfoot 2005), was not important despite the presence of large amounts of game in the region (Cuinet 1894: 426, 452). In conclusion, the reports on the Cide region by travellers through the ages present us with some remarkable continuities in settlement locations, place names, and industries. At the same time, the lack of certain information is also interesting, suggesting for example that the numerous castles in the area mainly postdate the Roman era and that fishing was not an important industry. Such insights, together with those from more recent periods discussed in chapter 2, provided an excellent starting point for our archaeological assessment of the Cide and Şenpazar region.
Bibliography Greek and Roman sources Ammianus Marcellinus, Rerum Gestarum Libri. [Translated by J. C. Rolfe]. Harvard University Press, Cambridge Mas., 1935-1940.
Bibliography
51
Apollonius of Rhodes, Argonautica. [Translated by R. C. Seaton]. London and New York, 1919. Arrian, Periplus, Ponti Euxini. [Translated and annotated by A. Liddle]. Bristol Classical Press, London, 2003. Catullus, Carmina. [Edited by C. Stuttaford]. G. Bell and Sons, London, 1912. Claudius Ptolemy, Geographia. [Translated by E. L. Stevanson]. New York, 1932. Homer, Illiad. [Translation by A.T. Murray]. Harvard University Press, Cambridge, M.A., 1924. Pliny, Naturalis Historiae. [Translated by J. Bostock]. H. T. Riley, London, 1955. Pomponius Mela, De Chorographia. [Translated and annotated by F.E. Romer]. The University of Michigan Press, Ann Arbor, 1998. Pseudo Skylax, Periplus. [Edited by G. Shipley]. University of Exeter Press, Exeter, 2002. Ovid, Metamorphoses. [Translated by A.S. Kline]. Borders Classics, New York, 2004. Strabo, Geographica. [Translated by H.C. Hamilton and W. Falconer]. London, G. Bell & Sons, 1903. Theophrastus, Historia Plantarum. [Translated by I. B. Stapel], Henricum Laudentium, Amsterdam, 1644. Valerius Flaccus, Argonautica. [Translated by J. H. Mozley]. Loen Classical Library No 286, Harvard University Press, Cambridge MA, 1934. Virgil, Georgics. [Edited by J. B. Greenough]. Ginn & Co., Boston, 1900. published sources Anonymous 1970. Türkiye Mülki İdare Bölümleri ve Bunlar Bağli Köyler Belediyer. Ankara, Türk Cümhüriyeti İçişleri Bakanlığı. Avram, A., Hind, J., and Tsetskhladze, G. 2004. The Black Sea area. In M. H. Hansen and N. T. Heine, eds., An Inventory of Archaic and Classical Poleis. Oxford, Oxford University Press: 924-28; 954-64. Belke, K. 1996. Paphlagonien und Honorias. Vienna, Verlag der Österreichische Akademie der Wissenschaften. Bijişkyan, P.M. 1969. Karadeniz Kıyıları Tarih ve Coğrafyası 1817-1819.[Çeviren Hrand D. Andreasyan]. İstanbul, İstanbul Üniversitesi. Bore, E. 1840. Correspondence et mémoires d’un voyageur en Orient. Tome I. Paris, Olivier Fulgence. Bryer, A., and Winfield, D. 1985. The Byzantine Monuments and Topography of the Pontos. Washington DC, Dumbarton Oaks. Çelebi, E. 1970. Evliya Çelebi Seyahatnamesi Cilt 3. [Çeviren Z. Danışman]. Crow, J., and Hill, S. 1995. The Byzantine fortifications of Amastris in Paphlagonia. Anatolian Studies 45: 251-65. Cuinet, V. 1894. La Turquie d’Asie. Géographie administrative, statistique descriptive et raisonnée de chaque province de l’Asie-Mineure. Tome Quatrième. Paris, E. Leroux. D’Anville, J.B.B. 1814 [1791]. Compendium of Ancient Geography by Monsiur D’Anville. [Translated from the French by L. J. Horsley]. New York, R. McDermut and D.D. Arden. de Clavijo, R.G. 2010 [1859]. Narrative of the Embassy of Ruy González de Clavijo to the Court of Timur at Samarkand AD 1403-1406 [Translated by Clements R. Markhom], Cambridge, Cambridge University Press. de Marigny, E.T. 1847. New Sailing Directions for the Dardanelles, Sea of Marmara, Bosphorus, Black Sea and the Sea of Azov. London, Charles Wilson. de Tournefort, J.T. 2003 [1741]. A Voyage into the Levant Vol. III. [Translated by J. Ozell]. Farmington Hills, Thomson Gale. Di Cosmo, N. 2010. Black Sea emporia and the Mongol empire: A reassessment of the pax mongolica. Journal of the Economic and Social History of the Orient 53: 83-108.
52
Traveller’s Notes on the Cide Region through the Ages
Drakoulis, D.P. 2012. Regional transformations and the settlement network of the coastal Pontic provinces in the Early Byzantine period. In G. Tsetskhladze, ed., The Black Sea, Paphlagonia, Pontus and Phrygia in Antiquity. Aspects of Archaeology and Ancient History. Oxford, Archaeopress: 79-95. Heiden, B. 2008. Homer’s Cosmic Fabrication: Choice and Design in the Iliad. New York, Oxford University Press. Hommaire de Hell, X. 1854. Voyage en Turquie et en Perse exécuté par ordre du gouvernement français. Paris, Bertrand. Jaupert, P.A.E.P. 1821. Voyage en Armenie et en Perse 1805-1806. Paris, Chez Pélicier et Nepveu. Kalinka, E. 1933. Aus Bithynien und Umgebung. Jahreshefte des Österreichischen Archäologischen Institutes in Wien 28 (Beiblatt): 45-111. Kortepeter, C.M. 1966. Ottoman imperial policy and the economy of the Black Sea region in the sixteenth century. Journal of the American Oriental Society 86: 86-113. Leake, W.M. 1824. Journal of a Tour in Asia Minor with Comparative Remarks on the Ancient and Modern Geography of that Country. London, John Murray. Lightfoot, K.G. 2005. Indians, Missionaries, and Merchants. The Legacy of Colonial Encounters on the California Frontiers. Berkeley, University of California Press. Madsen, J.M. 2009. Eager to be Roman: Greek Responses to Roman Rule in Pontus and Bithynia. London, Duckworth. Rennell, J. 1831. A Treatise on the Comparative Geography of Western Asia. London, Rivington. Saïd, E.W. 2003 [1978]. Orientalism. London, Penguin. Texier, C. 1862. Asie mineure: description géographique, historique et archéologique des provinces et des villes de la Chersonnèse d'Asie. Paris, Firmin Didot Frères. West, S. 2003. The most marvellous of all seas: The Greek encounter with the Euxine. Greece & Rome 50: 151-67. Westenholz, A. 2006. Does Assyriology have a future? Kaskal 3: 275-83. Yücel, Y. 1980. Çoban-oğullları, Candar-oğulları Beylikleri. Ankara, Türk Tarih Kurumu.
Claudia Glatz, Bleda S. Düring and Toby C. Wilkinson, with contributions by Bernard Gratuze, Richard Jones, Effie Photos-Jones, and Victor Klinkenberg
4 Developing an Adaptive Field Methodology for Challenging Landscapes “In reality, there is no single formula for the maximum recovery of early landscape data. Rather, it is necessary to harness a combination of remote-sensing techniques, detailed field-survey, and local information, together with a certain amount of luck.” (Wilkinson 2003: 43).
The Turkish western Black Sea coast includes some of the most challenging landscapes for survey archaeology in Turkey. As has been described in detail in chapter 2, much of the region is dominated by steep, rugged and densely forested mountains. It is also affected by both natural and anthropogenic landscape transformations that influenced past subsistence and settlement strategies and that, importantly for our purposes, have significantly impacted archaeological site preservation and visibility (section 2.3). From the outset, one of the central research aims of the Cide Archaeological Project (CAP) was to find ways to identify and document the very ephemeral archaeological traces left by past communities in the landscapes of Cide and Şenpazar. The themes of ephemerality and visibility will recur throughout the following discussion on methodologies, and frequently in subsequent period-based and synthetic chapters, since dealing with both was a key challenge for and, we argue, a key strength of the survey methodology we developed. This chapter will describe in some detail the origin, development and execution of CAP’s sampling strategies, its methods of data collection and its associated postsurvey analysis. These were founded on well-established intensive and extensive archaeological survey methodologies (section 4.3), but with modifications or flexible adaptations to suit both the specific topographic and environmental conditions of the survey area and our research questions (chapter 1). A detailed outline of the factors and rationales guiding decisions on sampling strategies and the process of developing and optimising approaches and methods is important so that other researchers can evaluate our results and our interpretations as they are presented in this volume, as well as make use of our raw data, which is published online (http:// www.dans.knaw.nl/nl), with some knowledge of how these data were collected (see also Mattingly 2000: 5). This process also makes our results more easily comparable with past and future surveys and facilitates the incorporation of CAP data into synthetic studies of regional and supra-regional scope (Alcock and Cherry 2004). Finally, we hope that future archaeologists working in the Turkish Black Sea region, or similar environments elsewhere, might find our approach and the lessons learnt in the course of its development of use in devising their own research strategies.
54
Developing an Adaptive Field Methodology for Challenging Landscapes
4.1 Traditions of Archaeological Survey in Turkey and Beyond Working in Turkey means working at the interface of two well-established schools of regional archaeological studies, whose sampling strategies and field methods differ substantially in scope and form from each other. Broadly characterised, these are the extensive or ‘targeted reconnaissance’ school, which has long been a feature of research in (southeastern) Turkey and much of the Near East, and the intensive or ‘siteless’/‘continuous’ school, which was developed and applied particularly in the Aegean and Mediterranean. Whilst each was developed for landscapes very different from those we encountered in Cide and Şenpazar, their theoretical frameworks, methodologies, and the debates surrounding them cast long shadows onto the Black Sea coast. The following section outlines and assesses the strengths and weaknesses of these schools in general, before turning to their applicability to the Black Sea region and their incorporation, adaptation or rejection in CAP.
4.1.1 Extensive and Intensive Survey Traditions Extensive systematic archaeological survey of the Near Eastern type first emerged during the 1950s and 1960s. Methods included a combination of top-down reconnaissance (using topographic maps and aerial photographs), vehicle-based touring, and use of local knowledge to identify archaeological sites. First formalised in the alluvial plains of Mesopotamia (e.g. Adams 1965; 1981; Adams and Nissen 1972), this practical incarnation of contemporary ‘settlement archaeology’ was developed to answer ‘big questions’ about long-term socio-cultural developments in large regional units (often in the range of tens of thousands of square kilometres). Its investigative methods were designed for, and were very successful in, the flat or alluvial landscapes in which sites were normally easily recognisable at a distance, particularly in the form of mounded sites known as tells or, in Turkey, höyüks. Typically, such landscapes had a high degree of ground visibility (with little vegetation) so that during ‘groundtruthing’ of remotely identified sites, artefact densities stood in relative contrast to their surroundings and provided a clear indicator of human occupation (even if later agricultural practices may have distributed cultural material more widely in the surrounding landscape). Recent and innovative applications in this tradition have taken advantage of the ever-improving quality and quantity of remote sensing data (including high-resolution satellite imagery and remote radar/LiDAR topography data) and the capabilities of modern digital mapping/GIS for both manual or semiautomated site identification (Ur 2003; Menze et al. 2006; Casana et al. 2012). The first intensive ‘siteless’ or continuous surface surveys were developed by archaeologists working in the Aegean and eastern Mediterranean from the 1970s onwards (Cherry 1994). Typical methods emerged, in the first instance, from an attempt to apply principles of probabilistic sampling in order to provide a more representative
Traditions of Archaeological Survey in Turkey and Beyond
55
understanding of ancient landscapes. Such sampling was normally implemented through ‘tract-walking’, whereby teams of evenly spaced walkers record or collect archaeological material encountered along linear or zonal ‘tracts’. This pedestrian-led method allowed investigators to detect – in areas with relatively high ground visibility - small, inconspicuous surface scatters of cultural materials and to document the near continuous spread of pottery typical of many Mediterranean landscapes (Cherry 1983). Initially there was a strong concern with the collection of ‘random’ or rather statistically representative samples of archaeological assemblages from the surface to enable long-term demographic reconstructions and the investigation of land-use strategies (Banning 2002; Bintliff and Howard 2004). In recent years, landscape taphonomy has become an important methodological concern in Mediterranean surveys (Barker et al. 2007). Increasingly, emphasis is also shifting away from methodological and analytical concerns to the development of interpretive frameworks which tread a middle ground between rigorously assembled distribution maps and the desire to understand past human perception of, existence in, and interaction with particular landscapes (Given and Knapp 2003; Athanassapolous and Wandsnider 2004; Given 2013b). As in the Near Eastern tradition, digital mapping or GIS plays an ever more central role in the implementation and publication of intensive survey results. It also serves as an important interpretative tool that allows researchers to inspect their survey data at a range of spatial scales and temporal resolutions at the click of a mouse.
4.1.2 Regional Survey in Turkey In an ideal world, both the selection and size of a study region and the methods with which it is investigated should follow directly from, and be a function of, a project’s research questions. In practice, this is rarely the case and the selection of a research region and a project’s investigative methods are constrained by a variety of logistical and financial factors, as well as the official rules and procedures governing archaeological research in a national context. In Turkey, the traditional Near Eastern model of extensive survey informs official perception about the aims and outcomes of archaeological surface investigations, and government practice with regards to the issuing of research permits and their spatial and methodological scopes. As a result, the survey permits issued to both foreign and local archaeological teams reinforce or encourage a particular model of research. Customarily, permissions to carry out archaeological surveys are issued on the basis of single or multiple modern administrative units (a province, il, amounting to several thousands of square kilometres; or a district, ilçe, of several hundreds of square kilometres) on a more or less exclusive basis. Although this may change in the near future, this practice has resulted in the division of much of Turkey into geographically enormous survey areas that encourage highly extensive modes of research.
56
Developing an Adaptive Field Methodology for Challenging Landscapes
The framework of rules that regulates archaeological research in Turkey also has been in constant flux in recent years as the Ministry of Culture and Tourism attempts to reform the system. As a result, surveys (and excavation) projects rely to a large degree on the interpretation of these rules - either in spirit or letter - by the acting government representative (temsilçi). Although the temsilçi has a certain interpretive discretion as to what is or is not permitted, a range of intensive investigative methods that are routinely applied in combination with surface survey in other countries, such as test-pit excavations or large-scale vegetation clearances, are not within the remit of survey permits in Turkey. Regional scale archaeological projects have seen a steady increase over the past half-century in Turkey as elsewhere in the east Mediterranean and Near East.1 Whilst the Near Eastern model has remained dominant, the exact form, intensity and duration of projects range widely. Extensive forms of survey in Turkey have their roots in the work of von der Osten (1927), Burney (1956), Mellaart (1954; 1959), Seton-Williams (1954) and French (1965; 1967; 1969). These pioneers were followed by reconnaissance surveys of varying intensity conducted in advance of major dam construction projects, mostly in the southeast of the country (Özdoğan 1977; Whallon 1979). In more recent decades, a series of long-term, extensive reconnaissance surveys conducted over one or more modern administrative provinces have dramatically increased the number of known mound settlements, especially around the central Anatolian plateau (Efe 1997; Omura 2005; Özsait and Özsait 2010). Shorter-term extensive surveys, some of which also include intensive sampling components, geoarchaeological and environmental research were also conducted in recent years and have added considerably to our understanding of the occupation history of Anatolia and how this interplays with environmental factors (Marro et al. 1996; 1998; Kuzucuoğlu et al. 1997; Baird 2001; Doonan 2004: 35-9; Matthews 2009; Bikoulis 2012). A few intensive surface surveys, usually radiating out from already-studied major sites, have been undertaken. They include the surveys around Alişar Höyük (Branting 1996), Boğazköy (Czichon 2000), Gordion (Kealhofer 2005), Klazomenai (Ersoy and Koparal 2010), Komana (Erciyas and Sökmen 2010), Pessinus (Tsetskhladze et al. 2012) and Avkat.2 In extending a site-centred method to surrounding landscapes, these surveys have been successful in locating smaller sites and retrieving site densities that are considerably higher than the norm for traditional Anatolian surveys (Düring 2008: 36). Despite these efforts, however, Turkey as a whole remains relatively poorly surveyed archaeologically in comparison to neighbouring countries such as Bulgaria,
1 See the MAGIS project for an attempt to catalogue past survey projects over the Mediterranean: http://cgma.depauw.edu/MAGIS/ 2 https://www.princeton.edu/avkat/
Ambitions for CAP: A Hybrid and Adaptive Method
57
Cyprus, Greece, and Syria. Moreover, northern Turkey and the Black Sea coast in particular are among the least well-researched parts of the country. In addition to a scarcity of excavated sites in the region for all periods but particularly before the Classical eras (chapters 5 to 9 for more detailed discussions), surveys in this area have tended to follow the extensive model outlined above, with large survey territories and a focus on conspicuous sites. Pioneered by Burney (1956), on-going and now concluded long-term extensive projects in the central Black Sea region include Dönmez’s survey of the provinces of Sinop, Samsun and Amasya (Dönmez 1999); Kızıltan (1992), and Bilgi and colleagues in Samsun (Bilgi et al. 2004); Özsait’s survey in Samsun and Amasya (Özsait 2005; Özsait and Özsait 2010); and Sipahi in Çorum and Çankırı (Sipahi 2002). Smaller regions investigated with similar methods include the Gökırmak valley (DonceelVoûte 1977), the Sinop promontory (Işın 1998) and the Devrek region of Ereğli province (Karauğuz 2007). All of these projects have recorded many new sites and provided initial chronological assessments, but they have done so in preliminary reports only. With a few exceptions, such as the well-illustrated output from the Dönmez surveys, these reports tend to lack detailed presentation of collected material or distribution maps, making comparisons at the artefact level and the integration of their results into cross-regional syntheses difficult. Besides these projects, a small number of surveys, some of which operate in comparatively small research areas, have included more intensive approaches as well as palaeoenvironmental and geoarchaeological research. Examples include the central Kastamonu survey (Marro et al. 1996; 1998; Kuzucuoğlu et al. 1997), which relied on local guides for the detection of archaeological sites, but intensively surveyed 15 of the sites initially identified by extensive means (Özdoğan et al. 2000: 41-2). The on-going Sinop Regional Archaeological Project uses a combination of extensive and intensive survey methods adapted from Mediterranean and Near Eastern approaches and operates at several spatial scales (Doonan 2004). Project Paphlagonia pioneered a hybrid method of systematic extensive survey in a very large survey region (ca. 8,000km2) in Çankırı and Karabük provinces combined with intensive transect survey (Matthews and Glatz 2009), which in many ways directly inspired CAP (chapter 1).
4.2 Ambitions for CAP: A Hybrid and Adaptive Method The Cide Archaeological Project was conceived in its planning stages as an attempt to bring the systematic and intensive methodology of the Mediterranean model to a relatively unexplored region of Turkey. Direct models for our methodological strategies included the Kythera (Broodbank 1999) and Antikythera (Bevan and Conolly 2013) surveys and the geographically closer Project Paphlagonia (Matthews and Glatz 2009). In particular, we wanted to apply the notion of the archaeological landscape
58
Developing an Adaptive Field Methodology for Challenging Landscapes
as a continuous entity, in which the full spectrum of evidence for past human activity is only really accessible through an intensive ‘siteless’ approach (Cherry 1994; Bintliff et al. 1999). A series of adaptations and additions, however, were necessary to implement an intensive approach in the specific environmental and topographic conditions of the Cide and Şenpazar survey ‘universe’, which differs significantly from the regions normally investigated by siteless surveys and, therefore, poses a range of challenges for which there exist no established methodological solutions. For instance, both Mediterranean and Near Eastern survey traditions generally lack an engagement with upland regions and marginal environments (Blanton 2000: 628; 2001), although some recent projects have made explicit efforts to address this issue (Casana and Wilkinson 2005; Given 2013a). The mountainous topography and low surface visibility characteristic of the Cide and Şenpazar region left us with little choice but to rethink some of the basic tenets of survey archaeology.
Fig. 4.1: Map showing the CAP survey area. Produced by Toby C. Wilkinson.
The CAP survey area encompassed the Cide and Şenpazar districts of northwest Kastamonu province and amounts to a total of 929km2 (Fig. 4.1). Our initial survey permit for 2009 included only the Cide district (694km2) and was restricted to the study of the region’s ‘Pre- and Protohistory’. This was expanded in 2010 to include the
Ambitions for CAP: A Hybrid and Adaptive Method
59
adjacent district of Şenpazar (235km2) and a permission to extend our chronological range to the Classical and later periods, which allowed us to investigate inland connections and to adopt a long-term perspective on the region’s development. A handful of sites in Cide, dating to the Roman, Byzantine and later periods, had been recorded previously (Belke 1996; Marek 2003), but there had been no systematic exploration of the region, most of which remained a terra incognita in archaeological terms at the start of the project in 2009. CAP’s initial focus was on the Early Holocene, the Chalcolithic and Bronze Ages. We expected sites dating to these periods to be smaller and more ephemeral than contemporary settlements in more populated regions in central Anatolia, and, thus, easily missed by extensive surveys. What became clear from preliminary scouting in the first days of the 2009 season was that much of the survey area was unsuitable or simply inaccessible for large-scale probabilistic sampling and intensive pedestrian survey due to topography, dense forest cover and the logistical constraints that mountainous regions place on team mobility (section 4.2.1). Although tiny in comparison to other surveys in Turkey and the wider Near East, the size of CAP’s research region was much larger than the areas of surveys following the intensive Mediterranean model. Even in less challenging topography, only a small fraction of the region could be ‘fieldwalked’ in the timeframe and with the financial means available to us. Thus, in the first days of the 2009 season it became very clear that a hybrid, adaptive and multi-scalar approach was required to investigate the CideŞenpazar region and its range of ecological zones and landscape types. Much of this approach was developed in the course of the first field season, but CAP’s methodology and sampling strategies continued to evolve throughout the course of the project and were adapted and improved as we grew increasingly familiar with the survey area and its archaeology as well as in response to the ongoing geoarchaeological survey of the area in 2010 and 2011 (Düring and Glatz 2010; Glatz et al. 2011; Düring et al. 2012). The importance of familiarity with the survey area through repeated driving, walking and simply being in the landscape, and on several occasions exploring some of its coast by boat, cannot be overstated in the process of developing and optimising CAP’s sampling strategies and field methods.
4.2.1 Sampling: Ecological and Topographic Zones, Visibility and Accessibility The sampling strategy adopted for the CAP survey area was shaped by a range of environmental and logistical factors specific to the Cide and Şenpazar landscapes, our project’s research agendas, time and financial constraints, as well as the scope of our research permit. Although defined by modern administrative boundaries, our survey area constitutes a representative part of the larger ecological zone that is the Turkish western
60
Developing an Adaptive Field Methodology for Challenging Landscapes
Black Sea region (section 2.1.1). In order to gain an overview of past human presence in, and interaction with the different landscapes within our assigned boundaries, a (nonstatistical) sample of the full range of the region’s ecological and topographic sub-zones had to be investigated (cf. Schiffer et al. 1978: 12; Tartaron 2003: 30). To achieve this, a number of broad but distinctive zones were defined. They included: 1. the coastal plain between Cide and Kalafat, 2. the narrow river valleys connecting the coast with the interior, 3. the coastal hinterland (including the foothills and upland valleys south of the coastal plain), 4. inland slopes and hilltops, 5. the high mountains framing the lower-lying landscapes. With the exception of high-mountains, which were impossible to access and explore in any meaningful way in the time and with the resources available to us, areas for intensive survey were selected from these different landscape types, mindful of geoarchaeological considerations, the likelihood of pre-modern surface archaeology, and prevailing conditions of surface visibility. A geoarchaeological assessment of the CAP survey area demonstrated that each of these landscape categories was characterised by different soil types and modern land-use practices, and was subject to distinctive taphonomic processes with different effects on archaeological surface assemblages (section 2.3; table 2.2). In landscapes as densely covered by vegetation as those of Cide and Şenpazar, the effects of long-term erosion on archaeological surface assemblages have to be weighed against the advantages of surface visibility. Thus, for instance, areas affected by erosion, which in other regions may be excluded from intensive survey due to the likelihood of destruction, burial or displacement of surface evidence for past settlement and land-use became focal points of CAP’s intensive sampling strategy due to the high surface visibility such areas afforded (Fig. 4.2). Ground visibility was also dramatically affected by precipitation and temperature in the months leading up to and during the field seasons. For example, a hotter, drier summer in 2009 resulted in a less dense grass and shrub coverage and higher ground visibility than we encountered in the two subsequent summers, which were characterised by higher average precipitation. As a result, in 2010 and 2011 many fields were covered in waist to head-high vegetation, often reducing ground visibility to almost zero. The problem of ground visibility is further exacerbated by the spectacular decline of agriculture in recent decades, which has resulted in the gradual reversion of fields and grazing areas back to forest (section 2.2.2). Extensive and dense forest cover, plus the mountainous character of our survey region, were undoubtedly the two most important limiting factors for the implementation of intensive (as well as in some parts extensive) surface investigations. About 73% or 680km2 of the survey region is currently covered by forest (Fig. 4.3) and for much of the past, this proportion could have been even higher (section 2.1.1).
Ambitions for CAP: A Hybrid and Adaptive Method
61
Fig. 4.2: Surveying parts of Okçular valley affected by erosion (CAP-P1/388).
Fig. 4.3: Map showing areas covered by forest in the survey region. Produced by Toby C. Wilkinson.
62
Developing an Adaptive Field Methodology for Challenging Landscapes
The majority of modern villages in the Turkish Black Sea region are generally located below 700m above sea level, avoiding gradients above 15 degrees of slope. The upper limit for modern settlement in the Black Sea region is approximately 1,000m, with few isolated hamlets located as high as 1,450m (Yakar 2000: 287) (Fig. 4.4). Much of the zones located above 700-1,000m and with gradients higher than 15 degrees of slope overlap with the forested area. If we take the distribution of modern settlements as an indicator of the likely limits of agricultural settlement in the more distant past, CAP’s effective survey area is further reduced. Landscapes beyond the reach of agricultural settlements are of course anything but devoid of human presence and activity in both the recent and distant past. Activities in these areas may range from herding and mining or ritual activities to the creation of locales for tax evasion and resistance to political regimes (Shaw 1990; Glatz and Matthews 2005; Given 2007). Our own targeted explorations in such zones, many of which required the use of a machete or the Turkish orak to cut a path into the undergrowth, yielded several important sites (chapters 5, 6, 8, and 11), but such forays were limited in number due to the time and energy expenditures involved. Thus logistical constraints of working in a mountain region severely limited the spatial extent of effective field-research on foot as well as by car. Travel times for return journeys from the project base in Cide harbour to the more remote areas of our survey region could occupy the best part of a working day. For instance, the
Fig. 4.4: Map showing areas above and below 15 degrees of slope. Produced by Toby C. Wilkinson.
Ambitions for CAP: A Hybrid and Adaptive Method
63
trip from Cide town to the eastern-most village included in our survey area, İlyasbey, could take between two to three hours on the coastal road for a Euclidean distance of 35km only. The trip from Cide to Şenpazar on the new and spacious Kastamonu–Cide highway would take in the region of 45 minutes. In practice, this meant that, with the exception of targeted excursions to the more remote parts of the survey region, both intensive and extensive survey concentrated within an orbit of a maximum of ca. 90 minute travel time from the project base (Fig. 4.5).
Fig. 4.5: Map showing CAP’s ‘effective’ survey region. Produced by Toby C. Wilkinson.
4.2.2 Recording Frameworks: Data Models, Forms, Digital Database and GIS Although it was not appropriate to sample the landscape statistically, it was nonetheless necessary to create suitable and robust data models to record the survey’s findings. This included the creation of standardised recording forms, a comprehensive digital database and an integral GIS. The data model of the survey was designed around a flexible relational foundation: common indexes (e.g. site number) allowed linking information about archaeological material, i.e. finds (ceramics, lithics, small finds, bones), to find contexts, i.e. places (sites, field units, tracts, grid-squares). Both finds and places were in turn also linked to metadata for relevant photographs and drawings (photos, drawings).
64
Developing an Adaptive Field Methodology for Challenging Landscapes
From the beginning, recording forms and database were conceived in tandem to reflect the methodological aims of the project. Detailed recording forms were created (section 16.2) to ensure that required information was really collected, and in as standard a fashion as possible: exhaustion and weather conditions can have a considerable effect on the quality of recording whilst in the field, and tick boxes and structured questions decrease the chances of errors or missing information. Field forms and the database input interface3 were designed to match each other as closely as possible both to ensure the same data was captured on paper and digitally, but also to facilitate fast input. Data-entry was done by all team members, with one team member, in particular Eva Boeijen, tasked specifically with various aspects of data-entry and record keeping. Field data was digitised daily in order to ensure best possible accuracy, while memory of the field day was still fresh in everyone’s mind and recording mistakes and disagreements could be resolved. As a form of additional back-up, scans were made of all paper documents used during the field seasons, including recording sheets, finds and field drawings (Fig. 4.6). All of this meant that there was little data-processing backlog to be carried forward into subsequent
Fig. 4.6: Data-entry and artefact documentation on the top floor of the Yalı Hotel. Photo by Imke van Hagen. 3 The interface was created using Microsoft’s Access database software: this allowed a quick design and an easy way to extract data into other forms later, though restrictions in its design (especially its inability to share data amongst multiple users simultaneously) caused problems in data cleaning later. In order to speed up data entry especially towards the end of field seasons, the database was split between several laptop computers and subsequently recombined. Any operation larger than CAP would have required the setting up of a central network.
Ambitions for CAP: A Hybrid and Adaptive Method
65
seasons. While logistically laudable, the need to process finds data within the season of their collection was spurred by the requirement to deposit each year’s collected finds in the Kastamonu Museum. Only for the final two-week study season in 2011 were we able to bring back the finds from previous seasons to the project base for (re-) analysis. The development of CAP’s database structure began with the design of its field recording forms ahead of the first field-season in 2009. Recording forms and database design were loosely based on prototypes from the Antikythera survey, kindly supplied by Prof. Andrew Bevan. New forms and database structure were added throughout the duration of the project to reflect improvements to our methodological repertoire (e.g. unit and grid forms). In the field, images (including photographs and scanned forms) and spatial data about the survey (GIS shapefiles and geodatabases, etc.) were mostly stored separately from the field database to keep data-processes efficient and stable. A web-version of the database was started in 2011 with the aim of making it easier for team-members to follow links in the data and integrate images and spatial data from the GIS.4 This provided a speedy access to the CAP datasets for our international team and greatly facilitated specialist work during the publication process. The functionality of this web-based interface gradually increased, and now forms part of the final publication of the data (www.cidearchaeology.com). Certain abbreviations, letter codes or standard referencing formats (table 4.1) were used consistently for documentation, or for marking objects, to allow quick cross-referencing between data. These abbreviations have been carried through to the web-database and also appear at appropriate places in this publication.
4.2.3 Spatial Frameworks: Sites and Units The spatial models of the project were designed to integrate both extensive and intensive practices. Two archaeological designators dominate the description of space or place in extensive and intensive survey paradigms: namely ‘sites’ and ‘transects/tracts’ respectively. In vector or topological terms sites are normally characterised as ‘point’ (or ‘centroid’) features and tracts as ‘polyline’ or ‘polygon’ features. Though the concept of an archaeological ‘site’ has been widely recognised as problematic (Dunnell and Dancey 1983), it remains an indispensable heuristic term in both specialist and wider literature. Administrative bodies, including museums and governments, are also deeply locked into the node-based concept of ‘sites’. Whilst intensive ‘siteless’ survey was intended to
4 The available version of Microsoft Access proved difficult to scale to the web, and was restricted by which team members had a copy of the software. Data from the original database tables were transferred to a MySQL database, and an HTML/PHP interface designed using the Symfony2 object-oriented MVC framework.
66
Developing an Adaptive Field Methodology for Challenging Landscapes
Tab. 4.1: Letter codes/abbreviations used in the field, in publication and in the CAP database. abbreviation
example
full form
equivalent database table name
S U G Gxyy T
S45 U567 G4 G4H4 T3
Site Field Unit Grid Grid-square Tract
cdb_Sites cdb_FieldUnits cdb_Grids cdb_GriddingSquares cdb_Tracts
C L F K
C1333 L15 F8 K6
Ceramic Lithic Small Find Bone
cdb_Ceramics cdb_Lithics cdb_SmallFindObjects included as part of cdb_ SmallFindObjects
P D
P2/366 D5/14
cdb_Photos included as part of cdb_Drawings
ds
ds4555
Photograph Drawing – field numbering Drawing – new numbering of drawings
cdb_Drawings
do away with the concept, most projects practicing tract-based fieldwork also continue to identify ‘sites’ as a way to clarify their results. But the interpretation of ‘sites’ from artefact distribution data gathered from ‘tracts’ is far from straightforward, as several decades of debate in the Mediterranean and elsewhere attest to (Bintliff and Snodgrass 1988; Dunnell 1992; Ebert 1992; Wandsider and Camilli 1992). Unsurprisingly, there is still no universally accepted definition of what constitutes a site in terms of relative and absolute artefact densities per given unit of analysis and its identification remains both an observational and interpretive procedure (Cherry 1994: 119; Kanter 2008: 45-6). For the purpose of CAP, we defined sites qualitatively to designate points or areas in the landscape that we felt were significant to the investigation of the Cide– Şenpazar region’s occupation history. Listed in table 4.2 are the four sub-categories of sites, which we defined based on site characteristics, method of discovery, and to enable the recording of cultural and natural locales whose significance were not clear at the time of investigation. The site category of cluster takes the form of a clearly identifiable concentration of cultural material relative to the surrounding landscape as it is normally encountered during intensive survey, and that is indicative of a spatially bounded past human activity or activities. CAP clusters were identified intuitively rather than statistically from surface density distributions derived from unit-walking. Low overall artefact counts and low visibility scores have thwarted all our attempts at a more rigorous quantitative approach. All our distribution maps of pottery, tile and lithic surface densities, therefore, represent absolute counts. The concentrations, which we
Ambitions for CAP: A Hybrid and Adaptive Method
67
designated as clusters stand out clearly in our GIS distribution maps. In one case at least, a follow-up geochemical investigation of soil samples from such a cluster (G6/S513) corroborated our interpretation of the concentration as a settlement locale (section 4.7, chapters 11 and 14, section 16.4). Tab. 4.2: Classes of sites identified in CAP and their characteristics. site type
type
examples
motives
Cluster
archaeological
sherds or lithic concentrations
Feature
archaeological
Spot
archaeological
walls, bridges, graves, mills, spolia, modified and occupied caves, castles. sherds/lithics/previously reported features
to document meaningful artefact clusters to document features of archaeological interest
geomorphological geomorphological sections
Group
natural
caves
archaeological
collective of clusters, features and/or spots
to document the location of spolia or other individual finds and structures to document locations with information on landscape formation to document locales of potential interest to identify spatial associations between clusters, features and/or spots
Features map onto what extensive surveys commonly identify as sites: conspicuous features in the landscape that are clearly of archaeological interest, ranging from standing structures, ruins and spolia to cemeteries and caves with signs of past human occupation. The spot is a primarily heuristic category created to allow us to record data, whose archaeological interest was not clear at the time of investigation. This includes individual artefacts or small concentrations encountered outside of regular fieldwalking, such as a scatter of pottery on the path leading to a field to be walked intensively, or a pipe fragment shown to us by local residents and found in their fields. Natural features were also recorded as spots. In the course of CAP we visited many caves in the hope of finding signs of past human activity (section 4.3.2); while some yielded rather spectacular finds (chapters 5, 6 and 8), the majority did not. The latter were recorded as natural features and designated as spots, as were the riverbanks, erosion gullies and construction pits which were investigated in the course of our geomorphological survey (section 4.6). The group is the means by which we combine clusters, features or spots which have a spatial association, but for a variety of reasons were recorded separately or investigated by different survey methods. For instance, we recorded a range of Late Roman/Early Byzantine artefacts and features from a range of disparate locales in the
68
Developing an Adaptive Field Methodology for Challenging Landscapes
Cide-Irmak plain at different points during the three field seasons and during both targeted visits and intensive survey. As a group, these finds define the outlines of a large Roman period coastal settlement. The region around the Gökçekale promontory provides a similar example (chapters 10 and 11). CAP also has a number of different types of polygonal units of investigation through which our data on artefact distributions was recorded (table 4.3). For reasons that will be outlined below (section 4.3.4), Mediterranean-type tracts were trialled at the beginning of the first season but quickly abandoned due to the difficulties that the rigidity of walker spacing and the fixed walking directions of this model posed for the identification of surface materials in our particular context. Tracts were replaced by field units, which describe a polygonal area of ground coverage, normally a square of 50 x 50m, through which our team walked in either a ‘systematic’ or random (‘grab’) fashion. Grid-squares were laid out over areas of particular interest to intensify the mapping of densities and/or the collection of surface materials. These spatial models formed the centre of the CAP GIS platform. Both commercial and open-source software was used during fieldwork and in subsequent analysis5 to integrate and compare the collected spatial data (sites, units, grids), all of a vector nature, with various sources of raster and vector geographical data, including high-resolution satellite imagery from Google Earth, topographical data from the SRTM (Shuttle Radar Topography Mission) and ASTER (Advanced Spaceborne Thermal Emission and Reflection Radiometer) missions, and simplified locations of settlements and roads from OpenStreetMap and NGA (National Geospatial-Intelligence Agency) databases for navigation. For the most part, CAP adopted the UTM system to record co-ordinates and to reference spatial data; with the whole of the survey area falling within the 36N zone, this simplified recording and allowed distances to be easily compared using metres in the field (which would not have been so easy using latitude/longitude-based coordinates). The GIS formed a central tool of planning and recording of the survey from the beginning, as will become obvious below.
4.3 Field Methods During CAP, four central methodological templates were followed to investigate the archaeology of the Cide-Şenpazar region. The first consisted of desk-based analyses of written sources, maps, and remotely-sensed landscape data. The second encompassed extensive, and mainly car-based, explorations of the region in order to gain a better understanding of different landscapes and their formation over
5 Including ArcGIS (versions 9.1–9.3), QGIS (versions 1.2–1.8), GlobalMapper (version 6), Google Earth, gpsbabel. Access to some of this software was dependent on team members’ institutional affiliation.
an arbitrarily-shaped and sized tract defined in the field using GPS points or drawn onto georeferenced map using landscape markers (field boundaries, etc.); arbitrary number of walkers, chosen according to size of area
grab
Grid & Gridsquares
square ‘tract’ of 50 x 50m, predefined using GIS and identified in the field using GPS; ~5 walkers walking in same direction
systematic
Unit
grid made up of a number of 5 x 5m or 10 x 10m grid-squares, each walked by one walker, the full grid is defined by the fall-off of archaeological material or landscape barrier
an area within a gridding-square
standard
vacuum
a rectilineal tract defined in the field by 4 cardinal GPS points; maximum of 5 walkers
–
Tract
parameters
subtype
tract type
Tab. 4.3: Surface collection area-units.
to increase the likelihood of recovering ephemeral but important materials (e.g. lithics)
to document with greater nuance and resolution the spatial extent of ceramic or lithic clusters, or to increase the recovery and representation of certain classes of material
to investigate areas which could not be examined using a standard-sized, ‘systematic’ unit
to create a flexible unit of investigation by which densities of archaeological surface material can be recorded to investigate archaeological distributions intensively and quickly in a standardised and a more landscapeappropriate way than normal ‘tracts’
to create a flexible unit of investigation by which densities of archaeological surface material can be recorded (deprecated after difficulty to implement in 2009)
motives
Field Methods 69
70
Developing an Adaptive Field Methodology for Challenging Landscapes
time, and to document features pointed out to us by local residents. Some of Cide and Şenpazar’s residents also volunteered information about the region’s more recent past, its inland and maritime connections and economy; two kindly agreed to be recorded in formal interviews. The third thread consisted of the intensive survey of selected areas that were both representative of the different landscape types and ecological zones characteristic of the Cide-Şenpazar region and likely to yield archaeological surface materials (section 4.2.1). A fourth strand included the re-investigation of concentrations of archaeological materials recorded during unitwalking more intensively through grid-square collections. In the following we will discuss these techniques in more detail.
4.3.1 Remote Sensing, Mapping and Desk-Based Survey Before setting foot in the survey area, work began on gathering relevant written sources and landscape data (including maps and remotely-sensed data) for the CideŞenpazar region. Some of this data, especially high-quality maps, proved difficult to acquire. At first, we were unable to obtain copies of the 1:25,000 government maps, which are sometimes used as the basic map for surveys in Turkey, but are closely controlled for security reasons. Alternative maps, including 1:100,000 Soviet army maps provided basic topographic information and showed the location of modern towns and villages, but given their scale, were of limited use on the ground. In 2010, we received several 1:25,000 topographic maps from the local Land Registry and Cadastre Office in Cide, which covered some, but not all of our survey area. For basic navigation between villages, the most useful map turned out to be a basic 1:100,000 line-drawn road-map from the Köy Hizmetleri (Village Services), which the Yalı Hotel, our project base, provides to its guests. This map provided the most accurate information on the local road network, the location of modern villages and their names. Although it contains no topographic information, overlaying the Köy Hizmetleri map onto satellite imagery from Google Earth provided us with a rough but useful tool for planning field visits. Following the traditions of extensive survey, the use of satellite imagery and remotely-sensed data for reconnaissance and pre-fieldwork planning has become common practice across the Near East, overtaking the earlier applications of aerial photography (Wilkinson 2003: 33-7). Such practices have been particularly effective in the exploration of semi-arid lowland plains, valley systems and high plateaus (Adams 1981; Wilkinson and Tucker 1995; Ur 2002; Pournelle 2007; Casana and Cothren 2008), where archaeological features are more visible through soil colour or three-dimensional characteristics on a flat background. To date at least, the same techniques have been less successful in regions with rugged topography and/or dense vegetation cover (Parcak 2009: 125-6), although LiDAR may offer help in the case of vegetation (Evans et al. 2013).
Field Methods
71
During CAP we made extensive use of satellite imagery to identify potential points and areas of interest to be subsequently ground-truthed, as well as to orient ourselves in the field. Before and during fieldwork, we gathered LANDSAT multispectral imagery, and SRTM and ASTER topographical data/DEMs (Digital Elevation Models) from NASA as well as declassified CORONA photographs (via the Oriental Institute’s CAMEL programme). Sadly, the mountainous and forested nature of the region, as well as cloud coverage when the photographs were taken, meant that the CORONA images were of no practical use. However, the 90m SRTM and 30m resolution ASTER DEMs were very important in mapping the topography of the overall region to establish altitudes and slopes, even if the scale was still too large for localised mapping of sites. Throughout the survey, we made heavy use of Google Earth and Google Maps, whose base imagery of the Cide region relied on basic pan-sharpened LANDSAT imagery (~15m) during 2009, and then was augmented in 2010 and 2011 by very high-resolution pan-sharpened version of Digital Globe’s Quickbird multispectral imagery (downgraded and calibrated to show visible light only, pan-sharpened from 2.4m to ~0.65m). The strips of Quickbird imagery in Google Earth available to us during that time were originally collected on 27.08.2003 (north/south of Cide centre), 11.07.2004 (to the east), 05.03.2006 (to the west). Whilst the vegetation cover of the region continued to be a limiting factor with regards to remote site identification, the availability of this resource provided a fantastic ‘scouting’ and planning tool for fieldwork, and, as will be discussed in more detail below, an orientation aid in the field. One site, Çoban Kalesi (S32), which was first reported by Kalinka (1933), was clearly visible on Google Earth imagery, and we initially identified and explored it using satellite data before investigating on the ground (chapter 12).
4.3.2 Targeted Reconnaissance Targeted reconnaissance formed the main ‘extensive’ component of CAP’s field methodology. It proved especially important for the discovery of the earliest phases of occupation of Cide and Şenpazar, which are underrepresented in the surface record (chapters 5 and 6). This method also yielded some of the most spectacular sites recorded by CAP. We came to fully appreciate the importance of this targeted component when, during the first season of fieldwork, local residents led us to a series of sites which in other regions might have been easily spotted from the road, such as the standing walls of Byzantine castles, but which here were completely obscured by dense forest canopy. Others, especially the many caves of the region, required long hikes into the mountains along paths no foreigner to the region could hope to identify alone. Much of our information about locally known places of potential archaeological interest came from Recai Yılmaz, a primary school teacher, photographer and local expert on Cide and its landscapes (Fig. 4.7, section 4.3.2). The sites, which Recai bey
72
Developing an Adaptive Field Methodology for Challenging Landscapes
showed us or told us about include several caves with extensive Chalcolithic, Bronze Age and later materials (chapters 5 to 8) and the standing remains of a series of castles (kaleler) along the coast and in the interior (chapters 11 and 12).
Fig. 4.7: Recai Yılmaz talking to Michele Massa in the Gideros İni I cave, which he brought us to in 2009. Photo by Renate de Boer.
Besides visiting locations highlighted by our local guides, we also undertook a number of non-directed ‘vehicular transects’ throughout the survey region, stopping in villages along the way and asking about places of historical or archaeological interest. Admittedly, much of the information volunteered during such conversations was either very vague (of the ‘uncle-so-and-so many years ago had been to a cave somewhere-up-this-or-the-other mountain’ sort), or else the locales in question, caves in particular, ultimately yielded no cultural material. Although we were usually welcomed in a very friendly manner (Fig. 4.8), in some cases, villagers took us for illegal treasure hunters or were understandably suspicious of the strangers walking through their fields. The production of our permit and explanations by our temsilçi or local supporters, sometimes over tea, normally assuaged such anxieties. Once trust was gained, villagers would generously take us on lengthy and difficult hikes of scrambling through steep and thorny terrain in order to show us places of potential archaeological interest.
Field Methods
73
Fig. 4.8: An important invitation to drink çay by locals at Kuşçu (CAP-P2/0033).
Although we visited many empty caves and natural hills as a result, the rewards of this approach could be great. Sites recorded in this manner include, for instance, the cave site of Derebağ Köy Mağarası (S188), which yielded a large sample of Chalcolithic, Bronze and Early Iron Age pottery (chapters 5 to 8); and the remains of several Roman or Early Byzantine structures (S89) which villagers at Kuşçu have incorporated into their houses (chapter 12). On a few occasions, portable finds, from glazed pottery fragments to a Roman period headstone (S208), were shown to us in the field or brought to the project base. As our familiarity with the survey region increased and the pattern of site distribution became more evident in the second and third field-seasons, we undertook a number of targeted investigations of locales where we suspected further sites to be located. We focused in particular on the detection of coastal castles as the distribution and regular spacing of known kaleler suggested the existence of more such sites within the survey area. An example of a site located in this manner is the group of features around Gökçekale (S174, S176 and S178). In landscapes as taphonomically active as those of Cide and Şenpazar (section 2.3), road-cuts, construction pits or river-sections can hold important clues about buried landscapes and sites (Banning 2002: 40-41). For this reason, road-cuts and river-sections were examined wherever the opportunity arose, but throughout the survey no archaeological materials were detected in this manner. More successful in
74
Developing an Adaptive Field Methodology for Challenging Landscapes
this respect was the investigation of gravel and rubbish pits and newly dug house foundations in the urban surroundings of Cide town and along the coastal plain (e.g. S70, S71, S73 or S196), which yielded glimpses of a substantial Late Roman and Byzantine occupation now buried under at least a metre of deposits (section 2.3, chapters 10 and 11). In the field, locations of interest were documented as sites (either features or spots) (section 4.2.3) using CAP’s site recording form (section 16.2). Collected information includes how the site was discovered, the type of site, its main characteristics, preservation condition, material density, amount of surface material collected, provisional dating, as well as its topographic setting and modern land-use in the vicinity. GPS points were taken at the centre and, where definable, on the edges of sites and a sketch drawn on the back of the site form. Basic scaled drawings were made of architectural remains (using either tape measures, handheld GPS, handheld differential GPS, or a total station depending on the complexity of the site), and photographs taken, which were subsequently linked in the database via the site’s ‘S-number’. The collection of surface material differed depending on the nature of the site. In the case of standing architectural remains, often located in the midst of dense forest and shrub, a selection of surface artefacts was collected by team members through a more-or-less random exploration of the immediate surroundings. Cave floors were scanned more systematically for cultural material. Depressingly, we soon discovered that treasure hunting is a local pastime as in many other parts of Turkey. Many of the sites we visited in the course of our extensive reconnaissance, in particular caves and churches, had been heavily disturbed and damaged by illegal digging or the removal of building materials. In some cases, it appears that explosives were even used to crack rocks. The results of these activities created a perverse insight into information otherwise inaccessible to us. Whilst the scale of destruction of cultural heritage is enormous, disturbing and thoroughly objectionable, the robber pits we encountered particularly in caves left visible on the surface cultural material which was crucial for our understanding of especially the earliest phases of occupation in the region (chapters 5 and 6). Bound as we were by the rules governing legitimate archaeological research in Turkey (section 4.1.2), this material would have been otherwise inaccessible to us. Continued illegal excavations and destruction of these sites are likely to result in their complete obliteration in the near future and perhaps have already done so. To stop treasure hunting and illegal excavations in the area will require a significant effort to inform local residents of the importance of their recent and more ancient cultural heritage as well as the consistent punishment of dealers and collectors of illicit antiquities locally and worldwide. We endeavoured to contribute to this effort through informal discussions with local residents in the field, a series of newspaper articles about the activities of CAP written by Murat Karasalihoğlu in the Kastamonu Gazetesi and Posta (Karasalihoğlu 2009, 2010, 2011), as well as our bilingual TurkishEnglish project website.
Field Methods
75
4.3.3 Local Memories: Oral History and Historical Photographs The collection of oral history data and historical photographs was undertaken in order to provide a further source of information about the study region, especially with regards to connections—both cultural links and communication routes—to the interior of Kastamonu province or along the Black Sea coast. We also wanted to learn more about the region’s economy in the more recent past and the changes it had undergone. In addition to ad hoc conversations with local residents, which took place during our field days and had no particular structure to them, two knowledgeable Cide residents, Ali Osman Tuğtepe (proprietor of the Yalı Hotel in Cide) and Recai Yılmaz, agreed to be interviewed about the recent history of the region based on their personal experiences and memories. Both interviewees agreed that their conversations be recorded and the information used in publications. The main questions asked can be found in table 4.4. Tab. 4.4: List of interview questions. themes
questions
Communications
When was the coastal road to Bartın/Sinop constructed? How did people go to these places before this road was constructed? When was the Cide harbour constructed? What mooring places for ships were used prior to that? How was coastal transport organised in the past? How long did it take to reach Amasra and Istanbul by boat? How long did it take to reach İnebolu and Sinop by boat? Was there any transport across the Black Sea (e.g. the Crimea/Russia)? When was the road to Kastamonu constructed? What type of road was there previously? What was its course? How long did it take in the past to reach Kastamonu? What types of farming were there in Cide in the past? How important for the local economy were timber shipbuilding coal hunting fishing mining linseed production craft production coastal trade Would you describe yourself as coming from Kastamonu or Cide? Do you feel more connected to Bartın/İnebolu/Sinop or Kastamonu? Where do people living in Cide now come from? Do you know who the people were who lived in this area in the past? Were there any non-Muslims? What language did they speak? Did they live in specific places or villages?
Economy
Society
76
Developing an Adaptive Field Methodology for Challenging Landscapes
Both interviewees and other local residents who volunteered information about their region’s recent past or showed us places of archaeological interest were men. This is a reflection of the cultural prominence, in public at least, of men in Turkey and the Black Sea, particularly in rural contexts. Planned follow-up interviews, including with local women, proved impossible to complete due to time restrictions which forced us to prioritise finds processing in the last season. Despite their limited number and gender bias, these interviews provided us with many useful insights, and the discussion on communication routes and the local economy of Cide and Şenpazar in chapter 2 (sections 2.2.1 and 2.2.2) is based in part on these conversations.
4.3.4 Tract-Walking and Unit-Walking Initially we had envisaged an intensive survey strategy modelled on Mediterranean ‘siteless’ or continuous surveys such as the Boeotia Survey (Bintliff et al. 2007), the Kythera Island Project (e.g. Bevan and Conolly 2004) and the Antikythera Survey (Bevan and Conolly 2012, 2013). Once in the field, the implementation of a methodology based on rigid or statistically formulated tract-walking proved more challenging than we had anticipated. Having scouted the survey region for suitable areas for intensive fieldwalking, we began the 2009 field season by testing the first stage of this intensive approach in the fields and meadows around the villages of Mencekli, Kuşçu and Sakallı in the Aydos valley east of Cide. Our two field teams consisted of 5 walkers each, who were spaced 20 metres apart and walking standard 100-metre tracts wherever possible (an idealised version of this system is shown in Fig. 4.9a). The first difficulty we encountered was the identification of suitable tract-walking areas. A decline in agriculture over the past decades (section 2.2.2) means that in most valleys there are no ploughed field systems that elsewhere lend themselves to intensive tract-walking. Abundant rainfall means that abandoned fields are quickly overgrown. Former fields that had not reverted to forest include meadows, grazing areas and orchards. These were normally covered in lush green grass and shrub
Fig. 4.9: Diagram showing (a) traditional linear transects, (b) undulating transects and (c) CAP’s meandering approach.
Field Methods
77
resulting in 0-10% visibility and zero sherd counts in over three quarters of the tracts walked. Modest peaks in artefact counts in areas with higher ground visibility, such as ploughed fields and areas affected by erosion, however, suggested that our problem was ground visibility and not the absence of cultural materials. After a number of unsatisfactory forays, we realised that we had to adapt our intensive methodology in order to be able to focus on parts of the landscape that afforded higher ground visibility. It also meant doing away with the rigid walker spacing and linear directionality of standard tract-walking, as this rigidity had forced our walkers to bypass high-visibility areas such as erosion patches and therefore ignore archaeological material. Our experience resonates with those working in similarly challenging landscapes, such as parts of Corsica: “The generation of stratified random samples within each geographical unit was abandoned; instead, attention was focused on maximizing retrieval of surface material by targeting areas deemed to have greater surface visibility and accessibility. Such a shift marked a change from an idealistic, statistically based, but ultimately impractical approach, to a more pragmatic one. We retained the morphometric classification as a means of dividing the landscape and to enable us to target areas that Corsican archaeologists have not traditionally considered, e.g. valleys.” (Llobera et al. 2010: 174).
In order to maximise the probability of artefact detection, which as our experience had shown fell off exponentially in contexts of low ground visibility, we modified our intensive strategy to a method resembling what Banning has called ‘undulating transects’ (Banning 2002: 91-92; figure 16b). Unlike the model he describes, which specifies walkers undulating in parallel bands (Fig. 4.9b), we allowed our walkers to meander in individual shallow zigzags within their hypothetical tract lines in order to gain access to all high-visibility areas in their path and inspect them from a range of angles (Fig. 4.9c). Similar undulating transects were employed by Mortensen (1974) in Luristan and by Rupp (2004) in the Paphos Survey in western Cyprus to balance out low visibility areas and allow walkers to circumvent obstacles in the landscape. This change in survey approach allowed not only specific types of surfaces to be targeted by walkers, but also dramatically increased the overall area surveyed by each walker. Since tracts no longer seemed an appropriate description of our field method, suggesting as the term does the traditional model, the term unit was adopted (section 4.2.3; table 4.3). We further intensified our approach by reducing the walker spacing and hence size of the area covered from 100 x 100m for tracts to 50 x 50m for units. Field team sizes remained the same, so that each walker effectively covered a 10m-wide strip. Areas amenable to this intensive approach were surveyed more-or-less in their entirety, rendering our approach hyper-intensive in those areas that permitted fieldwalking. The first test area for this new approach was the Okçular region in the coastal hinterland of Cide. Located around 200m above the coastal plain, this small upland valley is dominated by the rock outcrop of Okçular Kale with its heavily overgrown remains of a Byzantine stronghold (S22). Okçular Kale received cursory mentions by travellers and
78
Developing an Adaptive Field Methodology for Challenging Landscapes
Classical and Byzantine historians (Belke 1996: 255), and was shown to us by Recai bey in 2009 (chapters 11 and 14). The valley stretching south from the kale (which also sits on top of an important cave site, Okçular Kale İni (S24), is dissected by old field boundaries in various stages of abandonment and disrepair. Large-scale erosion has cut deep gullies into the eastern valley flank and denuded its surface vegetation (Fig. 4.10). The combination of high surface visibility and the presence of a known archaeological site made Okçular the ideal testing ground for our modified survey strategy.
Fig. 4.10: Google Earth satellite image of Okçular valley, showing the large-scale erosion on the eastern valley flanks.
An initial experiment was to map the units in the field ‘free-style’ as they were being walked. The outermost walkers of two adjacent field teams used the track path function on our hand-held GPS receivers (Garmin eTrex) with a compass to record the edges of units. This was quickly abandoned, however, because it led to very uneven coverage. Instead, to standardise the shape of units, maps of survey areas were produced using a base map (either LANDSAT or Google Earth) with a grid of units (of 50 metre square each) and a combination of GIS applications (including Hawth’s Tools in ArcGIS) (Fig. 4.11). For this first zone covered by units, Okçular east, the grid was aligned according to the local topography but subsequent grids were laid out in a north-south orientation for simplicity. The corner points of each unit were given numbered labels and fed into hand-held GPS receivers. A printed version of the map (with unit corners and base map) was given to team leaders. In the field, one walker from each survey team (often the outer-most walker) located the corner points in his/her path using the GPS and marking them with biodegradable survey tape to guide the direction of walking.
Field Methods
79
Fig. 4.11: Map of Okçular valley showing units surveyed in 2009-2011. Produced by Toby C. Wilkinson.
On occasion, in terrain where these arbitrary units proved too cumbersome and time-consuming to implement, such as in the small fields and vegetable gardens surrounding modern villages, field boundaries, roads and other distinct landscape features were used to define unit shapes (Fig. 4.12). GPS readings of unit corners were taken and sketches of unit shape and size drawn onto the printed grid maps. Both regular and irregular shapes were subsequently digitised at the project base and integrated back into our GIS platform. For this, Google Earth and its imagery was a considerable help in identifying landscape markers back at the project base. Depending on the size and accessibility of a survey unit, all or part of the field team would walk the unit, and, wherever feasible, in a north or south direction (the direction of walking was recorded using the unit forms). For small or partially accessible survey units such as fenced-off garden patches or fields with young crops in which standard-spaced undulating/meandering walking was not practical or appropriate, edges were walked and the investigation described as a ‘grab’ sample. Each field team consisted of between four and six, typically five, walkers and included one of the co-directors, one other experienced survey archaeologist and two to four graduate and undergraduate students with varying archaeological field experience. Photographs were taken of each unit. Unit organisers made strategic decisions about the location of units and recorded walker specific information on visibility and artefact counts as well as general information about the survey unit. Standard information recorded included the topographic setting of the survey unit, degree and type of erosion, information on current land-use, and any other features and observations of interest (section 16.2).
80
Developing an Adaptive Field Methodology for Challenging Landscapes
Key Year of investigation 2011
0
0.5
1 Km
Fig. 4.12: Map of the Loç valley showing a combination of arbitrary unit and survey areas defined by field boundaries and other landscape features. Produced by Toby C. Wilkinson.
Walkers used tally counters to record sherd densities and collected all diagnostic ceramics, potential Prehistoric, Bronze and Iron Age body sherds, lithic and small finds. In the first season, we also collected diagnostic tiles, but given their size and ubiquity this proved to be impractical. The decision was taken in the second year to draw and describe diagnostic tiles in the field instead and only collect chronologically sensitive examples. In total, we surveyed ca. 3.4km2 of the Cide and Şenpazar landscapes in this manner (table 4.5). Tab. 4.5: Intensive fieldwalking: number of units and area walked. area
2009
2010
2011
total area (ha)
units
Okçular
55
42
21
119
514
Abdulkadir
4
46
8
58
258
Aybasan
17
31
48
171
Sakallı Köy
8
8
45
Cide town
0.2
12
Coastal plain Çamdibi Total
59
125
12
45
41
41
130
61
16
227
162
348
1430
81
Field Methods
4.3.5 Grid-Square Collection In 2010, we introduced a follow-up sampling methodology, grid-square collection, to increase the intensity of investigation of a few artefact scatters of particular interest, which were first identified during unit-walking and often later classified as cluster sites. In each case, a grid consisting of 5 x 5m or 10 x 10m squares was flagged out with tape measures and compass around a base point (the co-ordinates for which were recorded using a GPS receiver to allow subsequent digitisation in the GIS platform). In each standard grid, one walker spent 5 minutes counting all pottery and collecting diagnostic ceramics and any Prehistoric, Bronze and Iron Age body sherds, lithics, metals and any other small finds (see sections 16.2) within one grid-square before moving onto another square. Where appropriate, an additional fixed-time ‘vacuum’ or total collection of all surface material was undertaken by the same walker in a smaller 1m radius at the centre of each grid-square. In total, we investigated eight grids, five in Okçular valley and three at Aybasan (Fig. 4.13).
0
0.5
1 Km
G8
G3 G1
G2 G6
Key Gridding Squares Units
Fig. 4.13: Location of CAP grid-squares in Okçular valley. Produced by Toby C. Wilkinson.
This method of high-intensity grid-square collection was applied under two different circumstances and for two slightly different but complementary research objectives. In the first model, our aim was to obtain a larger sample of diagnostics and a more
82
Developing an Adaptive Field Methodology for Challenging Landscapes
representative spread of materials from an area of special interest. This was the case in places such as the fields below Okçular Kale (G1, G2, G3) and a second, small rockoutcrop to the northeast (G8), where the lower intensity unit-walking had identified sherds of particular interest, but only very few. In these cases we opted for small grid squares (5 x 5m) and, in the case of the Okçular Kale area, additional vacuum time. In the second model, the primary objective was to attempt to define the extent of a particular artefact scatter better in addition to gaining a larger sample of diagnostic finds. Grid square size in the latter instance varied according to the extent of the surface scatter. The lithic assemblage at Aybasan was gridded in 5 x 5m squares (G4, G5, G7). A tile and pottery scatter in the Okçular erosion area was investigated using 10 x 10m squares (G6). In addition to the collection of archaeological material in Grid 6, we also took 48 soil samples from the grid-square centre points, which were subsequently exported to Glasgow for laboratory-based chemical and trace element analyses (sections 4.7 and 16.4). In total, we (re-)investigated 2.5ha of the Cide and Şenpazar landscapes with this high intensity method (table 4.6). Tab. 4.6: CAP grid-square collection areas. grid
year
area
size (ha)
5m
1 2 3 4 5 6 7 8
2010 2010 2010 2010 2010 2011 2011 2011
Okçular Okçular Okçular Aybasan Aybasan Okçular Aybasan Okçular
0.25 0.17 0.25 0.09 0.01 1.5 0.08 0.15
x x x x x x x
10m
x
vacuum
main finds
x x x
pottery pottery pottery chipped stone chipped stone pottery and tile chipped stone pottery, chipped stone
4.4 Analytical Recording Procedures and Approaches to Finds In the course of the Cide Archaeological Project, we set up a range of standardised procedures to record different artefact categories such as ceramics, lithics and small finds, as well as for the description of soils and for organising the GIS and database. These procedures will be described here.
4.4.1 Pottery Procedures 4.4.1.1 Pottery and Tile Recording and Macroscopic Analysis The central aim of macroscopic pottery analysis in the context of field survey is relative dating (section 4.5). A second aim, which in practical terms is closely related
Analytical Recording Procedures and Approaches to Finds
83
to the first and forms part of CAP’s main research questions (chapter 1) is the degree of overlap or resemblance of the CAP material with the cultural traditions attested in other parts of Turkey and along the Black Sea coast. The ceramic spectrum represented in the CAP assemblage overall as well as at specific finds locations, moreover, may permit tentative conclusions about particular activities and functions represented at such locales. Recorded artefact surface densities are consistently low across the Cide-Şenpazar region, even when low surface visibility and a range of taphonomic processes are taken into consideration. The total sherd material collected and recorded in the field in the course of three field seasons is accordingly modest: the final CAP database holds a total of 4125 ceramic entries, of which around 900 are tiles. All diagnostic pottery collected in the field, which include rims, handles, bases and decorated sherds as well as body sherds of likely pre-Classical date, were photographed, registered into the CAP database and assigned a ceramic finds accession number designated by a ‘C’ followed by a unique consecutive number. Standard details about each find’s metric dimensions, manufacturing technique, macroscopically observed fabric properties and coarseness, colour, surface treatment and firing condition, were recorded in separate database fields. Also recorded were the vessel type each find was likely to represent (table 4.7), which allows conclusions about the broad functional spectrum of individual finds and larger assemblages. All diagnostic pottery and tile fragments were also drawn, as were body pieces likely dating to the Chalcolithic, Bronze or Iron Ages. Procedures for the recording of diagnostic tile fragments were more or less the same as those for collected sherds. The main difference was that for practical reasons tiles were no longer collected in the 2010 and 2011 seasons, but were drawn and described (but not photographed) in the field.
4.4.1.2 Laboratory-based Pottery Analyses (Richard Jones and Claudia Glatz) Following the final field season, a total of 31 pottery fragments were exported to Glasgow for petrographic, chemical composition and residue analyses by Richard Jones. With tight restrictions on the number, size and nature of export material from the local authorities, we focused on Chalcolithic and Bronze Age pottery from the cave sites of Ballıcı İni (S17), Okçular Kale İni (S24), Kılıçlı Mağarası (S88) and Derebağ Köy Mağarası (S188), units from Abdulkadir (U7), Okçular (U219, U524, U554, U646) and Cide (U814). The export sample also included a small selection of Iron Age, Hellenistic and later sherds (table 4.8). The aim of the laboratory-based pottery analyses was to shed more light on Prehistoric and Bronze Age pottery traditions, including the identification of locally produced and imported material, overall assemblage diversity, and continuity and
84
Developing an Adaptive Field Methodology for Challenging Landscapes
Tab. 4.7: Abbreviations used for the description of vessel types in the CAP database. acronym
vessel type
description
AMPH
amphora
HEL/ROM/BYZ medium-sized storage/transport container
BOTTLE
bottle
BOWL
bowl
BRICK
brick
small to medium-sized container with elongated, constricted neck small to medium open vessel used for the preparation, serving and consumption of food building material
CHAFING DISH
chafing dish
medium-sized flat vessel likely used for chaffing
CPOT
cooking pot
CUP
cup
vessel whose (usually coarse) fabric and signs of secondary burning identify it as a cooking instrument small, open consumption vessel
HH
horizontal handle
horizontal handle
HV
vertical handle
vertical handle
IMBREC
imbreces
curved terracotta roofing tile
JAR
jar
medium-sized storage container
JCON JUG
jar with constricted medium-sized storage container with narrow/constricted neck neck dating to any period jug small to medium vessel used for pouring liquid
LID
lid
MINI
miniature
flat/shallow vessel (possibly with lug/handle on the outside) used to cover e.g. storage containers miniature vessel
PITH
pithos
large storage vessel
PLATE
plate
medium to large shallow vessel with rim articulation
TEAPOT
teapot
globular vessel with elongated spout
TILE
tile
ceramic roof tile
UNKNOWN
unknown
vessel type unknown/unidentifiable
WASTER
kiln waster
over-fired (and usually deformed) vessel, tile or fragment
change in local technological traditions. Another goal was to determine to what degree macroscopically observed differences between ‘wares’ were borne out by petrography and X-ray fluorescence (XRF) analysis. At Glasgow University, the pottery samples were cut with a diamond saw and samples ground to a thickness of 0.03mm and mounted onto glass slides. The thin sections were then examined with a polarising microscope. The chemical composition of the samples was determined non-destructively by XRF, using a Thermo-Niton XL3t instrument. Each sherd was placed on a Niton-manufactured test stand allowing constant distance and geometry between the X-ray beam and the selected locations on the cut section of each sherd. The analysis area was estimated at ca. 10mm2. The
Analytical Recording Procedures and Approaches to Finds
85
Tab. 4.8: Ceramic samples exported for analysis (see table 4.10 for period abbreviations). C-number
site
unit
grid
approximate date
56
24
MCHA
170
17
CHA
358
7
MBA-LBA? (Okçular ware)
633
219
HEL
750
524
HEL
790
24
MCHA
797
24
MCHA
803
24
MCHA
812
24
MCHA
1088
554
2509
1/A10
MBA-LBA? (Okçular ware)
2856
646
MBA
2956
1/A11
MBA-LBA? (Okçular ware)
3114
88
EB II-III
3124
814
MROM-LROM
3132
814
MROM-LROM
3168
88
EB II-III
3186
88
EB II-III
3615
88
EB II-III
3644
188
EB II-III
3646
188
EB II-III
3621
88
EB II-III
3740
88
EB II-III
4023
188
EB II-III?
4027
188
EB II-III?
4032
188
EIA
4169
188
EIA
4185
8/I6
LBA
4221
188
CHA/EB I?
4222
188
EIA
4228
188
EIA
MBA-LBA? (Okçular ware)
instrument’s calibration algorithms, Soil and Testall Geo, gave results that were generally in good agreement with each other; the results reported here were obtained with the latter calibration because it gave a wider range of elements (Jones and Campbell in press). NIST standards – 2709a, 2780 and Till 4 - were analysed. The count time in each analysis was 60 seconds. The chemical data for fourteen elements was treated with SPSS v. 21. For a discussion of the results of this analysis see section 16.3.
86
Developing an Adaptive Field Methodology for Challenging Landscapes
4.4.2 Lithic Procedures 4.4.2.1 Lithic Macroscopic Inspection and Analysis (Victor Klinkenberg and Bleda S. Düring) Chipped stone constitutes one of the most ubiquitous categories of surface artefacts in the CAP survey region. Stone suitable for chipping has been an important material for creating artefacts through the millennia. Given the fact that the technologies that were used in chipped stone industries changed considerably over time, technological properties can be used to date these artefacts. Further, the types of objects used and how they were produced can shed light on the activities that took place in the Cide and Şenparzar landscapes over time. Whereas organic material, and in some circumstances also ceramics, may deteriorate rapidly on the surface, lithics are typically well preserved. Moreover, in periods lacking ceramic traditions, lithic artefacts are often the only source of information, especially in the context of surface survey. All lithic artefacts were collected in the intensive and extensive components of CAP. In unit-walking and in grid-square collections, every identified lithic object was counted and collected. The density and spatial distribution of lithic material is highly varied throughout the different landscape types. On the one hand, the Abdulkadir region displays an even but low density distribution of lithics throughout the area, the Aybasan hillside, on the other hand, exhibits a clearly clustered pattern and a high density of lithic artefacts. In total, we documented 284 lithic objects, 120 of which were modern threshing sledge flints and 164 were considered ancient artefacts. All lithics were cleaned, drawn, photographed and described. Drawing of the lithic artefacts was conducted by selected researchers, since this requires a certain level of knowledge and training. Visual inspection and description was done in sunlight on a macro level with the use of a small ocular with a 10 x magnification by Victor Klinkenberg. The information recorded for the majority of the lithics is based on the technological and functional aspects of the objects. An exception was made for a number of threshing sledge elements. These objects are slotted in the base of a thick wooden board, which is pulled across harvested crops, separating straw and ears from the cereal grain in the process. Threshing sledge lithics are easily recognised as such because their edges are heavily rounded and glossy as a result of their use. They are typically produced from a medium coarse to coarse yellow coloured flint. All threshing sledge elements were photographed but only a selection was drawn and described. In the description of lithics, a major division is made between tools, blades and flakes, debitage and cores, and recorded characteristics include dimensions, completeness, grain size, colour, use wear, weathering and traces of burning (table 4.9). Furthermore, an attempt was made to classify the raw materials used. We identified obsidian, one type of quartz and 25 different types of flint. With this approach, we attempted to define as many ‘objective’ features of the artefacts as possible.
Analytical Recording Procedures and Approaches to Finds
87
This information was subsequently used in a GIS-based spatial analysis. Analysing the spatial distribution of all individual characteristics of lithics in a find spot (site) makes it possible to discern patterning, which otherwise may have been obscured. This has led for instance to the discovery of clear spatial patterning of lithics in the Okçular valley (section 5.3.1). Objects which are commonly associated with settlements such as cores and tools are located primarily around the rocky outcrop of Okçular Kale (S22), while flakes and blades, which are associated with sharpening and repairing flint objects, dominate in the surrounding landscape. At Aybasan Mahallesi, a technological assessment of the lithics indicates that, although the cluster contained a large variety of lithic materials, only one type of activity, flintknapping, occurred at the site. Traces of burning on some flints may indicate that the people working the stone here had set up a small camp with a fire. Apart from their use in a techno-functional analysis, general characteristics of flint artefacts can be used to date the objects to a broad time period such as the Palaeo-, Meso- and Neolithic (chapter 5). For more fine-grained dating, a local, and stratigraphically anchored, typology is needed, which is as yet absent for the CAP study area. The use of comparanda from more distant regions is problematic (section 5.1.1). With the exception of a few formal tools (projectile points and a backed bladelet), dates were assigned to the CAP lithics on the basis of whole assemblages rather than the typological characteristics of individual flint artefacts. In addition to their role in dating and the elucidation of activity patterns, lithics can also be used to provide evidence for long distance contact between communities in the CAP survey region and beyond. For this purpose, several obsidian fragments were exported for further investigation, which revealed a Galatian provenance for the objects (see below). Tab. 4.9: Main database fields and their explanation for the description of lithic artefacts. database field
description
Material
What material is the object made of? flint / quartz / obsidian.
Retouch
Is the object retouched? Yes: tool, no: non-tool.
Category
Tool / flake / blade / core / core trimming element / non-artefact.
ToolDescription In case the object is a tool: common designation of function based on shape (eg. ‘burin’ or ‘scraper’ ‘retouched blade’). GrainSize Average grain size of the lithic material. Very coarse to very fine. Weathering
None / low / medium / high.
UseWear
The presence of the following possible use wear traces: rounding / use retouch / gloss.
Potlids Craquelee
The presence of potlids (small round splinters sprung from surface of object due to heating). Is there any sign of craquelee (hairline cracks) on the surface of the object?
Patinated
Is there any sign of patination of the surface of the object?
Distal End
In case the distal end of the object is present, is it feather, step or hinge shaped?
Flint Type
Is the object similar to other lithics in colour, grainsize and inclusions?
Cortex
The percentage of the dorsal side covered in cortex.
88
Developing an Adaptive Field Methodology for Challenging Landscapes
4.4.2.2 Sourcing Obsidian (Bleda S. Düring and Bernard Gratuze) A total of six pieces of chipped obsidian were found during CAP. Four derive from Aybasan Yukarı Mahallesi (L136; L137; L193; and L224), and two from the cave of Okçular Kale İni (S24: L186 and L197). The discovery of obsidian chipped pieces, albeit in small numbers, is of considerable interest as it provides the opportunity to identify geological sources and trace geographical connections. Of the six pieces found, we were allowed to export three obsidian samples: two from Aybasan Mahallesi (L136 and L224), and one piece of obsidian which derives from the Chalcolithic site of Okcular Kale İni (L1974, see section 5.4.1 for a discussion of the site). Starting with the seminal publications by Renfrew and colleagues (1965; 1966; 1968), the reconstruction of obsidian distribution networks in the Prehistoric Mediterranean without doubt has been one of the great success stories of scientific archaeology in the 20th century (Williams-Thorpe 1995). In terms of mineral composition, each obsidian source is unique. This unique chemical ‘fingerprint’ makes it possible to identify the source of obsidian artefacts and debitage. There are many obsidian sources in Anatolia, but only a few appear to have been exploited in Prehistory (Yeğingil et al. 1998: 825-6). The most important known obsidian sources with evidence for Prehistoric exploitation include the Cappadocian sources at Göllüdağ, Nenezidağ and Acıgöl, and the Galatian outcrops at Sakaeli, Yağlar and Galatia-X (Keller and Seifried 1990; Yeğingil et al. 1998). There are further sources in eastern Anatolia, the Caucasus, and the Aegean (Cauvin et al. 1998). In order to investigate the origins of the obsidian found in Cide-Şenpazar the three samples were analysed by Bernard Gratuze at the Centre Ernest-Babelon, IRAMAT (CNRS/Université d’Orléans) by Laser Ablation Inductively Coupled Plasma Mass Spectrometry (LA-ICP-MS) (see Düring and Gratuze 2013 for full details). A comparison with the source reference dataset showed that the chemical composition of the three CAP samples (by combining successively the Ba-Zr, Y/ Zr-Nb/Zr and Fe-Ti contents or ratios) matches most closely the signatures of the Galatian sources (Fig. 4.14). In particular, the composition of the two artefacts from Aybasan (L136 and L224) match the composition of the Sakaeli source, near the town of Orta in Çankırı, whereas comparison with data published by Keller and colleagues (Keller and Seifried 1990; Keller et al. 1996; Poidevin 1998) shows that the piece from Okçular Kale İni (L197) probably derives from the Yağlar source, near the modern town of Kızılcahamam.
Dating Procedures
89
Fig. 4.14: Binary diagram of Ba/Zr-La/Th ratios for the Yağlar-Sakaeli/Orta obsidian outcrops and for the three studied artefacts. This diagram shows that one of the artefacts (L197) comes from the Yağlar region, while the two others (L136 and L224) come from the Sakaeli/Orta district. Produced by Bernard Gratuze.
4.5 Dating Procedures In this section we outline the methods used to date our recorded artefacts and features. Broadly speaking they were of two kinds: comparative typological and absolute dating.
4.5.1 Comparative Dating of Pottery, Lithics and Features Most of the artefacts found in the course of CAP were dated by comparisons with welldated assemblages from adjacent regions. The degree to which this was possible and the chronological resolution this provides differ dramatically from period to period. For example, most Prehistoric pottery traditions were produced over several hundreds of years, whereas certain Roman and Byzantine ceramic types can be dated within a few decades or a particular century. Such variations in chronological resolution between as well as within periods result in uneven and sometimes overlapping chronological ranges. This is reflected in the chronological labels used for the classification of CAP finds and the overlap between them (table 4.10). With no previous systematic archaeological work in Cide and Şenpazar, a major obstacle to the establishment of a fine-grained chronological sequence was the absence of excavated contexts, not only in the survey region and its immediate
90
Developing an Adaptive Field Methodology for Challenging Landscapes
Tab. 4.10: Period designations and chronological ranges. period
sub-period
Palaeolithic
Early Holocene
5
-1,000,000
-10,000
-1,000,000
-300,000
Middle Palaeolithic
MPAL
-300,000
-40,000
Upper Palaeolithic
UPAL
-40,000
-20,000
Epipalaeolithic
EPAL
-20,000
-10,000
Mesolithic
MESO
-10,000
-6000
Neolithic / Early Chalcolithic
NEO
-10.000
-5500
CHA
-5500
-3000
Middle Chalcolithic
MCHA
-5500
-4000
Late Chalcolithic
LCHA
-4000
-3000
EBA
-3000
-2000
Early Bronze Age I
EBAI
-3000
-2600
Early Bronze Age II
EBAII
-2600
-2300
Early Bronze Age III
EBAIII
-2300
-2000
Middle and Late Bronze Age
MBALBA
-2000
-1200
Middle Bronze Age
MBA
-2000
-1700
Late Middle to early Late Bronze Age
MBA-LBAI
-1700-
-1500
Late Bronze Age
LBA
-1600
-1200
Late Bronze Age II
LBAII
-1400
-1200
Iron Age
IRON
-1200
-330
Early Iron Age
EIA
-1200
-950
Middle Iron Age / Phrygian
PHRY
-950
-550
Late Iron Age / Achaemenid
LIAACH
-550
-330
HEL
-330
-1
9 10
Roman
ROM
AD 1
AD 650
Early Roman
EROM
AD 1
AD200
Middle Roman
MROM
AD 200
AD 400
Late Roman
LROM
AD 400
AD 650/700
Late Roman/Early Byzantine
LROM/EBYZ
-
-
Byzantine
6
7
8
BYZ
ca. 650
1500
Early Byzantine
EBYZ
ca. 650
900
Middle Byzantine
MBYZ
900
1300
Late Byzantine or Beylik period
LBYZ
1300
1500
11 and 12
OTT
1500
1920
13
MOD
1920
2012
OTHER
NULL
NULL
Ottoman
Other/unknown
chapter
LPAL
Hellenistic
Modern
end
PAL
Early Bronze Age
Iron Age
start
Lower Palaeolithic
Chalcolithic
Second Millennium BC
database code
Modern
11
Dating Procedures
91
surroundings, but across much of the Turkish western Black Sea region. Imported pottery, especially distinctive types with well-defined chronological spans such as Late Roman African Red Slip or Late Byzantine Sgraffito wares, are rare in the survey material. Thus, the vast majority of CAP pottery was dated through typological comparisons with stratified and published assemblages further afield. The closest relevant excavated sites especially for the Prehistoric, Bronze and Iron Ages (e.g. İkiztepe, Fikirtepe, Boğazköy, Troy, etc.) are located hundreds of kilometres away along the Black Sea coast or inland in central and western Turkey. This geographical distance, and the differences in cultural traditions this brings with it in many of the periods concerned, frequently render the dates derived from such comparisons tentative. Ceramic finds were dated, therefore, with degrees of confidence, expressed in percentages (cf. Bevan et al. 2013), by period specialists throughout the three field seasons and re-adjusted during a final study season in 2011. A further issue is the particular cultural perspective produced by a comparative approach that focuses attention on those ceramic types with obvious formal and stylistic connections to the well-documented cultural traditions of other regions and the excavated sites they are associated with. Local cultural traditions tend to fall by the wayside, since the identification and especially the dating of previously undocumented pottery types from surface assemblages is challenging to say the least and potential candidates tend to end up in the pile of ‘unknown’ sherds rejected by all period-specialists. One method which has been applied in a range of Mediterranean surveys to date local or otherwise undocumented types, is to treat the repeated co-occurrence with more securely dated finds as a proxy for more direct dating. The underlying hypothesis is that there is a high probability of chronological correlation of pottery types that are repeatedly found together with the same types of well-dated materials (Hayes 2000: 106; Given 2013a). The comparatively small sample size of CAP surface pottery and the limited number and rarity of closely datable types, however, make this technique unsuitable for the CAP material. The relatively manageable numbers of pottery collected by CAP made it possible for period-specialists, several of which were present throughout the duration of each field-season and during the final study season, to watch out for unusual pottery types or fabrics and particular patterns in their spatial distribution. It is in this manner that one previously undocumented and highly localised tradition was identified. Named ‘Okçular ware’ due to its overwhelming concentration in the fields surrounding Okçular Kale (S22), this pottery is discussed in more detail in chapter 7. The dating of lithic artefacts followed a very similar procedure to the pottery: comparison with assemblages from excavated or better understood collections from surrounding regions. However, whereas in the case of pottery formal characteristics form the main axis of comparison, in lithic studies the technical properties of artefacts and the way they were produced constitute the basis for comparison. In general, dating lithics through the comparative method is a very problematic exercise in Turkey, permitting only broad periodisations (see 5.1.2).
92
Developing an Adaptive Field Methodology for Challenging Landscapes
Some features such as castles and churches were also dated through comparative methods. Byzantine masonry, for example, changes considerably over time and certain construction elements, such as the presence and size of tiles, can be used as chronological indicators. Similarly, the size and layout of churches is chronologically sensitive (see chapter 11 for a detailed discussion).
4.5.2 Absolute Dating Four methods of absolute dating were used in CAP, including epigraphic, radiocarbon (C14), optically stimulated luminescence (OSL) and rehydroxilation methods. Absolute dates could be ascribed to some features on the basis of epigraphic evidence such as the Islamic calendar dates inscribed on Ottoman period gravestones. In 2011, we collected a series of radiocarbon samples from a section through a midden deposit left by looters at the back end of the cave of Okçular Kale İni (S24). Two AMS C-14 samples, which were analysed at the Centre for Isotope Studies, Groningen University, yielded consistent dates in the first half of the fifth millennium BC (section 5.4.1). The landscape taphonomy of the Cide coastal plain was investigated through a series of geological sections in 2010 (section 4.6) and two OSL samples of a palaeobeach deposit situated two metres below the present surface of the coastal plain were analysed at the Netherlands Centre for Luminescence dating, Delft University (section 16.1). The results indicate that the Cide coastal plain in its current form developed after 1850 BC (section 16.1). Rehydroxylation is a relatively new technique for dating ceramics (Wilson et al. 2009). An experimental pilot study on Prehistoric pottery from CAP, which was carried out by Eva Boeijen (2011) as part of her undergraduate dissertation research, did unfortunately not yield consistent results due to a range of unresolved technical uncertainties.
4.6 Geomorphological Research In the 2010 and 2011 season, Geuch de Boer undertook a geomorphological survey of parts of the Cide district in order to investigate alluvial, colluvial and erosion processes, and their chronology in the research region. The aim was to establish how the different landscapes of the Cide region had evolved and how this affected the preservation and accessibility of archaeological assemblages. In his work, de Boer used standard procedures of soil description as they are employed in the Netherlands by companies such as RAAP, augmented by knowledge from his training as soil engineer and his previous research in Asia. Exposed river sections, cuts for construction, gravel extraction and erosion gullies were investigated and described
Geochemical Research
93
in terms of texture (the balance between sand, clay and loam, and the sorting and granularity of the soil), colour (using a classification based on the Munsell Chart), and inclusions (e.g. shells). This geomorphological survey informed field strategies and methodological choices and its results are incorporated in discussions in chapter 2.
4.7 Geochemical Research (Richard Jones and Claudia Glatz)
The relationship between surface and sub-surface assemblages is a long-standing issue in survey archaeology (e.g. Schofield 1991) and a range of approaches have been developed to investigate this relationship. Geophysical remote sensing, for instance, can – in the right conditions – detect subsurface architectural remains and features. Chemical analysis and magnetic susceptibility measurements have also proven useful in the identification of human habitation and activity areas (Cavanaugh et al. 2005; Sarris and Jones 2000: 41-46). At the end of the 2011 field season, we collected 48 soil samples from a surface cluster of tiles, pottery and some glass fragments, which we sampled intensively in grid G6 in the previous year. The collected pottery from the cluster, together with two more concentrations along the northeastern edge of the valley, dates to the Roman and into the Byzantine period (chapter 11). Rims of medium sized jars and amphora handles and other diagnostic ceramics point towards habitation sites such as farmsteads, although fragments of what appear to have been glass bracelets could indicate the location of graves. The soil samples were collected in order to shed more light on the nature of the Roman to Early Byzantine presence in this part of the Okçular landscape (Fig. 4.15). The samples were taken using an iron augur, designed by the team and made by a local blacksmith, which was 5.5cm in diameter and was sunk into the ground to a depth of ca. 25cm. The first ca. 5cm of topsoil was discarded, the rest packed into plastic bags and taken to the project base. Core samples were taken from the centre points of the grid-squares of grid G6. We sampled two rows of grid-squares in N-S direction (down-hill) and three shorter series of squares in E-W direction (parallel with the contour lines) (Fig. 4.15). The eastern-most N-S row of samples cross-cuts two areas of high artefact densities, as does the northern most E-W sampling row. The other samples were taken in areas of low artefact densities and two outlier samples were collected on the way back to the road to the west of G6. At the project base, the soil samples were broken up and allowed to dry on newspaper for two days, before they were re-bagged ready for export to the UK. The weight of each sample was ca. 0.5kg. In Glasgow, a 100gm subsample was taken for sieving with a 2mm Endicott sieve and the powder collected in a small plastic bag. For three samples (15, 19 and 24) larger subsamples were prepared and analysed
94
Developing an Adaptive Field Methodology for Challenging Landscapes
Fig. 4.15: Map showing soil sampling locations in grid G6 superimposed over the sherd count distribution showing high density in dark shades.
separately to check for comparability. Chemical composition was determined by XRF using the same system and the same standards as for the pottery described above; the powdered sample was placed in a plastic vial covered with a Mylar window and analysed in at least two positions. Quantitative determinations of Zr, Sr, Rb, Pb, Cu, Zn, Fe, Mn, V, Ti, Sc, K, Ca and Ti were made. The molybdenum blue colorimetric method was employed to measure the total extractable phosphate concentration. Magnetic susceptibility was determined with a Bartington MS2 system at low and high frequencies on a known weight of sample. The results of these analyses are discussed in chapter 14 and section 16.4.
4.8 Archaeometallurgy
(Effie Photos-Jones and Claudia Glatz) The Pontic mountains are one of the most metal-rich areas of Turkey. However, in the Cide and Şenpazar region, mining for metals appears to have played little or no part in the local economy at least in recent history (sections 2.2.2 and 3.2). Few metal objects were recorded during surface survey, but intensive survey in the Okçular region in 2010 identified a concentration of vitreous material around 100-150m to the northeast of Okçular Kale (S22). Large quantities of small (d= ca. 1-5cm), rounded
Concluding Comments: Assessing the CAP Methodology
95
pieces resembling iron slag with smooth exterior surfaces and medium sized pores in section were found strewn across the area of Unit 593 and spilling into adjacent units, especially where erosion had exposed sub-surface layers. Three pieces were exported to Glasgow and the chemical composition of both sample surfaces and polished sections were analysed by Effie Photos Jones (Analytical Services for Art and Archaeology Scotland) using p-XRF (portable X-ray fluorescence analyser) and Scanning Electron Microscopy with Energy Dispersive Analyser (SEM-EDAX) respectively. The results of the analyses revealed iron oxides or oxy-hydroxides, namely iron ore, rather than iron slag, which, however, was also arsenic rich (1-2% As). These iron oxides would have been a coveted source of iron ore (c.55% -60% Fe), but would have generated, on smelting in a bloomery, arsenical iron rather than wrought iron. Arsenical iron can have good anti-corrosion properties but requires expertise in manufacture (PhotosJones 2012; see also chapter 14 for more detail).
4.9 Concluding Comments: Assessing the CAP Methodology The aims of the CAP methodology were, from the beginning, to maximise the information we could glean from what can only be described as a challenging landscape for archaeological investigation. What we needed was a sampling strategy that could yield sufficient systematic data to gain a meaningful understanding of the long-term cultural development of this archaeologically unknown area and answer a host of period-specific questions about the regions cultural identity, social organisation and connections with inland Anatolia and the Black Sea regions. To achieve these goals, our field methodology had to be capable of detecting the ephemeral traces of the region’s earliest inhabitants, which have so far eluded archaeological investigations in much of northern Turkey. From an initial aspiration to conduct a ‘siteless’ survey that was capable of revealing a more in-depth, landscape-scale understanding of past human presence and activity than traditional surveys in Turkey have been able to capture - tempered by a series of trials, errors and lessons learnt - we developed a new, hybrid and adaptive strategy. CAP’s repertoire of methods, we believe, builds on the strengths and buffers against some of the weaknesses of both Mediterranean-style intensive fieldwalking and extensive Near Eastern traditions to provide a multi-scalar window into the archaeological surface record of Cide and Şenpazar. The success of any method must be measured against the degree to which it permits a project to achieve its research goals (chapter 1). The richness and ‘diagnosticity’ (Bevan and Conolly 2013) of our collected artefacts and sites differ dramatically from one period to the next, as discussions in the following chapters will show in more detail. Overall, however, we are confident that the combination of extensive methods of investigation with an adapted intensive approach has served us well in achieving our research goals. The results from our extensive and intensive surveys complement
96
Developing an Adaptive Field Methodology for Challenging Landscapes
each other well with respect to the types of sites yielded by each method, enabling us to document a much broader spectrum of past human activities than either method would have allowed us to do on its own. Both approaches, for instance, yielded different types of locales with evidence for a Prehistoric and Bronze Age presence that has gone largely undocumented in the Turkish western Black Sea region thus far. Although highly variable in terms of the type, quality and quantity of data available for each period, we were able to collect sufficient information to begin to reconstruct the region’s long-term cultural development, or particular aspects thereof, and make some crucial observations about the region’s changing relationship with other parts of Turkey and the Black Sea (see chapters 14 and 15). Extensive reconnaissance involved both local guides and unguided vehicularbased explorations of the survey region. This method proved highly successful in that it yielded some of the most important sites recorded in the course of CAP, whose finds have fundamentally altered what we know about the region in the Chalcolithic, the Early Bronze Age and Early Iron Age, and dramatically increased our knowledge of the region’s Roman and Byzantine past. Its weakness lies in the narrow range of sites this approach tended to produce. Although we were interested in recording any type of human activity in the landscape and during all periods of occupation, we tended to be shown or came across a specific sub-set of sites that included caves (some with and others without signs of human occupation) and architectural remains (including castles, churches, bridges, wells and spolia). No surface scatters were shown to us by local residents, nor were any discovered during undirected extensive survey. The archetypical archaeological site that most extensive surveys in Turkey and the wider Near East tend to focus on, the highly visible mounded settlement sites known as höyük in Turkey, are entirely absent in the survey area and other parts of the central Turkish Black Sea region, most likely for cultural and environmental reasons (see discussions in chapters 5 to 7). The results from intensive unit-walking proved complementary to our extensive reconnaissance in that it allowed us to identify locales of past human activity and settlement that would have been entirely missed otherwise. The surface clusters of pottery and lithic materials as well as features otherwise hidden by vegetation, which we recorded during unit-walking date from the Early Holocene to the Ottoman period and again include finds that have dramatically altered our understanding of the region’s past. These include the discovery, for instance, of an Early Holocene knapping station at Aybasan (S82, S98, section 5.3.2), the only Late Bronze Age ceramic evidence recorded on the Black Sea coast (G8, section 7.2.2) as well as a series of Byzantine farmhouses or farmsteads at Okçular (S512-16; chapter 11). A follow-up grid-square collection at these locales allowed us to define more accurately these numerically low but culturally significant assemblages. On the downside, intensive survey was possible only in landscapes with particular topographic and ecological characteristics that provided higher groundvisibility than usual in the lush, green, mountainous environment typical for Cide
Concluding Comments: Assessing the CAP Methodology
97
and Şenpazar. Such conditions were encountered in recently occupied or abandoned open areas with low relief. In entirely flat areas, such as the Cide coastal plain and valley bottoms, much of the archaeological record, in particular remnants of the more distant past, lie buried beneath deposits accumulated through a range of different sedimentation processes. CAP’s two methodological strands were not only complementary, but throughout the three field seasons, results from extensive and intensive survey interactively fed back into the strategies for both. In particular, the results of targeted extensive reconnaissance were used to inform the selection of areas for intensive sampling. This facilitated a better understanding of the long-term development of key sites and their relationship with surrounding landscapes (chapter 14). A crude comparative measure of the viability of our approach is the density of sites recorded, overall survey coverage, and the representativeness of the recovered sample of sites for the wider region. As regards recorded site densities, large-scale Near Eastern surveys, whose research areas frequently cover 20,000km² or more and which report average site densities of one site per 60km² (Düring 2008: 36) occupy one end of the spectrum. Mediterranean-style intensive models which record average densities of about 4 sites per km² (Cherry 1983: fig. 1) traditionally cover survey areas of 100km² or less (Keos: 18km², Cherry et al. 1991; Antikythera: 20km², Bevan and Conolly 2012; Wadi Faynan: 40km², Barker et al. 2007; Kythera: 100km², Broodbank 1999). Exceptional larger Mediterranean surveys include the Troodos (182km², Given 2013a: 19) and the Nikopolis surveys (1,200km², Tartaron 2003). CAP, with its 929km² survey area, falls (at least in terms of size) in between the Mediterranean and Near Eastern model, and is perhaps most closely comparable to the larger Mediterranean surveys such as the Nikopolis project. With 83 recorded sites in the cluster and feature categories, CAP site density is on average one site per 11km2 and falls at the lower end of the distribution of intensive Mediterranean projects (Cherry 1983: fig. 1; Wilkinson et al. 2004: 190) but by far exceeds the number of sites typically recorded by traditional extensive surveys. In this regard CAP’s results are comparable to small, intensive, site-centred survey projects in Turkey such as the Gordion and Alişar surveys (Düring 2008: 36). CAP’s intensive coverage, however, is much lower than those of most Mediterranean surveys. The Kythera survey, for instance, intensively surveyed over 40% of its survey area,6 the Troodos survey covered ca. 20% (Given 2013a), while Antikythera was sampled in its entirety (Bevan and Conolly 2013). Such figures certainly dwarf the area covered by intensive fieldwalking by CAP, which amounts to 3.44km2 or 0.4% of the total survey area. Although minute in absolute terms, when considered in the context of the region’s topographic and environmental context
6 http://www.ucl.ac.uk/kip
98
Developing an Adaptive Field Methodology for Challenging Landscapes
and in combination with the spatial coverage of our more extensive explorations, this figure becomes somewhat more relevant. Major limiting factors include the mountainous terrain and the dense forest cover, which make about three-quarters of the Cide and Şenpazar ilceler inaccessible to intensive methods of investigation and severely restrict extensive exploration in logistical terms. Of course there are always areas to be improved upon or avenues to be explored more fully than time, budget or other factors allowed for. Such areas include a more systematic extensive coverage less focused on the knowledge of Cide residents, the use of surface clearance and test pits, more extensive use of soil analysis as well as canopy-penetrating LiDAR imaging. Overall, however, we feel that we have gained a good understanding of the archaeological record of our research region that is detectable with the range of methods employed and in the specific environmental and topographic contexts of Cide and Şenpazar.
Bibliography websites http://cgma.depauw.edu/MAGIS/ (last accessed 25.09.2013) https://www.princeton.edu/avkat/(last accessed 25.09.2013) http://www.ucl.ac.uk/kip/(last accessed 25.09.2013) http://www.ucl.ac.uk/asp/(last accessed 25.09.2013) published sources Adams, R. McC. 1965. Land Behind Baghdad. A History of Settlement on the Diyala Plains. Chicago, University of Chicago Press. — 1981. Heartland of Cities. Surveys of Ancient Settlement and Land Use on the Central Floodplain of the Euphrates. Chicago, University of Chicago Press. Adams, R. McC. and Nissen, H.J. 1972. The Uruk Countryside. The Natural Setting of Urban Societies. Chicago, University of Chicago Press. Alcock, S. and Cherry J.F., eds. 2004. Side-by-Side Survey: Comparative Regional Studies in the Mediterranean World. Oxford, Oxbow. Athanassapolous, E., and Wandsnider, L., eds. 2004. Mediterranean Archaeological Landscapes: Current Issues. Philadelphia, University of Pennsylvania Museum of Archaeology and Anthropology. Baird, D. 2001. Settlement and landscape in the Konya Plain, south central Turkey. From the Epipalaeolithic to the Medieval period. In A. Basim, ed., Çatalhöyük’ten Günümüze Çumra Kongresi. Çumra, Çumra Belediyesi: 269-76. Banning, E.B. 2002. Archaeological Survey. New York, Kluwer. Barker, G., Gilbertson, D., and Mattingly, D., eds. 2007. Archaeology and Desertification. The Wadi Faynan Landscape Survey, Southern Jordan. Oxford, Oxbow. Belke, K. 1996. Paphlagonien und Honōrias. Vienna, Österreichische Akademie der Wissenschaften. Bevan, A. and Conolly. J. 2004. GIS, archaeological survey and landscape archaeology on the island of Kythera, Greece. Journal of Field Archaeology 29: 123-38.
Bibliography
99
— 2012. Intensive survey data from Antikythera, Greece. Journal of Open Archaeology Data, 1: DOI: http://dx.doi.org/10.5334/4f3bcb3f7f21d — 2013. Mediterranean Islands, Fragile Communities and Persistent Landscapes. Cambridge, Cambridge University Press. Bevan, A., Conolly, J., Hennig, C., Johnston, A., Quercia, A., Spencer, L., and Vroom, J. 2013. Measuring chronological uncertainty in intensive survey finds: A case study from Antikythera, Greece. Archaeometry 55: 312-28. Bikoulis, P. 2012. Revisiting prehistoric sites in the Göksu Valley: A GIS and social network approach. Anatolian Studies 62: 35-59. Bilgi, Ö., Dönmez, S. and Ulugergerli, E.U. 2004. Samsun ili 2003 dönemi yüzey araştırması. Araştırma Sonuçları Toplantısı 22/1: 115-24. Bintliff, J., and Howard, P. 2004. A radical rethink on approaches to surface survey and the rural landscape of central Greece in Roman times. In F. Kolb, ed., Chora und Polis. München, Oldenbourg Verlag: 43-78. Bintliff, J., Howard, P., and Snodgrass, A. 1999. The hidden landscape of prehistoric Greece. Journal of Mediterranean Archaeology 12: 139-68. — 2007. Testing the Hinterland. The Work of the Boeotia Survey (1989-1991) in the Southern Approaches to the City of Thespaiai. Cambridge, McDonald Institute/British School at Athens. Bintliff, J., and Snodgrass, A. 1988. Off-site pottery distributions: A regional and interregional perspective. Current Anthropology 29: 506-13. Blanton, R.E. 2000. Hellenistic, Roman and Byzantine Settlement Patterns of the Coast Lands of Western Rough Cilicia. Oxford, Archaeopress. — 2001. Mediterranean myopia. Antiquity 75: 627-9. Boeijen, E. 2011. Rehydroxylatie. Een Aardewerkdateringsmethode Toegepast op het Aardewerk van het Cide Archaeological Project. Unpublished BA thesis, Leiden University. Branting, S. 1996. The Alişar regional survey 1993-1994: A preliminary report. Anatolica 22: 145-58. Broodbank, C. 1999. Kythera Survey: Preliminary report on the 1998 season. Annual of the British School at Athens 94: 191-214. Burney, C. 1956. Northern Anatolia before Classical times. Anatolian Studies 6: 179-203. Casana, J., and Cothren, J. 2008. Stereo analysis, DEM extraction and orthorectification of CORONA satellite imagery: Archaeological applications from the Near East. Antiquity 82: 732-49. Casana, J., Cothren, J. and Kalayci, T. 2012. Swords into ploughshares: Archaeological applications of CORONA satellite imagery in the Near East. Internet Archaeology 32(2): http://intarch. ac.uk/journal/issue32/2/toc.html Casana, J., and Wilkinson, T.J. 2005. Settlement and landscapes in the Amuq region. In K.A. Yener, ed., The Amuq Valley Regional Projects I. Surveys in the Plain of Antioch and Orontes Delta, Turkey, 1995-2002. Chicago, University of Chicago Press: 25-65. Cauvin, M.-C., Gourgaud, A., Gratuze, B., Arnaud, N., Poupeau, G., Poidevan, J.-P., and Chataigner, C., eds. 1998. L’obsidienne au Proche et Moyen Orient: Du volcan à l’outil. Oxford, Archaepress. Cavanaugh, W., Jones, R. E. and Sarris, A. E. 1996. The phosphate geophysical surveys. In W. Cavanaugh, J. Crouwel, R. Catling, and G. Shipley, eds.,The Laconia Survey: Continuity and Change in a Greek Rural Landscape 2. London, British School at Athens: 235-262. Cherry, J.F. 1983. Frogs round the pond: Perspectives on current archaeological survey projects in the Mediterranean region. In D.R. Keller and D.W. Rupp, eds., Archaeological Survey in the Mediterranean Region. Oxford, British Archaeological Reports: 375-416. — 1994. Regional survey in the Aegean: The ‘new wave’ (and after). In P.N. Kardulias, ed., Beyond the Site: Regional Studies in the Aegean Area. Lanham, University Press of America: 91-122.
100
Developing an Adaptive Field Methodology for Challenging Landscapes
Cherry, J.F., Davis, J.L. and Mantzourani E., eds., 1991. Landscape Archaeology as Long-Term History: Northern Keos in the Cycladic Islands from Earliest Settlement until Modern Times. Los Angeles, Cotsen Institute of Archaeology. Czichon, R. 2000. Das Hattusa / Boğazköy Survey Project. In P. Matthiae, A. Enea, L. Peyronel and F. Pinnock, eds., Proceedings of the First International Congress of the Archaeology of the Ancient Near East. Rome, La Sapienza: 269-77. Donceel-Voûte, P. 1977. Gökırmak valley, 1977. Anatolian Studies 27: 196-7. Dönmez, S. 1999. Sinop-Samsun-Amasya İlleri 1998 yüzey araştırması. Araştırma Sonuçları Toplantısı 17/2: 229-44. Doonan, O.P. 2004. Sinop Landscapes: Exploring Connections in a Black Sea Hinterland. Philadelphia, University of Pennsylvania Museum of Archaeology and Anthropology. Dunnell, R.C. 1992. The notion site. In J. Rossignol and L.A. Wandsnider, eds., Space, Time and Archaeological Landscapes. New York, Plenum Press: 21-37. Dunnell, R.C., and Dancey, W.S. 1983. The siteless survey: A regional scale data collection strategy. Advances in Archaeological Method and Theory 6: 267-87. Düring, B.S. 2008. The early Holocene occupation of north-central Anatolia between 10,000 and 6,000 BC cal.: Investigating an archaeological terra incognita. Anatolian Studies 58: 15-46. Düring, B.S., and Glatz, C. 2010. The Cide Archaeological Project 2009: First results. Anatolia Antiqua 18: 203-13. Düring, B.S., Glatz, C., and Şerifoğlu, T.E. 2012. The Cide Archaeological Project 2011: Third preliminary report. Anatolia Antiqua 20: 167-75. Düring, B.S., and Gratuze, B. 2013. Obsidian exchange networks in prehistoric Anatolia: New data from the Black Sea region. Paléorient 39 (2): 173-182. Ebert, J.I. 1992. Distributional Archaeology. Albuquerque, University of New Mexico Press. Efe, T. 1997. 1995 yılında Kütahiya, Bilecik ve Eskişehir İllerinde yapılan yüzey araştırmaları. Araştırma Sonuçları Toplantısı 14/2: 215-32. Erciyas, D.B., and Sökmen, E. 2010. An overview of Byzantine period settlements around Comana Pontica in north-central Turkey. Byzantine and Modern Greek Studies 34/2: 119-41. Ersoy, Y., and Koparal, E. 2010. Urla ve Seferihisar İlçeleri yüzey araştırması 2008 yılı çalışmaları. Araştırma Şonuçları Toplantısı 27/2: 127-43. Evans, D., Fletcher, R., Pottier, C., Chevance, J.-B., Soutif, D., Tan, B.S., Im, S., Ea, D., Tin, T., Kim, S., Cromarty, C., De Greef, S., Hanus, K., Bâty, P, Kuszinger, R., Shimoda, I., and Boornazian, G. 2013. Uncovering archaeological landscapes at Angkor using Lidar. Proceedings of the National Academy of Sciences: doi 10.1073/pnas.1306539110 French, D.H.1965. Prehistoric sites in the Göksu valley. Anatolian Studies 15: 177-201. — 1967. Prehistoric sites in northwest Anatolia, I. The İznik area. Anatolian Studies 17: 49-100. — 1969. Prehistoric sites in northwest Anatolia, II. The Balıkesır and Akhısar / Manısa areas. Anatolian Studies 19: 41-98. Given, M. 2007. Mountain landscapes on early modern Cyprus. Hesperia 40: 137-48. — 2013a. Methodology and analytical procedures. In M. Given, A.B. Knapp, J. Noller, L. Sollars, and V. Kassianidou, eds., Landscape and Interaction: The Troodos Archaeological and Environmental Survey Project, Cyprus I. Methodology, Analysis and Interpretation. London, Council for British Research in the Levant: 16-54. — 2013b. Commotion, collaboration, conviviality: Mediterranean survey and the interpretation of landscape. Journal of Mediterranean Archaeology 26(1): 3-26. Given, M., and Knapp, A.B., eds. 2003. The Sydney Cyprus Survey Project: Social Approaches to Regional Archaeological Survey. Los Angeles, The Cotsen Institute of Archaeology. Glatz, C., Düring, B.S., and Şerifoğlu, T.E. 2011. The Cide Archaeological Project 2010: Second preliminary report. Anatolia Antiqua 19: 279-88.
Bibliography
101
Glatz, C., and Matthews, R. 2005. Anthropology of a frontier zone: Hittite-Kaska relations in Late Bronze Age North-Central Anatolia. Bulletin of the American Schools of Oriental Research 339: 21-39. Hayes, J.W. 2000. The current state of Roman ceramic studies in Mediterranean survey, or, handling pottery from surveys. In R. Francovich and H. Patterson, eds., Extracting Meaning from Ploughsoil Assemblages. Oxford, Oxbow: 105-109. Işın, M.A. 1998. Sinop region field survey. Anatolia Antiqua 6: 97-139. Jones, R. and Campbell, L. in press. Non-destructive analysis of Samian ware from Scottish military sites. In S. Coxon, E. Sibbesson and B. Jervis, eds., Insights from Innovation. Proceedings of Conference in Honour of David Peacock. Oxford, Oxbow. Kealhofer, L. 2005. The Gordion regional survey, settlement and land use. In L. Kealhofer, ed., The Archaeology of Midas and the Phrygians: Recent Work at Gordion. Philadelphia, University of Pennsylvania Museum of Archaeology and Anthropology: 137-48. Kalinka, E. 1933. Aus Bithynien und Umgebung. Jahreshefte des Österreichischen Archäologischen Institutes in Wien 28: 45-122. Kanter, C.J. 2008. The archaeology of regions: From discrete analytical toolkits to ubiquitous spatial perspective. Journal of Archaeological Method and Theory 15: 167-89. Karasalihoğlu, M. 2009. Cide’nin tarihi yeniden yazılacak. Kastamonu Gazetesi, 20 Ağustos (August): 1, 11. — 2010. Cide Arkeoloji Projesi, Cide ve Şenpazar’da çok çarpıcı keşifler imza attı: Buluntular MÖ 20 bini gösterdi. Kastamonu Gazetesi, 24 Temmuz (July): 1, 6 -7. — 2011. Antik Cide’nin kapısı aralandı. Kastamonu Posta, 28 Temmuz (July): 7. Karauğuz, G. 2007. Zonguldak İli ve İlçeleri yüzey araştırması 2006. Araştırma Şonuçları Toplantısı 25/1: 43-54. Keller, J., and Seifried, C. 1990. The present status of obsidian source identification in Anatolia and the Near East. PACT 25(4): 57-87. Keller, J., Bigazzi, G., and Pernicka, E. 1996. The Galatia-X source. A combined major-element, trace-element and fission-track characterization of an unknown obsidian source in northwestern Anatolia. In S. Demirçi, A.M. Özer, and G.D. Summers, eds., Archaeometry 94, The Proceedings of the 29th International Symposium on Archaeometry. Ankara, Tübitak: 543-51. Kızıltan, Z. 1992. Samsun bölgesi yüzey araştırmaları. Belleten 56(215): 213-41. Kuzucuoğlu, C., Marro C., Özdoğan, A., and Tibet, A. 1997. Prospection archéologique francoturque dans la région de Kastamonu (mer Noire). Deuxième rapport préliminaire. Anatolia Antiqua 5: 275-306. Llobera, M., Wilkinson, K.N., Weiss, M.C., Flaming, R.J., Marini, N.A.F., and Mazet, S. 2010. Into the maquis: Methodological and interpretational challenges in surveying La Balagne, northwest Corsica. Journal of Mediterranean Archaeology 23: 169-96. Marek, C. 2003. Pontus et Bithynia, Die römischen Provinzen im Norden Kleinasiens. Mainz am Rhein, Philipp von Zabern. Marro, C., Özdoğan A., and Tibet, A. 1996. Prospection archéologique franco-turque dans la région de Kastamonu (mer Noire). Premier rapport préliminaire. Anatolia Antiqua 4: 273-90. — 1998. Prospection archéologique franco-turque dans la région de Kastamonu (mer Noire). Troisième rapport préliminaire. Anatolia Antiqua 6: 317-35. Matthews, R. 2009. Project Paphlagonia: Research issues, approaches and methods. In R. Matthews and C. Glatz, eds., At Empires’ Edge: Project Paphlagonia. Regional Survey in North-Central Turkey. London, British Institute at Ankara: 1-25. Matthews, R., and Glatz, C., eds. 2009. At Empires’ Edge: Project Paphlagonia. Regional Survey in North-Central Turkey. London, British Institute at Ankara.
102
Developing an Adaptive Field Methodology for Challenging Landscapes
Mattingly, D.J. 2000. Methods of collection, recording and quantification. In R. Francovich, H. Patterson, and G. Parker, eds., Extracting Meaning from Ploughsoil Assemblages. Oxford, Oxbow: 5-15. Mellaart, J. 1954. Preliminary report on a survey of pre-Classical remains in southern Turkey. Anatolian Studies 4: 175-239. — 1959. Archaeological survey of the Konya plain. Anatolian Studies 9: 31-3. Menze, B., Ur, J., and Sherratt, A. 2006. Detection of ancient settlement mounds: Archaeological survey based on the SRTM terrain model. Photogrammetric Engineering and Remote Sensing 72(3): 321-7. Mortensen, P. 1974. A survey of prehistoric sites in northern Luristan. Acta Archaeologica 45: 1-47. Omura, S. 2005. Preliminary report of the 2004 general survey in central Anatolia. KamanKalehöyük - Anatolian Archaeological Studies 14: 55-84. Özdoğan, A., Marro, C., Tibet A., and Kuzucuoğlu, C. 2000. Kastamonu yüzey araştırması 1998 yılı çalışmaları. Araştırma Sonuçları Toplantısı 17/2: 41-56. Özdoğan, M. 1977. The Lower Euphrates Basin 1977 Survey. Istanbul, Middle East Technical University Lower Euphrates Publications. Özsait, M. 2005. 2004 yılı Samsun ve Amasya illeri yüzey araştırmaları. Araştırma Sonuçları Toplantısı 23/2: 249-58. Özsait, M. and Özsait, N. 2010. Amasya yüzey araştırmaları. Araştırma Sonuçları Toplantısı 28/2: 25-40. Parcak, S.H. 2009. Satellite Remote Sensing for Archaeology. London, Routledge. Photos-Jones, E. 2012. Non-destructive analysis of slag from Cide, Turkey. Unpublished analytical report. Poidevin, J.-L. 1998. Les gisements d’obsidienne de Turquie et de Transcaucasie: géologie, geéochimie et chronométrie. In M.-C. Cauvin, A. Gourgaud, B. Gratuze, N. Arnaud, G. Poupeau, J.-L. Poidevin and C. Chataigner, eds., L’obsidienne au Proche et Moyen Orient: du volcan à l’outil. Oxford, Archaeopress: 105-203. Pournelle, J.R. 2007. KLM to CORONA: A bird’s-eye view of cultural ecology and early Mesopotamian urbanization. In E.C. Stone, ed., Settlement and Society: Essays Dedicated to Robert McCormick Adams. Los Angeles, Cotsen Institute of Archaeology: 29-62. Renfrew, C., Cann, J.R., Dixon, J.E. 1965. Obsidian in the Aegean. Annual of the British School at Athens 60: 225-47. Renfrew, C., Dixon, J.E., and Cann, J.R. 1966. Obsidian and early cultural contact in the Near East. Proceedings of the Prehistoric Society 32: 30-72. — 1968. Further analysis of Near Eastern obsidians. Proceedings of the Prehistoric Society 34: 319-31. Rupp, D. 2004. Evolving strategies for investigating an extensive terra incognita in the Paphos district by the Canadian Palaipaphos Survey Project and the Western Cyprus Project. In M. Iacovou, ed., Archaeological Field Survey in Cyprus: Past History, Future Potentials. London, British School at Athens: 63-76. Sarris, A. and Jones, R.E. 2000. Geophysical and related techniques applied to archaeological survey in the Mediterranean: A review. Journal of Mediterranean Archaeology 13(1): 3-75. Schiffer, M.B., Sullivan, A.P., and Klinger, T.C. 1978. The design of archaeological surveys. World Archaeology 10: 1-28. Schofield, A.J., ed. 1991. Understanding Artefact Scatters: Contributions to Ploughzone Archaeology. Oxford, Oxbow. Seton-Williams, M.V. 1954. Cilician Survey. Anatolian Studies 4: 121-74. Shaw, B.D. 1990. Bandit highlands and lowland peace: The mountains of Isauria-Cilicia. Journal of the Economic and Social History of the Orient 33: 199-233.
Bibliography
103
Sipahi, T. 2002. 2001 yılı Çorum ve Çankiri bölgeleri yüzey araştırması. Araştırma Sonuçları Toplantısı 20/2: 275-84. Tartaron, T.F. 2003. The archaeological survey: sampling strategies and field methods. In J. Wiseman and K. Zachos, eds., Landscape Archaeology in Southern Epirus, Greece I. The Nikopolis Project. Athens, American School of Classical Studies in Athens: 23-45. Tsetskhladze, G.R., Adams, J., Avram, A., Avram, S., Dandrow, E., Madden, A., Maranzana, P., Naserdin, L. and Schmidt, A. 2012. Pessinus in Phrygia: Brief preliminary report of the 2011 field season. In G.R. Tsetskhladze, ed., The Black Sea, Paphlagonia, Pontus and Phrygia in Antiquity: Aspects of Archaeology and Ancient History. Oxford, Archaeopress: 337-341. Ur, J.A. 2002. Settlement and landscape in northern Mesopotamia: The Tell Hamoukar Survey 2000-2001. Akkadica 123: 57-88. — 2003. CORONA satellite photography and ancient road networks: A northern Mesopotamian case study. Antiquity 77: 102-15. von der Osten, H.H. 1927. Explorations in Hittite Asia Minor: A Preliminary Report. Chicago, University of Chicago Press. Wandsider, L.A., and Camilli, E. 1992. The character of surface archaeological deposits and its influence on survey accuracy. Journal of Field Archaeology 16: 31-46. Whallon, R. 1979. An Archaeological Survey of the Keban Reservoir Area of East-Central Turkey. Ann Arbor, University of Michigan, Museum of Anthropology. Wilkinson, T.J. 2003. Archaeological Landscapes of the Near East. Tucson, University of Arizona Press. Wilkinson, T.J. and Tucker, D. 1995. Settlement Development in the North Jazira, Iraq: A Study of the Archaeological Landscape. Warminster: British School of Archaeology in Iraq. Wilkinson, T.J., Ur, J. and Casana, J. 2004. From nucleation to dispersal: Trends in settlement pattern in the northern Fertile Crescent. In Cherry, J. and Alcock, S., eds., Side-by-Side Surveys in the Mediterranean World: Oxford: Oxbow: 198-205. Williams-Thorpe, O. 1995. Obsidian in the Mediterranean and the Near East, A provenancing success story. Archaeometry 37: 217-48. Wilson, M.A., Carter, M.A., Hall, C., Hoff, W.D., Ince, C., and Savage, S.D. 2009. Dating fired-clay ceramics using long-term power law rehydroxylation kinetics. Proceedings of the Royal Society online: 1-9. Yakar, J. 2000. Ethnoarchaeology of Anatolia. Rural Socio-Economy in the Bronze and Iron Ages. Tel Aviv, The Institute of Archaeology. Yeğingil, Z., Bigazzi, G., Poupeau, G. and Bellot-Gurlet, L. 1998. Provenance studies of obsidian artefacts in Anatolia: The contribution of the fission track analyses. In M.J. Mellink, G. Arsebük, and W. Schirmer, eds., Light on Top of the Black Hill: Studies Presented to Halet Çambel. Istanbul, Ege Yayınları: 823-44.
Bleda S. Düring and Victor Klinkenberg
5 The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC One of the prime research objectives for the Cide Archaeological Project (CAP) was to identify Prehistoric assemblages predating the Early Bronze Age, which are poorly documented in northern Turkey for a variety of reasons (see below). In this chapter, we will discuss the sites and assemblages documented by the Cide Archaeological Project that can be dated to the period between about 10,000 and 3000 BC. Although relatively small assemblages were found in the three years of archaeological surface survey in the Cide region, these artefacts are nonetheless important given the present paucity of data on the Prehistory of northern Turkey. Moreover, in this chapter, we will discuss cultural and material links of these assemblages with inland Anatolia and the Circumpontic.
5.1 The Problem of the Early to Mid-Holocene Occupation of Northern Turkey The Prehistory of northern Asia Minor, which includes both the Pontus and the interior plateau north of the Tüz Gölü, is very poorly documented. In particular the period between 10,000 and 5500 BC, for which we will use the label ‘Early Holocene’, is effectively terra incognita in northern Turkey. The database of the Türk Arkeoloji Yerleşmeleri project (www.tayproject.org) provides very few sites dating to the Palaeolithic and Neolithic periods. Likewise, the period between 5500 – 3000 BC, the ‘Middle Holocene’, is poorly documented as indexed by the scarcity of Chalcolithic sites in the Türk Arkeoloji Yerleşmeleri database. In this chapter, we will be using a terminology that differs from that traditionally used in Anatolian archaeology (see table 5.1). Conventionally, archaeologists have dated artefacts to periods such as the Epipalaeolithic, Mesolithic and Neolithic. The reason why such terminologies are avoided here is that they conflate typochronological information and assessments of the economy and society of the people that produced them. Thus, for example calling an artefact ‘Neolithic’ implies that we are dealing with a more or less sedentary farming community. In the Cide-Şenpazar region we found lithic artefacts that have parallels with both Neolithic sites in south-central Anatolia and others that appear similar to those found in the Ağaçlı Mesolithic of the Marmara region. In the absence of any excavated sites in northcentral Anatolia, it is impossible, however, to assess whether the lithics which we found in the Cide-Şenpazar region were produced by farmers or hunter-gatherers. It should be emphasised that we completely lack any ceramics from this period, and that in terms of typology we have very few microliths – the criterion commonly used
The Problem of the Early to Mid-Holocene Occupation of Northern Turkey
105
to assign artefacts to the Mesolithic. Some of our colleagues would prefer to label our Early Holocene assemblages as ‘Aceramic Neolithic’ but this terminology has its own problems. Apart from the fact that the Aceramic Neolithic remains a poorly defined and possibly non-existent entity in western and northern Turkey (Düring 2011: 125-7), we have no data at present to suggest that people were indeed farming in the Cide-Şenpazar region in the Early Holocene. Therefore, in order to avoid suggestive terminologies, this chapter uses the somewhat awkward ‘Early Holocene’ as a neutral period designation to discuss assemblages that in all likelihood date to the period between 10,000 and 5500 BC. Tab. 5.1: Period terminologies used in this chapter compared to traditional terminologies employed in Anatolian archaeology.
absolute dates BC
traditional terminology
this chapter
3000-2000 4000-3000 5500-4000 10,000-5500 10,000-6000 20,000-10,000 1,000,000-20,000
Early Bronze Age Late Chalcolithic Middle Chalcolithic Neolithic Mesolithic Epipalaeolithic Palaeolithic
Early Bronze Age Late Chalcolithic Middle Chalcolithic Early Holocene Early Holocene Palaeolithic
In an earlier study Bleda Düring (2008) scrutinised the possible reasons for why we know little to nothing about the Early Holocene occupation of north-central Anatolia. Many of the arguments presented in that study are equally pertinent for the Turkish Black Sea region, and therefore the discussion will be briefly recapitulated here. Various arguments can be put forward to explain this hiatus in our knowledge. These include: first, a real absence of occupation, an argument that is usually linked with the idea that the region had an adverse climate; second, that the area was inhabited but that various factors, such as a mobile lifestyle and the use of wood have resulted in an ephemeral archaeological ‘footprint’; third, that the visibility and preservation of sites in the north is poor; and fourth, that too little (intensive) archaeological surface survey work had been undertaken in the area. A genuine absence of Early Holocene occupation in northern Anatolia due to an adverse climate is implausible, given that the climate was not less favourable than that of other regions for which we have evidence of occupation in the Early Holocene (Düring 2008: 25-30; 2011: 229-30). Instead, it is plausible that Early Holocene sites in much of northern Anatolia were most likely small, ephemeral, and short-lived. A good model for what such sites might have looked like is now available through the excavations of Çamlıbel Tarlası, a Late Chalcolithic site located near the Hittite capital Hattusa (Schoop 2011). Çamlıbel Tarlası is a small hamlet site measuring about 50
106
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
by 50 metres and consisted of a few houses at any one point in time. The settlement was regularly abandoned and reoccupied in the course of a few centuries, suggesting periodic relocation of this small settlement. The site deposits were relatively shallow and have been severely affected by erosion. It is clear that sites of this type are difficult to find and can only be located through intensive survey techniques in which landscape formation processes are taken into account. If the site of Çamlıbel Tarlası, rather than the large mound site of İkiztepe on the Bafra plain (Alkım et al. 1988; 2003), is representative of what sites across much of northern Turkey would have looked like in Prehistory, it is perhaps not surprising that few such sites have been discovered in the relatively extensive surveys that have typically been undertaken in Turkey. Many survey cover territories of some 20,000km² in size and have site densities of 1 site per 60km² (Düring 2008: 36, chapter 4). With such methodologies only large conspicuous sites can be documented. Obviously, if large Prehistoric sites were present in northern Turkey, they would have been identified by now. The working hypothesis in our survey work therefore was that the only way that the small and ephemeral Prehistoric sites could be located in an environment with low visibility is through intensive and targeted survey techniques: something which had not been previously trialled in the Turkish Black Sea region. This idea can be supported by evidence from Cyprus, where the Aceramic Neolithic was discovered relatively recently due to intensive investigations (Swiny 2001; Peltenburg and Wasse 2004), and western Turkey, where Neolithic sites have also been discovered recently as a result of intensive investigations and urban and industrial developments (Lichter 2005).
5.1.1 Dating Prehistoric Assemblages Apart from the issue of how small Prehistoric sites can best be discovered in the challenging landscapes of the Turkish western Black Sea region, a second problem is how Prehistoric assemblages of this poorly investigated area should be dated. The problem is equally pertinent for lithic and ceramic assemblages. Ceramic assemblages predating the Late Chalcolithic are very difficult to date in northern Anatolia. This is due to: first, the lack of robust excavated Prehistoric sequences in the area (Düring 2011: 229-35); second, the fact that assemblages that have been found in the region are difficult to relate to each other because of a lack of overlapping formal criteria that would allow for seriation; third, the distinctiveness of the ceramic assemblages in the region which has hampered dating through comparison with better known collections from the Marmara region and south-central Anatolia; and, fourth, the lack of absolute dates to determine the chronological position of the various assemblages. Various efforts have been undertaken to create a chronological ordering of the ceramic assemblages of northern Anatolia predating the Late Chalcolithic by comparing
The Problem of the Early to Mid-Holocene Occupation of Northern Turkey
107
these assemblages with excavated sequences in other parts of Anatolia. On the basis of the same assemblages from north-central Anatolia and the same methodology of formal comparison with adjacent regions, a series of mutually exclusive chronologies have been put forward (Bittel 1934; von der Osten 1937; Bittel 1950; Orthmann 1963; Parzinger 1993; Thissen 1993; Schoop 2005). Obviously, the methodology of formal comparisons of ceramic assemblages with those of surrounding regions is of limited use for establishing the Prehistoric sequence of northern Anatolia. To illustrate this point, the thorough study by Schoop (2005) Das anatolische Chalkolithikum can be mentioned, in which he dates the distinct assemblages of Büyükkaya and Yarıkkaya to the Early Chalcolithic, between 6000-5500 BC. Subsequently, Schoop excavated a site with Yarıkkaya ceramics at Camlıbel Tarlası, and updated his chronological assessment to the mid-fourth millennium BC on the basis of new radiocarbon dates (Schoop 2008), thus two millennia later than previously thought. The lithic assemblages of northern Anatolia are likewise difficult to date. There are of course many Prehistoric sequences in other parts of Turkey, but our data are biased in numerous ways. For example, in south-central Anatolia, lithic assemblages have been well investigated for the Neolithic periods, but very little has been published for the Chalcolithic and Early Bronze Age periods, when the focus shifts to ceramic assemblages. In the Lake District and Aegean Anatolia, little has likewise been published on the lithic industries. By contrast, a good synthetic study is available for Neolithic lithic assemblages in the Marmara region and Thrace (Gatsov 2009), although, again, we know little about the subsequent Chalcolithic and Early Bronze Age lithic industries. Although some publications exist (Balkan-Atlı 1994; Shimelmitz 2011), comprehensive and synthetic studies of lithic assemblages in Prehistoric Anatolia are absent, and this makes their comparison challenging. The lack of synthetic studies is perhaps the main reason why comparative dates on the basis of chipped stone artefacts, such as that of Sokukayası in the Aceramic Neolithic, are controversial (Marro 2000: 951). Moreover, it is unclear to what degree lithic industries in northern Anatolia are similar to those in adjacent regions, and whether formal similarities can be used to date assemblages through a comparative methodology. A firmer chronology of Prehistoric ceramic assemblages in northern Anatolia could be achieved through the use of absolute dating techniques, preferably in combination with excavations. Unfortunately, most excavated sequences in the region are of limited use for various reasons. Some Middle Chalcolithic material has been excavated at Çadır Höyük (Steadman et al. 2008), but this was unfortunately obtained from a small sounding. Unlike Çadır Höyük, at the site of İkiztepe, located on the Black Sea coast in the Bafra delta, considerable Chalcolithic deposits have been excavated (Alkım et al. 1988; 2003). Although the site has been excavated since 1974, considerable confusion exists regarding its stratigraphy and chronology. In the publications, three main periods of occupation are distinguished: period 1 is dated to the ‘Early Hittite Period’ (actually to be dated primarily to the Early Bronze Age 3); period 2 is assigned to the Early Bronze Age; and period 3 is dated to the Chalcolithic.
108
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
The main problem with this stratigraphy is that period 2 includes heterogeneous assemblages, and that much of this material is in fact to be dated to the Chalcolithic rather than the Early Bronze Age (Thissen 1993; Schoop 2005: 308; Welton 2011). Due to these considerable stratigraphic and chronological problems, the İkiztepe sequence is of limited use for dating Prehistoric assemblages in northern Turkey. Likewise, sequences such as those of Büyük Güllücek and Orman Fıdanlığı remain poorly dated and their assemblages derive from less than ideally excavated or preserved contexts (Efe 2001; Schoop 2005; Düring 2011: 229-40). In this situation, dating Prehistoric lithics and artefacts found in surveys remains challenging. Various efforts have been undertaken to try and bring some clarity into the chronological position of poorly dated Prehistoric assemblages. Ulf Schoop has excavated the site of Çamlıbel Tarlası, in order to better understand and date the type of pottery known from the site of Yarıkkaya, and has sampled previously excavated materials from Büyükkaya to obtain radiocarbon dates, and as a result the chronological position of these assemblages is now much better understood (Schoop 2008; 2009; 2010). Alternatively, some scholars have tried to directly date surface assemblages. One possibility that has been explored is that of using thermoluminescence dating on the early ceramic assemblages of north-central Anatolia (Liritzis and Galloway 1982). This approach has been taken up by our colleagues from the Sinop Regional Survey and the İkiztepe excavations (Göksu-Ögelman 1986; Doonan et al. 2007). Although thermoluminescence dating, and related techniques such as OSL (optically stimulated luminescence) and IRSL (infrared stimulated luminescence) have the potential to date artefacts directly, the problem with dating surface assemblages or isolated sherds has been their wide chronological range, which is caused by the fact that we cannot adequately control for background radiation in surface finds (Aitken 1997: 189). Consequently, the luminescence dates obtained for surface assemblages in the Sinop survey and for İkiztepe often have ranges of several millennia and are therefore of limited use for dating Prehistoric assemblages. In recent years, better results have been reported but these have yet to be published (Bauer personal communication 14-12-2012). Yet another technique that might be considered for dating Prehistoric ceramic assemblages is rehydroxylation dating, a technique first put forward in 2009 (Wilson et al. 2009). A pilot was done to date a number of Prehistoric sherds from CAP with this technology (Boeijen 2011). However, due to various technical difficulties in how measurements were obtained, no convincing results have been produced so far. In sum, dating Prehistoric artefacts in northern Anatolia in general, and the Cide-Şenpazar region in particular, remains challenging due to the lack of reliable excavated sequences in the region, and uncertainties as to the degree to which this part of Anatolia was part of broader cultural horizons. To exacerbate matters, at many sites we found only a few undiagnostic sherds or lithic artefacts, and such artefacts can only be dated in very broad terms by looking at fabrics or technologies which
The Palaeolithic (1,000,000-10,000 BC)
109
were in use over millennia in adjacent regions. Thus, in the following discussions, it should be borne in mind that dates are tentative and open to re-evaluation for all assemblages, except for the site of Okçular Kale İni (S24), where we have radiocarbon dates to secure the chronological position of one of our Prehistoric assemblages.
5.2 The Palaeolithic (1,000,000-10,000 BC) No artefacts that could be dated to the Palaeolithic were recognised in the three survey campaigns undertaken in the Cide-Şenpazar region. This lack of data for a period lasting about one million years can be explained either as a real absence or as a result of landscape taphonomies. The idea of a real absence of Palaeolithic occupation is not the most plausible explanation in our minds. Palaeolithic artefacts have been found in many locations across northern Turkey. In the Çankırı region a good sample of Middle as well as some possible Lower Palaeolithic artefacts were found in surface surveys (Matthews 2009: 75-7). In the Türk Arkeoloji Yerleşmeleri database several find spots are mentioned in the Turkish western Black Sea region: including Zonguldak Yöresi, where a handaxe was found; İnceburun near Sinop, with material possibly dating to the Palaeolithic (Işin 1998: 95, plate 1); and various sites in central Kastamonu including Daday Çevresi, Gölköy Enstitüsü, and Akarca Sekileri, of which we saw some materials in the Kastamonu Archaeological Museum and these appeared authentic Palaeolithic artefacts, including various handaxes. One of the specific aims of the 2011 survey campaign was to try and find a Palaeolithic presence in the region. Given the landscape characteristics of the Cide-Şenpazar region, this is of course a great challenge. Up to 2010 we had been visiting caves in the hope that one of these would yield Palaeolithic implements. However, chances are that any layers with Palaeolithic materials in caves are deeply buried under later deposits, and one could only hope to find evidence for the Palaeolithic when cave deposits were disturbed in some major way. It was also clear, given the enormous movements of sediments and rocks in the Cide-Şenpazar region (chapter 2), that the chances of preservation of Palaeolithic implements on the surface would be very slim. Vangelis Tourloukis (2010) has argued that Palaeolithic find spots are determined largely by geological circumstances, such as uplift and erosion, and that locating Palaeolithic sites is most successfully done by investigating the appropriate geological windows. In 2011, Vangelis joined us during our survey season in order to investigate whether such windows - Pleistocene deposits and rock shelters, and in particular old river terraces - existed in the Cide-Şenpazar region. We devoted a number of days to finding such localities, but without success. Therefore, we can state with some confidence that we did not simply miss Palaeolithic sites in our admittedly small sample of the Cide-Şenpazar landscape, but that the appropriate geological windows, where one would expect to find artefacts from this period, are most probably not present in the area.
110
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
5.3 The Early Holocene (10,000-5500 BC) The material that we can date to the period between about 10,000 and 5500 BC in the Cide-Şenpazar region consists exclusively of lithic artefacts. To date, no ceramics can be assigned to this period, although we cannot exclude the possibility that some of the ceramics we found may in fact date back to this period. The second characteristic of the finds dating to the Early Holocene period from CAP is that they were found exclusively on the surface, rather than in caves. In this respect these assemblages differ from those dating to later Prehistoric periods, such as the Chalcolithic and Early Bronze Age, which were found almost exclusively in caves and are dominated by ceramics, although they include a few lithic artefacts as well. These marked differences in find locations and materials found raise various issues. One idea could be that lithic artefacts preserve better on the surface than pottery. Prehistoric pottery is generally baked to relatively low temperatures (e.g. Van As and Wijnen 1995: 87) and, once the burnished surfaces flake off, will quickly disintegrate into small and unrecognisable pieces of pottery. Indeed we did find some tiny pieces of such fabrics in CAP, especially during grid-square collections, but such fragments could not be dated with any certainty (cf. Bintliff et al. 1999). If the paucity of pottery from the Early Holocene period on the surface is related to preservation issues, it would explain why we find no associated ceramics with the Early Holocene chipped stone artefacts. This idea can be supported by the fact that, barring a few small fragments, all Prehistoric ceramics found in CAP derive from caves, and these include Chalcolithic and Early Bronze Age (chapter 6) assemblages, as well as handmade and low fired ceramics from the Early Iron Age (chapter 8). Given that a relatively limited number of caves with good Prehistoric assemblages were found, it is possible that Early Holocene pottery assemblages might be discovered in the future. Alternatively, the Early Holocene in the Cide-Şenpazar region might have been an aceramic period. This would mean that the use of ceramics would start up to 1500 years later than in the surrounding regions of central, western, and northwestern Anatolia (Düring 2011: 122-99). The issue whether or not ceramics were in use in the Cide-Şenpazar region prior to 5500 BC should not be conflated with the question of whether we are dealing with farmers or sedentary populations, given that the lithics could also have been used by aceramic farmers or more or less sedentary Mesolithic groups (Özdoğan and Gatsov 1998; Düring 2011: 31-46). At present, it is impossible to decide which of the two possibilities, an aceramic phase lasting until about 5500 BC, or the loss of Early Holocene surface ceramics due to poor preservation, is the accurate one. We can only hope that the issue will be solved in future investigations. The chipped stone artefacts dating to the Early Holocene have been found mainly in three landscapes investigated during CAP: at Aybasan Mahallesi, in the
The Early Holocene (10,000-5500 BC)
111
Okçular valley, and in the Abdulkadir valley (Fig. 5.1). This is remarkable because extensive unit-walking also took place in other landscapes, such as Çamdibi and the Cide coastal plain, which suggests that some landscapes are more likely to yield Early Holocene artefacts than others (section 2.3). In the case of the Okçular and Abdulkadir valleys we are most likely dealing with the erosion of hill shoulders that exposed Early Holocene materials, which were found near to, or down slope from these shoulders in relatively low densities. By contrast, at Aybasan Mahallesi we came across a dense and variegated in situ cluster of chipped stone artefacts.
0
5
10 Km
Fig. 5.1: Early Holocene finds locations in the CAP survey area. Produced by Toby C. Wilkinson.
5.3.1 Okçular and Abdulkadir The Okçular and Abdulkadir landscapes share many similarities. Both consist of relatively shallow valleys with the following components in cross section. First, there are limestone outcrops at the top, in which large caves are located that have yielded materials dating to the Chalcolithic and later periods, such as Okçular Kale İni (S24) in the Okçular valley and Ballıcı İni (S17) in the Abdulkadir valley. Second, there are sediment shoulders directly beneath these outcrops facing south, which were in use both in the Byzantine period when various farmsteads where located on them, and in the modern period, with clustered villages located in similar localities. Third, there
112
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
are the hill slopes, which are in many places badly affected by gully erosion exposing bedrock, a problem which must have been controlled to some degree in the past, possibly by obstructing the formation of gullies. Fourth, we have the valley floors, which are flat, have deep sediment pockets and can be marshy in places. In both the Okçular and Abdulkadir valleys the surface assemblages were heavily dominated by ceramics and tile fragments dating to the Byzantine period (chapter 11). Another ubiquitous find category consisted of chipped stone artefacts, which upon closer inspection were predominantly threshing sledge elements (chapter 4), which were usually clustered downslope from the ceramic and tile concentrations on hill shoulders, the latter of which most likely indicate the locations of farmsteads. Initially, while unit-walking the Abdulkadir and Okçular landscapes, we recognised only Late Roman and Byzantine materials in the form of tiles, ceramics, and, presumably, threshing sledge elements. The latter, of course are notoriously difficult to date as they have been in use in more or less similar shapes from the Bronze Age up to the sub-recent past (Anderson et al. 2004; Whittaker et al. 2009). However, in the course of our unit-walking we also encountered a few isolated chipped stone artefacts that were clearly Prehistoric in date. These include a number of retouched arrowheads and one backed knife. Throughout the first two seasons we considered these Prehistoric chipped stone artefacts as isolated stray finds. It was only in the final season, when the analysis of the complete chipped stone assemblages took place that we realised that there were numerous Prehistoric artefacts in the Abdulkadir and Okçular landscapes and that their distribution was far from random. The clearest pattern emerged in the Okçular valley (Fig. 5.2). Here, both the greatest numbers of artefacts and the greatest diversity of artefact types, as well as the majority of the cores were all found at the base of the Okçular Kale (S22) rock outcrop. This location has excellent views both to the interior and towards the sea, is sheltered from sea winds, is south facing and, thus, receives a good amount of sunshine, and for all these reasons was in use in many periods (chapter 14). Elsewhere in the Okçular landscape there were fewer artefacts, and most were found on or not far downslope from the hill shoulders from which these artefacts had probably eroded. In the Abdulkadir valley the situation is more or less comparable, although the spatial distribution of artefacts is less clear (Fig. 5.3). To the south of the village of Abdulkadir two cores, a core trimming element and a retouched point were found. To the west of the village a number of flakes and blades, some with retouch were found. Like at Okçular, most of these artefacts were found on or close to the hill shoulder, and there is one locality south of the village that might have been a knapping station, judging by the presence of the cores. While in both the Okçular and Abulkadir valleys the total number of Prehistoric chipped stone artefacts is relatively small, they do seem to represent an ephemeral Prehistoric landscape, hidden beneath the carpet of materials dating to later periods (see also Bintliff et al. 1999). The fact that similar patterns occur in both landscapes reinforces this interpretation.
The Early Holocene (10,000-5500 BC)
113
Fig. 5.2: The distribution of Prehistoric chipped stone artefacts in the Okçular landscape. Produced by Victor Klinkenberg.
Fig. 5.3: The distribution of Prehistoric chipped stone artefacts in the Abdulkadir landscape. Produced by Victor Klinkenberg.
114
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
The lithic assemblages of Okçular and Abulkadir are similar in their technical features and will therefore be presented jointly here. The raw material used for chipped stone artefacts is diverse, including stones of various colours and granularities and fracture properties. In the immediate surroundings of these valleys, no flint outcrops appear to exist, and it seems that stones, often of less than optimal quality for chipping, were brought to this area and chipped here. Alternatively, some artefacts found in these valleys might have been produced elsewhere. Remarkably, some of the most elaborate tools, such as large retouched points, were produced on flint with poor fracture properties. The ratio of tools versus non-tools in the lithic assemblage is almost 1:2. No chips were found, all blades and flakes are larger than 15mm. Unretouched flakes and blades in the assemblage are crude and usually fragmented, they vary in maximum dimension from 15 to 50mm. The distal ends are generally either broken off or stepped or hinged. It therefore can be concluded that the discarded flakes and blades are of low quality. Four core trimming elements were found:1 these were made of the same material as the blades and flakes but have a larger average size. This can be related to their function in the reduction process: they were struck with the intention to remove imperfections and therefore cover a large area of the nodule. Cores from the Okçular and Abdulkadir valleys on the contrary are very small (Fig. 5.4). Equally, the negatives on these cores have an average width of 7mm – far smaller than the blades and flakes from the same valleys. The cores were reduced even beyond the point of exhaustion: all cores are fragmented and were broken in an attempt to strike off a flake when the core was already too small. The tools found at Okçular and Abdulkadir (Fig. 5.5) were made from the same raw materials as the non-tools but are generally larger and were produced using more sophisticated flaking and reduction methods. Half of the tools are informal tools: simple notched or irregularly retouched flakes and blades. These informal tools are predominantly present on rather crude blanks, whereas the formal tools are produced on better quality flakes and blades. These tools were thus produced in a highly pragmatic manner. A crudely struck flake is easily altered into a tool by means of retouching or notching. The production of the formal tools, by contrast, was based on proper blanks. Amongst the two assemblages are nine formal tools: four scrapers,2 three projectile points,3 one backed knife4 and a burin.5
1 L58; L130; L172; L189. 2 L7; L45; L104; L249. 3 L127; L144; L288. 4 L3. 5 L175.
The Early Holocene (10,000-5500 BC)
115
Fig. 5.4: Chipped stone cores found in the Okçular and Abudulkadir valleys. Produced by Lorraine McEwan.
Three of the four scrapers were produced on small flakes with steep retouch on a lateral side.6 The fourth was produced on the distal end of a blade.7 The burin was made on thick flake.8 Interestingly, the flake is only partly patinated; the retouch which is applied to produce the burin was struck on a later moment in time than when the flake was struck from its core, again suggesting a rather pragmatic approach to chipping using any available materials. The backed knife9 is a small implement, heavily but finely retouched on one lateral side, creating a steep back. The sharp and
6 L7; L104; L249. 7 L45. 8 L175. 9 L3.
116
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
Fig. 5.5: Chipped stone tools found in the Okçular and Abdulkadir valleys. Produced by Lorraine McEwan.
straight edge on the other side shows some signs of wear. The projectile points were produced with great skill. The projectile point from Abdulkadir10 could be considered to differ slightly from the Okçular specimens11 in that it is thicker and rounder in section. All three points are heavily retouched over their whole surface and represent sturdy projectile points. It is not entirely clear how they were hafted and used.
10 L127. 11 L144; L288.
The Early Holocene (10,000-5500 BC)
117
The Okçular and Abdulkadir lithic assemblages comprise a relatively small collection in terms of formal tools which can be used for dating through comparison with assemblages from surrounding regions. The problem that the study of lithic assemblages of northern Anatolia is hampered by the lack of synthetic studies encompassing the very diverse chipped stone industries of various regions of Anatolia has already been mentioned. Nonetheless, typological comparison with often far removed excavated assemblages is the only means of dating our chipped stone materials. The three retouched points found in the Okçular and Abdulkadir valleys have good parallels in Ceramic Neolithic and Early Chalcolithic sites in central Anatolia dating to the seventh and sixth millennia BC, such as Çatalhöyük, Tepecik-Çiftlik, and Köşk Höyük (Balkan-Atlı 1994; Conolly 1999; Bıçakçı et al. 2007), although these are produced on better quality lithic materials than those of the Cide region. Another parallel for these points, and closer to our region, can be found at the site of İkiztepe, where similar fully retouched points have been found (Alkım et al. 1988, plate 63). This assemblage has been labelled ‘assemblage BB’ by Ulf Schoop and was dated by him to ca. 4500 BC (Schoop 2005: 307-322; Welton 2011: 84). The İkiztepe points, however, are more geometric in shape, have a much thinner profile, and were thus produced and used in ways that differ from the artefacts found at Okçular and Abdulkadir. The İkiztepe points are also found in association with a ceramic assemblage that is completely absent in the Cide-Şenpazar region. Finally, the type of retouched points found in the Okçular and Abdulkadir valleys are completely absent in the lithic assemblages of the Marmara region (Özdoğan 2002; Gatsov 2009) and Orman Fıdanlığı (Baykal-Seeher 2001). Of the remaining tools only the backed knife can be considered as a diagnostic type which we can compare with artefacts from other regions. Interestingly, this artefact has very good parallels in the Mesolithic ‘Ağaçlı’ sites from the Marmara region (Özdoğan 2007: fig. 18.6), which can tentatively be dated to the eighth and seventh millennia BC (Düring 2011: 40-42). Thus, the few parallels we can draw for the lithic assemblages found at Okçular and Abdulkadir suggest a tentative date for these assemblages in the seventh millennium BC. Given the circumstance that this date rests on three retouched points and one backed knife, and the questionable assumption that they can be compared with types found in far removed central Anatolia and the Marmara region, the hypothetical nature of this dating should be clear.
5.3.2 Aybasan Mahallesi The landscape at Aybasan Mahallesi is among the more spectacular investigated by CAP. Situated in the central part of the Şenpazar district it consists of a steep mountain, which is covered from top to bottom by agricultural fields, with the
118
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
village perched halfway along the slope, a situation probably related to the presence of a spring in this position. Given that most of the mountains surrounding that of Aybasan Mahallesi were covered by forests rather than by fields, we hypothesised that the mountain is probably geologically more suitable for cultivation, with a geological make up that makes it less susceptible to erosion. Indeed, when we unitwalked the area we found hardly any evidence for erosion, whether in the form of sheet or gully erosion. In the course of unit-walking the Aybasan Mahallesi landscape we picked up sherds of Byzantine date in the vicinity of the village, but generally found little further afield. Amidst a large number of more or less empty units, we unexpectedly found clusters of chipped stone artefacts (S82 and S98), dating to the Early Holocene. These clusters are located halfway on the slope between the village and the river below at the spot where meadows give way to denuded bedrock. Why people would have chosen this particular spot on the slope of the mountain at Aybasan Mahallesi to chip their artefacts is not clear. At present there are no particular elements that make this spot more favourable than others, like for example a water source. It is possible, of course, that there were large trees providing shade, and that that is all there is to it. The chipped stone artefacts found at Aybasan are diverse in nature. A total of 73 pieces of chipped stone were found, and these consist mostly of debitage, such as large flakes, and chips. There are a few blades and bladelets, but in general there is little that can be described as a tool. There are (exhausted) unipolar cores, as well as irregular cores that were worked from various angles (Fig. 5.6). Further, various core trimming elements and a crested blade – the first blade struck after core preparation - suggests that we might be dealing with in situ knapping at this location. The interpretation of an in situ knapping assemblage is supported by two properties of the assemblage. First, the chipped stone artefacts were distributed in clear clusters, rather than randomly spread over an erosion interface. Second, the chipped stone artefacts from S82 and S98 vary considerably in size and weight. If they would have eroded downslope one could expect a sorting by size/weight, with the smaller objects moving further downslope, but this is clearly not the case (Fig. 5.7). Like at Okçular and Abdulkadir, the chipped stone artefacts of Aybasan were produced from various types of flint – macroscopically eleven different types of flint were distinguished, as well as from obsidian, of which four pieces were found. This variability suggests that we are dealing with mobile groups that obtained some of their materials through exchange. The obsidian derives from the Galatian sources (Düring and Gratuze 2013 and section 4.4.2.2) about 150km from the Cide-Şenpazar region, whereas the flint presumably derived from more local sources: one such source seems to have been at the Fakas Çay near the villages of Kışla and Ziyaret, in the east of the Cide district, where we encountered lithic materials similar to some of the material
The Early Holocene (10,000-5500 BC)
119
Fig. 5.6: Chipped stone cores found at Aybasan Mahallesi. Produced by Lorraine McEwan.
from Aybasan. Interestingly, the obsidian was chipped more carefully and for smaller artefacts than the flint.12
12 A very similar pattern occurs at the site of Kayabaşı in central Kastamonu (Kuzucuoğlu et al. 1997: 287, plate 9).
120
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
Fig. 5.7: Size distribution of the chipped stone artefacts found at Aybasan Mahallesi. Produced by Michele Massa.
The debitage from the Aybasan site ranges in size from seven to 44mm in maximal dimension. Two obsidian bladelets with a maximal dimension of seven and eight millimetres represent the smallest faction of debitage in the assemblage. Small flint objects have also been found, however. Perhaps this smaller size category is underrepresented for flint objects because they are harder to spot in the field. Obsidian chips are relatively easy to spot in the field because their glossy surface stands out amongst the gravel soil. The large unretouched flakes from Aybasan indicate a relative abundance of lithic material at the site; apparently the material was plentiful enough to discard rather than having to be processed into tools. it should, however, be noted that also unretouched flakes may have been used as tools. In terms of tools both informal tools as well as formal tools have been found (Fig. 5.8). The ratio between formal and informal tools at Aybasan is 1:2.7. This high proportion of informal tools in the Aybasan assemblage contrasts with the 1:1 ratio from the Okçular and Abdulkadir valleys. Informal tools include notched blades and
The Early Holocene (10,000-5500 BC)
121
Fig. 5.8: Chipped stone tools from Aybasan Mahallesi. Produced by Lorraine McEwan.
simply retouched blades and flakes. Characteristically, the many informal tools from Aybasan were produced on high quality flint, indicating once more that flint quality does not seem to relate to tool type in our archaeological assemblage. Formal tools in the Aybasan assemblage are restricted to scrapers13 and (fragments of) knives.14 The scrapers are made on thick and large flakes by steep retouching one side. Knife L131 is a very large curved flake with retouch mainly
13 L183; L190; L201; L248; L252; L263. 14 L131; L221.
122
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
on the inside bend. It is made from high quality, fine grained material and has striking green and purple colours. Fragment L221 is a small blade segment with flat retouch on one lateral side. This artefact may represent a part of a composed knife. A small notch on the non-cutting side seems to indicate that the object was deliberately broken, a practice which is common in lithic technology when several fragments are hafted together. Four core fragments have been found at Aybasan, one of these is the unipolar obsidian core.15 The other cores from the assemblage differ greatly from this obsidian example. Cores L133 and L227 have a characteristic cube-like shape due to multidirectional reduction. L162 is a heavily burned core fragment and it is unclear what mode of reduction it represents. Beside the core fragments four core trimming elements have been found. Fragments L207 and L254 have been struck off a core in order to rejuvenate the striking surface. L165 is a large flat fragment of flint which has been interpreted as a striking platform renewal flake. The large size of the artefact seems conspicuous within a assemblage of on average smaller lithic flakes but the edges of this core trimming element have been battered in such a way that further reduction or retouching is nearly impossible. L225 is a heavily burned core fragment representing the distal end of a unipolar core. All surfaces of the fragment have been heavily burned; therefore it might have been split from the core through heating rather than striking. The particular shape of the object, with an extruding point in one corner, may indicate that the base of the core was faulted and needed repair. If the latter is the case, L225 might indicate that cores were unipolarly reduced by preference until the core was exhausted. At this point the remaining core could be multidirectionally reduced as is indicated by the fully exhausted cores L133 and L227. The fact that we probably found a knapping station at Aybasan Mahallesi explains why relatively few formal tools were found, such as points or burins. Unfortunately, this lack of formal tools makes comparative dating a challenge for the Aybasan assemblage. The presence of unipolar bullet cores for producing narrow bladelets, suggest a very general date in the Early Holocene, given that such bladelets and cores do not occur on Middle Chalcolithic sites such as Orman Fıdanlığı (Baykal-Seeher 2001) and elsewhere (Shimelmitz 2011: 516-9).16 This very coarse date is further supported by the absence of associated ceramics, although this absence could also be explained as resulting from poor preservation or the nature of the site: a knapping station rather than a settlement.
15 L193. 16 However, similar cores do occur in EBA levels at Demircihüyük (Baykal-Seeher 1996).
The Early Holocene (10,000-5500 BC)
123
5.3.3 Çamdibi The Çamdibi valley, locally known also as ‘Loç’, is located in the southeastern corner of Cide district on the Koca Dere/Devrekani Çay river (both names are used interchangeably in the area), just after it emerges from the large Valla canyon to the south. At one edge of the valley is Kılıçlı Mağarası (S93/S88), which has yielded good Prehistoric ceramic assemblages (chapters 6 and 7, and below). One of the objectives for the Loç valley intensive survey was, therefore, to contextualise the cave in its surrounding landscape (chapter 14). In our unit-walking, we found predominantly Late Roman and Byzantine period ceramics (chapter 11) and tile material, as well as spolia from the same periods. Prehistoric finds in the Çamdibi valley were few, apart from those found in a single location. A most remarkable find comes from a flat area adjacent to the village mosque. Here, a polished chisel and an accompanying flint knife (a large blade retouched laterally and on the proximal end) were found next to a large electricity pole (S132). These artefacts (Fig. 5.9) were probably dug up from a subsurface level in the process of creating a pit for positioning the pole, after which the soil from the pit was used for refilling it. The spot also contained many tile fragments, one Byzantine sherd, and a pivot stone. The flint knife17 consists of a large blade retouched laterally and on the proximal end from the dorsal side and at the distal end from the ventral side. It was made of a brown fine grained flint. The accompanying chisel,18 which was just over 5cm long, is a polished tool, and made from a so far undetermined green coloured stone with a symmetrical point which is rounded in the top view. Both the chisel and flint knife have good parallels elsewhere in Anatolia. The flint knife can be compared with similar examples from Early Neolithic Thrace dating to the sixth millennium BC, for example at Čavdar and Hoca Çeşme (Gatsov 2009), and Orman Fıdanlığı (Baykal-Seeher 2001: plate 12), which would also suggest a late sixth or early fifth millennium BC date (Düring 2011: 212-3).19 Polished axes and chisels have a much broader chronological distribution, which starts in the Neolithic: they are for example common in Ilıpınar and Ulucak (Roodenberg 1999: fig. 19; Çilingiroğlu and Çilingiroğlu 2007: fig. 14), but continue to occur up to the Early Bronze Age (Korfmann et al. 1996: Tafel 81). Very tentatively these two artefacts could be dated to the sixth to early fifth millennium BC, but later dates are possible.
17 L265. 18 F56. 19 Parallels at Neolithic Çatalhöyük (Carter et al. 2005), and Early Bronze Age Demircihüyük (BaykalSeeher 1996) are much less convincing.
124
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
Fig. 5.9: Flint knife and chisel found at site S132 in the Çamdibi valley. Produced by Lorraine McEwan.
5.3.4 Obsidian Exchange Networks A total of six pieces of chipped obsidian were found in CAP: of which four derive from Aybasan Mahallesi,20 and two from the cave of Okçular Kale İni (S24).21 The discovery of obsidian chipped pieces, albeit in small numbers, is of considerable interest as it provides us with the opportunity to investigate from which source these pieces originated. Of the six pieces we were allowed to export three obsidian samples:
20 L136; L137; L193; and L224. 21 L186 and L197.
The Early Holocene (10,000-5500 BC)
125
two from Aybasan Mahallesi,22 and one piece of obsidian which derives from the Chalcolithic site of Okçular Kale İni (S24)23 (see section 5.4.1, for a discussion of the site). These pieces were analysed by Bernard Gratuze from the CNRS at Orléans, France (see section 4.4.2.2 and Düring and Gratuze 2013 for more details). The comparison of the artefacts’ composition with the source reference dataset shows that two artefacts originating from the Aybasan area match the composition of the Sakaeli source,24 near the town of Orta in Çankırı, whereas the third one, recovered from Okçular Kale İni25 probably derives from the Yağlar source, near the modern town of Kızılcahamam. Obsidian from the Sakaeli and Yağlar sources has also been identified in Neolithic sites of the Marmara region, such as Ilıpınar, Pendik and Fikirtepe, where they co-occur with obsidian from the Cappadocian sources (Bigazzi et al. 1995, Bigazzi et al. 1998: 80-86). So far, obsidian from the Galatian sources has not been identified at sites in south central Anatolia, the Aegean, or Thrace, and as such this material seems restricted in distribution to northwestern Anatolia (Chataigner 1998: 275-7). The artefacts from Ilıpınar come from levels X and IX, and can be dated to approximately 6000 BC. Those from Pendik in all likelihood predate Ilıpınar X, and can be tentatively assigned to the second half of the seventh millennium BC (Özdoğan 1999: 213). Thus, the distribution of Galatian obsidian is restricted at present to northwestern Anatolia, and to a short time period in the early Ceramic Neolithic, ca. 6500-5800 BC. Remarkably, recent excavations at Late Chalcolithic Çamlıbel Tarlası, located not far from the Galatian sources, have produced obsidian deriving from four distinct Cappadocian sources, but not from the nearby Galatian sources (Schoop 2011: 64). This suggests that by the Late Chalcolithic, the Galatian sources were no longer at the centre of an obsidian exchange network. On the other hand, the sample from Okçular Kale İni, which we can date with some confidence to the early fifth millennium BC, indicates that this obsidian exchange network might have continued to exist at least in part during the Middle Chalcolithic. We are far removed from understanding the Galatian massif obsidian exchange networks in anything but its barest outlines. The data presented here from the CideŞenpazar region, show that obsidian was exchanged northwards as well as to the Marmara region. This alone is significant. Transport connections with the Marmara region are much easier than with the Black Sea littoral to the north. Although the distance between the Galatian obsidian sources and the Cide-Şenpazar region is only about 150 kilometres as the crow flies, there are no natural routes that connect the coast with the interior and the mountains form a formidable obstacle for travelling
22 L136 and L224. 23 L197. 24 L136 and L224. 25 L197.
126
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
in this terrain. Thus the presence of obsidian from the Galatian sources suggests that the communities in northern Anatolia were in contact with each other despite the difficult nature of the intervening terrain. The alternative suggestion would be that the Galatian obsidian was brought to the Marmara region first, and was then brought to the Cide region by boat along the Black Sea coast, but this seems less plausible at the moment, given that there is no evidence at present for long distance maritime activities on the Black Sea in the Early Holocene (section 2.2.1).
5.3.5 The Early Holocene in the Cide-Şenpazar Region: By Way of Summary The Prehistoric chipped stone industries found in the landscapes of Okçular, Abdulkadir, Aybasan, and Çamdibi all have parallels suggesting a date in the seventh and sixth millennia BC in central Anatolia and the Marmara region. One could argue that this could simply reflect the fact that the lithic assemblages of the ensuing Chalcolithic period have been much less well investigated and published. However, some scholars have argued that lithic technologies change considerably in the Chalcolithic, in which unsystematic flake production becomes the norm (BaykalSeeher 2001: 159-167; Shimelmitz 2011: 516), and it is only in the Late Chalcolithic and Early Bronze Age that blade technologies are documented again at sites like Orman Fıdanlığı and Demircihüyük (Baykal-Seeher 1996; 2001: 167). The mere fact that we clearly have blade technologies could therefore be used to support an Early Holocene date for our surface assemblages in Cide-Şenpazar. Further, the complete absence of associated Prehistoric ceramics in Okçular, Abdulkadir, Aybasan, and Çamdibi also provides tentative evidence for an Early Holocene date, assuming that people in the area had not yet adopted ceramics in this period. As always, absence of evidence is not evidence of absence, but the complete lack of ceramics in all these landscapes does suggests that we are dealing with an aceramic horizon. Finally, the fact that obsidian has been found among these assemblages lends further support to an Early Holocene date, as the obsidian exchange networks originating from the Galatian sources seem to have dropped out of use in the course of the Chalcolithic. Thus, while our dating of the lithic assemblages of Okçular, Abdulkadir, Aybasan, and Çamdibi to the Early Holocene must remain tentative due to the small number of artefacts and the difficulties of comparative dating with sites that are far removed, the comprehensive study of these artefacts does suggest that an Early Holocene dating for these assemblages is highly plausible. Only future research can determine whether in this period we are dealing with communities relying on hunting, gathering, and fishing, or whether they practiced agriculture, and how sedentary they were.
The Chalcolithic (5500-3000 BC)
127
5.4 The Chalcolithic (5500-3000 BC) The period between 5500 and 3000 BC is exclusively known in the Cide-Şenpazar district from cave sites, and thus the dataset differs completely from that of the preceding Early Holocene period, for which data derived only from exterior surfaces. By far the best dataset derives from the site of Okçular Kale İni (S24), and smaller assemblages have been found at Kılıçlı Mağarası (S88/93), Çadır İni (S20), Gideros Mağara I (S59), Ballıcı İni (S17), and Derebağ Köy Mağarası (S188). To this general find location there is but one exception, which consists of Unit 3: a field located due south of Okçular Kale (S22), where we later carried out a gridsquare collection (G1). Unit 3 yielded four small undiagnostic sherds,26 and two base fragments,27 for which the mineral tempered fabric and firing suggested a (very broad) Prehistoric date. Three similar examples were subsequently found in grid G1,28 at the same location. the only diagnostic shape is a large vessel with a flat base with a diameter of about 20cm (Fig. 5.10).29 The shape is too generic for drawing specific parallels with assemblages known from other sites.
Fig. 5.10: C293 found in Unit 3. Handmade, mineral tempered. Produced by Lorraine McEwan.
In each campaign, we visited many caves that were difficult to locate and access. Most of these caves did not yield any archaeology, but in each campaign we were lucky enough to discover one cave with assemblages dating to the Chalcolithic, Bronze Age and Iron Ages. Chalcolithic deposits had in most cases been disturbed by looters looking for materials they could sell on the antiquities market. In many of the caves, which we investigated only a few Prehistoric sherds were found in spoil heaps of looters. Examples are Ballıcı İni (S17), where we found a total
26 C242; C244-5; C296. 27 C293; C295. 28 C1082; C2467; C2468. 29 C293.
128
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
of six Prehistoric sherds;30 Çadır İni (S20), which yielded one sherd;31 and Gideros Mağara I (S59), which also yielded one sherd.32 These were all buff to brown unevenly fired, mineral tempered and handmade body sherds. Some of them had smoothed or coarsely burnished surfaces. Material of this sort is of course notoriously difficult to date, given that handmade mineral tempered unevenly fired fabrics could date to many millennia of Prehistory. However, given that these fabrics are not dissimilar to Middle Chalcolithic pottery from the Devrek region (Karauğuz and Düring 2009) and the Late Chalcolithic central Kastamonu survey ‘Koçaç group’ (Kuzucuoğlu et al. 1997: 287-92),33 a broad date to the fifth and fourth millennium BC seems plausible for these sherds. More diagnostic ceramic assemblages have been found at the three sites of Derebağ Köy Mağaraşı (S188), Kılıçlı Mağarası (S88/S93) and Okçular Kale İni (S24). Each of these caves is different in its archaeological signature and the periods that are represented in it, and each will therefore be discussed separately.
5.4.1 Okçular Kale İni The most spectacular Chalcolithic assemblage found by CAP is that of the cave of Okçular Kale İni (S24). The landmark rock outcrop of Okçular, on top of which are located the remains of a Byzantine castle (S22), was shown to us by local residents, alongside with the entrance to the cave. After crawling through a narrow passage one needs to descend a steep slope of some five meters. Below is a large cave, which contains stalactites and stalagmites, and whose floor is covered with rocks from episodes of roof collapse. At the back of the cave we found an area with brown crumbly moist loamy clay with large numbers of Prehistoric burnished ceramics and lithics: the spoil heaps of an illegal excavation. Beyond these spoil heaps we encountered a profile with stratified horizontally laminated midden deposits containing ash, charcoal, animal bone fragments, flints and substantial amounts of pottery which were horizontally deposited (Fig. 5.12). These deposits had been exposed by looters and were situated at the bottom of shafts. It seems that we are dealing with a former cave entrance that had been used as a garbage dump in Prehistory. This would imply that in the Chalcolithic people were settled on the rock outcrop of Okçular, which in Byzantine times was fortified as a
30 C135; C170-2; C174-5. 31 C279. 32 C783. 33 Bleda Düring was graciously allowed access to the central Kastamonu survey collections by Aslı Özdoğan, Aksel Tibet, and Catherine Marro. His thanks also go out to Mehmet Özdoğan for facilitating access to the Istanbul Prehistory laboratories, where these materials are stored.
The Chalcolithic (5500-3000 BC)
129
Fig. 5.11: Plan of Okçular Kale İni (S24). Produced by Michele Massa and Victor Klinkenberg.
stronghold, but which would have been a natural stronghold with a good view of the surrounding landscape in earlier periods. On this rock outcrop, which is mostly
130
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
covered with dense shrub vegetation, no Prehistoric artefacts or features are visible today. However, the density of midden deposits and the extraordinary amount of stratified Prehistoric pottery in the cave shaft below suggests extended occupation at this site. Interestingly, similar settlement locations are documented elsewhere in Anatolia during the Middle Chalcolithic (the date of Okçular Kale İni will be discussed below): with good examples at Güvercinkayası (Gülçür and Firat 2005) and Kanlıtaş (Türkcan 2010).
Fig. 5.12: View of section through midden deposits at the back end of Okçular Kale İni (S24) containing charcoal, bone, ash, lithics, and pottery in horizontal orientation. In the foreground Imke van Hagen, project member of CAP. Photo by Victor Klinkenberg.
The Prehistoric pottery from Okçular discussed here predominantly derives from three spoil heaps of disturbed midden deposits at the back end of the cave. From the entrance of the cave two woolen threads led directly to these excavation areas, where digging seems to have taken place in recent years. Elsewhere in the cave (Fig. 5.11), pottery dating to Byzantine period and the Early Bronze Age was found, as well as some Chalcolithic sherds in smaller quantities. It is not clear, how these sherds ended up in this part of the cave: in part they might have been left by looters digging in the Chalcolithic deposits at the back end of the cave, but other materials, such as larger fragments of Byzantine pots, testify to later uses of this part of the cave, the
The Chalcolithic (5500-3000 BC)
131
physical shape of which might have changed considerably over the millennia – former entrances might for example have collapsed. In total 97 sherds were collected from Okçular Kale İni (S24) that can be dated with some confidence to the Chalcolithic period.34 The Prehistoric pottery found in Okçular Kale İni is handmade and contains small mineral inclusions such as quartz and calcareous grits. The pottery is fired to low temperatures, either in reduced atmospheres, or with differential degrees of oxidation in the vessel wall. A large component of the pottery has wall thicknesses below five millimetres and surface colours range from brownish red to jet black. The surfaces of these Okçular Kale İni ceramics were finished in various techniques but with a great deal of effort aimed at producing conspicuous vessels. Many vessels were burnished to a jet black or brownish red high lustre. Five ‘wares’ were distinguished among the Prehistoric sherds from Okçular Kale İni, on the basis of their surface treatment and colours (table 5.2). The latter are probably a function of firing conditions, for example an open vessel placed upside down in reducing conditions would result in a black outer surface and a brown interior surface. Thus, the ‘wares’ in table 5.2 largely reflect small differences in production and/or firing conditions. Tab. 5.2: Five Prehistoric ‘wares’ found at Okçular Kale İni (S24).3536373839 ware
description
A
exterior: highly burnished jet black, burnishing marks horizontal interior: vertical scraping marks – suggesting thinning of vessel in leather hard condition, possible with flint or bone implement.35 exterior: highly burnished and jet black interior: smoothed and orange brown.36
B C
exterior: burnished and reddish brown interior: smoothed and black.37
D
exterior: high burnish and jet black interior: smoothed (sometimes lightly burnished), colours ranging from dark grey to brown grey.38 both surfaces reddish brown and smoothed, exterior lightly burnished.39
E
34 C206;C747-9; C752-3; C756-9; C761-7; C786; C788-93; C795-842; C2886-91; C2899; C2904; C2909; C2911; C2922; C2925; C2927; C2961; C2994-5; C2997; C2999; C3000; C3002; C3004-6; C3010; C3078. 35 Examples include: C790; C804; C824; C828; and C830. 36 Examples include: C797; C821; C823; C825; C829; C831-2; and C834. 37 Examples include: C788; C805; C810; and C827. 38 Examples include: C786; C789; C792; C795; C803; C806; C808-9; C815; C822; C826; C837; C842; C2994-5; C3000; C3002; and C3006. 39 Examples include: C206; C796; C798; C802; C807; C812-4; C819; C833; C836; and C840-1.
132
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
These differences are significant however in the consideration of possible comparanda for the Okçular Kale İni assemblages. One initial idea was that this pottery could be related to ‘Red-Black Burnished Ware’ a ceramic tradition wellknown from eastern Anatolia and the Caucasus which appears around the midfourth millennium BC (Palumbi 2011: 216), and for which parallels are sometimes postulated in northern and central Anatolia, for example at Dündartepe and Çadır Höyük (Thissen 1993: 213-5; Steadman et al. 2007: 391; Steadman et al. 2008: 64). This idea can be discounted, however, as this type of pottery is consistently black on the exterior and red on the interior. At Okçular, our assemblage does not match this description. Initially, the high quality black and red burnished ceramics found in Okçular were interpreted as belonging to the Late Chalcolithic and possibly Early Bronze Age I (Düring and Glatz 2010: 208), on account of the fact that similar highly burnished handmade fabrics occur at sites such as Beycesultan (Lloyd and Mellaart 1962) Demircihüyük (Efe 1987), and the Early Bronze Age sites of the Ankara region (Bertram 2008); the same is true for a twisted handle (Fig. 5.13). There are however, no typical Early Bronze Age shapes in the Okçular Kale İni assemblage – which could be explained in part by the fact that the Early Bronze Age I period is very poorly known in north-central Anatolia (Düring 2011). The one sequence in northern Anatolia that covers much of the Chalcolithic, that of İkiztepe, is poorly published (Thissen 1993; Schoop 2005) and provides few good parallels. Thus we were left with a possible date somewhere in the (Late) Chalcolithic. In 2011 Ulf-Dietrich Schoop, one of the prime specialists on the Anatolian Chalcolithic, visited us and investigated our materials. He was particularly interested in the ceramic assemblage of Okçular Kale İni, arguing that some of this material, in particular that classed as ‘ware E’ during the analysis (table 5.2), is comparable to that of Büyükkaya Upper Plateau (Schoop 2005), which has recently been radiocarbon dated to 5600-5500 Cal. BC (Schachner 2012: 118). All these divergent dates simply demonstrate how difficult it is to date Chalcolithic ceramic assemblages in northern Anatolia through comparative dating (also Schoop 2005; Düring 2008). For this reason we returned to Okcular Kale İni in 2011, to obtain radiocarbon samples from the exposed section for absolute dating. Two of the eleven radiocarbon samples collected, both consisting of bone samples, were submitted for dating to the Centre for Isotope Studies at Groningen University. They yielded two consistent dates between about 4900-4700 Cal. BC (table 5.3). These dates thus place the Okçular Kale İni ceramic assemblages in the early fifth millennium BC, rather than the mid-sixth millennium as suggested by Schoop, or the fourth millennium as hypothesised initially by Düring!
The Chalcolithic (5500-3000 BC)
133
Fig. 5.13: Chalcolithic ceramics found at Okçular Kale İni (S24). Produced by Lorraine McEwan. Tab. 5.3: Radiocarbon dates from Okçular Kale İni (S24). sample number
uncalibrated date
calibrated date (Oxcal 4.1)
GrA-51787
5890 +/- 35 BP
4840-4692 BC (95%)
GrA-51788
6005 +/- 35 BP
4991-4799 BC (95%)
The pottery shapes include: first, flat bases of large vessels with vertical burnishing on the exterior and wiping traces on the interior and relatively large mineral inclusions;40
40 C805; C818; C2911.
134
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
second, open bowls that were smoothed or burnished,41 in one case with impressed decoration on the exterior,42 and in another with the same on the inside;43 third, large tapering necks, presumably of jars with a small amount of temper and signs of scraping and smoothing;44 and finally, there are two pots with particular rim shapes that are so far unique,45 both of which were burnished on the exterior. Handles took various shapes, including one large burnished handle with a false twist,46 and two vertical flat strap handles.47 Finally, there is a large flat burnished piece that might have been a handle or the corner of a plate.48 The parallels that can be found for the Okçular Kale İni ceramic shapes are of a rather generic nature: some shapes can be compared with elements of assemblages elsewhere, but it is not possible to link sets of features. Thus, for example the tapering necks found at Okçular Kale İni are reminiscent of those found in Orman Fidanlığı levels V and VI (Efe 2001: figs. 15 and 16) but other features at Orman Fidanlığı, such as painted and incised decoration are absent, as are the majority of the shapes. Resemblances are also evident for Büyükkaya, in particular the groups published as ‘Mittlere und Untere Plateau’, were somewhat similar large flat bases and necks occur. However, at Büyükkaya handles are more or less absent, and the knobs on rims and carinations common at Büyükkaya do not occur at Okçular Kale İni. Perhaps the most convincing parallels published are those of the central Kastamonu survey, at sites like Çatalkaya, Kayabaşı, and Sokukayası (Marro et al. 1996: plate 1: 3-5), where similar flat bases, tapering necks, and vertical strap handles were found, although often in association with (Late Chalcolithic) fruit stands and knobs, which are both lacking at Okçular Kale İni. Finally, an isolated parallel can be drawn between the handle with false twist and a similar one found at Alacalıgöl, a Middle Chalcolithic site in the Troy region (Gabriel et al. 2004: fig. 16). It is also useful to consider features that do not occur in the Okçular Kale İni assemblage. Remarkable is the complete absence of any parallels in the Middle Chalcolithic AA and BB complexes at İkiztepe as defined by Schoop (2005). Likewise, the complete absence of horned handles, which occur from the Aegean up to İkiztepe (Schoop 2005; Düring 2011) and are well attested in Zonguldak province (Karauğuz and Düring 2009) in the mid-fifth millennium BC. This probably implies that the Okçular Kale İni assemblage predates the ‘horned handle horizon’ starting at ca. 4500 BC.
41 C206; C799; C817; C836; C2904; C2925; C3078. 42 C3078. 43 C2925. 44 C802; C2888; C2922; C2999; C3010. 45 C756 and C813. 46 C816. 47 C2890; C3005. 48 C2887.
The Chalcolithic (5500-3000 BC)
135
Twelve lithic artefacts were also retrieved from the Chalcolithic deposits at Okçular Kale İni.49 They comprise two obsidian pieces50 and ten flint artefacts, and include mostly blades and a few flakes. Some of the blades and flakes show use retouch, others were intentionally retouched to serve as tools, for cutting or hafting (Fig. 5.14). None of the chipped stone artefacts are of a formal type that can be used for comparative dating.
Fig. 5.14: Selected chipped stone artefacts from Okçular Kale İni (S24). Produced by Lorraine McEwan.
49 L108-11; L173; L179; L184-6; L196-8. 50 Both L186 and L197 were tested in Lyon and derive from the Galatian sources.
136
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
Of interest in this lithic assemblage, however, are the following issues. First, in the early fifth millennium BC the obsidian exchange networks from the Galatian sources to northern Anatolia that were in place already in the Early Holocene (section 5.3.4) were still in operation, testifying to the fact that the region was connected to other parts of Asia Minor. Second, small bladelets and regular blades are largely absent from Okçular Kale İni, and the chipped stone consisted mainly of flakes, both large and small and large irregular blades, which were retouched in a pragmatic rather than systematic manner. In this respect, the chipped stone industry resembles that of contemporary Orman Fidanlığı (Baykal-Seeher 2001). Summarising the data from Okçular Kale İni: at this site we have a sizeable collection of well dated early fifth millennium BC ceramics and associated chipped stone artefacts that can be linked in a very general manner with assemblages in other parts of Turkey, such as those of Büyükkaya and Orman Fidanlığı, but which as a complete assemblage does not have clear comparanda and must therefore for the time being be understood primarily as a local phenomenon. It is precisely because of the local repertoire of the assemblage, that is that the ceramics differ from those known from other regions, that comparative dating is tenuous, and that absolute dating is key for anchoring Chalcolithic assemblages in northern Anatolia.
5.4.2 Derebağ Köy Mağarası Derebağ Köy Mağarası (S188) is a large cave (ca. 220m deep) whose entrance is located on the slope of a steep mountain above the village of Derebağ Köy in the easternmost part of the Cide district. The cave is more or less linear and generally has a regular width and height, although it does have some narrower passages, and descends and ascends considerably. Throughout the cave there are collapsed stones, which have fallen from the roof. Below these collapsed rocks is a sediment from which we obtained sherds dating to numerous periods. Thus, Derebağ Köy Mağarası is probably our only cave site that does not show clear signs of looting. This makes the find of numerous sherds dating to various periods including the Chalcolithic, Early Bronze Age, the early second millennium BC, the Iron Age, and Byzantine periods all the more remarkable. We do not understand how the ceramics were deposited in the cave throughout the ages and how they came to be present on the surface. At present the cave floor is covered with stone collapsed from the roof, but it is not clear whether the cave was equally instable in the past. Derebağ Köy Mağarası contained a significant amount of sherds of handmade, poorly fired dark grey pottery. At first these sherds were all dated to the Chalcolithic, but some characteristics of these sherds did not fit easily with this assignation. One thing is that many of the sherds contained plant temper, which is absent in the Okçular Kale İni assemblage. Further, many vessels were produced in a very coarse technology, with poorly sorted pastes, large inclusions, uneven wall thicknesses, and very rudimentary surface treatment. In all these aspects, the pottery differs from typical Chalcolithic pottery in the
The Chalcolithic (5500-3000 BC)
137
Cide-Şenpazar region and beyond, which is produced with care and nicely smoothed and burnished, even if poorly fired. Dr. Ulf Schoop suggested that at least some of the Derebağ Köy Mağarası ceramics could fit in the Early Iron Age, and this idea was subsequently confirmed by Dr. Herman Genz (personal communication Schoop 20-06-2011; personal communication Genz 09-05-2012). This material will be discussed in chapter 8 and will not concern us here. Later Early Bronze Age and transitional early second millennium BC finds from Derebağ Köy Mağarası will be discussed in chapters 6 and 7. In the final analysis only three sherds51 found at Derebağ Köy Mağarası qualify as possibly dating to the Chalcolithic period,52 two of which are rim fragments and allow the reconstruction of vessel shapes with some confidence (Fig. 5.15). C4030 is part of a simple globular bowl shape, whereas C4125 is part of a jar or pot. Both shapes are of a very common type, and occur at many sites, including Middle and Late Chalcolithic assemblages at Alaca Höyük, Büyükkaya, Orman Fidanlığı (Efe 2001; Schoop 2005),
Fig. 5.15: A selection of Prehistoric sherds from Derebağ Köy Mağarası (S188). Produced by Lorraine McEwan. 51 No chipped stone artifacts were found in this cave. 52 C4030; C4125; C4167.
138
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
and even Neolithic assemblages such as Ilıpınar (Thissen 2001). Perhaps the best fit would be the Late Chalcolithic assemblage of Çadır Höyük (Steadman et al. 2008), but it must be stressed that this is a very tentative date.
5.4.3 Kılıçlı Mağarası The cave of Kılıçlı Mağarası (S88/93) is located in the Loç area close to the village of Çamdibi. The main gallery (S93) of the cave measures approximately 240 metres in length and is full of beautiful stalactites, stalagmites, and pools. In this cave there is ongoing formation of stalactites, stalagmites, sedimentation, and the cave turns into an active river in spring and is therefore clearly not suitable for habitation. In one location some Prehistoric artefacts were found, but these had obviously fallen down from a dry and smaller upper area of the cave, which was labelled site S88. Here we encountered a great amount of robber pits and substantial remains of artefacts. At present, the upper gallery of Kılıçlı Mağarası can only be reached via a hazardous climb and descend, and its geographic position suggests that this gallery was difficult to reach also in the past. This raises the question about the nature of this site, which is very rich in artefacts, and how it is best understood. Kılıçlı Mağarası is probably not a settlement site. Instead, we could think of a usage as a burial or storage cave during Prehistory and in the Bronze Age, and a possible reuse as a burial or storage location in the Byzantine period. In Kılıçlı Mağarası we found some bone fragments that were possibly human, and this in conjunction with isolated finds of a bead and a spindle whorl could support the idea that this is a disturbed cemetery. Regardless of how the site of Kılıçlı Mağarası should be understood, there was a remarkably dense distribution of materials in the cave dating to numerous periods including the Chalcolithic; Early Bronze Age, which constitutes the dominant phase; the second millennium BC; some Iron Age; and Byzantine period ceramics. In this assemblage there were a total of 53 sherds that appear to date to the Chalcolithic on the basis of their fabrics, firing, and surface treatment.53 The fabrics are predominantly tempered with mineral temper - quartz is especially common - but also include plant temper in a minority of sherds. The larger vessels especially have relatively poorly executed surface treatments, frequently consisting of smoothing or wiping and often have large inclusions. On the other hand some finer vessels have better sorted pastes, smaller inclusions and are burnished. Colours range from light brown to grey and firing is often uneven and generally under reducing conditions.
53 These are C2031; C2033; C2037; C2040-2; C2045; C2047; C2923; C2928-30; C2932; C2944; C2968; C2971; C3088-93; C3095-7; C3099-105; C3109-13; C3117; C3161; C3165; C3175; C3195; C3244; C3246-7; C3250-2; C3256; C3289; C3806; C3808-9.
The Chalcolithic (5500-3000 BC)
139
Among the Chalcolithic sherds from Kılıçlı Mağarası only nine have diagnostic features (Fig. 5.16). These include a number of simple bowl shapes,54 of which one has three parallel white diagonal lines painted on the interior.55 There are also a number
Fig. 5.16: Chalcolithic ceramics and chipped stone artefacts found at Kılıçlı Mağarası (S88/S93). Produced by Lorraine McEwan.
54 C2932; C3096. 55 C3096.
140
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
of pots,56 of which one has a clear carination.57 There are two examples of flat bases.58 Finally, there are two vertical rims that appear to have belonged to jars.59 Many parallels can be drawn for these Kılıçlı Mağarası ceramics especially for the Late Chalcolithic. At Çadır Höyük similar bowl and jar shapes occur as well as similar flat bases in the Late Chalcolithic, and one pot has very similar white painted decoration albeit on the exterior rather than the interior of the vessel (Steadman et al. 2008). Very similar white painted decoration on the exterior is also in evidence from Dündartepe, where there are also similar carinated vessels, jar shapes, and bases (Thissen 1993; Alkım et al. 2003: plate 7; Schoop 2005). White painted decoration on the interior of bowls also occurs in İkiztepe complex ‘BB’ dated to the mid-fifth millennium BC (Schoop 2005), but the ceramic shapes of this assemblage do not match those of Kılıçlı Mağarası, where the associated knobs, horned handles, and incised decorations typical for this assemblage are completely absent. Much better parallels exist at Orman Fidanlığı level VII, where similar parallel diagonal white painted decoration was found on the interior of bowls, as are similar jar necks and bases (Efe 2001). The most plausible date therefore for the Kılıçlı Mağarası ceramics discussed here is in the Late Chalcolithic. The two lithics found: L228, a back blade, and L168, a blade, cannot be used to date the assemblage more accurately, and these two chipped artefacts could in fact be dated to later periods of use in Kılıçlı Mağarası. However, it appears that the final part of the Chalcolithic, transitional into the Early Bronze Age, is lacking at Kılıçlı Mağarası, and in our Cide-Şenpazar assemblages in general, given that the typical fruit stands and jars with rim handles documented at a range of sites, such as Çadır Höyük (Steadman et al. 2008), Alaca Höyük, Alişar Höyük, and Çengeltepe (Schoop 2005), and the central Kastamonu survey (Kuzucuoğlu et al. 1997), are completely absent in our collections.
5.4.4 The Chalcolithic in the Cide-Şenpazar Region: By Way of Summary In the three survey seasons in the Cide-Şenpazar region, numerous cave sites were found with Chalcolithic assemblages, whereas surface material was only found in the fields below Okçular Kale. Substantial Chalcolithic collections have only been found in two caves: Okçular Kale İni (S24) and Kılıçlı Mağarası (S88/93). In the preceding discussion it has become clear that dating these assemblages on the basis of comparison with assemblages from other regions is far from straightforward
56 C2033; C2968; C3101. 57 C2033. 58 C3092; C3195. 59 C3111; C3808.
Conclusion
141
due to the fact that much of the Chalcolithic has not been systematically studied in excavations in northern Anatolia and that there are many regional and local features in these assemblages, which of course makes comparison more difficult. In these circumstances absolute dates are of great importance especially for the sixth and fifth millennia BC (Schoop 2005; Düring 2011). On the one hand our survey has been successful in documenting the Chalcolithic period in the Cide-Şenpazar region in that we now have assemblages dating to the early fifth and the early fourth millennia BC, but on the other hand, it is also clear that this understanding relies heavily on a few chance finds of sites where these periods were present. For example, without Okçular Kale İni, we would not have a good Middle Chalcolithic phase in our assemblage. In this light we should also consider the fact that some periods are not represented in our data. In particular we lack data for: first, the second half of the sixth millennium BC; second, the later part of the fifth millennium BC – the so-called ‘horned handle horizon’ discussed above; and, third, the later part of the Late Chalcolithic and the Early Bronze Age I. The lack of data for these periods probably tells us more about how chance factors have influenced our dataset than about patterns of occupation in the past. Thus, our understanding of the Chalcolithic in the Cide-Şenpazar region remains very partial and incomplete.
5.5 Conclusion In the course of CAP we have managed to document assemblages dating to the Early Holocene and Chalcolithic periods, both of which are poorly documented across the Turkish western Black Sea region. Even if our understanding of the societies who have left their imprint on the Cide-Şenpazar region in these periods remains hazy, some tentative conclusions can nonetheless be drawn. One interesting conclusion is that the Cide-Şenpazar region was connected with the interior of Anatolia, as demonstrated by the presence of obsidian from the Galatian sources from the Early Holocene up to the fifth millennium BC. In our material culture, including both the ceramic and lithic assemblages, there are links with assemblages both in central Anatolia and the Marmara region. Remarkably, however, the Cide-Şenpazar assemblages do not show good comparanda with the published data from İkiztepe. It is unclear whether this result is in part due to chance factors. For example, the horned handle horizon of the later fifth millennium BC does stretch from the Aegean to İkiztepe and there is every reason to suggest that similar materials await discovery in the Cide-Şenpazar region. If we can establish the fact that the Cide-Şenpazar region was connected to other regions, such as the Marmara region, the means of these connections remain difficult to assess. In theory seafaring could have been one of the mechanisms, especially given that the same has been argued for Thrace in the fifth millennium BC (Ivanova
142
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
2012), but we lack any kind of positive evidence for the time being. Certainly land based movements can be postulated both in the Early Holocene and the Chalcolithic. The nature of the occupation of the Cide-Şenpazar region in the Early Holocene must remain elusive for the time being, as the evidence is restricted to lithics collected from the surface. To us, it seems plausible that we are dealing with an aceramic period, given the complete absence of associated ceramics. The economy and the degree of sedentism must remain open for now, but the evidence for knapping locations at Aybasan Mahallesi fits best with mobile hunter-gatherer groups. Also we lack strong evidence for implements used for cereal harvesting in the lithic assemblages. By contrast, the Middle Chalcolithic midden deposits from Okçular Kale İni contain large amounts of ash, charcoal, bone, ceramics, and chipped stone deposited over a fairly short period of time, suggesting that we are dealing with domestic refuse of a more or less sedentary community, which need not have been very large. The fact that similar locations were occupied in the Middle Chalcolithic elsewhere in Anatolia, for example at Güvercinkayası and Kanlıtaş by sedentary farming populations, adds credence to this interpretation. Ideally excavations at Okçular Kale İni could shed further light on this issue, for example, to which animals the bones in the midden layers belonged and how they were hunted or managed, and whether there is any evidence for crops. The data for the Late Chalcolithic from Kılıçlı Mağarası are less easy to interpret. If we are dealing with a burial or storage cave, as has been suggested, then the least we can say is that people were present in the area using caves such as Kılıçlı Mağarası and Derebağ Köy Mağarası as repositories as part of their land-use strategies.
Catalogue of Illustrated Artefacts Ceramics C206. S24. Rim fragment of closed pot. Reddish brown homogenous paste with some mineral temper. Surfaces smoothed. Oxidised firing. Handmade. Fabric 7.5 YR 6/4, surfaces 10 YR 5/6. D= unknown. C293. U3. Base fragment. Grey paste with calcareous, black and red grit, and possibly some organic temper. Exterior burnished. Reduced firing. Handmade. Fabric 7.5 YR 6/1, exterior surface 10 YR 5/4, interior surface 7.5 YR 4/1. Base D= 20cm. C756. S24. Rim fragment of closed (cooking?) pot. Dark red paste with quartz and some calcareous grits. Differential firing. Handmade. Fabric 7.5 YR 2.5/1, exterior 7.5 YR 3/1. D=16cm. C799. S24. Rim fragment of bowl. Light brown homogeneous paste with small mineral temper. Both surfaces wiped horizontally, exterior uneven. Differential firing. Handmade. Fabric light brown grading into dark grey, no Munsell. Exterior 7.5 YR 5/2. D= 25cm. C802. S24. Rim fragment of a large jar. Grey homogeneous paste with minute mineral temper. Surfaces scraped and wiped. Reduced firing. Handmade. Fabric Gley 2 4/1, exterior 2.5 YR 5/6, interior 7.5 YR 5/4. D=16cm. C805. S24. Base fragment. Reddish brown paste with grog. Exterior smoothed, interior wiped. Oxidised firing. Handmade. Fabric Gley 3 2/1, exterior 2.5 YR 5/6. D= 9 cm.
Catalogue of Illustrated Artefacts
143
C813. S24. Rim fragment of closed pot. Greyish brown paste with mineral temper and grog. Exterior burnished, interior wiped. Differential firing. Handmade. Fabric 2.5 YR 5/6 and 5 YR 5/1, exterior 7.5 YR 6/4. D= 24cm. C816. S24. Handle fragment with false twist. Grey paste with some mineral temper. Burnished. Reduced firing. Handmade. Fabric 10 YR 5/1, surface 10 YR 5/3. C817. S24. Rim fragment of plate. Dark grey paste with small mineral temper. Exterior burnished, interior wiped. Reduced firing. Handmade. Fabric 5 YR 3/1, exterior surface 5 YR 2.5/1, interior surface 5 YR 3/3. D= 24cm. C818. S24. Base fragment. Grey paste with some mineral temper up to 4 mm in size. Exterior has wash and is burnished, interior wiped. Reduced firing. Handmade. Fabric 5 YR 3/1, exterior surface Gley 2 2.5/1, interior surface 7.5 YR 6/4. D= 9cm. C836. S24. Rim fragment of bowl. Grey paste with mineral temper, up to 2 mm in size. Both surfaces have a wash. Reduced firing. Handmade. Fabric Gley 2 3/1, surfaces 5 YR 5/4. D= unknown. C2033. S93. Body fragment of bowl(?). Brown paste with small to medium subangular quartz temper. Exetrior burnished, interior smoothed. Differential firing. Handmade. Fabric 7.5 YR 7/6, surface 7.5 YR 3/1. C2887. S24. Fragment of plate or lid. Dark brown paste with medium-coarse dense angular quartz and organic material. Surfaces have streaky burnish on slib. Reduced firing. Handmade. Fabric 2.5 YR ¾, surfaces 10 R 4/6. C2888. S24. Rim fragment of closed pot. Dark grey paste with dense medium angular quartz. Interior and exterior burnished. Reduced firing. Handmade. Fabric variable colour, no Munsell, surface 5 YR 25/1-2. D= unknown. C2890. S24. Rim fragment of closed pot with handle/knob scar. Dark grey paste with dense medium angular quartz. Interior and exterior burnished. Reduced firing. Handmade. Fabric variable colour, no Munsell, surface 5 YR 25/1-2. D= 16cm. C2891. S24. Body fragment with handle/lug scar. Dark brown paste with small to medium angular and subangular quartz. Exterior burnished. Reduced firing. Handmade. Variable colours, no Munsell. C2904. S24. Rim fragment of bowl. Dark grey paste with small to medium angular quartz. Exterior and interior burnished. Differential firing. Handmade. Fabric has variable colours, no Munsell, surface 5 YR 2.5/1. D= unknown. C2911. S24. Base fragment. Reddish brown coarse and brittle fabric with angular stone grit. Exterior smoothed. Differential firing. Handmade. Fabric 2.5 YR 3/1, surfaces 2.5 YR 5/1. D= 14 cm. C2922. S24. Rim fragment of a jar. Reddish brown paste with small to medium angular and subangular quartz. Exterior has wash and some burnishing, interior is smoothed. Differential firing. Handmade. Fabric 10 R 5/8 to 5 R 6/4, exterior 5 YR 5/3. D= 14cm. C2925. S24. Rim fragment of bowl with impression decoration below rim. Dark brown paste with many grits, of dense medium to large angular quartz. Surface is smoothed / has wash. Reduced firing. Handmade. Fabric 2.5 YR 4/4, surfaces 6 YR 5/3 -5 YR 3/1. C2932. S93. Rim fragment of bowl. Brown paste with small grits and possible plant temper. Exterior and interior burnished. Reduced firing. Handmade. Fabric has variable colours, no Munsell, surface 2.5 YR 5/8. D= 14cm. C2968. S93. Rim fragment of jar. Greyish brown paste with some small to medium black and red grit. Exterior and interior burnished. Reduced firing. Handmade. Fabric and surface colours variable, no Munsell taken. D= 14cm. C2997. S24. Fragment of rim of jar. Dark grey paste with dense small to medium angular quartz temper. Exterior burnished and wiped. Reduced firing. Handmade. Fabric Gley 2 /5BP, exterior surface 7.5 YR 5/2-4/2, interior surface Gley2 4/5BP. D= 18cm.
144
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
C2999. S24. Fragment of rim of jar. Grey-brown paste with dense small to medium angular to subangular quartz, not well sorted. Surfaces have a wash and are burnished. Reduced firing. Handmade. Fabric Gley 1/3, surfaces 5 YR 3/1. D= 18cm. C3005. S24. Body fragment with handle attachment. Buff brown paste with large amounts of small quartz and grits. Exterior and interior smoothed. Reduced firing. Handmade. Fabric has variable colours, no Munsell, surfaces 7.5 YR 5/4. C3010. S24. Rim fragment of closed pot. Brownish grey paste with dense small angular quartz. Exterior and interior surfaces burnished. Reduced firing. Handmade. Fabric has variable colours, no Munsell taken, surfaces 5 YR 6/6. D= unknown. C3069. S88. Rim fragment of bowl with white painted decoration. Grey paste with some quartz and plant temper. Exterior burnished and painted. Reduced firing. Handmade. Fabric 5 YR 2.5/1, surface 5 YR 2.5/1. D= unknown. C3078. S24. Rim fragment of closed pot. Surfaces smoothed and decorated with impressions. Oxidised firing. Handmade. No Munsell colours recorded. D= unknown. C3092. S88. Fragment of a base. Light grey poorly sorted paste with some quartz and lime. Surfaces wiped and scraped. Differential firing. Handmade. Fabric 2.5 YR 5/2, surfaces 7.5 YR 6/4. D= 14cm. C3101. S88. Rim fragment of cooking (?) pot. Grey poorly sorted paste with large lime inclusions. Surface burnished. Differential firing. Handmade. Fabric 7.5 YR 4/1, surface 2.5 YR 5/6. D= 19cm. C3111. S88. Rim fragment of a jar. Grey well sorted with some minute lime. Surfaces well burnished throughout. Reduced firing. Handmade. Fabric Gley 2 5/10G, surfaces Gley 2 5/10G. D= 16cm. C3195. S88. Fragment of a base. Grey paste with dense medium subangular quartz temper. Exterior and interior burnished. Reduced firing. Handmade. Fabric 5 YR 4/1, exterior surface 7.5 YR 6/4, interior surface 6 R 5/6. D= unknown. C3808. S88. Fragment of rim of jar. Grey paste with dense quartz and grit. Surfaces smoothed. Reduced firing. Handmade. Fabric has variable colours, no Munsell taken, surface 2.5 YR 6/6. D= 16cm. C4030. S188. Rim fragment of a bowl. Grey paste with some lime. Exterior burnished, interior smoothed. Differential firing. Handmade. Fabric 2.5 YR 4/8 to 2.5 YR 2.5/1, surface 5 YR 3/2 to 5 YR 5/6. D= unknown. C4125 S188. Rim fragment of jar. Grey poorly sorted paste with some lime and grit. Surfaces burnished. Differential firing. Handmade. Fabric 7.5 YR 7/6 to 7.5 YR 3/1, surface 5 YR 3/1 to 5 YR 7/8 to 7.5 YR 6/2. D= approximately 14cm. Object F56. S132.Stone chisel, polished. Stone variety undetermined: fine grained crystalline green stone with small white quartz-like inclusions and one or two narrow white veins. Some use wear. Oval section, polished to a shine. Facets hardly visible and edge still sharp. 53 by 20 by 12 mm. Lithics L2. S18. Flat round flake with some surface retouch. Denticulate retouch on 50% of lateral sides. Denticulate. 40 by 32 by 7 mm. Brown medium fine flint. L3. U3. Crescent-shaped blade. Clear step retouch. Dorsal surface pattern: simple. Soft percussion. Gloss on sharp side. Backed knife. 41 by 14 by 7mm. Grey medium fine flint. L6. U8. Long regular blade. Dorsal surface pattern: simple. Steep retouch on one lateral side (notching?). Soft percussion. Fe-spots on ventral side and ribs. Cutting tool? Retouched blade. 75 by 21 by 7mm.
Catalogue of Illustrated Artefacts
145
L16. U7. Finely shaped bladelet. Material very fine grained but relatively soft. Broken on distal and proximal ends. Dorsal surface pattern simple with 4 negatives. Micro retouche on both lateral sides. Soft percussion. Retouched blade.33 by 18 by 4mm. Brown fine flint. L23. U30. Distal end of narrow bladelet. Stepped retouche on lateral side. Dorsal surface pattern simple. Gloss on non-retouched lateral side. Retouched blade. 14 by 15 by 2mm. Yellow fine flint. L27. U30. Approximately half of a very small core. Bladelet negatives 4-2mm wide. Core. 21 by 14 by 7mm. White fine flint. L28. U30. Part of possible bullet core. Three negatives 9-12 mm wide. Core. 24 by 23 by 13mm. Grey medium flint. L41. U80. Bottom part of micro core, probably overshoot flake. Negatives 5-11 11 mm wide. Core. 17 by 11 by 7mm. Blueish fine flint. L45. U187. Distal end of blade. Distal side step-retouched into scraper-like tool. Lateral sides small notches. Proximal end upturned hinge. Dorsal surface pattern: simple. Trapezoidal profile. End scraper. 21 by 18 by 5mm. Brown fine flint. L51. U205. Distal and medial part blade. Dorsal surface pattern: centripetal. Ventral side one negative same direction as flake. One notch on lateral side. Notched flake. 36 by 25 by 5mm. Grey fine flint. L58. U177. Diagonally struck blade. Soft percussion. Core preparation blade. Dorsal surface pattern: simple and side (centripetal?). Patinated. Core trimming element. 36 by 35 by 7mm. Brown fine flint. L104. U174. Hard percussion. Dorsal surface pattern: simple + contra. 40% cortex. Ventral side large scar. Distal end missing due to retouch. Lateral and distal side step-retouched into “convex scraper”-like tool. Scraper. 35 by 30 by 7mm. Grey fine flint. L108. S24. Soft hammer percussion. Finely struck blade. Distal end partly hinged. Dorsal surface pattern: simple. Blade. 51 by 21 by 10mm. Brown fine flint. L109. S24. Soft hammer percussion. Possibly core-rejuvenation blade. Very fine glass-like material. Dorsal surface pattern: simple. Blade. 48 by 20 by 4mm. Grey fine flint. L110. S24. Medial part of blade. Proximal end broken horizontally, distal end broken off diagonally. Use retouch on longer lateral side, a lot of gloss. Use wear on both lateral sides. Dorsal surface pattern: simple. Blade. 26 by 26 by 5mm. Brown medium fine flint. L114. U43. Very thin, finely struck bladelet. Distal part missing. Soft hammer percussion. Dorsal surface pattern: simple. Surface retouch on proximal end of dorsal side. Knife. 26 by 12 by 3mm. Grey fine flint. L123. U469. Flake. Soft percussion. Flake 26 by 21 by 5mm. Grey fine flint. L127. U469. Tip of point. A lot of surface retouch. Point / spearhead. 32 by 22 by 11mm. Brown medium fine flint. L131. Gr4. Cutting implement. Well made. Crescent shaped lateral side. Knife. 51 by 51 by 1 mm. Purple fine flint. L132. Gr4. Notched blade. 60 by 28 by 12mm. Brown fine flint. L133. Gr4. Discoid core. Core. 34 by 33 by 19mm. Yellow fine flint. L136. Gr4. Flake. 8 by 6 by 2mm. Obsidian, transparent black. L137. Gr4. Flake. 6 by 6 by 2mm. Obsidian, transparent black. L140. Gr4. Retouched flake. 56 by 45 by 10mm. Grey medium fine flint. L144. U559. Part of large spear/arrowhead. Lots of surface retouch. Point / arrowhead. 45 by 28 by 7mm. Brown fine flint. L146. Gr4. Backed knife; back retouched from lateral side and rib. Backed knife. 36 by 9 by 6mm. Green fine flint.
146
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
L147. Gr4. Distal part of blade. Notched blade. 32 by 17 by 4mm. Brown fine flint. L149. Gr4. Notched blade. 38 by 17 by 6mm. Grey fine flint. L162. Gr5. Part of core, heavily burnt. Core. 44 by 25 by 15mm. Green fine flint. L164. Gr4. Very fine retouching on one lateral side. Cruder on other lateral side. Retouched blade. 29 by 20 by 4mm. Purple fine flint. L165. Gr4. Platform rejuvenation flake. Ventral side on natural fissure. Core trimming element. 46 by 36 by 11mm. Brown fine flint. L168. S93. Dorsal surface pattern simple. 6 parallel negatives. Soft percussion. Blade. 30 by 19 by 3mm. Brown fine flint. L171. U434. Crude core with 3 negatives. Core. 44 by 36 by 27mm. Brown medium coarse flint. L172. U434. Proximal end of core. Platform rejuvenation flake. Core trimming element. 32 by 29 by 8mm. Brown medium coarse flint. L174. Gr1. Dorsal surface pattern simple. On blade core? One large hinged negative on ventral side. Retouched flake. 39 by 36 by 5mm. Grey fine flint. L175. U624. Patinated burin. Later used as bladecore and scraping/cutting implement. Burin. 35 by 30 by 12mm. Yellow fine flint. L179. S24. Retouched flake. 44 by 38 by 7mm. Brown fine flint. L184. S24. Backed knife? Retouched blade. 27 by 17 by 3mm. Green medium coarse flint. L186. S24. Curved bladelet. Dorsal surface pattern simple. six negatives. Blade. 29 by 9 by 3mm. Obsidian, transparent black. L187. Gr1. Part of small core. Core. 22 by 13 by 12mm. Brown medium coarse flint. L190. U722. Steep retouch on one side. Denticulate retouch on other side. Scraper. 38 by 30 by 5mm. Green fine flint. L193. U722. Six negatives. Original nodule probably small: cortex on three sides. Core. 26 by 20 by 9mm. Obsidian, transparent black. L196. S24. Knife. 73 by 21 by 4mm. Brown fine flint. L197. S24. Flake. 16 by 14 by 1mm. Obsidian, transparent black. L208. U714. Retouched on both lateral sides. Proximal and distal end missing. Retouched flake. 42 by 35 by 2mm. Purple fine flint. L221. Gr4. Part of backed knife? Knife. 25 by 23 by 6mm. Green fine flint. L224. Gr4. Flake. 17 by 9 by 1 mm. Obsidian, transparent black. L225. Gr4. Core trimming element. 43 by 32 by 8mm. Brown fine flint. L227. S82/U720. Core. 36 by 30 by 22mm. Brown fine flint. L248. S98/Gr7. Small flake, pressure-retouched on distal end only on ventral side to form, a very straight edge. Scraper. 29 by 26 by 5mm. Brown fine flint. L258. U1053. Small bladelet, medial part, proximate and distal ends deliberately broken off. Three parallel negatives on dorsal side. Use retouch all over both lateral sides. Retouched blade. 29 by12 by 4mm. Brown fine flint. L263. U1052. Flake with lots of surface retouch. Long microblades running over dorsal side. Steep retouch on one lateral side, some retouch on proximal side as well, perhaps for hafting. Most rounding to be found here as well. Scraper. 32 by 25 by 1 mm. Brown fine flint. L265. S132/U1031. Long, heavily steep retouched blade. Retouched on all sides. Rounded at the very edges of lateral sides. Retouched blade. 78 by 21 by 8mm. Brown fine flint. L288. Gr8. Almost symmetrical triangle arrowhead with white patination. Surface retouch by pressure flaking. Base of arrowhead is distal end with cortex. Point / arrowhead. 41 by 24 by 6mm. Brown fine flint.
Bibliography
147
Bibliography Aitken, M.J. 1997. Luminescence dating. In R. E. Taylor and M. J. Aitken, eds., Chronometric Dating in Archaeology. New York, Plenum Press: 183-215. Alkim, U.B., Alkim, H., and Bilgi, Ö. 1988. Ikiztepe I, Birinçi ve İkinçi Dönem Kazıları (1974-1975). Ankara, Türk Tarih Kurumu. — 2003. Ikiztepe II, Üçüncü, Dördüncü, Beşinci, Altıncı, Yedinci Dönem Kazıları (1976-1980). Ankara, Türk Tarih Kurumu. Anderson, P.C., Chabot, J., and van Gijn, A., 2004. The functional riddle of ‘glossy’ Canaanean blades and the Near Eastern threshing sledge. Journal of Mediterranean Archaeology 17: 87-130. As, A. van, and Wijnen, M. 1995. The Neolithic and Chalcolithic pottery from Ilıpınar’s phases X-V: A technological study. In J.J. Roodenberg, ed., The Ilıpınar Excavations I. Istanbul, Nederlands Historisch Archaeologisch Istituut te Istanbul: 77-107. Balkan-Atlı, N. 1994. La néolithisation de l’Anatolie. Istanbul, Institut Français d’Etudes Anatoliennes d’Istanbul. Baykal-Seeher, A. 1996. Die lithischen Kleinfunde. In M. Korfmann, ed., Demircihüyük: die Ergebnisse der Ausgrabungen, 1975-1978. Band 4: Die Kleinfunde. Mainz am Rhein, Philip von Zabern: 7-206. — 2001. Die Lithik von Orman Fidanlığı. In T. Efe, ed., The Salvage Excavations at Orman Fidanlığı, A Chalcolithic Site in Inland Northwestern Anatolia. Istanbul, Tarih Arkeoloji Sanat ve Kültür Mirasını Koruma Vakfı: 159-86. Bertram, J.K. 2008. Ahlatlıbel, Etiyokuşu, Koçumbeli – Zur Neubewertung der Ankara-Gruppe. TÜBA-AR 9: 73-84. Bıçakçı, E., Altınbilek-Algül, C., Balcı, S., and Godon, M. 2007. Tepecik-Çiftlik. In M. Özdoğan and N. Başgelen, eds., Anadolu’da Uygarlığın Doğuşu Avrupaýa Yayılımı Türkiye’de Neolitik Dönem. Yeni Kazılar, Yeni Bulgular. Istanbul, Arkeoloji ve Sanat Yayınları: 237-54. Bigazzi, G., Oddone, M., and Yengigil, Z. 1995. A provenance study of obsidian artifacts from Ilıpınar. In J.J. Roodenberg, ed., The Ilıpınar Excavations I. Istanbul, Nederlands Historisch Archaeologisch Istituut te Istanbul: 143-50. Bigazzi, G., Poupeau, G., Yeğingil, Z., and Bellot-Gurlet, L. 1998. Provenance studies of obsidian artefacts in Anatolia using the fission-track dating method: an overview. In M.-C. Cauvin, A. Gourgaud, B. Gratuze, N. Arnaud, G. Poupeau, J.-L. Poidevin and C. Chataigner, eds., L’obsidienne au Proche et Moyen Orient, du volcan á l’outil. Oxford, Archaeopress: 69-89. Bintliff, J.L., Howard, P., and Snodgrass, A. 1999. The hidden landscape of prehistoric Greece. Journal of Mediterranean Archaeology 12: 139-68. Bittel, K. 1934. Prähistorische Forschung in Kleinasien. Istanbul, Universum. — 1950. Grundzüge der Vor- und Frühgeschichte Kleinasiens. Tübingen, Verlag Ernst Wasmuth. Boeijen, E. 2011. Rehydroxylatie. Een Aardewerkdateringsmethode toegepast op het Aardewerk van het Cide Archaeological Project. Unpublished BA dissertation, Leiden University. Carter, T., Conolly, J., and Spasojevíc, A. 2005. The chipped stone. In I. Hodder, ed., Changing Materialities at Çatalhöyük, Reports from the 1995-1999 Seasons. Cambridge, McDonald Institute: 221-83. Chataigner, C. 1998. Sources des artefacts du Proche Orient d’apres leur characterisation geochemique. In M.-C. Cauvin, A. Gourgaud, B. Gratuze, N. Arnaud, G. Poupeau, J.-L. Poidevin and C. Chataigner, eds., L’obsidienne au Proche et Moyen Orient, du volcan á l’outil. Oxford, Archaeopress: 273-350. Çilingiroğlu, A., and Çilingiroğlu, C. 2007. Ulucak. In M. Özdoğan and N. Başgelen, eds., Anadolu’da Uygarlığın Doğuşu Avrupaýa Yayılımı Türkiye’de Neolitik Dönem. Yeni Kazılar, Yeni Bulgular. Istanbul, Arkeoloji ve Sanat Yayınları: 361-72.
148
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
Conolly, J. 1999. The Çatalhöyük Flint and Obsidian Industry, Technology and Typology in Context. Oxford, Archaeopress. Doonan, O., Casson, A., and Gantos, A. 2007. Sinop province archaeological project: Report on the 2006 field season. Araştırma Şonuçları Toplantışı 25/2: 133-50. Düring, B. S. 2008. The Early Holocene occupation of North-Central Anatolia between 10,000 and 6,000 BC cal: Investigating an archaeological terra incognita. Anatolian Studies 58: 15-46. — 2011. The Prehistory of Asia Minor: From Complex Hunter-Gatherers to Early Urban Societies. Cambridge, Cambridge University Press. Düring, B., and Glatz, C. 2010. The Cide Archaeological Project 2009: First results. Anatolia Antiqua 18: 203-13. Düring, B.S., and Gratuze, B. 2013. Obsidian exchange networks in Prehistoric Anatolia: New data from the Black Sea region. Paléorient 39(2): 173-82. Efe, T. 1987. Demircihüyük: die Ergebnisse der Ausgrabungen, 1975-1978, Die frühbronzezeitliche Keramik der jüngeren Phasen (ab Phase H). Mainz am Rhein, Philipp von Zabern. Efe, T. ed. 2001. The Salvage Excavations at Orman Fidanlığı. A Chalcolithic Site in Inland Northwestern Anatolia. Istanbul, Tarih Arkeoloji Sanat ve Kültür Mirasını Koruma Vakfı. Gabriel, U., Aslan, R., and Blum, S.W.E. 2004. Alacaligöl: Eine neuentdeckte Siedlung des 5. Jahrtausend v. Chr. in der Troas. Studia Troica 14: 121-33. Gatsov, I. 2009. Prehistoric Chipped Stone Assemblages from Eastern Thrace and the South Marmara Region 7th-5th mill. B.C. Oxford, Archaeopress. Göksu-Ögelman, H.Y., and Ukav, I. 1986. Age determination of Ikiztepe sherds by thermoluniscence. PACT 15: 111-5. Gülçur, S., and Firat, C. 2005. Spatial analysis of Güvercinkayası, A Middle Chalcolithic hilltop settlement in northwestern Cappadocia, A preliminary report. Anatolia Antiqua 13: 41-52. Işın, M.A. 1998. Sinop regional field survey. Anatolia Antiqua 6: 95-139. Ivanova, M. 2012. Perlious waters: Early maritime trade along the western coast of the Black Sea (fifth millenium BC). Oxford Journal of Archaeology 31: 339-65. Karauğuz, G., and Düring, B.S. 2009. A note on the prehistory of the Devrek region, Northern Turkey. Anatolica 35: 153-65. Korfmann, M., Baykal-Seeher, A., and Obladen-Kauder, J. 1996. Demircihüyük: die Ergebnisse der Ausgrabungen, 1975-1978. Band 4: Die Kleinfunde. Mainz am Rhein, Philip von Zabern. Kuzucuoğlu, C., Marro, C., Özdoğan, A., and Tibet, A. 1997. Prospection archéologique franco-turque dans la région de Kastamonu (Mer Noire), Deuxième rapport préliminaire. Anatolia Antiqua 5: 275-306. Lichter, C. 2005. Western Anatolia in the Late Neolithic and Early Chalcolithic, the actual state of research. In C. Lichter, ed., How did Farming reach Europe? Anatolian-European Relations from the Second Half of the 7th through the First Half of the 6th Millennium BC. Istanbul, Ege Yayınları: 59-74. Liritzis, Y., and Galloway, R.B. 1982. Thermoluminiscence dating of Neolithic Sesklo and Dimini, Thessaly, Greece. PACT 6: 450-9. Lloyd, S., and Mellaart, J. 1962. Beycesultan Volume I, The Chalcolithic and Early Bronze Age Levels. London, British Institute of Archaeology at Ankara. Marro, C. 2000. Archaeological survey in the Kastamonu Region, Turkey, Remarks on the Preclassical geography of the Southern Black Sea. In P. Matthiae, A. Enea, L. Peyronel, and F. Pinnock, eds., Proceedings of the First International Congress of the Archaeology of the Ancient Near East. Rome, Università degli studi di Roma ‘La Sapienza’: 945-62. Marro, C., Özdoğan, A., and Tibet, A. 1996. Prospection archéologique franco-turque dans la région de Kastamonu (Mer Noire), Premier rapport préliminaire. Anatolia Antiqua 4: 273-90.
Bibliography
149
Matthews, R. 2009. Silent centuries: Pahlagonia from the Palaeolithic to the Early Bronze Age, 200,000-2000 BC. In R. Matthews and C. Glatz, eds., At Empires’ Edge: Project Paphlagonia. Regional Survey in North-Central Turkey. London, British Institute at Ankara: 75-105. Orthmann, W. 1963. Die Keramik der frühen Bronzezeit aus Inneranatolien. Berlin, Gebr. Mann Verlag. Özdoğan, M. 1999. Northwestern Turkey: Neolithic cultures in between the Balkans and Anatolia. In M. Özdoğan and N. Başgelen, eds., The Neolithic of Turkey. Istanbul, Arkeoloji ve Sanat Yayınları: 203-24. — 2002. On arrows and sling missiles: What happened to the arrows? In R. Aslan, S. Blum, G. Kastl, F. Schweizer, and D. Thumm, eds., Mauerschau, Festschrift für Manfred Korfmann. RemshaldenGrunbach, Bernhard Albert Greiner: 437-44. — 2007. Coastal changes of the Black Sea and Sea of Marmara in archaeological perspective. In V. Yanko-Hombach, A. S. Gilbert, N. Panin, P. M. Dolukhanov, eds., The Black Sea Flood Question: Changes in Coastline, Climate and Human Settlement. Dordrecht, Springer: 651-69. Özdoğan, M., and Gatsov, I. 1998. The Aceramic Neolithic period in Western Turkey and in the Aegean. Anatolica 24: 209-32. Palumbi, G. 2011. The Chalcolithic of Eastern Anatolia. In S. Steadman and G. McMahon, eds., Oxford Handbook of Anatolian Studies, 8000-323 BC. Oxford, Blackwell: 205-26. Parzinger, H. 1993. Zur Zeitstellung der Büyükkaya-Ware, Bemerkungen zur vorbronzezeitlichen Kulturfolge Zentralanatoliens. Anatolica 19: 211-29. Peltenburg, E., and Wasse, A., eds. 2004. Neolithic Revolution: New Perspectives on Southwest Asia in Light of Recent Discoveries on Cyprus. Oxford, Oxbow. Poidevin, J.-L. 1998. Les gisements d’obsidienne de Turquie et de Transcaucasie, Geologie, geochimie et chronometrie. In M.-C. Cauvin, A. Gourgaud, B. Gratuze, N. Arnaud, G. Poupeau, J.-L. Poidevin and C. Chataigner, eds., L’obsidienne au Proche et Moyen Orient, du volcan á l’outil. Oxford, Archaeopress: 105-203. Roodenberg, J. 1999. Ilıpınar, An early farming village in the Iznik Lake basin. In M. Özdoğan and N. Başgelen, eds., The Neolithic of Turkey. Istanbul, Arkeoloji ve Sanat Yayınları: 193-202. Schachner, A. 2012. Die Ausgrabungen in Boğazköy-Hattusa 2011. Archäologischer Anzeiger: 85-137. Schoop, U.-D. 2005. Das anatolische Chalkolithikum. Remshalden, Albert Greiner Verlag. — 2008. Ausgrabungen in Çamlıbel Tarlası 2007. Archäologischer Anzeiger: 148-157. — 2009. Ausgrabungen in Çamlıbel Tarlası 2008. Archäologischer Anzeiger: 56-67. — 2010. Ausgrabungen in Çamlıbel Tarlası 2009. Archäologischer Anzeiger: 191-201. — 2011. Çamlıbel Tarlası, ein metallverarbeitender Fundplatz des vierten Jahrtausends v. Chr. im nördlichen Zentralanatolien. Anatolian Metal 5: 53-68. Shimelmitz, R. 2011. Lithic industries of prehistoric Anatolia: An overview. In Y. Yakar, ed., Reflections of Ancient Anatolian Society in Archaeology, From Neolithic Village Communities to EBA Towns and Polities. Istanbul, Homer: 477-536. Steadman, S.R., McMahon, G., and Ross, J.C. 2007. The Late Chalcolithic at Çadır Höyük in central Anatolia. Journal of Field Archaeology 32: 515-58. Steadman, S.R., Ross, J.C., McMahon, G., and Gorny, R.L. 2008. Excavations on the north-central plateau: The Chalcolithic and Early Bronze Age occupation at Çadır Höyük. Anatolian Studies 58: 47-86. Swiny, S., ed. 2001. The Earliest Prehistory of Cyprus, From Colonization to Exploitation. Boston, American Schools of Oriental Research. Thissen, L. 1993. New insights in Balkan-Anatolian connections in the Late Chalcolithic, Old evidence from the Turkish Black Sea littoral. Anatolian Studies 43: 207-37. — 2001. The pottery of Ilıpınar, Phases X to VA. In J.J. Roodenberg and L. Thissen, eds., The Ilıpınar Excavations II. Istanbul, Nederlands Historisch Archaeologisch Istituut te Istanbul: 3-154.
150
The Early Prehistory of the Cide-Şenpazar Region: ca. 10,000-3000 BC
Tourloukis, V. 2010. The Early and Middle Pleistocene Archaeological Record of Greece, Current Status and Future Prospects. Leiden, Leiden University Press. Türkcan, A.U. 2010. Kanlıtaş Höyük ve cıvarı (İnönü, Eskişehir) yüzey araştırması. Araştırma Sonuçları Toplantısı 28/1: 303-28. von der Osten, H.H. 1937. The Alishar Hüyük, Seasons of 1930-32. Part I. Chicago, University of Chicago Press. Welton, L. M. 2011. Mobility and Social Organization on the Ancient Anatolian Black Sea Coast: An Archaeological, Spatial and Isotopic Investigation of the Cemetery at İkiztepe, Turkey. Unpublished PhD thesis, University of Toronto. Whittaker, J., Kamp, K., and Yilmaz, E. 2009. Çakmak revisited: Turkish flintknappers today. Lithic Technology 34: 93-110. Wilson, M.A., Carter, M.A., Hall, C., Hoff, W.D., Ince, C., and Savage S.D. 2009. Dating fired-clay ceramics using long-term power law rehydroxylation kinetics. Proceedings of the Royal Society online: 1-9.
Claudia Glatz, with contributions by Michele Massa
6 The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC The Early Bronze Age was a period of profound social transformation across many parts of Anatolia. Beginning in the later part of the EB II (ca. 2600 – 2300 BC) and culminating in the EB III (ca. 2300-2000 BC), western and central Anatolia experienced a dramatic population increase manifested in peaks in settlement numbers and the emergence of settlement hierarchies. Internal spatial differentiation of settlements, incipient monumentality, defensive architecture, hoard deposits, ostentatious burials and evidence for feasting all point to the competitive strategies of display, commensality and confrontation through which elites establish and maintain themselves (e.g. Sahoğlu 2005; Rahmstorf 2006; Çevik 2007; Bachhuber 2009; Zimmerman 2009). These developments took place at different points during the late third millennium BC in different parts of Anatolia and are documented and understood to varying degrees in different regions (Düring 2011; Schoop 2011). Early Bronze Age communities in the mountainous northern fringes of Turkey, their social organisation and cultural traditions as well as their involvement in, or isolation from, the above developments remain little explored. In this chapter, I present the evidence for the later Early Bronze Age of the Cide-Şenpazar region and attempt to sketch a tentative picture of Early Bronze Age society in the area.
6.1 Chronological Considerations As is the case for preceding and subsequent periods, the lack of well-dated assemblages all along the southern shores of the Black Sea presents a major obstacle for the dating of survey finds (also chapter 5). The dating of the Cide Archaeological Project’s (CAP) later Early Bronze Age ceramic data, therefore, relies exclusively on typological comparison with type-sites located at several hundred kilometres’ distance, such as Troy and Küllüoba in the west, or Alaca Höyük and Alişar Höyük on the central Anatolian plateau. Excavated Early Bronze Age sites in the closer vicinity of Cide-Şenpazar include the Yassıkaya cave in Ereğli, Kınık-Kastamonu, Boyabat-Kovuklukaya in Sinop province and İkiztepe in the Bafra plain. They are, however, all plagued by chronological problems and uncertainties, resulting in an increasingly confusing chronological maze. This includes the regional and supra-regional reshuffling of relative chronological correlations following several different, but equally substantial, backward revisions of İkiztepe and other central Anatolian materials (in Orthmann 1963a) from the Early Bronze Age into the Chalcolithic (Thissen
152
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
1993; Schoop 2005: 308) and the introduction of absolute dates at Troy (Kromer et al. 2003: 48, figs. 4-5). Comparison with other regions bordering the Black Sea holds yet more complications due to the traditional use of regional cultural chronologies, which generally lack absolute chronological anchors. As a result, our survey evidence can only be dated in general terms. The pottery here identified as belonging to the later Early Bronze Age is handmade, with very occasional hints of the use of a rotary device. It is almost exclusively mineral tempered with fabrics varying from very coarse to fine. Fabric colours range from yellowish-buff to brown and red. Many pieces have dark yellow, brown or red burnished slips on one or both surfaces. Also characteristic are a range of plastic decorations such as finger impressions, applied bands arranged in geometric motifs and twisted handles. A handful of body sherds, which may belong to the Early Bronze Age, carry painted decorations, but none have easily identifiable parallels elsewhere. Vessel types range from holemouth pots and bowls to jars of varying sizes, jugs and cups. No painted Intermediate/Cappadocian wares which are characteristic of the final Early Bronze Age at Kültepe (Emre 1964; Özgüç 1986), Alişar Höyük (von der Osten 1937: 230-58) and to a lesser extent also at Boğazköy (Orthmann 1963b), were found in the Cide-Şenpazar region. Pinpointing the final Early Bronze Age is, therefore, difficult and several pieces discussed in this and the following chapter may fall on either side of the transition from the third to the second millennium BC.
6.2 Ceramic Assemblages from the Later Early Bronze Age in the Cide-Şenpazar Region All the pottery dating to the later part of the Early Bronze Age found during CAP was recorded in cave contexts. This parallels the situation for the preceding periods at least with regards to ceramic data (chapter 5) and contrasts with the following second millennium BC (chapter 7). The most extensive collection of EB II-III pottery comes from the cave of Kılıçlı Mağarası (S88/S93); smaller quantities were recorded at Okçular Kale İni (S24) and Derebağ Köy Mağarası (S188) (Fig. 6.1).
6.2.1 Kılıçlı Mağarası By far the richest and most diverse assemblage that can be dated to the later part of the Early Bronze Age was found at the inland cave site of Kılıçlı Mağarası in the Loç/Çamdibi area of Cide. With the exception of a few fragments from a lower part of the cave (S93), pottery finds concentrated in a second, upper gallery (S88) (Fig. 6.2). Here illicit excavations have extensively disturbed cultural deposits consisting of densely packed sandy soils mixed with reddish/yellowish clayey lenses, including
Ceramic Assemblages from the Later Early Bronze Age in the Cide-Şenpazar Region
0
5
153
10 Km
Derebağ Köy Mağarası
Okçular Kale İni Kılıçlı Mağarası
Fig. 6.1: Early Bronze Age sites in the CAP survey area. Produced by Toby C. Wilkinson.
some possible plaster layers, and fragments of charcoal. The majority of the ceramic fragments from this rich archaeological deposit can be assigned a date in the second half of the third millennium BC. The cultural connections, which these finds attest to are surprisingly wide-ranging and include western and central Anatolia, Thrace and the Balkans. The Kılıçlı assemblage includes a rim fragment of a handmade globular bowl with simple tapering rim,1 which finds parallels across Anatolia during the third millennium BC (e.g. von der Osten 1937: fig. 248.14; Topbaş et al. 1998: 37.1-3). Simple bowls of this type are attested in the survey and excavation records of neighbouring areas, including inner Kastamonu (Marro et al. 1998: pl. III 1, 11), Çankırı (PS015, PS111 Matthews 2009: figs. 33.9; 3.34:2/3), the cave site of Yassıkaya in Zonguldak (Efe 2004: fig. 1: 1-5) and the site of Koçagöz near Sinop (Bauer 2006: fig. 3). The rim sherd of a small, handmade cup with s-shaped profile and red-burnished surfaces2 and that of a similar larger bowl3 are typical of the Early Bronze Age across Anatolia. The form - with or without loop handle - is attested at Alişar Höyük (von der Osten 1937: 247, fig. 238) and Alaca Höyük (Orthmann 1963a: Tafel 42.11.27), Horoztepe
1 C2970 (fig. 6.3 and fig. 6.4). 2 C3055 (fig. 6.3 and fig. 6.4). 3 C3739 (fig. 6.3).
154
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
Fig. 6.2: Plan of the upper chamber of Kılıçlı Mağarası (S88). Hatched areas are spoil heaps, illicit excavations are marked in black. Produced by Michele Massa and Victor Klinkenberg.
(Orthmann 1963a: Tafel 71.20.27) and Kaklık Mevkii (Topbaş et al. 1998: pl. 51.117). S-shaped bowls and cups also form part of survey assemblages in the central Turkish Back Sea region. In inner Kastamonu, the sites of Eylekderesi, Kâhintepe and Kayabaşı have yielded such bowls (Marro et al. 1998: 323, pl. III 2-7; Kuzucuoğlu et al. 1997: 292, pl. V. 3-4), as have two sites in Çankırı (PS013, PS016: Matthews 2009: Fig. 3.33.7, 11-12). Another ubiquitous find at Kılıçlı Mağarası are rim fragments of handmade, slipped and burnished jars with vertical necks. There is considerable variation in rim shapes, which range from simple, tapering or rounded,4 to squared5 externally slightly profiled6 to more pronounced, externally thickened square rims.7 Fabrics are generally medium to medium-coarse with almost exclusive use of mineral temper. Fabric and surface colours range from buff-yellow to orange, brown and red. To this group of vessels also belongs a body piece of the neck and upper body of a jar with light brown fabric and parallel dark brown burnishing marks running diagonally
4 C2030, C2070, C2967, and C3248 (fig. 6.3). 5 C2036 and C3616 (all fig. 6.3 and 6.4) and C3608 (fig. 6.3). 6 C3605 and C3606 (all fig. 6.3 and 6.4). 7 C3598 (fig. 6.3 and 6.4).
Ceramic Assemblages from the Later Early Bronze Age in the Cide-Şenpazar Region
155
Fig. 6.3: Early Bronze Age pottery from Kılıçlı Mağarası. Produced by Lorraine McEwan.
across the vessel.8 Similar jars with vertical necks and simple rims are well attested at west Anatolian sites. They first appear in Troy I but become ubiquitous during Troy II and continue into subsequent phases (Blegen et al. 1950: types B1, 3-5, 7 and C5, 9, 10-12; Çalış-Sazcı 2006; Jablonka 2011: 719-20). They are also attested, for instance,
8 C3610 (fig. 6.3 and 6.4).
156
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
at Demircihüyük-Sarıket (Seeher 2000: Abb. 50: G509a; Abb. 58.15; 60.5), Külüoba (Efe 2007: fig. 9.c) and Beycesultan (Lloyd and Mellaart 1962: levels xiv and ix: pls. 42.1, 3/4; 54.5, 6/9). Necked jars of this type do not form part of the central Anatolian repertoire. Neither are they attested in surveys and excavations in adjacent areas along the Turkish Black Sea coast. This vessel type, therefore, appears to attests to a more direct westward connection. Such a link is further underscored by the presence at Kılıçlı Mağarası of several body fragments whose decorative patterns also find their closest parallels in the west. They include a body piece from a handmade globular jug, probably originally with cut-away spout, made of medium-fine, mineral tempered yellowish-brown clay, which is decorated with two raised nested chevrons9 and another fragment of a jug with three parallel horizontal bands below a spout (or handle base).10 Several more body sherds of medium to large jars feature raised decorations such as two or more parallel bands11 and raised circular patterns.12 Also present are fragments of a large jar with a combination of rectangular and circular applied bands on the exterior and a highly burnished interior surface.13 Chevron and zigzag patterns as well as parallel vertical and horizontal lines and ‘moustaches’, either incised or in the form of raised bands, are hallmarks of the EB II in western Anatolia. Comparable examples to the CAP finds can be found at Troy II and III (Blegen et al. 1950: figs. 389, 35.116. 35.561, 35.486; 390, 35.390), DemircihüyükSarıket (Seeher 2000: G8; G31; G53a; G57a; G58; G85, G88b; G161e; G162a; G203a; G222; G330; G294a; G427; G552), Külüoba (Şahoğlu 2011: 304. no. 262), Beycesultan (Lloyd and Mellaart 1962: pl. xxii.7) and the Yortan cemetery (Kâmil 1982: fig. 56-69). A metal jug with nested zigzag pattern comes from the Alaca Höyük cemetery (Toker and Öztürk 1992: no. 30). A bowl with an applied moustache is attested at Boğazköy (Orthmann 1963a: Tafel 56.14.04) and bowls with raised parallel lines and crescentshaped lugs are found at Alişar Höyük (von der Osten 1937: fig. 237 c1796 and c1273). Applied or broadly incised patterns are attested on necked jars from Troy (Korfmann 2001: 351, fig. 375) and Demircihüyük-Sarıket (Seeher 2000: Abb. 17, G25a). Comparisons within the central Black Sea region include several pitchers and a jar from Yassıkaya cave, which are decorated with raised bands and incised motifs (Efe 2004: figs. 2a/2b). Applied vertical bands and a moustache are attested on a bowl from an Early Bronze Age tumulus at Boztepe in Sinop and fragments with ‘raised eyebrow’ motifs come from Güllüavlu (Doonan 2004: fig. 3.11) and Harmantepe (Işın 1998: pl. 4.27). A bowl with two short vertical bands on the rim was also found at
9 C3047 (fig. 6.5 and 6.6). 10 C3620 (fig. 6.5 and 6.6) and a painted piece made from the same fabric: C3604 (fig. 6.5 and 6.6). 11 C3602, C3639 and C3641 (all fig. 6.5); C3061, C3186, C3194, C3242 and C3601 (all fig. 6.5 and 6.6). 12 C3053 (fig. 6.5 and 6.6) and C3168 (fig. 6.6). 13 C3190 (fig. 6.5 and 6.6).
Ceramic Assemblages from the Later Early Bronze Age in the Cide-Şenpazar Region
Fig. 6.4: Early Bronze Age pottery from Kılıçlı Mağarası. Produced by Lorraine McEwan.
157
158
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
Fig. 6.5: Early Bronze Age pottery from Kılıçlı Mağarası. Produced by Lorraine McEwan.
Kınık (Çınaroğlu and Genç 2004: çizim 3b) and body sherds with incised, nested chevrons, some associated with raised knobs, are attested at Eylekderesi in inner Kastamonu (Marro et al. 1998: 325, pls. III: 13-15; IV: 5). A red-burnished body piece
Ceramic Assemblages from the Later Early Bronze Age in the Cide-Şenpazar Region
159
with applied hand motif and with parallels at Demircihöyük comes from Yazıboy in Çankırı (Mathews 2009: fig. 3.20). A final indication of Cide’s Early Bronze Age connections with western Anatolia are several twisted handles from Kılıçlı Mağarası. They include a large, heavy horizontal twisted handle and attached wall fragment,14 which is made of medium, hard-fired, buff mineral tempered clay. The vessel interior is polished to a dark reddish brown. Two smaller and much more abraded twisted handles15 also come from the same cave. West-central and west Anatolian parallels for the later Early Bronze Age include twisted and faux-twisted handles on cups, jugs and jars from Beycesultan (Lloyd and Mellaart 1962, numerous examples figs. P23-P68), and the cemeteries of Küçükhöyük (Gürkan and Seeher 1991: Abb. 14.3.; 15.6.; 16.3; 17.4; 18.2-4) and Kaklık Mevkii (Topbaş et al. 1998: figs. 53.133; 55.142), the nearby settlement of Karaoğlan (Topbaş et al. 1998: figs. 37.3; 38.20; 39.30; 45.68) as well as the Yortan cemetery (Kâmil 1982: figs. 82.b; 37.126; 45.162). More sporadic evidence comes from, for instance, Polatlı on the westcentral plateau (Orthmann 1963a: Tafel 35.847; 37.865). Sites in inner Kastamonu and Çankırı did not yield any twisted handles. Other, less distinctive handles from Kılıçlı include a small, rounded horizontal handle16 and a vertical handle with elongated profile, which may have belonged to a cup or pouring vessel.17 One of the most distinctive pottery finds from Kılıçlı Mağarası is the rim-fragment of a medium-coarse, handmade bowl or small jar with a light brown, smoothed surface and a band of finger-impressed ridges just below the vessel opening.18 An even coarser handmade jar with finger-impressed indentations was also found at the same site.19 The fabric is heavily mineral tempered with large, angular crushed quartz protruding from both smoothed interior and exterior surfaces. Parallels, some specific, others more generic, come from several sites in the central Turkish Black Sea region, but finger-impressed bands of this type are also attested in the western and northern Black Sea and may thus point towards connections beyond Anatolia. Along the Turkish Black Sea coast, the Kılıçlı vessels can be compared to a redslipped and burnished bowl with buff fabric from Yassıkaya cave (Efe 2004: fig. 1.7) and a similar type of impressed decoration is also found at İkiztepe (Alkım et al. 1988: xi, 13, 15- 17; xii; xiii, 1; xiv; xvi; Thissen 1993: 212; for mound III, Level II – EB III see Bilgi 1999a: pl. 3: 9; Bilgi 1999b: 173, pl. 2: 9). Further inland, comparable pieces come
14 C3177 (fig. 6.5 and 6.6). 15 C3603 (fig. 6.5) and C3618 (fig. 6.5 and 6.6). 16 C2931 (fig. 6.6). 17 C3597 (fig. 6.5. and 6.6). 18 C3183 (fig. 6.7 and 6.8). 19 C3045 (fig. 6.7 and 6.8).
160
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
Fig. 6.6: Early Bronze Age pottery from Kılıçlı Mağarası. Produced by Lorraine McEwan.
from Eylekderesi and Gâvurevleri in inner Kastamonu (Marro et al. 1998: 323, pl. ii 1-2, 4) and the site of Yazıboy in northwest Çankırı (Matthews 2009: figs. 3.18; 3.19). They are absent from Early Bronze Age levels at Troy and central Anatolian assemblages (Thissen 1993: 212). Such vessels and their distinctive decoration, however, form part of a long-lasting ceramic tradition in eastern Thrace and the western Black Sea coast.
Ceramic Assemblages from the Later Early Bronze Age in the Cide-Şenpazar Region
161
Fig. 6.7: Early Bronze Age pottery from Kılıçlı Mağarası. Produced by Lorraine McEwan.
Examples from Turkish Thrace come from Hasköy Höyüğü (Başaran 1999: lev. 3: 6-7) and Salhane and last from the later Early Bronze Age to the Iron Age (Özdoğan 1982: 45, lev. XVII: 4). Good parallels for the rim piece from Kılıçlı also come from Sitagroi Vb (ca. 2700-2200 BC) (Sherratt 1986: 440; figs. 13.16.4; 13.22.5; 13.23). Earlier examples come from nearby Dikili Tash (Séfériadès 1995) as well as Cernavoda III sites along the Lower Danube, Foltešti and Late Ezero in central Bulgaria (Georgiev
162
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
Fig. 6.8: Early Bronze Age pottery from Kılıçlı Mağarası. Produced by Lorraine McEwan.
Ceramic Assemblages from the Later Early Bronze Age in the Cide-Şenpazar Region
163
and Merpert 1966: fig. 1.23 and 24; Merkyte 2007). Later examples of this tradition can be found at the Late Bronze Age site of Vratitsa in the coastal hinterland of the Bulgarian Burgas region (Hristova 2011: figs. 8, 9, 10.1-2) and at Troy VIIa such pottery is referred to as ‘barbarian ware’ (Blegen et al. 1958; Koppenhöfer 1997: 334-7, Abb. 18, 24). Coarse, brown and black coloured, chaff-tempered and burnished jar rims with impressed cordons are also reported from a rescue excavation at the foot of Sinop Kale, but dated to the Iron Age on the basis of proposed parallels with pre-colonial sites on the north and western Black Sea such as Berezan and Histira (Doonan 2004: 57, fig. 3-5.7 and 8). Although this means that both an earlier and later date than the Early Bronze Age is possible for the Kılıçlı examples, the chronological focus of most of the pottery from the cave in the later third millennium BC would point towards an Early Bronze Age date for the finger-impressed pieces. In any case, the coarse finger-impressed jars would suggest a participation in a western and central Black Sea interaction zone, possibly along the coast (also Matthews 2009: 87). With regards to vessel function, Sherratt (1986: 439) suggested that the: “Continuous cordons … may have a functional significance in securing flexible organic (e.g., skin) covers with a string tied around the neck of the vessel, or they may be simply skeuomorphic reflections of such strings or ropes.”
At Late Bronze Age Vartitsa, the majority of pottery decorated in this manner is classified as cooking equipment (Hristova 2011). The vessels from Kılıçlı show no signs of secondary burning indicative of a use as cooking equipment, which makes a function in the short-term storage or consumption of food more likely. Kılıçlı Mağarası also yielded several fragments of handmade vessels with holes for spouts20 for which a date in the later Early Bronze Age seems possible, but by no means certain. One of the vessels shows the dense burnishing typical of the period. From the surviving pieces it is not possible to reconstruct vessel shapes, but they appear to belong to rather large vessels with spouts or openings in the middle or lower body. Similarly difficult to interpret are two painted body sherds, which may also belong to this period, but for which no ready comparisons could be found. The first belongs to a relatively thick-walled, handmade vessel made of medium, mineral tempered clay. A decorative ridge runs across the upper end and a series of parallel yellowish-orange painted bands outlined in brown start just below the ridge.21 A second body piece is made of the same mineral tempered, reddish fabric as one of the jug fragments with
20 C2945, C3054, C3106 and C3245 (fig. 6.7 and 6.8). 21 C3600 (fig. 6.7 and 6.8).
164
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
raised bands22 already discussed. A one centimetre thick band outlined in dark brown and filled with a bright orange paint runs across the top of the sherd.
Fig. 6.9: Early Bronze Age pottery from Okçular Kale İni and Derebağ Köy Mağarası. Produced by Lorraine McEwan.
Several rim and body pieces from Kılıçlı point towards a date in the late Early Bronze Age or the early second millennium BC. A delicate red-slipped and polished cup with a thin, rounded horizontal handle23 finds parallels at Boğazköy (Orthmann 1963b). The rim-fragment of a jar with tapering upper body and upward pointing rim, powdery yellowish clay and brown slip24 may also belong to the final part of the Early Bronze Age as does a body sherd of a large jar or bowl with highly burnished surfaces and horizontal handle base.25 A large body sherd of a jar or pithos with dark purple
22 C3620. 23 C3642 (+C3643) (fig. 6.7 and 6.8). 24 C3087 (fig. 6.7 and 6.8). 25 C3050 (fig. 6.7).
Ceramic Assemblages from the Later Early Bronze Age in the Cide-Şenpazar Region
165
burnished slip,26 a red slipped and polished piece27 and a further fragment with a pinkish-buff smoothed outer surface and a well-polished red surface on the interior28 also seem to point to this transitional phase. As do a dark red slipped and polished piece29 and red slipped and smoothed sherds from large jars.30
6.2.2 Okçular Kale İni Evidence for the third millennium BC at Okçular Kale İni (S24) is much less extensive than at Kılıçlı Mağarası. The Okçular assemblage consists of a handful of diagnostic pieces and body sherds. Fabric, firing and surface treatment of these vessels can be assigned to the third millennium BC, with some examples sitting more comfortably in its second half. These include the rim fragment of a bowl with simple rounded rim,31 a large horizontal handle belonging to a jar or pithos32 (for comparisons, see Matthews 2009: fig. 3.37.9-10), a smaller rounded handle,33 and two flat bases of medium to large handmade jars.34 Another flat base of a small jar with a highly burnished exterior surface may perhaps also belong to this period, but its fine fabric and hard firing would also fit with a later date.35 Also found at Okçular was a sherd, which appears to have circular raised decorative features reminiscent of those at Kılıçlı36 and several body sherds whose slipped and/or burnished surfaces point towards an Early Bronze Age presence at the site.37
6.2.3 Derebağ Köy Mağarası The handmade pottery from Derebağ Köy Mağarası is difficult to interpret, but it would appear that many date to the Early Iron Age rather than the Early Bronze Age or earlier (chapter 8). A body fragment of a handmade bowl or globular jar with a partial, highly
26 C2966 (fig. 6.8). 27 C2035. 28 C2933 (fig. 6.8). 29 C3114 (fig. 6.8). 30 C3049 and C3115. 31 C2908 (fig. 6.9 and 6.10). 32 C2903 (fig. 6.9 and 6.10). 33 C2902 (fig. 6.9 and 6.10). 34 C2906 and C2996 (fig. 6.9 and 6.10). 35 C0787 (fig. 6.9 and 6.10). 36 C3068 (fig. 6.9 and 6.10). 37 C2901, C2916 (fig. 6.10) and C3011.
166
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
Fig. 6.10: Early Bronze Age pottery from Okçular Kale İni and Derebağ Köy Mağarası. Produced by Lorraine McEwan.
polished red slip on the exterior and dark brown polish on the interior,38 however, may belong to the late Early Bronze Age or the earlier part of the second millennium BC.
6.2.4 Summary To sum up, a rather substantial number of pottery fragments collected during CAP can be dated to the later part of the Early Bronze Age. The site of Kılıçlı Mağarası in particular appears to have been a focal point of activity, while less substantial evidence for a later Early Bronze Age presence also comes from Okçular Kale İni and Derebağ
38 C4193 (fig. 6.8 and 6.9).
Early Bronze Age Communities in the Central Turkish Black Sea Region
167
Köy Mağarası. The ceramic evidence from the three caves, simultaneously displays far-flung cultural connections against a background of localised technological traditions. Formally, the Kılıçlı material in particular displays affinities with the EB II-III ceramic traditions of west-central and western Anatolia, Thrace and the Balkans. More generic Early Bronze Age forms also find parallels in adjacent regions of the central Turkish Black Sea and inner Anatolia. A preliminary petrographic and chemical analysis points to a varied geological and chemical composition (section 16.3). At the moment, it is not possible to qualify this observation further and pinpoint more precisely the provenience of the samples. The results of this analysis, however, add a further dimension to, and underscore the significance of, this locale and the activities taking place within it, which may have been of regional or supra-regional importance (see below). The almost exclusive use of mineral temper connects the later Early Bronze Age pottery from Cide-Şenpazar with the local Chalcolithic material and earlier traditions further to the west along the Black Sea coast. A similar preference for mineral temper can be found in the Yassıkaya assemblage (Efe 2004: 30) and the Zonguldak survey material (Karauğuz and Düring 2009). Early Bronze Age traditions in neighbouring regions to the east and south, by contrast, display rather different technological characteristics. Prehistoric pottery in Sinop is described as primarily shell-tempered (Bauer 2006: 241); and a mix of mineral and vegetal temper characterises the pottery from sites further inland in Kastamonu (Marro et al. 1996; 1998; Kuzucuoğlu et al. 1997) and Çankırı (Matthews 2009). On the basis of this data, and evidence from surrounding areas, I will now make an attempt to sketch a picture of the later Early Bronze Age in Cide-Şenpazar and neighbouring regions.
6.3 Early Bronze Age Communities in the Central Turkish Black Sea Region The evidence for later Early Bronze Age communities in the central Turkish Black Sea region comes from four types of sites with distinctive regional distributions: mounded and flat settlement sites, cemeteries and caves. Mound settlements as well as cemeteries concentrate in the open hilly and plain landscapes of the eastern part of the central Black Sea coast, such as Dündartepe, Tekkeköy and İkiztepe (see e.g. Dönmez 2002). Further to the west, in inland regions such as Çankırı and Karabük (Burney 1956; Matthews 2009) as well as in the plains of Eskişehir (Efe 1997) mounded sites also predominate. As we move north into the more mountainous and forested regions of Kastamonu, true höyük sites become rare, while many flat sites are found on top of, or associated with, natural hills. The majority of Chalcolithic and Early Bronze Age sites in inner Kastamonu are located on terraces above valley floors (Marro et al.
168
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
1996: 273, 279). The excavated site of Kınık in the Devrekani district of Kastamonu, which yielded evidence of metal and pottery production in the Early Bronze Age, is located on a natural hill (Çınaroğlu and Genç 2004; 2005). In Sinop too Early Bronze Age sites show a preference for natural hills, which distinctively resemble settlement mounds, but are not normally the result of long-term human occupation; there are also some notable exceptions such as the mound of Koçagöz (personal communication Alexander Bauer, 13-12-2012). A good example for the former type of site is the recently excavated Boyabat-Kovuklukaya, which is located on top of a rock-outcrop (Dönmez 2004: 36-37). A similar picture has begun to emerge from Zonguldak (Karauğuz and Düring 2009: 254-255). Known Early Bronze Age cemeteries track the spatial distribution of mounded settlement sites, and are associated with the emergence of increasingly large-scale and socially stratified, sedentary communities with claims to the land that surrounds their settlements. EB II cemeteries are attested inland at, for instance, Balıbağı (Süel 1989) and Salur North (Matthews 2004; 2009: 88-90) in Çankırı, and at Kaklık Mevkii (Topbaş et al. 1998), Demircihöyük-Sarıket (Seeher 2000) and Küçükhöyük (Gürkan and Seeher 1991) in west-central Turkey. A later Early Bronze Age burial mound was excavated at Boztepe on the Sinop promontory (Işın 1989), and the illegally dug cemetery at Hıdırlı also in Sinop dates to the late third and early second millennium BC. Further to the south, the ostentatious burials of Mahmatlar (Koşay and Akok 1950), Horoztepe (Özgüç and Akok 1958), Kalınkaya (Zimmerman 2006) and Alaca Höyük (Koşay and Akok 1944; Kosay 1951) date to the late EB II or early EB III (GürsanSalzman 1992; but see Bachhuber 2011). Neither mounded sites, indicative of long-term settlement continuity and the building techniques otherwise popular in contemporary western and central Anatolia, nor cemetery sites dating to the later Early Bronze Age have been found to date in the mountainous and forested coastal areas. All the available evidence for this period comes from cave sites. There is little doubt that third millennium BC settlements and other activity sites existed in the Cide, Şenpazar and adjacent landscapes and that their absence in our survey record is due at least in part to a combination of poor surface visibility and taphonomic processes. Lithic surface finds, for instance, indicate an earlier presence throughout the region (chapter 5), while geoarchaeological work and chance finds, thrown up by recent construction work, demonstrate that most Early to Mid-Holocene surfaces have either eroded away or lie buried underneath erosion material (chapters 2 and 4). Anecdotal evidence would also suggest that surface abrasion of pottery constitutes an obstacle in the identification of finds dating to these periods. Once the smoothed or burnished surfaces of coarse, handmade and low-fired vessels are fully or partially abraded, it becomes very difficult to assign a date to them. The lack of mounded sites can be explained through the use of abundantly available stone and timber as construction materials rather than mud-brick. Evidence from excavated sites in the wider region points to a range of construction methods that include plastered timber at Yassıkaya (Efe 2004: 28), and stone slabs
Early Bronze Age Communities in the Central Turkish Black Sea Region
169
at Boyabat-Kovuklukaya (Dönmez 2004: 37; figs. 4-8). Early Bronze Age communities in this mountainous region also can be expected to have been of a much smaller scale than those developing in the plains and open valleys to the south. The mountainous setting, which for the most part lacks larger areas of agriculturally viable land, makes a mixed agricultural and pastoral economy with transhumant elements and a continued reliance on gathered and hunted forest resources highly likely. Recent isotopic analysis of Chalcolithic skeletons at İkiztepe, for instance, indicates a transhumant element among its population (Welton 2010). LBA textual sources also imply a high(er) degree of group mobility in the region (chapter 7) and transhumant practices are attested in the Turkish Black Sea region until recent times (Yakar 2000). Small-scale social organisation and relative mobility of at least part of the population can be expected to result in small settlements and potentially low settlement continuity, which translates into ephemeral archaeological assemblages that are easily obscured by natural processes and difficult to detect through surface survey. What then was the function of these cave sites and how do they fit into the socioeconomic scenario proposed above? The morphology of the three Cide caves with evidence for a third millennium BC presence as well as the artefacts recovered from them would suggest that none of them were used for long-term habitation. The lower chamber of Kılıçlı Mağarası (S93) can be accessed with relative ease by both humans and their livestock, but space is limited (Fig. 6.11). Accessible only via a ladder, an upper gallery continues deep into the interior and appears to be subject to regular flooding. The pottery found in S93 among otherwise clean sandy deposits appears to have been washed down from a higher-lying chamber (S88), possibly through a sinkhole, which is connected to the lower part of the cave through at least one small opening in the rock and which yielded the majority of finds. The upper chamber can be reached only through a short but steep climb further up the mountain, followed by a ca. six-metre vertical drop that leads to the current cave floor (Fig. 6.12). The upper chamber is light and well protected from wind and weather, but a series of large boulders on the cave floor attest to the distinct possibility of roof collapse. The difficulty with which the upper chamber is accessed makes it an unlikely habitation site. While livestock could be sheltered in the lower part of the cave, there is - at least today - no easy access to and from the upper chamber, which is where most activities appear to have taken place. Okçular Kale İni is similarly cumbersome to access, requiring visitors to crawl through a narrow entrance passage and subsequently climb down a vertical rock face. Late Roman and Mid- to Late Byzantine pottery attest to later activities inside the cave (chapters 10 and 11). The spatial clustering of Chalcolithic and Bronze Age pottery at the very back of the cave and the midden-like deposits they were found in, however, would suggest that they were thrown into the cave from the top of Okçular Kale through a now-blocked passage (also chapter 5). At Okçular too, large boulders on the cave floor, some of which have crushed Late Roman/Byzantine pots, attest to
170
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
Fig. 6.11: Kılıçlı Mağarası, lower chamber (S93) (CAP-P2/1001).
Fig. 6.12: Downward climb into Kılıçlı Mağarası, upper chamber (S88) (CAP-P2/1010).
Early Bronze Age Communities in the Central Turkish Black Sea Region
171
frequent ceiling collapse. This is also the case for Derebağ Köy Mağarası, a long but narrow cave that requires a long, steep climb to access it. The samples from Okçular Kale İni and Derebağ Köy Mağarası are too small to draw any conclusions about the nature of their use during the later Early Bronze Age, but they do suggest the presence of small-scale communities throughout the region. The occupation of the top of Okçular Kale appears to continue from the Chalcolithic, or at least echo an earlier settlement practice (chapter 5). A functional characterisation of the Kılıçlı assemblage alongside the range of cultural connections it attests to, however, can provide further clues about this particular cave’s function. The ceramic types found at Kılıçlı include small drinking cups, hole-mouth bowls, pitchers and possibly also other pouring vessels, and small to medium necked jars. The existence of larger jars is indicated by several large body pieces and a large, heavy handle. The majority of these vessels are slipped and burnished and many have decorative elements. There is no unequivocal cooking equipment among the finds. I would, therefore, propose that the primary activities suggested by the pottery assemblage are short-term storage, serving and consumption, especially of liquids. A possible interpretation, thus, would seem to be that the cave served as a symbolically charged space, due to its dramatic natural setting (e.g. Bradley 2000) and the potentially long-term cultural memory of its significance from at least the Chalcolithic period (section 5.4.3), in which socially significant consumption activities took place. Fragmentary bone material, some of which may be human and may or may not relate to the Early Bronze Age, was recovered from the cave. It is, therefore, possible that funerary rituals may have formed at least part of such consumption events. The formal characteristics of the cave pottery trace a number of regional and supra-regional cultural networks, with a particular focus on western Anatolia and the western Black Sea coast. Consumption practices in the cave and their equipment may have been inspired by traditions of commensality that were developing in western Anatolia during this period. Communal drinking has been argued to be part and parcel of the emergence of social complexity during the later Early Bronze Age (e.g. Schoop 2011 for a review), but access to alcoholic beverages may not have been socially restricted and drinking equipment appears to have formed an integral part of a significant proportion of graves in Early Bronze Age cemeteries (Whalen 2014). The evidence from Kılıçlı, therefore, need not imply the presence of an emerging local elite, but it does attest to the attempts of this remote region to engage with distant cultural practices, which are associated with trajectories towards greater social divisions and complexity. The only good comparison for Kılıçlı Mağarası in Anatolia is the cave of Yassıkaya, located in nearby Ereğli/Zonguldak (Efe and Mercan 2001; Efe 2004). The cave is located ca. 20km inland from the bay of Ereğli and can be reached
172
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
from the coast via a broad open valley, a setting not unlike that of Kılıçlı Mağarası, although access from the coast is somewhat more difficult in our case. Small-scale excavations revealed the remains of what Efe (2004) has called the ‘Filyos culture’ complex. This includes wood architecture covered by mud and a collection of cups, bowls and pitchers, whose decorative attributes and formal characteristics resonate with later third millennium BC west Anatolian and western Black Sea and Balkan traditions. The cave site itself, which lacks both hearths and, with the exception of ground stones, definitive cooking equipment, is interpreted as a camp-site (Efe 2004: 33). With regards to the functional types represented, the assemblage at Yassıkaya closely resembles that of Kılıçlı Mağarası (albeit not in formal and decorative detail), making a similar use likely. Further cave comparisons are difficult to come by. A second cave site in Zonguldak, İnönü cave, produced only one possible later Early Bronze Age sherd (Karauğuz and Düring 2009: 155, fig. 10.7). Similarly, the majority of cave sites in south and west Anatolia have either yielded no or only limited evidence for use in the later Early Bronze Age or lack sufficiently detailed publications (e.g. Mellaart 1961; Yalçinkaya 1987: 24; Meriç 1992: 238; Peschlow-Bindokat and Gerber 2003; Peschlow-Bindokat 2006; 2007; Aydin 2006; French 2008). An ever-growing appetite for metals, bronze in particular, during the second half of the third millennium BC, was likely an important catalyst for the connections that we see manifested in the material culture of Kılıçlı Mağarası and other sites in the wider region. The central Turkish Black Sea region is rich in metal sources, especially copper, and the geographically closest copper outcrops to the CAP survey area are the Küre deposits of north-central Kastamonu, located some 50km as the crow flies from Cide. A Prehistoric exploitation of the Küre deposits may be suggested by the presence of narrow, irregular mining shafts (Wagner and Öztunalı 2000). Lead isotope analyses from stratified artefacts from Kaman-Kalehöyük, Troy, Mersin and further afield appear to confirm the export of Küre copper to distant parts of Turkey, the Aegean and to Crete during the Bronze Age (Stos-Gale 1993: 124-5; Sayre et al. 2001). Overall, however, archaeometric approaches to Bronze Age metal artefacts, which would allow us to trace metal sources more accurately, are only beginning to gain momentum in Anatolian archaeology (Muhly 2011; Zimmerman 2012). Archaeological evidence for the involvement of Early Bronze Age communities in metal processing comes from the site of Kınık in the Devrekani district of Kastamonu, located some 25km southeast of the Küre mines. In addition to an eclectic collection of silver and bronze vessels, some of which can be dated with certainty to the Late Bronze Age (Emre and Çınaroğlu 1993; see chapter 7), excavations at the site revealed evidence for Early Bronze Age metallurgical activities including kilns, crucibles, slag and tools (Çınaroğlu and Genç 2004; 2005). Boyabat-Kovuklukaya in the Sinop hinterland also points to a rather substantial metallurgical enterprise during the EB III-early Middle Bronze Age. In addition to smelting ovens, finds include stone and
Conclusions
173
clay moulds for casting axes and arrowheads and a range of processing tools (Dönmez 2004: 44-5; pls. 6-9). With the exception of a possible iron-smelting site at Okçular, probably dating to the Roman or Byzantine period (chapter 14), we recorded no mining sites or metallurgical activities in Cide. Thus, while perhaps not involved in metal extraction or processing, the more mobile elements of Early Bronze Age communities in the Cide region may well have been instrumental in channelling metals and other goods to and from the coast. There is as yet no conclusive evidence for large-scale seafaring in the Black Sea before the seventh century BC. This stands in stark contrast to the Mediterranean, where it gathers momentum during what Broodbank calls the ‘long third millennium’ and by the end of which long-distance maritime routes become well established (Broodbank 2000: 44, 283; 2010: 250). Even if the Black Sea is a much more unpredictable maritime environment than the Mediterranean (chapter 2) and there is no reliable archaeological evidence (Ivanova [2012: 343, note 6] suggests a third millennium BC date for a longboat from Lake Varna and boat models from Sozopol and Urdoviza), it is hard to imagine the Bronze Age Black Sea entirely devoid of small crafts and fishing boats. While impossible to prove at this point, it seems not unreasonable to propose short-distance coastal transfers of goods and people from Cide to destinations with easier access to inland routes such as Amasra or Sinop. Further inland, the topography of the Pontic mountains no doubt played some part in channelling movement of people, ideas and objects in east-west rather than north-south direction (Matthews 2009: 87). At the same time, however, it is worth keeping in mind that what may seem a cumbersome journey today over winding mountain roads would have appeared less so to shepherds or traders with donkey caravans intimately familiar with the terrain or equipped with local guides.
6.4 Conclusions Evidence for the later Early Bronze Age from the CAP survey area concentrates on three cave sites, and Kılıçlı Mağarası in the coastal hinterland in particular. The collected material attests to both geographically extensive cultural connections and a locally rooted technological tradition. The link with EB II-III western Anatolian sites, as well as Thrace and the Balkans is particularly prominent in the Kılıçlı assemblage. This outlook to the west, however, is not exclusive as more general forms find widespread parallels in central Anatolia and the more southerly and eastern parts of the Turkish central Black Sea region. In this way, communities in the Cide region appear rather well connected to a wider world in the second half of the third millennium BC, in which significant social processes were taking place and far-reaching economic relationships developed. Both western and central
174
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
Anatolia were moving towards steeper social hierarchies and in both regions these developments were fuelled by newly developed practices of commensality. The nature of the Kılıçlı assemblage, with its predominance of cups, bowls, small jars and pouring vessels, together with its evocative natural setting, suggest a special purpose site, used perhaps for socially charged consumption events – whether in the context of funerals or not - whose material components were modelled on those used by the emerging elites of western and central Anatolia. The chemical and geological diversity of the assemblage (section 16.3) also points to the existence of imports, brought to the cave either for their intrinsic value or their contents as traded goods, or brought along by participants in the activities taking place in the cave. Metals, especially copper, may have provided the catalyst for the involvement of local communities in the wider exchange networks of the time. The mechanisms and means through which the inhabitants of Cide participated in these wider spheres of interaction, however, are less easily defined. It seems clear though, that mobility in rugged and densely forested mountains and along the coast depended on knowledge of highland pathways and coastal weather and currents on the one hand and social connections to adjacent valleys and coastal communities on the other. Whatever the impetus for these connections and the modes of communication that sustained them, Cide, unlike what has been proposed for neighbouring Sinop (Bauer 2006: 237) was not an island in the Early Bronze Age.
Catalogue of Illustrated Artefacts C0787. S24. Fragment of flat base. Red, very fine fabric with no visible inclusions. Exterior surface burnished, interior wiped. Oxidised firing. Hand/wheel(?)made. Exterior 5 YR 6/6, interior 10 R 5/6. D= unknown. C2030. S93. Rim fragment of a medium-sized jar with simple rim. Brown fine fabric with almost no inclusions, some small white and red grit. Exterior burnished, parallel grooves on neck. Handmade. Exterior surface 5 YR 5/4 – 6, interior surface 5 YR 6/4. D= 10cm. C2035. S93. Body fragment of a jar. Red to brown medium coarse fabric with small to large rounded and angular quartz. Exterior surface burnished, interior smoothed. Differential firing. Handmade. Surfaces 2.5 YR 5/8. D= unknown. C2036. S93. Body sherd. Reddish orange medium coarse vessel with small to medium black and red grit, some large angular quartz (?), possibly some grog. Parallel grooves on exterior. Differential firing. Handmade. Exterior surface 2.5 YR 4/4, interior surface 5 YR 4/4. D= unknown. C2070. S93. Rim fragment of a jar with simple rim. Reddish medium coarse soft fabric with very small red grit and calcareous inclusions. Surfaces smoothed. Differential firing. Handmade. Surfaces 5 YR 5/6, core 5 YR 4/1. D= unknown. C2902. S24. Small handle fragment. Dark purplish brown medium coarse fabric with dense small to medium grit and pebbles, some mica. Surface burnished. Oxidised firing. Handmade. Fabric 2.5 YR 4/4, surface 2.5 YR 6/8.
Catalogue of Illustrated Artefacts
175
C2903. S24. Large horizontal handle of a large jar. Pale brown coarse fabric with small dense white grit and large shell (?) and red grit. Exterior surface smoothed, some burnishing, interior smoothed. Oxidised firing. Handmade. Fabric 10 R 6/8, surfaces 5 YR 7/6. C2906. S24. Fragment of a flat base of a medium to large jar. Dark orange to brown coarse brittle fabric with medium to large grit and pebbles. Interior and exterior surfaces smoothed. Differential firing. Handmade. Fabric 10 R 6/8, interior surface 10 R 6/6, exterior surface 5 YR 7/3. D= unknown. C2908. S24. Rim fragment of a large, hemispherical bowl with simple rim. Purplish-grey, very coarse, gritty and brittle fabric. Very dense large angular quartz. Both surfaces smoothed/ wiped. Reduced firing. Handmade. Fabric 10 R 4/3, surfaces 10 R 6/6 to 10 R 4/2. D= 17cm. C2916. S24. Body fragment. Buff-brown medium fabric, dense small round white and black grit and some quartz, sparse gold mica. Exterior red slipped and polished, interior burnished. Oxidised firing. Handmade. Fabric 5 YR 5/6, exterior 10 R 4/8, interior 7.5 YR 7/4. D= unknown. C2933. S88. Small body fragment. Dark red medium fine fabric with small black and white grit. Interior slipped (?) and burnished. Oxidised firing. Handmade. Fabric and interior surface 10 R 4/8. D= unknown. C2945. S93. Fragment of a body sherd and spout. Pale red to dark brown fine fabric with occasional small red and black grit. Exterior smoothed, interior slipped (?) and traces of burnish. Reduced firing. Handmade. Exterior surface 2.5 YR 5/6, interior surface 2.5 YR 4/4. D= unknown. C2967. S93. Rim fragment of a jar with simple, thickened rim. Medium coarse, dark red fabric, small to medium grit and quartz. Burnished exterior. Reduced firing. Handmade. Exterior surface 10 R 5/8, interior surface 10 R 4/6. D = 12cm C2970. S93. Rim fragment of a bowl with simple rim. Medium coarse reddish brown gritty fabric, medium sub-angular quartz (?) and black grit. Some burnishing on exterior. Oxidised firing. Hand/wheelmade (?). Fabric 10 R 5/8, surfaces 3.5 YR 6/8. D= 12 cm. C2996. S24. Fragment of flat base of a medium to large jar. Light brown medium-coarse fabric with very dense small angular and sub-angular grit. Surfaces smoothed. Reduced firing. Handmade. Fabric 5 YR 6/8 to greenish black, exterior surface 5 YR 5/3, interior surface 2.5 YR 5/2. D= 10cm. C3011. S24. Body fragment. Brown medium fabric, small sparse sub-angular quartz and grit. Surfaces washed, some burnishing. Oxidised firing. Handmade(?). Fabric 5 YR 5/6, surfaces 10 R 5/8. D= unknown. C3045. S88. Rim fragment of a bowl or jar with slightly everted rim and finger impressed decoration below rim. Reddish brown very coarse, brittle fabric with, small to very large angular quartz, large pieces visible on surface. Exterior and interior surface slipped, exterior burnished. Oxidised firing. Handmade. Fabric and exterior surface 5 YR 6/6, interior surface 2.5 YR 3/6. D= unknown. C3047. S88. Large body fragment of a round-bellied jug or jar decorated with two raised nested chevrons on the exterior. Yellowish brown to grey medium fine fabric with small black, white and red grit (grog?) inclusions. Traces of polished slip. Oxidised firing. Handmade. Fabric 2.5 YR 6/6, surfaces 2.5 YR 4/4. D= unknown. C3049. S88. Small body fragment of a bowl. Buff brown, fine fabric with sparse small black and white mineral inclusions. Both surfaces are red slipped and burnished. Oxidised firing. Handmade. Fabric 2.6 YR 6/6, surfaces 10 R 5/8. D= unknown. C3050. S88. Body fragment of a large jar with handle base (?). Red very coarse and brittle fabric with dense large grit, calcareous inclusions and some vegetal temper. Exterior surface burnished, interior smoothed. Oxidised firing. Handmade. Fabric and interior surface 2.5 YR 6/8, exterior surface 2.5 YR 8/6. D= unknown.
176
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
C3053. S88. Body fragment of a jar with parallel, raised circular bands on the exterior. Reddish orange medium fabric with dense small white and black grits. Soft fabric. Exterior smoothed, interior some burnishing. Traces of red slip or paint on exterior. Oxidised firing. Handmade. Fabric 2.5 YR 6/8, surfaces 2.5 YR 6/6. D= unknown. C3054. S88. Body fragment with spout. Orange red medium fine fabric, occasional small black grit. Interior surface burnished, exterior abraded. Reduced, hard fired. Handmade. Core 5 YR 4/1, exterior surface 2.5 YR 6/6, interior surface 5 YR 6/8. D= unknown. C3055. S88. Rim fragment of a small s-shaped bowl or cup. Dark reddish brown to orange brown coarse, very brittle fabric. Soft, dense small to medium quartz and/or shell (?) and grit. Burnished interior and exterior. Differential firing. Handmade. Surfaces 6 R 5/6, core 2.5 YR 5/6. D= unknown. C3061. S88. Body fragment of a medium-sized jar with parallel raised bands on the exterior. Medium-fine pale to orange brown fabric with dense small white and dark mineral inclusions. Surfaces abraded. Handmade. Differential firing. Fabric and surfaces 5 YR 6/6. D = unknown. C3068. S24. Body fragment with raised circular features. Dark grey/black medium mediumcoarse fabric with sparse small rounded calcareous inclusions and some grit. Exterior burnished, interior smoothed (and burnished?). Reduced firing. Handmade. Fabric 5 YR 2.5/1. D= unknown. C3087. S88. Rim fragment of a jar with everted rim. Pale pinkish yellow medium fine fabric with sparse, very small calcareous inclusions. Surfaces smoothed, possibly burnished, interior abraded. Oxidised-firing. Handmade. Fabric 5 YR 8/4, surfaces 2.5 YR 5/6. D= 30cm. C3106. S88. Fragment of a body sherd and spout. Light red medium fabric with very small calcareous particles. Surfaces smoothed, signs of burnishing. Oxidised firing. Handmade. Fabric 2.5 YR 6/6, exterior surface 5 YR 7/6, interior surface 10 R 5/8. D= unknown. C3114. S88. Body fragment. Red fine fabric with sparse and very small calcareous inclusions. Exterior and interior burnished. Oxidised firing. Handmade. Fabric 2.5 YR 6/8, surfaces 10 R 4/6. D= unknown. C3115. S88. Body fragment. Red medium-fine fabric with sparse and very small calcareous inclusions. Exterior smoothed. Differential firing. Handmade? Fabric 2.5 YR 5/6, exterior surface 5 YR 5/3. C3168. S88. Body fragment of a jar with traces of parallel raised circular bands on the exterior. Red medium fabric with small mineral inclusions. Interior surface burnished, exterior abraded. Oxidised firing. Handmade. Fabric 6 R 5/6. D= unknown. C3177. S88. Twisted horizontal handle of medium to large jar. Pale red medium fabric with sparse white and black grit, some larger pebbles. Interior surface burnished, handle surface smoothed. Reduced, hard fired. Handmade. Fabric and surfaces 2.5 YR 5/4. C3183. S88. Rim fragment of a bowl or jar with finger impressed decoration below rim. Reddish brown/orange coarse gritty fabric with medium to large dense angular and sub-angular quartz. Wash/wipe marks on exterior. Reduced firing. Handmade. Core 2.5 YR 3/1, exterior surface 2.5 YR 7/6. D= unknown. C3186. S88. Body sherd with raised band on exterior. Pale orange brown to brown medium fabric with fine mineral temper. Interior surface burnished. Oxidised firing. Handmade. Fabric 5 YR 6/6, exterior surface 5 YR 8/4, interior surface 5 YR 4/4. D= unknown. C3190. S88. Large body fragment of a jar with raised geometric decoration on the exterior. Pale to brownish red medium fine fabric with sparse small to large white grit and pebbles. Exterior smoothed, interior burnished. Oxidised, hard fired. Handmade. Fabric 2.5 YR 6/6, exterior surface 2.5 YR 7/6, interior surface 2.5 YR 6/6. D= unknown. C3194. S88. Body fragment of a medium-sized jar with three parallel raised bands on the exterior. Medium red to reddish yellow fabric with small dense white and black mineral
Catalogue of Illustrated Artefacts
177
temper. Interior surface slipped and polished. Reduced firing. Handmade. Fabric 2 YR 5/1, exterior surface 5 YR 7/6, interior surface 2.5 YR 4/6. D= unknown. C3242. S88. Body fragment of a medium-size jar with raised band on the exterior. Pale brown to reddish yellow fabric with medium density of well-sorted small calcareous specks and some vegetal temper. Surfaces smoothed. Differential firing. Handmade. Fabric 10 YR 7/3, surfaces 5 YR 6/6. D= unknown. C3245. S88. Fragment of a body sherd and spout (?). Fine reddish brown fabric with very sparse small mineral inclusions. Interior burnished, exterior abraded. Reduced firing. Handmade. Fabric 5 YR 6/1, surfaces 5 YR 5/4. D= unknown. C3248. S88. Rim fragment of a jar with simple rim. Dark grey coarse gritty fabric with abundant large quartz (?) inclusions. Interior smoothed surface light brown, rim reddish yellow, exterior surface abraded. Differential firing. Handmade. Fabric 7.5 YR 3/1, exterior surface 2.5 YR 5/6, interior surface 7.5 YR 6/4. D= unknown. C3597. S88. Small flat handle attached to vessel rim, probably a jug. Fine orange brown fabric with sparse, very small rounded mineral inclusions. Surfaces burnished. Differential firing. Handmade. Fabric and surfaces 2.5 YR 5/6-4/6. C3598. S88. Rim fragment of a jar with squared, everted rim. Greyish brown medium fine fabric with sparse rounded purplish grit/pebbles. Appears to have been slipped and burnished, but exterior surface is very abraded. Reduced firing. Handmade. Fabric 2.5 YR 5/4, surfaces 2.5 YR 4/8. D = 16cm. C3600. S88. Body fragment with raised band and painted decoration. Pale orange to brownish medium-fine fabric with sparse small black grit and calcareous inclusions. Horizontal raised band, one horizontal and several light brown orange buff painted vertical lines. Surfaces smoothed. Reduced firing. Handmade. Surfaces 2.5 YR 5/6-6/6, paint 2.5 YR 6/8 5/3. D= unknown. C3601. S88. Body sherd with ribbed decoration on exterior. Dark reddish brown medium fabric with dense small sub-angular quartz and grog (?), occasional larger pebbles. Exterior slipped and burnished. Oxidised firing. Handmade. Fabric 2.5 YR 5/8, surfaces 2.5 YR 6/6-6/8. D= unknown. C3602. S88. Body fragment of a jar with traces of three parallel raised bands on exterior. Orange, medium-fine fabric with sparse small grit and calcareous inclusions. Exterior slipped. Oxidised firing, very hard. Handmade. Fabric 2.5 YR 6/8, surfaces 2.5 YR 6/8. D= unknown. C3603. S88. Part of a twisted handle. Medium coarse light orange brown fabric with sparse small to medium rounded red and purplish grit, some white specs. Surfaces abraded. Oxidised firing. Handmade. Fabric 2.5 YR 6/8. C3604. S88. Body fragment of a jar or jug. Red very brittle medium fabric with sparse medium to large sub-angular purplish grit/pebbles. Exterior surface smoothed with painted dark brown and light red parallel band across the top. Oxidised firing. Handmade. Fabric 10 R 4/8, surfaces 2.5 YR 5/8. D= unknown. C3605. S88. Rim fragment of a necked jar with externally profiled rim. Medium fine fabric with sparse small purplish/black rounded grit, occasional larger pebbles, sparse small angular quartz. Exterior and interior burnished purplish brown slip. Oxidised firing. Handmade. Fabric 2.5 YR 7/6 - 6/8, surfaces 2.5 YR 5/3 - 4/3. D= 18cm. C3606. S88. Rim fragment of a narrow-necked jar with externally profiled rim. Medium fine fabric with sparse small to medium rounded black grit, some larger inclusions. Exterior and upper interior slipped and polished. Oxidised firing. Handmade. Fabric 2.5 YR 6/6, exterior and upper interior 2.5 YR 4/6. D= 10cm. C3610. S88. Body sherd from the neck and shoulder of a jar. Brown fine fabric with sparse very small quartz, small to medium rounded calcarious inclusions and sparse medium rounded
178
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
purplish grit/pebbles. Surface has dark strayations from burnishing. Oxidised firing. Handmade. Fabric 2.5 YR 6/6. D= unknown. C3616. S88. Rim fragment of a jar with squared rim. Grey medium fabric with sparse to medium small to medium black and purplish grit and pebbles. Surface is reddish brown, with reddish brown polished slip on rim and light brown slip on neck. Groove or ribbed decoration running along colour break. Reduced firing. Handmade. Fabric 2.5 YR 5/8, rim 2.5 YR 6/8, neck 7.5 YR 7/6. D= 20cm. C3618. S88. Part of a small handle. Medium fine light orange brown fabric with sparse small rounded dark grit, sparse elongated voids (chaff?). Surfaces very abraded. Oxidised firing. Handmade. Fabric 2.5 YR 6/8. C3620. S88. Large body fragment of a jug with three parallel raised bands below a handle or spout base. Red very brittle medium fabric with sparse medium to large sub-angular purplish grit/pebbles. Oxidised firing. Handmade. Fabric 10 R 7/6 - 7/8, exterior surface 2.5 YR 8/3, interior surface 2.5 YR 6/6. D= unknown. C3639. S88. Small body fragment with traces of ribbed decoration on the exterior. Red, medium coarse fabric with small to medium angular to sub-angular quartz and medium rounded pebbles. Signs of burnishing. Differential firing. Handmade. Fabric 7.5 YR 5/4, surfaces 2.5 YR 6/6. D= unknown. C3641. S88. Small body fragment with traces of ribbed decoration on exterior. Orange brown to light red medium fabric with sparse, small to medium rounded purplish mineral inclusions. Exterior surface abraded, interior burnished. Oxidised firing. Handmade. Fabric 2.5 YR 5/6, surfaces 2.5 YR 6/6. D= unknown. C3642+C3643. S88. Almost complete profile of straight-sided, handled cup. Medium-fine dark orange fabric with sparse, small to medium rounded mineral inclusions. Surfaces are red to purplish in colour, exterior is slipped (?) and burnished. Oxidised firing. Handmade. Fabric 2.5 YR 5/6, surfaces 2.5 YR 4/8. D= 14cm. C3739. S88. Rim fragment of an s-shaped bowl. Reddish-brown to grey very coarse fabric with dense small to large angular to sub-angular quartz. Surfaces abraded. Reduced firing. Handmade. Fabric 5 YR 5/2, interior surface 5 YR 3/1, rim 5 YR 6/4. D= 16cm. C4193. S188. Rim fragment of a hemispherical bowl. Dark brown medium fabric with fine calcareous specks, fine to medium light brown grits. Upper exterior surface is brownish red slipped and polished, lower exterior smoothed, interior brown polished slip. Reduced firing. Handmade. Upper exterior surface 7.5 YR 6/6, lower exterior 2.5 YR 4/6, interior 5 YR 5/6. D= unknown.
Bibliography Alkım, U. B., Alkım H. and Bilgi, Ö. 1988. İkiztepe I: Birinci ve ikinci dönem kazilari. Ankara, Türk Tarih Kurumu. Aydin, B. 2006. A survey of Clazomenaean close territory in Turkey-Izmir-Urla 1997. In A. Erkanal-Öktu, E. Özgen, S. Günel, A. T. Ökse and T. Halil, eds., Studies in Honor of Hayat Erkanal: Cultural Reflections. Istanbul, Homer Kitabevi: 95-105. Bachhuber, C. 2009. The treasure deposit of Troy: Rethinking crisis and agency on the Early Bronze Age citadel. Anatolian Studies 59: 1-18. — 2011. Negotiating metal and metal form in the royal tombs of Alacahöyük in north-central Anatolia. In T.C. Wilkinson, S.E. Sherratt and J. Bennet, eds., Interweaving Worlds. Systemic Interactions in Eurasia, 7th to 1st millennia BC. Oxford, Oxbow: 158-74.
Bibliography
179
Başaran, S. 1999. Enez çevresinde yapılan arkeolojik araştırmalar. In N. Başgelen, G. Çelgin and A. V. Çelgin eds., Anatolian and Thracian Studies in Honour of Zafer Taşlıklıoğlu. Istanbul, Arkeoloji ve Sanat Yayınları: 169-97. Bauer, A. A. 2006. Between the steppe and the sown: Prehistoric Sinop and inter-regional interaction along the Black Sea coast. In D.L. Peterson, L.M. Popova, and A.T. Smith, eds., Beyond the Steppe and the Sown. Proceedings of the 2002 University of Chicago Conference on Eurasian Archaeology. Leiden, Brill: 225-46. Bilgi Ö. 1999a. Samsun-İkiztepe arkeolojik kazıları Tepe III çalişmaları (1993-1994) dönemlari sonuçları. Anadolu Araştırmaları 15: 137-65. — 1999b. Samsun-İkiztepe arkeolojik kazıları Tepe III çalişmaları (1995) dönemlari sonuçları. Anadolu Araştırmaları 15: 167-90. Blegen, C. W., Caskey, J. L., Rawson, M. and Sperling, J. 1950. Troy. General Introduction - The First and Second Settlements. Princeton, Princeton University Press. Blegen, C. W., Boulter, C. G., Caskey, J. L. and Rawson, M. 1958. Troy: Settlements VIIa, VIIb, and VIII. Princeton, Princeton University Press. Bradley, R. 2000. An Archaeology of Natural Places. London, Routledge. Broodbank, C. 2000. An Island Archaeology of the Early Cyclades. Cambridge, Cambridge University Press. — 2010. ‘Ships a-sail from over the rim of the sea’: Voyaging, sailing and the making of Mediterranean societies c. 3500–800 BC. In A. Anderson and G. Barker, eds., The Global Origins of Seafaring. Cambridge, McDonald Institute: 249-64. Burney, C. 1956. Northern Anatolia before Classical times. Anatolian Studies 6: 179-203. Çalış-Sazcı, D. 2006. Die Troianer und das Meer – Keramik und Handelsbeziehungen der sog. Maritimen Troia-Kultur. In M.O. Korfmann, ed., Troia: Archäologie eines Siedlungshügels und seiner Landschaft. Mainz am Rhein, von Zabern: 201-8. Çevik, Ö. 2007. The emergence of different social systems in Early Bronze Age Anatolia: Urbanisation versus centralisation. Anatolian Studies 57: 131-40. Çınaroğlu, A., and Genç, E. 2004. Kastamonu-Kınık 2002 yılı kazısı. Kazı Sonuçları Toplantısı 25/1: 355-66. — 2005. 2003 yılı Kastamonu-Kınık kazısı. Kazı Sonuçları Toplantısı 26/1: 277-90. Dönmez, Ş. 2002. The 2nd millennium BC settlements in Samsun and Amasya provinces, central Black Sea region, Turkey. Ancient West and East 1(2): 243-93. — 2004. Boyabat-Kovuklukaya: A Bronze Age settlement in the central Black Sea region, Turkey. Ancient Near Eastern Studies 41: 38-82. Doonan, O. P. 2004. Sinop Landscapes: Exploring Connection in a Black Sea Hinterland. Philadelphia, University of Pennsylvania Museum of Archaeology and Anthropology. Düring, B. S. 2011. The Prehistory of Asia Minor: From Complex Hunter-Gatherers to Early Urban Societies. Cambridge, Cambridge University Press. Efe, T. 1997. 1995 yılında Kütahiya, Bilecik ve Eskişehir İllerinde yapılan yüzey araştırmaları. Araştırma Sonuçları Toplantısı 14/2: 215-32. — 2004. Yassıkaya, an Early Bronze Age site near Heraclea Pontica (Kdz Ereğli) on the Black Sea Coast. In B. Hansel and E. Studenikova, eds., Zwischen Karpaten und Ägäis. Neolithikum und ältere Bronzezeit. Gedenkschrift für Nemejcova-Pavukova. Rahden, Leidorf: 27-38. — 2007. The theories of the ‘Great Caravan Route’ between Cilicia and Troy: The Early Bronze Age III period in inland western Anatolia. Anatolian Studies 57: 47-64. Efe, T. and Mercan, A. 2001. Yassıkaya: Karadeniz Ereğli (Heraclea Pontica) yakınlarinda bir Tunç Çağı yerleşmesi. Kazı Sonuçları Toplantısı 23/1: 361-74.
180
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
Emre, K. 1964. The pottery of the Assyrian Colony Period according to the building levels of the Kaniş Karum. Anatolica 7: 87-99. Emre, K. and Çinaroğlu, A. 1993. A group of metal Hittite vessels from Kınık-Kastamonu. In M.J. Mellink, E. Porada, and T. Özgüç, eds., Aspects of Art and Iconography: Anatolia and its Neighbors. Studies in Honour of Nimet Özgüç. Ankara, Türk Tarih Kurumu: 675-717. French, D. H. 2008. Chalcolithic and Early Bronze Age pottery of southwest Anatolia. In H. Erkanal, H. Hauptman, V. Şahoğlu, and R. Tuncel, eds., The Aegean in the Neolithic, Chalcolithic and the Early Bronze Age. Ankara, Ankara University Press: 197-202. Georgiev, G. I., Merpert, N., Katincharov, R., Dimitrov, D. 1979. Ezero. Eine Siedlung aus der Frühbronzezeit. Sofia, Bulgarian Academy. Gürkan G., and Seeher J. 1991. Die frühbronzezeitliche Nekropole von Küçükhöyük bei Bozüyük. Istanbuler Mitteilungen 41: 39-96. Gürsan-Salzman, A. 1992. Alaca Höyük. A Reassessment of the Excavation and Sequence of the Early Bronze Age Settlement. Unpublished PhD Thesis, University of Pennsylvania. Hristova, R. 2011. Late Bronze Age pottery from the site of Vratitsa, eastern Bulgaria: Definition, chronology and its Aegean affinities. In B. Horejs and P. Pavúk, eds., Aegean and Balkan Prehistory: http://www.aegeobalkanprehistory.net/article.php?id_art=19 (last accessed 22 Jan. 2013). Işın, M. A. 1998. Sinop region field survey. Anatolia Antiqua 6: 95-139. Ivanova, M. 2012. Perilous waters: Early maritime trade along the western coast of the Black Sea (fifth millennium BC). Oxford Journal of Archaeology 31: 339-65. Jablonka, P. 2011. Troia in regional context. In S. Steadman and G. McMahon, eds., The Oxford Handbook of Ancient Anatolia. Oxford, Oxford University Press: 717-34. Kâmil, T. 1982. Yortan Cemetery in the Early Bronze Age of Western Asia. Oxford, Archaeopress. Karauğuz, G. and Düring. B.S. 2009. A note on the prehistory of the Devrek Region, northern Turkey. Anatolica 35: 153-65. Koppenhöfer, D. 1997. Troia VII: Versuch einer Zusammenschau einsließlich der Ergebnisse des Jahres 1995. Studia Troica 7: 295-353. Korfmann, M. 2001. Der prähistorische Siedlungshügel Hisarlık. In Troia: Traum und Wirklichkeit (Exhibition catalogue). Darmstadt, Archäologisches Landesmuseum Baden-Württemberg: 345-68. Koşay, H. Z. 1951. Türk Tarih Kurumu Tarafından Yapılan Alaca Höyük Kazısı: 1937-1939 dakı Çalışmalara ve Keşiflere ait İlk Rapor. Ankara, Türk Tarih Kurumu. Koşay, H. Z. and Akok, M. 1944. Ausgrabungen von Alaca Höyük: Ein Vorbericht über die im Auftrage der türkischen Geschichtskommison im Sommer 1936 durchgeführten Forschungen und Entdeckungen. Ankara, Türk Tarih Kurumu. — 1950. Amasya Mahmatlar köyü definesi. Belleten 14: 481-87. Kromer, B., Korfmann, M., and Jablonka, P. 2003. Heidelberg radiocarbon dates for Troia I to VIII and Kumtepe. In G.A. Wagner, E. Pernicka, H.-P. Uerpmann, eds., Troia and the Troad: Scientific Approaches. London, Springer: 43-64. Kuzucuoğlu, C., Marro, C., Özdoğan, A. and Tibet, A. 1997. Prospection archéologique francoturque dans la région de Kastamonu (Mer Noire): Deuxième rapport préliminaire. Anatolia Antiqua 5: 275-306. Lloyd, S. and Mellaart, J. 1962. Beycesultan I. The Chalcolihic and Early Bronze Age Levels. London, British Institute of Archaeology at Ankara. Marro, C., Özdoğan, A. and Tibet, A. 1996. Prospection archéologique franco-turque dans la région de Kastamonu (Mer Noire). Premier rapport préliminaire. Anatolia Antiqua 4: 273-90.
Bibliography
181
Matthews, R. 2004. Salur North: An Early Bronze Age cemetery in north-central Anatolia. In A. Sagona, ed., A View from the Highlands: Archaeological Studies in Honour of Charles Burney. Herent, Peeters: 55-66. — 2009. Silent centuries: Paphlagonia from the Palaeolithic to the Early Bronze Age 200,000 to 2000 BC. In R. Matthews and C. Glatz, eds., At Empires’ Edge: Project Paphlagonia. Regional Survey in North-Central Turkey. London, British School at Ankara: 75-106. Mellaart, J. 1961. Early cultures of the south Anatolian plateau I. Anatolian Studies 11: 159-184. Meriç, R. 1992. Torbalı-Metropolis Kazısı 1990. Kazı Sonuçları Toplantısı 13/1: 237-40. Merkyte, I. 2007. Ezero-Kale: From the Copper Age to the Bronze Age in the southern Balkans. Acta Archaeologica 78(2): 1-78. Muhly, J. D. 2011. Metals and metallurgy. In S. Steadman and G. McMahon, eds., The Oxford Handbook of Ancient Anatolia. Oxford, Oxford University Press: 858-76. — 1998. Prospection archéologique franco-turque dans la région de Kastamonu (Mer Noire). Troisième rapport préliminaire. Anatolia Antiqua 6: 317-35. Orthmann, W. 1963a. Die Keramik der frühen Bronzezeit aus Inneranatolien. Berlin, Gebr. Mann. — 1963b. Frühe Keramik von Boğazköy-Hattuša aus den Ausgrabungen am Nordwesthang von Büyükkale. Berlin, Gebr. Mann. Özdoğan, M. 1982. Doğu Marmara ve Trakya araştırmaları. Türk Arkeoloji Dergisi 26: 37-61. Özgüç, T. 1986. New Observations on the relationship of Kültepe with southwest Anatolia and north Syria during the third millennium B.C. In J.V. Canby, E. Porada, B.S. Ridgeway, and T. Stech, eds., Ancient Anatolia. Aspects of Change and Cultural Development, Essays in Honour of Machteld J. Mellink. Madison, University of Wisconsin: 31-47. Özgüç, T., and Akok, M. 1958. Horoztepe: Eski Tunç Devri Mezarliği ve İskân Yeri. Horoztepe: An Early Bronze Age Settlement and Cemetery. Ankara, Türk Tarih Kurumu. Peschlow-Bindokat, A. 2006. Die Arbeiten des Jahres 2004 im Latmos. Araştırma Şonuçları Toplantısı 30/3: 269-78. — 2007. Tarihöncesi insan resimleri. Latmos dağları’ndaki prehistorik kaya resimleri. Istanbul, Zero Books. Peschlow-Bindokat, A., and Gerber, C. 2003. Die Arbeiten des Jahres 2002 in Herakleia am Latmos und Umgebung (Bafa Gölü/Beşparmak). Araştırma Şonuçları Toplantısı 21/2: 201-10. Rahmstorf, L. 2006. In search of the earliest balance weights, scales and weighing systems from the eastern Mediterranean, the Near and Middle East. In M.E. Alberti, E. Ascalone and L. Peyronel, eds., Weights in Context: Bronze Age Weighing Systems of Eastern Mediterranean: Chronology, Typology, Material and Archaeological Contexts. Roma, Istituto Italiano di Numismatica: 9-45. Şahoğlu, V. 2005. The Anatolian Trade Network and the Izmir region during the Early Bronze Age. Oxford Journal of Archaeology 24: 339-61. Şahoğlu, V., ed. 2011. Across. The Cyclades and Western Anatolia during the 3rd Millennium BC. Istanbul, Sakıp Sabancı Museum. Sayre, E. V., Joel, E. C., Blackman, M. J., Yener, K. A., and Özbal, H. 2001. Stable lead isotope studies of Black Sea Anatolian ore sources and related Bronze Age and Phrygian artefacts from nearby archaeological sites. Archaeometry 43(1): 77-115. Schoop, U.D. 2005. Das anatolische Chalkolithikum. Remshalden, Albert Greiner Verlag. — 2011. Some thoughts on social and economic development in western Anatolia during the fourth and third millennia BC. In A.N. Bilgen, R. von den Hoff, S. Sandalcı and S. Silek, eds., Archaeological Research in Western Central Anatolia. Kütahya, Üçmart Press: 29-45.
182
The Later Prehistory of the Cide-Şenpazar Region: ca. 2600-2000 BC
Seeher, J. 2000. Die bronzezeitliche Nekropole von Demircihüyük-Sarıket: Ausgrabungen des Deutschen Archäologischen Instituts in Zusammenarbeit mit dem Museum Bursa, 1990-1991. Tübingen, Wasmuth. Séfériadès, M. 1995. Deshayes’ excavations at Dikili Tash: The Early Bronze Age levels. In L. Nikolova, ed., Early Bronze Age Settlement Patterns in the Balkans (ca. 3500-2000 BC, Calibrated Dates). Part 2. Sofia, Prehistory Foundation: 95-128. Sherratt, A. 1986. The pottery of phases IV and V: the Early Bronze Age. In C. Renfrew, M. Gimbutas, and E.S. Elster, eds., Excavations at Sitagroi. A Prehistoric Village in Northeast Greece. Los Angeles, Cotsen Institute of Archaeology, University of California Press: 429-76. Stos-Gale, Z. A. 1993. The origin of metal used for making weapons in Early and Middle Minoan Crete. In C. Scarre and F. Healy, eds., Trade and Exchange in Prehistoric Europe. Oxford, Oxbow: 115-30. Süel, M. 1989. Balıbağı / 1988 kurtarma kazisi. Türk Arkeoloji Dergisi 28: 145-63. Thissen, L. 1993. New insights in Balkan-Anatolian connections in the Late Chalcolithic, Old evidence from the Turkish Black Sea littoral. Anatolian Studies 43: 207-37. Toker, A., and Özturk, J. 1992. Metal Vessels. Ankara, T.C. Kültür Bakanlığı, Anıtlar ve Müzeler Genel Müdürlüğü. Topbaş, A., Efe, T. and İlaslı, A. 1998. Salvage excavations of the Afyon Archaeological Museum, Part II: The settlement of Karaoğlan Mevkii and the Early Bronze Age cemetery of Kaklık Mevkii. Anatolia Antiqua 6: 21-94. von der Osten, H. H. 1937. The Alishar Hüyük: Seasons of 1930-32. Chicago, University of Chicago Press. Wagner, G. A. and Öztunalı, Ö. 2000. Prehistoric copper sources in Turkey. Ü. Yalçin, ed., Anatolian Metal I. Bochum, Deutsches Bergbau Museum: 31-61. Welton, M. L. 2010. Mobility and Social Organization on the Ancient Anatolian Black Sea Coast: An Archaeological, Spatial and Isotopic Investigation of the Cemetery at İkiztepe. Unpublished PhD thesis, University of Toronto. Whalen, J. 2014. Feasting and Shared Drinking Practices in the Early Bronze Age II-III (2650-2000 BC) of North-Central and Western Anatolia. Unpublished PhD thesis, University of Edinburgh. Yakar, J. 2000. Ethnoarchaeology of Anatolia. Rural Socio-Economy in the Bronze and Iron Ages. Tel Aviv, The Institute of Archaeology. Yalçinkaya, N. 1987.1986 yılı Karain kazısı. Kazı Sonçları Toplantısı 9/1: 15-38. Zimmerman, T. 2006. Kalinkaya: A Chalcolithic-Early Bronze Age settlement in northern Central Anatolia, first preliminary report: The burial evidence. Anadolu Medeniyetleri Müzesi Koferansları 2005 yilliği: 271-311. — 2009. Frühmetallzeitliche Eliten zwischen Ostägäis und Taurusgebirge im 3. Jahrtausend v. Chr. – Versuch einer kritischen Bestandsaufnahme. In M. Egg and D. Quast, eds., Aufstieg und Untergang: Zwischenbilanz des Forschungsschwerpunktes ‘Eliten’. Mainz, RömischGermanisches Zentralmuseum: 1-30. — 2012. Lapanu – Let (it) glow! – Recent archaeometric analyses of Hattian and Hittite metalwork. In R. Matthews, J. Curtis, M. Saymour, A. Fletcher, A. Gascoigne, C. Glatz, St J. Simpson, H. Taylor, J. Tubb and R. Chapman, eds., Proceedings of the 7th International Congress on the Archaeology of the Ancient Near East. Wiesbaden, Harrassowitz: 303-12.
Claudia Glatz
7 Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region The second millennium BC is one of the most intriguing as well as archaeologically challenging periods in the early history of northern Turkey. Ample textual sources document an enduring struggle between the polities of the central Anatolian plateau and the inhabitants of the Turkish Black Sea region, but scant archaeological work and the difficulty of identifying contemporary local material culture traditions have resulted in an exclusively southern, central Anatolian, perspective of both Middle and Late Bronze Age encounters. A central focus of the Cide Archaeological Project’s (CAP) research agenda has been to contribute to a local, archaeological understanding of the second millennium BC occupation of the Turkish Black Sea region and the relationships of its communities with contemporary societies on the central plateau and elsewhere along the Black Sea coast (chapter 1). Here, I present the second millennium BC material recorded by CAP, which - although small in absolute numbers - is significant in that it includes the only Late Bronze Age assemblage recorded to date along Turkey’s central Black Sea coast.
7.1 Issues and Open Questions The Turkish Black Sea region as a whole is poorly understood in the second millennium BC. This is due to a general lack of archaeological work in the region, especially in its forested and mountainous coastal strip, which poses a number of challenges for archaeological reconnaissance. One of these, which carries particular significance for the Bronze Age, is the apparent absence of mounded sites. This is most likely a consequence of the social organisation and subsistence strategies of local Bronze Age communities and the preferential use of stone and wood architecture, but whose effects are no doubt amplified by taphonomic processes resulting in the destruction or burial of sites (chapters 2 and 6). From the point of view of field-survey, the absence of easily visible höyüks significantly reduces the chances of identifying Bronze Age settlements, especially during extensive survey. Flat sites and surface scatters are readily recognised during intensive surface survey, but this is complicated in the Cide-Şenpazar and wider Turkish Black Sea region by rugged topography and dense vegetation cover. An active coastal landscape, moreover, means that any second millennium BC coastal activity now lies covered by several metres of alluvial and colluvial deposits (section 2.3). Unsurprisingly then, there are many unresolved questions regarding the Turkish central Black Sea region during the second millennium BC beyond the
184
Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region
limited information that can be gleaned from contemporary textual sources and their conflict-centred narrative (von Schuler 1965; Glatz and Matthews 2005). For instance, we know next to nothing about potential local Middle Bronze Age and Late Bronze Age cultural traditions, their pace of development and their relationship with those of neighbouring regions. Pertinent, if impossible to answer with surface data, is the question of whether technological changes such as the transition from hand- to wheel-production took place at the same time in the remote valleys of the Black Sea region as among the increasingly urban and socio-politically centralised communities further to the south, or whether handmade traditions continued to be produced throughout the second millennium BC in rural and marginal regions (Seeher 2010; Schoop 2011: 265-6). If technological adoption did take place, which in the case of the potter’s wheel requires long-term apprenticeship or specialist mobility (Roux 2015), under what cultural and social conditions did this process occur? In the absence of reliable alternative chronological indicators and dating methods (also chapter 5), the identification of second millennium BC assemblages in the region has to rely on the presence of material derived either directly or inspired by the cultural traditions of archaeologically better-known regions and sites, in particular the central Anatolian plateau. This comes with a whole set of chronological and interpretive problems. For instance, it is difficult to define the transitional phase between the final Early Bronze Age and the Middle Bronze Age due to the absence of distinctive handmade and painted Intermediate/ Cappadocian wares in the Turkish Black Sea region (e.g. Dönmez 2008: 89; Glatz et al. 2009: 108). By contrast, Middle Bronze Age and Late Bronze Age central Anatolian pottery types with their distinctive buff fabrics, signs of wheelproduction and high firing temperatures survive well in surface assemblages and are easily recognised. As ceramic studies at Boğazköy-Hattusa and elsewhere have demonstrated, however, inherent formal and technological conservatism pose significant problems for the construction of a relative chronology based on northcentral Anatolian pottery (Müller-Karpe 1988; Parzinger and Sanz 1992; Schoop 2003, 2006, 2011). Chronological variation in central Anatolian assemblages can be identified with confidence only on the basis of the relative abundance of formal types and through morphological developments in vessel attributes, which require a quantitative approach supported by absolute dates (Schoop 2006). As a consequence, only broad chronological labels can be applied to surface assemblages with varying degrees of confidence (see also Glatz et al. 2009). I here use a modified version of Schoop’s (2011: 242) tripartite chronological division of (1) an early phase, which I have extended to include the period from the 20th to the 17th century BC to account for the added difficulty of dating surface material; (2) a middle phase spanning the 16th and 15th centuries BC; and (3) a late phase which includes the 14th and 13th centuries BC.
The Middle and Late Bronze Age Evidence from Cide-Şenpazar
185
7.2 The Middle and Late Bronze Age Evidence from Cide-Şenpazar The pottery assemblage collected in the course of CAP, which can be assigned a second millennium BC date with some confidence is very small and, thus, any conclusions based on this assemblage have to be tentative and provisional (Fig. 7.1). There is a relatively clear chronological division with regards to the locations where we found early and later second millennium BC material. Fragments of vessels seemingly belonging to the transitional phase from the third to the second, or perhaps falling into the first half of the second millennium BC, derive for the most part from the cave sites of Kılıçlı Mağarası (S88/S93) and Derebağ Köy Mağarası (S188). In contrast, all the pottery securely datable to the Late Bronze Age as well as fragments of a Middle Bronze Age vessel come from surface scatters in the vicinity of Okçular Kale (S22). Chance of discovery, and the better preservation of the later material on the surface may help explain this difference, which is even more pronounced in the Early Bronze Age (chapter 6) and earlier periods (chapter 5). It is, however, also possible, that this pattern tracks past social and cultural developments in the Turkish Black Sea region, perhaps associated with, or in response to, the major transformations taking place further to the south, which include the emergence and expansion of hierarchical societies and centralised political structures.
0
5
10 Km
Cide Derebağ Köy Mağarası
Okçular Kılıçlı Mağarası
Fig. 7.1: Map showing the locations of second millennium BC finds. Produced by Toby C. Wilkinson.
186
Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region
7.2.1 The Middle Bronze Age (2000-1600 BC) A small number of pottery fragments from Kılıçlı Mağarası (S88) may belong to the difficult to define transitional Early-Middle Bronze Age phase or to the first phase of the second millennium BC sequence. For instance, the small rim fragment of a red slipped and burnished jar with slightly everted rim1 finds general parallels in the early Middle Bronze Age levels at Boğazköy (Fischer 1963: Tafel 9. 124, M/18; Orthmann 1963: Tafel 14. 139, NW Hang 9). Triangular, slipped and burnished horizontal handles belonging to bowls with inverted upper bodies appear in the first part of the ceramic sequence and go out of use in the second (Schoop 2011: 246). The CAP material does not include any exact parallels for this type of horizontal handle, but an example with faceted profile2 from Kılıçlı Mağarası may represent a local take on this general theme. A similar date is also likely for a small, slipped and polished open spout3 probably belonging to a juglet with steep, upward-facing beak with parallels at BoğazköyHattusa (Fischer 1963: Tafel 23: 268, Tafel 26: 300 but with more pronounced beak; Müller-Karpe 1988: Taf. 1: RK 1-4); and two small rims, one of a jar with constricted neck4 (e.g. Parzinger and Sanz 1992: E2.4) and one of a bowl5 (for parallels see Fischer 1963: Tafel 89: 770; Parzinger and Sanz 1992: I2.2), both in medium-fine light brown fabric and with smoothed surfaces. A hand- (and wheel?) made, slipped and roughly burnished bowl with everted rim and horizontal handle6 from Derebağ Köy Mağarası (S188) shares formal elements with the larger bowls with inverted or straight upper bodies and horizontal handles found in MBA layers at Boğazköy (Büyükkale IVd: Orthmann 1984: Abb. 14; Fischer 1963: Tafel 105, 940; Tafel 106, 943). The formal similarities, however, are rather general and there are differences in surface treatment. The Boğazköy bowls tend to be partially slipped and polished on the upper exterior, whereas the Derebağ Köy example is burnished on both sides. A similar bowl type is also attested at Troy VI and into Troy VII, where it appears in Anatolian Grey, Red Washed and Tan Wares (Pavúk 2002: 40, fig. 3). The bowl fragment is mineral tempered, including small black glassy inclusions, which may be volcanic sand from the Sinop area and an indication of the origin of the vessel.
1 C3056 (Fig. 7.2 and 7.3). 2 C3051 (Fig. 7.2 and 7.3). 3 C3254 (Fig. 7.2 and 7.3). 4 C2969 (Fig. 7.2 and 7.3). 5 C3057 (Fig. 7.2). 6 C4235 (Fig. 7.2 and 7.3).
The Middle and Late Bronze Age Evidence from Cide-Şenpazar
Fig. 7.2: Second millennium BC pottery from the Cide region. Produced by Lorraine McEwan.
187
188
Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region
Fig. 7.3: Second millennium BC pottery from the Cide region. Produced by Lorraine McEwan.
The Middle and Late Bronze Age Evidence from Cide-Şenpazar
189
Two small fragments of a very coarse closed vessel with a raised band and parallel incisions7 were collected in the meadows immediately to the south of Okçular Kale (S22) (Fig. 7.4). The vessel is handmade and the low-fired brittle matrix is heavily tempered with angular crushed calcite (marble, see section 16.3: C2056). The interior vessel wall is covered in dark soot. The fragment itself is very small and lacks a rim, which renders any identification and dating difficult. The coarse, handmade character, internal residue and decorated band, however, point to similarities with a Middle Bronze Age central Anatolian cooking pot tradition found at Boğazköy (NW- Slope Level 8: Orthmann 1963: Tafel 27: 247-51, Tafel 33: 330-33; Büyükkale: Orthmann 1984: Abb. 1: 10-11, Abb. 76-7; Büyükkaya – Ulf-Dietrich Schoop personal communication 13-12-2012), Alaca Höyük (Orthmann 1963: Tafel 40: 3); and Kültepe (Özgüç 1949: 184 Abb. 266 no provenance). Some of the cooking pots from Boğazköy are also reported to have black or brown internal surfaces (Orthmann 1963: 36). Several examples of similar vessels were recorded at the site of Gâvurevleri in Daday district of Kastamonu province (Marro et al. 1998: 325, pl. I, 2-4).
Okçular wares
0
100
200 m
Other 2nd m. wares Other pottery finds
G8
Field units Gridding squares
Okçular Kale İni
G3 Kale
G2
G1
Fig. 7.4: Distribution of second millennium BC pottery in the Okçular valley. Produced by Toby C. Wilkinson.
7 C2854 (Fig. 7.2 and 7.3).
190
Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region
7.2.2 The Late Bronze Age (1600-1200 BC) Evidence for a Late Bronze Age occupation of the Okçular valley comes from a pottery concentration at the foot of a smaller rock outcrop three hundred metres to the northeast of Okçular Kale (Fig. 7.5). Intensive survey (Unit 683) and grid-square collections (G8) in this area yielded several rim fragments that can be dated with confidence to the Late Bronze Age. Among the fragments are the rims of two carinated bowls with everted rims8 (for parallels see Müller-Karpe 1988: S10; Parzinger and Sanz 1992: I8), the fragment of a bowl with internally profiled, blackened rim9 (Müller-Karpe 1988: S5; Parzinger and Sanz 1992: I3), a second internal rim bowl with fine, pinkish micacious fabric10 (Müller-Karpe 1988: S5; Parzinger and Sanz 1992: I3), the rim-fragment of a mediumsized funnel-neck jar11 (Müller-Karpe 1988: T10-12; Parzinger and Sanz 1992: A4) and another similar jar type.12 A further possible Late Bronze Age jar fragment13 comes from the coastal plain in the Irmak/Cide area.
Fig. 7.5: View north from Okçular valley showing the locations of grids G1 and G8 (CAP-P2/916).
8 C1089 (Fig. 7.6 and 7.7) and C4186 (Fig. 7.6). 9 C4052 (Fig. 7.6). 10 C2251 (Fig. 7.6 and 7.7). 11 C4185 (Fig. 7.6). 12 C3949. 13 C0961 (Fig. 7.6).
The Middle and Late Bronze Age Evidence from Cide-Şenpazar
Fig. 7.6: Late Bronze Age pottery from the Cide region. Produced by Lorraine McEwan.
Fig. 7.7: Late Bronze Age pottery from the Cide region. Produced by Lorraine McEwan.
191
192
Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region
Several of the Late Bronze Age sherds show signs of wheel-production, although the surface on some pieces was too abraded to assess manufacturing technique. In terms of their shape, the pieces conform to standard north-central Anatolian types, showing no signs of production by a potter unfamiliar with central Anatolian pottery shapes or production methods. Even though the Late Bronze Age sample is exceedingly small, a range of different types are represented that point to a mixture of food preparation and presentation as well as short to medium term storage. All of the above shapes are characteristic of the middle part of the central Anatolian ceramic sequence, although they continue to be produced in the 14th and 13th centuries BC.
7.2.3 A Local Ceramic Tradition Evidence from surface survey is ill-suited in principle for the identification and especially the dating of local and previously unknown cultural traditions. It is therefore with caution (and possible future regret) that I include in this chapter what appears to be a highly localised and previously undocumented pottery tradition, whose concentration around Okçular Kale (Figs. 7.4 and 7.5) has earned it the name ‘Okçular ware’. Okçular ware fragments are handmade and heavily tempered with small to large angular and sub-angular crushed calcite (section 16.3), not dissimilar to the fabric of the cooking pot with impressed band discussed above and the Chalcolithic pottery from the cave of Okçular Kale İni (S24) (chapter 5). However, Okçular ware examples differ from both in terms of firing, surface colours and the formal types represented. Despite its coarse character, most sherds are very hard fired in both oxidising and reduced atmospheres. Surfaces range from reddish brown to bright red. Among the diagnostic pieces are a grooved rim bowl,14 a small s-shaped bowl or cup,15 a large jar with everting rim16 and a smaller jar with rounded and slightly everted rim.17 Grid G8 to the northwest of the kale also yielded a flat base made from this fabric type.18 Hemispherical bowls and hole-mouth jars with simple, rounded and everting rims are among the earliest ceramic shapes attested in Anatolia. By contrast, the grooved rim bowl is a much more unusual shape, which is not found in earlier assemblages in the region. Grooved rim bowls are known from Level II (end of the Late Bronze Age and Early Iron Age) Kilise Tepe (Hansen and Postgate 2007: figs. 736-8) in Cilicia and other south-Anatolian sites, but there they are wheelmade in medium to fine fabrics and any connection with the Okçular example is unlikely.
14 C0893 (Fig. 7.7. and 7.8). 15 C2276 (Fig. 7.7. and 7.8). 16 C1088 (Fig. 7.7. and 7.8). 17 C2509. 18 C3948 (Fig. 7.7).
The Middle and Late Bronze Age Evidence from Cide-Şenpazar
193
Fig. 7.8: ‘Okçular ware’ diagnostic types. Produced by Lorraine McEwan.
The provisional assignment of this pottery to the second millennium BC, while not impossible on the basis of the above arguments, is by no means certain. Handmade traditions, many as yet undocumented, exist also in much later periods (see e.g. Vroom 2005). Only further research, and ideally excavation, will provide an answer about the date of this distinctive and enigmatic pottery tradition. What we can be certain of is that it represents an element of local material culture, whose technological characteristics are clearly related to better datable Chalcolithic and Bronze Age traditions from the same location. Such traditions are normally very difficult to pin down in surface assemblages, but they may hold important insights about local societies, which may or may not ever come to feature in historical sources. The Kaska of the Late Bronze Age are a prime example for such a society, but the mountainous region of northern Turkey no doubt was home to many more people without histories.
194
Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region
Fig. 7.9: ‘Okçular ware’ diagnostic types. Produced by Lorraine McEwan.
7.3 The Turkish Central Black Sea Region in the Middle and Late Bronze Ages Central Anatolian textual sources dating to or referring to the first half of the second millennium BC do not concern themselves much with the Turkish Black Sea region. Those that do provide little detail about the societies inhabiting the region and their relationships with communities further to the south. The geographical focus of two Late Bronze Age texts, whose historical setting appears to be the later part of the Middle Bronze Age, is the city of Zalpa (or Zalpuwa). Zalpa is conventionally located on or near the mouth of the Kızılırmak in the Bafra plain and equated with the site of İkiztepe (Alkım 1973; 1977: 199-200). It is not entirely clear whether earlier Middle Bronze Age texts already refer to this Black Sea Zalpa (Barjamovic 2011: 118-20). Several cities with the name Zalpa are mentioned in Old
The Turkish Central Black Sea Region in the Middle and Late Bronze Ages
195
Assyrian texts; one is securely located in upper Mesopotamia and another in eastern Turkey. The existence of an Anatolian Zalpa in the Middle Bronze Age has been proposed on the basis of a small number of texts, which detail the transport of lapislazuli and shrimps to and from trade locations on the central plateau (e.g. Dercksen 2001: 59-60; Gelb 1955: Alişar Texts 5 and 6; kt n/k 10:5). There are no toponyms in Old Assyrian sources that can be associated with the northwestern part of the Turkish central Black Sea region. In part this may be due to the lack of archaeological research in this area and a resulting absence of sites to pin toponyms onto. Old Assyrian texts, moreover, rarely mention settlements without resident Assyrian communities (Barjamovic 2011: 390-1), thus providing a selective perspective of Middle Bronze Age political geography which focuses on the central plateau and its western and eastern extensions, including the Bafra plain (Barjamovic 2008: Map 2). Archaeological evidence points to well-established connections between the central Anatolian cultural sphere and the eastern part of the Turkish central Black Sea region during the early part of the Middle Bronze Age (also chapter 6). Pottery from İkiztepe Mound 1 (Alkım et al. 1988), Dündartepe and Tekkeköy (Kökten et al. 1945) in Samsun province and Boyabat-Kovuklukaya (Dönmez 2004) in Sinop and other sites in the region share affinities with the ceramic assemblages of Middle Bronze Age Kültepe and other contemporary central Anatolian sites (V. Müller-Karpe 2001; Dönmez 2004). However, the former assemblages tend to be associated with Early Bronze Age ceramics in excavated contexts at İkiztepe and Boyabat-Kovuklukaya, posing a number of stratigraphic and chronological questions. Inner Kastamonu too has yielded pottery typical of the central Anatolian Middle Bronze Age (Marro et al. 1996: 282-4; Kuzucuoğlu et al. 1997: 292-3; Marro et al. 1998: 325). Marine resources, such as the shrimp mentioned in the Old Assyrian texts, as well as the rich metal sources of the Pontic mountains were likely catalysts for this interaction. There is evidence for metal smelting and casting at Boyabat-Kovuklukaya (Dönmez 2004: 42-3) and the metal finds from the nearby looted Hıdırlı cemetery as well as the finds from İkiztepe are comparable to items from central Anatolian sites (Bilgi 2001; Dönmez 2012). In Hittite texts of the Late Bronze Age, Zalpa plays an important role in the process of Hittite state and identity formation (Klinger 2010: 229). The Old Hittite Anitta text (CTH 1;19 Neu 1974) recounts the conflict and competition between the central Anatolian city of Kanes with Zalpa ‘on the sea’. The text describes how the king of Zalpa had destroyed Kanes and removed its city god’s statue to Zalpa, and how Anitta subsequently returned the statue to Kanes and took the rival king hostage. Although the preserved copies of the text date to the 16th and 13th centuries BC respectively, the events described in the Anitta text are generally thought to take place in the preceding Middle Bronze Age (Neu 1974: 1), when a great-king Anitta is mentioned in Old
19 CTH = Catalogue des textes hittites (Laroche 1971).
196
Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region
Assyrian sources as well as on a bronze dagger from Kültepe. The story of the ‘Queen of Kanes’ (CTH 3; Otten 1973) also deals with the complicated relationship between Kanes and Zalpa, albeit from a mythological angle. A rebellion of Zalpa, possibly the one on the Black Sea, is mentioned in the Annals of Hattusili (Bryce 2009: 785-6). From the 15th century BC, Hittite engagement with the Black Sea region as reflected in textual sources widens in geographical scope as the north of the Land of Hatti and the Upper Land became a primary security concern. A broad range of textual sources details the long-lasting conflict between the Hittite state and the socio-politically more loosely organised Kaska. Fifteenth-century BC texts recount the destruction of important Hittite cult centres, including Zalpa, while later sources detail annual military campaigns against Kaska groups, the rebuilding of destroyed towns and the resettlement of abandoned landscapes by Hittite and deportee populations. They also mention occasional, and inevitably ill-fated, attempts to resolve the conflict through diplomatic means (von Schuler 1965; Klinger 2005) and provide insights into the day-to-day management of frontier settlements and regions (Alp 1991; Hoffner 2009: 91-251). Hittite sources portray Hittite-Kaska relations as primarily hostile in character. The Kaska are the inevitable aggressor, deploying hit-and-run tactics on Hittite settlements and rarely facing the Hittite army in open combat, instead retreating into the high mountains where they are unreachable for Hittite troops. How far north towards the Black Sea coast this conflict between the Hittite state and the Kaska tribes extended depends on the reconstruction of Hittite historical geography. A close match between Strabo’s (7.3.40) Blaënê and Domanitis with the northern most Hittite provinces of Pala and Tumanna and the identification of the Dahara river with the Gökırmak would place much of the conflict described in the texts as far north as central Kastamonu and southern Sinop (e.g. Forrer 1928-1929; Güterbock 1961; Houwink ten Cate 1967). More recently, however, the Dahara river has been equated with the Devrez Çay, located to the south of the Ilgaz mountains (Forlanini 1977; Forlanini and Marazzi 1986), which matches not only the descriptions of Hittite troop movements more comfortably, but is supported by archaeological evidence in the form of a string of heavily fortified Late Bronze Age settlements dotted along the Devrez Çay (Matthews and Glatz 2009). Further to the north, evidence for a second millennium BC presence is sparse, especially for the Late Bronze Age. Survey in the Gökırmak, Araç, Daday Çay and Karaçomak valleys yielded a small number of sites with evidence for mainly the first part of the north-central Anatolian ceramic sequence (Marro et al. 1996: 282; fig. 9; pl. III). Pottery typical of the middle part of the sequence is much more rare and individual pieces are reported from Höyükdoruğu, Kayabaşı, Üyüktepe and possibly also Gâvurevleri. Chance finds from Kınık in Devrekani and a cave in Pınarbaşı, both in Kastamonu province, further hint at Late Bronze Age activities in the region. At Kınık, an eclectic collection of metal vessels, figurines and instruments were discovered during road constructions. Some of the vessels can be dated with certainty to the Late Bronze Age, including a silver bowl with an engraved Luwian hieroglyphic inscription which dates
The Turkish Central Black Sea Region in the Middle and Late Bronze Ages
197
to the 14th-13th centuries BC (Hawkins 1993), a second bowl whose decorative hunting frieze fits well into the Late Bronze Age north-central Anatolian iconographic tradition, as well as a large, bull-headed cup (Emre and Çınaroğlu 1993). Other finds, including a small metal warrior figurine (Gates 1997), displayed in the Kastamonu Archaeological Museum as part of the Kınık hoard, however, appear to be much later in date, which puts in question the integrity of the hoard and when and how it came to be deposited. Excavations at Kınık revealed an Early Bronze Age workshop area including crucibles and slag as well as evidence for pottery production and storage. A wallfeature was dated to the early second millennium BC and an Iron Age occupation level was also identified by the excavators (Gates 1997; Bilgen 1999; Çınaroğlu and Genç 2004; 2005). A lack of published ceramic evidence, however, makes it currently impossible to fully assess the dating of the site and its cultural affiliation. None of the ceramics on display in the Kastamonu Archaeological Museum are of recognisable Late Bronze Age types. Another recent metal find comes from the cave of Mızrak/Buz Mağarası in Pınarbaşı, which is located just outside of the southern boundary of CAP’s survey area. Inside the cave, a British speleological expedition found a small bronze sword (Harris 1992: 35, 41; 1993: 25). Based on similarities with a sword from BoğazköyHattusa, which carries an inscription mentioning Tudhaliya I/II, the Mızrak/Buz Mağarası sword may be dated to the Late Bronze Age. Conventionally, the Boğazköy sword is thought to be of Mycenaean or west Anatolian origin (e.g. Mellink 1993: 112-3; Hansen 1994) and by extension, Ünal (1999) proposed an Aegean origin for the Pınarbaşı sword. More recently, however, the Boğazköy sword is thought of as Anatolian in tradition (Taracha 2003) and a more local origin for the Pınarbaşı example seems also more convincing. A similar sword found in Bulgaria and now housed in the Varna Museum (Althanassov et al. 2012) may provide a connection across the Black Sea. In view of the eclectic collection of artefacts from Kınık and the seeming lack of recognisable Late Bronze Age occupation at the site, it is possible, if not provable, that they represent loot from Hittite sites or temples by Kaska raiders (Matthews 2000: 1017), as may be the Mızrak/Buz Mağarası sword, but other scenarios are also plausible. The narrative of perpetual conflict with the Kaska north, which dominates not only Hittite sources but also the more recent discourse, overshadows and downplays other forms of interaction between central Anatolia and the mountainous regions to its north. However, especially in the border zone, intensive, ‘unofficial’ forms of interaction will have taken place on a daily basis (Lightfoot and Martinez 1995). Brief references in Hittite sources at attempts to regulate Kaska-Hittite interaction, including mixed marriages and the conscription of Kaska into the Hittite military (von Schuler 1965: 56-60, 146-7) hint at such relations. They, no doubt, also included the exchange of agricultural produce, raw materials and value-added items, such as swords and prestige drinking equipment and, in the case of the latter, perhaps also associated ritual and consumption practices.
198
Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region
Central Anatolian textual and archaeological evidence seems to suggest that two important characteristics of the Kaska economy, at least from the Hittite point of view, flax production and pig farming, were of relatively little interest to the Hittite south (Glatz and Matthews 2005: 57-9). Metal, especially copper, and timber are important resources of the Turkish central Black Sea region, but there is as yet no unequivocal evidence for either the exploitation of the Küre and other mines in the region during the Late Bronze Age or Hittite interest in them. There is also no direct evidence for the procurement of timber in the region or their transport to the south. The copious timber used in Late Bronze Age central Anatolian architecture was procured from local sources at least in the earlier part of the Late Bronze Age (Dörfler et al. 2011: 103). Intensification of agricultural production, forest clearing to create grazing areas, and the logging of construction timbers, however, did have dramatic effects on central and north-central Anatolian landscapes from about 1500 BC onwards (Bottema and Wording 1990; Marsh et al. 2009; Dörfler et al. 2011: 101103). Woodlands across the region degraded and retreated to higher altitudes as a result, forcing central Anatolian communities to search further afield for construction material (Dörfler et al. 2011: 103). The more northerly regions of the Pontic mountains still await a detailed archaeobotanical and geoarchaeological assessment of this kind, but it would seem that they remained largely unaffected by these developments in the Bronze Age (Bottema et al. 1993-4) and, thus, might have become a more attractive timber source as the Late Bronze Age progressed. Other, equally tentative hints at a connection between the Black Sea coast and central Anatolia come in the form of the shells of oysters and marine snails, which were found on Büyükkaya (von den Driesch and Pöllath 2004: 6-8) and in the 15th century BC Square building (Quadratgebäude) (Seeher 2003: 9; 2004: 65) at Boğazköy as well as in Building E at Kuşaklı-Sarissa (von den Driesch 2009: 144). They may equally, however, have come from the Mediterranean. Fish, although mentioned in a range of ritual texts, appears not to have formed an important part of Hittite diet (Hoffner 1974: 124-5). A small number of fish bones from Boğazköy-Hattusa include local fresh-water fish as well as shark vertebrae (von den Driesch and Boessneck 1981; von den Driesch and Pöllath 2004: 6-8; van Neer et al. 2005: 149). To date, no Bronze Age shipwrecks that would point to maritime trade within and beyond the Black Sea have been discovered. Yet, Mycenaean traders have been envisioned in the past as operating within the Black Sea region on the basis of stone anchors and metal ingots found at Bulgarian sites, the appearance of double axes in the river deltas of the northwest and northern Pontic and a series of swords and spearheads from Trialeti graves in Georgia (e.g. Buchholz 1983: 90-94; Bouzek 1985: 35-45; Hillier 1991; Korfmann 1995). With regards to Turkey, French (1982), for instance, proposed a Black Sea route for the transport of a small number of Mycenaean and Cypriot vessels found at Maşat Höyük (Özgüç 1978: 66, 127-28, pls. 83-84; Özgüç 1982: 102-03, pl. 47.5-6.). However, chronological issues, recent provenance studies and the lack of convincing morphological similarities with Mycenaean material culture have
Settlement Trends in the Turkish Central Black Sea Region
199
led to the rejection of a Mycenaean-run maritime trade network in the Late Bronze Age Black Sea in recent years (e.g. Tsetskhladze 1999: 473, note 6; de Boer 2006-7; Leshtakov 2007). There is also no evidence from the Turkish Black Sea coast and no indication in Late Bronze Age or earlier sources for the existence of a burgeoning Black Sea maritime trade or Anatolian participation in it. Hittite sources no doubt would have mentioned a profitable maritime trade and sought control over it, but from Late Bronze Age texts it would seem that at least the Hittite king and army avoided or rather lacked the means of sea travel. With the exception of Hittite-manned ships, probably loaned from Ugarit, on a mission to Cyprus at the very end of the Late Bronze Age (Güterbock 1967), there is no evidence to suggest the existence of a Hittite fleet in the Mediterranean and especially not the Black Sea. Only riverine boat traffic is mentioned in Hittite texts (e.g. KUB 31.79: Hoffner 2009: 81-4). This does not mean, of course, that there was no sea-based travel along the shores of the Black Sea in this period, but that it was probably of a more local and small-scale nature. To sum up, Middle Bronze Age textual sources offer little in the form of detailed information about Black Sea societies and their relationships with the polities and economic networks centred on the central Anatolian plateau. A much larger corpus of texts is available for the Late Bronze Age, the central focus of which is the Hittite-Kaska conflict. This conflict, however, appears not to have involved the coast and Hittite texts show little knowledge or concern with the Black Sea beyond the founding myths which feature the city of Zalpa in the role of ‘the other’ against which Nesite/Hittite identity is defined. The Black Sea, its resources and potential maritime connections played an insignificant part in the textually recorded concerns of the Hittite state. What they do hint at, however, are a series of unofficial forms of sustained or habitual personal and inter-cultural forms of interaction between Hittite and Kaska individuals and groups, which would have been mediated by, and resulted in, a range of new and hybrid material culture traits and associated practices.
7.4 Settlement Trends in the Turkish Central Black Sea Region Survey projects across the Turkish central Black Sea region, but particularly in its eastern half, have identified an attrition of settlement with recognisable north-central Anatolian ceramic traits sometime during the Middle Bronze Age in what appears to be a retreat of settled communities with central Anatolian cultural affinities to the south of the Devrez-Kızılırmak line. There is widespread evidence for transitional Early-Middle Bronze Age settlement in Samsun and Sinop, but with the exception of Oymaağaç Höyük in inner Samsun, there is little to suggest a Late Bronze Age presence in the region. İkiztepe, Dündartepe, Tekkeköy and Kavak and about 30 additional surveyed sites attest to a relatively densely settled transitional and early Middle Bronze Age landscape in
200
Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region
Samsun (Dönmez and Yurtsever Bayazıt 2008: 103-4). In Sinop, sites with Middle Bronze Age pottery have also been recorded (Işın 1998: pl. 10; Doonan 2004: 52 and fig. 3.1), but again none have yielded north-central Anatolian-style pottery indicative of a Late Bronze Age occupation. Bilgi (2001: nos. 76-7) proposed a Late Bronze Age date for two unprovenanced lugged axes in Sinop Museum. A Late Bronze Age - Early Iron Age occupation has been suggested for Gerze and Sinop Bus Station (Doonan 2004: 52; Bauer 2006: 236) on the basis of rope-impressed pottery and its affinity with north-Pontic steppe traditions, but these connections and their proposed dates await confirmation through more extensive excavation. To the west, extensive survey and excavations in Zonguldak province yielded no indications of a second millennium BC presence (Efe 2004; Karauğuz and Düring 2009). Surveys and excavations further inland suggest more widespread patterns of later Middle and Late Bronze Age settlement. Since 2004, excavations at the inland site of Oymaağaç Höyük (Czichon 2012) have begun to shed light on a substantial, but seemingly isolated (Czichon et al. 2006) Late Bronze Age settlement of distinctively central Anatolian character, probably the Nerik of Hittite texts (Forlanini 1977: 207; Macqueen 1980: 81). Over 50 sites with traces of occupation in the earlier part of the second millennium BC have been recorded in Amasya province, 11 with signs of occupation during the Late Bronze Age (Dönmez 2002; Dönmez and Yurtsever Bayazıt 2008). Eight sites with pottery dating to the first part of the second millennium BC ceramic sequence were recorded during survey in inner Kastamonu (Marro et al. 1996; 1998; Kuzucuoğlu et al. 1997; Marro 2000). Pottery typical of the middle part of the sequence is much rarer and found at three sites only: Höyükdoruğu, Kayabaşı, Üyüktepe and possibly Gâvurevleri (Marro et al. 1996; Kuzucuoglu et al. 1997). By contrast, Çankırı province to the west experienced an increase in settlement sites during the second millennium BC from 14 dating to the early part of the sequence to 21 in the Late Bronze Age. Of these, six sites show evidence for substantial defensive architecture and, with one exception, their linear spatial arrangement follows the main east-west communication route of the Devrez valley. Small and medium-sized settlements are located to the south of this defensive line (Glatz et al. 2009).
7.5 Middle and Late Bronze Age Cide in Context The finds from Cide, on the one hand, fit well within this cross-regional pattern of widespread activity and settlement in the Turkish central Black Sea region during the early part of the second millennium BC and the participation of its inhabitants in a wider Anatolian cultural, and presumably also economic, sphere of interaction. More surprising, on the other hand, is the surface cluster at Okçular, which to date is the only convincing evidence for Late Bronze Age activity. This pottery shows clear typological links with the central Anatolian ceramic tradition and thus, provides evidence for some form of contact and interaction with Hittite communities to the south.
Middle and Late Bronze Age Cide in Context
201
It is of course arguable whether or not an ephemeral scatter of rim pieces constitutes a site, and if so, what type of site. What we can say is that the range of pottery types represented, which include vessels for the preparation and consumption of food as well as storage, are compatible with a settlement context. The clustering of the diagnostic shapes in the middle of the ceramic sequence, the 15th century BC, ties the Okçular material in with four roughly contemporary sites in inner Kastamonu and several more in Çankırı province, including the site of İnandıktepe (Mielke 2006 for a revision of the traditional dating of the site; Özgüç 1988). In historical terms, a 15th century BC date situates the Okçular cluster roughly in the period during which the Hittite-Kaska conflict appears to have flared up for the first time and important Hittite settlements and cult centres are reported to have been destroyed and looted and northern territories lost to the Kaska (Klinger 2005; von Schuler 1965 suggests an earlier date). On a very general level, the events described in Hittite prayers and other sources, may explain the lack of 14th and 13th century BC settlement in the wider region and in Cide and Kastamonu more particularly, but this would both overstretch the available archaeological evidence and unduly restrict interpretive possibilities based on equally scant textual information. As outlined above, the formal types represented at Okçular, while suggesting a date around the 15th century BC, continue to be produced throughout the Late Bronze Age and a later date cannot, therefore, be discounted. The inscribed metal bowl from Kınık, for instance, would suggest some activity in the region in the last centuries of the Late Bronze Age. What this means with regards to the identity of Okçular’s Late Bronze Age inhabitants or occupants is difficult to say without more data. Petrographic analysis of a sample of the pottery from Okçular (section 16.3) indicates that the plain vessels, which are typical of the central Anatolian Late Bronze Age, were probably produced using local raw materials. This means that we are either dealing with a local potter, who was most likely trained in the use of the potter’s wheel and the production of typical north-central Anatolian-style pottery, or a resident or itinerant central Anatolian specialist, who produced either for a local population interested in the consumption of material culture associated with Hittite urban life, a resident Hittite population, or a combination of the two. Okçular Kale and its surroundings are certainly strategically placed to exert visual control over the Cide coastline, to tap into route networks leading into the interior and to access good agricultural land (chapter 14). The location of the cluster on the slope of a natural hill also echoes 16th and 15th century BC central Anatolian trends of establishing new settlements away from traditional centres, mounds in the more southerly regions, and perhaps Okçular Kale in our context. Regardless of how we want to interpret the material from Okçular, it seems very likely that it is not an entirely isolated phenomenon and that more Late Bronze Age settlements are probably waiting to be found in Cide and the central Black Sea region more generally, but for a range of reasons, which I will return to now, have not been identified to date.
202
Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region
7.6 Filling Empty Landscapes Extant survey data, as I have outlined earlier, currently suggests a rather empty Late Bronze Age archaeological landscape and two main hypotheses have been put forward to explain this apparent absence of Late Bronze Age settlements in the Turkish central Black Sea region. Due to the seemingly sudden appearance of the Kaska on the Hittite historical stage in the second half of the 15th century BC, it has been widely assumed that they are newcomers to the region, who, upon arrival, began to displace communities with central Anatolian cultural connections (e.g. von Schuler 1976-80: 451; Klinger 2005: 451). In archaeological terms, this would mean the sudden appearance of new cultural elements, which we have not been able to identify so far. Others see the Kaska as autochthonous to the region. Based on prevalent suffixes in Kaska place, personal and divine names, Singer (2007 following Giorgadze 2000) suggested that they constitute the mobile elements of Hattian society, which were pressured into the mountainous margins of central Anatolia by increasingly dominant Hittite populations in the course of the second millennium BC. Zimansky (2007), using Lattimore’s (1940; 1962) frontier model, moves beyond questions of ethnic or cultural difference and suggests that both the northern frontier of the Hittite empire and, at least to some degree, the Kaska as the ‘barbarous’ northern enemy are the products of Hittite imperialism. With the Pontic mountains a zone of greatly diminishing returns for Hittite military ventures and governance, a frontier and the enemies beyond were created to limit Hittite expansion more so than keep the Kaska out, much like Hadrian’s Wall marked the northern limit of the Roman empire. Whether Hattian or not, an autochthonous Anatolian population, which comes into historical focus when hostile conflict with its neighbours flares up or increases, can be expected to show some degree of cultural continuity as well as affinity with neighbouring regions. This scenario would fit with recent suggestions that handmade pottery normally associated with Early Bronze Age horizons, may have continued to be produced in Anatolia throughout the second millennium BC. This is based on the observation that handmade Early Iron Age (1200/1180-800 BC) vessels found at Boğazöky and other central Anatolian sites often resonate with Early Bronze Age traditions (Genz 2004: 37; see also chapter 8). Tentative corroborative evidence for the continuity of aspects of Early Bronze Age material culture comes from the so-called Quadratgebäude located in the Upper City of Boğazöky-Hattusa (Seeher 2004: 65). Among the finds from the building are bronze weapons, oyster and marine snail shells, as well as a handmade spouted jug reminiscent of vessels produced in northwest central Anatolia during the Early Bronze Age. Seeher (2010: 222) concludes that: “there must therefore have been regions where the Hittite influence on the material culture was so slight that ancient traditions were able to survive for many centuries.” If correct, this scenario of cultural continuity would make an already patchy and disputed ceramic sequence even less reliable, and the distinction between
Filling Empty Landscapes
203
Early Bronze Age and later material in surface assemblages difficult at best. It would, however, potentially fill an empty Late Bronze Age archaeological landscape north of the Devrez-Kızılırmak line. Schoop (2011: 266) has expressed doubts about this hypothesis on the basis that centuries of co-existence between Hittite and Kaska populations, not all of which was hostile (von Schuler 1965, 63-4, 73-4, 81-2; Glatz and Matthews 2005), would have resulted in the adoption by the Kaska of central Anatolian pottery types and production techniques. There is no doubt that continued interaction across cultural and socio-political boundaries produces hybrid cultural traditions (e.g. van Dommelen 2006: 108; 118-20). Some degree of adoption and adaptation of northcentral Anatolian cultural elements by some Kaska – in particular by those in regular contact with Hittite groups – is, therefore, highly likely. The wholesale adoption of the potter’s wheel as a superior technology and Hittite pottery as the desired material culture of a more powerful socio-political entity, however, do not happen as a matter of course (e.g. Dietler 2010). Cultural hybridisation, moreover, works both ways, or rather it creates new cultural traditions, which both Hittite and Kaska groups in the contact zone would have engaged in. Beyond regions of daily interaction, however, a continuity of earlier traditions is entirely possible, as is the deliberate rejection of imperial culture as a form of resistance (see Glatz 2013). A third possibility is the existence and long-term continuity of an entirely local cultural tradition or – encouraged by the rugged topography of the region – many different localised traditions. In this case, we would be looking for material culture, which does not fit any standard period repertoires, but has links with earlier local cultural or technological features. Among the CAP survey material, one group of pottery stands out in this regard. This is the highly localised, coarse, handmade tradition, which I here refer to as ‘Okçular ware’, and which shares close technological links with local Chalcolithic pottery and a Middle Bronze Age cooking pot. It is, therefore, at least a plausible candidate for the material culture of a community which the Hittites would have described as Kaska, or another, historically obscure people of the Pontic mountains. While contemplating what we ought to be looking for to shed new light onto the Late Bronze Age past of the Turkish central Black Sea region, the how deserves equal consideration. Whether newcomers or indigenous to the region, the little that we know about Kaska economic organisation would point towards a combination of sedentary farmers and pastoralists with an element of mobility such as seasonal transhumance (Yakar 2000: 295-300; Glatz and Matthews 2005). It is not possible from textual sources to reconstruct the size and nature of Kaska settlements, but a seemingly low degree of social centralisation and the ease with which whole communities disappear from the reach of approaching Hittite armies, point towards small to very small settlements. Settlement sizes across northern Anatolia tend to be much smaller during the Bronze Age than in the south. In Çankırı province, for instance, Middle and Late Bronze Age sites range between 0.5 and 1.5ha in size (e.g. Glatz et al. 2009) with regional centres
204
Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region
around 6ha (Glatz 2009: 134). Along the mountainous coastal zone of the Turkish central Black Sea region, we are, thus, likely looking for sites below 1ha for the smallscale and semi-mobile communities of the second millennium BC. The absence of mounded sites in Cide and neighbouring regions has already been touched upon. Very small, flat sites are routinely missed by surveys relying on extensive approaches (chapter 4), which account for the majority of survey strategies employed in Turkey. Add to this the potential longevity of local ceramic traditions and the possibility of highly localised pottery types, it is not surprising that so few Middle and Late Bronze Age sites have been recorded in the region to date. A lack of sufficiently intensive field research rather than the absence of evidence would appear to be at the heart of the problem. All the surface material that is presented in this chapter was recorded during intensive unit-walking or subsequent grid-square collections, and would not have been identified by a more extensive survey strategy. The CAP team undertook several visits to Okçular Kale and surrounding fields prior to intensive investigations, and no Middle or Late Bronze Age pottery was recorded during these random, extensive walkovers. It is, therefore, likely that additional Middle and Late Bronze Age sites are waiting to be discovered in the region by future intensive work.
7.7 Conclusions A decade ago, Dönmez (2002: 247) proposed a 1000-year absence of settlement occupation in the Turkish central Black Sea region’s coastal zone. The finds presented in this chapter (and the following chapter) would suggest otherwise. A two-stage intensive survey approach in the surroundings of Okçular Kale yielded a small but diagnostic sample of Middle and Late Bronze Age pottery, with parallels in the ceramic repertoires of Boğazköy-Hattusa and other north-central Anatolian sites. These finds indicate some form of contact between the Turkish west-central Black Sea coast and the Anatolian plateau during the Late Bronze Age. With absolute finds numbers exceedingly small, all interpretations of the nature of this interaction and the type and function of the Okçular material has to be tentative. The range of pottery vessels represented and their strong affinity with central Anatolia, however, may suggest ties that are more direct, intensive and perhaps also long-term, than the conflict-driven scenarios implied by Hittite textual sources. The existence of a highly localised, and perhaps contemporary, pottery tradition at Okçular, provides an avenue for future enquiry into local communities that together with more intensive survey strategies may help fill hitherto empty archaeological landscapes and provide a more balanced perspective on the region’s relationship with the Hittite south. Thus, while the new finds from Cide tentatively challenge the standard narrative of settlement retreat in the Turkish central Black Sea region and open up a series of tantalising possibilities for the role of this area in the events and processes unfolding
Catalogue of Illustrated Artefacts
205
in Anatolia during this period, the ephemeral nature of our evidence means that there are yet more questions to grapple with than answers to offer.
Catalogue of Illustrated Artefacts C0893. G1/A29. Rim fragment of a bowl. Red coarse fabric with dense calcite inclusions. Surfaces smoothed. Oxidised firing. Exterior surface 2.5 YR 6/8, interior surface 2.5 YR 5/8, fabric 5 YR 7/8. D= unknown. C0961. S67. Rim fragment of a jar. Dark reddish brown fine fabric with sparse, very small calcareous inclusions. Surfaces smoothed. Reduced firing. Wheelmade. Fabric 5 YR 3/2. D= 20cm. C1088. U554. Rim fragment of a jar. Buff very coarse fabric with dense small to large angular mineral inclusions. Differential firing. Handmade. Fabric 2.5 YR 8/4, surfaces 10 YR 5/8. D= unknown. C1089. U683. Rim fragment of a carinated bowl with everted rim. Orange brown medium coarse fabric with dense medium grit and possibly some shell and grog. Differential firing. Probably wheelmade but surfaces abraded. Fabric 5 YR 7/6, surface 7.5 YR 7/4. D= 30cm. C2251. U683. Rim fragment of a bowl with inverted rim. Pink fine, powdery fabric with small to medium sparse black grit and calcareous inclusions, some silver mica. Oxidised firing. Surfaces abraded. Probably wheelmade. Fabric 2.5 YR 7/6. D= unknown. C2276. G1/D3. Rim fragment of a bowl. Reddish orange coarse fabric with dense white angular mineral inclusions (calcite). Oxidised firing. Handmade. Fabric 10 R 5/8-6/8. D= 18cm. C2509. G1/A10. Rim fragment of a large jar. Reddish brown to dark grey very coarse fabric with dense medium to large angular and sub-angular calcite, shell (?) on surface. Surfaces abraded. Reduced firing. Handmade. Surfaces 10 R 5/8. D= unknown. C2851?. U646. Body fragment and joining piece of cooking pot with parallel incised pattern on raised band. Reddish to dark brown coarse fabric with dense small to large angular and sub-angular calcite inclusions. Exterior surface smoothed. Handmade. Exterior surface 2.5 YR 5/6, interior surface dark brown almost black. D= unknown. C2969. S93. Rim fragment of jar with constricted neck. Dark brown medium fine fabric with small sparse black and white grit. Exterior smoothed, some burnishing. Oxidised firing. Wheelmade. Surfaces 5 YR 6/6, fabric 5 YR 5/4. D= 6cm. C3051. S88. Pointed horizontal handle with faceted profile. Reddish brown medium fine fabric with small sparse white grit, some pebbles. Interior (vessel) slipped and burnished, exterior smoothed. Differential firing. Handmade. Exterior surface 2.5 YR 4/6, interior surface 10 R 6/6, fabric 2.5 YR 5/8. C3056. S88. Rim fragment of red-slipped and burnished jar with everted rim. Dark red medium fine fabric with low red, black and white mineral inclusions. Exterior and interior slipped (?) and burnished. Oxidised firing. Wheelmade (?). Fabric 6 YR 4/6. D= unknown. C3057. S88. Rim fragment of bowl. Light red to dark brown fine fabric with small sparse white and dark grit. Interior traces of burnishing, exterior abraded. Reduced firing. Wheelmade (?). Fabric 5 YR 6/4, surfaces 5 YR 5/2. D= unknown. C3254. S88. Fragment of a spout. Reddish brown very fine fabric with no visible inclusions. Surfaces burnished. Oxidised firing. Handmade. Surfaces 5 YR 4/3, fabric 5 YR 6/3. C3948. G8/L5. Fragment of a flat base. Reddish orange to brown coarse fabric with dense small to medium angular and sub-angular white mineral (calcite). Reduced firing. Handmade. Exterior surface 2.5 YR 6/6, interior surface 5 YR 7/6. D= unknown. C3949. G8/L5. Rim fragment of a jar. Yellowish orange medium fabric with dense small to medium grog (?). Oxidised firing. Wheelmade. Fabric 7.5 YR 7/6. D= unknown.
206
Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region
C4052. G8/L6. Rim fragment of a bowl with inverted rim. Pale brown medium fine fabric with fine calcareous specs, quartz and grit and a few medium-sized inclusions. Oxidised firing. Wheelmade. Rim GLEY 1 6/5 GY, fabric 5 YR 6/6. D= 16cm. C4185. G6/I6. Rim fragment of jar. Reddish yellow medium fabric with fine to medium calcareous specks, grog and dark grits. Differential firing. Wheelmade. Fabric 2.5 YR 8/1, surfaces 5 YR 7/6. D= 45cm. C4186. G8/I6. Rim fragment of a carinated bowl with everted rim. Reddish yellow medium fabric with calcareous specs, grog and dark grit. Differential firing. Wheelmade. Fabric GLEY 1 6/5G, surfaces 2.5 YR 7/6. D= 12cm. C4235. S188. Rim fragment of large bowl with everted rim and horizontal handle. Dark brown fabric with dense medium dark brown/black and dark red grit and small rounded stones, sparse medium calcareous specs and sparse small to medium angular volcanic sand. Surfaces burnished. Oxidised firing. Handmade, wheel-finished (?). Fabric 2.5 YR 4/8, surfaces 2.5 YR 4/6 and 3/3. D= 22cm.
Bibliography Greco-Roman source Strabo, Geographica. [Translated by H.L. Jones]. Cambridge, MA., Harvard University Press, 1917-1932. published sources Alkım, U. B. 1973. Tilmen Hüyük and the Samsun region. Anatolian Studies 23: 62-5. — 1977. Ikiztepe 1976. Anatolian Studies 27: 41-4. Alkım, U. B., Alkım H. and Bilgi, Ö. 1988. İkiztepe I. Birinci ve İkinci Dönem Kazıları. Ankara, Türk Tarih Kurumu. Alp, S. 1991. Hethitische Briefe aus Maşat-Höyük. Ankara, Türk Tarih Kurumu. Althanassov, B., Krauß, R., and Slavčev, V. 2012. A bronze sword of the Aegean-Anatolian type in the museum of Varna, Bulgaria. In B. Horejs and P. Pavúk, eds., Aegean and Balkan Prehistory: http://www.aegeobalkanprehistory.net/article.php?id_art=20 (last accessed 22 Jan. 2013). Barjamovic, G. 2008. The geography of trade. Assyrian colonies in Anatolia c. 1975-1725 and a study of early interregional networks of exchange. In J.G. Dercksen, ed., Anatolia and the Jazira in the Old Assyrian Period. Leiden, Nederlands Instuut voor het Nabije Oosten: 87-100. — 2011. A Historical Geography of Anatolia in the Old Assyrian Colony Period. Copenhagen, Museum Tusculanem Press. Bauer, A. A. 2006. Between the steppe and the sown: Prehistoric Sinop and inter-regional interaction along the Black Sea coast. In D.L. Peterson, L.M. Popova, and A.T. Smith, eds., Beyond the Steppe and the Sown. Proceedings of the 2002 University of Chicago Conference on Eurasian Archaeology. Leiden, Brill: 225-46. Bilgen, A. N. 1999. Kastamonu Kınık kazısı 1994–1995 metalürjik buluntuları. Anadolu Araştırmaları 15: 269-93. Bilgi, Ö. 2001. Orta karadeniz bölgesi Protohistorik çağ maden sanatının kökeni ve gelişimi. Belleten 65: 1-35. Bottema, S., and Woldring, H. 1990. Anthropogenic indicators in the pollen record of the eastern Mediterranean. In S. Bottema, G. Entjes-Nieborg and W. van Zeist, eds., Man’s Role in the Shaping of the Eastern Mediterranean Landscape. Rotterdam, Balkema: 231-64. Bottema, S., Woldring, H., and Aytuğ, B. 1993-4. Late Quaternary vegetation history of northern Turkey. Palaeohistoria 35-36: 13-72.
Bibliography
207
Bouzek, J. 1985. The Aegean, Anatolia and Europe: Cultural Interrelations in the Second Millennium B.C. Göteborg, P. Åström. Bryce, T. 2009.The Routledge Handbook of the Peoples and Places of Ancient Western Asia: The Near East from the Early Bronze Age to the Fall of the Persian Empire. London, Routledge. Buchholz, H.-G. 1983. Doppeläxte und die Frage der Balkanbeziehungen des ägäischen Kulturkreises. In A.G. Poulter, ed., Ancient Bulgaria. Papers Presented at the International Symposium on the Ancient History and Archaeology of Bulgaria. Nottingham, University of Nottingham: 43-134. Çınaroğlu, A., and Genç, E. 2004. Kastamonu-Kınık 2002 yılı kazısı. Kazı Sonuçları Toplantısı 25/1: 355-66. — 2005. 2003 yılı Kastamonu-Kınık kazısı. Kazı Sonuçları Toplantısı 26/1: 277-90. Czichon, R. M. 2012. Arbeitsbericht der Kampagne 2011: http://nerik.de/artikel/artikel. php?artikelid=148 (last accessed 21 January 2012). Czichon, R. M., Flender, M. and Klinger, J. 2006. Interdisziplinäre Geländebegehungen im Gebiet von Oymaağaç-Verziköprü, Provinz Samsun. Mitteilungen der Deutschen Orient-Gesellschaft 138: 157-76. de Boer, J. G. 2006-7. Phantom Mycenaeans in the Black Sea. TALANTA 38-39: 277-302. Dercksen, J. G. 2001. ‘When we met in Hattuš’. Trade according to Old Assyrian texts from Alishar and Boğazköy. In W.H. van Soldt, ed., Veenhof Anniversary Volume. Leiden, Nederlands Instituut voor het Nabije Oosten: 39-66. Dietler, M. 2010. Archaeologies of Colonialism: Consumption, Entanglement, and Violence in Ancient Mediterranean France. Berkely, University of California Press. Dönmez, Ş. 2002. The 2nd millennium BC settlements in Samsun and Amasya provinces, central Black Sea region, Turkey. Ancient West and East 1(2): 243-293. — 2004. Boyabat-Kovuklukaya: A Bronze Age settlement in the central Black Sea region, Turkey. Ancient Near Eastern Studies 41: 38-82. — 2008. An overview of the 2nd millennium BC cultures of the central Black Sea region in the light of new research. TÜBA-AR 11: 85-104. — 2012. An overview of the 2nd millennium BC and Iron Age cultures of the province of Sinop in the light of new research. In D.K. Tezgör, ed., Sinope: The Results of Fifteen Years of Research. Leiden, Brill: 153-74. Dönmez, Ş., and Yurtsever Beyazıt, A. 2008. A general look at the central Black Sea region during the Middle Bronze Age and a new approach to the Zalpa problem in the light of new evidence. In J.G. Dercksen, ed., Anatolia and the Jazira in the Old Assyrian Period. Leiden, Nederlands Instuut voor het Nabije Oosten: 101-136. Doonan, O. P. 2004. Sinop Landscapes: Exploring Connection in a Black Sea Hinterland. Philadelphia, University of Pennsylvania Museum of Archaeology and Anthropology. Dörfler, W., Herking, C., Neef, R., Pastenrak, R., and von den Driesch, A. 2011. Environment and economy in Hittite Anatolia. In H. Genz and D.P. Mielke, eds., Insights into Hittite History and Archaeology. Leuven, Peeters: 99-124. Efe, T. 2004. Yassıkaya, an Early Bronze Age site near Heraclea Pontica (Kdz Ereğli) on the Black Sea Coast. In B. Hänsel and E. Studenikova, eds., Zwischen Karpaten und Ägäis. Neolithikum und ältere Bronzezeit. Gedenkschrift für Němejcová Pavúková. Rahden, Leidorf: 27-38. Emre, K., and Çinaroğlu, A. 1993. A group of metal Hittite vessels from Kınık-Kastamonu. In M.J. Mellink, E. Porada, and T. Özgüç, eds., Aspects of Art and Iconography: Anatolia and its Neighbors: Studies in Honour of Nimet Özgüç. Ankara, Türk Tarih Kurumu: 675-717. Fischer, F. 1963. Die hethitische Keramik von Boğazköy-Hattuša. Berlin, Gebr. Mann. Forlanini, M. 1977. L’Anatolia nordoccidentale nell’impero Eteo. Studi Micenei ed Egeo-Anatolici 18: 197−225.
208
Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region
Forlanini, M. and Marazzi, M. 1986: Atlante Storico del Vicino Oriente Antico. Fascicolo 4.3. Anatolia: L’impero Hittita. Rome, La Sapienza. Forrer, E. 1928−1932. Balâ. Reallexikon der Assyriologie und Vorderasiatischen Archäologie I. Berlin, de Gruyter: 392-3. French, D. H. 1982. Mycenaeans in the Black Sea. Thracia Pontica 1: 19-28. Gates, M.-H. 1997. Archaeology in Turkey. American Journal of Archaeology 101: 241–305. Gelb, I. J. 1955. Inscriptions from Alishar and Vicinity. Chicago, Chicago University Press. Genz, H. 2004. Büyükkaya I. Die Keramik der Eisenzeit. Funde aus den Grabungskampagnen 1993 bis 1998. Mainz am Rhein, von Zabern. Giorgadze, G. 2000. Non Indo-European ethnic groups: The Hattians and the Kaskeans in ancient Anatolia according to Hittite cuneiform texts. Achalziche: 58-61. Glatz, C. 2009. Empire as network: Spheres of material interaction in Late Bronze Age Anatolia. Journal of Anthropological Archaeology 28: 127-141. — 2013 Negotiating empire. A comparative investigation into the responses to Hittite imperialism by the vassal state of Ugarit and the Kaska peoples of the Anatolian Black Sea region. In G. Areshian, ed., Empires and Complexity: On the Crossroads of Archaeology, History and Anthropology. Los Angeles, Cotsen Institute of Archaeology: 21-56. Glatz, C., and Matthews, R. 2005. Anthropology of a frontier zone: Hittite-Kaska relations in Late Bronze Age north-central Anatolia. Bulletin of the American Schools of Oriental Research 339: 47-65. Glatz, C., Matthews, R., and Schachner, A. 2009. A landscape of conflict and control: Paphlagonia during the second millennium BC. In R. Matthews and C. Glatz, eds., At Empires’ Edge: Project Paphlagonia. Regional Survey in North-Central Turkey. London, British Institute at Ankara: 107-147. Güterbock, H. G. 1961. The north-central area of Hittite Anatolia. Journal of Near Eastern Studies 20: 85-97. — 1967. The Hittite conquest of Cyprus reconsidered. Journal of Near Eastern Studies 26: 73-81. Hansen, C. K. and Postgate, J. N. 2007. Pottery from Level III. In J.N. Postgate and D.C. Thomas, eds., Excavations at Kilise Tepe, 1994-1998: From Bronze Age to Byzantine in Western Cilicia. Cambridge, McDonald Institute: 329-41. Hansen, O. 1994. A Mycenaean sword from Boğazköy-Hattuša found in 1991. Annual of the British School at Athens 89: 213-5. Harris, S. 1992. U.E.A. speleological expedition, Turkey 1992. Norwich, unpublished report, University of East Anglia fell and cave club. — 1993. FACC in Turkish delight. Caves and Caving 62: 24-7. Hawkins, J. D. 1993. A bowl epigraph of the official Taprammi. In M.J. Mellink, E. Porada and T. Özgüç, eds., Aspects of Art and Iconography: Anatolia and its Neighbors. Studies in Honor of Nimet Özgüç. Ankara, Türk Tarih Kurumu: 715-7. Hillier, S. 1991. The Mycenaeans in the Black Sea. In R. Laffineur and L. Basch, eds., THALASSA L’Egée préhistorique et la mer. Liège, Université de Liège: 207-16. Hoffner, H. A. 1974. Alimenta Hethaeorum. Food production in Hittite Asia Minor. New Haven, American Oriental Society. — 2009. Letters from the Hittite kingdom. Atlanta, Society of Biblical Literature. Houwink ten Cate, P. H. J. 1967. Mursili’s north-western campaigns, A commentary. Anatolica 1: 44-61. Işın, M. A. 1998. Sinop region field survey. Anatolia Antiqua 6: 95-139. Karauğuz, G. and Düring. B.S. 2009. A note on the prehistory of the Devrek Region, northern Turkey. Anatolica 35: 153-65. Klinger, J. 2005. Das Korpus der Kaskäer-Texte. Altorientalische Forschungen 32: 347-59.
Bibliography
209
— 2010. Ethnogenèse et identité dans l’Anatolie de la période hittite ou: Qui étaient les Hittites? In I. Klock-Fontanille, S. Biettlot, and K. Meshoub, eds., Identité et altérité culturelles: Le cas des Hittites dans le Proche-Orient ancient. Bruxelles, Safran: 227-40. Kökten, K., Özgüç, N., and Özgüç, T. 1945. 1940 ve 1941 yılında Türk Tarih Kurumu adına yapılan Samsun bölgesi kazıları hakkında ilk kısa rapor. Belleten 9: 361-400. Korfmann, M. 1995. Troia: A residential and trading city at the Dardanelles. In R. Laffineur and W.-D. Niemeier, eds., POLITEIA: Society and State in the Aegean Bronze Age. Liège, Université de Liège: 173-83. Kuzucuoğlu, C., Marro, C., Özdoğan, A. and Tibet, A. 1997. Prospection archéologique franco-turque dans la région de Kastamonu (Mer Noire): Deuxième rapport préliminaire. Anatolia Antiqua 5: 275-306. Laroche, E. 1971. Catalogue des textes hittites. Paris, Klincksieck. Lattimore, O. 1940. Inner Asian frontiers of China. New York, American Geographical Society. — 1962. Studies in Frontier History. London, Oxford University Press. Leshtakov, K. 2007. The eastern Balkans and the Aegean economic system during the LBA. Ox-hide and bun-ingots in Bulgarian Lands. In I. Galanki, H. Thomas, Y. Galanakis, and R. Laffineur, eds., Between the Aegean and the Baltic Seas - Prehistory Across Borders. Liège, Université de Liège: 447-58. Lightfoot, K. G., Martinez, A. 1995. Frontiers and boundaries in archaeological perspective. Annual Review of Anthropology 24: 471-92. Macqueen, J. G. 1980. Nerik and its ‘weather-god’. Anatolian Studies 30: 179-87. Marro, C. 2000. Archaeological survey in the Kastamonu region, Turkey: Remarks on the pre-classical cultural geography of the southern Black Sea. In P. Matthiae, A. Enea, L. Peyronel and F. Pinnock, eds., Proceedings of the First International Congress on the Archaeology of the Ancient Near East. Rome, La Sapienza: 945-65. Marro, C., Özdoğan, A. and Tibet, A. 1996. Prospection archéologique franco-turque dans la région de Kastamonu (Mer Noire). Premier rapport préliminaire. Anatolia Antiqua 4: 273-90. — 1998. Prospection archéologique franco-turque dans la région de Kastamonu (Mer Noire). Troisième rapport préliminaire. Anatolia Antiqua 6: 317-35. Marsh, B., Roberts, N., Toprak, V., Matthews, R., Eastwood, W., Carolan, J., Arcasoy, A. and Lütfi Süzen, M. 2009. Contexts of human interaction: geology, geography, geomorphology and environment. In R. Matthews and C. Glatz, eds., At Empires’ Edge. Project Paphagonia. Regional Survey in North-Central Turkey. London, British Institute at Ankara: 27-73. Matthews, R. 2000. Time with the past in Paphlagonia. In P. Matthiae, A. Enea, L. Peyronel, and F. Pinnock, eds., Proceedings of the First International Congress of the Archaeology of the Ancient Near East. Rome, La Sapienza: 1013–1024. Matthews, R. and Glatz, C. 2009. The historical geography of north-central Anatolia in the Hittite period: Texts and archaeology in concert. Anatolian Studies 59: 51-72. Mellink, M. J. 1993. Archaeology in Anatolia. American Journal of Archaeology 97: 106-113. Mielke, D. P. 2006. İnandıktepe und Sarissa. Ein Beitrag zur Datierung althethitischer Fundkomplexe. In D.P. Mielke, U.-D. Schoop and J. Seeher, eds., Strukturierung und Datierung der hethitischen Archäologie: Voraussetzungen – Probleme – Neue Ansätze. (BYZAS 4). Istanbul, Ege Yayınları: 251-76. Müller-Karpe, A. 1988. Hethitische Töpferei der Oberstadt von Hattuša. Ein Beitrag zur Kenntnis spät-großreichszeitlicher Keramik und Töpferbetriebe unter Zugrundelegung der Grabungsergebnisse von 1978-82 in Boğazköy. Marburg, Hitzeroth. Müller-Karpe, V. 2001. Zur frühhethitischen Kultur im Mündugsgebiet des Maraššantija. In G. Wilhelm, ed., Akten des IV. internationalen Kongresses für Hethitologie. (Studien zu den Boğazköy-Texten 45). Wiesbaden, Harrassowitz: 430-42. Neu, E. 1974. Der Anitta-Text. (Studien zu den Boğazköy-Texten 18). Wiesbaden, Harrassowitz.
210
Beyond the Frontier: The Second Millennium BC in the Cide-Şenpazar Region
Orthmann, W. 1963. Frühe Keramik von Boğazköy-Hattuša aus den Ausgrabungen am Nordwesthang von Büyükkale. Berlin, Gebr. Mann. — 1984. Keramik aus den ältesten Schichten von Büyükkale. In K. Bittel, ed., Boğazköy VI. Funde aus den Grabungen bis 1979. Berlin, Gebr. Mann: 9-62. Otten, H. 1973. Eine althethitische Erzählung um die Stadt Zalpa. (Studien zu den Boğazköy-Texten 17). Wiesbaden, Harrassowitz. Özgüç, T. 1949. 1947 Türk Tarih Kurumu Tarafindan Yapılan Karahöyük Hafrıyatı Raporu. Ankara, Türk Tarih Kurumu. — 1978. Excavations at Maşat Höyük and Investigations in its Vicinity. Ankara, Türk Tarih Kurumu. — 1982. Maşat Höyük II. A Hittite Center Northeast of Boğazköy. Ankara, Türk Tarih Kurumu. — 1988. İnandıktepe. An Important Cult Center in the Old Hittite Period. Ankara, Türk Tarih Kurumu. Parzinger, H. and Sanz, R. 1992. Die Oberstadt Keramik von Hattuša. Hethitische Keramik aus dem zentralen Tempelviertel: Funde aus den Grabungen 1982-1987. Berlin, Gebr. Mann. Pavúk, P. 2002. Troia VI and VIIa. The Blegen pottery shapes: Towards a typology. Studia Troica 12: 35-71. Roux, V. 2015. Cultural transmission, migration and plain wheelmade pottery in the Middle Bronze Age II southern Levant. In C. Glatz, ed., The Production, Consumption and Social Significance of Plain Pottery in the Ancient Near East. Walnut Creek, Ca., Left Coast Press. Schoop, U.-D. 2003. Pottery traditions of the later Hittite empire: Problems of definition. In B. Fischer, H. Genz, É. Jean and K. Köroğlu, eds., Identifying Changes: The Transition from Bronze to Iron Ages in Anatolia and its Neighbouring Regions. Istanbul, Türk Eskiçağ Bilimleri Enstitüsü: 167-78. — 2006. Dating the Hittites with statistics. Ten pottery assemblages from Boğazköy-Hattuša. in D.P. Mielke, U.-D. Schoop and J. Seeher, eds., Strukturierung und Datierung der hethitischen Archäologie: Voraussetzungen – Probleme – Neue Ansätze. (BYZAS 4). Istanbul, Ege Yayınları: 215-39. — 2011. Hittite pottery: A summary. In H. Genz and D.P. Mielke, eds., Insights into Hittite History and Archaeology. Leuven, Peeters: 241-73. Seeher, J. 2003. Die Ausgrabungen in Boğazköy-Hattuša 2002. Archäologischer Anzeiger 2003:1-24. — 2004. Die Ausgrabungen in Boğazköy-Hattuša 2003. Archäologischer Anzeiger 2004: 59-76. — 2010. After the empire: Observations on the Early Iron Age in central Anatolia. In I. Singer, ed., Ipamati kistamati pari tumatimis. Luwian and Hittite studies presented to J. David Hawkins on the Occasion of his 70th Birthday. Tel Aviv, Institute of Archaeology: 220-29. Singer, I. 2007. Who were the Kaška? Phasis, Greek and Roman Studies 10(2): 166-81. Taracha, P. 2003. Is Tuthaliya’s sword really Aegean? In H. Hoffner, G.M. Beckman, R.H. Beal, and J.G. McMahon, eds., Hittite Studies in Honour of Harry A. Hoffern Jr. on the Occasion of his 65th Birthday. Winona Lake, Eisenbrauns: 367-77. Tsetskhladze, G.R. 1999. Between west and east: Anatolian roots of local cultures of the Pontus. In G.R. Tsetskhladze, ed., Ancient Greeks West and East. Leiden, Brill: 469-96. Ünal, A. 1999. A Hittite Mycenaean Type B Sword from the vicinity of Kastamonu, northwest Turkey. In H.I.H. Prince Takahito Mikasa, ed., Essays on Ancient Anatolia. Wiesbaden, Harrassowitz: 207-66. van Dommelen, P. 2006. Colonial matters: Material culture and postcolonial theory in colonial situations. In C. Tilley, W. Keane, S. Kuechler, M. Rowlands, and P. Spyer, eds., Handbook of Material Culture. London, Sage: 104-124. van Neer, A., Zohar, I. and Lernau, O. 2005. The emergence of fishing communities in the eastern Mediterranean region: A survey of evidence from pre- and protohistoric periods. Paléorient 31: 131-57. von den Driesch, A. 2009. Tierknochenabfall aus Gebäude E. In S. Arnold, ed., Das hethitische Gebäude E auf der Akropolis von Kuşaklı. Rahden, Leidorf: 143-60.
Bibliography
211
von den Driesch A., and Boessneck, J. 1981. Reste von Haus- und Jagdtieren aus der Unterstadt von Boğazköy-Hattuša. Grabungen 1958-1977. In K. Bittel, ed., Boğazköy-Hattuša. Ergebnisse der Ausgrabungen 11. Berlin, Gebr. Mann: 1-71. von den Driesch A., and Pöllath, N. 2004. Vor- und frühgeschichtliche Nutztierhaltung und Jagdt auf Büyükkaya in Boğazköy-Hattuša, Zentralanatolien. Mainz am Rhein, von Zabern. von Schuler, E. 1965. Die Kaškäer: Ein Beitrag zur Ethnographie des alten Kleinasien. Berlin, de Gruyter. — 1976-80. Kaškäer. Reallexikon der Assyriologie und Vorderasiatischen Archäologie V. Berlin, de Gruyter: 460-63. Vroom, J. 2005. Byzantine to Modern Pottery in the Aegean: 7th to 20th Century. An Introduction and Field Guide. Utrecht, Parnassus. Yakar, J. 2000. Ethnoarchaeology of Anatolia. Rural Socio-Economy in the Bronze and Iron Ages. Tel Aviv, The Institute of Archaeology. Zimansky, P. E. 2007. The Lattimore model and Hatti’s Kaska frontier. In E.C. Stone, ed., Settlement and Society: Essays Dedicated to Robert McCormick Adams. Los Angeles, Cotsen Institute of Archaeology: 157-72.
T. Emre Şerifoğlu
8 The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC) The Iron Age (1200-330 BC) is one of the least understood periods in our survey area, which is located in the northwest corner of the region known as Paphlagonia in Greek and Roman sources.1 The archaeological evidence for the Iron Age of Cide and Şenpazar consists of a small and difficult to interpret assemblage of pottery fragments from caves and surface scatters, whose dating has been largely marred by a lack of strong typological comparanda. Instead, technological and morphological characteristics distinct from securely datable earlier and later ceramic traditions had to be used to assign pottery, with varying degrees of certainty, to the Iron Age. The situation in Cide and Şenpazar is not very different to most other regions of northern Anatolia, whose Iron Age past remains similarly little understood. However, excavations further south at Alişar Höyük, Boğazköy, Gordion, Çadır Höyük and Kerkenes Dağ (von der Osten 1930a; 1937; Summers and Summers 1998; Genz 2001; 2004; 2006a; 2006b; 2011: 332-335; Gorny 2006; Ross 2010; Omura 2011), and regional surveys in Çankırı and Kastamonu (Matthews and Glatz 2009; Johnson 2011) provide us with a general idea of the material culture and chronology of the north-central Anatolian Iron Age and constitute important reference points for the assessment of the Cide and Şenpazar material. The Iron Age coastal cultures of Sinop, Samsun and Amasya have been explored through surveys and excavations at İkiztepe and Oluz Höyük (Alkım et al. 1988, 2003; Işın 1998; Özsait 2002; Dönmez 2010; Doonan 2004; Özsait and Özsait 2010). In the following discussion, I use the common tripartite division of the Iron Age. Following the Gordion chronology, the Early Iron Age spans from 1200 to 950 BC, the Middle Iron Age from 950 to 550 BC and the Late Iron Age from 550 to 330 BC (Voigt 2002; 2005; De Vries et al. 2005). However, it should be noted that these dates are subject to an on-going debate on the absolute chronology of the Anatolian Iron Ages (Kealhofer and Grave 2011: 418-20).
1 Paphlagonia is delimited by the river Halys (Kızılırmak) to the east, Galatia and Phrygia to the south, Bithynia and Mariadi to the west, and the Black Sea to the north during the Roman period, and this broad description can also be used to define the borders of this cultural area in the Iron Age (Strab. 12. 3. 9).
North and North-Central Anatolia in the Iron Age – The Historical Perspective
213
8.1 North and North-Central Anatolia in the Iron Age – The Historical Perspective After the collapse of the Hittite empire at the end of the Late Bronze Age, some scholars see Anatolia descend into a ‘dark age’, which corresponds to what we define as the Early Iron Age (1200-950 BC) (Drews 1993: 3-11; Muscarella 1995: 91-2; Bryce 1998: 374-9). Not much is known about the historical events and political developments that occurred during this phase because no textual sources have been preserved. However, one can speculate that Anatolia was fragmented among local polities and that some of these managed to transform themselves into regional kingdoms during the course of Iron Age. The regions of Cide and Şenpazar are unlikely to have been conquered or controlled by the Hittites and this mountainous coastal area may have already been ruled by local polities during the Late Bronze Age (chapter 7). The collapse of the Hittite empire must have had some impact on the area as political and military pressure coming from the south came to an end. However, with few sites excavated in north-central Anatolia that span this transitional phase, we know as yet very little of the Early Iron Age in the region (Genz 2003; Kealhofer et al. 2009: 275-80). The Middle Iron Age (950-550 BC) is marked by the formation of the Phrygian kingdom in northwest Anatolia and the beginning of Greek colonisation in the Black Sea, which gained momentum towards the end of the period and resulted in the establishment of colonies all along the Anatolian Black Sea coast. Although the origins of the Phrygians are still debated, there is a general consensus that they came to westcentral Anatolia during the Early Iron Age and settled along the banks of the Sakarya (the Classical Sangarios) river (Barnett 1975: 417-23; Carrington 1977; Tsetskhladze 2005: 212; Bryce 2009: l-lI). Shortly after their arrival, these groups managed to unite politically and establish a polity centred around the town of Gordion (Roller 2011: 562-3). The rapid development and political and military expansion of this polity fundamentally shapes the Anatolian Middle Iron Age. Phrygia became a regional power, controlling most of central and western Anatolia during the eighth century BC (Muscarella 1995; Voigt and Henrickson 2000). In the late-eighth century BC, during the reign of King Midas, the most famous ruler of this kingdom whose name lives on in Greek mythology, the kingdom was strong enough to fight the Neo-Assyrian empire for control over Cilicia (Akurgal 1962: 372-3; Barnett 1975: 428; Prag 1989; Mellink 1991: 622-4; Vassileva 1997; Roller 2011: 563). However, even at this point in history, when Phrygia controlled large parts of the region of Paphlagonia, it is highly unlikely that their kingdom included the western and central Pontic mountains and the coast to the north. The Phrygian kingdom could not halt the waves of Cimmerians, who invaded Anatolia in the early seventh century BC (Hellmuth 2008; De Vries 2011a: 53-4; Tsetskhladze 2011: 115-120). As a result of this regional instability and possibly attacks on major Phrygian centres, Phrygia lost control over much of Anatolia and
214
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
shrank to its core territory around Gordion (Grayson 1991: 92; Mellink 1991: 624, 629; Ivantchik 2005: 110-12; Bryce 2009: 166; De Vries 2011b: 13-5). The resulting power vacuum allowed the Lydians, centred on Sardis, to gradually expand their territories eastwards (Barnett 1975: 438-9). Simultaneously, Lydia was in conflict with the Ionians along the Aegean coast (Mellink 1991: 647-8). After Lydia defeated the Cimmerian invaders in the late seventh century BC, during the reign of king Alyattes, Lydia managed to expand into Phrygia and by the beginning of the sixth century BC, it reached as far as the Kızılırmak river, the Classical Halys, which remained its eastern border until 547 BC (Hdt. 1. 28; Harmatta 1996: 181; Mellink 1991: 647-53; Bryce 2009: lI). During the final third of the seventh century BC, Ionians from the western coast of Anatolia began to establish colonies on the Black Sea coast (Tsetskhladze 1994; 2002: 81; 2011: 119; Greaves 2007; Mitchell 2010). It is probably not a coincidence that colonisation began after Lydia became the main power in western Anatolia (Akurgal 1962: 373-5). The vast majority of the Black Sea colonies were founded during the sixth century BC, when north-central Anatolia was a border area first between the Lydians and the Medes and subsequently between the Lydians and the Persians. For the Ionian cities, an expanded Lydian realm limited their possibilities to expand their agricultural hinterlands, which they needed in order to sustain their growing populations (Tsetskhladze 1994; 2010: 199). Although the Phrygian kingdom appears not to have controlled the mountains and the coasts of north-central Anatolia, its collapse removed all potential military and political pressures on the region. For Lydia, centred on the Aegean, Cide and Şenpazar must have been extremely peripheral and of little interest. This very remoteness made the area a suitable target for Ionian colonisation. Colonial settlements were typically founded on sheltered locations, which could be easily defended against attacks from land and sea, were located close to fertile agricultural lands, and provided a base from whence expeditions to gather natural resources could be mounted (Carpenter 1948; Graham 1971; Butyagin 2007: 8-11). Greek political and cultural influence in the region appears to have gradually increased. However, a cultural ‘Hellenisation’ of the coastal areas did not occur prior to the mid-late first millennium BC. The Greek colonists of the Iron Age appear to have adapted to local circumstances and became an integral part of the local economy (Tsetskhladze 2002). With the exception of a small number of diagnostic Archaic Greek pottery types, there is little that would allow the definition of a distinctive Greek Black Sea cultural tradition and, as a result, the identification of early colonial populations until the Late Iron Age. By contrast, Greek mythology provides considerable detail about the establishment of colonies (Braund 2002). One interesting example is the story of Herakles expelling the Amazons from Sinop in order to settle Greeks there (Tabula Albana FGH 40 F.1a; Braund 2010: 18-9). This city also had a foundation myth about a nymph named Sinope, who gave it her name (Hind 1988; Braund 2002: 74). However,
North and North-Central Anatolia in the Iron Age – The Historical Perspective
215
Strabo (Strab. 12. 3. 11) claims that the people of Sinop thought of Autolycus as its founder, and that the Milesians settled there much later. The Milesian colony at Sinop, which can be safely identified with modern Sinop, was one of the first Ionian colonies founded on the Turkish central Black Sea coast (Drews 1976: 18; Doonan 2004: 69-92). As a major commercial centre, Sinop subsequently also founded its own colonies along the Black Sea coast (Bryce 2009: 655; Tsetskhladze 2009). Kytoros, which can be located at Gideros harbour to the west of modern Cide, was one of Sinop’s new foundations (Avram et al. 2004: 955, 959; Doonan 2006; Mitchell 2010: 93) and trading centre (emporion) (Strab. 12. 3. 10). Others have suggested that it was founded directly by Milesians (Burstein 1976: 12-5, 76, 79). Kytoros was famous for its boxwood, which is mentioned in various ancient Greek sources (Strab. 12. 3. 10, Apollon 2. 942; McGing 1986: 1; Ersal 1994: 28; Avram et al. 2004: 959; Şenol 2006: 40; Arslan 2007: 52) and in the following Hellenistic period, the high-quality boxwood tools and furniture from this area became an important trade commodity (Hannestad 2007: 96-7; see also chapter 9). The copper sources of the Pontic mountains may also have been exploited by Greek colonists (De Jesus 1978: 98; Koçak 2006: 44-5, 60). During the Late Iron Age (550-330 BC), much of Anatolia was under Achaemenid control, which ended with the conquests of Alexander the Great. Northern Anatolia became a part of the Achaemenid empire after Croesus of Lydia was defeated by Cyrus in 547 BC. The Lydian kingdom was then turned into two satrapies: one with Sardis and the other with Daskyleion as their capitals (Dandamaev 1989: 20-27; Briant 2002: 36-7). The parts of Paphlagonia, which were under Persian control, were incorporated into the ‘major’ satrapies of Phrygia and Cappadocia (Cook 1983: 78; Brosius 2010: 29, n.3). Quintus Curtius Rufus mentions the existence of a ‘minor’ satrapy of Paphlagonia, with Ancyra (modern Ankara) at its southern edge, suggesting an administrative re-organisation of Achaemenid Paphlagonia (Curt. Alex. 3. 1. 22). However, Persia most likely exerted relatively little direct political control over the Pontic mountains and the Paphlagonian coast, but instead ruled the area through local dynasts, who were more or less autonomous and mostly loyal to their imperial overlords (Cook 1983: 182, 201; Burtstein 1976: 26-72). It should also be noted that due to its peripheral location, some Paphlagonian rulers could resist imperial orders and demands. One example is Otys, king of the Paphlagonians, who did not heed the order of Artaxerxes II to attack the Spartans headed by Agesilaos that were roaming in the area (Cook 1983: 214). At the same time, there are no reports hinting at any sort of trouble for Achaemenid rule on the Black Sea. Herodotus reports that Paphlagonians of the minor satrapy of Daskyleion (Hellespontine Phrygia, also Xen. Hell. 4. 1. 15) served in the army of Xerxes I in 480 BC in his Greek campaign (Hdt. 7. 72; Cook 1983: 201). Artoxares, a Paphlagonian eunuch, is known to have become a wealthy high official, and part of the private circle of king Artaxerxes (Cook 1983: 129-30, 136). Paphlagonian soldiers fought to take back Lydia from Antigonus, a general of Alexander, even after Darius
216
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
was defeated at Issus (Curt. Alex. 4. 1. 34-35; Briant 2002: 829). Textual sources also indicate that Persian and Paphlagonian nobles intermarried (Sekunda 1991: 141). At the end of the sixth century BC, Persia gained control over the Hellespont and with it, maritime trade between the Mediterranean and the Black Sea (Dandamaev 1989: 157). This might have contributed to the Ionian revolt of the early fifth century BC, as it affected shipments between the Ionian cities and their Black Sea colonies (Barron 1962; Evans 1963; Walinga 1984). Persian control over the Hellespont did not equate, however, with Persian control over the Black Sea coast. In the early fourth century BC, Datame, the satrap of Cappadocia, defeated Thuys, the king of the Paphlagonians, and possibly a cousin of Datame, and expanded his control as far as the Black Sea coast (Cook 1983: 182, 217, 221; Hornblower 1994: 221; Briant 2002: 198-9, 650-51). Datame had coins minted at Sinop following his conquest, which suggests a temporary foothold in the area. He subsequently revolted against Achaemenid rule, but was eventually defeated (Cook 1983: 221; Alram 1986: 110, no. 344; Moysey 1986: 25, pls. 5, 61; Briant 2002: 666). The coins struck by Persian generals at Sinop around 332 BC as well as amphorae bearing sealings with Persian names and Sinopean subjects of Darius, indicate that this part of the Black Sea coast was under firm Achaemenid control (Xen. Anab. 4. 24. 4; Curt. Alex. 6. 5. 6; Akurgal and Budde 1956; Burstein 1976: 70; Hornblower 1994: 221; Briant 2002: 699). Alexander the Great, for instance, did not perceive Sinop as a part of the Greek league (Burstein 1976; Hornblower 1994: 221; Briant 2002: 699). Alexander never conquered Paphlagonia as he turned south to Cilicia from Phrygia (Cook 1983: 202, 226; Hornblower 1994: 805). Even after the battle of Issus the area was partially under Persian control (Curt. Alex. 4. 1. 34-35; Hornblower 1994: 808; Briant 2002: 829, 831). Only with Persia’s defeat at Gaugamela did the Achaemenid political and military hold over the region begin to wane (Cook 1983: 227-8; Dandamaev 1989: 326-7; Hornblower 1994: 860). At the end of the fourth century BC, Kytoros was united with the colonies at Sesamos, Kromna and Tieion to form the polis of Amastris (Strab. 12. 3. 10; Burstein 1976: 76, 79; Otto et al. 1991: 95-6; Cohen 1995: 383; Stolba 2003: 279; Avram et al. 2004: 925-6, 960). This union is a clear indication that Kytoros already had close ties with the other Greek colonies in the area at the end of the Iron Age and that it was important enough to be a peer in this union.
8.2 North and North-Central Anatolia in the Iron Age – The Archaeological Evidence The study of the central and north-central Anatolian Early Iron Age, its material culture and relationship with the preceding period’s cultural traditions is as yet in its infancy. At Boğazköy, wheelmade Late Bronze Age pottery continued to be produced following the abandonment of the city by the Hittite court, but now these vessels were
North and North-Central Anatolia in the Iron Age – The Archaeological Evidence
217
also burnished (Genz 2004; 2005). At Gordion, Late Bronze Age pottery is gradually replaced by a new, undecorated handmade tradition of simple bowls, jars and pots, which are mostly burnished with occasional incisions/impressions (Kealhofer and Grave 2011: 431). Based on formal and technological similarities with Early-Middle Bronze Age pottery, some scholars have proposed the possibility of cultural continuity in marginal regions of the Hittite empire, such as the Pontic mountains, and the re-emergence of these traditions following Hittite collapse (Genz 2004: 49; 2005: 82; Glatz and Matthews 2005; chapter 7). By contrast, surveys and excavations in the Turkish central Black Sea region, to date, have not yielded any distinctive Early Iron Age material. This could be interpreted in a number of ways. The first possibility is that local ceramic traditions existed, but have yet to be recognised; the second is that we are dealing with an aceramic period in this area; and another still is that the region was de facto devoid of settlement during the Early Iron Age (Işın 1998: 110; Dönmez 2003: 214; 2005: 68; Matthews 2009: 152). Whatever the underlying cause of this intriguing pattern, the evidence suggests dramatic cultural changes in the wider region following the end of the Late Bronze Age (Sherratt 1998; Singer 2000; Yakar 2006). These are not restricted to pottery alone, but can be traced also in house architecture at sites such as Boğazköy, Çadır Höyük, Gordion and Kaman-Kalehöyük, where the Early Iron Age is characterised by semi-subterranean structures. At Gordion, the paucity of Early Iron Age pottery has often been related to the migration of the Phrygians into the area (Voigt 1994: 276; Henrickson 1994; Muscarella 1995: 94; Voigt and Henrickson 2000: 46; Voigt 2002: 192). Red wares decorated with triangles, trees, and dots make their appearance at the end of the Early Iron Age at both Çadır Höyük and Boğazköy-Büyükkaya. Ceramics with similar motifs from Oluz Höyük, dated to the end of the Early Iron Age or the start of the Middle Iron Age, may also belong to this tradition (Dönmez 2010: 292, fig. 42-3). A popular type of cooking pot from the later part of the Early Iron Age has distinctive horseshoe shaped handles and strainers (Seeher 2000; Kealhofer and Grave 2011: 427). The assemblages from Boğazköy and Gordion provide us with a good understanding of the Middle Iron Age pottery of north-central Anatolia. In this period, wheelmade pottery becomes common and a great variety of forms were manufactured with a special emphasis on jugs and juglets (Kealhofer and Grave 2011: 420-21; Genz 2011: 346, 349). It seems that some Early Iron Age traditions also survived into this period, but new wares and forms including trefoil jugs, whole-mouth jars with interior rim-ledge, and craters also appear (Henrickson 1994: 111; Bossert 2000; Genz 2004: 48). Painted wares became popular at Gordion (Sams 1994; Kealhofer and Grave 2011: 431-2), Pazarlı (Koşay and Akok 1957), and are also attested at surveyed sites in Tokat and Sivas (Durbin 1971). No painted ceramics were found during surveys in the regions further to the north, however, including the Project Paphlagonia survey,
218
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
where other Middle Iron Age pottery types were found. This might hint at regional cultural differences (Burney 1956; Marro et al. 1998; Matthews 2009: 153). The only exception to this are the painted beak-spouted jugs from Amisos near Samsun, which together with examples of local pottery, can be compared to central Anatolian types (Bossert 2000; Summerer 2005a: figs. 8-10; 2005b). Undecorated north-central Anatolian ceramics are also reported from Sinop (Nohutluk) and found together with early sixth century BC east Greek and Attic vases. These finds are probably best understood as indicators of a trade link with the Kızılırmak basin rather than as a north-central Anatolian cultural sphere extending into this coastal area (Akurgal and Budde 1956: 49, pls. 2-3; Doonan 2004: 4-14, 88-9). The common brown on buff pottery of this period were mostly decorated either with geometric designs or with stylised animals (Kealhofer and Grave 2011: 421). The Alishar IV silhouette style pottery decorated with stylised animals and circles, which appear at the beginning of the Middle Iron Age, can also be linked with this pottery tradition which is found at Alişar Höyük, Çadır Höyük, Gordion and Kaman-Kalehöyük (Akurgal 1955: pls. 1-7; Young 1960: 235, pl. 58, fig. 15; Kenneth 1994; Matsumura 2001; Genz 2001; 2011: 349). Pottery with decorations similar to examples from Late Iron Age levels at Oluz Höyük possibly belong to a later continuation of the same tradition (Dönmez 2010: 290, figs.36-40). The ceramic hallmark of this period is the so-called Phrygian grey ware, which refers to vessels with a fine, mineral tempered fabric and a burnished, soapy textured shiny surface (Summers 1994: 241; Matsunaga and Nakai 2000). The best examples of this pottery come from Gordion, but some types are also found in north-central Anatolia, in particular vessels with crater rims with interior ledges (Sams 1994; Summers 1994; Voigt et al. 1997: 14; Matthews 2009: 153-4). Examples from BoğazköyBüyükkale, Maşat Höyük, Kültepe, and Kaman-Kalehöyük can be dated to the eighth century BC, although they may have appeared slightly earlier at Boğazköy (Genz 2004: table 6). This ware was most common in west-central Anatolia. Sites to the east and south of the Kızılırmak river have yielded relatively few examples (Sams 1994: 177; Matsumura 2000: 121, 2001: 101; Omura 2000: 47, 2001: 48). In areas where grey wares are not common, brown on buff decorated wares were more popular (Kealhofer and Grave 2011: 420-21). In the north, the Delice river is generally thought to delimit the extent of this tradition (Omura 2002: 61), and although grey ware has been found further to the north, finds densities are much lower than in the more southerly regions and close to the Phrygian heartland (Kealhofer 2005). There are no grey ware finds reported from Sinop (although a small number of sherds were identified as Late Phrygian: Işın 1998: 98, 110, table 1). A vessel from Oluz Höyük, probably dating to the very end of the Middle Iron Age, may belong to this ware (Dönmez 2010: 284, fig. 32), as does pottery from Yüklütepe in the Gökırmak valley of Kastamonu (Johnson 2011: 196). The area surrounding Yüklütepe also has many tumuli, which have been assigned to the Phrygian period and tradition. The majority of surveyed sites with evidence for Phrygian grey ware in Kastamonu
North and North-Central Anatolia in the Iron Age – The Archaeological Evidence
219
however, tend to be located near the Araç Çay, to the north of the Ilgaz mountains (Marro et al. 1996: 284; Kuzucuoğlu et al. 1997: 288; Marro et al. 1998: 322; Dönmez 2006: 17). Grey ware pottery is reported from nine sites in Çankırı (Matthews 2009: 153-4) and in the adjacent province of Çorum, grey wares and painted Iron Age pottery increase further in frequency (Yıldırım and Sipahi 2004: 310). The process of consolidation of the Ionian Black Sea colonies, which were first established during the late seventh century BC, seems to have started during the sixth century BC. Amphorae from Heraclea Pontica, and Sinopan amphorae, amphora stamps, and coins dating to subsequent centuries, can be found all along the Black Sea coast and beyond in the Mediterranean, but not in inland Anatolia. Greek colonial settlements in the Black Sea, thus, maintained close maritime contact, but contact and exchange appears not to have significantly penetrated the mountainous hinterland of the Turkish coast (Garlan 2000; Monakhov 2003: 123-60; Tsetskhladze 2009; Finkielsztejn 2011). Native settlements such as Amisos and İkiztepe had closer cultural ties with inland Anatolia (Tsetskhladze 2007: 173, 186-91; Summerer 2007: 29). However, Archaic Greek pottery found at sites in the Kızılırmak bend indicates that there was some interaction between the north-central Anatolian and the Aegean worlds in this period (Summerer 2007). Examples of Greek pottery include a bird-bowl, dating to the mid-seventh century BC, now in Amasya Museum, Milesian and Corinthian pottery from Boğazköy, a fragment of a Milesian vase from Alişar Höyük, and a Protogeometric sherd from Kaman-Kalehöyük. Earlier imports may well have arrived from the Aegean rather than the Black Sea (Boysal 1959; Summerer 2005a: fig. 1-3; Tsetskhladze 2007: 165-73, 2010: 200). Mycenaean pottery found at Boğazköy and Maşat Höyük point to an overland connection in the Late Bronze Age (Bittel 1976: 9-14; Özgüç 1980). At Akalan, which is 18km inland from Amisos, two late seventh century BC Milesian jugs (middle wild goat style II) were found together with local pottery, which are similar to the examples found at Amisos (Macridy 1907: 5; Summerer 2005b: 135). The early east Greek bird-bowl mentioned above was probably part of a funerary assemblage and was found in the village of Dalsaray near Mecitözü, further inland (Summerer 2007: 30, fig. 1). The Milesian and Corinthian pottery from Boğazköy and a Milesian vase fragment from Alişar Höyük are from the mid-seventh century BC. A fragment of an Attic crater from Kaman-Kalehöyük is from the sixth century BC. At Kaman-Kalehöyük, however, much earlier protogeometric sherds are also attested (von der Osten 1937: pls. 3, 9; Bossert 2000: pl. 143, 1351, 1350a-c; Matsumura 2000: fig. 28). Late Archaic style architectural terracottas found at sites along the course of the Kızılırmak also give an idea about the nature of north-central Anatolian-Greek cultural interactions in this period (Summerer 2005b; 2007: 31). The motifs on clay revetment plaques found with roof tiles and simas2 at Akalan include guilloches,
2 These are architectural elements placed at the edges of roofs, and have an upturned shape and act as a gutter.
220
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
meanders and lotus-palmettes similar to the Amisos examples. One example also contained scenes depicting Herakles and Pholos (Macridy 1907; Summerer 2005a: 1-2, fig. 4-6, pl. 69; 2007: fig. 3). Lion simas resembling Milesian examples were both found at Akalan and Köyiçi Tepesi, which is further inland in Ladik district (Summerer 2007: fig. 4; 2008: 269-70). However, Köyiçi Tepesi, where revetment plaques painted with guilloche and lotus motifs similar to those at Akalan were found, also yielded a relief plaque depicting the Anatolian theme of two antithetical wild goats with a tree in between. This is probably a scene adapted from Ionian vase paintings (Summerer 2005b; 2007: 32, fig. 5) and may reflect increasing Anatolian cultural influences as one moves away from the sea. At Pazarlı Tepe, however, only Greek-style roof tiles, including painted terracotta relief plaques, antefixes, spouted eave tiles and square shaped revetment plaques, were encountered (Özsait 2003: 276; Summerer 2005b). This makes Pazarlı, which is located 40km north of Boğazköy, clearly different from the examples mentioned above and more similar to Gordion and Düver (Koşay 1941). The plaques at the site, which also yielded terracotta peg mosaics similar to the ones from the Archaic levels of Gordion, do not depict floral decorations but only warriors with Ionian helmets and the scene with wild goats flanking a tree (Koşay 1941: 14-5, pl. 8, 21-3; Young 1965: 12; Summerer 2007: fig. 6-7). This seems to suggest a cultural watershed between coastal and inland traditions along the Tavşan mountains and with Pazarlı closely associated with the coastal sphere. Kaman-Kalehöyük also yielded painted simas and spouted eave tiles similar to those from Pazarlı, but the Boğazköy roof tiles lack decoration. No tiles were found at Alişar Höyük, Maşat Höyük, Eskiyapar, Kerkenes Dağ and Kuşaklı (Meyer 1940: fig. 18-20; Åkerström 1966: 190-91, fig. 63; Summerer 2007: 33-4). Archaic red slip cover tiles found together with grey ware sherds at Yüklütepe in the Gökırmak valley to the north of the Ilgaz mountains (Johnson 2011: 196) also form part of the north-central Anatolian or Phrygian tile tradition, which reached its maturity in the sixth century BC (Glendinning 2002). This is a hybrid style incorporating Ionian and Anatolian elements. Examples from Gordion, for instance, are decorated with geometric designs, hunting scenes, hybrid creatures and lions. The general absence of such tiles from the region to the north of the Kızılırmak points to limited Phrygian cultural presence or influence. An exception is the Samsun area further north, which constitutes a strategic communication corridor to and from the Black Sea, where Phrygian roof tiles are attested. To sum up, we can recognise an increasing Greek cultural presence in and influence on the material culture of the Turkish Black Sea coast and to a lesser degree also the northern regions of central Anatolia. Greek colonists were interested in acquiring minerals such as sinopis (a type of red ochre) and a yellowish red pigment called sandarak from the Samsun area and the Kızılırmak basin. The resulting trade may in part explain the spread of Greek pottery and architectural features along this part of the coast and up-river (Lepik-Kopaczynska 1958: 86; Mehl 1987: 119-20; Summerer 2007: 35). Archaeological evidence for the Late Iron Age is scarce from northern Anatolia. The increasing level of west Anatolian ceramic influence manifests itself in painted
The Iron Ages in Cide and Şenpazar
221
decorations on white panels, polychrome decorations, red banded wares and distinctive ceramic types found at Gordion, Boğazköy, Çadır Höyük and Kerkenes (Voigt et al. 1997: fig. 30; Genz 2001: fig. 4; Kealhofer et al. 2010: 75-6; Kealhofer and Grave 2011: 421, 427, 432). In Gordion, fine wares reflect a level of Achaemenid influence but the more common utilitarian wares show continuity from the Middle Iron Age (Henrickson 1993: 147; 1994: 113-4; Voigt et al. 1997: 16; Voigt and Young 1999: 192). The most diagnostic pottery type from this period is the Achaemenid-style carinated bowl, which was found in surface assemblages around Gordion and in inner Paphlagonia, and which resembles Achaemenid metal vessels from Hasanlu in western Iran (Henrickson 1994: fig. 10.9b; Dyson 1999: 102, fig. 1; Matthews 2009: 156, fig. 5.14:3, 5.16:9-10). A Late Iron Age painted pottery tradition from İkiztepe appears to be confined to the coastal region (Bilgi 1999). Another popular Late Iron Age pottery group, which appears to have its roots in the Middle Iron Age and may well have evolved from grey ware traditions, consists of black polished (black lustrous) wares. Examples found at Boğazköy, Çadır Höyük, Gordion and Kaman-Kalehöyük were produced in and distributed from several regional production centres (Henrickson et al. 2002; Kealhofer et al. 2008: 202-3, 2009, 2010; Matsumura 2008; Kealhofer and Grave 2011: 426, 428, 432). Vessel surfaces are not burnished or polished but achieve their lustrous finish through a sintered slip. The majority of attested examples are table wares with bowls, goblets, small pots and jugs the most popular forms (Kealhofer et al. 2008: 203). The Achaemenid fortress of Kaletepesi in the Gökırmak valley has yielded a fourth century BC black gloss kantharos and a fine black burnished phiale (Johnson 2011: 197-8). The use of this ware to produce Achaemenid pottery types shows that both Achaemenid and northwestcentral Anatolian ceramic traditions were known north of the Ilgaz mountains. At Gordion, spouted eave tiles became common during this period but no clear parallels are attested from adjacent regions (Glendinning 1996). A piece showing a part of a lion with a curled tail, which was found during the Project Paphlagonia survey, was dated to the Late Iron Age (Matthews 2009: 156, fig. 5.7, 5.14:17).
8.3 The Iron Ages in Cide and Şenpazar Archaeological evidence from Cide and Şenpazar, which can be dated with varying degrees of confidence to the Iron Ages, comes from both cave and surface contexts. Iron Age pottery was found in the Okçular and Abdulkadir valleys, Okçular Kale İni (S24), the Çamdibi area, Kılıçlı Mağarası (S88/S93), Derebağ Köy Mağarası (S188) and the Aybasan area. Most notable in terms of overall patterns is the almost complete absence of Iron Age pottery finds from the coastal area, which may well be due to taphonomic processes. Some geologists, on the basis of data from the northern shores of the Black Sea, have proposed a dramatic sea level rise at the end of the Iron Age that would have submerged many first millennium BC sites (Fedorov 1972; 1977;
222
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
Kondrashov 1995; Porotov 2007; Balabanov 2009). More recent studies, however, suggest that sea levels did not change quite so dramatically overall, but that parts of the Black Sea coast are affected by microtectonics, causing localised coastal changes (Brückner et al. 2010; Fouache et al. 2012; section 2.1.3). Iron Age finds are more readily attested in the hinterland valleys of Çamdibi, Aybasan and the cave of Derebağ Köy Mağarası in the northeast of the survey region. Kılıçlı Mağarası in the Çamdibi area is the only location, which appears to be in use throughout the Iron Age. At all other sites, such as Okçular Kale İni and Derebağ Köy Mağarası, occupation or use appears to have been more punctuated. Before embarking on a more detailed discussion of the Cide and Şenpazar ceramic data, it is worth reiterating here that the Iron Age material collected in the course of the Cide Archaeological Project (CAP) are few in number and very difficult to date with accuracy. As a result, all dates put forward here are to be regarded as tentative and the division into Early, Middle and Late Iron Age as provisional. Much of the dating presented here is based on the generalisation that handmade pottery was produced primarily during the Early Iron Age, and that wheelmade pottery dates to the later phases of the Iron Age. Iron Age dates were also assigned through comparisons with earlier and later pottery traditions from the Cide region and, where possible, published comparanda. 8.3.1 Iron Age I Evidence for the Early Iron Age comes primarily from the cave sites of Derebağ Köy Mağarası (S188) and Okçular Kale İni (S24) (Fig. 8.1). The pottery is coarse, dark purple brown3 and dark brownish red4 in colour (Fig. 8.2). Medium coarse examples have a brownish grey5 colour and were found at Derebağ Köy, and also occur in finer fabrics. Other finer examples, possibly dating to this period are brown in colour6 and were found at the Kılıçlı and Derebağ Köy caves and at Aybasan, a finer orange coloured sherd7 and several reddish brown ceramics8 also come from Derebağ Köy Mağarası. Jars with ledge or simple rounded rims are the most common formal type. Pots usually have an everted rim, which are often flattened on the top, and can be thickened. A carinated bowl shape with an externally thickened rim was also recovered. Derebağ Köy Mağarası (S188) is the only place where all of these formal types were found in the same assemblage. Some of the pottery from Derebağ Köy is very
3 C4155. 4 C2919 (Fig. 8.2), C4197, C4209-4210 (Fig. 8.2), C4212, C4228 (Fig. 8.2), C4229 (Fig. 8.2). 5 C3972, C4170, C4171, C4194, C4196, C4199, C4213. 6 C3611, C3638 (Fig.8.2), C3965, C4031-4032, C4033, C4035 (Fig. 8.2), C4036, C4037, C4165 (Fig. 8.2). 7 C4198. 8 C3971, C4166 (Fig. 8.2), C4168, C4169, C4220 (Fig.8.2).
The Iron Ages in Cide and Şenpazar
0
5
223
10 Km
Derebağ Köy Mağarası
Okçular Kale İni
Aybasan Çamdibi
Fig. 8.1: Map of Cide and Şenpazar showing the locations of Early Iron Age finds. Produced by Toby C. Wilkinson
distinctive and includes a dark brownish horizontal handle with black mineral temper reminiscent of the volcanic sand of the Sinop region. Another brown sherd from Derebağ Köy is a horned handle, with a protruding tip on the carination (Fig. 8.3).9 Although it is hard to pinpoint exact parallels, pots with horned handles are known from Greece and Italy during the Late Bronze Age and the Early Iron Age and our example could point to contact with the west (Hencken 1968: 478, 480, figs. 425-6; Vagnetti 1985: 31-2; Moschos 2009: 380; MacIntoch Turfa 2011: 119; Romanos 2011: 72, 94, 111, 145). Another brown sherd was decorated with grooves on the exterior. The most interesting specimen is an almost complete brownish grey, crude one-handled cup,10 which resembles Early Iron Age pottery from BoğazköyBüyükkale (Genz 2003; 2004). Like the ceramics known from Early Iron Age north-central Anatolia, the Derebağ Köy vessels are simple, handmade, and often burnished but otherwise undecorated. We have no evidence in Cide or Şenpazar for the red painted pottery, which became popular in north-central Anatolia towards the end of the Early Iron Age.
9 C4037 (Fig. 8.2 and 8.3). 10 C3972 (Fig.8.3).
224
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
Fig. 8.2: Early Iron Age pottery from Cide. Produced by Lorraine McEwan.
The Iron Ages in Cide and Şenpazar
225
Fig. 8.3: Early Iron Age cup and handle from Derebağ Köy Mağarası (S188). Produced by Lorraine McEwan.
8.3.2 Iron Age II Most of our Iron Age ceramics seem to belong to the middle phase, and can be differentiated from Early and Late Iron Age ceramics in terms of fabrics, manufacturing technique and firing (Fig. 8.4). Fabric colours and forms are much more varied compared to the Early Iron Age, with grey the dominant fabric colour. With the exception of one fragment from Çamdibi, which has an externally thickened rim that was flattened on the top and slightly sloping down towards the edge, no formal types could be identified (Fig. 8.5).11 Some sherds with grooves are, however, of interest.
11 C3599 (Fig. 8.5).
226
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
0
5
10 Km
Derebağ Köy Mağarası Okçular Kale İni
Çamdibi Abdulkadir Okçular Kılıçlı Mağarası
Fig. 8.4: Map of Cide and Şenpazar showing the locations of Middle Iron Age finds. Produced by Toby C. Wilkinson.
Fig. 8.5: Middle Iron Age pottery from Cide. Produced by Lorraine McEwan.
A fragment of what is probably Phrygian grey ware was found in the Abdulkadir valley.12 The dark brownish grey13 and light reddish brown14 sherds from the Çamdibi area and the Kılıçlı and Derebağ Köy caves appear to be the main coarse ware ceramics
12 C1037. 13 C3060 (Fig. 8.5), C4003, C4022. 14 C3179, C4222, C4223.
The Iron Ages in Cide and Şenpazar
227
from this period. The dark brown grey ceramics may be related to the Early Iron Age brownish grey pottery, although the earlier ceramics only occurred in medium and fine fabrics. One light reddish brown sherd from Derebağ Köy15 appears to have faded traces of red paint on its exterior surface. Fragments of light grey ceramics,16 a handmade pale red pottery17 from Okçular Kale İni, together with examples of orange18 and light orange pottery19 from Kılıçlı Mağarası and the Okçular valley respectively appear to constitute the medium fine fabrics of the period. The medium fine orange ceramics might be a continuation of similar Early Iron Age fabrics, and examples of both occur in Kılıçlı Mağarası. A brownish grey20 and a reddish brown21 fragment from Kılıçlı appear to represent the only fine ware of the Middle Iron Age. However, reddish brown ceramics also occur in medium and coarse fabrics at Kılıçlı,22 and brownish grey medium fine fabrics are attested at Kılıçlı and the Okçular valley.23 A single body fragment of what appears to be Phrygian grey ware is impossible to interpret in cultural terms, but it is most plausibly an import from further south. The red painted piece from Derebağ Köy hypothetically represents another link, perhaps a local version of Middle Iron Age north-central Anatolian painted wares. We recorded no Archaic Greek pottery or Middle Iron Age roof tiles in Cide and Şenpazar. This is remarkable, given that the Greek colonisation of the Black Sea coasts occurs in this period. This may be due to a number of factors. In light of the repeated mentions of anchorages such as Kytoros/Gideros in Greek sources, the most plausible explanations are that, for reasons to do with our survey methods and focus as well as environmental conditions and taphonomy, we did not record any Greek colonial sites. Alternatively, such sites may have used primarily local material culture.
8.3.3 Iron Age III As with previous phases, only a small number of sherds could be assigned to the Late Iron Age (Fig. 8.6), and once again it is hard to talk about popular forms. Pottery shapes we can identify include flat-based jars and deep bowls with simple curved rims (Fig. 8.7).
15 C4223. 16 C2900. 17 C2893. 18 C3191. 19 C1415. 20 C3167. 21 C3182 (Fig. 8.5). 22 C3162, C3169. 23 C931, C3599.
228
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
0
5
10 Km
Cide Okçular Kale İni
Okçular Kılıçlı Mağarası
Fig. 8.6: Map of Cide and Şenpazar showing the locations of Late Iron Age finds. Produced by Toby C. Wilkinson.
Fig. 8.7: Late Iron Age pottery from Cide and Şenpazar. Produced by Lorraine McEwan.
Locales of the Cide Iron Age
229
Fabric colours vary greatly: a reddish brown sherd with a coarse to medium fine fabric from Kılıçlı Mağarası resembles examples from the preceding period.24 Sherds of medium fine fabric occur in greyish brown25 at Kılıçlı, and light-red26 in the vicinity of Cide. The greyish brown ceramics might represent a continuation of the fine brownish grey and even the dark brown grey ceramics of the Middle Iron Age from Kılıçlı. A pale pink sherd27 from Okçular valley and a yellow buff fragment28 from Aybasan have both fine fabrics. The yellow buff sherd is the only potentially Late Iron Age example found in the Şenpazar district. The Late Iron Age ceramics of Cide and Şenpazar, if correctly identified as such, seem to have a completely local character with elements of continuity tentatively recognisable in fabric types and colours. No examples of black polished wares or painted pottery types popular in western Anatolia were found; also not attested are elements of Achaemenid culture such as pottery, architectural remains or rockcut tombs, no doubt indications of the peripheral nature of the region to both the west Anatolian and Achaemenid worlds. More surprising, as mentioned above, is the complete absence of indications of Greek colonial activities in the area.
8.4 Locales of the Cide Iron Age 8.4.1 Gideros The natural harbour of modern Gideros is the most plausible location for the Greek trading post of Kytoros (Avram et al. 2004: 959; Tsetskhladze 2007: 177, note 35). While no unequivocal Iron Age architectural or material remains were found at the cove of Gideros or in its surroundings, the monumental walls, which form the base of the defences of Gideros Kale West (S12) (chapters 11 and 14), are constructed without the use of mortar and may, therefore, date back to the Iron Age (Fig. 8.8).
8.4.2 Iron Age Caves Most of the Iron Age material collected during CAP comes from the cave sites of Okçular Kale İni (S24, Fig. 5.11), Derebağ Köy (S188, Fig. 8.9) and Kılıçlı Mağarası (S88/S93, Fig. 7.4). Okçular Kale İni and Derebağ Köy Mağarası yielded pottery that may be assigned a date in the Early and Middle Iron Ages, whereas the Kılıçlı finds
24 C3172 (Fig. 8.7). 25 C3607 (Fig. 8.7), C3609. 26 C3795. 27 C3908. 28 C3743 (Fig. 8.7).
230
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
Fig. 8.8: Wall built of large dry-stone blocks at the base of Gideros Kale West (S12). (CAP-P1/1092).
include pottery from all three sub-phases. Among these, Derebağ Köy Mağarası, is of particular interest because of the connections some of its material provide with the Early Iron Age assemblage at Boğazköy on the one hand, and, more tentatively, with Greece on the other. The fact that most of our evidence comes from cave sites does not mean, of course, that the Iron Age inhabitants of the area lived in caves. There must have been contemporary settlements in Cide and Şenpazar, which we have not been able to locate during surface survey. Evidence from other caves in Anatolia attest to both habitation and the ritual use of caves by Iron Age communities. For example, the famous Prehistoric cave of Karain, in southern Turkey, has an Iron Age level (Storch et al. 1992; Atici 2009; Kartal 2009: 24). The Köroğlu (Tauraz) caves near Sivas and the
Locales of the Cide Iron Age
231
Fig. 8.9: Derebağ Köy Mağarası (S188) (CAP-P15/300).
İnbükü cave near Samsun were also in use during the Iron Age (von der Osten 1930b: 55-6; Yılmaz 2012: 14-6). Of interest also is the nearby Mızrak Mağarası, in which a Late Bronze Age sword was found (Ünal 1999) and which may have served a ritual purpose. The Early Bronze Age assemblage from Kıllıçlı Mağarası may also point at ritual or other socially significant uses (chapter 6). The ritual use of caves in the wider region may have continued into the Iron Age and inspired the construction of rock-cut monuments and chambers, which were used both for offerings and burials (Fol 2008; Vassileva 2012). The Phrygian goddess Kybele was associated with sacred caves and neighbouring Paphlagonian communities may have also regarded caves as holy places (Cox 1990: 395). According to Xenophon, Herakles entered the netherworld through caves located in the vicinity of Heraclea Pontika, and this might reflect older ideas about caves in the region (Xen. Anab. 6. 2. 2; Asheri and Hoepfner 1972:38-41). Changing climatic conditions during the Iron Age, which resulted in drier and colder winters (Göktürk 2011: 56, 77), may also have resulted in an increased use of caves as habitation sites. On the Crimean peninsula, on the opposite side of the Black Sea, a cave site of the Kizil-Koba culture was inhabited from the ninth until the sixth century BC (Golomshtok 1938: 433-4; Burov 1997: 129-30; Carter et al. 2000: 709, Bilde et al. 2007-2008: 140). Chokurcha is another cave located in this area and was occupied
232
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
during the same period (Golomshtok 1938: 259-61). Location and morphology of the Cide caves, however, would make longer-term habitation rather unlikely (chapters 5 and 6).
8.5 Final Remarks To sum up, the archaeological evidence from Cide and Şenpazar, sparse as it is, suggests that the historically attested events and processes taking place elsewhere in Anatolia during the final centuries of the second and much of the first millennium BC had little effect on this region. The Iron Age inhabitants of our survey area seem to have been effectively beyond the cultural and political spheres of influence of both west and central Anatolian kingdoms and more distant imperial and colonial powers, and retained a strongly local cultural character throughout this period. The Early Iron Age pottery from Derebağ Köy Mağarası points to some form of contact with north-central Anatolia, and perhaps with the eastern Mediterranean, as does a coarse Phrygian grey ware sherd for the Middle Iron Age. Puzzling is the absence of evidence for any Greek colonial presence at Kytoros/Gideros and elsewhere in the region. Our Late Iron Age materials also do not reflect any Greek or Achaemenid presence or influences, despite a rather strong Achaemenid cultural presence just to the south of the Küre mountains in the form of rock monuments and pottery (Summerer and von Kienlin 2010).
Catalogue of Illustrated Artefacts Early Iron Age C2919. S24. Rim. Dark red medium coarse vessel with medium grit and some medium angular quartz. Black slipped and burnished on the exterior and smoothed on the interior. Differential firing. Wheelmade. Exterior surface 5 YR 5/3, interior surface 10 R 4/4. D= 12cm. C3611. S88. Body sherd. Grey fine vessel with sparse, small, rounded white grit. Smoothed on the exterior and light orange brown burnished on the interior. Reduced firing. Handmade. Exterior surface 2.5 YR 6/6, interior surface 2.5 YR 6/8. D= unknown. C3638. S88. Rim. Brown medium fine vessel with dense small to medium well-sorted sub-angular quartz and grit. Smoothed and reddish brown burnished on the exterior. Reduced firing. Handmade (?). Exterior surface 2.5 YR 5/2, interior surface 2.5 YR 5/3. D= unknown. C3965. U2127. Handle. Brown coarse vessel with very dense small to medium sub-angular quartz, and purplish and white grit. No surface treatment. Oxidised firing. Handmade. Exterior surface 5 YR 6/6, interior surface 5 YR 6/6. D= unknown. C3971. S188. Body sherd. Reddish brown coarse vessel with a dark grey core, biscuit fabric, and large lime particles. Smoothed and reddish brown burnished on both sides. Differential firing. Handmade. Exterior surface 2.5 YR 6/8, interior surface 2.5 YR 5/6. D= unknown. C3972. S188. Whole profile (rim, body, base and handle). Brownish grey coarse vessel with a medium hard fabric, some lime and fine grit. Smoothed and brownish grey burnished on both sides. Reduced firing. Handmade. Exterior surface 7.5 YR 7/6, interior surface 7.5 YR 7/6. D= 10cm.
Catalogue of Illustrated Artefacts
233
C4031-4032.S188. Body sherd. Dark reddish brown coarse vessel with a very hard fabric, minute grit and mica/quartz. Smoothed on the exterior. Oxidised firing. Handmade. Exterior surface 5 YR 3/2, interior surface 5 YR 4/4. D= unknown. C4033. S188. Body sherd. Dark yellowish brown coarse vessel with a medium hard fabric, lime and small grits. Burnished on both sides. Differential firing. Handmade. Exterior surface 5 YR 3/1, interior surface 5 YR 5/4. D= unknown. C4035. S188. Rim. Dark brown coarse vessel with dense fine to large grits and fine to medium lime specks. Burnished on both sides. Reduced firing. Handmade. Exterior surface 10 YR 6/4, interior surface 10 YR 6/4. D= 20cm. C4036. S188. Rim. Brown coarse vessel with dense fine to medium lime specks and yellow/red grits. Smoothed and burnished on the exterior. Reduced firing. Handmade. Exterior surface 5 YR 6/6, interior surface 5 YR 5/6. D= unknown. C4037. S188. Handle. Dark Brown very coarse vessel with dense fine to large lime specks and grits in different colours. Burnished on both sides. Reduced firing. Handmade. Exterior surface 5 YR 5/4, interior surface 5 YR 5/4. D= unknown. C4155. S188. Handle. Dark reddish brown very coarse vessel with dense medium to large angular reddish/purplish stones and quartz, smaller calcareous specs. Red slipped (traces), smoothed and burnished on both sides. Oxidised firing. Handmade. Exterior surface 2.5 YR 2.5/2, interior surface 2.5 YR 2.5/2. D= unknown. C4165. S188. Rim. Light brown coarse vessel with a medium hard poorly sorted paste, mineral temper up to 3mm, and some chaff. Burnished on both sides. Reduced firing. Handmade. Exterior surface 7.5 YR 5/6, interior surface 7.5 YR 5/6. D= 18cm. C4166. S188. Rim. Reddish brown medium quality vessel with a poorly sorted medium hard paste and grit up 3 mm. Burnished on the exterior and smoothed on the interior. Reduced firing. Handmade. Exterior surface 10 R 5/8, interior surface 2.5 YR 7/6. D= unknown. C4168. S188. Body sherd. Reddish brown medium quality vessel with a moderately well sorted medium hard paste and minute lime inclusions smaller than 0.5mm. Smoothed and reddish brown burnished on the exterior. Oxidised firing. Handmade. Exterior surface 5 YR 4/6, interior surface 5 YR 6/4. D= unknown. C4169. S188. Body sherd. Reddish brown medium quality vessel with a moderately well sorted medium hard paste and minute lime inclusions smaller than 0.5mm. Smoothed and reddish brown burnished on the exterior. Oxidised firing. Handmade. Exterior surface 5 YR 4/6, interior surface 5 YR 6/4. D= unknown. C4170. S188. Body sherd. Light brown medium quality vessel with a moderately well sorted medium hard paste and minute lime inclusions smaller than 0.5mm. Dark grey burnished on the exterior and smoothed on the interior. Differential firing. Handmade. Exterior surface 5 YR 3/1, interior surface 5 YR 6/8. D= unknown. C4171. S188. Body sherd. Light brown coarse vessel with a poorly sorted medium hard paste, mineral temper up to 3 mm. Smoothed and light brown burnished on the exterior and smoothed on the interior. Differential firing. Handmade. Exterior surface 7.5 YR 3/1, interior surface 10 YR 8/6. D= unknown. C4194. S188. Body sherd. Dark grey coarse vessel with some fine lime specks and dense fine to medium light grit. Brownish red burnished on the exterior and smoothed on the interior. Reduced firing. Handmade. Exterior surface 2.5 YR 5/6, interior surface 2.5 YR 3/1. D= unknown. C4196. S188. Body sherd with lug. Dark brown to dark red coarse vessel with a very gritty fabric, lots of very fine gold mica specks, some fine lime specks and fine to medium reddish grit. Yellowish brown burnished on the exterior and polished on the interior. Reduced firing. Handmade. Exterior surface 7.5 YR 7/4, interior surface 2.5 YR 2.5/2. D= unknown. C4197. S188. Body sherd. Dark brown to dark red coarse vessel with a very gritty fabric, lots of very fine gold mica specks, some fine lime specks and fine to medium reddish grit. Yellowish brown
234
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
burnished on the exterior and polished on the interior. Reduced firing. Handmade. Exterior surface 7.5 YR 7/4, interior surface 2.5 YR 2.5/2. D= unknown. C4198. S188. Body sherd. Orange coarse vessel with a very gritty fabric, lots of fine lime specks, some fine quartz and fine to medium reddish grit. Brown slipped on the exterior and smoothed on the interior. Oxidised firing. Handmade. Exterior surface 7.5 YR 3/3, interior surface 2.5 YR 5/3. D= unknown. C4199. S188. Body sherd. Coarse vessel with a dark grey core and lots of inclusions including fine to large reddish grit, fine to medium lime specks and some plant temper. Dark brown burnished on the exterior and black burnished on the interior. Reduced firing. Handmade. Exterior surface 2.5 YR 5/6, interior surface 2.5 YR 3/1. D= unknown. C4209-4210. S188. Rim and body sherd. Orange brown coarse vessel with fine to medium lime and grit. Black burnished on the exterior and dark brown burnished on the interior. Reduced firing. Handmade. Exterior surface 2.5 YR 3/1, interior surface 2.5 YR 5/4. D= 14cm. C4212. S188. Body sherd. Orange brown coarse vessel with a dark grey core, and fine to large lime and grit. Smoothed on both sides. Reduced firing. Handmade. Exterior surface 2.5 YR 4/1, interior surface 2.5 YR 5/4. D= unknown. C4213. S188. Body sherd. Pale brown coarse vessel with a dark grey core, and fine to large lime and grit. Smoothed on both sides. Differential firing. Handmade. Exterior surface 7.5 YR 7/8, interior surface 7.5 YR 7/6. D= unknown. C4220. S188. Rim. Reddish brown to grey medium quality vessel with a poorly sorted medium hard paste, dark grey core, and lime inclusions smaller than 0.5 mm. Dark grey burnished on the exterior and smoothed on the interior. Differential firing. Handmade. Exterior surface 7.5 YR 3/1, interior surface 7.5 YR 4/2. D= unknown. C4228. S188. Rim and body sherd. Dark reddish brown very coarse vessel with dense very fine to medium lime specks and medium to large grit, and some very fine gold mica specks. Dark reddish brown burnished on both sides. Reduced firing. Handmade. Exterior surface 7.5 YR 5/6, interior surface 2.5 YR 5/6. D= 11cm. C4229. S188. Body sherd and handle. Dark brown to reddish brown very coarse vessel with dense fine to medium lime specks and fine to large dark grit. Dark brown burnished on the exterior. Reduced firing. Handmade. Exterior surface 10 YR 6/4, interior surface 7.5 YR 5/6. D= unknown. Middle Iron Age C931. U521. Rim. Grey brown fine vessel with black grit inclusions and voids. Light brown burnished on both sides. Reduced firing. Handmade. Exterior surface 5 YR 5/3, interior surface 5 YR 5/3. D= unknown. C1037. U328. Body sherd. Dark grey fine vessel with a medium number of white limestone inclusions. Smoothed on both sides. Oxidised firing. Handmade. Exterior surface 5 YR 5/1, interior surface 5 YR 5/1. D= unknown. C1415. G1D12. Body sherd. Orange red medium quality vessel with limestone, shell and grog inclusions. No surface treatment. Differential firing. Handmade. Exterior surface 2.5 YR 6/8, interior surface 2.5 YR 6/8. D= unknown. C2893. S24. Body sherd. Orange coarse to medium quality vessel with limestone inclusions. Smoothing on both surfaces. Oxidised firing. Wheelmade. Exterior surface 10 R 6/8, interior surface 10 R 6/8. D= unknown. C2900. G1D12. Body sherd. Yellowish red medium quality vessel with a grey core, and limestone and quartz inclusions. Smoothing on both surfaces. Reduced firing. Wheelmade. Exterior surface 5 YR 6/6, interior surface 5 YR 6/6. D= unknown. C3060. S88. Body sherd. Brownish grey medium quality vessel with small to medium size angular and sub-angular quartz and red grog. Smoothing and pale brown slip on the exterior and
Catalogue of Illustrated Artefacts
235
smoothing on the interior. Parallel raised bands on the exterior. Reduced firing. Wheelmade. Exterior surface 5 YR 6/4, interior surface 5 YR 4/1. D= unknown. C3162. S88. Body sherd. Reddish brown coarse vessel with small to medium grit and grog. Dark purple slip on the exterior. Differential firing. Wheelmade. Exterior surface 5 YR 5/2, interior surface 5 YR 5/4. D= unknown. C3167. S88. Body sherd. Greyish brown medium quality vessel with small black and white grit. Dark grey slip and burnish on the interior. Oxidised firing. Wheelmade. Exterior surface 2.5 YR 5/3, interior surface 10 YR 4/1. D= unknown. C3169. S88. Body sherd. Reddish brown medium quality vessel with dense small white grit, larger purple mineral inclusions and grog. Smoothed on both sides. Oxidised firing. Wheelmade. Exterior surface 2.5 YR 5/2, interior surface 2.5 YR 5/2. D= unknown. C3179. S88. Body sherd. Light reddish brown coarse vessel with a grey core, dense angular and sub-angular medium quartz and yellow grog. Smoothing and dark purplish brown slip on the exterior and smoothing on the interior. Oxidised firing. Wheelmade. Exterior surface 5 YR 2.5/2, interior surface 5 YR 6/4. D= unknown. C3182. S88. Body sherd. Grey medium to fine vessel with very small white and black inclusions. Reddish brown slip on the exterior. Parallel grooves on the exterior. Reduced firing. Wheelmade. Exterior surface 5 YR 6/6, interior surface 2.5 Y 6/1. D= unknown. C3191. S88. Body sherd. Orange medium vessel with small to medium size quartz, grog and black grit. Brown slip on the exterior and buff slip on the interior. Oxidised firing. Wheelmade. Exterior surface 5 YR 6/6, interior surface 5 YR 7/6. D= unknown. C3599. S88. Rim. Brownish grey medium vessel with dense small to medium rounded grog and black grit. Smoothed on both sides. Reduced firing. Wheelmade. Exterior surface 2.5 YR 4/6, interior surface 2.5 YR 4/6. D= 14cm. C4003. S132. Body sherd. Dark brown coarse vessel with very dense dark grit, voids and plant temper. No surface treatment. Reduced firing. Wheelmade. Exterior surface 7.5 YR 3/3, interior surface 7.5 YR 3/3. D= unknown. C4022. S188. Body sherd. Dark grey coarse vessel with a medium hard fabric, lime particles smaller than 1mm and some plant temper. Smoothed on both sides. Reduced firing. Wheelmade. Exterior surface 7.5 YR 3/3, interior surface 7.5 YR 3/3. D= unknown. C4222. S188. Body sherd. Light brown to red coarse vessel with a poorly sorted medium hard paste with limestone pieces up to 3mm, and chaff temper. Brown burnished on the exterior. Reduced firing. Handmade. Exterior surface 2.5 YR 4/8, interior surface 10 R 5/8. D= unknown. C4223. S188. Body sherd. Light brown coarse vessel with a poorly sorted medium hard paste and limestone pieces smaller than 1mm. Light brown burnished on the exterior. Reduced firing. Handmade. Exterior surface 2.5 YR 5/8, interior surface 2.5 YR 5/8. D= unknown. Late Iron Age C3172. S88. Body sherd. Reddish orange medium to fine vessel with very small white and small black grit. Smoothed on both sides. Differential firing. Wheelmade. Exterior surface 2.5 YR 5/4, interior surface 10R 6/6. D= unknown. C3607. S88. Rim. Greyish brown medium vessel with dense small to medium rounded grog and black grit. Smoothed on both sides. Reduced firing. Wheelmade. Exterior surface 2.5 YR 5/2, interior surface 2.5 YR 5/2. D= unknown. C3609. S88. Body sherd. Greyish brown medium vessel with dense small to medium rounded grog and black grit. Smoothed on both sides. Reduced firing. Wheelmade. Exterior surface 2.5 YR 5/2, interior surface 2.5 YR 5/2. D= unknown.
236
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
C3743.S88. Base. Yellowish buff fine vessel with dense small dark grit. Yellowish buff wash on both sides. Oxidised firing. Wheelmade. Exterior surface 5 YR 7/6, interior surface 5 YR 7/6. D= unknown. C3795. U3148. Body sherd. Light red medium vessel with small and a few large grit inclusions. Grey wash on the exterior. Oxidised firing. Wheelmade. Exterior surface GLEY 2 6/1, interior surface 10 R 5/6. D= unknown. C3908. G8K6. Body sherd. Pale pink fine vessel with very fine lime specks. Dark olive brown slip on the exterior. Oxidised firing. Wheelmade. Exterior surface 2.5 Y 1/2, interior surface 2.5 R 9/2. D= unknown.
Bibliography Greco-Roman Sources Apollonius of Rhodes, Argonautica. [translated by G. W. Mooney: The Argonautica of Apollonius Rhodius]. Dublin, Hodges, Figgis, & Co, 1912. Arrian, Anabasis Alexandri. [Translated by A. G. Roos: Flavii Arriani Anabasis Alexandri]. Leipzig, B.G. Teubner, 1907. Herodotus, Historiae. [Translated by A. D. Godley: Herodotus]. Cambridge / London, Harvard University Press / William Heinemann, 1920. Quintus Curtius Rufus, Historiae Alexandri Magni. [Translated by E. Hedicke: Historiae Alexandri Magni Macedonis libri qui supersunt]. Berlin, Weidmann, 1867. Strabo, Geographica. [Translated by H. L. Jones: The Geography of Strabo]. Cambridge / London, Harvard University Press / William Heinemann, 1924. Tabula Albana. [Translated by F. Jacoby and P. Bonnechere: Die Fragmente der griechischen Historiker Part I-III]. Leiden, Brill, 1999. Xenophon, Anabasis. [Translated by C. L. Brownson: Xenophon. Xenophon in Seven Volumes, III: Anabasis]. Cambridge MA / London, Harvard University Press / William Heinemann, 1968. Xenophon, Hellenica. [Translated by C. L. Brownson: Xenophon Hellenica, Books I-V]. Cambridge MA / London, Harvard University Press / William Heinemann, 1961. published sources Åkerström, Å. 1966. Die Architektonischen Terrakotten Kleinasiens. Lund, CWK Gleerup. Akurgal, E. 1955. Phrygische Kunst. Ankara, Archäologisches Institut der Universität Ankara. — 1962. The early period and the golden age of Ionia. American Journal of Archaeology 66: 369-79. Akurgal, E., and Budde, L. 1956. Vorläufiger Bericht über die Ausgrabungen in Sinope. Ankara, Türk Tarih Kurumu. Alkım, U. B., Alkım, H., and Bilgi, Ö. 1988. İkiztepe I: Birinci ve İkinci Dönem Kazıları. Ankara, Türk Tarih Kurumu. — 2003. İkiztepe II: Üçüncü, Dördüncü, Beşinci, Altıncı, Yedinci Dönem Kazıları 1976-1980. Ankara, Türk Tarih Kurumu. Alram, M. 1986. Nomina Propria Iranica In Nummis: Materialgrundlagen zu den iranischen Personennamen auf antiken Münzen. Vienna, Österreichische Akademie der Wissenschaften. Arslan, M. 2007. Roma’nın Büyük Düşmanı Mithradates VI Eupator. Istanbul, Odin Yayınları. Asheri, D., and Hoepfner, W. 1972. Herakleia Pontike: Forschungen zur Geschichte und Topographie. Vienna, Hermann Böhlaus.
Bibliography
237
Atici, L. 2009. Specialisation and diversification: Animal exploitation strategies in the terminal Pleistocene, Mediterranean Turkey. Before Farming 2009: 136-52. Avram, A., Hind, J., and Tsetskhladze, G. 2004. The Black Sea Area. In M. H. Hansen and N. T. Heine, eds., An Inventory of Archaic and Classical Poleis: An Investigation Conducted by the Copenhagen Polis Centre for the Danish National Research Foundation. Oxford, Oxford University Press: 924-73. Balabanov, I. P. 2009. Paleogeograficheskie predposylki formirovaniya sovremennyh prirodnyh usloviy i dolgosrochnyy prognoz razvitiya golotsenovyh terras Chernomorskogo poberezhya Kavkaza. Moscow, Dal’nauka Barnett, R. D. 1975. Phrygia and the Peoples of Anatolia in the Iron Age. In I. E. Edwards, C. J. Gadd, N. G. Hammond, and E. Sollberger, eds., The Cambridge Ancient History: II, Part 2, History of the Middle East and the Aegean Region c. 1380–1000 B.C. Cambridge, Cambridge University Press: 417-42. Barron, J. P. 1962. Milesian Politics and Athenian Propaganda c. 460-440 B.C. The Journal of Hellenic Studies 82: 1-6. Bilde, P. G., Bøgh, B., Handberg, S., Højte, J. M., Nieling, J., Smekalova, T., Stolba, V., Baralis, A., Bîrzescu, I., Gergova, D., Krapivina, V. V., Krusteff, K., Lungu V., and Maslennikov, A. A. 2007-2008. Archaeology in the Black Sea Region in Classical Antiquity 1993-2007. Archaeological Reports 54: 115-73. Bilgi, Ö. 1999. İkiztepe in the Late Iron Age. In A. Çilingiroğlu and R. Matthews, eds., Anatolian Iron Ages 4. Anatolian Studies 49. London, British School of Archaeology at Ankara: 27-54. Bittel, K. 1976. Die Hethiter. Munich, Beck. Bossert, E.-M. 2000. Die Keramik phrygischer Zeit von Boğazköy. Funde aus den Grabungskampagnen 1906, 1907, 1911, 1912, 1931-1939 und 1952-1960. Mainz am Rhein, Philipp von Zabern. Boysal, Y. 1959. Über die älteren Funde von Sinope und die Kolonisationsfrage. Archäologischer Anzeiger: 8-20. Braund, D. 2002. The myths of Panticapaeum: Construction of colonial origins in the Black Sea Region. In G. R. Tsetskhladze and A. M. Snodgrass, eds., Greek Settlements in the Eastern Mediterranean and the Black Sea. Oxford, Archaeopress: 73-80. — 2010. Myth and ritual at Sinope: From Diogenes the Cynic to Sanape the Amazon. Ancient Civilizations from Scythia to Siberia 16: 11-23. Briant, P. 2002. From Cyrus to Alexander: A History of the Persian Empire. Winona Lake, Eisenbrauns. Brosius, M. 2010. Pax Persica and the peoples of the Black Sea region: Extent and limits of the Achaemenid imperial ideology. In J. Nieling and E. Rehm, eds., Achaemenid Impact in the Black Sea: Communication of Powers. Aarhus, Aarhus University Press: 29-40. Brückner, H., Kelterbaum, D., Marunchak, O., Porotov, A., and Vött, A. 2010. The Holocene sea level story since 7500 BP - Lessons from the eastern Mediterranean, the Black and the Azov Seas. Quaternary International 225: 160–79. Bryce, T. 1998. The Kingdom of the Hittites. Oxford, Oxford University Press. — 2009. The Routledge Handbook of The People and Places of Ancient Western Asia: The Near East from the Early Bronze Age to the Fall of the Persian Empire. New York, Routledge. Burney, C. 1956. Northern Anatolia before Classical times. Anatolian Studies 6: 179-203. Burov, G. M. 1997. The human use of caves and rock shelters in the mountain Crimea (Ukraine) in Prehistoric, Classical and Medieval times. In C. Bonsall and C. Tolan-Smith, eds., The Human Use of Caves. Oxford, Archaeopress: 127-35. Burstein, S. M. 1976. Outpost of Hellenism: The Emergence of Heraclea on the Black Sea. Berkeley and Los Angeles, University of California Press.
238
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
Butyagin, A. M. 2007. The history of the northern Black Sea Region. In A. Trofimova, ed., Greeks on the Black Sea: Ancient Art from the Hermitage. Los Angeles, Getty Publications: 8-17. Carpenter, R. 1948. The Greek penetration of the Black Sea. American Journal of Archaeology 52(1): 1-10. Carrington, P. 1977. The heroic age of Phrygia in ancient literature and art. Anatolian Studies 27: 117-26. Carter, J. C., Crawford, M., Lehman, P., Nikolaenko, G., and Trelogan, J. 2000. The Chora of Chersonesos in Crimea, Ukraine. American Journal of Archaeology 104: 707-41. Cohen, G. M. 1995. The Hellenistic Settlements in Europe, the Islands, and Asia Minor. Berkeley and Los Angeles, University of California Press. Cook, J. M. 1983. The Persian Empire. New York, Schocken Books. Cox, R. 1990. A history of music. The Journal of Aesthetics and Art Criticism 48: 395-409. Dandamaev, M. A. 1989. A Political History of the Achaemeid Empire. Leiden, Brill. De Jesus, P. S. 1978. Metal resources in ancient Anatolia. Anatolian Studies 27: 97-102. De Vries, K. 2011a. Textual evidence and the destruction level. In C. B. Rose and G. Darbyshire, eds., The New Chronology of Iron Age Gordion. Philadelphia, University of Pennsylvania Museum of Archaeology and Anthropology: 49-58. — 2011b. The creation of old chronology. In C. B. Rose and G. Darbyshire, eds., The New Chronology of Iron Age Gordion. Philadelphia, University of Pennsylvania Museum of Archaeology and Anthropology: 13-22. De Vries, K., Sams, G. K., and Voigt, M. M. 2005. Gordion re-dating. In G. Darbyshire and A. Çilingiroğlu, eds., Anatolian Iron Ages 5. London, British Institute of Archaeology at Ankara: 45-6. Dönmez, Ş. 2003. The Early Iron Age problem in the central Black Sea region. In B. Fischer, H. Genz, E. Jean, and K. Köroğlu, eds., Identifying Changes: The Transition from Bronze to Iron Ages in Anatolia and its Neighbouring Regions. Istanbul, Ege Yayınları: 213-28. — 2005. Amasya Province in the Iron Age. In A. Çilingiroğlu and G. Darbyshire, eds., Anatolian Iron Ages 5. London, British School of Archaeology at Ankara: 65-74. — 2006. Some observations on the socio-economic structure and ethnic make-up of the central Black Sea region of Turkey during the Iron Age in the light of new evidence. Ancient West and East 5: 13-43. — 2010. Oluz Höyük kazısı üçüncü dönem (2009) çalışmaları: Değerlendirmeler ve sonuçlar. Colloquium Anatolicum 9: 275-306. Doonan, O. P. 2004. Sinop Landscapes: Exploring Connection in a Black Sea Hinterland. Philadelphia, University of Pennsylvania Museum of Archaeology and Anthropology. — 2006. Exploring community in the hinterland of a Black Sea port. In P. G. Bilde and V. F. Stolba, eds., Surveying the Greek Chora. The Black Sea Region in a Comparative Perspective. Aarhus, Aarhus University Press: 47-58. Drews, R. 1976. The earliest Greek settlements on the Black Sea. The Journal of Hellenic Studies 96: 18-31. — 1993. The End of the Bronze Age: Changes in Warfare and Catastrophe ca. 1200 B.C. Princeton, Princeton University Press. Durbin, G. E. 1971. Iron Age pottery from the provinces of Tokat and Sivas. Anatolian Studies 21: 99-124. Dyson, R. H. 1999. The Achaemenid painted pottery of Hasanlu IIIA. A. Çilingiroğlu and R. Matthews, eds., Anatolian Iron Ages 4 (Anatolian Studies 49). London, British Institute of Archaeology at Ankara: 101-10.
Bibliography
239
Ersal, Y. 1994. Antik Çağdan Günümüze Karadeniz Ekonomisi. Ankara, Türkiye Odalar ve Borsalar Birliği Yayınları. Evans, J. A. 1963. Histiaeus and Aristagoras: Notes on the Ionian revolt. The American Journal of Philology 84: 113-28. Fedorov, P. V. 1972. Postglacial transgression of the Black Sea. International Geology Review 14(2): 160-164. — 1977. Pozdnechetvertichnaia istoriia Chernogo moria i razvitie iuzhnykh morei Evropy. In P. A. Kaplin and F. A. Shcherbakov, eds., Paleogeografiia i otlozheniia pleistotsena iuzhnykh morei SSSR. Moscow, Nauka: 25-32. Finkielsztejn, G. 2011. Black Sea amphora stamps found in the southern Levant: Most probably not a trade. In C. Tzochev, T. Stoyanov, and A. Bozkova, eds., Production and Trade of Amphorae in the Black Sea II. Sofia, St. Kliment Ohridski University Press; National Institute of Archaeology and Museum: 55-61. Fol, V. 2008.The rock as a topos of faith: The interactive zone of the rock-cut monuments - from Urartu to Thrace. In R. I. Kostov, B. Gaydarska and M. Gurova, eds., Geoarchaeology and Archaeomineralogy. Sofia, St. Ivan Rilski: 153-62. Fouache, E., Kelterbaum, D., Brückner, H., Lericolais, G., Porotov, A., and Dikarev V. 2012. The Late Holocene evolution of the Black Sea - A critical view on the so-called Phanagorian Regression. Quaternary International 266: 162-74. Garlan, Y. 2000. Amphores et Timbres Amphores et timbres amphoriques grecs: Entre erudition et idéologie. Paris, Académie des inscriptions et belles-lettres. Genz, H. 2001. Iron Age pottery from Çadır Höyük. Anatolica 27: 159-70. — 2003. The Early Iron Age in central Anatolia. In B. Fischer, H. Genz, É. Jean, and K. Köroğlu, eds., Identifying Changes. The Transition from Bronze to Iron Ages in Anatolia and its Neighbouring Regions. Istanbul, Türk Eskiçağ Bilimleri Enstitüsü: 179-92. — 2004. Büyükkaya I: Die Keramik der Eisenzeit. Funde aus den Grabungskampagnen 1993 bis 1998. Mainz am Rhein, Philipp von Zabern. — 2005. Thoughts on the origin of the Iron Age pottery traditions in Central Anatolia. In A. Çilingiroğlu and G. Darbyshire, eds., Anatolian Iron Ages 5. London, British Institute of Archaeology at Ankara: 75-84. — 2006a. Die eisenzeitliche Besiedlung im Bereich der Grabungen an den Ostteichen 1996-1998. In J. Seeher, ed., Ergebnisse der Grabungen an den Ostteichen und am mittleren BüyükkakaleNordwesthang 1996-2000. Mainz am Rhein, Philipp von Zabern: 26-38. — 2006b. Die eisenzeitliche Besiedlung im Bereich der Grabungen an den Ostteichen 1998-2000. In J. Seeher, ed., Ergebnisse der Grabungen an den Ostteichen und am mittleren BüyükkakaleNordwesthang 1996-2000. Mainz am Rhein, Philipp von Zabern: 98-158. — 2011. The Iron Age in central Anatolia. In G.R. Tsetskhladze, ed., The Black Sea, Greece, Anatolia and Europe in the First Millennium BC. Leuven, Peeters: 331-368. Glatz, C., and Matthews, R. 2005. Anthropology of a frontier zone: Hittite-Kaska relations in Late Bronze Age north-central Anatolia. Bulletin of the American Schools of Oriental Research 339: 21-39. Glendinning, M. R. 1996. A mid-sixth century tile roof system at Gordion. Hesperia 65: 99-119. — 2002. Recovering the lost art of Phrygian roof tiling: Practical and aesthetic elements converge in clay, reflecting Greek artistic temperament. Expedition 44(2): 28-35. Göktürk, O. M. 2011. Climate in the Eastern Mediterranean through the Holocene Inferred from Turkish Stalagmites. Bern, University of Bern (Unpublished Sc.D. Thesis). Golomshtok, E. A. 1938. The Old Stone Age in European Russia. Transactions of the American Philosophical Society, New Series 29(2): i-vii, 190-460.
240
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
Gorny, R. L. 2006. The 2002-2005 excavation seasons at Çadır Höyük. Anatolica 32: 29-54. Graham, A. J. 1971. Patterns in early Greek colonisation. The Journal of Hellenic Studies 91: 35-47. Grayson, A. K. 1991. Assyria: Tiglath-Pileser III to Sargon II (744-705 B.C.). In J. Boardman, I. E. Edwards, N. G. Hammond, and E. Sollberger, eds., The Cambridge Ancient History: III, part 2, The Assyrian and Babylonian Empires and Other States of the Near East, from the Eighth to the Sixth Centuries B.C. Cambridge, Cambridge University Press: 71-102. Greaves, A. 2007. Milesians in the Black Sea: Trade, settlement and religion. In V. Gabrielsen and J. Lund, eds., The Black Sea in Antiquity: Regional and Interregional Economic Exchanges. Aarhus, Aarhus University Press: 9-21. Hannestad, L. 2007. Timber as a trade resource of the Black Sea. In V. Gabrielsen and J. Lund, eds., The Black Sea in Antiquity. Regional and Interregional Economic Exchanges. Aarhus, Aarhus University Press: 85-99. Harmatta, J. 1996. The Scythians. In J. Herrmann, E. Zürcher, J. Harmatta, J. Litvak King, R. Lonis, T. Obenga, and Z. Y. Romila Thapar, eds., History of Humanity: Volume III, From the Seventh Century BC to the Seventh Century AD. Paris; London, United Nations Educational, Scientific and Cultural Organisation; Routledge: 181-2. Hellmuth, A. 2008. The chronological setting of the so-called Cimmerian and early Scythian material from Anatolia. Ancient Near Eastern Studies 45: 102-122. Hencken, H. 1968.Tarquinia, Villanovans and Early Etruscans. Cambridge, MA, Peabody Museum of Archaeology and Ethnology. Henrickson, R. C. 1993. Politics, economics, and ceramic continuity at Gordion in the late second and first millennia B.C. In W. D. Kingery, ed., The Social and Cultural Contexts of New Ceramic Technology. Westerville, American Ceramic Society: 89-176. — 1994. Continuity and discontinuity in the ceramic tradition of Gordion during the Iron Age. In A. Çilingiroğlu and D. H. French, eds., Anatolian Iron Ages 3. London, British Institute of Archaeology at Ankara: 95-129. Henrickson, R. C., Vandiver, P. B., and Blackman, M. J. 2002. Lustrous black fine ware at Gordion, Turkey: A distinctive sintered slip technology. In P. B. Vandiver, M. Goodway, and J. L. Mass, eds., Materials Issues in Art and Archaeology VI. Warrendale, Materials Research Society: 391-400. Hind, J. 1988. The colonisation of Sinope and the southeast Black Sea area. In O. D. Lordkipanidze, ed., Mestnye Ethno-Politicheskie Ob’edineniya Prichernomor’ya v VII-IV vv. do n.e. Tblisi, Metsniereba: 207-23. Hornblower, S. 1994. Asia Minor. In D. M. Lewis, J. Boardman, S. Hornblower, and O. Martin, eds., The Cambridge Ancient History: VI, The Fourth Century B.C. Cambridge, Cambridge University Press: 209-33. Işın, M. A. 1998. Sinop regional field survey. Anatolia Antiqua 6: 95-139. Ivantchik, A. I. 2005. Early Eurasian nomads and the civilizations of the ancient. In R. Amitai and M. Biran, eds., Mongols, Turks, and Others: Eurasian Nomads and the Sedentary World. Leiden, Brill: 103-26. Johnson, P. 2011. Topographies of urbanization: Survey in and around Pompeiopolis. In L. Summerer, ed., Pompeiopolis I. Eine Zwischenbilanz aus der Metropole Paphlagoniens nach fünf Kampagnen (2006 – 2010). Langenweißbach, Beier & Beran: 195-206. Kartal, M. 2009. Karain Mağarası 2008 yılı kazıları. ANMED 7: 22-5. Kealhofer, L. 2005. The Gordion Regional Survey. Settlement and land use. In L. Kealhofer, ed., The Archaeology of Midas and the Phrygians: Recent Work at Gordion. Philadelphia, University of Pennsylvania Museum of Archaeology and Anthropology: 137-48.
Bibliography
241
Kealhofer, L., Grave, P., Genz, H., and Marsh, B. 2009. Post-Collapse: The re-emergence of polity in Iron Age Boğazköy, central Anatolia. Oxford Journal of Archaeology 28: 275-300. Kealhofer, L., and Grave, P. 2011. The Iron Age on the central Anatolian plateau. In S. R. Steadman and G. McMahon, eds., The Oxford Handbook of Ancient Anatolia. Oxford, Oxford University Press: 415-42. Kealhofer, L., Grave, P., Marsh, B., and Matsumura, K. 2008. Analysis of specialized Iron Age wares at Kaman-Kalehöyük. Kaman-Kalehöyük - Anatolian Archaeological Studies 17: 201-23 Kealhofer, L., Grave, P., Marsh, B., Steadman, S., Gorny, R. L., and Summers, G. D. 2010. Patterns of Iron Age interaction in central Anatolia: Three sites in Yozgat Province. Kaman-Kalehöyük Anatolian Archaeological Studies 60: 71-92. Kenneth, S. G. 1994. The Early Phrygian Pottery, Volume IV. The Gordion Excavations 1950-1973: Final Reports. Philadelphia, University of Pennsylvania Museum. Koçak, Ö. 2006. Eskiçağ’da Orta Karadeniz Bölümü Madenciliği. Mining at the Black Sea Region in the Ancient Period. In D. B. Erciyas and E. Koparal, eds., Karadeniz Araştırmaları Sempozyum Bildirileri. Black Sea Studies Symposium Proceedings. Istanbul, Ege Yayınları: 1-62. Kondrashov, A. 1995. Underwater investigations at Cape Panagia on the Taman Peninsula, Straits of Kerch. The International Journal of Nautical Archaeology 24: 109-19. Koşay, H. Z. 1941. Pazarlı Hafriyati Raporu. Les Fouilles de Pazarli: Entreprises par la Société d’Histoire Turque. Ankara, Türk Tarih Kurumu. Koşay, H. Z., Akok, M. 1957. Türk Tarih Kurumu Tarafından Yapılan Büyük Güllücek Kazısı: 1947 ve 1949 daki Çalışmalar Hakkında İlk Rapor. Ausgrabungen von Büyük Güllücek. Ausgeführt durch die Türkische Historische Gesellschaft: Vorbericht über die Arbeiten von 1947 und 1949. Ankara, Türk Tarih Kurumu. Kuzucuoğlu, C., Marro, C., Özdoğan, A., and Tibet, A. 1997. Prospection archéologique franco-turque dans la région de Kastamonu (Mer Noire). Deuxième rapport préliminaire. Anatolia Antiqua 5: 275-306. Lepik-Kopaczynska, W. 1958. Colores floridi und austeri in der antiken Malerei. Archäologischer Anzeiger 72: 79-99. MacIntoch Turfa, J. 2011. Catalogue of the Etruscan Gallery of the University of Pennsylvania Museum of Archaeology and Anthropology. Philadelphia, University of Pennsylvania Press. Macridy, T. 1907. Une citadelle archaïque du Pont. Mitteilungen der Vorderasiatischen Gesellschaft 4: 1-9. Marro, C., Özdoğan, A., and Tibet, A. 1996. Prospection archéologique franco-turque dans la région de Kastamonu (Mer Noire). Premier rapport préliminaire. Anatolia Antiqua 4: 273-90. — 1998. Prospection archéologique franco-turque dans la Région de Kastamonu (Mer Noire). Troisième rapport préliminaire. Anatolia Antiqua 6: 317-35. Matsumura, K. 2000. On the manufacturing techniques of Iron Age ceramics from Kaman-Kalehöyük. Kaman-Kalehöyük - Anatolian Archaeological Studies 9: 119-35. — 2001. On the manufacturing techniques of Iron Age ceramics from Kaman-Kalehöyük (II): the cultural influence of Phrygia at Kaman-Kalehöyük. Kaman-Kalehöyük - Anatolian Archaeological Studies 10: 101-10. — 2008. A note on Anatolian Iron Age ceramic chronology: Black Lustrous Ware with diamond faceting. Kaman-Kalehöyük - Anatolian Archaeological Studies 17: 175-84. Matsunaga, M., and Nakai, I. 2000. Study on the origin of the silver luster on grey Iron Age pottery. Kaman-Kalehöyük - Anatolian Archaeological Studies 9: 207-11. Matthews, R. 2009. A dark age, grey ware and elusive empires: Paphlagonia through the Iron Age, 1200-330 BC. In R. Matthews and C. Glatz, eds., At Empire’s Edge: Project Paphlagonia. Regional Survey in North-Central Turkey. London, British Institute at Ankara: 149-72.
242
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
Matthews, R., and Glatz, C. eds. 2009. At Empire’s Edge: Project Paphlagonia. Regional Survey in North-Central Turkey. London, British Institute at Ankara. McGing, B. C. 1986. The Foreign Policy of Mithridates VI Eupator King of Pontus. Leiden, E.J. Brill. Mehl, A. 1987. Der Überseehandel von Pontos. In E. Olshausen, ed., Stuttgarter Kolloquium zur historischen Geographie des Altertums 1, 1980. Bonn, Rudolf Habelt: 104-86. Mellink, M. 1991. The native kingdoms of Anatolia. In J. Boardman, I. E. Edwards, N. G. Hammond, and E. Sollberger, eds., The Cambridge Ancient History: III, part 2, The Assyrian and Babylonian Empires and Other States of the Near East, from the Eighth to the Sixth Centuries B.C. Cambridge, Cambridge University Press: 619-65. Meyer, H. W. 1940. Die phrygische Dachdeckung. Mitteilungen der Deutschen Orient-Gesellschaft 78: 69-75. Mitchell, S. 2010. The Ionians of Paphlagonia. In T. Whitmarsh, ed., Local Knowledge and Micro Identities in the Imperial Greek World. Cambridge, Cambridge University Press: 86-110. Monakhov, S. I. 2003. Grecheskie Amfory v Prichernomor'e: Tipologii͡a Amfor Vedushchikht͡ sentrovėksporterov Tovarov v Keramicheskoĭ Tare. Katalog-Opredelitelʹ. Moscow, Kimmerida. Moschos, I. 2009. Evidence of social re-organization and reconstruction in Late Helladic IIIc Achaea and modes of contacts and exchange via the Ionian and Adriatic Sea. In E. Borgna and P. Cassola Guida, eds., Dall’ Egeo all’ Adriatico: organizzazioni sociali, modi di scambio e interazione in età postpalaziale (XII-XI sec. a.C.). Rome, Edizioni Quasar di Severino Tognon s.r.l.: 345-414. Moysey, R. A. 1986. The silver stater issues of Pharnabazus and Datames from the mint of Tarsus in Cilicia. Museum Notes (American Numismatic Society) 31: 7-61. Muscarella, O. 1995. The Iron Age background to the formation of the Phrygian state. Bulletin of the American Schools of Oriental Research 299/300: 91-101. Omura, S. 2000. Preliminary report of the general survey in central Anatolia (1999). KamanKalehöyük - Anatolian Archaeological Studies 9: 37-96. — 2001. Preliminary report of the general survey in central Anatolia (2000). Kaman-Kalehöyük Anatolian Archaeological Studies 10: 37-86. — 2002. Preliminary report of the general survey in central Anatolia (2001). Kaman-Kalehöyük Anatolian Archaeological Studies 11: 45-112. — 2011. Kaman-Kalehöyük Excavations in central Anatolia. In S. R. Steadman and G. McMahon, eds., The Oxford Handbook of Ancient Anatolia. Oxford, Oxford University Press: 1095-111. Otto, M., Westermark, U., and Grierson, P. 1991. Early Hellenistic Coinage from the Accession of Alexander to the Peace of Apamaea (336-188 B.C.). Cambridge, Cambridge University Press. Özgüç, T. 1980. Excavation at the Hittite site, Maşat Höyük: Palace, archives, Mycenaean pottery. American Journal of Archaeology 84: 305-9. Özsait, M. 2002. 2001 yılı Samsun ve Amasya yüzey araştırmaları. Araştırma Sonuçları Toplantısı 20/2: 127-40. — 2003. 2002 yılı Samsun ve Amasya yüzey araştırmaları. Araştırma Sonuçları Toplantısı 21/2: 273-84. Özsait, M., and Özsait, N. 2010. Amasya yüzey araştırmaları. Araştırma Sonuçları Toplantısı 28/2: 25-40. Porotov, A. 2007. Relative sea-level changes and submersion of archaeological sites along the northern shoreline of the Black Sea. Mediterranée 108(1): 29-36. Prag, A. J. 1989. Reconstructing King Midas: A first report. Anatolian Studies 39: 159-65. Roller, L. E. 2011. Phrygian and the Phrygians. In S. R. Steadman and G. McMahon, eds., The Oxford Handbook of Ancient Anatolia. Oxford, Oxford University Press: 560-78.
Bibliography
243
Romanos, C. L. 2011. Handmade Burnished Ware in Late Bronze Age Greece and Its Makers. Birmingham, University of Birmingham (Unpublished Ph.D. Thesis). Ross, J. C. 2010. The Iron Age at Çadır Höyük. Anatolica 36: 67-87. Sams, G. K. 1994. The Gordion Excavations, 1950-1973, Final Reports Volume IV: The Early Phrygian Pottery. Philadelphia, University of Pennsylvania Museum of Archaeology and Anthropology. Seeher, J. 2000. Die Ausgrabungen in Boğazköy-Hattuša 1999. Archäologischer Anzeiger: 355-76. Sekunda, N. 1991. Achaemenid settlement in Caria, Lycia and Greater Phrygia. In H. SancisiWeerdenburg and A. Kuhrt, eds., Achaemenid History VI. Asia Minor and Egypt: Old Cultures in a New Empire. Leiden, Nederlands Instituut voor Het Nabije Oosten: 83-143. Sherratt, S. 1998. ‘Sea Peoples’ and the economic structure of the Late Second Millennium BC in the Eastern Mediterranean. In S. Gitin, A. Mazar, and E. Stern, eds., Mediterranean Peoples in Transition. Jerusalem, Israel Exploration Society: 292-313. Singer, I. 2000. New evidence of the end of the Hittite Empire. In E. D. Oren, ed., Sea Peoples and Their World. Philadelphia, University of Pennsylvania Press: 21-35. Stolba, V. F. 2003. Some reflections on the Amphora stamps with the name Amastris. In V. F. Stolba, J. M. Højte, and V. F. Stolba, eds., The Cauldron of Ariantas. Studies Presented to A.N. Sčeglov on the Occasion of his 70th Birthday. Aarhus, Aaarhus University Press: 279-301. Storch, G., Schoch, W., Rähle, W., Moser, J., Burger, D., Berke, H., Albrecht, B., Albrecht, G., and Uerpmann, H.-P. 1992. Late Pleistocene and Early Holocene finds from Öküzini: A contribution to the settlement history of the Bay of Antalya, Turkey. Paléorient 18(2): 123-41. Summerer, L. 2005a. Architektur und Akkulturation. Griechische Tondächer im kappadokischen Kontext. Habilitationsschrift. Munich, University of Munich. — 2005b. Griechische Tondächer im kappadokischen Kontext: Die Architektur-Terrakotten aus Akalan. In F. Fless, ed., Bilder und Objekte als Träger kultureller Kommunikation im Schwarzmeergebiet. Rahden, Verlag Marie Leidorf: 125-39. — 2007. Greeks and natives on the southern Black Sea coast in Antiquity. In G. Erkut and S. Mitchell, eds., The Black Sea: Past, Present and Future. London, British Institute at Ankara: 27-36. — 2008. Indigenous responses to encounters with the Greeks in Northern Anatolia: The reception of architectural terracottas in the Iron Age settlements of the Halys Basin. In B. P. Guldager, and P. J. Hjarl, eds., Meetings of Cultures: Between Conflicts and Coexistence. Aarhus, Aarhus University Press: 263-86. Summerer, L., and von Kienlin, A. 2010. Achaemenid impact in Paphlagonia: Rupestal tombs in Amnias valley. In J. Nieling and E. Rehm, eds., Achaemenid Impact in the Black Sea. Communication of Powers. Aarhus, Aarhus University Press: 195-221. Summers, G. D. 1994. Grey Ware and the eastern limits of Phrygia. In A. Çilingiroğlu and D. H. French, eds., Anatolian Iron Ages 3. London, British Institute of Archaeology at Ankara: 241-52. Summers, G., and Summers, F. 1998. The Kerkenes Dağ Project. In R. Matthews, ed., Ancient Anatolia: Fifty Years’ Work by the British Institute of Archaeology at Ankara. London, British Institute of Archaeology at Ankara: 177-94. Şenol, G. Ş. 2006. Klasik ve Hellenistik Dönem’de Mühürlü Amphora Üreten Merkezler ve Mühürleme Sistemleri. Istanbul, Ege Yayınları. Tsetskhladze, G. R. 1994. Greek penetration of the Black Sea. In G. R. Tsetskhladze, ed., The Archaeology of Greek Colonisation. Essays Dedicated to Sir John Boardman. Oxford, Oxford University Committee for Archaeology: 111-36. — 2002. Ionians abroad. In G. R. Tsetskhladze and A. M. Snodgrass, eds., Greek Settlements in the Eastern Mediterranean and the Black Sea. Oxford, Archaeopress: 81-96.
244
The Cide-Şenpazar Region in the Iron Age (ca. 1200 to 325/300 BC)
— 2005. Early Iron Age societies of the Black Sea region and Anatolia: some observations. In G. Darbyshire and A. Çilingiroğlu, eds., Anatolian Iron Ages 5. London, British Institute of Archaeology at Ankara: 211-16. — 2007. Greeks and locals in the southern Black Sea littoral: A re-examination. In G. Herman, I. Shatzman, and D. Asheri, eds., Greeks Between East and West: Essays in Greek Literature and History in Memory of David Asheri. Jerusalem, Israel Academy of Sciences and Humanities: 160-95. — 2009. Secondary colonisers in the Black Sea: Sinope and Panticapaeum. In M. Lombardo, ed., Colonie di Colonie: Le Fondazioni Sub-Coloniali Greche tra Colonizzazione e Colonialismo. Galatina, Congedo: 229-54. — 2010. Black Sea trade: Some further general observations. Anadolu Araştırmaları/ Jahrbuch für Kleinasiatische Forschung 19 (1): 197-212. — 2011. The Scythians: three essays. In G.R. Tsetskhladze, ed., The Black Sea, Greece, Anatolia and Europe in the First Millennium BC. Leuven, Peeters: 95-140. Ünal, A. 1999.A Hittite Mycenaean type B sword from the vicinity of Kastamonu, northwest Turkey. In T. Mikasa, ed., Essays on Ancient Anatolia. Wiesbaden, Harrassowitz Verlag: 207-26. Vagnetti, L. 1985. Late Minoan III Crete and Italy: Another view. La Parola del Passato 40: 29-33. Vassileva, M. 1997. King Midas: Between the Balkans and Asia Minor. Dialogues d’Histoire Ancienne 23(2): 9-20. — 2012. The rock-cut monuments of Phrygia, Paphlagonia and Thrace: a comparative overview. In G. R. Tsetskhladze, E. Laflı, J. Hargrave, and W. Anderson, eds., The Black Sea, Paphlagonia, Pontus and Phrygia in Antiquity: Aspects of Archaeology and Ancient History. Oxford, Archaeopress: 243-52. Voigt, M. M. 1994. Excavations at Gordion 1988-89. The Yassıhöyük stratigraphic sequence. In A. Çilingiroğlu and D. H. French, eds., Anatolian Iron Ages 3. London, British Institute of Archaeology at Ankara: 265-93. — 2002. Gordion: the rise and fall of an Iron Age capital. In D. C. Hopkins, ed., Across the Anatolian Plateau: Readings in the Archaeology of Ancient Turkey. Boston, American Schools of Oriental Research: 187-96. — 2005. Old problems and new solutions. Recent excavations at Gordion. In L. Kealhofer, ed., The Archaeology of Midas and the Phrygians: Recent Work at Gordion. Philadelphia, University of Pennsylvania Museum of Archaeology and Anthropology: 22-35. Voigt, M. M., De Vries, K., Henrickson, R. C., Lawall, M., Marsh, B., Gürsan-Salzmann, A., and Young, T. C. 1997. Fieldwork at Gordion 1993-1995. Anatolica 23: 1-59. Voigt, M. M., and Henrickson, R. C. 2000. Formation of the Phrygian state: The Early Iron Age at Gordion. Anatolian Studies 50: 37-54. Voigt, M. M., and Young, T. C. 1999. From Phrygian capital to Achaemenid entrepot: Middle and Late Phrygian Gordion. Iranica Antiqua 34: 191-241. von der Osten, H. H. 1930a. The Alishar Hüyük. Chicago, University of Chicago Press. — 1930b. Explorations in Hittite Asia Minor, 1929. Chicago, University of Chicago Press. — 1937. The Alishar Hüyük Seasons of 1930-1932, Part 2. Chicago, The University of Chicago Press. Walinga, H. T. 1984. The Ionian revolt. Mnemosyne 37(3): 401-37. Yakar, J. 2006. Dating the sequence of the final destruction/ abandonment of LBA settlements: Towards a better understanding of events that led to the collapse of the Hittite kingdom. In D. Mielke, U.-D. Schoop, and J. Seeher, eds., Structuring and Dating in Hittite Archaeology. Istanbul, Ege Yayınları: 33-51. Yıldırım, T., and Sipahi, T. 2004. 2002 Yılı Çorum ve Çankırı İlleri yüzey araştırmaları. Araştırma Sonuçları Toplantısı 21/2: 305-13.
Bibliography
245
Yılmaz, E. 2012. Salıpazarı Yeşilköy Kaya Mezarları ve Garpu Kale üzerine bir değerlendirme. Samsun Sempozyumu, 13-16 Ekim 2011, Samsun’a Bir Parantez Açıyoruz, Cilt I. Samsun, Ceylan Ofset: 1-20. Young, R. S. 1960. Gordion campaign of 1959: Preliminary report. American Journal of Archaeology 64: 227-44. — 1965. Early mosaics at Gordion. Expedition 7: 4-13.
T. Emre Şerifoğlu and Caner Bakan
9 The Cide-Şenpazar Region during the Hellenistic Period (325/300 - 1 BC) Greek colonies were established along the Turkish Black Sea coast during the Archaic and Classical periods. The toponym of Kytoros,1 probably the cove of Gideros located to the west of the modern town of Cide, is first mentioned in Homer’s Illiad (Hom. Il. 2. 853). This could potentially point to an even earlier Greek presence in the region (Drews 1976: 18; de Boer 2006), even if the interpretation of the Homeric epics, their chronology and the accuracy of geographical and other information, is wrought with difficulties and the subject of an ongoing debate (e.g. Tsetskhladze 1997). There is also as yet no archaeological evidence to corroborate Greek colonial settlement prior to ca. 800 BC in the Black Sea or a Late Bronze Age Mycenaean connection with the southern shores (Carpenter 1948: 1; Bouzek 1990: 13; also chapter 7, this volume). In the Cide and Şenpazar region, elements of Greek cultural traditions are attested for the first time in the late first millennium BC. In this chapter, we present the Hellenistic period finds recorded in the course of the Cide Archaeological Project (CAP) and their wider archaeological and historical context.
9.1 The Hellenistic Period on the Turkish Black Sea Coast The Hellenistic period in Paphlagonia and northwestern Anatolia is still very much a ‘dark age’ with regards to our knowledge of the region’s history and cultural traditions. From historical sources we know that the army of Alexander the Great never marched north of Ancyra to conquer Paphlagonia or the Black Sea coast, but left these areas under the control of Achaemenid satraps, who eventually accepted Alexander’s authority (Ashley 2004: 215-6; Bosworth 2006: 805). At the same time, Greek sources indicate that the coast and its colonial settlements, as well as the maritime trade network that tied them together, were firmly in Greek hands, and that Greek had established itself as the main language of the coastal region (Tsetskhladze 2002: 86-7, note 35). However, the situation in areas further inland, which would have been under the control of local dynasts, must have been rather different (Mitchell 2010: 102-6).
1 The toponym is attested in several different forms: Ps-Skylax, Arr. peripl. m. Eux. 90: Kytoris; Arr. peripl. m. Eux 20: Kytoros; Anonymi, Arr. peripl. m. Eux 17: Kytoros; Ptolemy 5. 1. 7: Kytoron; Pom. Mel. de situ orbis 19. 8: Cytoros; Apollon. 2. 942: Kytoros; Steph. Byz. PeriPoleon 41: Kytoros.
Hellenistic Settlement Patterns
247
With Achaemenid control over northern Anatolia fading in the wake of Alexander’s campaign, Dionysius, the ruler of Heraclea and husband to Amastris, Darius III’s niece (Heckel and Yardley 2004: 183), seized this opportunity to expand his realm as far as Kytoros (Strab. 12. 3. 10; Burstein 1976: 74, 79). After his death in 305 BC, Amastris married Lysimachos, one of Alexander’s generals, and following the end of a short-lived marriage, she became the sole ruler of Heraclea. Amastris had coins struck under her name in a Persian style (Mørkholm 1991: 95-6; Mitchell 2004: 145). During this period she founded the polis of Amastris by uniting the towns of Sesamus, Cromna, Kytoros and Tieum (Mitchell 2010: 93). After queen Amastris was killed by her sons in 284 BC, Lysimachos re-took control of the city and gave it to his wife Arsinoe, appointing Eumenes as the acting ruler (McGing 1986a: 18; Cohen 1995: 383; Stolba 2003: 280). Mithridates I (281-266 BC), who might have been a descendent of a royal Persian satrap family, founded the kingdom of Pontus after a rebellion that he started in Paphlagonia against the Hellenistic kingdoms (McGing 1986b; Bosworth and Wheatley 1998). The Cide and Şenpazar region would have formed the western most part of Pontus, although much of the coast remained under Greek authority. Following the battle of Corupedium in 281 BC, Mithridates successor, Ariobarzanes, expanded the realm of Pontus to include the city of Amastris, and with it, control over the Black Sea coast (McGing 1986a: 18; Cohen 1995: 383). After its conquest by Pharnaces in 183 BC, the capital of the kingdom of Pontus was transferred from Amaseia to Sinop, strengthening the latter’s central position in the trade networks of the second and first centuries BC (McGing 1986a: 23-7, 39), which, however, also attracted Roman imperial attention and military presence (Koromila 1991: 198-9). Nicomedes V of Bithynia invaded Pontus in 89 BC with Roman military support but was defeated by Mithridates VI Eupator, who proceeded to enlarge his territories after his victory (McGing 1986a: 68-109). The Romans, who saw Mithridates as a threat to their interests in Anatolia, attempted to subjugate him, but Sulla and Lucullus could not achieve a decisive victory. Only after Pompey defeated Mithridates in 65 BC could the Romans achieve some degree of control over the Black Sea coast (Magie 1950: 341; Wellesley 1953; McGing 1986a: 79-169; Koromila 1991: 199; Cohen 1995: 383; Højte 2006: 22; Arslan 2007: 340-88). Mithridates VI Eupator’s successor, Pharnaces II, revolted against Roman rule, but was defeated by Julius Caesar in the battle of Zela in 47 BC (Adcock 1937; McGing 1986a: 166-78; Goldsworthy 2006: 446-7).
9.2 Hellenistic Settlement Patterns Ancient Greek sources focus almost exclusively on Greek settlements along the Black Sea coast and largely lack information about native communities and settlements.
248
The Cide-Şenpazar Region during the Hellenistic Period (325/300 - 1 BC)
Greek colonists involved in maritime trade favoured coastal and urban locations, while local Anatolian communities may well have preferred sheltered positions in the coastal hinterland, which could be more easily defended, and from where local resources could be exploited and traded (Hind 1984; 1993). Based on Greek sources, the Black Sea trade route which started from Amastris/ Sesamos passed along the harbour towns of Eryt(r)inoi, Kromna, Kytoros, Aigialos, Klimax, Thymena/Teuthrania, Timolaion, Karambis, Marsylla/Kallistratia, Zephyrion, Garios, Abonouteichos/Ionopolis, Aiginetes, Kinolis, Antikinolis, Koloussa, Stephane, Potamoi and finally reached Sinop (Talbert 2000: 86). During the Hellenistic period, Sesamos, which became Amastris, modern Amasra (Cohen 1995: 383) functioned as the acropolis of Amastris following the unification of the polis (Strab. 12. 3. 10; Avram et. al. 2004: 960). Kytoros, which can be equated with the natural harbour of Gideros, according to Strabo, was a Sinopean marketplace and trading zone (emporion) (Strab. 12. 3. 10; Burstein 1976: 76, 79; Cohen 1995: 383; Mørkholm 1991: 95-6; Avram et. al. 2004: 925-6). As a result, some scholars see Kytoros as a Sinopean foundation (Doonan 2006: 49), but it may well have been founded earlier by Milesian colonists (Boardman 1973: 236-53; Tsetskhladze 1994; 1998a: 9-68; Avram et. al. 2004: 955; Greaves 2002; Whitmarsh 2010: 93). Strabo lists Kytoros as one of four katoikiai or colonial settlements in the area, which were united to form the polis of Amastris (Strab. 12. 3. 10). There is some disagreement about whether or not Kytoros was a polis already before this unification (Strab. 12. 3. 10; Ps-Skylax 90), but based on its earlier relations with Sinop, this would seem to be the case (Avram et. al. 2004: 959).
9.3 Trade and Economy Hellenistic Kytoros would have been oriented primarily towards the sea and tied into the extensive maritime trade networks of the time (for an illustration of their extent, see Archibald 2004: 1-2; 2007: 253-4) rather than the Anatolian interior, whose mountainous character would have further hindered communication. Kytoros and other small coastal settlements would have been closely connected with the major trading centres of Amastris and Sinop, and would have brought into circulation their own local products through these networks. Although the southern and eastern Black Sea coast was an important source of metals, including iron, silver and copper, the main commodity traded from the emporion of Kytoros was boxwood, and even today the mountains of Cide are covered with these trees (Smith 1870: 739; Rostovtzeff 1941: 572-3; De Jesus 1978; Griffith 1983: 124-6; Ersal 1994: 28; Avram et. al. 2004: 959; Koçak 2006; Hannestad 2007). Strabo, for instance, mentions that the boxwood trees of Amastris were of the highest quality and that Kytoros was an important source for this popular material for furniture and tools (Strab. 12. 3. 10; Texier 1862: 622-3; McGing 1986a: 1; Şenol 2006: 40; Arslan 2007:
Archaeological Evidence from Cide and Şenpazar
249
52, note 232). Apollonius of Rhodes (Apollon 2. 294) also mentions that Kytoros was well-known for its trees. Slave trading may have been another important part of the local economy during the Hellenistic period (Finley 1999: 70-79; Avram et al. 2004: 959; Gabrielsen 2005: 391). A slave market is attested at Tanais on the northern coast of the Black Sea, and an increase in piracy in the Aegean and the eastern Mediterranean suggests a heightened demand for slaves in this period (Strab. 11. 2. 3; Tsetskhladze 1998b: 67-8). The degree to which the Turkish Black Sea coast partook in this slave trade is hard to gauge, but a stele from the second half of the fourth century BC of a Paphlagonian slave, who worked at the Laurion mines in Attica, suggests some involvement (Avram 2007: 245; Gabrielsen 2007: 315; Tsetskhladze 2008). Stamped amphorae and coins can provide information about the economic and political importance of Hellenistic settlements. Stamped amphorae were produced in Heraclea, Amastris and Sinop between the fourth and the second century BC (Tezgör Kassab 1995, 1996; Şenol 2006: 35-47). No workshops manufacturing these amphorae have been located to date in the region of Amastris, and there is no evidence that Kytoros produced its own stamped amphorae (Şenol 2010: 126). We do know, however, that the nearby towns of Sesamos, Cromna and Tieum minted their own coins during the fourth century BC and - enjoying some level of autonomy - continued to do so after unification under Amastris (Smith 1870: 118; Mørkholm 1991: 95-6; Erciyas 2006a: 34). Kytoros did not mint its own coins, which suggests that the settlement was of a rather modest size and economically closely tied to Amastris (Smith 1870: 118; McGing 1986a: 1; Hill 1990: 82).
9.4 Archaeological Evidence from Cide and Şenpazar As for the preceding Iron Age (chapter 8), evidence for the Hellenistic period in the Cide region is very limited (Düring and Glatz 2010; Glatz et al. 2011; Düring et al. 2012). Hellenistic pottery and other materials were found mainly around Okçular Kale (S22) and the neighbouring Abdulkadir valley in the immediate coastal hinterland (Fig. 9.1). There are hints of a Hellenistic presence also inland at Çamdibi (S95) and Aybasan. Along the coast, locations near Kalafat, Gideros and the Gökçekale promontory have yielded evidence for a Hellenistic presence (section 10.3.2). Diagnostic pottery types include bowls, jars, pithoi and amphora fragments (Fig. 9.2). Just over half of this assemblage is made up of medium fine fabrics, the remainder is split between fine and coarse wares. Most are well fired and fabric colours range from light red to brown, and beige or grey (Fig. 9.3). The fabrics and the quality of the Hellenistic pottery from Cide are quite distinct from those common in the Aegean and the Mediterranean. During the Hellenistic period, very fine table-wares were in use at settlements along the Black Sea coast
250
The Cide-Şenpazar Region during the Hellenistic Period (325/300 - 1 BC)
0
5
10 Km
Kuşçu (Kurucal Tepe)
Gideros Kalesi
Abdulkadir Okçular
Fig. 9.1: Map showing the locations of Hellenistic-period finds. Produced by Toby C. Wilkinson.
and their hinterlands (Domżalski 2007: 161-81). Aegean and Mediterranean table wares or local imitations are reported from sites in the lower Dnieper (Bylkova 2006: 217-47; Karjaka 2007: 133-41) and the lower Dniester regions (Ochotnikov 2006: 81-98), settlements in the Crimea, such as Chersonesos (Zolotarev 2006: 193216) and Panskoe I (Hannestad 2006: 179-92), and Scythian Neapolis near modern Simferopol (Zajcev 2006: 259-73), and at Olbia on the Bug estuary in Ukraine (Krapivina 2006: 249-57). Such table wares are not attested in the CAP survey record. There are, however, several examples of Hellenistic to Roman period amphorae (section 10.2.2), which suggest the region’s participation in Circumpontic maritime networks of interaction. Based on their formal characteristics and fabrics, several roof tiles can also be assigned a tentative date in the Hellenistic period (Fig. 9.4). Some of these were found at Kuşçu2 (S10), which is located approximately 15km northeast of the modern town of Cide, some at Kuruçalı Tepesi3 (S16) in the near-coastal location of the Abdulkadir valley and close to the Gideros bay (see chapter 14), and one4 comes from Okçular valley (U89).
2 C5, C7 and C50 (Fig. 9.3). 3 C62 and C63 (Fig. 9.3). 4 C506 (Fig. 9.3).
Archaeological Evidence from Cide and Şenpazar
251
Fig. 9.2: Hellenistic pottery from Cide-Şenpazar.5 Produced by Lorraine McEwan.
5 C1222, C2852, C2853 and C3292 were found in the Okçular valley, C929, C1774, C1893, C2204 and C2207 were found in the vicinity of Okçular Kale, C2172 and C2244 were found at the northern edge of Abdulkadir valley.
252
The Cide-Şenpazar Region during the Hellenistic Period (325/300 - 1 BC)
Fig. 9.3: Hellenistic pottery from Cide-Şenpazar. Produced by Caner Bakan.
In the light of historical sources, which connect the Turkish Black Sea region, and Kytoros/Gideros more specifically, with the Greek and Hellenistic worlds of the first millennium BC, as well as abundant archaeological evidence from neighbouring regions, the scarcity of archaeological materials from Cide is rather striking. As with other underrepresented periods and materials in the survey region, poor ground visibility caused by dense vegetation may well have prevented the collection of a largeenough surface sample to identify more confidently Hellenistic period settlement and activity. In addition, the thick layer of alluvium, which accumulated along the coast over millennia, may have buried Hellenistic settlements along the littoral. However,
Archaeological Evidence from Cide and Şenpazar
253
Fig. 9.4: Possible Hellenistic roof tiles from Cide-Şenpazar. Produced by Lorraine McEwan.
the low quantities of Hellenistic material, could equally be reflective of a real scarcity of settlement in the Cide region, or, more likely, low numbers of individuals partaking in behaviours and the use of material culture, which we have come to recognise as ‘Hellenistic’ in the archaeological record. Again, as in many other periods, we know little about local cultural traditions and their long-term development (chapters 7 and 14). Surveys along the Turkish Black Sea region have recorded numerous settlements dating to this period, but the quantities of unequivocal Hellenistic pottery finds are consistently low even at sites where Iron Age and Roman ceramics are abundant (Dönmez 2000; Doonan et. al. 2000, 2001, 2008; Bilgi et. al. 2002; Özsait 2004; Doonan and Bauer 2005). All this may mean that pottery from the Cide region, and from elsewhere along the Turkish Black Sea coast, which are assigned dates in the Iron Age or the Roman period, or indeed have gone unrecognised altogether, may actually date to the final third of the first millennium BC. Clearly more research is needed to better understand the Cide assemblage.
254
The Cide-Şenpazar Region during the Hellenistic Period (325/300 - 1 BC)
9.5 Conclusion Historical sources would suggest that during the final centuries of the first millennium BC, the areas surrounding the Black Sea underwent a process of cultural transformation that is referred to as ‘Hellenisation’ in much of the literature, and that denotes the adaptation and incorporation by local populations of Greek practices and elements of material culture (Bouzek 1990: 10; Tsetskhladze 2002: 83). For instance, Hellenistic-period finds unearthed at the city of Vani, ancient Colchis, in modern Georgia, and at many other settlements around the Sea of Azov, in the lower Don region and in Thrace, testify to the popularity of Greek artistic styles (Kern 1955; Tsetskhladze 1999: 469-96; 2002: 86; Cholakov 2006). Another example of this process is the incorporation of Greek stylistic elements into Scythian anthropomorphic gravestones and graves (Tsetskhladze 2002: 87, note 38; Erciyas 2006b: 229). Cide, and more specifically Kytoros - at least according to historical sources - became increasingly tied into the wider political and economic networks of the Black Sea when it was unified with several other coastal towns to form the polis of Amastris and became an important source of boxwood. However, unless important Hellenistic-period settlements lay now buried in the coastal plains, the archaeological evidence from Cide suggest only limited settlement and activity of a recognisably Greek or Hellenistic character during this period. As might be expected, the evidence that we do have, points to a focus of activity in the coastal region and its immediate hinterland. Important questions concerning local cultural and social developments, the relationships of native populations with colonial settlements and settlers, and the region’s involvement in Black Sea trade await future investigations.
Catalogue of Illustrated Artefacts Ceramics C929. U521. Rim. Brownish red fine vessel with black and red grit and limestone inclusions, and some voids. Light reddish brown burnished on the exterior. Oxidised firing. Wheelmade. Exterior surface 5 YR 5/4, interior surface 5 YR 5/6. D= 24cm. C1222. G6F15. Rim. Buff orange fine vessel with white, black, and red grit. Smoothed on both sides. Oxidised firing. Wheelmade. Exterior surface 5 YR 7/8, interior surface 5 YR 7/8. D= 7cm. C1774. G1A9. Rim. Light red coarse vessel with a well sorted paste and black mineral inclusions up to 1mm in size. No surface treatment. Oxidised firing. Wheelmade. Exterior surface 2.5 YR 7/8, interior surface 2.5 YR 7/8. D= unknown. C1893. G1D16. Rim. Red medium quality vessel with a well sorted powdery paste and quartz inclusions smaller than 0.5mm. No surface treatment. Oxidised firing. Wheelmade. Exterior surface 2.5 YR 5/6, interior surface 2.5 YR 5/6. D= 20cm.
Bibliography
255
C2172. U288. Rim. Reddish yellow very coarse vessel with dense small mineral inclusions. No surface treatment. Oxidised firing. Handmade. Exterior surface 5 YR 6/8, interior surface 5 YR 6/8. D= 38cm. C2204. U593. Rim. Light red fine vessel with a grey core, very small grey mineral inclusions, calcareous specks, and few grit. No surface treatment. Oxidised firing. Wheelmade. Exterior surface 2.5 YR 6/6, interior surface 2.5 YR 6/6. D= 18cm. C2207. U593. Rim. Light red very coarse vessel with dense very small mineral inclusions, calcareous specks, grit, and chaff temper. No surface treatment. Oxidised firing. Handmade. Exterior surface 2.5 YR 6/6, interior surface 2.5 YR 6/6. D= 40cm. C2244. U325. Base. Light beige medium quality vessel with small white mineral inclusions. No surface treatment. Oxidised firing. Handmade(?). Exterior surface 5 YR 6/6, interior surface 5 YR 6/6. D= unknown. C2852. U609. Rim. Light red medium to fine quality vessel with small mineral inclusions. No surface treatment. Oxidised firing. Wheelmade. Exterior surface surface 2.5 YR 7/8, interior surface 2.5 YR 7/8. D= 8cm. C2853. U655. Rim and handle. Light red fine vessel with small black mineral inclusions. No surface treatment. Oxidised firing. Wheelmade. Exterior surface 10 R 6/8, interior surface 10 R 6/8. D= 34cm. C3292. U554. Rim. Pinkish buff medium to fine vessel with small limestone inclusions, grog and grit. No surface treatment. Oxidised firing. Wheelmade. Exterior surface 7.5 YR 7/4, interior surface 7.5 YR 7/4. D= 10cm. Roof Tiles C5. S10. Tile. Light reddish brown coarse tile with small to large red grit and vegetable temper. Light reddish brown slipped. Oxidised firing. Handmade. Exterior surface 7.5 YR 8/3, interior surface 7.5 YR 8/3. C7. S10. Tile. Orange coarse tile with medium size grit and vegetable temper. No surface treatment. Oxidised firing. Handmade. Exterior surface 5 YR 6/8, interior surface 5 YR 6/8. C50. S10. Tile. Orange very coarse tile with medium to large size red and black grit. No surface treatment. Oxidised firing. Handmade. Exterior surface 5 YR 6/8, interior surface 5 YR 6/8. C62. S16. Tile. Light red coarse tile with medium to large size red grit and limestone specks. No surface treatment. Oxidised firing. Handmade. Exterior surface 10 YR 6/8, interior surface 10 YR 6/8. C63. S16. Tile. Light red coarse tile with medium to large size red grit, limestone specks and grog. No surface treatment. Oxidised firing. Handmade. Exterior surface 10 R 7/6, interior surface 10 R 7/6. C506. U89. Tile. Pale red very coarse tile with large dark brown grit and grog. No surface treatment. Oxidised firing. Handmade. Exterior surface 10 R 6/8, interior surface 2.5 YR 7/6.
Bibliography Greco-Roman sources Anonymi, Periplus Ponti Euxini. [Edited by H. Dodwell, F. Osanni, and A.G.R. Naumann: Periplus ponti Euxini: Anonymi Periplus ponti Euxini, qui Arriano falso adscribitur. Anonymi Periplus ponti Euxini et Maeotidis paludis. Anonymi Mensura ponti Euxini. Agathemeri Hypotyposes geographiae. Fragmenta duo geographica.] Leipzig, C.L. Fritzsche, 1842. Apollonius of Rhodes, Argonautica. [Translated by G.W. Mooney: The Argonautica of Apollonius Rhodius.] Dublin, Hodges, Figgis, & Co., 1912.
256
The Cide-Şenpazar Region during the Hellenistic Period (325/300 - 1 BC)
Arrian, Periplus Ponti Euxini. [Edited by A. Liddle: Arrian: Periplus Ponti Euxini.] Bristol, Bristol Classical Press, 2003. Homer, Iliad. [translated by A.T. Murray: The Iliad.] Cambridge / London, Harvard University Press / William Heinemann, 1924. Pomponius Mela, De Situ Orbis. [Edited by K.H. Tzschucke: De Situ Orbis.] Leipzig, Siegfried Lebrecht Crusius, 1807. Pseudo-Scylax, Periplus Ponti Euxini. [Translated by G. Shipley: Pseudo-Skylax’s Periplous: The Circumnavigation of the Inhabited World - Text, Translation and Commentary.] Liverpool, Liverpool University Press, 2011. Ptolemy, Geographia. [Edited by C. Müller and C.T. Fischer: Claudii Ptolemaei Geographia.] Paris, Alfred Firmin-Didot, 1901. Strabo, Geographia. [Translated by H.L. Jones: The Geography of Strabo.] Cambridge / London, Harvard University Press / William Heinemann, 1924. Stephanus of Byzantium, Peri Poleon. [Edited by J. Gronow and T. De Pinedo: De Urbibus.] Amsterdam, Rud. & Gerh. Wetstenios, 1725. published sources Adcock, F. E. 1937. Lesser Armenia and Galatia after Pompey’s settlement of the east. The Journal of Roman Studies 27: 12-7. Archibald, Z. H. 2004. In-groups and out-groups in the Pontic cities of the Hellenistic age. In: C.J. Tuplin, ed., Pontus and the Outside World, Studies in Black Sea History, Historiography and Archaeology. Leiden, Brill: 1-15. — 2007. Contacts between the Ptolemaic kingdom and the Black Sea in the early Hellenistic age. In V. Gabrielsen and J. Lund, eds., The Black Sea in Antiquity: Regional and Interregional Economic Exchanges. Aarhus, Aarhus University Press: 253-71. Arslan, M. 2007. Roma’nın Büyük Düşmanı Mithradates VI Eupator. Istanbul, Odin Yayınları. Ashley, J. R. 2004. The Macedonian Empire: The Era of Warfare Under Philip II and Alexander the Great, 359-323 B.C. Jefferson, McFarland. Avram, A. 2007. Some thoughts about the Black Sea and the slave trade. In V. Gabrielsen and J. Lund, eds., The Black Sea in Antiquity: Regional and Interregional Economic Exchanges. Aarhus, Aarhus University Press: 239-51. Avram, A., Hind, J., and Tsetskhladze, G. 2004. The Black Sea area. In M.H. Hansen and T. H. Nielsen, eds., An Inventory of Archaic and Classical Poleis. Oxford, Oxford University Press: 924-73. Bilgi, Ö., Atasoy, S., Dönmez, Ş., and Summerer, L. 2002. Samsun (Amisos) Bölgesinin Kültürel Gelişimi Projesi ile İlgili Yüzey Araştırması 2000. Araştırma Sonuçları Toplantısı 19/1: 279-96. Boardman, J. 1973. The Greeks Overseas: The Archaeology of Their Early Colonies and Trade. Harmondsworth, Penguin. Bosworth, A. B. 2006. Alexander the Great part I: The events of the reign. In D.M. Lewis, J. Boardman, S. Hornblower, and M. Ostwald, eds., The Cambridge Ancient History, IV, The Fourth Century B.C. Cambridge, Cambridge University Press: 791-845. Bosworth, A. B. and Wheatley, P. V. 1998. The origins of the Pontic house. The Journal of Hellenic Studies 118: 155-64. Bouzek, J. 1990. Studies of Greek Pottery in the Black Sea Area. Prague, Charles University. Burstein, S. M. 1976. Outpost of Hellenism: The Emergence of Heraclea on the Black Sea. Berkeley, University of California Press. Bylkova, V. P. 2006. The chronology of settlements in the Lower Dnieper region (400-100 BC). In V.F. Stolba and L. Hannestad, eds., Chronologies of the Black Sea Area in the Period c. 400-100 BC. Aarhus, Aarhus University Press: 217-47. Carpenter, R. 1948. The Greek penetration of the Black Sea. American Journal of Archaeology 52: 1-10.
Bibliography
257
Cholakov, I. M. 2006. Bronze bust of Maenad from southeast Bulgaria. Journal of the Serbian Archaeological Society 22: 265–80. Cohen, G. M. 1995. The Hellenistic Settlements in Europe, the Islands, and Asia Minor. Berkeley / Oxford, University of California Press. de Boer, J. G. 2006. The Greek colonies in the Pontic area during the 5th century BC, Athens, the rise of the ‘barbarian’ kingdoms and the Ionian revolt: An economic approach. Talanta: Proceedings of the Dutch Archaeological and Historical Society 36-7: 269-87. de Jesus, P. S. 1978. Metal resources in ancient Anatolia. Anatolian Studies 27: 97-102. Domżalski, K. 2007. Changes in Late Classical and Hellenistic fine pottery production in the eastern Mediterranean as reflected by imports in the Pontic area. In V. Gabrielsen and J. Lund, eds., The Black Sea in Antiquity: Regional and Interregional Economic Exchanges. Aarhus, Aarhus University Press: 161-81. Dönmez, Ş. 2000. Sinop-Samsun-Amasya İlleri 1998 Yüzey Araştırması. Araştırma Sonuçları Toplantısı 17/2: 229-44. Doonan, O. 2006. Exploring community in the hinterland of a Black Sea Port. In P.G. Bilde and V.F. Stolba, eds., Surveying the Greek Chora: The Black Sea Region in a Comparative Perspective. Aarhus, Aarhus University Press: 47-58. Doonan, O., and Bauer, A. 2005. Sinop province archaeological project: Report on the 2003 field season. Araştırma Sonuçları Toplantısı 22/2: 275-84. Doonan, O., Casson, A., and Gantos, A. 2008. Sinop province archaeological project: Report on the 2006 field season. Araştırma Sonuçları Toplantısı 25/2: 133-50. Doonan, O., Gantos, A. J., Hiebert, F. J, Besonen, M., and Watters, M. 2001. Sinop ve Çevresi Arkeolojik Yüzey Araştırması. Araştırma Sonuçları Toplantısı 18/1: 137-48. Doonan, O., Gantos, A. J., Hiebert, F. J., Yaycıoğlu, A., and Besonen, M. 2000. Sinop Bölgesel Arkeoloji Araştırması 1998: Karasu Vadi Araştırması. Araştırma Sonuçları Toplantısı 17/2: 345-56. Drews, R. 1976. The earliest Greek settlements on the Black Sea. The Journal of Hellenic Studies 96: 18-31. Düring, B. S., and Glatz, C. 2010. The Cide Archaeological Project 2009: First results. Anatolia Antiqua 18: 203-13. Düring, B. S., Glatz, C., and Şerifoğlu, T. E. 2012. The Cide Archaeological Project 2011: Third preliminary report. Anatolia Antiqua 20: 167-75. Erciyas, D. B. 2006a. Wealth, Aristocracy and Royal Propaganda Under the Hellenistic Kingdom of the Mithradatids in the Central Black Sea Region of Turkey.Leiden, Brill. — 2006b. The Black Sea region in the Hellenistic period: The Pontic kingdom, its settlements, monuments and coins. In D.B. Erciyas and E. Koparal, eds., Karadeniz Araştırmaları Sempozyum Bildirileri. Istanbul, Ege Yayınları: 219-30. Ersal, Y. 1994. Antik Çağdan Günümüze Karadeniz Ekonomisi. Ankara, Türkiye Odalar ve Borsalar Birliği. Finley, M. I. 1999. The Ancient Economy. Berkeley, University of California Press. Gabrielsen, V. 2005. Piracy and the slave-trade. In A. Erskine, ed., A Companion to the Hellenistic World. Oxford, Blackwell: 389-404. — 2007. Trade and tribute: Byzantion and the Black Sea straits. In V. Gabrielsen and J. Lund, eds., The Black Sea in Antiquity: Regional and Interregional Economic Exchanges. Aarhus, Aarhus University Press: 287-324. Glatz, C., Düring, B. S., Şerifoğlu, T. E. 2011. The Cide Archaeological Project 2010: Second preliminary report. Anatolia Antiqua 19: 279-88. Goldsworthy, A. 2006. Caesar: Life of a Colossus. New Haven / London, Yale University Press / Weidenfeld and Nicolson. Greaves, A. M. 2002. Miletos: A History. London, Routledge.
258
The Cide-Şenpazar Region during the Hellenistic Period (325/300 - 1 BC)
Griffith, J. G. 1983. Catullus, poem 4: A neglected interpretation revived. Phoenix 37(2): 123-28. Hannestad, L. 2006. The dating of the monumental building U6 at Panskoe I. In V.F. Stolba and L. Hannestad, eds., Chronologies of the Black Sea Area in the Period c. 400-100 BC. Aarhus, Aarhus University Press: 179-92. — 2007. Timber as a trade resource of the Black Sea. In V. Gabrielsen and J. Lund, eds., The Black Sea in Antiquity: Regional and Interregional Economic Exchanges, Aarhus, Aarhus University Press: 85-99. Heckel, W. and Yardley, J. C. 2004. Alexander the Great: Historical Texts in Translation. Oxford, Blackwell. Hill, S. J. 1990. Preliminary survey at Amasra, Zonguldak. Araştırma Sonuçları Toplantısı 7: 81-6. Hind, J. G. F. 1984. Greek and barbarian peoples on the shores of the Black Sea. Archaeological Reports (The Society for the Promotion of Hellenic Studies) 30: 71-97. — 1993. Archaeology of the Greeks and barbarian peoples around the Black Sea (1982-1992). Archaeological Reports (The Society for the Promotion of Hellenic Studies) 39: 82-112. Højte, J. M. 2006. From kingdom to province: Reshaping Pontos after the fall of Mithridates VI. In T. Bekker-Nielsen, ed., Rome and the Black Sea Region: Domination, Romanisation, Resistance. Aarhus, Aarhus University Press: 15-30. Karjaka, A. V. 2007. Amphora finds of the 4th century BC from the settlements of the Lower Dnieper Region. In V. Gabrielsen and J. Lund, eds., The Black Sea in Antiquity: Regional and Interregional Economic Exchanges. Aarhus, Aarhus University Press: 133-41. Kern, J. H. C. 1955. Three relief-plaques from Olbia (South Russia). Mnemosyne, Fourth Series 8(3): 219-21. Koçak, Ö. 2006. Mining at the central Black Sea region in the ancient period. In D. B. Erciyas and E. Koparal, eds., Karadeniz Araştırmaları Sempozyum Bildirileri. Istanbul, Ege Yayınları: 39-62. Koromila, M. 1991. The Greeks in the Black Sea from the Bronze Age to the Early Twentieth Century. Athens, Panorama. Krapivina, V. V. 2006. Problems of the chronology of the Late Hellenistic strata of Olbia. In V.F. Stolba and L. Hannestad, eds., Chronologies of the Black Sea Area in the Period c. 400-100 BC. Aarhus, Aarhus University Press: 249-57. Magie, D. 1950. Roman Rule in Asia Minor to the End of the Third Century After Christ. Princeton, Princeton University Press. McGing, B. C. 1986a. The Foreign Policy of Mithridates VI Eupator King of Pontus. Leiden, Brill. — 1986b. The kings of Pontus: Some problems of identity and date. Rheinisches Museum für Philologie, Neue Folge 129(3/4): 248-59. Mitchell, S. 2004. Doğu ve batı arasında Anadolu Hellenistik çağ’da Attalos ve Mithridates krallıklarının paralel yaşamları. Anadolu Araştırmaları 17: 140-53. — 2010. The Ionians of Paphlagonia. In T. Whitmarsh, ed., Local Knowledge and Microidentities in the Imperial Greek World. Cambridge, Cambridge University Press: 86-110. Mørkholm, O. 1991. Early Hellenistic Coinage from the Accession of Alexander to the Peace of Apamea (336-186 B.C.). Cambridge, Cambridge University Press. Ochotnikov, S. B. 2006. The chorai of the ancient cities in the Lower Dniester area (6th century BC3rd century AD). In P.G. Bilde and V.F. Stolba, eds., Surveying the Greek Chora: The Black Sea Region in a Comparative Perspective. Aarhus, Aarhus University Press: 81-98. Özsait, M. 2004. 2002 Yılı Samsun-Amasya yüzey araştırmalarının: İlk sonuçları. Araştırma Sonuçları Toplantısı 21/2: 273-84. Rostovtzeff, M. 1941. The Social and Economic History of the Hellenistic World. Oxford, Oxford University Press. Şenol, G. C. 2006. Klasik ve Hellenistik Dönem’de Mühürlü Amphora Üreten Merkezler ve Mühürleme Sistemleri. İstanbul, Ege Yayınları.
Bibliography
259
— 2010. Amphora üretim merkezleri ışığında Hellenistik dönemde Anadolu’da şarap üretiminin izleri. In Ü. Aydınoğlu and A. K. Şenol, eds., Antik Çağda Anadolu’da Zeytinyağı ve Şarap Üretimi. Istanbul, Ege Yayınları: 123-34. Smith, W. 1870. Dictionary of Greek and Roman Geography I: Abacaenum-Hytanis. Boston, Little and Brown. Stolba, V. F. 2003. Some reflections on the amphora stamps with the name of Amastris. In P. Guldager Bilde, J.M. Højte and V.F. Stolba, eds., The Cauldron of Ariantas. Studies Presented to A.N. Sčeglov on the Occasion of his 70th Birthday. Aarhus, Aarhus University Press: 279-301. Talbert, R. J. A. 2000. Barrington Atlas of the Greek and Roman World. Princeton, Princeton University Press. Texier, F. M. C. 1862. Asie Mineure: Description géographique, historique et archéologique des provinces et des villes de la Chersonnèse d’Asie. Paris, Didot. Tezgör Kassab, D. 1995. Sinop’ta seramik ve anfora atölyelerinin incelenmesi. Araştırma Sonuçları Toplantısı 12/2: 177-90. — 1996. Demirci-Sinop anfora atölyelerinin kazısı. Kazı Sonuçları Toplantısı 17/2: 277-91. Tsetskhladze, G. R. 1994. Greek penetration of the Black Sea. In G.R. Tsetskhladze and F. De Angelis, eds., The Archaeology of Greek Colonisation: Essays Dedicated to Sir John Boardman. Oxford, Oxford University School of Archaeology: 111-35. — 1997. Argonautica, Colchis and the Black Sea: Myth, reality and modern scholarship. Thracia Pontica 6(1): 337-47. — 1998a. Greek colonisation of the Black Sea area: Stages, models, and native population. In G. R. Tsetskhladze, ed., The Greek Colonisation of the Black Sea Area: Historical Interpretation of Archaeology. Stuttgart, Franz Steiner Verlag: 9-68. — 1998b. Trade on the Black Sea in the Archaic and Classical periods: Some observations. In H.M. Parkins and C.J. Smith, eds., Trade, Traders and the Ancient City. London, Routledge: 52-74. — 1999. Between west and east: Anatolian roots of local cultures of the Pontus. In G.R. Tsetskhladze, ed., Ancient Greeks West and East. Leiden, Brill: 469-96. — 2002. Ionians abroad. In G.R. Tsetskhladze and A.M. Snodgrass, eds., Greek Settlements in the Eastern Mediterranean and the Black Sea. Oxford, Archaeopress: 81-96. — 2008. Pontic slaves in Athens: Orthodoxy and reality. In P. Mauritsch, W. Petermandl, R. Rollinger and C. Ulf, eds., Antike Lebenswelten. Konstanz- Wandel- Wirkungsmacht. Festschrift für Ingomar Weiler zum 70. Geburtstag. Wiesbaden, Otto Harrassowitz Verlag: 309-19. Wellesley, K. 1953. The extent of the territory added to Bithynia by Pompey. Rheinisches Museum für Philologie, Neue Folge 96(4): 293-318. Whitmarsh, T. 2010. Local Knowledge and Microidentities in the Imperial Greek World. Cambridge, Cambridge University Press. Zajcev, J. P. 2006. Absolute and relative chronology of Scythian Neapolis in the 2nd century BC. In V.F. Stolba and L. Hannestad, eds., Chronologies of the Black Sea Area in the Period c. 400-100 BC. Aarhus, Aarhus University Press: 259-73. Zolotarev, M. I. 2006. A Hellenistic ceramic deposit from the north-eastern sector of Chersonesos. In V.F. Stolba and L. Hannestad, eds., Chronologies of the Black Sea Area in the Period c. 400-100 BC. Aarhus, Aarhus University Press: 193-216.
Philip Bes
10 The Cide-Şenpazar Region in the Roman Period Turkey’s northern coast is often perceived as peripheral from a Mediterranean perspective. Apart from the well-documented connections, economic and other, with the Aegean and Mediterranean, the region was, however, also culturally part of the Black Sea basin as well as inland Asia Minor (Braund 2005: 115, 117). The Black Sea basin and northwestern Asia Minor have long attracted the attention of scholars, particularly in more recent years (for instance Knipowitsch 1929; Asheri and Hoepfner 1972; Marek 1993; Belke 1996; Gabrielsen and Lund 2007; Kassab Tezgör 2010; Tsetskhladze 2012). Fairly little is known, however, from Turkey’s Black Sea region in terms of ceramic assemblages dating to the Hellenistic and Roman periods, although more and more data are emerging in published form. This chapter discusses the Roman-period data, ceramic and architectural, that were gathered by the Cide Archaeological Project (CAP); the Roman period was a particular focus during the third and final year of fieldwork in 2011 (Düring et al. 2012). These artefactual data provide the basis for the exploration of spatial and chronological patterns of occupation within the study area, as well as the reconstruction, however tentative, of aspects of local economic and socio-cultural life. Given the scattered and restricted nature of the collected evidence, these matters cannot be investigated in any great detail, yet fourteen find clusters nonetheless offer indications that the Cide-Şenpazar region was fairly urbanised and relatively densely occupied during the (Late) Roman period. Artefactual data also offer clues for contacts with other parts of the Pontic, the Aegean and the wider Mediterranean. An important archaeological anchor point for northern Turkey is Sinop, wellknown for the manufacture of amphorae during the Hellenistic and Roman periods, which were widely — if somewhat thinly — distributed throughout the central and eastern Mediterranean (Pieri 2007: 8-9; Gabrielsen and Lund 2007; Kassab Tezgör 2010; Reynolds 2010; de Boer 2013). Fragments of Sinopean amphorae are now also attested in Boeotia (central Greece), Sagalassos (southwest Turkey) and Horvat Kur (northern Israel) (pers. obs.). Constantinople, capital of the eastern Roman empire from AD 330 onwards and the thoroughfare between the Aegean, and by extension the Mediterranean, and the Black Sea, remains comparatively poorly investigated, apart from the excavations at the church of St. Polyeuktos and sondages beneath the mosaic floors of the Great Palace (Hayes 1992; Bardill and Hayes 2002; Jobst 2005). In the future, results from the recent excavations at Yenikapı and the Theodosian Harbour will add significantly to our understanding of Roman-period ceramic developments. Constantinople and Sinop, thus, serve as points of reference for the analysis of the pottery collected by CAP. The survey area is located about two-thirds along the coast coming from Constantinople toward Sinop, which further allows us to compare the pottery with newly published evidence from Pompeiopolis (Zhuravlev
The Cide-Şenpazar Region in the Roman Period
261
2011; Domżalski 2011), Hadrianopolis (Laflı and Kan Şahin 2012a, b; Laflı and Christof 2012), and the Paphlagonia survey (Matthews and Glatz 2009) (Fig. 10.1).
Fig. 10.1: Roman-period sites in the Black Sea area mentioned in the text. Produced by Toby C. Wilkinson.
In total 211 ceramic fragments of Hellenistic, but above all of Roman to Late Roman date were identified, representing ca. 5% of the total of 4,125 sherds collected. This comparatively modest signal for activity and occupation during (mainly) the Roman imperial period in the study area, which is partly complemented by the architectural finds (see below), means that the interpretations presented here should be considered with due caution. Below, the pottery will first be discussed by period and functional class, followed by an evaluation within its regional context in terms of ‘local’‘regional’-‘import’, and will conclude with a spatial-chronological discussion of both ceramic and architectural evidence. The chronological ranges used below differ from those used initially by CAP (Düring and Glatz 2010; Düring et al. 2012: table 1), and a periodisation more common in Hellenistic and Roman pottery studies is used (table 10.1). The fact that not every pottery fragment can be chronologically attributed to a single period poses significant challenges for survey archaeology. To some extent this also applies to the architectural fragments, which I studied from photographs. As a result, some of the artefacts presented in this chapter will also feature in preceding and following discussions (chapter 9 and 11).
262
The Cide-Şenpazar Region in the Roman Period
Tab. 10.1: Chronological framework used in this chapter. period
approximate dates
Hellenistic
325/300-1 BC
Early Roman
AD 1-200
Mid-Roman
AD 200-400
Late Roman (-Early Byzantine)
AD 400-650/700
10.1 Historical Background Here a short historical introduction will be provided that will allow us to situate our archaeological data in the broader administrative and socio-cultural developments of the period (Jones 1937: 148-74; Magie 1950: chapters 8, 13, 15; Marek 1993, 2009: 35-46). During the Late Republican (Late Hellenistic in an eastern context) period, northwest Asia Minor was divided into several kingdoms. The region did not remain unaffected by, among others, the Mithridatic Wars between Rome and the Pontic kingdom, that were in effect stages of Rome’s eastern expansion. These wars eventually culminated in the defeat of Mithridates and the gradual annexation of these kingdoms, which began with Nikomedes IV bequeathing his kingdom of Bithynia to Rome in 74 BC (Mitchell 1993b: 62). These developments accelerated following Pompey the Great’s defeat of Mithridates VI in 66 BC, and in 63 BC Pontos was joined to Bithynia to form a single provincia (Madsen 2009: 27, 29; McGing 2005: 88; Wesch-Klein 2001). The Paphlagonian kingdom was finally usurped in 6 or 5 BC and joined to the provincia Galatia (Mitchell 1993a: 152; 1993b: 92). The dynamic developments of provincial-administrative boundaries in the north, as elsewhere in Asia Minor during the Late Hellenistic and Early Roman periods, will not be exhaustively explored here (for example Mitchell 1993a: appendix 1). The Late Hellenistic administration was increasingly organised via urban centres in this part of Asia Minor. Even if the region had a distinct rural-administrative character, settlements with urban features were nonetheless significant. On the whole, however, Paphlagonia appears to have been little urbanised around the time it passed into Roman hands. Even if Pompey the Great’s rather exceptional plans, legally laid down in the lex Pompeia, may have been directed more towards administratively reorganising existing settlements and communities rather than founding cities/poleis from scratch, the effects for existing communities must have permeated throughout the region (Madsen 2009: 29-40; Mitchell 1993b: 91, 116, 162; Matthews et al. 2009: 178-9). Existing rural districts, which did not disappear altogether, were grouped and became part of poleis’ territories (Mitchell 1993b: 91). The position of the major urban centres in Paphlagonia during this period, Gangra/Germanikopolis, Hadrianopolis, Neoklaudiopolis, and in particular Pompeiopolis, along an important west-east route contributed to their expansion into significant cities (Mitchell 1993b: 93, 127). This is further underscored by Bithynia’s and
The Pottery — Breakdown by Period and Functional Class
263
Paphlagonia’s logistical importance in providing grain for the eastern legions during the military campaigns of the late second to third centuries AD, if not earlier (Mitchell 1993b: 119-42, 232, 252-3). Bithynia and Pontos continued to form a provincia in the Early Roman period, with coastal centres among others at Heraclea Pontike, Amastris and Sinop. Well-known is Pliny the Younger‘s legateship of the provincia during Trajan’s reign. Paphlagonia at that time was an administrative unit of the large provincia Galatia and Kappadokia, which existed from Flavian to late Trajanic times: the administrative history of Paphlagonia is obscure for the period following the Severan dynasty (Talbert 2000: maps 100-1; Mitchell 1993a: 154-5, 158-9). Certainly by the early fourth century AD, following Diocletian’s provincial reorganisation, Paphlagonia had become a provincia within the diocese of Pontika, though it cannot be ruled out that Paphlagonia had formed a separate provincia between the Severan dynasty and Diocletian’s reforms (Mitchell 1993a: 160). Under Theodosius I, the new provincia Honorias was created, formerly part of Bithynia. This division seems to have continued into the late fifth, early sixth century AD (both being provinciae in the diocese of Pontika), and whilst Honorias was joined to Paphlagonia under Justinian, sometime later in his reign Honorias may have been reinstated as a separate provincia (Talbert 2000: map 102; Mitchell 1993a: 160, 162, map 7). Finally, the tumultuous seventh and earlier eighth centuries AD may have given rise to a shift in both urban and rural settlement patterns. Gangra/Germanikopolis may have been the only larger urban settlement that survived in one form or the other, continuing as such into the Byzantine period. Evidence from parts of Asia Minor suggests a break in existing rural settlement patterns during or following the seventh century AD, and the emergence of a series of ‘fortified hilltop sites’ (Matthews et al. 2009: 190-99, figs. 6.29, 6.35). To answer how urban and rural life was affected by the incorporation in the Roman provincial system is a question beyond the scope of this chapter. We should imagine, however, that for example taxes, laws, new forms in material culture, religious practices (such as Christianity later on), the emergence of the imperial cult, and other changes that the new political system brought with it, did affect — probably to varying degrees within the social hierarchy — the lives of the region’s inhabitants. The incorporation into the Roman republic and subsequently the empire probably resulted, as is observed elsewhere, in an amalgamation of ‘old’ and ‘new’ rather than the ‘new’ replacing the ‘old’. Regardless of the exact nature and effects of these changes, the Black Sea, possibly more tangibly than before, was now connected to the Aegean and Mediterranean areas. The Roman-period pottery collected by CAP, as far as this allows any firm statements, also seems to reflect this.
10.2 The Pottery — Breakdown by Period and Functional Class Observations made in the field were cursory given the limited time available, and were mostly directed towards determining the fabric, type, and/or chronological
264
The Cide-Şenpazar Region in the Roman Period
identification. Methodologically a common approach was used whereby insofar as possible the following characteristics were determined, in some cases incorporating a certain degree of uncertainty: 1. the identification of ware, or fabric. Only macroscopic observations, with the naked eye, were made; 2. shape: the type of fragment (rim, handle, etc.); 3. surface treatment; 4. recognition of function: the (original) purpose for which the vessel was made. A recent synthesis has made it amply clear that, at least theoretically, one should not apply this method too rigidly (Peña 2007); and 5. chronology. This approach permits to capture the data as shown in table 10.2. Tab. 10.2: Absolute counts and relative proportions of the pottery discussed in this chapter, by functional category, and type and/or fabric. functional category
category
n
%
approx. date
Table wares
Phocaean LRC
8
8.53
LROM
Phocaean LRC (/Çandarlı Ware)
3
ROM
Jugs
3
HEL-LROM
Other
4
Local/Regional amphorae/-iskoi
95
Sinopean pâte rosée
4
AD 100-300
Sinopean pâte rouge/colorée
10
AD 300-500
Sinopean pâte blanche/claire
5
AD 500-650
Sinopean
2
AD 100-650
LRA 2
13
LROM
Central North African
4
ROM-LROM
Aegean (Rhodian, Knidian, Chian, Dressel 2-4) Agora M273/Samos Cistern Type
11
HEL-LROM
2
ROM-LROM
Ephesian Nikandros Group
1
CLA-MHEL
Aegean/Black Sea?
1
ROM-LROM
Kapitän II (and related)
5
M-LROM
Zeest 80/Knossos Type 39
2
M-LROM
Unidentified
25
LR Aegean Micaceous Ware Northern Turkey?
1 1
Unidentified
9
Amphora/Cooking vessel
1
0.47
LROM-EBYZ
Small pithos
1
0.47
ROM-LROM
211
100
Amphorae
Cooking Wares
Other Total
CLA-LROM 85.31
(M)ROM-LROM
HEL- (E)BYZ 5.21
LROM ROM-LROM HEL-LROM
The Pottery — Breakdown by Period and Functional Class
265
10.2.1 Table Wares As the name implies, table wares is a ceramic functional category that comprises vessels (open and closed) used for the serving and consumption of food and beverages, supposedly the more representative of the instrumentum domesticum. In addition to shape, the main characteristics of table wares are the presence of slip and/or certain kinds of decoration. Naturally the equation slip = table ware/table ware = slip does not always apply. We should allow for a fair degree of variability, one that was likely dependent on socio-economic context and the desires and means that came with a particular social status. This would have determined the quantity and variety of vessels in use as table ware, but also the materials these were made of. Also, vessels were used as table wares that we cannot or do not recognise as such. Eighteen sherds were classified as table wares, four of which are chronologically assigned to too broad a period to be significant. These include a possible coarse ware echinus (or incurving rim) bowl, if correct most likely of Classical-Hellenistic date,1 a Hellenistic to Late Roman bowl fragment,2 and two ring bases, one datable to the Roman period,3 the other is Late Roman.4 Secondly, jugs (n=3) are relatively small closed vessels for the (temporary) storage and pouring of liquids, though plain sherds attributed to this functional class could equally belong to small-sized amphorae (amphoriskoi) that were rather appreciated for their contents. Two pieces were too fragmented to date precisely: Hellenistic to Early Roman,5 and Hellenistic to Late Roman.6 Only a small handle7 shows strong morphological similarities with the handles of Mid- to Late Roman jugs from Athens (Hayes 2008: 93), now also known to have been manufactured at Koroneia in Boeotia. Terra sigillata, or red slip (table) ware, was an extremely common feature of Roman society, though certainly not exclusively, from the mid-second century BC until the late seventh, early eighth century AD (Bes 2015). Although the products of about a dozen major table ware producers reached sites across the Mediterranean and sometimes beyond, an increasing amount of evidence shows that this scenario was much more varied: only the tip of the proverbial iceberg is gradually emerging with regard to table ware manufacturing centres that catered for smaller (regional?) markets. Late Hellenistic and Early Roman terra sigillata was not identified among the pottery collected. Only Late Roman Phocaean Red Slip Ware (PRSW) was recognised (Hayes 1972: 323-70; Vaag 2005). Originally this ware was associated exclusively
1 C2204. 2 C4042. 3 C3873. 4 C4183. 5 C3862. 6 C4114. 7 C3819.
266
The Cide-Şenpazar Region in the Roman Period
with ancient Phokaia in western Turkey; recently however it has become evident that several other places in western Turkey were producing similar shapes, and as a result the old, more neutral term Late Roman C (LRC), which serves as an umbrella term, is gaining in popularity. Eleven fragments were recognised, eight of which can be attributed to the Phokaia workshops, and include Hayes’ forms 3C,8 3E,9 3E-F,10 3F,11 3F/10A12 and 10A13 (for C3295: type 3 E-F, and C3815: type 3F see Fig. 10.2). The number is too small to perform a quantitative analysis, yet it is noteworthy that only form 3F occurs more than once, which is in line with more general, Mediterranean, trends. Chronologically, these forms span roughly the period between AD 425 and the early seventh century AD. Three fragments remain ambiguous,14 being classified as Çandarlı Ware/LRC, in line with Hayes’ observation concerning the morphological and macroscopic similarities between these two classes of red slip (table) ware (Hayes 1972: 317, 369). To further complicate matters, recent archaeometric analyses have shown that it is no longer feasible to confidently distinguish between Çandarlı Ware produced during the first three centuries AD at Çandarlı, ancient Pitane, close to Pergamon, and Late Hellenistic to Early Roman products from Pergamon proper, the pergamenische Sigillata (Japp 2009; Mommsen and Japp 2009; Schneider and Japp 2009). Also here there is a need for an umbrella term, and Eastern Sigillata C seems appropriate. Despite these small numbers, it is noteworthy that no other class of (Late) Roman table ware was recognised in the CAP material. Above all the absence of Pontic Sigillata and Pontic Red Slip Wares in the survey data, which are otherwise well attested at Hadrianopolis, Pompeiopolis and Sinop (see below), is unexpected.
10.2.2 Amphorae The best represented vessel type in the CAP assemblage are Hellenistic to Late Roman amphorae (n=180), vessels used for the storing but above all the transport of particular foodstuffs. The CAP evidence only modestly reflects the large variety in morphology and provenance of amphorae manufactured during this period and most types and fabrics are attested in single or a few occurrences only.
8 C2822. 9 C3732. 10 C3295. 11 C2226, C3815, C4233. 12 C4073. 13 C3311. 14 C548, C3912, C3978.
The Pottery — Breakdown by Period and Functional Class
267
Among the CAP amphora sample are Aegean classes such as Rhodian15 and Chian16 amphorae, the latter are known from the sea around Sinop (Kassab Tezgör et al. 2003: 172-3, nos 7-9). Slightly more common are Hellenistic to Early Roman Knidian amphorae.17 Further finds from the Aegean proper include three sherds from western Turkey. A rim, probably of the so-called Nikandros group18 from the region of Ephesos, is most distinctive (Fig. 10.2); its profile resembles examples dating to the third century BC (Lawall 2004: 180, figs 3, 4 left). From the same general area comes an amphora toe,19 possibly associable with the Mid- to Late Roman Agora M273/ Samos Cistern Type family (Robinson 1959: 109-10, pl. 29; Reynolds 2010: 97), as well as a small rim of possible Roman date.20 Examples of the Agora M273/Samos Cistern Type family are also known from shipwreck finds around Sinop (Kassab Tezgör et al. 2003: 181-3, nos 26-7; Pieri 2005: 132-7). Further finds from the Aegean include thirteen sherds that belong to Late Roman Amphora 2 (LRA 2)21 (for C3711 see Fig. 10.2). This is an amphora type that given the considerable fabric diversity observed elsewhere was probably manufactured at a number of places, yet remains poorly understood (Karagiorgiou 2001; Reynolds 2004: 231-2; Pieri 2005: 85-93; Slane and Sanders 2005: 286-7; Reynolds 2010, 95-7). Three fragments22 can be associated with a production site in the Argolid, although based on macroscopic fabric descriptions some confusion appears to exist whether or not it ought to be associated with Chios rather than mainland Greece (Reynolds 2004: 231-2). One fragment has a colour that does not fit the Argolid (?) products too well; the horizontal combing on the other two fragments tentatively suggests a date prior to the mid-sixth century AD (Karagiorgiou 2001). If LRA 2 was indeed used predominantly for the transport of olive oil, its presence in the Cide region and elsewhere may well signal the Black Sea’s ‘enormous thirst’ for olive oil (Braund 2005: 122). The provenance of another group of fragments remains unknown. Seven fragments bear a strong, but not conclusive, resemblance to both the Kapitän II (one handle fragment resembles the Kapitän I) and Zeest 80/Knossos Type 39 amphorae types and can be dated to the Roman era.23 Due to the absence of archaeological evidence for manufacture (i.e. kilns, workshops, pottery wasters), the origin(s) of these types, which share macroscopic characteristics (Reynolds 2010: 90; Tanagra and Hyettos, central Greece, pers. obs.), remain(s) poorly understood. Reynolds favours a northern
15 C1135. 16 C3952. 17 C1226, C2166, C2959, C3259. 18 C1215. 19 C2595. 20 C2682. 21 C671, C713, C1132, C1203, C2864-5, C3266, C3711, C3716, C3750. 22 C2408, C3489, C3827. 23 C2647, C3472, C3475, C3823, C3845, C4103.
Fig. 10.2: A selection of Roman-period sherds. Produced by Lorraine McEwan.
268 The Cide-Şenpazar Region in the Roman Period
The Pottery — Breakdown by Period and Functional Class
269
Black Sea origin (Reynolds 2010: 90), whereas Opaiţ recently postulated an origin on Chios for the Kapitän II (pers. comm.). These types are nevertheless fairly easily recognisable, in particular because of their characteristic handle and rim profiles as well as their fabric, although some fabric variety does exist and may reflect different production centres, and/or the use of different clays by the same workshop(s). For instance, one handle resembles that of the Zeest 80/Knossos Type 39 (Fig. 10.2), yet it is not in the common coarse red-grey fabric.24 A small but interesting group are four fragments of what appear to be central North African amphorae (Tunisia and Tripolitania), presumably dating to the Mid- to Late Roman periods,25 one fragment is tentatively attributed to the Late Roman period (Fig. 10.2).26 However, the group that stands out are the Sinopean amphorae, more correctly amphorae from Sinop and environs that are tentatively associated with the transport of wine, (olive?) oil, and non-liquid contents such as fish products (Demir 2007; Lund 2007; Kassab Tezgör 2010: 121-2, 127, 133-4, 137). This attribution is based on both colour and the presence of black volcanic inclusions (mostly pyroxene), though recent research, particularly for the pâte rosée, indicates that: ‘[c]aution is also due [...] when using pyroxene as a criterion. All the parts of the Sinopean amphorae do not contain the same density of pyroxene: while it is quite high in the neck and the handles of some containers, it can be lower, if not nearly absent in the rest of the body [...]. Just as the presence of pyroxene does not necessarily mean that an amphora originates from Sinop, its absence does not mean that it does not.’ (Erten et al. 2004: 105; also Kassab Tezgör 2010: 121, 123, 134-7).
Twenty-one fragments were identified as belonging to this category, four of which were assigned to the pâte rosée,27 ten to the pâte rouge/colorée,28 and five to the pâte blanche/claire,29 which approximately date to the second-third, fourth-fifth and sixth(+?) centuries AD respectively. Only one fragment could be typologically identified: in pâte blanche/claire,30 it belongs to type D Snp III (Fig. 10.2), the manufacture of which presumably started not later than the second half of the sixth century AD (Kassab Tezgör 2010: 135-7). The pâte blanche/claire need not have originated from Sinopean workshops alone (Reynolds 2010: 96; Erten et al. 2004: 104-5). A Sinopean provenance is postulated for two fragments — this tentative identification rests on the characteristic black (pyroxene) inclusions (Kassab Tezgör 2010; Erten et al. 2004).
24 C3020. 25 C4056, C4090, C4131. 26 C4083. 27 C2800, C2875, C3988, C4182. 28 C669-70, C3697, C3729, C3833, C3889, C3915, C3923, C4093, C4147. 29 C352, C1982, C2228, C3018. 30 C2782.
270
The Cide-Şenpazar Region in the Roman Period
10.2.3 Cooking Wares and Other One fragment of likely Roman date is the rim of a (small) pithos,31 a large vessel type principally used for storing dry or liquid foodstuffs. Our understanding of the production, use, life-span and socio-economic context of pithoi (Latin: dolia) is relatively limited due to a paucity in research. Some work was carried out on pithoi from Pessinus in central Anatolia (De Paepe and Vermeulen 1999; Devos et al. 1999; De Paepe et al. 2001), and a recent study, thematically and chronologically broad, offers a useful synthesis (Giannopoulou 2010). Eleven fragments were confidently classified as cooking wares based on fabric characteristics. Nine remain unidentified, and the scarcity of reference points other than fabric allows for only broad chronological sub-divisions: one is probably Hellenistic,32 six are of Roman date,33 and a further two could well be post-Roman.34 One fragment is tentatively thought to have come from northern Turkey,35 since its fabric contains black volcanic inclusions similar to those in Sinopean (and related) fabrics. A handle fragment contains abundant silvery micaceous inclusions, which points to a likely source in western Turkey. This class predominantly tends to be fired grey to dark grey throughout, yet the CAP fragment in question is orange.36 We can tentatively associate this with a related class of Late Roman micaceous Aegean Ware (Slane and Sanders 2005: 255-6).
10.3 Discussion The comparatively limited quantity of collected Roman pottery precludes any detailed quantitative analyses or study of surface scatters. Nonetheless some interesting observations can be made concerning the functional and chronological dimensions of the pottery and, combined with the architectural remains, their spatial distribution within the survey area.
10.3.1 Provenance and Proportions A significant share of the CAP pottery comes from sources that lie well beyond northwest Asia Minor. This pertains for example to amphorae (LRA 2, central North
31 C3723. 32 C3963. 33 C3027, C3029-30, C3315, C3733, C4085. 34 C4116, C4118. 35 C3632. 36 C3694.
Discussion
271
African examples) and table wares (Phocaean LRC, western Turkey). The provenance of a small portion of the assemblage remains unknown: an Aegean or Black Sea source is assumed for those fragments that were ascribed to Kapitän II, Zeest 80/Knossos Type 39 and possibly related shapes. From a closer source come Sinop-region amphorae, and in particular a group of smaller amphorae that constitute a considerable share of the CAP material (see below). Insofar as their provenance could be established with some confidence, nearly none of the categories attested come as much of a surprise: only a small proportion does not originate from a Black Sea or Aegean source. The absence of certain well-known classes is also noteworthy. No fragments of Late Roman Amphora 1 (LRA 1) were identified, a type that was manufactured predominantly in Cilicia and Cyprus and which is attested far and wide at Late Roman Mediterranean sites and beyond (Decker 2001). It circulated possibly in conjunction with LRA 2 in the Aegean and/or Black Sea (Karagiorgiou 2001). Not recognised were table ware products that were supposedly manufactured at or around Hadrianopolis (Laflı and Kan Şahin 2012a: esp. 51-3). Further, the absence of African Red Slip Ware (ARSW) from Tunisia, but above all of products from the Pontic region proper (Pontic Sigillata, Pontic Red Slip Wares) comes unexpectedly. Both wares are commonly attested at Sinop (Fırat 2010) and Pompeiopolis (Domżalski 2011), and ARSW has been attested at Hadrianopolis (Laflı and Kan Şahin 2012b). One possible explanation for the dominance of Phocaean LRC, which is, however, not entirely satisfactory since it focuses too narrowly on one category, is that the survey area lies in greater proximity to Constantinople than, for example, Sinop where a more varied table ware scenario can be observed. Constantinople was the point of transition for seaborne traffic between the Black Sea and the Aegean and by extension the Mediterranean and cabotage may have been one way through which goods were brought to settlements along Asia Minor’s northern shores (Belke 1996: 135-7). Perhaps the region of Cide-Şenpazar was therefore located within the area in which Phocaean LRC dominated, a class that is well documented at Constantinople (Hayes 1972: 418; 1992: 5, 7).
10.3.2 Spatial and Chronological Distribution With the exception of the odd scatter of finds datable to the Hellenistic to Early Roman periods (see above), the majority of the collected pottery belongs to the Mid- but above all the Late Roman period, a predominance that is well documented elsewhere, e.g. in rural Greece (Pettegrew 2007). There, this predominance is considered to signal an important phase of land use and occupation, which is explained through, among other things, historical factors, changes in the socioeconomic organisation and exploitation of the land, as well as the relative ease with which we can identify Late Roman pottery types, specifically amphorae, even if the latter point is certainly not as straightforward as is sometimes postulated.
272
The Cide-Şenpazar Region in the Roman Period
In the Cide-Şenpazar region, geomorphological factors have altered the landscape considerably since antiquity (chapter 2), which (partly) explains the scarcity of Hellenistic and Roman-period finds. This seems to be the case particularly for the coastal and near-coastal zones, where Roman occupation is suspected to have been predominantly concentrated and where sedimentation appears to have had an even greater impact than elsewhere in the region. Despite the hidden or obscured character of the Hellenistic and Roman cultural landscapes, which greatly hampers the interpretation of the survey finds, some observations can be made. In the following, I will discuss a number of locations where Roman features or finds that merit separate discussion, were found. In some cases these are of the group site type, in others we are dealing with features, or spot sites (section 4.2.3). It should be emphasized that most of the locations that will be discussed, are not significant as centres of past human activities, instead in many cases we are dealing with chance discoveries and spolia that are of interest. The artefact concentrations are mostly small to very small, making it not only difficult to investigate the association between certain finds, but also, more importantly, to conclude whether these clusters can be referred to as activity or habitation sites (Fig. 10.3). The characteristics of each location nevertheless merit discussion here. Location 1,37 which includes group site S602, consists of the promontory of Gökçekale. This site did not deliver much in the way of artefacts. Three fragments of imported amphorae (Sinop, possibly North Africa) are accompanied by (ancient?) building traces and the remains of at least 16 to 17 opened cist graves (S100, S102, S104, S106 and S108) (Fig. 10.4) and many more undisturbed ones, that point to a necropolis of some significance. These graves are of possible Roman date, as the rescue excavation of two comparable cist graves at Türbe Tepe Üste (Cide) suggests (Bleda Düring, pers. comm.). Also, cist graves seemingly constructed of four slabs were found in the eastern part of Hadrianopolis, which are dated, albeit without further explanation, to the Late Antique period (Laflı and Christof 2012: 7-8, fig. 13a-b). A younger date (Mid- and/or Late Roman) can indeed be postulated considering the more common practice of cremation during the Early Roman period. Another interesting feature is a rectangular depression (S176) on a rocky outcrop right next to the beach. Both an outlet at its western end and a possible small rectangular hole up at its eastern end provide indications that this once was part of a (lever?) press installation (Fig. 10.5), most likely to produce olive oil. Worth mentioning are two more features, the first ‘ein Stück antiker Pflasterung’ near Kapısuyu, located some 2.5km west of location 1 (Belke 1996: 129) and, secondly, a (limestone?) quarry that was identified at S182, about 1km eastwards along the coast (Fig. 10.6).
37 Sites 100; 102; 104; 106; 108; 176; 178; Units 3231; 3233. Pot count 5. Sinopean pâte colorée; MROMLROM North African amphora (?); Sinopean pâte blanche/claire; Zeest 80.
Discussion
273
Fig. 10.3: Concentrations of predominantly Roman-period ceramic finds in the study area. Produced by Toby C. Wilkinson.
Fig. 10.4: Cluster of robbed stone cist graves on the Gökçekale promontory. Part of a larger cemetery. Produced by Lorraine McEwan.
274
The Cide-Şenpazar Region in the Roman Period
Fig. 10.5: Basin of a possible (lever?) press installation, presumably for the production of olive oil, at Gökçekale (S176; location 1) (CAP-P1/5111).
Location 238 is located to the west of, and around, Yenice. Diagnostic ceramic finds are sparse, and include a Hellenistic handle fragment, single fragments of Phocaean LRC and a North African amphora. Architectural features comprise an ashlar (S168), and two large E-W oriented cist graves (one at least with a monolithic lid) at S170 (Fig. 10.7) some 800m west of Yenice and also not too far from Gökçekale, which were originally dated to the Hellenistic period, but in light of the observations made above a Mid- and/or Late Roman date can plausibly be postulated. Location 3,39 around Gideros, actually comprises three finds concentrations: coastal location 3a around the bay of Gideros; 3b is a non-coastal location some 1.7km to the southeast of this bay; near-coastal location 3c is located to the east of
38 Sites S168; S170; Units U1108; U1109. Pot count 3. Hellenistic handle; Phocaean LRC; MROM-LROM North African amphora; Hellenistic cist graves; Roman (?) building blocks. 39 Location 3a: Sites S12, S17; S16; S21, S59, S123; location 3b: Site 17; location 3c: Sites S16; S19; S66, and Units U7; U253; U276; U289; U298; U381; U461. Knidian amphora (?); Sinopean pâte blanche/ claire (3x); Phocaean LRC Hayes form 3F; Roman roof tiles (12, 21?, 59); (Early?) BYZ spolia (123). Possibly ancient Kytoros, Cythero, Cytorus, Cytorum (Gideros?) (see Belke 1996: 50, 245-6; Talbert 2000: map 86; Tabula Peutingeriana Section VIII: http://www.tabula-peutingeriana.de/tp/tpx.html).
Discussion
Fig. 10.6: Quarry face along the coast, between Gökcekale and Yenice (S182; between locations 1 and 2) (CAP-P15/268).
275
Fig. 10.7: One of two cist graves near Yenice (S170; location 2) (CAP-P1/5091).
Gideros. Finds were rare at 3b, with a scatter of likely Roman pottery fragments. At 3a the ceramic finds included a Hellenistic to Early Roman Knidian amphora fragment, Roman roof tiles, and imported Sinopean amphorae and Phocaean LRC table ware. At S123 the architectural finds include a rectangular basin, a base or top moulding, perhaps from a podium (Fig. 10.8), a door threshold with a dowel, a pivot and a bolt hole (CAP P2/1867), an ashlar (?) with a dowel hole (i.e. originally it was standing upright), and a now largely buried column fragment. Ceramic finds at 3c include Roman roof tile fragments, a Knidian amphora fragment, and imported Phocaean LRC table ware and Aegean, North African and Sinopean amphorae. At S66 a rock-cut grave (chamosorion) was encountered, thus possibly of Mid- to Late Roman date (Fig. 14.21). The spatial distribution and relative scarcity of the finds preclude a detailed interpretation, but the Byzantine monastic complex identified at Çadır İni/Kilise (S20/S21), which is discussed in detail in chapter 11, may originate in the Late Roman-Early Byzantine period. One or more of these locations can be associated tentatively with ancient Kytoros, which was equipped with a (natural) harbour: two ‘Landzungen’ are still visible today (Belke 1996: 245-6; Marek 1993: 17-8; Talbert 2000:
276
The Cide-Şenpazar Region in the Roman Period
Fig. 10.8: A base or top moulding, perhaps from a podium, from the bay of Gideros (S123; location 3a) (CAP-P2/1856).
map 86).40 Kytoros was a colony and trading post of Sinope and was founded in pre-Hellenistic times (chapters 3, 8 and 9), and later controlled by Heraclea Pontike (Jones 1937: 149). Modern Gideros likely echoes the Classical name and, despite the absence of a geographical link, Cide possibly also derives from Kytoros. Kytoros is not to be confused with Cotyora, also considered to have been a Sinopean colony; the Cytorus mentioned by Strabo (Strab. 12. 3. 17) pertains to Cotyora (Erciyas 2007: 1196; Commentary to Arr. Peripl. m. Eux. 117). It is noteworthy that the distance between Cythero and Egilan as indicated on the Tabula Peutingeriana is 8 Roman miles (8* ca. 1,479m = ca. 11.8km),41 which roughly corresponds to that between Gideros (≈Kytoros) and Cide (≈Egilan/Aigialos), even if the Barrington Atlas gives a distance of only about 6km, as well as to the 60 stadia (*200m = 12km) given by Arrian, who mentions a ‘mooring for ships’ (Arr. Peripl. m. Eux. 14). Location 4,42 near-coastal Unit 2213 close to Kalafat, some 4km southeast of location 3c, comprises two pottery fragments, one of which is possibly Hellenistic in date. Of note is that this location, or more broadly this hill range, might be identified
40 http://mail.nysoclib.org/BarringtonAtlas/BATL086.pdf 41 http://www.tabula-peutingeriana.de/tp/tpx.html 42 Unit U2213. Pot count 2. HEL fragment?
Discussion
277
geographically with the ‘Kleinere Burg’ or ‘Festung’ at Okçular, which Belke locates some 4km from the coast (Belke 1996: 255). However, the village of Okçular and the rock outcrop of Okçular Kale (S22) is located much closer to the coast (ca. 1.3km), as is Gebeş (ca. 800m). Belke (1996: 255) considers his Okçular to be ‘identisch mit der von Kalinka unter Gebeş (2km n. von Okçular) erwähnten Festung’. Location 543 is situated some 800m east of Kalafat. Ceramic finds include a possible Rhodian amphora handle, a Mid- to Late Roman jug fragment, imported Phocaean LRC table ware, a few amphorae from the Aegean (LRA 2) and/or Black Sea (Kapitän II/Zeest 80/Knossos Type 39), and an unidentified eastern Mediterranean source. A large Roman spolium, now used to cover a well, is a cornice block or, more likely, an upper moulding (of a podium?) (Fig. 10.9). Further worked stones are reported alongside a nearby stream.
Fig. 10.9: Upper moulding, possibly of a podium, east of Kalafat (S160; location 5) (CAP-P1/5033).
43 Sites S73; S160; Units U1067; U1075; U1088. Pot count 8. Rhodian (?) handle; Phocaean LRC Hayes form 3F; MROM-LROM jug handle; LRA 2 amphora (wavy combing); Kapitän II/Zeest 80; east Mediterranean import?; large (EROM?) spolium.
278
The Cide-Şenpazar Region in the Roman Period
Location 644 consists of the Okçular valley. Here, 29 pottery fragments were found to the south and southeast of Okçular Kale (S22) which date to the (Hellenistic-) Roman period (also chapters 11 and 14). Except for a possible Classical-Hellenistic echinus bowl, the diagnostic fragments are Roman, and include imported red slip table ware and amphorae. The architectural finds include a column fragment (S64), and two ancient cistern mouths, purpose made, at S61 (perhaps of white marble) (Fig. 10.10) and S77 (CAP P2/700); both are nowadays reused as well heads (see chapter 11, this volume, for a different interpretation). Such cistern mouths already occur in Hellenistic times, for instance at Pergamon. Location 745 is located on the coast west of Kumluca, with Phocaean LRC, and amphorae from Sinope, the Aegean and the Aegean/Black Sea.
Fig. 10.10: A (marble?) cistern head, from around Okçular (S61; location 6) (CAP-P1/404). 44 Sites S24; S61; S64; S77; Units U44; U89; U147; U174; U176; U185; U187; U189; U593; U673; U679; U683; U685; U2239. Pot count 29. Echinus bowl (?); Sinopean pâte rose, pâte colorée (2x), pâte blanche/claire Type D Snp III (6th century?); Phocaean LRC; LRA 2 (3x); MROM-LROM North African amphora; column (base?) (64); column base (77). 45 Site S109; Units U2104; U2185; U2189. Pot count 6. LRA 2 amphora; Sinopean pâte colorée; Phocaean LRC Hayes form 3E; related to Kapitän II/Zeest 80 (2x).
Discussion
279
Location 8,46 near the village of Irmak, remarkably contains no diagnostic ceramic and architectural finds. Location 947 comprises various types of artefacts and features. Only Roman roof tiles are among the diagnostic pottery from near-coastal location 9a, between Irmakköy and Cide. Architecturally, a block with a circular hole (d= ca. 30cm) was found at S140 (CAP-P1/4922), and a similar block (hole d= ca. 40cm) at S148 (CAPP1/4936). Both were originally interpreted as mortars for pounding grain, yet their exact purpose remains unclear. Perhaps these were collecting vats, or elements used in pressing activities (Belke 1996: 227, 229, figs 72, 74-5, 119); since the block at S140 holds water it was probably not a press weight. At S142, a basin was found constructed of ancient slabs, held together by iron clamps (CAP-P1/4923). A marble Ionic capital was found at S150 (Fig. 10.11). It is not very well executed, and being adorned with only one large egg consequently lacks the standard egg list. Finally, some 200m from the beach at S152, several large dressed blocks were found (Fig. 10.12): the block on the left is possibly a top moulding, the other two blocks might be parts of a door frame. A clamp and lifting hole were clearly visible on one block, and a roof-gutter on another. At S70-S71 for example, at coastal location 9b some 2.3km west of Cide, pottery finds were rather numerous, and included a Kapitän II handle, Phocaean LRC table ware, and imported Sinopean and Aegean amphorae. Architectural finds include two columns, originally identified as possible milestones, a possible door lintel (CAP P2/9), and two more unidentified dressed blocks. Perhaps most significant is the upper part of an Early Roman (?) tombstone with some epigraphic significance, documented in an ex situ location, which reads ΓΛΑΥΚΟΣ ΜΕΝΕΦΡΟΝΟΣ ΚΡωMΝΕΙΤΕΣ (Glaukos, son of Menephron, from Kromna) (Fig. 10.13). The inscription is flanked by a rosette on either side of the panel, and the stone is topped by a pediment that holds a rosette in its centre (Marek 1993: 66-7). Both Glaukos and Menephron are known from northwest Asia Minor (Corsten 2010). Significantly, Kromneites, if the reading is correct, most likely is derived from Kromna, a coastal city thought to be located at present-day Tekkeönü, some 27km west of Cide (Belke 1996: 241-2; Marek 1993: 17, 185 for epigraphy). Arrian gives a distance of 150 stadia between Kromna and Egilan/Aigialos/Cide, which corresponds to ca. 30km (Arr. peripl. m. Eux. 14). If indeed Roman, Kromneites implies that Glaukos was a native of Kromna. However, using an ethnikon in the Roman period
46 Units U3193; U 3199; U 3205; U 3210; U 3216; U 3217; U 3225. 47 Location 9a: Sites S1; S140; S142; S148; S150; 1S52; Roman (?) roof tile fragments; large mortar (?) press weight (?); Roman (?) basin (reused?); large grain mortar (?); capital, marble and other stones (unclear); Roman building blocks. Location 9b: Sites S2; S3; S67; S70; S71; S127; S129; S208; Units U3135; U 3142; U 3143; U3146; U 3147; U 3148. Pot count 63(+?). Sinopean pâte rose (?); Kapitän II handle; LRA 2 amphora (2x); Phocaean LRC; Sinopean pâte colorée (5x); blocks (Roman? Byzantine?); gravestone (Roman); Roman block; Roman (?) milestone (2x).
280
The Cide-Şenpazar Region in the Roman Period
Fig. 10.11: Marble (?) Ionic capital, of Roman date, between Irmakköy and Cide (S150; location 9a) (CAP-P1/4943).
Fig. 10.12: Dressed blocks found between Irmakköy and Cide (S152; location 9a) (CAP-P1/4952). On the left a possible top moulding, the other two blocks could be elements of a door frame.
Discussion
281
was rather unusual. Alternatively, it could signify a demothikon, which could confirm that Kromna had been joined to a larger administrative unit, which appears to have been Amastris (Marek 1993: 21, 33; Strab. 12. 3. 10). It should be noted that the exact provenance of this tombstone is not known.
Fig. 10.13: Top part of an Early Roman (?) tombstone, reading ΓΛΑΥΚΟΣ ΜΕΝΕΦΡΟΝΟΣ ΚΡωMΝΕΙΤΕΣ (Glaukos, son of Menephron, from Kromna) (S208; location 9b?) (CAP-P15/402).
Location 1048 is located near Cide. At the near-coastal location 10a west of Cide, the ceramic finds include a Knidian amphora fragment, Phocaean LRC, an imported Sinopean amphora, as well as fragments of presumably locally-made amphorae (U814, S114) (see below). The ceramic finds at coastal location 10b ― comprising modern-day Cide ― did not contain any recognisable fragments, yet the majority found at S196 again belongs to the locally-made group. Architectural fragments include large blocks
48 Location 10a: Site S114; Units U814; U842; U846; U852; U856; U858. Pot count 31. Knidian amphora (?); Phocaean LRC Hayes forms 3E-F, 10A; Sinopean pâte colorée. Possibly ancient Aigialos, Egilan? (see Belke 1996: 158; Talbert 2000: map 86; Tabula Peutingeriana Section VIII: http://www.tabulapeutingeriana.de/tp/tpx.html). Location 10b: Sites S196; S206; S502; S519. Roman amphorae; Roman cist graves; Roman building blocks; column (period?); Roman milestone. Possibly ancient Aigialos, Egilan? (see Belke 1996: 158; Talbert 2000, map 86; Tabula Peutingeriana Section VIII: http://www. tabula-peutingeriana.de/tp/tpx.html).
282
The Cide-Şenpazar Region in the Roman Period
presumably of a monumental/public building (CAP P1/5198), two Roman cist graves excavated by the Turkish authorities at Türbe Tepe Üste, a column fragment, and a column or possible milestone. Interestingly, a sanctuary dedicated to Zeus has been postulated about 1km southwest of Cide (Belke 1996: 158; Marek 1993: 98-9, n. 685). Belke puts forward the idea that Classical Aigialos might be situated ‘in der Nähe des heutigen Cide, vielleicht Karaağaç Limanı ö. von Cide’ (Belke 1996: 158; Talbert 2000: map 86;49 Marek 1993: 185-7 for epigraphy). As a matter of fact, Apollonius of Rhodes and Strabo use the name Aigialoi for a 100(+)-stadia long stretch of the shore (Commentary to Periplus Ponti Euxini: 112). In addition, Strabo (12. 3. 8, 10) mentions a village of the same name ‘ten schoeni distant from Amastris’ that roughly equals 55km, which in turn approximates the distance between modern Amasra and Cide. On the Tabula Peutingeriana, both Cythero and Egilan are situated inland50 (see below). Location 11,51 in the inland Loç valley, consists of only two diagnostic pottery fragments (Aegean LRA 2) that were identified between Çamdibi and Şenköy. Also from this area are a (Late?) Roman or Byzantine square well lined with bricks, a large irregular block with a depression (S120), a fluted column fragment, a stone basin with an outlet and a rosette-like relief carved asymmetrically on one of its short sides (S126) (Fig. 10.14) (cf. Belke 1996: 279, fig. 119), perhaps a reused sarcophagus, and a large roof tile fragment that possibly belonged to a tile grave. Location 12,52 also located inland at Aybasan, includes a Hellenistic cooking ware sherd and an imported amphora from western Turkey. These are the only diagnostic fragments found in this large surveyed area southwest of Aybasan, some 14-15km from the coast. Location 1353 is located on the coast around Kuşçu and has also yielded limited diagnostic ceramic finds. This is to some degree compensated by the architectural finds that include several large ashlars, now reused in modern buildings, that probably belonged to a monumental/public building. Furthermore, a ceiling cassette fragment (Fig. 10.15) that could have belonged to a variety of buildings, including funerary monuments, a stone gutter or perhaps a stone imbrex, a gutter block or possible cancelli pier with a slit to hold a chancel screen (CAP-P2/947) (though see Adam 1984: 281, fig. 601), several moulded fragments that may have belonged to a doorpost, and two small columns were recognised. The latter taper slightly and so
49 http://mail.nysoclib.org/BarringtonAtlas/BATL086.pdf 50 http://www.tabula-peutingeriana.de/tp/tpx.html. 51 Sites S88; S90; S120; S122; S126; S134; Units U2086; U2108; U3049; U3086. Pot count 6. LRA 2 (2x); Roman and/or Byzantine well; pressing stone (?); fluted column (?) statue fragment (?); stone basin (reused sarcophagus?); brick/tile (?) grave? 52 Units U707; U2127; U2140; U3097; U3098. Pot count 12. HEL cooking ware (?); MROM-LROM Agora M273/Samos Cistern Type amphora. 53 Sites S10; S11; S89. Roman roof tile fragments, spolia (chancel screen?; church?), HEL-EROM ashlars?
Discussion
283
Fig. 10.14: Photograph of a stone basin, with a wheel-like engraving on the short end and an outlet, perhaps a reused sarcophagus (S126; location 11) (CAP-P1/4890).
Fig. 10.15: A ceiling cassette fragment from Kuşçu (S89; location 13) (CAP-P2/940).
284
The Cide-Şenpazar Region in the Roman Period
possibly were either not columns or subsequently recut (CAP-P2/947). In concert, these architectural elements appear to reflect the multi-period character of this part of the survey area. Location 14,54 or Site S110, is also located on the coast around Uğurlu, where diagnostic ceramic finds are again scarce, but significant is another Knidian amphora fragment. No Roman architectural finds were identified, though a Roman wall is suspected (Belke 1996: 274-5). These scarce finds may be tentatively brought in connection with Classical Thymaina, located some 16km east of Cide, alternatively known probably as Thymena, Teuthrania, or Timle (Talbert 2000: map 86;55 Belke 1996: 274-5; Marek 1993: 84). The latter name is used to refer to the two Byzantine castles (S110 and S112) recorded by CAP (see chapter 11). Arrian gives a distance of 90 stadia, about 18km (Arr. peripl. m. Eux. 14). Three further features are of interest. First, an Ionic-Attic column base (S133) of probable Early or Mid-Roman date, an isolated find in the village of Ovacık (CAPP2/1997). Second, Kazallı Kalesi (S91), ca. 6km east of Cide, is a fortified stronghold that includes a Roman-period structure (or possibly structures), which was later reused in the Mid- or Late Byzantine period (chapter 11). Third, a rather spectacular find was a well preserved bridgehead, rather than an aqueduct because of the topography of the location, which is perched dramatically against the western cliff alongside the Devrekani river at Atköprü (S95) (Marek 1993: 10, 92) (Fig. 10.16). The original construction can tentatively be dated to the Late Hellenistic-Early Roman period (Adam 1984: 311, fig. 662; see chapter 11, this volume, for a somewhat later date). The abutment is constructed of bossed ashlars with drafted edges, set off above by a row of imposts. Nearly nothing remains of the arch proper: at either end two low walls are preserved that possibly represent the lower courses of the arch, but it is difficult to distinguish springers and/or voussoirs. Although this bridge may have consisted of multiple arches — the Devrekani is about 10 to 15m wide at this point — single arches with a span of 20m or more are known from Late Republican-Early Roman Italy (Quilici 2008: 569-73). What is preserved of the bridge proper on the eastern side could not be fully investigated, but here also part of a wall may be part of the springer and lower voussoirs segment. More significant, however, is the irregular pavement that is in line with the bridgehead on the western side, and is preserved right to the point where the bridge/arch proper began. Wooden beams protrude from underneath the pavement, and are parallel to it (Adam 1984: 302, fig. 646). Its location is interesting, not in the least because it suggests the existence of a (paved?) road running 10-11km inland at this point. Atköprü is, however, not unique in this regard. Despite the region’s mountainous character, (paved) roads are also known from elsewhere in northern Paphlagonia (Belke 1996: 127-9).
54 Site S110. Pot count 2. Knidian amphora?; LROM-EBYZ fragment. 55 http://mail.nysoclib.org/BarringtonAtlas/BATL086.pdf.
Discussion
285
Fig. 10.16: Bridgehead along the Devrekani at Atköprü. Note the wall in the lower part of the picture, which is part of the bridgehead on the eastern side (S95) (CAP-P3/262).
10.3.3 Interpretation of the Data The scarce, if occasionally monumental, finds of the Roman period collected and recorded during three years of CAP should be seen as the remnants of a now largely vanished cultural landscape, which, however, appears to have been rather densely settled during the Roman period. Although this scarcity makes it impossible to identify, among other things, well-defined urban and non-urban zones, some interesting observations can nevertheless be made. First and foremost we can note the coastal or near-coastal character of most locations. Only two locations (11-12) ― ca. 15km and 14km from the present coastline
286
The Cide-Şenpazar Region in the Roman Period
respectively ― are situated inland. Even though the coverage by CAP was necessarily selective (chapter 4), this geographical pattern seems to add weight to the hypothesis that the occupation/habitation of Cide and environs during the Roman period was primarily concentrated in the coastal zone, which is now largely hidden due to human and natural factors. The non-coastal locations are obviously also of note, and it is not unthinkable that the thin but extensive scatter around Okçular, for instance, represents permanent occupation. In times of military, social or other unrest, this location may also have served as a refuge (chapters 11 and 14 for more detail). Secondly, the majority of pottery fragments that are typologically, chronologically or otherwise identifiable belong to the Mid- to Late Roman period. To what degree these locations were occupied contemporaneously, however, remains as yet unanswerable. The pottery generally reflects exchange trends with a west (Aegean-Mediterranean) east (such as Phocaean LRC, Aegean amphorae) and an east-west (for example Sinopean amphorae) direction. For the non-Black Sea pottery it can be conjectured that most if not all flowed into the Black Sea basin via Constantinople, and probably the actual transport of goods, particularly ‘international’ exchange, mostly took place by ship rather than over land, since ‘im allgemeinem der Transport von Waren und Menschen auf dem Seeweg [war] zu allen Zeiten billiger, schneller und bequemer’ (Belke 1996: 127; Braund 2005: 115, for Asia Minor’s northern coast). It remains unclear, why of all the major Late Roman red slip table wares, only Phocaean LRC is identified in the survey area. That red slip table wares present a more mixed picture at Sinope might well mirror its strong economic basis, which include the manufacture of widely traded amphorae, and which could well have promoted a more dominant link with Constantinople and a more central role for Sinope in the redistribution of this pottery within the Black Sea basin. With an intermediate distance of ca. 2km (except between locations 2-3, where it is ca. 6km, and the ‘outlier’ locations 13-14), Roman-period coastal locations occur with some regularity in the Cide region, even if only selected parts of the coastal zone could be surveyed. One or more of these were surely anchorages that played a role in coastal exchange systems, such as cabotage. In addition to the partly conjectural coastal road, Mitchell rightly points out that the ― also partly unknown ― noncoastal road network played a crucial role for the movement, both economic and military, between and within inland regions and the coast (Mitchell 1993b: 245-6). For instance, a north-south route ran from Ankyra via modern Kastamonu to the Black Sea (Belke 1996: 127-30; Talbert 2000: map 86; Matthews et al. 2009: 187-9, figs 6.25-7). Finally, the scarcity of finds makes it difficult to attribute particular function(s) to each of the locations identified here (Belke 1996: 138-51, for mostly post-Roman ‘Wirtschaft’ in Paphlagonia). Historical sources suggest that the region played a role in supplying timber, predominantly perhaps for shipbuilding (Belke 1996: 139; Bittner 1998: 122-3; Meiggs 1982: 47, 372, 393; Broughton 1975: 836-7; Braund 2005: 122; also chapter 2, this volume). It remains unclear, however, whether ships were constructed in this region during the Roman period. If so, it would have also featured a lively scene
Discussion
287
of ancillary ‘industries’, although individual parts could also have been imported from further afield (Mitchell 2005: 102-3, on related activities at Korykos in Cilicia). The commercial exploitation and working of boxwood was also likely associated with the timber industry. Boxwood (Buxus sempervirens L.) is ideally suited for the manufacture of tools and other objects for use in daily life and Classical authors (a.o. Theophrastus, Pliny the Elder, Strabo) mention Kytoros and particularly Amastris as important sources of boxwood (Belke 1996: 140, esp. n. 12, 245-6; Marek 1993: 100; Bittner 1998: 125; Meiggs 1982: 282; Braund 2005: 124). Also noteworthy is a group (some 95 fragments) of amphorae fragments, in particular at locations 8-10 and with a thin but extensive scatter throughout the survey area, that includes a probable waster (Fig. 10.17). Macroscopically this is a fairly homogeneous and relatively sizeable group that, judging by the size of handles and rims, also comprises narrow-necked amphorae.56 Based on the
Fig. 10.17: Fragments of narrow-necked amphorae from Cide, probably locally manufactured (S196; location 10b) (CAP-P1/5898).
56 This group was further substantiated through close study of the photographs and descriptions of the fragments in question. Identifying sherds entirely or mostly on macrophotographs is a less than ideal practice; it implies that options need to be kept open regarding homogeneity, provenance, etc.
288
The Cide-Şenpazar Region in the Roman Period
quantity of material and the presence of waster(s), this group is postulated to have a local origin. Kiln wasters of vessels and possibly also of tiles were noted during the survey and large quantities of pottery fragments, handles especially, were observed particularly in U814 and adjacent units as well as at S114. A production context is therefore likely, even if we ought to take into account that part of this material may well represent habitation debris (Claudia Glatz, pers. comm.). That said, the vast majority of inspected fragments share a fairly homogeneous fabric, generally fine to very fine and buff to orange in colour. Inclusions usually comprise small to medium-sized dark (black) and some red grit, some (red) grog, calcitic and/or limestone bits; pores/voids are also not uncommon. Insofar observations can be made concerning morphology, most regularly occurring are handles that often have pronounced ridges, and are sometimes made rather crudely. The few rims thought to belong to this group are plain (undecorated), and lightly thickened; body sherds are generally ribbed. Some fragments suggest that some of the smaller amphorae represent so-called narrow-necked vessels. As the name implies, these vessels have very narrow and often high necks, features not untypical for amphora manufacturers in this area (Heraclea Pontike, Sinope; for the latter: Kassab Tezgör 2010: planches 8.4, 22). If, as the above archaeological evidence would suggest, amphorae were indeed manufactured in the Cide area during the Mid- and Late Roman period, questions arise concerning their content and distribution patterns. The manufacture of amphorae in Cide would have formed part of a wider regional phenomenon of amphorae production along Asia Minor’s northwest coast. Given the wide distribution of Sinopean and other regional amphorae in the Black Sea, Aegean and Mediterranean zones, it is likely that Cide amphorae would have also been channelled into these established lines of exchange (Braund 2005). Illustrative, though perhaps not directly relevant to the Cide region, are contacts between Sinope and the regions of Colchis and the Bosporan kingdom (Braund 2005: 120, 127-30). Of greater relevance perhaps was the stationing of troops along the Danubian frontier that provided one potential market to which amphorae and other products from northwest Asia Minor and elsewhere could be shipped; LRA 2, from a variety of Aegean workshops, is thought to represent an element of official army supplies (Salmeri 2005: 197; Karagiorgiou 2001; Swan 2004, 2007).
10.4 Conclusion In summary, it can be said that the archaeological evidence collected by CAP is too scarce to allow a comprehensive reconstruction of the Roman-period settlement landscape of this part of northwest Asia Minor. The combined sources at our disposal, archaeological and literary, do however provide important hints for a vibrant and most likely rather well-populated cultural landscape during the Roman period. This
Catalogue of Illustrated Artefacts
289
is most clearly documented for the Late Roman period and seems to continue into the Byzantine era (chapter 11). Some of the architectural remains suggest that at times Cide’s urban landscape must have been rather monumental and (in)directly connected to exchange networks operating in the Black Sea, but also connecting the region with the Aegean and the Mediterranean ‘markets’. Though little seems certain at present, the region appears to have participated in these networks primarily because of its timber and boxwood industries and through the manufacture of amphorae. However, even if coastal communities may have been largely oriented towards the sea, there are also a number of indications in the CAP survey data that point towards connections with the interior. Acknowledgements: I am most grateful to Bleda Düring, Claudia Glatz and T. Emre Şerifoğlu for their invitation to study the Roman-period finds from the Cide Archaeological Project, and to provide me with valuable information. Toby C. Wilkinson was most helpful with questions concerning the database and related matters, as well as making maps. Whether certain finds could be or were Roman and/or Byzantine was happily discussed with Marica Cassis. Also I would like to thank Inge Uytterhoeven and Julian Richard who provided valuable comments on the architectural finds, and Willy Clarysse for his reading and interpretation of the tombstone.
Catalogue of Illustrated Artefacts C1215. G6. Large fragment of an amphora rim, most likely of the Nikandros group, third century BC, from the region of Ephesos. Fine powdery, light red fabric with a reddish yellow surface, with abundant micaceous inclusions (‘mica-dusted’), and some small black inclusions. Smooth surfaces. Oxidised firing, wheelmade. Fabric 2.5YR 6/6, surfaces 5YR 6/6. D= 14cm. C2782. U624. Amphora handle and neck from Sinope in so-called pâte blanche/claire, possibly Typ D Snp III. Hard-fired, light yellow fabric with black volcanic inclusions, and white mineral inclusions. Oxidised firing, handmade. Fabric and surfaces 2.5Y 8/4. C3020. S108. Large handle fragment, most likely of a Zeest 80/Knossos Type 39 amphora, non-typical fabric. Reddish yellow fabric, with small light and dark mineral inclusions. Oxidised firing, handmade. Fabric and surfaces 7.5YR 6/6. C3295. U852. Rim fragment with rouletting, from a bowl, Phocaean LRC, Hayes form 3E-F. Fine fabric, with occasional small dark and white inclusions. Traces of slip in rouletting and on interior. Oxidised firing, wheelmade. Fabric and surfaces 10R 6/6. D= 34cm. C3711. U3143. Large body fragment, with horizontal combing, belonging to a Late Roman Amphora 2 (LRA 2). Fine, light orange fabric containing some tiny quartz, and some possible mica. Fabric and surfaces 5YR 7/8. Smooth surfaces. Oxidised firing, wheelmade. C3815. U1067. Small rim fragment with dot rouletting, Phocaean LRC, Hayes form 3F. Light orange fabric with smooth surfaces, containing very few small inclusions. Fabric and surfaces 2.5YR 7/6-2.5YR 7/8. D= 28cm?
290
The Cide-Şenpazar Region in the Roman Period
C4083. U2239. Small rim fragment of an amphora, possibly Late Roman, from central North Africa. Light reddish fabric, with small to medium rounded black, and small calcareous inclusions. Exterior abraded, rough feel. Oxidised firing, wheelmade. Fabric and surfaces 10R 6/8. D= 8cm.
Bibliography websites http://mail.nysoclib.org/BarringtonAtlas/BATL086.pdf. (last accessed 12 October 2013). http://www.tabula-peutingeriana.de/tp/tpx.html. (last accessed 5 September 2014). Greco-Roman sources Arrian, Periplus Ponti Euxini. [Translated by A.T. Liddle]. Bristol, Bristol Classical Press, 2003. Strabo, Geographica. [Translated by H.L. Jones]. Cambridge, MA., Harvard University Press, 1917-1932. published sources Adam, J.-P. 1984. La construction romaine. Materiaux et techniques. Grands Manuels Picard, Paris. Asheri, D., and Hoepfner, W. 1972. Forschungen an der Nordküste Kleinasiens 1. Vienna, Österreichische Akademie der Wissenschaften. Bardill, J., and Hayes, J.W. 2002. Excavations beneath the peristyle mosaic in the Great Palace of the Byzantine emperors: The pottery from site D, 1936. Cahiers Archéologiques fin de l’Antiquité et Moyen Age 50: 27-40. Belke, K. 1996. Tabula Imperii Byzantini 9. Paphlagonien und Honorias. Vienna, Verlag der Österreichischen Akademie der Wissenschaften. Bes, P.M. 2015. Once upon a Time in the East. The Chronological and Geographical Distribution of Terra Sigillata and Red Slip Ware in the Roman East. Oxford, Archaeopress. Bittner, A. 1998. Gesellschaft und Wirtschaft in Herakleia Pontike. Eine Polis zwischen Tyrannis und Selbstverwaltung. Bonn, Rudolf Habelt Verlag. Braund, D. 2005. Across the Black Sea. Patterns of maritime exchange on the northern periphery of Roman Asia Minor. In S. Mitchell and C. Katsari, eds., Patterns in the Economy of Roman Asia Minor. Swansea, Classical Press of Wales: 115-38. Broughton, T.R.S. 1975. [reprint] Part 4: Roman Asia. In T. Frank, ed., An Economic Survey of Ancient Rome. New York, Octagon: 499-918. Corsten, T. ed. 2010. A Lexicon of Greek Personal Names 5A. Coastal Asia Minor: Pontos to Ionia. Oxford, Clarendon Press. de Boer J. 2013. Stamped amphorae from the Greek Black Sea colony of Sinope in the Mediterranean during the Hellenistic period. In G. R. Tsetskhladze, S. Atasoy, A. Avram, S. Dönmez and J. Hargrave, eds., The Bosporus: Gateway between the Ancient West and East (1st Millennium BC-5th Century AD). Proceedings of the Fourth International Congress on Black Sea Antiquities, Istanbul, 14th-18th September 2009. Oxford, Archaeopress: 109-14. De Paepe, P., and Vermeulen, F. 1999. Etude microscopique et chimique de pithoi, matériaux de construction et autres terres cuites antiques de Pessinonte (Turquie). Anatolia Antiqua 7: 115-27. De Paepe, P., Devos, G., and Vermeulen, F. 2001. Les pithoi de Pessinonte (Anatolie centrale). Donnees analytiques complementaires (Annexe 4). Anatolia Antiqua 9: 89-101. Decker, M.J. 2001. Food for an empire: Wine and oil production in north Syria. In S.A. Kingsley and M.J. Decker, eds., Economy and Exchange in the East Mediterranean during Late Antiquity. Oxford, Oxbow Books: 69-86.
Bibliography
291
Demir, M. 2007. The trade in salt-pickled hamsi and other fish from the Black Sea to Athens during the Archaic and Classical periods. In G. Erkut and S. Mitchell, eds., The Black Sea. Past, Present and Future. London, British Institute at Ankara: 57-64. Devos, G., De Paepe, P. and Vermeulen, F. 1999. The pithoi from the ancient Anatolian city of Pessinus: An integrated archaeological and petrographical analysis. Bulletin Antieke Beschaving 74: 79-110. Domżalski, K. 2011. Late Roman pottery from Pompeiopolis. In L. Summerer, ed., Pompeiopolis 1. Eine Zwischenbilanz aus der Metropole Paphlagoniens nach fünf Kampagnen (2006-2010). Langenweißbach, Beier & Beran: 163-78. Düring, B.S., and Glatz, C., 2010. The Cide Archaeological Project 2009: First results. Anatolia Antiqua 18: 203-13. Düring, B.S., Glatz, C., and Şerifoğlu, T.E. 2012. The Cide Archaeological Project 2011: Third preliminary report. Anatolia Antiqua 20: 167-75. Erciyas, D.B. 2007. Cotyora, Kerasus and Trapezus: The three colonies of Sinope. In D.V. Grammenos and E.K. Petropoulos, eds., Ancient Greek Colonies in the Black Sea 2. Oxford, Archaeopress: 1195-206. Erten, H.N., Kassab Tezgör, D., Türkmen, I.R., and Zararsız, A. 2004. The typology and trade of the amphorae of Sinope. Archaeological study and scientific analyses. In J. Eiring and J. Lund, eds., Transport Amphorae and Trade in the Eastern Mediterranean. Athens, Danish Institute at Athens: 103-15. Fırat, N. 2010. Les céramiques sigillées mises au jour dans l’atelier de Demirci. In D. Kassab Tezgör, ed., Les fouilles et le matériel de l’atelier amphorique de Demirci près de Sinope. Paris, Institut Francais d’Etudes Anatoliennes: 187-95. Gabrielsen, V., and Lund, J., eds. 2007. The Black Sea in Antiquity. Regional and Interregional Economic Exchanges. Aarhus, Aarhus University Press. Giannopoulou, M. 2010. Pithoi: Technology and History of Storage Vessels Through the Ages. Oxford, Archaeopress. Hayes, J.W. 1972. Late Roman Pottery. London, The British School at Rome. — 1992. Excavations at Saraçhane. in Istanbul 2. The Pottery. Princeton, Princeton University Press. — 2008. Roman Pottery. Fine-Ware Imports. Princeton, Princeton University Press. Japp, S. 2009. Archäometrisch-archäologische Untersuchungen an Keramik aus Pergamon und Umgebung. Istanbuler Mitteilungen 59: 193-268. Jobst, W. 2005. Das Palastmosaik von Konstantinopel. Chronologie und Ikonographie. In H. Moulier, ed., La mosaïque gréco-romaine 9.2. Rome, École française de Rome: 1083-101. Jones, A.H.M. 1937. Cities of the Eastern Roman Provinces. Oxford, Oxford University Press. Karagiorgiou, O. 2001. LR2: A container for the military annona on the Danubian border? In S.A. Kingsley and M.J. Decker, eds., Economy and Exchange in the East Mediterranean during Late Antiquity. Oxford, Oxbow Books: 129-66. Kassab Tezgör, D. 2010. Typologie des amphores sinopéennes entre les IIe-IIIe s. et le VIe s. ap. J.-C. In D. Kassab Tezgör, ed., Les fouilles et le matériel de l’atelier amphorique de Demirci près de Sinope. Paris, Institut Français d’Etudes Anatoliennes: 121-40. Kassab Tezgör, D., Lemaître, S., and Pieri, D. 2003. La collection d’amphores d’Ismail Karakan à Sinop. Anatolia Antiqua 11: 169-200. Knipowitsch, T. 1929. Die Keramik römischer Zeit aus Olbia in der Sammlung der Eremitage. Frankfurt, Joseph Baer. Laflı, E., and Christof, E. 2012. Hadrianopolis 1. Inschriften aus Paphlagonia. Oxford, Archaeopress. Laflı, E., and Kan Şahin, G. 2012a. Terra sigillata and red-slipped ware from Hadrianopolis in southwestern Paphlagonia. Anatolia Antiqua 20: 45-120. — 2012b. Pottery from southwestern Paphlagonia 1: terra sigillata and red-slipped ware. In B. Ramminger and O. Stilborg, eds., Naturwissenschaftliche Analysen vor- und frühgeschichtlicher
292
The Cide-Şenpazar Region in the Roman Period
Keramik 2. Methoden, Anwendungsbereiche, Auswertungsmöglichkeiten. Bonn, Rudolf Habelt Verlag: 179-216. Lawall, M.L. 2004. Archaeological context and Aegean amphora chronologies: A case study of Hellenistic Ephesos. In J. Eiring and J. Lund, eds., Transport Amphorae and Trade in the Eastern Mediterranean. Athens, Danish Institute at Athens: 171-88. Lund, J. 2007. A fishy business: Transport amphorae of the Black Sea region as a source for the trade in fish and fish products in the Classical and Hellenistic periods. In T. Bekker-Nielsen, ed., Ancient Fishing and Fish Processing in the Black Sea Region. Aarhus, Aarhus University Press: 161-9. Madsen, J.M. 2009. Eager to be Roman. Greek Response to Roman Rule in Pontus and Bithynia. London, Duckworth. Magie, D. 1950. Roman Rule in Asia Minor to the End of the Third Century after Christ. Princeton, Princeton University Press. Marek, C. 1993. Stadt, Ära und Territorium in Pontus-Bithynia und Nord-Galatia. Tübingen, Ernst Wasmuth Verlag. — 2009. Hellenisation and romanisation in Pontos-Bithynia: An overview. In J.M. Højte, ed., Mithridates VI and the Pontic Kingdom. Aarhus, Aarhus University Press: 35-46. Matthews, R., and Glatz, C., eds. 2009. At Empire’s Edge: Project Paphlagonia. Regional Survey in North-Central Turkey. London, British Institute at Ankara. Matthews, R., Metcalfe, M., and Cottica, D. 2009. Landscapes with figures: Paphlagonia through the Hellenistic, Roman and Byzantine periods, 330 BC-AD 1453. In R. Matthews and C. Glatz, eds., At Empire’s Edge: Project Paphlagonia. Regional Survey in North-Central Turkey. London, British Institute at Ankara: 173-226. McGing, B. 2005. Subjection and resistance: To the death of Mithridates. In A. Erskine, ed., A Companion to the Hellenistic World. Malden, Blackwell: 71-89. Meiggs, R. 1982. Trees and Timber in the Ancient Mediterranean World. Oxford, Oxford University Press. Mitchell, S. 1993a. Anatolia. Land, Men, and Gods in Asia Minor 1. The Celts in Anatolia and the Impact of Roman Rule. Oxford, Clarendon Press. — 1993b. Anatolia. Land, Men, and Gods in Asia Minor 2. The Rise of the Church. Oxford, Clarendon Press. — 2005. Olive cultivation in the economy of Roman Asia Minor. In S. Mitchell and C. Katsari, eds., Patterns in the Economy of Roman Asia Minor. Swansea, Classical Press of Wales: 83-113. Mommsen, H., and Japp, S. 2009. Neutronenaktivierungsanalyse von 161 Keramikproben aus Pergamon und Fundorten der Region. Istanbuler Mitteilungen 59: 269-86. Peña, J.T. 2007. Roman Pottery in the Archaeological Record. Cambridge, Cambridge University Press. Pettegrew, D.K. 2007. The busy countryside of Late Roman Corinth. Interpreting ceramic data produced by regional archaeological survey. Hesperia 76: 743-84. Pieri, D. 2005. Le commerce du vin oriental à l’époque byzantine (Ve-VIIe siècle). Le témoignage des amphores en Gaule. Paris, Institut Français du Proche-Orient. — 2007. Béryte dans le grand commerce méditerranéen. Production et importation d’amphores dans le Levant protobyzantin (Ve-VIIe s. ap. J.-C.). In M. Sartre, ed., Productions et échanges dans la Syrie gréco-romaine. Lyon, Maison de l’Orient méditerranéen: 297-327. Quilici, L. 2008. Land transport, part 1: Roads and bridges. In J.P. Oleson, ed., The Oxford Handbook of Engineering and Technology in the Classical World. Oxford, Oxford University Press: 551-79. Reynolds, P. 2004. The Roman pottery from the Triconch Palace. In R. Hodges, W. Bowden, and K. Lako, eds., Byzantine Butrint: Excavations and Surveys 1994-99. Oxford, Oxbow Books: 224-69. — 2010. Trade networks of the east, 3rd to 7th centuries: The view from Beirut (Lebanon) and Butrint (Albania) (fine wares, amphorae and kitchen wares). In S. Menchelli, S. Santoro, M.
Bibliography
293
Pasquinucci, and G. Giuducci, eds., Late Roman Coarse Wares 3: Cooking Wares and Amphorae in the Mediterranean. Archaeology and Archaeometry. Comparison between Western and Eastern Mediterranean. Oxford, Archaeopress: 89-114. Robinson, H.S. 1959. Pottery of the Roman Period. Chronology. Princeton, Princeton University Press. Salmeri, G. 2005. Central power intervention and the economy of the provinces in the Roman empire. In S. Mitchell and C. Katsari, eds., Patterns in the Economy of Roman Asia Minor. Swansea, Classical Press of Wales: 187-206. Schneider, G., and Japp, S. 2009. Röntgenfluoreszenzanalysen von 115 Keramikproben aus Pergamon, Çandarlı, Elaia und Atarneus (Türkei). Istanbuler Mitteilungen 59: 287-306. Slane, K.W., and Sanders, G.D.R. 2005. Corinth: Late Roman horizons. Hesperia 74: 243-97. Swan, V.G. 2004. Dichin (Bulgaria) and the supply of amphorae to the Lower Danube in the Late Roman-Early Byzantine period. In J. Eiring and J. Lund, eds., Transport Amphorae and Trade in the Eastern Mediterranean. Athens, Danish Institute at Athens: 371-82. — 2007. Dichin (Bulgaria): Interpreting the ceramic evidence in its wider context. In A.G. Poulter, ed., The Transition to Late Antiquity on the Danube and Beyond. Oxford, Oxford University Press: 251-80. Talbert, R.J.A., ed. 2000. Barrington Atlas of the Greek and Roman World. Princeton, Princeton University Press. Tsetskhladze, G.R., ed. 2012. The Black Sea, Paphlagonia, Pontus and Phrygia in Antiquity. Aspects of Archaeology and Ancient History. Oxford, Archaeopress. Vaag, L.E. 2005. Phocaean Red Slip Ware – Main and secondary productions. In M. Berg Briese and L.E. Vaag, eds., Trade Relations in the Eastern Mediterranean from the Late Hellenistic Period to Late Antiquity: The Ceramic Evidence. Odense, University Press of Southern Denmark: 132-8. Wesch-Klein, G. 2001. Bithynia, Pontus et Bithynia, Bithynia et Pontus: ein Provinzname im Wandel der Zeit. Zeitschrift für Papyrologie und Epigraphik 136: 251-6. Zhuravlev, D. 2011. Early Roman fine ware from Pompeiopolis. In L. Summerer, ed., Pompeiopolis 1. Eine Zwischenbilanz aus der Metropole Paphlagoniens nach fünf Kampagnen (2006-2010). Langenweißbach, Beier & Beran: 149-62.
Marica Cassis
11 The Cide-Şenpazar Region in the Byzantine Period At the end of the 19th century, the German archaeologist Gustav Hirschfeld travelled westward across the western Black Sea coast of Turkey, beginning in İnebolu. This journey marked one of the first archaeological forays into the region, and Hirschfeld was struck by the difficulty of the landscape and the dispersed settlements, commenting that: “The little harbours and places along the coast, so easily accessible to one another by sea, are separated by ridges of the most repellent character, so that communication between the coast and the interior is restricted to a very few lines, which were used in ancient times as they are to this day for the exportation of timber and the importation of the few and simple necessaries required by the autochthones.” (Hirschfeld 1883: 276).
Hirschfeld saw the isolation of the landscape as a longstanding feature of the region, noting that ‘ancient remains…are scanty’ (1883: 275-6). Yet, Hirschfeld’s observation that little archaeological material remains was incorrect, a result of his bias in favour of the Classical world. Indeed, it is because the western Black Sea Region in Turkey obscures the more distant past under both a shifting coastline and a harsh landscape that the remains of the Byzantine era stand out, peering out from beneath the veneer of modernity in the region. It is precisely because of the isolation of this area, both now and in the Byzantine period, that we have the remains of a continuous Byzantine presence in the area from the seventh to the 15th century. This continuity, which involves primarily fortresses and farmsteads undiscussed in Byzantine historical literature, provides a glimpse into a world largely unknown in Byzantine history, and generally ignored in our interpretations of it. Given how little is known about the region from the Byzantine period, as there are virtually no texts which speak of this area specifically (although, as noted below, there are many references to the theme of Paphlagonia), or make reference to identifiable towns or kastra, the corpus of data from the Cide Archaeological Project (CAP) takes on a larger significance. The Byzantine remains range from spolia built into walls, gardens and cemeteries, to fortresses and farms sitting on top of earlier settlements and strongholds. The remains consist of examples of military and religious architecture as well as the detritus of domestic life. The volume and diversity of this data stresses the continuity of habitation in the region throughout the Byzantine period. Northern Paphlagonia offers an exceptional view of such continuity, given its protected location on the Black Sea, separated from the calamities of the south by the Pontic mountains. Thus, while still affected by the political upheavals of the Byzantine world, this coastal area was able to remain a protected enclave longer than other areas of the empire. While not highly urbanised in the Byzantine period, it was nonetheless continuously inhabited and, at times, protected by the empire.
Methodological Problems
295
By looking at the local and the imperial in tandem, we can gain a complete sense of the settlement patterns in this region. To this end, the survey material is organised around types of sites and settlements, because it is within these groupings that the usage – and change of usage – of sites and landscapes becomes visible.
11.1 Methodological Problems Any survey project covering the Byzantine period is beset by a number of problems of interpretation, particularly concerning dates and site identification. To begin with, both defining periods and assigning dates to architectural remains continue to be serious issues in Byzantine studies. Consistent period definitions are unclear in Byzantine studies for a variety of reasons (Anderson 2012:14; Vroom 2005: 15-16). There is an inconsistency between definitions of what is Byzantine between Anatolia and Greece, and the Levant and North Africa. Second, there is no common consensus on when Late Roman stops and starts, a problem compounded by the usage of the terms Late Antiquity and, more recently, ‘Medieval’ (Whittow 2009). Finally, this lack of consensus on dates has led to a vague use of terminology for the Early, Middle, and Late Byzantine periods. As Adam Izdebski (2013: 21-22) has pointed out, this has created an impossible issue in terms of comparing data, since often these dates are not defined anywhere in the works in question. Rather, these terms are used as if there is consistency in terminology in the field. All of these problems mean that it is incumbent upon researchers to be clear about their dates up front. To that end, I will use the term Late Roman to refer to the material up to the seventh century (roughly AD 650), as there is important scholarly support for the idea that the period up to the rise of Heraclius can still be seen as representing the remnants of the Late Roman worldview (Haldon 1997: 31-40). Haldon has convincingly argued that it is after the reign of Phocas (AD 602-610) that the Late Roman world became more Medieval and less dependent on the Roman past, largely due to a shift in the way the population interacted with and believed in its emperors. This period also reflects an accepted change in the ceramic assemblage for Byzantine sites in western Anatolia and Greece, which saw the rise of the ‘first post-Roman lead glazed wares’ (Vroom 2005: 15-6) and marks the beginning of the decline of traditional Roman-type fine wares. To this end, the term Early Byzantine will be roughly defined as the seventh to ninth centuries, following Vroom’s ceramic chronology. There will necessarily be some cross-over with the Roman dates. The tenth - early 13th centuries represent the Middle Byzantine period revival of the empire, particularly under the Macedonian and Komnenian emperors, as well as the corresponding rise of new pottery types outlined by Vroom (2005: 15-6). The protected nature of northern Paphlagonia left this area in the hands of the Byzantine empire longer than other areas, such as southern Paphlagonia or Bithynia, and consequently the later end of Vroom’s sequence can, I think, be safely applied here. The 13th-15th centuries represent
296
The Cide-Şenpazar Region in the Byzantine Period
the Late Byzantine period, which needs to be seen as a period of disunity. The term ‘Late Byzantine’ is problematic because it largely ignores the fragmented nature of Byzantium in this period, given that by this period the former empire had fragmented into the empire of Nicaea, the kingdom of Trebizond, and the expanding Seljuk and Ottoman kingdoms. Further, Italian merchants left a legacy of Medieval Italian pottery and construction in the region as well (chapters 12 and 13). Ultimately, the terminology for dating this period remains unclear because of a lack of clearly dateable material culture. As Izdebski has pointed out, sites are often termed Byzantine in surveys with no really clear evidence for this designation other than a few sherds of often undiagnostic pottery and unsubstantiated beliefs about site usage and structures in the Byzantine period (2013: 21-23). Although Byzantine fortresses and remains dot the Anatolian landscape, the lack of stratigraphic excavation has left us with little clear comparanda through which to apply secure dates. Some attempts have been made to date construction types, with limited success (Bryer and Winfield 1985: 76-8; Foss and Winfield 1986), and often such dates are designated based simply on unproven ideas of the ebb and flow of Byzantine history (Izdebeski 2013: 24, 47-51). Further, attempting to assign strict dates to structures disregards the issue of continuity in the Byzantine world, even though the majority of Byzantium was agrarian and was consequently defined by exactly this – the reuse and rebuilding of earlier structures and sites. This problem is coupled with difficulties in pottery analysis. While progress is being made in the assessment of Byzantine ceramics (Degeest 2000; Vroom 2003; 2004; 2005; Armstrong 2008; 2009; Vionis et al. 2010), Byzantine survey remains limited by a number of factors. First, as Vroom has shown: “surface survey is in the end dependent on the construction of detailed ceramic chronologies and typologies. These can only be obtained through stratigraphic excavation of sites that are local to the survey area. Too often the dating and analysis of surface assemblages is reliant on a handful of ceramics publications, many of which relate to sites which lie some considerable distance from the survey area.” (Vroom 2004: 326).
The comparanda available for the CAP survey region are limited in various ways. While there are important excavated comparanda from Tios (Atasoy 2012; Atasoy and Yıldırım 2011) and Pompeiopolis (Domżalski 2011), these are largely focused on the Late Roman period. The work at Komana has the potential of providing important parallels (Erciyas and Sökmen 2010: 123-41), but is relatively far from our target area. More extensive evidence exists from other surveys, including those of Sinop (Doonan 2004), Amastris (Crow and Hill 1990; 1995), inland Paphlagonia (Matthews and Glatz 2009: 173-226) and Euchaita/Avkat,1 but this is beset by the same problems that CAP faces. The ceramic evidence is a particular problem in this area since the majority
1 http://www.princeton.edu/avkat/reports/
Methodological Problems
297
of the ceramic finds are utilitarian wares, and thus represent – except in the case of amphorae – pottery assemblages that remained largely unchanged over long periods of time. There is very little fine ware. One change in the way that ceramics are assessed for this period will be included here, which is the preference for the terms kitchen ware and table ware instead of the traditional fine ware and coarse ware divisions (see Vroom 2005: 19-21; Armstrong 2009: 429-43). The lack of fine ware, coupled with the fact that many of the so-called kitchen wares are, in fact, quite fine, makes these more appropriate terms. Too much dependence is placed on a few glazed sherds in Byzantine archaeology, which, in the region of northern Paphlagonia, almost certainly represent only one occupational phase at sites used over long periods of time. Thus, while these identify one phase of usage, they do not define the entire life of the site. This was brought home to me while we were surveying. On a couple of occasions, we were shown large numbers of glazed Byzantine and Islamic-period sherds collected by locals who had been picking them up for years because of their beauty. The removal of such sherds obviously changes the nature of how sites are identified and dated and creates a notable handicap for these later periods that, I would argue, probably does not affect earlier periods. In many ways, this makes a statistical analysis of the ceramics pointless. Further, an overdependence on imported fine ware largely disregards the fact that ceramic production in much of the Byzantine period was almost certainly local, as noted, for example, at Sagalassos (Poblome et al. 2000; 2001). While survey material is not really conducive to defining this pottery more substantially, the clusters present in the Cide and Şenpazar material indicate new questions about local production that could be defined through excavations. Finally, one of the issues that has long plagued Byzantine studies is the issue of an overreliance on monumental remains as representative of empire. The most visible elements in a landscape are, of course, the fortresses and the churches. Yet, these are only half of the equation in understanding the life cycle of a region or site. Traditional approaches to the extant structures in northern Paphlagonia have been limited to these monumental remains. This approach characterises the monumental Byzantine Monuments and Topography of the Pontos (Bryer and Winfield 1985), or works about the Classical periods, which are much better represented by monumental remains (Marek 1993; 2003). While Bryer and Winfield’s work is an important survey of the eastern Black Sea region in Turkey, it provides a somewhat skewed vision because it largely negates the significance of the remains of the ordinary population. Further, even when the survey material is more comprehensive, as for example in Belke’s Paphlagonia und Honorias volume of the Tabula Imperii Byzantini (1996), the dates tend to be either too broad, or do not adequately reflect the issues in the field outlined above. As Crow (2009: 26) has shown, this means that more often than not, there has been little understanding of these monumental structures in terms of their place in wider society. Further, there has been even less consideration of the place that smallscale remains, such as farms and villages, held in the Byzantine empire.
298
The Cide-Şenpazar Region in the Byzantine Period
11.2 Historical Survey Northern Paphlagonia, the region covered by CAP, is entirely absent from Byzantine sources, and, until Italian merchants began creating portolans for their trade routes in the region, none of the villages or towns are identifiable beyond those noted in earlier Roman sources (for example, Belke 1996: 245-6). Although the theme of Paphlagonia is very visible in the sources, we must be careful not to overgeneralise about the entire province. The Pontic mountains provided a powerful barrier between the northern and southern parts of Paphlagonia. Yet, through a combination of settlement patterns, pollen analysis, and careful extrapolation from the sources about the entire theme, it is possible to gain some sense of the continuity of habitation in this particular region. It is somewhat ironic that a region that was probably never densely populated or highly urbanised in the Byzantine period, thus resulting in its absence from the sources, should provide one of our clearest examples of continuity of habitation. However, the Black Sea coast and its immediate hinterland fared better than many other regions of Anatolia, precisely because of its protected peripheral location, and thus provides an important example of the continuity of Byzantine settlements. Analysis of this region needs to begin with consideration of why it might have been important to the Byzantine empire. In fact, this is foreshadowed by the role of this area in the Roman period, when, particularly under Trajan, and recorded in the letters of Pliny the Younger, the southern coast of the Black Sea became the focus of intense building and consolidation (Pliny the Younger, Epistles [Williams 1990: 10-16]). Control of the prosperous Pontic coast, stretching as far as Sinop and Amisus, was seen as significant both in terms of the ongoing conflicts and negotiations with the Parthians and other kingdoms in the east, as well as a source of resources, such as grain, nuts, and olives (Anderson 2012: 14; Levick 1979: 127-9; Wittke et al. 2010: 201). The strategic importance of the area lead to a large amount of money being invested in the infrastructure of several places, including in Amastris (Pliny Book X: 98-9). While the specific centres of our survey area are not identifiable from these letters, Wilken (2003: 15) points out that it was at an unnamed place on the road between Amisus and Amastris that Pliny was called on to deal with local Christian populations (Book X: 96-7). Thus, it is no surprise that there was a vibrant coastal urban life in the Roman period, something confirmed to some degree by the ceramic analysis from the survey (chapter 10). Indeed, this settlement is confirmed to some extent by the appearance of a few sites known from the survey of Roman itineraries. Thus, the second-century writer Arrian identifies Kyteros (our Gideros), Aegialus (which probably refers to Cide), Thymena, and Cape Carambis in the Periplus Ponti Euxini (Arrian 14.2. [Silberman 1995: 12, 42]), as does Ptolemy of Alexandria, who wrote the Geographia, a secondcentury gazetteer and mapping guide. As outlined by Belke, Late Roman sources such as the fourth or fifth century Peutinger Atlas also list Kyteros and Cape Carambis, as do other anonymous periploi (Hongimann 1939; Belke 1996: 245-6). It is not, however,
Historical Survey
299
until the portolans of the Italian period that we once again find references to specific towns. Although there are no other specific references to these sites in the Late Roman and Byzantine periods beyond those listed above, there is a great deal in the sources about the province of Paphlagonia which hints at the significance of this coastal region. While the region covered by CAP never boasted large cities, it remained significant, as did the entirety of the Turkish Black Sea coast, as a trade route and a source of food, particularly wheat and timber. In the two Late Roman reorganisations of the provinces, first under Diocletian (AD 284-305) and then under Justinian in the early sixth century, the coastal area of Paphlagonia generally required protection. Indeed, there are two commonalities which link these periods. First, both were periods of reorganisation and, ultimately, refortification in the face of unrest. Diocletian ruled an empire that he had led out of chaos, and he felt it was necessary to impose order on it. To that end, one of his major contributions was the reorganisation of the fifty imperial provinces into a hundred new provinces, all under a reorganised administrative structure. Diocletian had much to worry about – the growing uncertainty in Europe as large numbers of tribes crossed the Danube and the increased Persian threat were only two of the many problems he faced (Treadgold 1997: 18-9). What became clear was that there had to be an effective defence of the entire empire, particularly in those places which were strategically important either for defence, or as in the case of the north shore of Paphlagonia, for food and trade (Wittke et al. 2010: 182). A similar reconsolidation of the Late Roman world occurred under Justinian, who was concerned with fortifying and protecting all areas considered part of his reconstituted Roman empire (see Procopius, Buildings III: vii.24 and IV: xi.20 for the Euxine Sea). Further, in Novel 29, Justinian reorganised the province of Paphlagonia (Justinian, Novels [Scott 1932]). While literary references remain silent after this point, archaeological and scientific data suggests a continuity which conforms to wider developments in the Byzantine world. While the traditional view of the Byzantine ‘dark ages’ (roughly the sixth and seventh centuries AD) has been one of decline and collapse, more recent scholarship suggests that this is not entirely accurate. Sites like Anemurium (Russell 1986), Sagalassos (Vionis et al. 2010), and Çadır Höyük (Cassis 2009; Steadman et al. 2013) indicate that there was continuity of occupation in these periods, but that the communities concerned were considerably smaller than they had been in the Late Roman period. The standard theory centres around the idea of a widespread ruralisation of the empire as people abandoned diminished urban centres. On a formal level, this has been noted as a shift from the polis (city) to the kastron (rural fortifications) (Dunn 1994; Crow 1996: 23), and the reasons behind the change are complex. First, we now understand that environmental changes began from approximately the beginning of the sixth century in Anatolia (Haldon 2007: 220-7). The warmer, wetter, Late Roman world gave way to a period of droughts which could have been weathered, perhaps, if it were not for a series of invasions, first by the
300
The Cide-Şenpazar Region in the Byzantine Period
Persians (in the early seventh century), and then by the Arabs (in the late seventh and early eighth centuries; see in particular Haldon 2007). This in turn led to the collapse of cities which were already in decline. Areas not directly affected by these changes, such as our survey region, suffered collateral damage from the overall turmoil in the empire (Haldon 1997: 104-5). This in turn led to a ‘ruralisation’ of the empire, that was reflected to a lesser or greater extent in specific areas. It did not, however, result in the complete collapse of Byzantine society. While the wider region of Paphlagonia suffered greatly under the Arab incursions into Anatolia (Belke 1996: 69-82; Matthews and Glatz 2009: 190-1), new analysis of pollen data suggests that the coastline area north of the Pontic mountains, although in decline, still maintained a substantial human population (Izdebski 2012: 57-8). Thus, Izdebski uses the pollen data from northern Turkey, and Abant Lake in particular (Bottema et al. 1993-4), to illustrate that, while there is a decline in the Late Roman agricultural indicators (primarily olive and grape), the evidence for human agricultural activities does not disappear completely, in contrast to sites on the other side of the Pontic mountains (Izdebski 2012: 57-8; Mathews and Glatz 2009: 190-91). Rather, agriculture continues consistently throughout the Byzantine period, albeit at a reduced level and with an increase in pasturing. Throughout Anatolia: “what remained was instead a pattern of defended villages and fortresses, the strongest of which often came to serve as the administrative and military centres; and on the coasts of the Black Sea, the Aegean, the Adriatic and in south-west Asia Minor, there are a few isolated ports and emporia.” (Haldon 1997: 120).
This becomes apparent when we consider the finds (discussed in detail below) from the region. Although little can be very specifically dated, the extant material does indicate that there was extensive occupation of the region throughout the period. At the end of the Late Roman period and into the Early Byzantine period, settlement remained scattered throughout the region, while by the end of the Early Byzantine period and into the Middle Byzantine period, the upland hinterland away from the coastline seems to have housed the majority of the population. The majority of Roman, Late Roman, and Early Byzantine material, both ceramics and dateable architecture, hugs the coastline (also chapter 10). This is not to say that it does not appear elsewhere, quite the contrary, in fact, but that the Roman period occupation seems to have followed a standard pattern of urban centres with hinterland villas and farms. At the start of the ‘dark ages’ and into the Middle Byzantine period, the ceramic assemblage largely disappears on the coastline. This is a problematic assertion, of course, since the Byzantine period pottery is largely represented by coarse wares which are often indistinguishable from Late Roman parallels. However, there is also very little Middle or Late Byzantine glazed or fine table ware from these coastal sites. This is roughly paralleled in the data from the Sinop Regional Survey (Doonan 2004: 117, 159). Thus, the ‘dark ages’ in this region represents the same pattern as is seen
Historical Survey
301
throughout Anatolia: the diminishing of urban centres, the rise of the village, and the disappearance of the traditional urban Roman way of life. Such observations are confirmed by the pollen analysis discussed by Izdebski (2012: 53-8), which suggests that cereals and livestock grazing remained constant in the area throughout this period, possibly until as late as about AD 1000, indicating a substantial rural habitation. The Middle Byzantine period is equally poorly illuminated in terms of historical sources concerning this specific region, although there are an increasing number of references to the larger region of Paphlagonia. More importantly, however, historians have created a narrative for Anatolia in the period stretching from roughly the late eighth century through to the 11th century which accounts for both the continuation of habitation of regions such as that covered by CAP and for the evolution of thematic power. Several prominent scholars have argued that, following the move to village organisation in the sixth and seventh centuries, as military leaders were given lands in the provinces as compensation, they settled into their roles as landowners, creating a system of provincial nobles (dynatoi) who amassed large amounts of land and subsequently became somewhat of a threat to the government in Constantinople (Harvey 1989: 30; Angold 1997: 29). Scholars cite a number of examples and events which suggest that this was the case. Most notable for our region is St. Philaretus, a Paphlagonian saint who was also a landowner of some means in the ninth century (Nesbitt 1969; Haldon 1997: 139). His control over large amounts of lands and livestock suggests that such notables could create significant power bases in rural Anatolia. As Whittow (1995: 62-3) has argued, these families were not housed behind fortifications (as in the west), but rather held their power through the accumulation of large amounts of generally unfortified land. The shift away from a village based society to one where land was controlled by these notables led to a decentralisation of power from the capital (Harvey 1989: 41) and the loss of peasant autonomy (Harvey 1989: 41; Angold 1997: 29-30). That this was happening behind the scenes becomes even more apparent with the rise to prominence of families from particular regions at various times. Thus, for example, Magdalino (1998; see also Crow 1996) argues that under the Komnenoi, themselves from Kastamonu, prominent Paphlagonians yielded considerable influence at the imperial court. This type of development cannot be documented in our specific area. However, there are three pieces of circumstantial evidence which suggest continuity and even rural settlement growth in the region covered by CAP during the early part of the Middle Byzantine period. First, the ceramic assemblages for several areas, most notably Okçular and Abdulkadir, indicate continuous settlement. Also, at the end of the Middle Byzantine period, we begin to find glazed wares again, representing both external trade and more affluence in the region. The growth in the trade routes along the southern Black Sea coast also suggests an increased interest in the area (Vryonis 1971: 14-5; Wittke et al. 2010: 246). Secondly, the fortifications along the Cide coastline were rebuilt and utilised, probably because of the increased trade in the
302
The Cide-Şenpazar Region in the Byzantine Period
area. Finally, Izdebski (2012: 59) draws attention to increased pollen representing human activity in the ninth and tenth centuries in particular. Certainly the Middle Byzantine period appears to show continuous occupation and some growth, even as the Turkic tribes (starting with the Danishmendids) usurped the area to the south of the Pontic mountains (chapter 13). By the late 11th century, however, things had begun to change substantially for the Byzantine empire. This is confirmed by the pollen data which has been discussed for this period and region by Izdebski (2012: 59), who notes that the decline in human indicators really starts for this region in the 11th century. Indeed, this is no wonder. As the entire province of Paphlagonia became caught up in regional disputes between the Byzantines and the Turkic groups flooding into Anatolia (Booth 2004), the coastal region was no longer protected. While it does not seem likely that the Danishmendids, who took Kastamonu in the late 11th century, ever really held the northern part of the region, there was almost certainly widespread disruption of trade and travel networks in the whole area. Although it was reconsolidated by the Komnenoi in the early 12th century, the region was shortly thereafter enveloped by the raids of the Seljuks, the second Turkic group to move into the area. Again, the earliest effects of the Turks in this region were largely raids and continuous disruption (Vryonis 1971: 110), but this was enough to bring the region into crisis. Vryonis claims, based largely on his interpretation of the Alexiad of Anna Komnene, that: “The whole of the coastal regions from Heracleia to Trebizond as well as the hinterland...were turned into areas of frequent raiding and plundering, and were in many cases partially abandoned by the Christian population...Though the Byzantines were able to recover the coastal regions by 1085 and even some of the hinterland, at least temporarily, northern Asia Minor remained a scene of continuous raiding and warfare throughout the twelfth century...There is very little specific information as to the fate of individual coastal areas.” (Vryonis 1971: 160).
If the 12th century saw this region caught in the crossfire of general strife, the early 13th century (and the transition to the Late Byzantine period) marked a period where the various competing powers recognised the inherent significance of controlling this region – for trade and as a source of food and security. Thus, the entire region was controlled by both the empire of Trebizond initially, and then the kingdom of Nicaea (Treadgold 1997: 713-20; Booth 2004). Identifying when and for how long these states controlled our area is impossible to determine in the survey record. However, within a short period of the retaking of Constantinople in AD 1261 from the Latins, this entire region, including the coastline, passed first to the Seljuks and subsequently the Ottomans (Vryonis 1971: 137, also chapter 13, this volume). The coastline, however, remained marked by significant possessions of the Genoese, including the city of Amastris, during the 13th and early part of the 14th century (Crow and Hill 1995; 1990; Kazhdan 1991: sv Amastris). While this did not result in large-scale settlements in the area covered by CAP, the Genoese presence did leave a notable archaeological signature in the region, largely connected to the fortresses on the coastline (chapter 12).
Byzantine Finds and Sites in Cide and Şenpazar
303
11.3 Byzantine Finds and Sites in Cide and Şenpazar The most exciting thing about the material discussed below is that it unequivocally shows the continuity of Byzantine communities from the Early Byzantine to the Late Byzantine period. As noted above, the relatively protected nature of this region, hidden behind the Pontic mountains, allowed it to survive and even thrive in ways not open to the southern part of the theme of Paphlagonia and to other parts of Anatolia. The remains discussed below fall into numerous categories, domestic, fortification, and religious, but they were all interconnected with one another and to the changing fortunes of the Byzantine empire (Fig.11.1).
0
5
İlyasbey 1
10 Km
Kuşçu 1
Sakallı Mezarlığı
Kumluca 1
Mençekli
Çamdibi Abdulkadir Ballıcı İni Okçular
Kılıçlı Mağarası
Pottery distribution (Byzantine) Tile distribution (all periods)
Fig. 11.1: Map of Byzantine settlements. Produced by Toby C. Wilkinson.
11.3.1 Ceramics As noted above, the ceramics of the Byzantine period are particularly problematic to identify and date for several reasons. First, we do not have an adequate pottery sequence from the region to clearly indicate what the development of kitchen wares (and non-glazed table wares) looked like in this period. Thus, although identifiably Byzantine sherds constitute about 11% of the overall ceramic assemblage (10.72%; 442 of 4125 sherds; see Figs. 11.2 and 11.3), the real number is probably much higher when one considers unidentifiable
304
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.2: Byzantine sherd density map. Produced by Toby C. Wilkinson.
Fig. 11.3: Byzantine tile density map. Produced by Toby C. Wilkinson.
Byzantine Finds and Sites in Cide and Şenpazar
305
body sherds and tiles. That is, there are many sherds that simply cannot be identified because we do not have a good sequence from the region. Second, the percentage of glazed wares is very low for the region covered by CAP. Of all of the ceramics recovered from the survey, the glazed wares constitute just over one percent (1.39 %; 59 sherds total), and only 13% of the identifiable Byzantine ceramics (13.35%). As noted above, removal of such ceramics from the fields by locals has certainly affected this statistic, and so this needs to be considered with caution. Finally, the presence of glazed ceramics does not necessarily mean that a site was only used during a particular period, but rather that it continued in use through this period, and so the dates of Byzantine glazed ceramics and their connection to particular sites also need to be used with some caution. Pamela Armstrong (2008: 430) has recently classified the types of ceramics for the Byzantine period into three categories, the international wares which travelled extensively throughout Byzantium, the regional wares which travelled mostly among neighbouring provinces, and the local wares produced within the context of small areas. By far the majority of recovered Byzantine ceramics seem likely to have been produced locally, as there are some distinct fabrics represented here. Some regional wares from throughout the Byzantine period are identifiable in the assemblage, primarily consisting of amphorae from the Black Sea and east coast of Turkey. The few international wares are restricted to later glazed wares, which appear in very small amounts. This speaks to the relative isolation of this region in the Byzantine period. Nevertheless, the ceramic assemblage from CAP contains representative samples from five categories: glazed table wares; unglazed table wares; kitchen wares; amphorae; and tiles. The majority of the extant evidence represents the domestic sites identified in the survey. Specific sherds used for dating are discussed in relation to the sites below.
11.3.1.1 Glazed Table Wares The majority of glazed table wares are in very fragmentary condition. They originate largely from either Constantinople or Thessalonica, as would be expected along the Black Sea coast. Although there are examples of glazed table wares from throughout the Byzantine period, the highest concentration come from the Late Byzantine period (Fig. 11.4). There is little that can clearly be identified as Early Byzantine, which is not surprising given the continued use of red-slip wares, which are often difficult to isolate from Late Roman red-slip wares in the survey record, and the decline in trade throughout the Early Byzantine period (Dark 2001: 54). Indeed, the only Early Byzantine glazed ware is an example of Glazed White Ware, which is known primarily from the excavations at the Saraçhane in Istanbul. Among our survey finds this is the only piece2 that almost certainly dates to the Early Byzantine period.
2 C3243 (Fig. 11.4).
306
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.4: Glazed table wares. Produced by Lorraine McEwan.
Byzantine Finds and Sites in Cide and Şenpazar
307
The Middle Byzantine wares are restricted to another piece of later white ware3 and a piece of Incised Sgraffito ware.4 The majority of the Late Byzantine sherds are variations of Sgraffito ware.5 An example of Zeuxippus ware6 and two sherds of slip-painted ware are also represented in the corpus.7
11.3.1.2 Unglazed Table Wares This section refers to (mostly) Early Byzantine ceramics which mirror earlier Roman forms (Fig. 11.5). These ceramics differ from earlier Roman table wares primarily in relation to their fabrics (Armstrong 2008: 429, 436; Dark 2001: 31-58), which contain more inclusions, making these pastes closer to medium fabrics than fine fabrics. While Roman table ware was largely imported, during the Early Byzantine period, these imports seem to have ceased and ceramics were produced locally. This resulted in forms which were very similar to Roman forms, but which have fabrics identifiable by an increase in grit, mica, and/or limestone, and a decrease in grog. These are hard, red fabrics, which no longer have the soft, powdery, orange paste of the earlier Roman ceramics. Without excavation, it is not possible to claim with complete certainty that these date to the Early Byzantine period. However, the majority of forms represented are common table wares for the period, and include a large number of bowls and jugs (Dark 2011: 32-3). Further, a majority of these are found in proximity to earlier Roman pottery, in locations such as coastal settlements, suggesting that these fabrics reflect the last period of intense habitation at the end of the Late Roman or the beginning of the Early Byzantine period. The best evidence for this transition in ceramics, both for this group and the following kitchen wares, comes from the deposits found along the coastline, particularly at Cide (S70; S71; S86)8 and at Kalafat (S72). There are also good examples of these types of ceramics at sites which show long periods of occupation, including Abdulkadir,9 Okçular,10 and Çamdibi.11
3 C2032 (Fig. 11.4). 4 C3322 (Fig. 11.4). 5 C184, C246, C650/651, C3082 (all Fig. 11.4). 6 C715 (Fig. 11.4). 7 C3076, C3077 (Fig. 11.4). 8 C984 (Fig. 11.5). 9 C1046, C1598, C2105 (all Fig. 11.5). 10 C1090, C1404 (all Fig. 11.5). 11 C3349 (Fig. 11.5).
308
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.5: Unglazed table wares. Produced by Lorraine McEwan.
Byzantine Finds and Sites in Cide and Şenpazar
309
11.3.1.3 Kitchen Wares This is a group of vessels which includes forms like large pitchers,12 pots, jars, and bulbous cooking pots13 characteristic of particularly the Middle and Late Byzantine periods (Dark 2001: 31-52; Vroom 2005: 17-8, 136-7; Armstrong 2008: 429-37) (Fig. 11.6). As with the ceramics above, the fabrics of these vessels also include larger and more frequent inclusions, making the fabrics coarser. These fabrics are often dark red or brown-red in colour. Cooking pots and pitcher handles are the most prevalent remaining ceramics. 11.3.1.4 Amphorae As would be expected along the coast of the Black Sea, there were also a number of amphora fragments (Fig. 11.7). The two most common types of fragments are ribbed body sherds and pieces of handle. The amphorae date to both the Early and the Middle Byzantine periods. Included below are the fragmentary pieces which do not fit nicely into the Late Roman categories, and which probably represent Early Byzantine variations.14 C942 is a Middle Byzantine amphora.15 11.3.1.5 Tiles Tiles make up a vast portion of the ceramic assemblage (Fig. 11.7), and are difficult to date for numerous reasons. As pointed out by Witte-Orr (2007: 296) in relation to the tile assemblage at Amorium, tiles were often reused. Second, they were in common usage throughout the Hellenistic, Roman, and Byzantine period, which makes a typology both difficult and necessary. While a typology is not possible from our survey data, some basic observations can be made. There appear to be two basic types of tiles present in our data set, one which can probably be identified with the Roman or Early Byzantine period, and one with later Byzantine sites. The first type is exemplified by the corpus in the region of Çamdibi, which contained a large selection of tiles that all seem to roughly belong to one type. The majority of these tiles have the tegula cut across the top, either diagonally or straight across, a typology16 as being identifiably Early Byzantine or possibly Roman based on the stratigraphy at Amorium (2007: 301-2). There were almost no finds at Çamdibi which were identifiably later than the Middle Byzantine period, suggesting a rough date for this type of tile in the Early Byzantine period. At the sites of Okçular and Abdulkadir, there was a much more varied corpus of tiles consisting of some like those found at Çamdibi, as well as large flat pan tiles, the second type of tile present in the assemblage, which are more commonly associated with later Byzantine sites.17
12 From Abdulkadir and Cide: C1602, C3268 (all Fig. 11.6). 13 From Ballıcı İni and Okçular Kale İni: C1722, F28, C3036-3039 (all Fig. 11.6). 14 C24, C992, C1027, C3867, C3934 (all Fig. 11.7). 15 C942 (Fig. 11.7). 16 C3451-53, C3453-62 (all Fig. 11.7). 17 C1330 (Fig. 11.7).
Fig. 11.6: Kitchen wares. Produced by Lorraine McEwan.
310 The Cide-Şenpazar Region in the Byzantine Period
Byzantine Finds and Sites in Cide and Şenpazar
Fig. 11.7: Amphorae and tiles. Produced by Lorraine McEwan.
311
312
The Cide-Şenpazar Region in the Byzantine Period
11.3.2 Byzantine Settlements Bryer and Winfield’s survey of the Pontic coast contains a telling statement in the introduction to the work: ‘[I]t is to churches and castles that we are largely confined, for we are, alas, spared the domestic debris of the medieval Pontic people.’ (1985: ix). Inherent in this statement is both sorrow and relief – Byzantine ‘domestic debris’ is almost impossible to date in survey and, until relatively recently, was perceived as unimportant by many Byzantine scholars. Yet, it is precisely the debris of settlements which provides a window into understanding continuity and change in the Byzantine period. A close look at the settlement patterns at three sites that underwent intensive survey, Okçular, Abdulkadir, and Çamdibi, indicates patterns of settlement which range from the Late Roman through to the Ottoman period (also chapter 14).
11.3.2.1 Okçular The area surrounding the modern village of Okçular, which sits on a gently sloping hillside leading away from a large fortified rocky outcrop, contains a number of features and small domestic sites which were occupied to varying degrees throughout the Byzantine period (Figs. 11.8 and 14.2). In the modern period, the area around Okçular provides fertile land for small-scale farming, and this was almost certainly the case throughout the Byzantine period. While the Early Byzantine period is visible in survey only through small-scale remains, the probable Middle and Late Byzantine remains speak to the overall viability of this landscape both for farming and for village life. The earliest Byzantine evidence at Okçular consists not of ceramics, in fact, but of spolia located near the modern village. Two column bases (S61 and S77; Figs. 11.9 and 11.10) and a column fragment (S64, CAP P02/548) are located ex situ. The column is lying in a ditch near the entrance to the village, while the two possible column bases have been reused to cover wells, one on the road in modern Okçular and the other in a field in the village (chapter 10 for a somewhat different interpretation). The two bases are different in form, which suggests two different structures. None of these finds can be linked to either an extant building, but do suggest that there was at least one large structure in the vicinity in the Late Roman or Early Byzantine period when basilical, columned churches were still common, even in the provinces (Mango 1976: 61). Other examples in the region, including a column capital identified by Kalinka at Gebeş (cited in Belke 1996: 255), further suggest this. In the Late Roman and Early Byzantine period, then, there was a population in this area sizeable and wealthy enough to sustain large structures. Beyond these spolia, however, nothing else remains in the vicinity of Okçular which suggests monumental or elite building in the subsequent periods. Rather, the remaining architecture indicates rural and agricultural habitation. The most striking structure of the Okçular region is the large roughly elliptical kale (Turkish for fortress) (S22; Figs. 11.8 and 14.2), which overshadows the valleys below (Fig. 11.11). Identified
Byzantine Finds and Sites in Cide and Şenpazar
313
Fig. 11.8: A reconstruction of Okçular Kale (S22). Produced by Victor Klinkenberg and Michele Massa.
Fig. 11.9: Column base (S61) in Okçular village (CAP-P1/404).
314
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.10: Column Base (S77) in Okçular village (CAP-P2/699).
as a kale by Belke, the fortification had at least two phases of building, including one that he dated to earlier than the Byzantine period (Belke 1996: 255). The second phase, which dates to the Byzantine period, consists of walls of mortared fieldstones, which have traditionally been considered the norm between the seventh and ninth centuries (Fig. 11.12; Mango 1976: 11-2; Foss and Winfield 1986: 26-8, 162). At the base of these walls are a number of beam slots used in construction (Fig. 11.13). Some go straight through the walls of the structure, and may have functioned as drainage holes.18 Given the large amount of precipitation in the region, a drainage system may have been necessary to ensure that water did not build up within the fortification. Although there do not appear to be any clearly articulated towers, some buildings are present in the centre of the outcrop, visible where a robber trench has been dug (CAPP1/138). The walls of these structures also seem to have been constructed of loosely mortared fieldstones, although the mortar in the internal walls is pinker and less yellow than that of the external walls (see also Belke 1996: 255), and contains more ground ceramic material. This may suggest two phases of Byzantine construction. Although such structures have traditionally been dated to the so-called ‘dark ages’, which roughly correspond to the seventh century, the date for this structure is uncertain. As noted above, common wisdom in Byzantine archaeology has
18 My thanks to Dr. M. Nikolic, Department of Classics, Memorial University of Newfoundland, for this suggestion.
Byzantine Finds and Sites in Cide and Şenpazar
Fig. 11.11: View of the valley from Okçular Kale (S22) (CAP-P-3/47).
Fig. 11.12: Mortared Field Stones, Okçular Kale (S22) (CAP-P1/131).
315
316
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.13: Beam hole, Okçular Kale (S22) (CAP-P1/128).
traditionally dated these types of rough hill fortifications to the Early Byzantine period based largely on the assumption that local populations needed such structures in the face of the Arab invasions (Foss and Winfield 1986: 165; Strobel 2002: 35-6). As Izdebski has pointed out, however, there is really no firm evidence to assign all hilltop fortresses to this period, “As there exists no reliable method of dating the standing elements of unexcavated early medieval sites which are devoid of any inscriptions, obvious ceramic finds or coins), there is no way of determining whether these sites were constructed, developed or even used during the earlier or the later part of the Middle Ages, or simply whether they were actually used at all in the Middle Ages...” (2012: 62).
Rather, it is far more useful to think of these hilltop fortresses in practical, as opposed to defensive or standardised, terms. Although there were certainly known Arab invasions into Paphlagonia (Matthews and Glatz 2009: 190-1), there is virtually no evidence to suggest that the region covered by CAP was affected by these raids in any serious way. As the populations became increasingly ruralised throughout the Early
Byzantine Finds and Sites in Cide and Şenpazar
317
Byzantine period (and into the Middle Byzantine period), such fortifications could have been used for storage, communication, and defence (Izdebski 2012: 62-3) over long periods of time. Indeed, in the case of the Okçular Kale, the distance from the coastline, as well as the somewhat blocked sight lines of the Black Sea, suggest that defence of/from the Black Sea was never the primary goal of the population which utilised this structure. Rather, clear views of the valley and proximity to the nearby agricultural communities provide more far-ranging practical explanations for this fortification. Such a use also extends the period of usage for the site. It may have been built in the Early Byzantine period, but it seems entirely likely that this kale continued to be used for safe storage or defence during the much more extensive Turkish incursions into the region (Matthews and Glatz 2009: 196-7). Unfortunately, there is no way to more specifically date the majority of these types of structures in terms of architectural style or ceramics, and the kale at Okçular is no exception. The vast majority of the ceramics found in the kale were coarse kitchen wares, and the dates for these ceramics ranged from the Hellenistic period to the Late Byzantine period. Most were not diagnostic, leaving us with no way of specifically dating the site. The only truly dateable piece was one small fragment of Sgraffito ware,19 dating to roughly the 13th or 14th centuries. A small piece of what is probably a piece of Middle Byzantine amphora20 was also found in the area. Beyond that, there are a number of pieces of Byzantine kitchen wares21 which cannot be more specifically dated. In some ways, however, this is not particularly important – there can be no doubt that such a structure was used throughout the Byzantine period for a multitude of purposes. The significance of this kale lies in its connection to the network of other sites in the vicinity. Ceramics and tile clusters, as well as material remains in the landscape (including caves and agricultural structures), speak to a continued Byzantine presence in the region over a very long period. While the dates for these remains are somewhat obscure, the consistency of the material speaks to the rural population that inhabited most of Anatolia throughout the Byzantine period. Among the more difficult sites to interpret in this region is Okçular Kale İni (S24; Fig. 5.11), a large cave located underneath the kale which is primarily significant for the large cache of Prehistoric artefacts and ceramics (chapters 5 and 6). However, near the entrance to the cave there was also a small cache of Byzantine pottery which probably had been displaced from elsewhere in the cave. Of this, three pieces are distinctly identifiable as Middle or Late Byzantine wares22 and probably date to the 12th through 14th centuries. There were also several large pieces of kitchen ware, including kitchen pots and jugs, which are less clearly dateable, but also probably come from
19 C183. 20 Günsenin I type: C179. 21 see above for discussion; C255, C257, C275. 22 C3076, C3077, C3082 (see Fig. 11.4).
318
The Cide-Şenpazar Region in the Byzantine Period
the middle to late period.23 As Bes (chapter 10) has pointed out, there was also Late Roman pottery from this cache, indicating that it was used continuously from then until the end of the Byzantine occupation of the area. The nature of the use of this cave is uncertain; its presence in proximity to the village at Okçular suggests that it was probably used for storage or safety. It may also have functioned as a hermit’s cave, although the lack of any sort of religious identification makes this less likely. Immediately south of the kale and Okçular Kale İni, the modern village of Okçular is situated on the hillside, and there was almost certainly a village there throughout the Byzantine period, an observation based on the ceramic assemblage and the material remains. While the spolia identified above date to the early period, there are also features around the village, including an undated circular stone well (S81) at the top of a sloping field behind one of the houses in the modern village. Further down, at the bottom of the valley, are two further features which stand relatively close to one another. Neither structure has any ceramics associated with them, and it is difficult to either date or identify them. Okçular 4 (S80; Figs. 11.14) is probably a silo or water feature. A circular structure about four metres in diameter, it was constructed on a
Fig. 11.14: Silo or water feature, Okçular 4 (S80) (CAP-P1/1223).
23 F28, C3026, C3036, C3037, C3038, C3039, C3083 (see Fig. 11.6).
Byzantine Finds and Sites in Cide and Şenpazar
319
rocky outcrop. The field stones were carefully mortared together with a large lip at the entrance, and the whole thing was plastered. This could be Middle Byzantine in date (based on the presence of plaster), although it could also be quite late. Slightly further down the valley from this is a large decayed pit, Okçular 6 (S83). The eastern wall of this pit is poorly preserved, but contains mortared fieldstones (Fig. 11.15). To the north is a large depression with a lot of collapse, which was probably a similar pit (Fig. 11.16). There is less clear evidence for plaster here, but it seems likely that these served as silos or storage bins during the Byzantine period. The proximity of these sites to fields that were likely agricultural in the Byzantine period suggests an agricultural purpose for both structures.
Fig. 11.15: Possible storage pit, Okçular 6 (S83) (CAP-P2/810).
Identifying specific farms or villages is difficult in survey, as the majority of the structures in such settlements would almost certainly have been constructed of mud and stone architecture, and little remains today. Instead, we must rely on ceramics and tile assemblages, particularly those found in conjunction with standing architecture (such as those discussed above). The Okçular assemblage shows the fairly consistent presence of Byzantine occupation throughout the region, although neither tiles nor ceramics appear in any huge quantities. Identification of this corpus is perhaps
320
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.16: Possible storage pit, Okçular 6 (S83) (CAP-P2/848).
the biggest problem. As noted above, there is very little fine ware, although this is perhaps not surprising. What little fine ware existed has long since been picked up, disappearing from the archaeological record. More pertinent to the Byzantine period, however, is the fact that rural communities were unlikely to be huge consumers of what was, essentially, a luxury good until the end of the Middle Byzantine period (Cassis 2009 for the lack of fine ware at the excavated site of Çadır Höyük). What does remain is a sample that is representative of table, kitchen, and storage wares used in rural communities, as well as a large number of tiles. The survey patterns do identify some trends, however. Although there are minor scatterings of identifiably later Byzantine pottery to the west of the kale (C1983 from U504 is a small piece of degraded Sgraffito ware), the majority of the ceramics from this area are Late Roman or Early Byzantine (S516, for example). Given the probable Late Roman occupation near the modern village of Okçular as identified by the presence of the spolia discussed above, this makes sense. Larger amounts of Byzantine ceramics and tiles occur to the northeast, south, and southeast of the kale. In places where more intensive survey took place – U30, U198, U683, and U193, for example – there is more identifiably Byzantine material, suggesting that these concentrations represent the locations of farms. However, even these clusters are relatively modest, indicating small-scale settlements.
Byzantine Finds and Sites in Cide and Şenpazar
321
This type of settlement is most evident in the very southeastern part of the valley surveyed. A number of units provide large scatterings of tiles and Byzantine kitchen wares (U25, U26, and U27, for example; Fig. 11.17). A water feature which, in its current form, probably dates to the Ottoman period (S105) and a small undated bridge or roadway at U2017 indicate later habitation. However, the section cut by a small stream associated with S105 also shows some stone walls (Okçular 8; S131) and tiles. There is no ceramic assemblage associated with the structure, but its location, taken in conjunction with the later features, suggests a farmstead. The identification of this area as a settlement is also encouraged by the presence of many Byzantine roof tiles in the vicinity.24
Fig. 11.17: Map of Okcular tile density. Produced by Toby C. Wilkinson.
11.3.2.2 Çamdibi In many ways, the area of Çamdibi can be seen as the exception to the widespread rural Byzantine occupation of the region. This is a settlement area that exhibits almost no evidence for later Byzantine habitation – or, if there was such habitation, it is largely undetectable. Rather, this seems to be a site that reflects only a Late Roman
24 For example: C4254, C4255, C4256, C4257, C4259, C4259; from U2226.
322
The Cide-Şenpazar Region in the Byzantine Period
and Early Byzantine usage. The isolated location of Çamdibi in the hinterland, away from both major transport routes and the sea, may have rendered it untenable during the Middle and Late Byzantine periods. We can consider three types of evidence for this area: tiles, spolia, and ceramics. To begin with, the major finds in this valley are huge amounts of tile. Tile clusters range from the north all the way to the south of the region, suggesting that there was domestic or agricultural occupation throughout the region. The difficulty lies in pinning down dates for this data set, as tile fragments are difficult to date. However, as noted above, there is a noticeable difference between the tile assemblage at Çamdibi and those of both Okçular and Abdulkadir. The tile assemblage here generally consists of small worn pieces of the cut flanges of tiles, which parallel the stratified deposits at Amorium dated to the Late Roman/Early Byzantine period (Witte-Orr 2007: 301-2). What is lacking in this region are large numbers of the larger flat thin Byzantine tiles noted in other locations (see Abdulkadir). While this is not definitive evidence because of the difficulty involved in dating tiles, the uniform nature of this assemblage suggests a different pattern of occupation than is seen elsewhere in the region, and is exemplified by Çamdibi 8 (S113), which is represented by a tile scatter and some pieces of painted plaster visible in a section, suggesting a Late Roman/Early Byzantine date. These dates are largely confirmed by the spolia and other remains in the area, all of which is of high quality cut stone dating to the Roman or Early Byzantine period. Çamdibi 12 (S122; Fig. 11.18), for example, is a piece of carved stone, possibly a degraded piece of a capital. Çamdibi 3 and 4 (S128, CAP-P1/5024 and S130) are what appear to be a reused sarcophagus with a drainage hole cut in the bottom, and two pillar fragments which must have come from a substantial building. Çamdibi 2 (S126; Fig. 11.19) is a large carved stone basin with a circular decoration on one side and a drain in the bottom of it. Neither date nor original use are clear. There are also several pieces of spolia which are simply building materials, including several large pieces of cut and carved stone from a substantial structure at Çamdibi 1 (S124). Çamdibi 10 Kuyu (S90) is a square well of uncertain date, but is probably Roman or Byzantine and may speak to a continuous agricultural presence in the area. All of these materials (except the well) are carved and cut stone and date to a period of large-scale construction. Finally, Atköprü (S95) is a small bridge that is probably a Late Roman or Early Byzantine construction (for comparanda, see Mango 1976: 126-7; also chapter 10, this volume, for an earlier date). Finally, the ceramic assemblage for the area is very sparse for the Byzantine period. There are no glazed wares from this region, and the majority of the later ceramic assemblage dates to the Hellenistic and Roman period. There are some sherds25 which represent a continuation of Roman forms in a medium fine fabric. An interesting parallel for the area comes from a cave to the south of the settlement,
25 For example, C3349.
Byzantine Finds and Sites in Cide and Şenpazar
Fig. 11.18: Carved stone fragment, Çamdibi 12 (S122) (CAP-P1/4885).
Fig. 11.19: Stone basin, Çamdibi 2 (S126) (CAP-P1/4890).
323
324
The Cide-Şenpazar Region in the Byzantine Period
Kılıçlı Mağarası (S93/S88). The cave contains primarily earlier material (chapters 5 to 8), but there is a little Late Roman or Early Byzantine coarse ware. However, there are also two sherds of Early or Middle Byzantine White Ware (probably eighth century).26 These are the only sherds of early glazed ware found in the entire survey, and their presence in relation to Çamdibi is particularly interesting, given the earlier dates discussed above, and may represent the last real settlement in the area.
11.3.2.3 Abdulkadir The extensive settlement around the modern village of Abdulkadir is characterised by scatters of Byzantine pottery, with a couple of more intensive clusters. There are no architectural remains in the region, and the assessment of this area as Byzantine largely rests on the tile and ceramic assemblage, as well as the position of this area in relation to other known sites. As at Okçular, the ceramic assemblage suggests continuity of usage throughout the Byzantine period. To begin with, although there are scattered tiles and sherds throughout the area, they do cluster together at certain points, suggesting that these were domestic units or villages. The difficulty lies in the fact that there is only one piece of clearly dateable Byzantine pottery27, which is a small piece of Middle Byzantine Sgraffito ware. Everything else from the Byzantine period is kitchen or table wares which cannot be dated more closely. Neverthless, there are clusters of both tiles and ceramics which suggest habitation. These are located in two general areas (S504 (Abdulkadir 2); S505 (Abdulkadir 3), and are represented by tiles and ceramics in units U263, U274, U285, U283, U291, U390, and U455, and by ceramics in unit U469. Unlike the tile and ceramic assemblage at Çamdibi, the remains from Abdulkadir reflect the long-term settlement of the region. The tile assemblage at Abdulkadir is represented by both small pieces of what are probably Roman tiles, as well as large flat Byzantine tiles. The ceramic assemblage consists primarily of table and kitchen wares of the Early and Middle Byzantine periods. S504 (Abdulkadir cluster 2) is surrounded by several survey units containing ceramics that can be identified as Byzantine based on form and fabric.28 The only piece of glazed fine ware comes from U459, which is immediately to the south of this cluster. Immediately to the south and east of this, in U469, is a large cluster of Early Byzantine fragments. Finally, Abdulkadir 3 (S63) is a site to the south of this entire settlement, and lies immediately below the large cave of Ballıcı İni (S17). There are a number of Early and Middle Byzantine fragments here, which speaks to the wider issue of interaction between sites. The inhabitants of the settlements at Abdulkadir would have been able
26 C2032, C3243 (see Fig. 11.4). 27 C2180; from U459. 28 see particularly C1602.
Byzantine Finds and Sites in Cide and Şenpazar
325
to interact with a number of significant defensive or religious sites in the near vicinity (discussed below), including the fortress at Gideros (S12, S121) and the probable monastic complex at Çadır İni Kilise (S21).
11.3.2.4 Other Settlements There are other locations which, based on the ceramic assemblage, can also be considered as likely settlements. For example, there is a small cluster of sites at Mencekli (S4, S5, and S7) which contains some Byzantine ceramic material and which, according to local stories, sits on a Byzantine settlement. Sakallı Mezarlığı (S9) is an upland site off the coast. Little was observed except a cluster of uncollected pottery which includes some Early Byzantine table ware. The location of the site has parallels with Okçular. Kuşçu 1 (S10) is a site identified because of illegal excavations. Among the finds were ceramics of the Byzantine period, a large number of tiles, and at least one human bone. The ceramics include several pieces of Byzantine kitchen and table wares. The visibility of the valley, the river, and the sea has parallels with the site at Okçular and suggests another upland settlement. Kumluca 1 (S109) is a site situated near the coast. It has a concentration of stone, mortar, tile, and pottery which measures roughly 20 x 10m, with scatters in the surrounding fields. The ceramics seem to date to the Roman and Byzantine periods, although there is no fine ware. Finally, İlyasbey 1 Su Değirmeni (S186) is a site which contains two water features of Ottoman date. However, the foundations of the site look like they may be Byzantine.
11.3.3 Byzantine Fortifications Understanding the fortifications covered by CAP (Fig. 11.20) is difficult as there is no established or completely foolproof typology for these types of constructions. Although such stone fortifications are commonly accepted as Byzantine, the justification for such an assessment is often unclear, as noted above. Scholars have tried to establish a chronology, most notably Bryer and Winfield (1985), Foss and Winfield (1986), and Lawrence (1983), but there are numerous problems with these absolute chronologies. Fortifications located in rural or isolated areas do not correspond to the much larger structures in urban environments which are often used as the models for establishing the chronology. Architectural techniques varied from region to region based on the availability of resources. Nevertheless, a broad framework, taken in conjunction with historical knowledge of the area, can be established for understanding these fortifications. Late Roman structures hold on to traditions established by the Romans, who tended to build well-faced structures surrounding a cemented core (Ousterhout 1999: 133-6). By the Late Roman period, walls usually consisted of a core, heavily bonded with mortar, faced with courses of brick, tile, or cut stone. In ideal circumstances, five
326
The Cide-Şenpazar Region in the Byzantine Period
0
5
10 Km
Çoban Kalesi Timle 1. & 2. Kaleleri
Kazallı Kale
Gideros Batı & Doğu Kaleleri Hıdır Kalesi
Aydos Kalesi
Okçular Kalesi
Koca Kale Fortified site
Akça Kale
Tile distribution
Fig. 11.20: Map of Byzantine fortifications. Produced by Toby C. Wilkinson.
courses of tiles would be separated by four or five courses of stone (Mango 1976: 11-4; Foss and Winfield 1986: 28). Tile is one of the features which differentiates both Late Roman and late Middle Byzantine construction, since it seems to be largely absent from the interim Early Byzantine period. By the beginning of the Early Byzantine period, according to Foss and Winfield (1986: 28) point out, there was a substantial lack of tile throughout the empire as production was disrupted by the Arab incursions. While our survey discovered no architecture adhering to that model, there are several examples of fortresses which contain single or double rows of tile (for example, Kazallı Kalesi, S91; Figs. 11.21 and 11.22) or which have well-constructed, well-faced mortared walls (for example, the two kales at Timle, S110 and S112; Fig. 11.23). Nevertheless, none of the fortresses in the survey area were particularly well-endowed with tile, suggesting that this cannot be a definitive characteristic for Late Roman or Middle Byzantine in the region. Rather, strong, well-constructed fortifications of mortared, faced stones suggest imperial construction in the Late Roman period. A second consideration in dating these fortifications is their purpose. In all of the examples which can probably be dated to the Late Roman period, the structures are located on roads and the coastal trade routes, suggesting an imperial defensive function. As noted at the beginning of this chapter, there was extensive defensive building along this coastline throughout the Late Roman period, and these fortifications on the coast may go back to the fortresses built by Justinian in the
Byzantine Finds and Sites in Cide and Şenpazar
327
Fig. 11.21: Interior walls of rectangular structure at Kazallı Kalesi (S91). Rows of mortared tile and stone. Possible Late Roman construction (CAP-P1/1507).
Fig. 11.22: Exterior of outer wall at Kazallı Kalesi (S91). Several courses of stone interspersed with single rows of tile. Possible Middle Byzantine construction (CAP-P1/1511).
328
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.23: External wall at Timle 1 (S110). Well-faced mortared stones. Possible Late Roman construction (CAP-P1/2673).
early sixth century (Mango 1976: 97), which would explain the fortifications at Timle, Aydos, and Kazallı Kalesi (see below). The purpose of such structures in this period was to defend the coastline and to secure the trade and food supply. At some point, probably during the Early Byzantine period, there was a major shift in the way fortifications were constructed or rebuilt. This is particularly well articulated in the major fortification at Gideros (S12; Figs. 11.24), which has lower levels with alternating tile and stone layers (which sit on even earlier foundations), but which has loosely mortared field stones at the top of the walls, a technique generally dated to the seventh century (Foss ad Winfield 1986: 26-7), although it almost certainly carried through to the early part of the Middle Byzantine period. These structures all show substantial use of cribwork, or a wooden framework, which Foss and Winfield (1986: 28) connect to the loss of tile as a building material. This type of construction is also noted regularly in rural sites, such as Okҫular Kale (S22; see above), which were often used for storage and local protection. Nevertheless, the presence of mixed architectural types reflects the fact that many of these fortifications, most notably that at Gideros, were used throughout the Medieval period.
Byzantine Finds and Sites in Cide and Şenpazar
329
Fig. 11.24: Upper levels at Gideros (S12). Loosely mortared fieldstones. Possible Early Byzantine rebuilding (CAP-P2/1578).
The late eighth and early ninth century marks a return to large-scale, well-constructed architecture, often including tile once again (Foss and Winfield 1986: 161-3). However, this is really only notable in the major fortifications of urban centres as the Byzantine empire worked to reconsolidate its position. By the 11th century, such construction had become more common. According to Foss (1982), the period of the Komnenoi was particularly distinctive, resulting in several major projects in many areas, including the Black Sea. However, there was still no uniformity in fortifications. When possible, tile was included in the construction, but there are several examples cited by Foss in which tile remains absent (1982: 154-5). What is clear, however, is that there is a return to well-faced, mortared architecture. Four structures in our region suggest this: the fortification at Çoban Kalesi (S32; chapter 12); Aydos Kale (S190; Fig. 11.25); Koca Kale (S154); and possibly some rebuilding at Kazallı Kalesi (S91; Fig. 11.22 and 11.26), but there is little else in the material culture to confirm this revival. As in the Late Roman Period, revival of these sites was tied to an imperial mandate. Thus, during the Komnenian revival, reconstruction would almost certainly have occurred at these sites deemed important to maintaining the sea routes, notably Gideros, Kazallı Kalesi, and perhaps Çoban Kalesi, which, given its proximity to Timle, may have replaced the Timle kales. The catalogue is arranged as follows. The first kales discussed are those along the coast, and reflect the Late Roman, and later Byzantine periods in which the coastline was of concern to the population. The fortresses located in the hinterland are discussed after this, and reflect the concern for defence and storage of the
330
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.25: External wall at Aydos Kale (S190). Stone courses interspersed with tile courses. Possible Middle Byzantine construction (CAP-P15/313).
Fig. 11.26: View of later walls at Kazallı Kalesi (S91). Probable Middle Byzantine construction (CAP-P1/1503).
Byzantine Finds and Sites in Cide and Şenpazar
331
agricultural communities or internal roads located away from the coastline. In almost all cases, however, these fortresses reflect either continued or reoccurring occupation throughout the Byzantine period.
11.3.3.1 Gideros Batı Kalesi (S12) The site of Gideros (Kyteros) (Fig. 11.24) has been known since antiquity, and there are a number of sites located in the vicinity of this natural harbour. It is, in fact, one of the few sites that can be identified from ancient sources. The name became visible again during the Late Roman and later Medieval periods, in portolans and in the accounts of both the empire of Trebizond and the kingdom of Nicaea (Belke 1996: 245-6). The site was relatively recently surveyed by Belke, who identified it as a significant fortress throughout the Medieval period, and also isolated the cisterns and later elements that are still visible. While it is impossible to date the structure closely, there do appear to be at least two phases of Roman/Byzantine construction. The lower levels contain very regular, well-mortared stone and tile walls, which is suggestive of an early date (Roman or Late Roman) (Adam 1994: 76-9). Above this are levels of roughly mortared fieldstones which probably date to the Early or Middle Byzantine period. The site remained in use throughout the Byzantine period. A number of ceramics from this site reflect both the Middle and Late Byzantine period, indicating its significance throughout this period. The site, consisting of the kale and two heavily surveyed areas (U1, U2) produced the largest collection of glazed sherds from the CAP survey, all of which date from the 12th-14th centuries.29 While it is likely the structure was used in the Early Byzantine period as well, the presence of fine ware in the quantity that appears here indicates clearly that this was a site that was refurbished and reused in the Middle Byzantine revival, and subsequently by the various empires which controlled it in the Late Byzantine period.
11.3.3.2 Gideros Doğu Kalesi (S121) Situated on the opposite promontory from the large fortification, this site (Fig. 11.27) consists of a number of disconnected artefacts from various periods, including an upright cannon and an 18th century gravestone. The two rounded towers were noted by Belke (Belke 1996: 245-6), who considered this to be part of a larger fortification complex at Gideros. The foundational construction of these towers is, however, notably different from that of the larger kale opposite. Constructed mainly of loosely stacked fieldstones mortared together, there is little here that suggests the organised construction of the earlier period. Rather, it would seem likely that these towers were put in place during the Early Byzantine period or into the Middle Byzantine
29 C2, C15, C16, C18, C34, C41, C46, C66, C84, C85.
332
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.27: Gideros Doğu Kalesi (S121). Loosely mortared fieldstones. Probable Early Byzantine construction (CAP-P2/1850).
period, based on the loose construction. While there is tile, it is not organised in any systematic way but is rather simply part of the loose mortared matrix. The ceramics were limited to a little kitchen ware, but the fabrics are Byzantine.30
11.3.3.3 Kazallı Kalesi (S91) This is one of the clearest examples of a defensive fortification in this part of the theme of Paphlagonia (Fig.11.22). The structure sits on an outcrop of rock right on the coastline and has an excellent view of the sea. The structure as it stands seems to be the product
30 C4106, C4107.
Byzantine Finds and Sites in Cide and Şenpazar
333
of at least three phases of construction, and Belke connects it with the site of Klimax, which is known from both ancient sources and later Medieval ones (Belke 1996: 237). The kale consists of two parts, a small rectangular internal structure within an outer semicircular wall. In addition, there were also walls which probably circled the entire plateau originally. The lower courses of the small rectangular structure may be Late Roman based on the construction, which includes some tile (Fig. 11.21). There does appear to be some rebuilding in the middle period, as there are some places which are simply mortared stones, which suggests either the Early Byzantine or the start of the Middle Byzantine period. The top levels of the structure are a bit different, however, and suggest Middle or Late Byzantine reconstruction based on the presence of irregular courses of stone and tile. This structure has been variously explained as a tower or a cistern and, based on parallels with Koca Kale, another possibility is that this may actually have been a beacon tower, if we date the rebuilding to the early part of the Middle Byzantine period (Kazhdan 1991: sv beacon). The external walls of the kale are a bit different, however, and suggest Middle or Late Byzantine reconstruction based on the presence of irregular courses of stone and tile (Figs.11.22 and 11.26). The outer wall has no tiles, and looks like later construction, possibly the beginning of the Middle Byzantine period (Fig. 11.28), although the date is uncertain. The ceramics sample is small, but reflects usage of the site in the later Byzantine period.31
11.3.3.4 Timle Kale 1 (S110) One of the challenges in understanding the defensive structures along this coastline is the fact that they all look slightly different (Fig. 11.29). Timle Kale 1 is a roughly triangular structure built to utilise the shape of the hill it sits on. Two of the postulated three towers are visible, and a small round cistern is visible at the base of the eastern tower. The construction of this kale is made of a rubble core with cemented or mortared well-faced stones. Beam holes and/or drainage holes are visible near the base of the walls (Fig. 11.30). The lack of tile suggests that it is neither Roman nor Late Byzantine, but the fact that it is so regularly constructed makes it difficult to date. There is little pottery here to help date this structure, and it ranges from the Hellenistic through to the Byzantine period. What little that is here is early, however, and this, combined with the well-faced stones, suggests that this had a Late Roman foundation, but this is highly speculative. The proximity of the Çoban Kalesi also suggests that this was an earlier structure that fell into disuse, since the latter structure was clearly used more prominently in the Medieval period.
31 C2785, C2786, C2787.
334
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.28: Outer walls at Kazallı Kalesi (S91). Possible Early or Middle Byzantine construction (CAP-P2/992).
Fig. 11.29: Plan of Timle Kale 1 (S110) and Timle Kale 2 (S112). Produced by Michele Massa and Victor Klinkenberg.
Byzantine Finds and Sites in Cide and Şenpazar
335
Fig. 11.30: Internal walls at Timle Kale 1 (S110). Well-faced mortared walls (CAP-P1/2679).
11.3.3.5 Timle Kale 2 (S112) This kale is located inland from the first one, and there can be little doubt that their origins were connected, based on the type of masonry (Fig. 11.31). This kale is larger than the first, with two gates (Fig. 11.29). This creates two enclosures, and the second of these includes a smaller square structure, possibly a cistern or a beacon tower. Like Timle Kale 1, the walls are created with a rubble core surrounded by mortared, wellfaced stones. There appears to have been some later rebuilding, as the tops of some of the walls are much more haphazard collections of field stones (Fig. 11.32). There were no ceramics here.
11.3.3.6 Aydos Kale (S190) This is perhaps the most difficult of the kales to explain (Fig. 11.33). The location of Aydos Kale is largely isolated from any other identifiable settlements in the Byzantine period, and yet it is a structure of particularly fine construction. It is a small fortification which consists of several round bastions approximately every eight metres (and each is about four metres across). Probably because of the isolation of this structure, it has been preserved intact up to about seven metres in places. The overall construction consists of a well-bonded internal core with well-faced stones and tiles on the outside of the structure (Fig. 11.34), and some herringbone patterning inside (Fig. 11.35). Beam holes are visible throughout the structure. At first glance, this would seem like a Late Roman structure. However, three elements suggest an 11th century (or later) date. First,
336
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.31: Lower external wall at Timle Kale 2. Well-faced stone construction (S112) (CAP-P1/2693).
Fig. 11.32: Later rebuilding at Timle Kale 2 (S112) (CAP-P1/2691).
Byzantine Finds and Sites in Cide and Şenpazar
337
Fig. 11.33: Plan of Aydos Kale (S190). Produced by Victor Klinkenberg.
Fig. 11.34: External wall at Aydos Kale. Stone with sparse tile layers. Probable Middle Byzantine construction (S190) (CAP-P15/319).
338
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.35: Interior Herringbone patterning at Aydos Kale (S190). Probable Middle Byzantine construction (CAP-P15/344).
the interspersing of tile and stone on the outside is a technique known from the later period (Foss and Winfield 1986: 163-7), as is the herringbone design inside. Secondly, although CAP discovered little pottery, the structure was surveyed by Belke (1996: 175) who found some glazed ware. Given that it looks very much like it was all constructed in one period (unusually for this region), it seems likely that the ceramics date to the period of its construction. Finally, the location of this structure suggests an attempt to protect the coast from the Turkic tribes, an idea discussed by Booth, who posits that there was a loose grouping of inland fortresses in the region in the Late Byzantine period (Booth 2004: 48). It is possible that this was a small structure placed along a Byzantine roadway.
11.3.3.7 Hıdır Kalesi (S62) Located away from the coastline, slightly inland from Cide, this kale is very similar in construction and form to Aydos Kale, although much larger. It is a well-built, elliptical fortress which consists of rounded bastions built with stone and tile courses (Figs. 11.36-7), and includes a number of beam holes, probably for cribwork. As noted above, such construction suggests either a Late Roman, or a date in the 11th century. The extant pottery from this site suggests the end of the Middle Byzantine or a Late
Byzantine Finds and Sites in Cide and Şenpazar
Fig. 11.36: External wall at Hıdır Kalesi (S62). Stone and tile courses. Probable Middle or Late Byzantine construction (CAP-P1/409).
Fig. 11.37: Detail of construction techniques at Hıdır Kalesi (S62) (CAP-P1/407).
339
340
The Cide-Şenpazar Region in the Byzantine Period
Byzantine usage.32 When considered in relation to Aydos Kale, it seems likely that there was a small network of inland fortresses in this period meant to keep the area safe from the invading Turkic armies.
11.3.3.8 Akça Kale (S135) Akça Kale is probably a similar structure to that at Okçular (Fig. 11.38). This is a substantial roughly elliptical structure which is quite difficult to reconstruct because it is so overgrown. Nevertheless, it seems to be largely constructed of roughly mortared fieldstones. It has a small apsidal wall in it as well (Fig. 11.39).
Fig. 11.38: Plan of Akça Kale (S135). Produced by Michele Massa.
11.3.3.9 Koca Kale (S154) Koca Kale is, by far, the most unusual of the kales explored. Located on a rocky outcrop, it is an oval fortress overlooking the Devrekani river to the north. The construction includes a rectangular structure (approximately 3 x 4m), which has a staircase and the remains of a second floor located within it. Cribwork beam slots are visible where the staircase meets the wall. The purpose of this structure is uncertain. Although a cistern has been suggested, the location of this tower and its second floor suggests that this may have been a beacon tower (Kazdhan 1991: sv beacon). The construction of both this internal wall, which is between 60-75cm thick, and the huge outer walls, which are about 120cm. thick, is also problematic to
32 C714, C715 (see Fig. 11.4).
Byzantine Finds and Sites in Cide and Şenpazar
341
Fig. 11.39: Internal apsidal wall at Akça Kale (S135) (CAP-P15/390).
parse (Fig. 11.40). Both the internal and external structure are constructed of large cut limestone blocks, which are largely undressed but covered with a thick reddish cement-like mortar which functions to face the walls (Fig. 11.41). There is mortar between these blocks, and it is a soft powdery mortar. There are two possibilities for dating this structure. The large blocks suggest a Roman construction, although the somewhat haphazard setting of some of the courses of stones does not support this. A more likely scenario – given the proximity of this structure to Çamdibi, is that this fortress was rebuilt at the end of the early Byzantine period or the beginning of the Middle Byzantine period using the remains of an earlier structure. The mortar on the walls is attested in other types of structures (Ousterhout 1999: 133-6) as a poor variation on earlier Roman cement. The second floor to the tower is also suggestive of a beacon tower, leading to the question of whether this was a construction put into place to attempt to preserve the settlement at Çamdibi, the material remains of which date no later than the early part of the Middle Byzantine period. Unfortunately, the ceramic assemblage from this structure provides nothing which gives a clear date.
342
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.40: External wall at Koca Kale (S154) (CAP-P1/4978).
Fig. 11.41: Detail of mortar at Koca Kale (S154) (CAP-P1/4985).
Byzantine Finds and Sites in Cide and Şenpazar
343
11.3.4 Religious Spaces and Hermit Caves As noted above, there is a long history of Christianity in this region, with an identifiable Christian community being named in the letters of Pliny the Younger. The archaeological visibility of this community would not be clear until after the widespread acceptance of Christianity in the early fourth century, when large basilical churches would have been built in every community (Krautheimer 1986 for comparanda). In locations where sites were largely abandoned after the Late Roman period, particularly in the south and southwestern parts of Turkey, such structures often remain extant (see, for example, the churches at Alahan; Gough 1985). In places where Byzantine life continued roughly unbroken, such structures were often abandoned, robbed out, or rebuilt, which makes it difficult to identify and date these buildings. The region covered by CAP contains the sparse remains of two basilicas, as well as a number of potential hermit caves and religious spolia that speak to the widespread presence of Christianity in the Byzantine period (Fig. 11.42). As the pottery sequence illustrates, the majority of the Roman and Late Roman population was almost certainly clustered along the coastline. Thus, it is not surprising to find at least one basilica in this region. At Kalafat Mezarlık (S116; Figs. 11.43) the remains of a basilica with several standing pillars has been incorporated into an Islamic graveyard. Although at least one of the pillars has been moved, several
Fig. 11.42: Map of Byzantine religious sites. Produced by Toby C. Wilkinson.
344
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.43: In situ column at Kalafat Mezarlık (S116) (CAP-P2/1287).
were found in situ (although disturbed by nearby robber pits), and it is possible to identify the probable nave of a seaside church. There is parallel evidence for such seaside churches at Anemurium in Cilicia (Campbell 1998: 23) and the sea would have been further out in antiquity. Nevertheless, given the possible size of the church and its location along the coastline, this church almost certainly dates to the Early Byzantine period. Unfortunately, no ceramics were associated with the church. A potential parallel to this structure was found at Gökçekale (S174; Fig. 11.44), which is an apsed building oriented east sitting on the edge of the sea. All that remains is the apse which is constructed of stone and packed mud. There is no clear date for the structure, and there were no ceramics associated with it. An inland parallel at Çayyakka (S79) consists of two Byzantine columns in a later cemetery. The other extant church is inland from the harbour at Gideros (Kyteros), and is known as Çadır İni Kilise (S21; Fig. 11.45). This structure has suffered from a great deal of illicit digging over the years, and is in imminent danger of disappearing. It is a particularly important structure, however, as it exhibits evidence for the overall transition of religious structures during the Byzantine period and because it is clearly connected to a number of other sites in the vicinity, including the settlement at
Byzantine Finds and Sites in Cide and Şenpazar
Fig. 11.44: Apse at Gökçekale (S174) (CAP-P15/276).
Fig. 11.45: Plan of the church complex of Çadır İni Kilise (S21) and Çadır İni (cave below, S20). Produced by Michele Massa and Victor Klinkenberg.
345
346
The Cide-Şenpazar Region in the Byzantine Period
Abdulkadir, the harbour at Gideros, and several caves in the region. Survey of the immediate region indicates that there were a number of other buildings associated with this structure, which suggests that this was not simply a basilica, but a monastic community. The earliest church was a substantial tripartite basilica with a (possible) trefoil apse and no pastophoria (side chambers). Little work has been done on the basilicas of the Black Sea coast, but the format of the church is in keeping with other structures from the Late Roman period in Turkey (Krautheimer 1986). One of the aisles, the apse (Fig. 11.46) and some of the chancel screens (Fig. 11.47) are still standing, illustrating that at least the sanctuary of this church remained in use throughout the history of the structure. At some point, however, the structure was cut down and shortened to focus on only the sanctuary area. A later grave, covered by later Byzantine tiles, was added in the west part of what would have been the original nave of the church (Fig. 11.48), a tendency found in other reused churches of the Middle Byzantine period (such as Amorium, Lightfoot 2009: 144). That it remained an important structure, however, is suggested by the presence of patches of painted plaster visible in churned up and disturbed soil near the bottom of the walls of the church. There were also architectural fragments scattered around the structure (Figs. 11.49 and 11.50).
Fig. 11.46: Apse of the Çadır İni Kilise (S21) (CAP-P2/62).
Byzantine Finds and Sites in Cide and Şenpazar
Fig. 11.47: In situ chancel screen at Çadır İni Kilise (S21) (CAP-P2/72).
Fig. 11.48: Grave in the nave of the Çadır İni Kilise (S21) (CAP-P2/1311).
347
348
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.49: Architectural fragment at Çadır İni Kilise (S21) (CAP-P2/1305).
Fig. 11.50: Architectural fragments at Çadır İni Kilise (S21) (CAP-P2/1308).
Byzantine Finds and Sites in Cide and Şenpazar
349
The identification of this site with a monastic centre is encouraged by outlying buildings and architectural remains which could be seen, if not fully articulated, in survey, as well as a remodelled natural cave located directly below the church. Çadır İni (S20; Fig. 11.51) was a substantial cave used for much longer than the Byzantine period, although at some point during its use in relation to the church, carved ledges, benches, and wall sockets were added (Fig. 11.52). Its proximity to the church, which is on top of it, suggests some connection between the two. The ceramics were nondescript, once again, but one tiny sherd of glazed ware33 speaks to the use of the cave into the Late Byzantine period. The use of caves as centres of religious devotion is known from several other cave sites in the region. Found in close proximity to the religious complex discussed above, Gideros Mağarası (S59; Fig. 11.53) is a large cave complex. At some point it was turned into a commemorative shrine, which is noted by an apsidal enclave at the back of the cave surrounded by crosses scratched into the wall (Figs. 11.54-55). There are also some Greek inscriptions which need further analysis, as they will almost certainly offer some indication of the purpose and dedication of this site (Fig. 11.56). Right now, however, it seems likely that this was a structure dedicated to a hermit saint who had originally lived within the cave in proximity to the monastery discussed above. Unfortunately, illegal excavation has disturbed the deposits and the photos do not provide enough detail to date the cave or identify its occupant. The only Byzantine ceramics are kitchen wares.34 Other examples of caves with this kind of iconography include Uçakdibi Mağarası (S40; Fig. 11.57), which is a double cave which contains several drawn and carved crosses. Some illegal digging has also occurred in the cave. No glazed ware was found in the cave, although there were two examples of Byzantine kitchen wares.35 There are several other examples of inhabited caves, although none of them possess anything in the way of discernibly Christian artefacts or decoration. While it seems likely that some of them were used as hermit caves, it is also entirely possible that they were used as safe storage for the local population. Ballıcı İni (S17; Figs. 11.58 and 11.59), for example, is located in a cliff face south of Abdulkadir, and contains evidence of occupation for several periods, including the Byzantine. Although only a small proportion of the pottery found in the cave is clearly of the Byzantine period, it is in very good condition and provides examples of kitchen ware, probably from the Middle Byzantine period.36 There was also a small nail, probably Byzantine in date.37 Likewise, the cave of Kılıçlı Mağarası (S88/S93; Figs. 6.11, 6.12 and 14.26) is located in
33 C207. 34 C640, C642, C785. 35 C368, C371. 36 C1722 (see Fig. 11.6). 37 F23.
350
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.51: Çadır İni (S20) (CAP-P2/79).
Fig. 11.52: Carved benches in Çadır İni (S20) (CAP-P2/84).
Byzantine Finds and Sites in Cide and Şenpazar
351
Fig. 11.53: Plan of Gideros Mağarası I (S59). Produced by Michele Massa and Victor Klinkenberg.
Fig. 11.54: Crosses in Gideros Mağarası I (S59) (CAP-P1/358).
352
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.55: Crosses in Gideros Mağarası I (S59) (CAP-P1/359).
Fig. 11.56: Crosses and inscriptions in Gideros Mağarası I (S59) (CAP-P1/365).
Byzantine Finds and Sites in Cide and Şenpazar
353
Fig. 11.57: Plan of Uçakdibi Mağarası (S40). Produced by Michele Massa and Victor Klinkenberg.
the southern part of the Çamdibi area. The cave contained a small amount of Early Byzantine pottery - including three sherds of Glazed White Ware38 and some kitchen ware39 which indicate some usage during the Early and Middle Byzantine period.40 Finally, Yedikapılı Mağarası (S192) had a minor amount of probably Early Byzantine pottery located near the entrance to the cave.
38 C2032, C3243, C3738. 39 C2946; C2947; C2428; C3178. 40 See Figures 11.4 and 11.6
354
The Cide-Şenpazar Region in the Byzantine Period
Fig. 11.58: Plan of Ballıcı İni (S17). Produced by Michele Massa and Victor Klinkenberg.
Byzantine Finds and Sites in Cide and Şenpazar
355
Fig. 11.59: Ballıcı İni (S17) (CAP-P2/45).
11.3.5 Spolia and Other Remains There is a large amount of spolia found throughout the region, primarily along the coastline. Most of this probably dates to the Roman period, given that there was far more large-scale construction in that period (chapter 10). However, there are several pieces which may have come from Late Roman or Early Byzantine churches. These are included here to illustrate how widespread the occupation was in the region (table 11.1), with a sampling of illustrations.
356
The Cide-Şenpazar Region in the Byzantine Period
Tab. 11.1: Spolia from the Cide Archaeological Project. site
site ID
architectural fragment
Cide 1
S1
column
Cide 5 (Miltaşı)
S2
column
Cide 9
S3
architectural fragment
Çayyakka 1
S79
2 columns
Kuşçu 2 [Fig. 11.60]
S89
8 fragments: chancel screen piece; columns; worked stone
İlyasbey Mezarlık
S101
reused columns as graves
Gideros 2 [Fig. 11.61]
S123
rectangular basin; chancel screen fragment; column fragment; worked stone.
Cide 7
S127
dressed stone
Ovacık 1
S133
column base rebuilt into house
Irmak 4
S144
column piece
Irmak 7
S150
dressed block; stone plate; ionic column capital
other
Kumluca Aşağı Mezarlık S166
Ottoman cemetery with column pieces
Yenice 1
S168
dressed stones (Roman?)
Cide 2
S196
dressed stones (Roman?)
Harmangeriş 1 and 2
S94/S96)
pithos and column fragment
Irmak 1
S138
stone basin
Irmak 2
S140
pounding stone
Irmak 6
S148
pounding stone or pillar post
Irmak 8
S152
Hamitli 2
S156
carved stone basin
Kalafat 1
S166
table top/other spolia
Gökçekale 6
S176
grape press
Gökçekale 7
S178
dressed stone
dressed stones
Byzantine Finds and Sites in Cide and Şenpazar
Fig. 11.60: Spolia (possible chancel screen) from Kuşçu 2 (S89) (CAP-P2/936).
Fig. 11.61: Spolia (basin) from Gideros 2 (S123) (CAP-P2/1855).
357
358
The Cide-Şenpazar Region in the Byzantine Period
11.4 Conclusions While the Black Sea retained significance throughout the Byzantine world, primarily as a transportation and trade route, little is said in the sources about the populations on its southern shore. Indeed, even the names of the villages have largely disappeared from the Byzantine sources. Bereft of urban centres, this section of the Black Sea coast lay largely isolated from the rest of the Byzantine world. Because this area lay largely protected from the trials and tribulations occurring on the other side of the Pontic mountains, the area encompassed by CAP provides a unique example of both an independent regional community in the Byzantine world, as well as an example of continuity. Within this continuity, undisturbed by major invasions and upsets, we are able to gain a glimpse of how the overall changes within the Byzantine empire affected its more isolated outposts. Thus, this region becomes a mirror for understanding the changes within the wider empire. More importantly, however, it raises questions about regional differences within the empire. Byzantium cannot be seen as a uniform entity. The Late Roman period reflected the shift from the imperial Roman settlements of villas and large scale urban centres to a more Medieval village economy that has now been recognised as largely indicative of the Early Byzantine period. In the Late Roman period there was a last attempt to consolidate and hold the coastline as a means of controlling trade and resources, and this is identifiable in some of the fortresses, such as the two at Timle. It is also identifiable in the large amount of Late Roman and early Byzantine pottery along the coastline. Even as the Late Roman period began to shift into the Early Byzantine period, and trade became less viable, there was some attempt to maintain this way of life. Local ceramics mimic earlier Roman fine wares, and large ecclesiastical structures, such as the church at Kalafat Mezarlık continue in use. As the imperial influence of the Byzantine empire shrunk in the seventh century, local economies retreated into themselves. This shift, previously seen as the complete disappearance of the Roman world, has more recently come to be seen as a contraction and ruralisation of the Roman world. The Cide Archaeological Project illustrates this clearly in the continued settlements at Okçular, Çamdibi, and Abdulkadir. This is particularly evident at Okçular, which continued to exist as rural settlements in the shadow of a large kale which looked inward over the valley, providing safety for that community as opposed to defence of the sea. The ceramic assemblage becomes largely restricted to unglazed table and kitchen wares, and we are left with what was, in essence, a peasant economy in the archaeological record. As the Byzantine world began to come out of crisis in the late eighth and ninth century, there was a return to prosperity in the region. The settlements were not abandoned, for the most part, even as more prosperity and trade moved back into the region. However, as trade increased again, the coastline had to be defended. This came to a climax particularly in the late 11th and 12th centuries under the Komnenian dynasty, which had roots in the area. Indeed, the majority of both coastal and hinterland
Catalogue of Illustrated Artefacts
359
fortresses show rebuilding in this period as the Byzantine empire attempted to hold what it had from the encroaching Turks. This was short-lived, of course, as the Turks began to move into the region. Nevertheless, this speaks to the growing presence of wealth in the region. A careful reading of the archaeological evidence visible along this portion of the southern Black Sea coast provides a far more nuanced view of the way the Byzantine empire changed and adapted over the course of, particularly, the Early and Middle Byzantine periods. The region remained significant in terms of agriculture and trade, and provides evidence for both the small-scale lives of Byzantine peasants and the grander narrative of empire. Further, it opens new questions about how significant localised areas were within the empire, and allows us to consider the variety which existed throughout this period and in this geographical region. Acknowledgements: I would like to thank Bleda Düring, Claudia Glatz, and Tevfik Emre Şerifoğlu for involving me in this project. I would also like to thank my former student Aurora Camaño for her hard work during the summer of 2011.
Catalogue of Illustrated Artefacts Glazed Table Wares C184. U4 (Okçular Field 2). Base and lower wall of bowl. Medium fine orange-light red powdery fabric (10R6/8) with mineral and small red grit inclusions. Olive yellow glaze (2.5 Y 6/6) with incised decoration. Wheelmade. Sgraffito ware (subtype?). Parallels: Vroom 2005 LBYZ 4. Date= 13-14 century. C246. U5 (Okçular Mound 1). Base and lower wall of small bowl. Medium fine light red-orange fabric (2.5 YR 7/6) with fine black mineral and organic inclusions. Olive yellow glaze (2.5 Y 6/8) with incised star. Wheelmade. Sgraffito ware. Parallels: Vroom 2005 LBYZ 4. Date= 13-14 century. C650/651. U201 (Okçular Grid 20). Rim and inner wall of bowl. Fine orange-light red fabric (5 YR 7/8) with fine black and red grit inclusions. Yellow glaze (5.5 Y 7/8) with incised lines. Wheelmade. Sgraffito Ware. Parallels: Vroom 2005 LBYZ 4. Date= 13-14 century. C715. S62. Base and walls of small bowl. Fine orange-red fabric (5 YR 7/6) with very fine calcareous and quartz (?) inclusions. Brown yellow glaze (10 YR 6/6) with incised circles in base. Wheelmade. Zeuxippus ware (subtype?). Parallels: Vroom 2005 LBYZ 1 or 2. Date= 13-14 century. C2032. S93. Rim of chafing dish (?). Very fine pink-buff fabric (5 YR 7/6-7.5 YR 8/4) with very fine white (limestone?) and black grit inclusions. Green(2.5 Y 4/3)/Yellow (10 YR 6/8) glaze. Wheelmade. Glazed White Ware II-V. Parallels: Vroom 2005 MBYZ 4. Date= 8-11 century. C3076/3077. S24. Body sherds of a bulbous jug. Fine orange fabric (2.5 YR 6/6). Brown-red glaze (2.5 Y 3/3) on body with yellow spirals (2.5 Y 7/8). Wheelmade. Perforation on C3076. Slip-painted ware. Parallels. Vroom 2005: LBYZ 9. Date= 13-14 century. C3322. S2.Body fragment of small bowl. Fine light red fabric (2.5 YR 6/6) with sparse mineral inclusions. Pale yellow glaze (2.5 Y 8/4) with incised decoration. Wheelmade. Incised Sgraffito ware. Parallels: Vroom 2005 MBYZ 11; Dark 2001:132. Date= 12-13 century.
360
The Cide-Şenpazar Region in the Byzantine Period
C3243. S88. Handle join of a small jug. Fine pale yellow fabric with very sparse black mineral inclusions. Interior surface covered with pitted yellow glaze (2.5 Y 7/6). External surface (7.5 YR 6/4) with incised lines. Wheel/handmade. Glazed White Ware I. Parallels: Vroom 2005 EBYZ 17; Harrison et al. 1968. Date= 7-8 century. C3082. S24. Rim of bowl. Fine pink-red fabric (2.5 YR 7/6) with fine mineral inclusions. Yellow glaze (5 Y 7/8) with painted lines (2.5 Y 5/6). Wheelmade. Coloured Sgraffito ware (?). Parallels: Dark 2001: 136-137. Date= 12-13 century. Unglazed Table Wares C984. S70. Rim and wall of a small bowl. Medium light orange fabric (5 YR 7/6) with black grit, limestone, and mica/quartz inclusions. Wheelmade. Date= Early Byzantine (?). C1046. U469 (Abdulkadir). Ridged body sherd. Medium fine light orange fabric (2.5 YR 6/8) with red and black grit inclusions. Wheelmade. Date= Early Byzantine (?). C1090. U683 (Okçular). Rim and handle of a small jug. Medium fine light orange fabric (2.5 YR 7/8) with grit inclusions. Wheel/handmade. Date= Byzantine. C1404. G1 (Okçular Kale). Handle for a large jug. Medium light red fabric (2.5 YR 6/8) with black grit, limestone, and quartz/mica inclusions. Handmade. Impressed design. Date= Byzantine. C1598. U283 (Abdulkadir). Rim and body of bowl. Medium orange fabric (10R5/8)with black grit inclusions. Wheelmade. Date= Early Byzantine (?). C2105. U241 (Abdulkadir). Handle of a large jar or pot. Medium light red fabric (2.5 YR 6/8-7/8) with red grit inclusions. Handmade. Impressed Design. Date= Early Byzantine (?). C3349. U2035 (Çamdibi). Rim and handle of a jug. Medium orange-tan fabric (5 YR 7/6) with black grit and organic inclusions. Wheel/handmade. Date= Byzantine. Kitchen Wares C1602. U390 (Abdulkadir). Handle and rim of a large jug. Medium light-red fabric (5 YR 7/4) with limestone inclusions. Wheel/handmade. Date= Middle/Late Byzantine. C1722. S17. Rim of a cooking pot. Medium red fabric (2.5 YR 6/8) with black and red mineral and/or grit inclusions. Wheelmade. Parallels: Vroom 2005 MBYZ 18. Date= 12-13 century. F28 (C3024-25, 3031-35). S24. Profile of a cooking pot. Medium fine brown-red fabric with grit, mineral and quartz inclusions. Wheelmade. Parallels: Vroom 2005: LBYZ15. Date= 13-14 century. C3036-3039. S24. Rim and wall of a pot or jar. Medium fine brown-red fabric with limestone inclusions.Wheelmade. Parallels: Vroom 2005 MBYZ 18. Date= 12-13 century. C3268. U842 (Cide). Rim and handle of a large jug. Medium red fabric 2.5 Y 7/8) with red grit or grog temper. Wheel/handmade. Date= Middle Byzantine (?). Amphorae C24. S10. Ribbed body sherd. Medium yellow-grey fabric (7.5 YR 7.1) with a light gray slip. LR Amphora I, later variation (?). Parallels: Vroom 2005 EBYZ 11. Date= 7-9 century (?). C942. U455 (Abdulkadir). Handle. Sandy buff-red fabric(5 YR 7/4) with limestone inclusions. Finger marks visible. Günsenin Amphora 1. Parallels: Vroom 2005 MBYZ 13. Date= 10-11 century. C992. S70. Neck. Medium light brown-orange fabric (2.5 YR 5/8) with black grit and limestone inclusions. LR Amphora 2. Parallels: Vroom 2005 EBYZ 12. Date= 4-7 century (?). C1027. S71. Rim and neck. Medium buff orange (5 YR 7/8) fabric with black, red and mica mineral inclusions. Globular Amphora. Parallels: Vroom 2005 EBYZ 16. Date= 7-8 century (?). C3867. U3147 (Cide). Medium red-orange fabric (2.5 YR 6/8) with black and grey grit and red grog inclusions. Globular Amphora (?). Parallels: Vroom 2005 EBYZ 16. Date= 7-8 century (?). C3934. U3097 (Aybasan). Medium red fabric (10R5/8) with limestone inclusions. LR Amphora 3 (?). Parallels: Vroom 2005: EBYZ 13. Date= 4-7 century (?).
Bibliography
361
Tiles C3451-3453. U3059 (Çamdibi). Tile. Buff fabric. Handmade. C3454-3462. U3059. (Çamdibi). Tile. Buff fabric. Handmade. C1330. U279 (Abdulkadir). Tile. Orange fabric. Handmade.
Bibliography website http://www.princeton.edu/avkat/reports/ (last accessed 9 May 2014). Greco-Roman sources Arrian, Periplus Ponti Euxini. [Edited and translated by A. Silberman]. Paris, Belles Lettres, 1995. Justinian, Novels. [Edited by S.P. Scott] Cincinnati, The Central Trust Company, 1932. Pliny the Younger, Epistles. [Translated by W. Williams]. Warminster, Aris and Phillips, 1990. Procopius, Buildings. [Edited and translated by H.B. Dewing]. Cambridge, MA: Harvard University Press, 1960. published sources Adam, J.-P. 1994. Roman Building: Materials and Techniques. London, Batsford. Anderson, W. and Robinson, A. 2012. Marginal or mainstream. The character of settlement in Late Roman Paphlagonia. In G. Tsetskhladze, ed., The Black Sea, Paphlagonia, Pontus and Phrygia in Antiquity. Oxford, Archaeopress: 13-27. Angold, M. 1997. The Byzantine Empire, 1025-1204. [2nd edition]. London, Longman. Armstrong, P. 2008. Ceramics. In E. Jeffries, J. Haldon, and R. Cormack, eds., Oxford Handbook of Byzantine Studies. Oxford, Oxford University Press: 429-43. — 2009. Trade in the east Mediterranean in the 8th century. In M.M. Mango, ed., Byzantine Trade, 4th-12th Centuries: The Archaeology of Local, Regional, and International Exchange. Farnham, Ashgate: 157-78. Atasoy, S. 2012. New exploration of the southern Black Sea coast: Filyos-Tios. In G. Tsetskhladze, ed., The Black Sea, Paphlagonia, Pontus and Phrygia in Antiquity. Oxford, Archaeopress: 29-33. Atasoy, S. and Yıldırım, Ş. 2011. Filyos-Tios 2009 yılı kazısı. Kazı Sonuçları Toplantısı 32: 1-16. Belke, K. 1996. Paphlagonien und Honorias. Vienna, Verlag der Österreichischen Akademie der Wissenschaften. Booth, I. 2004. The Sangarios frontier: The history and strategic role of Paphlagonia in Byzantine defence in the 13th century. Byzantinische Forschungen 28: 45-86. Bottema, S., Woldring, H., and Aytuğ, B. 1993-4. Late Quaternary vegetation history of northern Turkey. Palaeohistoria 35-36: 13-72. Bryer, A., and Winfield, D. 1985. The Byzantine Monuments and Topography of the Pontos. Washington DC, Dumbarton Oaks. Campbell, S. 1998. The Mosaics of Anemurium. Toronto, PIMS. Cassis, M. 2009. Çadır Höyük: A rural settlement in Byzantine Anatolia. In T. Vorderstrasse and J. Roodenberg, eds., Archaeology of the Countryside in Medieval Anatolia. Leiden, NINO: 1-24. Crow, J. 1996. Alexios Komnenos and Kastamon: Castles and settlement in Middle Byzantine Paphlagonia. In M. Mullett and D. Smythe, eds., Alexios I Komnenos I. Belfast: 12-36. — 2009. Byzantine castles or fortified places in Paphlagonia and Pontus. In T. Vorderstrasse and J. Roodenberg, eds., Archaeology of the Countryside in Medieval Anatolia. Leiden, Nino: 25-43. Crow, J., and Hill, S. 1990. Amasra, A Byzantine and Genoese fortress on the Black Sea. Fortress 2: 3-15.
362
The Cide-Şenpazar Region in the Byzantine Period
— 1995. The Byzantine fortifications of Amastris in Paphlagonia. Anatolian Studies 45: 251-65. Dark, K. 2001. Byzantine Pottery. Stroud, Tempus. Degeest, R. 2000. The Common Wares of Sagalassos. Turnhout, Brepols. Domżalski, K. 2011. Late Roman Pottery from Pompeiopolis. In L. Summerer, ed., Pompeiopolis I: Eine Zwischenbilanz aus der Metropole Paphlagoniens nach fünf Kampagnen (2006-2010). Langenweissbach, Beier & Beran: 163-77. Doonan, O. 2004. Sinop Landscapes: Exploring Connection in a Black Sea Hinterland. Philadelphia, University of Pennsylvania Museum of Archaeology and Anthropology. Dunn, A. 1994. The transition from polis to kastron in the Balkans (III-VIIcc.): General and regional perspectives. Byzantine and Modern Greek Studies 18: 60-80 Erciyas, B., and Sökmen, E. 2010. An Overview of Byzantine period settlements around Comana Pontica in north-central Turkey. Byzantine and Modern Greek Studies 34: 119-41. Foss, C. 1982. The defenses of Asia Minor against the Turks. The Greek Orthodox Theological Review 27: 145-205. Foss, C., and Winfield, D. 1986. Byzantine Fortifications: An Introduction. Pretoria, University of South Africa. Gough, M., ed., 1985. Alahan: An Early Christian Monastery in Southern Turkey. Toronto, PIMS. Haldon, J. 1997. Byzantium in the Seventh Century. [revised edition]. Cambridge, Cambridge University Press. — 2007. ‘Cappadocia will be given over to ruin and become a desert.’ Environmental evidence for historically-attested events in the 7th-10th centuries. In K. Belke, E. Kislinger, A. Külzer, and M. Stassinopoulou, eds., Byzantina Mediterranea: Festschrift für Johannes Koder zum 65. Geburtstag. Vienna, Böhlau: 215-30. Harrison, R. M., Firatli, N., and Hayes, J.W. 1968. Excavations at Sarachane in Istanbul: Fifth preliminary report, with a contribution on a seventh-century pottery group. Dumbarton Oaks Papers 22: 195-216. Harvey, A. 1989. Economic Expansion in the Byzantine Empire, 900-1200. Cambridge, Cambridge University Press. Hirschfeld, G. 1883. Notes of travel in Paphlagonia and Galatia. The Journal of Hellenic Studies 4: 275-80. Hongimann, E. 1939. Le Synekdèmos d’Hiéroklès et L’Opscule Géographicque de Georges de Chypre. Brussels, Institute de Philologie et d’Histoires Philologie et d’Histoire Orientales et Slaves. Izdebski, A. 2012. The changing landscapes of Byzantine northern Anatolia. Archaeologia Bulgarica 16(1): 47-66. — 2013. A Rural Economy in Transition: Asia Minor from Late Antiquity into the Early Middle Ages. Warszawa: Supplement to The Journal of Juristic Papyrology (Volume 18). Kazhdan, A. 1991. The Oxford Dictionary of Byzantium. Oxford: Oxford University Press. Krautheimer, R. 1986. Early Christian and Byzantine Architecture. [3rd Edition]. Harmondsworth, Penguin. Lawrence, A.W. 1983. A skeletal history of Byzantine fortification. The Annual of the British School at Athens 78: 171-227. Levick, B. 1979. Pliny in Bithynia – and what followed. Greece and Rome 26: 119-31. Lightfoot, C. 2009. Excavations at Amorium: Results from the last ten years (1998-2008). In T. Vorderstrasse and J. Roodenberg, eds., Archaeology of the Countryside in Medieval Anatolia. Leiden, NINO: 139-53. Magdalino, P. 1998. Paphlagonians in Byzantine high society. In S. Lampakes, ed., Byzantine Asia Minor (605-1205). Athens, Instituto Vyzantinon Ereunon: 141-50. Mango, C. 1976. Byzantine Architecture. New York, H.N. Abrams. Marek, C. 1993. Stadt, Ära und Territorium in Pontus-Bithynia und Nord-Galatia. Tübingen, Wasmuth.
Bibliography
363
— 2003. Pontus et Bithynia: Die Römischen Provinzen im Norden Kleinasiens. Mainz, Phillip von Zabern. Matthews, R., and Glatz, C., eds., 2009. At Empires’ Edge: Project Paphlagonia: Regional Survey in North-Central Turkey. London, British Institute at Ankara. Nesbitt, J. 1969. The Life of St. Philaretos and its significance for Byzantine agriculture. The Greek Orthodox Theological Review 14: 150-58. Ousterhout, R. 1999. Master Builders of Byzantium. Princeton, Princeton University Press. Poblome, J. Degryse, P, Cottica, D., and Fırat, N. 2001. A new early Byzantine production centre in western Asia Minor. Rei Cretariae Romanae Fautorum Acta 37: 119-26. Poblome, J., Degryse, P, Schlitz, M., Degeest, R., Viaene, W., Librecht, I., Paulissen, E, and Waelkens, M. 2000. The Ceramic Production Centre of Sagalassos. Rei Cretariae Romanae Fautorum Acta 36: 39-42. Russell, J. 1986. Transformations in early Byzantine urban life: The contribution and limitations of archaeological evidence. In 17th International Byzantine Congress, Major Papers. New Rochelle, A.D. Caratzas: 137-54. Steadman, S., McMahon, G., and Ross, J. 2013.The 2009 and 2012 seasons of excavation at Çadır Höyük on the Anatolian north central Plateau. Anatolica 39: 112-167. Strobel, K. 2002. State formation by the Galatians of Asia Minor. Anatolica 28: 1-46. Treadgold, W. 1997. A History of the Byzantine State and Society. Stanford, Stanford University Press. Vionis, A.K., Poblome, J., De Cupere, B., and Waelkens, M. 2010. A Middle-Late Byzantine pottery assemblage from Sagalassos: Typo-chronology and sociocultural interpretation. Hesperia 3: 423-64. Vroom, J. 2003. After Antiquity: Ceramics and Society in the Aegean from the 7th to the 20th Century A.C. A Case Study from Boeotia, Central Greece. Leiden, Archaeological Studies Leiden University. — 2004. Late Antique pottery, settlement and trade in the East Mediterranean: A preliminary comparison of ceramics from Limyra (Lycia) and Boeotia. In W. Bowen, L. Lavan, and C. Machado, eds., Recent Research on the Late Antique Countryside. Leiden, Brill: 282-331. — 2005. Byzantine to Modern Pottery in the Aegean. Bijleveld, Parnassus Press. Vryonis, Jr., S. 1971. The Decline of Medieval Hellenism in Asia Minor and the Process of Islamization from the Eleventh through the Fifteenth Century. Berkeley, University of California Press. Whittow, M. 1995. Rural fortifications in Western Europe and Byzantium, tenth to twelfth century. In S. Efthymiadis, C. Rapp, and D. Tsougarakis, eds., Bosphorus. Amsterdam, A.M. Hakkert: 57-74. — 2009. Early Medieval Byzantium and the end of the ancient world. Journal of Agrarian Change 9: 134-53. Wilken, R. L. 2003. The Christians as the Romans saw Them. New Haven, Yale University Press. Witte-Ore, J. 2007. Bricks and tiles form the triangular tower at Amorium. In B. Böhlendorf, R. Arslan, A.O. Uysal, and J. Witte-Orr, eds., Çanak: Late Antique and Medieval Pottery and Tiles in Mediterranean Archaeological Contexts. Istanbul, Deutsches Archäologisches Institut: 295-308. Wittke, A-M., Olshausen, E., and Szydlak, R., eds. 2010. Brill’s New Pauly: Historical Atlas of the Ancient World. Leiden, Brill.
Joanita Vroom
12 The Fortress of Çoban Kalesi, Late Medieval Ceramic and Small Finds The fortress of Çoban Kalesi (S32) is situated on the Black Sea coast to the north of Cide and near the modern village of Çayyakka (Fig. 12.1). The Medieval kale is located on a small rocky peninsula with steep slopes, which is presently connected to the mainland by a small land ridge on its southern side (Fig. 12.2), each side is flanked by a small, sheltered bay, where ships could probably moor. In the course of the Cide Archaeological Project, fortresses such as Çoban Kalesi were investigated as part of the project’s targeted reconnaissance activities (section 4.3.2). This involved the production of a basic map, the description of the site, its surroundings and any extant architectural features as well as the collection of surface artefacts. The fortification walls of the castle are placed directly on top of the northern sloping bedrock and run from the eastern to the western edge of the small peninsula (see also Belke 1996: 185). They are built in straight segments, interspersed at regular intervals by two round towers at the outer edges, four rectangular towers in between, and a number of smaller triangular towers/buttresses (Fig. 12.3). A rectangular
Fig. 12.1: Map showing the location of Çoban Kalesi (S32) in relation to other kales and contemporary sites. Produced by Toby C. Wilkinson.
The Fortress of Çoban Kalesi, Late Medieval Ceramic and Small Finds
Fig. 12.2: Schematic plan of Çoban Kalesi (S32). Produced by Victor Klinkenberg.
Fig. 12.3: A view of the fortifications of Çoban Kalesi (S32) (CAP-P2/198).
365
366
The Fortress of Çoban Kalesi, Late Medieval Ceramic and Small Finds
structure, probably a cistern, measuring circa 7 x 4m, can be distinguished at the top of the fortress. The walls (120cm wide) are made of flat squared stones in horizontal layers that are interspersed with thick layers of very hard mortar. No ceramic tiles or brickwork, typical of Byzantine building techniques, were used in the construction of these walls. Instead, unworked rubble set in a hard whitish lime mortar with fine aggregate (stones up to 1cm in size) would point towards a Genoese building technique (Crow and Hill 1995: 258-9). The total amount of surface material collected at Çoban Kalesi consists of approximately 30 pieces, the majority of which derive from the round western fortification tower and its surroundings. Among the finds are glazed and unglazed wares as well as a few small finds made of glass and metal.
12.1 The Glazed Wares The glazed pottery fragments from Çoban Kalesi include mostly open vessels, such as small-sized bowls for food consumption/distribution. They all have a thick whitish slip (or engobe) on the interior and on the upper part of the exterior surface. This white slip is often incised before firing with a fine sharp potter’s tool in order to get a delicate thin-lined decoration. Furthermore, the slip can be gouged with a blunt or broad-bladed tool for thicker engraved motifs. Two joined body fragments (C341) and one body fragment (C3200) are part of a so-called Monochrome Sgraffito bowl with a straight divergent upper body and straight symmetrical rim.1 Their fabric colour is reddish yellow (5 YR 7/6) to red (10 R 5/8); the lead glaze on the interior and exterior surfaces of these sherds has a pale yellow (2.5 Y 8/4) to greenish-yellow tinge (5 Y 7/8) (Fig. 12.4). All three pieces have an incised decoration on the interior, showing a combination of finely engraved and gouged lines. The design is abstract and simple. It consists of concentric gouged circles at the centre and under the rim, which are filled in with vertical groups of fine straight and wavy lines. The three fragments belong to the ‘Zeuxippus Ware Family (IA)’ and are variants of the so-called ‘Zeuxippus Ware’, a type of 13th-century AD glazed pottery with incised decoration that was first recognised during excavations at the Baths of Zeuxippus in Constantinople, modern Istanbul (e.g. Megaw 1968; Waksman and François 2004-2005; Vroom 2003: 164-5; 2005: 108-11; cf. for a similar shape, Inanan 2013: 76, table II-1, no. 6). The type of incised decoration on the three sherds is further known as ‘motifs à rayons’, and it is sometimes alternated with S-shaped designs (François 1995: pl. 16a).
1 The term ‘Sgraffito’ derives from the Italian word sgraffiare, which literally means ‘scratching’ (Vroom 2005: 203).
The Glazed Wares
367
Fig. 12.4: Sgraffito sherds from Çoban Kalesi. Produced by Lorraine McEwan.
This decorative style is common on Late Medieval/Late Byzantine ceramics found on sites in Romania, Bulgaria and along the Black Sea coast (Nicolescu and Popa 1965: 313). We may notice similar-looking abstract motifs on pottery finds from Turnu-Severin, Zimnicea and Curtea de Argeş (Nicolescu 1960-61: fig. 20), from Pacuiul-lui-Soare (Nicolescu and Popa 1965: figs. 3, 4b ‘l’onde exéctée au peigne’), from Carevgrad Tãrnov (Georgieva 1974: 71, pl. V, 81, fig. 55), from Drastar (Koleva
368
The Fortress of Çoban Kalesi, Late Medieval Ceramic and Small Finds
2010: pl. I, nos. 17-200), from Chersonesos (Romancuk 2003a: figs. 62, no. 198, 63, no. 201, 183, no. 494, 185, nos. 499; 2003b: fig. 4, no. 3, fig. 9, no. 2; 2005), and finally from Azak (Bocharov and Maslovskiy 2012: fig. 1, nos. 16-7, group 2). A few fragments with such an incised decoration were also found in northern Greece (Thrace) and in Ilion on the western coast of Turkey (Böhlendorf-Arslan 2004: pl. 46, nos. 16-7 for Thrace, pl. 84, no. 246 for Ilion). As yet, however, we do not know the provenance of this type of glazed decorated ware. These glazed engraved bowls can be generally dated in the second half of the 13th to the early 14th centuries AD (François 1995: 102). They were imported in various decorative styles and in substantial quantities to the northeastern Black Sea region and to the northern Azov region (especially to the site of Azak) from the middle to the third quarter of the 13th century AD. Apparently, towards the end of the 13th century AD, their import ceased entirely in both regions (Bocharov and Maslovskiy 2012: 272). The survey at Çoban Kalesi also yielded other, contemporary, glazed wares. These include, for instance, a rim fragment (C3207) and a body fragment (C3204) of Monochrome Sgraffito Ware. The fabric of both pieces is dark red (10 R 5/4 to 2.5 YR 6/6) and covered with a white slip (5Y 8/1). The lead glaze on top of the fabric and slip has a dark olive to light green-yellowish colour (5 Y 4/4 to 8/2). Gouged concentric circles may be recognised on the interior upper part or below the rim of the sherds. In addition, there are more finds of Monochrome Sgraffito Ware in various lead glaze colours, such as a body fragment (C3202) with a light red fabric (2.5 YR 6/6) and an olive greenish glaze (5 Y 5/6), and a rim fragment (C3203) of a dark red fabric (10 R 5/6) and a yellow glaze (5 Y 8/6). They all belong to the same date range as the above-mentioned fragments of the ‘Zeuxippus Ware Family (IA)’. An exception, however, is a rim fragment (C335) that is covered with a white slip in and out. Its shape is different from the rest, as it has a convex divergent upper body with a straight rim. This piece can probably be dated in the late 12th to early 13th centuries AD. Its fabric colour is brownish (10 YR 5/3); its lead glaze pale yellow (5 Y 8/4). In terms of its shape, this piece seems to belong to the so-called ‘Incised Sgraffito Ware’ group that can be dated in the late 12th and early 13th centuries AD (Vroom 2005: 90-1). This date would fit well with two late 12th to early 13th century AD Günsenin 3 amphora body fragments also from Çoban Kalesi.
12.2 The Unglazed Wares Examples of unglazed utilitarian pieces from Çoban Kalesi include mainly closed vessels, such as jars and pots that are probably used for cooking. We can distinguish a rim fragment of a thin-walled cooking pot with an everted rim and a straight convergent upper body (C3201). Its fine fabric has a reddish brown colour (2.5 YR
The Small Finds
369
4/4). Its shape is typical for Late Medieval cooking pots/jars found in the Black Sea region and in the northern Azov region. We see similar forms, for instance, in Chersonesos (Golofast and Ryzhov 2003: fig. 4, no. 1) and in Azak (Belinksi and Maslovski 2005: pl. 7, no. 20), both situated on the northern shorelines of the Black Sea. The pottery assemblage includes another rim fragment of a thin-walled jar (C3206). Its fine fabric has a light brown colour (2.5 YR 7/6), and its shape is related to Late Medieval pots recovered at the Genoese castle of Cembalo in Balaklava, on the southern coast of the Crimea (Adaksina et al. 2005: fig. 97, 1, 1). Other pieces of Late Medieval unglazed cooking pots are also worth mentioning. Two fragments have perforated holes either in the rim below the handle (C334) or in the broad oval-shaped handle itself (C342). Their fabric is medium and has a reddish colour (10 R 6/6 to 5/3). Their form is analogous to pots/jars recovered at sites on the northern littorals of the Black Sea region, such as Cembalo/Balaklava and Azak (cf. for shape, Adaksina et al. 2005: fig. 97, 1 (1); Belinksi and Maslovskiy 2005: pl. 7, no. 1 and Bocharov and Maslovskiy 2012: fig. 2, no. 17, Azak group 4, dated in the third quarter of the 14th century AD). The same handle shape can be detected in another cooking pot handle (C333), although its fabric is finer and has a light orange colour (7.5 YR 8/3). Finally, I mention here a rim fragment with an applied ‘pie-crust’ decoration (C3205). The fabric is fine and has a light brown colour (2.5 YR 5/6). This piece has a convex convergent body with a flattened rim. A rim diameter of only 18cm suggests a use as a lid, perhaps for a large storage jar. In fact, we can recognise similar Late Medieval/Late Byzantine lids from Lusta in the Crimea and from Carevgrad Tãrnov in Bulgaria (cf. for shape, Mytz 2002: fig. 22, no. 2 from Lusta; Georgieva 1974: figs. 34-5, 37 from Carevgrad Tarnov; Busuioc 1975: fig. 50, nos. 2-5.). The latter examples are generally dated in the 14th-15th centuries AD.
12.3 The Small Finds Among the glass finds, we can distinguish a ring base fragment of a green glass vessel (F13) as well as a body fragment with an applied knob or droplet (F31) (Fig. 12.5). Both pieces probably belong to Late Medieval/Late Byzantine glass beakers with applied small droplets (or ‘prunts’) on the outside. Unfortunately, the pieces from Çoban Kalesi are quite small, but they look very similar to prunted beakers from Corinth (Greece), from sites in southern Italy and in the Near East, and in particular to vessels from Anaia/Kadı Kalesi (nearby Kuşadası, in western Turkey) and from the Seljuk Palace inside the inner Citadel of Alanya (in southern Turkey). The parallels from these sites can all roughly be dated to the 13th and 14th centuries AD (e.g. Davidson 1952: fig. 14, nos. 742, 744, 746; Bakirer 2009: figs. 1-9; Oral Çakmakçı 2013: 139, fig. VIII-4).
370
The Fortress of Çoban Kalesi, Late Medieval Ceramic and Small Finds
Fig. 12.5: Selected sherds and small finds from Çoban Kalesi (S32). Produced by Lorraine McEwan.
12.4 Some Concluding Remarks To conclude, most of the surface finds from Çoban Kalesi can be dated to the Late Medieval/Late Byzantine period and especially in the 13th-14th centuries AD. Both the ceramic and the glass finds suggest that the Genoese fortification on the peninsula was in use only for a relatively short period. The only exceptions are a slightly earlier
Some Concluding Remarks
371
glazed rim fragment of the late 12th-early 13th century AD and an unglazed lid fragment that can be dated in the 15th century AD. For the table wares, the most common glaze colours are yellow and olive green, mostly covering red-bodied fabrics and white slipped surfaces. The decoration of these glazed fragments mainly consists of incised abstract designs, such as concentric circles. On the basis of these characteristics, a few fragments can be identified as ‘Zeuxippus Ware imitations’ (or ‘Zeuxippus Ware derivatives’). Pottery similar to the glazed decorated sherds from Çoban Kalesi has been found at several sites around the Black Sea, for instance in Bulgaria, Romania and along the Crimean coast, especially at Chersonesos and Azak. Both the glazed and unglazed pottery show that Çoban Kalesi was closely connected with other parts of the Black Sea region, and that these wares were imported from overseas. The fabric, style and date of the surface finds match the period in which - in my assessment - the Genoese fleet was very active in the Black Sea area and Genoese merchants were operating in many of the coastal markets in this region. By the 12th century AD most of inland Anatolia (including the Byzantine province of Paphlagonia) had come under Seljuk control and only a narrow coastal strip with natural strongholds along the Black Sea coast remained under Byzantine overlordship between AD 1181 and AD 1282 (Booth 2004: 59 and map 4). Once coastal Paphlagonia was cut off from its interior agricultural hinterland and only accessible from the sea, towns and castles were forced to import grain and goods from elsewhere, and Venetian and Genoese merchants were often involved in their transportation (Karpov 1993: 57). Cities and fortresses now became increasingly isolated and were gradually constructed in a more defensive mode, as was obviously the case at Çoban Kalesi. It is therefore not entirely surprising that the fortification walls on the peninsula were constructed using Genoese building techniques. The 13th-14th centuries AD is also the time in which Çoban Kalesi is described in Late Medieval portolan charts (navigational maps) as the station ‘Gira Petra’, although some scholars alternatively identify the site as ‘Docastelli’ (Belke 1996: 185). Whatever its Medieval name, it is perfectly plausible that the site was used in this period by the Genoese as a small trading post for a modest fleet of small- or medium-sized ships practicing cabotage navigation. This enabled the regional distribution of goods by petty merchants between harbours at short- or medium-range distances (Jacoby 2001: 229). Small boats with limited tonnage could moor quite easily in the sheltered bays south of the fortified kale. More generally, it is quite certain that harbours on the western coast of the Black Sea region such as Caffa/Feodosija and Cembalo/Balaklava were used by Italian traders as protective transit and transhipment stations for the stocking of provisions, money transactions and banking, as well as for ship repairs (François 1995: 143; Jacoby 2001: 229). Although the Black Sea was known as a hazardous water to navigate, the exchange of goods in the region was based on seafaring throughout the centuries (securely attested from the Iron Age to Ottoman times and beyond). It is therefore
372
The Fortress of Çoban Kalesi, Late Medieval Ceramic and Small Finds
important to note that Çoban Kalesi is situated at the point where the western, anticlockwise Black Sea current trends northwards, and facilitates a swift crossing of the Black Sea to the Crimea, and from there to Romania, Bulgaria and the Bosphorus (see section 2.2.1, this volume, and especially Fig. 2.5 for the Black Sea currents). For the Italian maritime powers, the Byzantine province of Paphlagonia was an important region. Not only was it a strategic location for navigating the Black Sea, but also a major source of agricultural produce that was exported to Constantinople/ Istanbul, the Crimea and other places in the Black Sea region. The shipping of agricultural products across the Black Sea is well documented at this time. It is wellknown, for instance, that by the end of the 1280s AD the Genoese were involved in substantial grain exports from the northern parts of the Black Sea (e.g. Crimea) to the southern parts of the region (Karpov 1993: 59). Apart from grain, there was also intra-regional trade and transportation of textiles, wax, honey and slaves between the Crimea (Caffa) and Constantinople/Istanbul (Jacoby 2001: 227). In short, the architecture and surface finds with parallels from the western and northern shores of the Black Sea and the northwest Aegean, as well as available historical information, seem to suggest that Çoban Kalesi was part of a Genoese intraregional distribution system of commodities in the Black Sea area.
Catalogue of Illustrated Artefacts Ceramics C334. S32. Rim fragment with handle attachment and perforation below handle. Red medium fine paste with some calcareous temper, sparse quartz and some brown grits. Oxidised firing. Wheelmade. Fabric 10 R 6/6, surface 10 R 6/6. D= 9cm. C335. S32. Rim fragment of bowl with glaze. Brownish medium fine paste with sparse calcareous and quartz inclusions, and some grit. Exterior and rim in part covered by glaze. Oxidised firing. Wheelmade. Fabric 10 YR 5/3, glaze 5 YR 8/4. D= 30cm. C341. S32. Sgraffito body sherd with carination. Brownish red fine paste with calcareous and sparse grog inclusions. Glazed with incised decoration. Oxidised firing. Wheelmade. Fabric 5 YR 5/6, glaze 10 YR 8/6. C342. S32. Handle fragment with perforation. Greyish brown paste with calcareous and mica inclusions. Differential firing. Wheelmade. No Munsell recorded. C3200. S32. Sgraffito body sherd. Brownish red fine paste with calcareous and sparse grog inclusions. Glazed with incised decoration. Oxidised firing. Wheelmade. Fabric 10 R 5/8, glaze 5Y 7/8. C3201. S32. Rim fragment of jar/cooking pot. Brownish red hard paste. Some smudging on rim Oxidised firing. Wheelmade. Fabric and surface 2.5 YR 4/4. D= 14cm. C3203. S32. Glazed rim fragment of bowl. Dark red fine paste. Glazed. Oxidised firing. Wheelmade. Fabric 10 R 5/6, glaze 5Y 8/6. D= 10cm. C3205. S32. Pedestal base with impressed decoration. Light brown paste. Finger impressed decoration. Oxidised firing. Wheelmade. Fabric and surface 2.5 YR 5/6. D= 18cm. C3206. S32. Rim fragment of jar/jug. Light brown paste with dense black grit. Oxidised firing. Wheelmade. Fabric and surface 2.5 YR 7/6. D= 4cm.
Bibliography
373
C3207. S32. Glazed rim fragment. Dark red medium fine paste. Glazed decoration with decoration in reserve. Oxidised firing. Wheelmade. Fabric 10 R 5/4, glaze 5 Y 4/4. D= unknown. Objects F13. S32. Fragment of pedestal base of glass vessel. Very light yellow-brown colouring, oily patina. Folded ends. D= 8cm. 3mm thick. F31. S32. Fragment of glass vessel with a knob attached. 1mm thick.
Bibliography Adaksina, S.B., Kirilko, V.P., and Mytz, V.L. 2005. Report of Archaeological Excavations in the Genoa Fortress of Cembalo (Balaklava, Crimea) in AD 2004. [in Russian]. St. Petersburg, Simferopol. Bakirer, Ö. 2009. Glass from the Seljuk Palace at Alanya. In E. Laflı, ed., Late Antique / Early Byzantine Glass in the Eastern Mediterranean. Izmir, Hürriyet Maatbaşı: 199-212. Belinski, I.V., and Maslovski, A.N. 2005. The imported glazed ceramics from Asak (XIV cent.). [in Russian]. In S.G. Bocharov and V.L. Mytz, eds., Glazed Pottery of the Mediterranean and Black Sea Region of the X-XVIII Centuries. Kiev, Stylos: 160-77. Belke, K. 1996. Paphlagonien und Honōrias. Vienna, Österreichischen Akademie der Wissenschaften. Bocharov, S. and Maslovskiy, A. 2012. Byzantine glazed pottery in the cities of the north Black Sea region in the Golden Horde period (second half of 13th century – end of 14th century). In S. Gelichi, ed., Atti del IX congress internazionale sulla ceramice medievale nel Mediterraneo. Venice, All’Insignia del Giglio: 270-5. Böhlendorf-Arslan, B. 2004. Glasierte byzantinische Keramik aus der Türkei. Istanbul: Ege Yayınları. Booth, I. 2004. The Sangarios frontier: The history and strategic role of Paphlagonia in Byzantine defence in the 13th century. Byzantinische Forschungen 28: 45-86. Busuioc, E. 1975. La céramique d’usage commun, non émaillée de Moldavie (XIVe siècle jusqu’au milieu du XVIe siècle). [in Romanian]. Bucharest, Editura Academiei Republicii Sociliste Romania. Crow, J., and Hill, S. 1995. The Byzantine fortifications of Amastris in Paphlagonia. Anatolian Studies 45: 255-65. Davidson, G.R. 1952. Corinth, vol. XII: The Minor Objects. Princeton, N.J., The American School of Classical Studies at Athens. François, V. 1995. La céramique byzantine à Thasos. Athens, École Française d’Athènes. Georgieva, S. 1974. La céramique du palais de Carcev. [in Serbian]. In K. Myatev, D.S. Angelov, and N. Angelov, eds., Carevgrad Tãrnov. Le palais des rois bulgares pendant le deuxième royaume bulgare (XIIe-XIVe s.). Vol. 2: Céramique objects domestiques et armament, parures et tissus. Sofia: 1-186. Golofast, L.A., and Ryzhov, S.G. 2003. Excavations in the quarter X in the northern district of Chersonesos. [in Russian]. Materialy po Arheologii, Istorii I Etnografii Tavrii VIII: 182-251. Inanan, F. 2013. Zeuksippus tipi seramikler ve Kuşadası/Anaia örnekeleri / Zeuxippus type ceramics and samples from Kadıkalesi/Anaia, Kuşadası. In Z. Mercangöz, ed., Bizanslı Ustular – Latin Patronlar / Byzantine Crafstmen – Latin Patrons, Istanbul, Ege Yayınları: 59-76. Jacoby, D. 2001. Changing economic patterns in Latin Romania: The impact of the West. In A.E. Laiou and R.P. Mottahedeh, eds., The Crusades from the Perspective of Byzantium and the Muslim World. Washington, D.C.: 197-233. Karpov, S.P. 1993. The grain trade in the southern Black Sea region: The thirteenth to the fifteenth century. Mediterranean Historical Review 8: 55-73.
374
The Fortress of Çoban Kalesi, Late Medieval Ceramic and Small Finds
Koleva, R. 2010. Data on the sgraffito ware production in the Medieval city of Drastar. [in Bulgarian]. In M. Milcheva, ed., Stephanos Archaeologicos in honorem Professoris Stephcae Angelova (Studia Archaeologica Universitatis Serdicensis suppl. V). Sofia, University of Sofia: 695-712. Megaw, A. 1968. Zeuxippos Ware. The Annual of the British School of Archaeology at Athens 63: 67-88. Mytz, W.L. 2002. Genoese Lusta – Administrative Gotii: 50-70 years of the 15th century. [in Russian] In Alushta and Alushtas Region. From Ancient Times up to Now. Kiev, Stylos: 139-89. Nicolescu, C. 1960-61. La céramique roumaine émaillée du Moyen Âge, à la lumière des dernières recherches. Byzantinoslavica 21-22: 260-73. Nicolescu, C., and Popa, R. 1965. La céramique émaillée des XIIIe et XIVe siècles de Pãcuiul-luiSoare. Dacia 9: 337-50. Oral Çakmakçı, Z. 2013. Anaia Camcılığının Geç Bizans Cam Dünyasındaki Yeri / The place of the Anaian glass production in the Late Byzantine glassworld. In Z. Mercangöz, ed., Bizanslı Ustular – Latin Patronlar / Byzantine Crafstmen – Latin Patrons. Istanbul, Ege Yayınları: 135-52. Romanchuk, A. 2003a. Glazed Pottery of Late Byzantine Cherson, Port Area. [in Russian]. Yekaterinburg, Volot. — 2003b. Befunde der glasierten Keramik der spátbyzantinischen Zeit in Chersonesos: örtliche Herstellung und Import. In Ch. Bakirtzis, ed., VIIe congrès international sur la céramique medieval en Mediterranée. Athens, Édition de la Caisse de Recettes Archéologiques: 101-14. — 2005. Studien zur Geschichte und Archäologie des byzantinischen Cherson. Leiden/ Boston, Brill. Vroom, J. 2003. After Antiquity. Ceramics and Society in the Aegean from the 7th to the 20th Century A.C. A Case Study from Boeotia, Central Greece. Leiden, Archaeological Studies Leiden University. — 2005. Byzantine to Modern Pottery in the Aegean: An Introduction and Field Guide. Utrecht, Parnassus Press. Waksman, Y, and François, V. 2004-2005. Vers une redéfinition typologique et analytique des céramiques byzantines du type Zeuxippus Ware. Bulletin de Correspondance Héllenique 128-129: 629-724.
A.C.S. Peacock, with contributions by Claudia Glatz and T. Emre Şerifoğlu
13 Cide and its Region from Seljuk to Ottoman Times The history of Cide in the Islamic period cannot yet be written. For the most part, its fortunes seem to have been bound up with that of the broader province of Kastamonu to which it was attached for administrative purposes. However, the history of Kastamonu itself is little researched and understood (Gökoğlu 1952: 22-3; Ergene 2003: 9). The Persian and Turkish literary sources, which constitute our main evidence for the Islamic history of Anatolia before circa AD 1500, seldom refer directly to Cide and its surroundings. Archival materials exist in significant quantities only from the 16th century, with a handful from the late 15th century, but their analysis for Kastamonu province has scarcely begun. This chapter, then, does not purport to offer a definitive view of Cide’s past under Turkish rule, for this can only be achieved when it can be contextualised and compared to developments in Kastamonu and north western Anatolia more broadly. This must be a task for future researchers. Rather, it is a preliminary sketch of Cide under Muslim rule in its regional context, from the arrival of the Turks to the end of the Ottoman empire from the 13th to 20th centuries AD (Fig. 13.1).
Fig. 13.1: Map showing locations and sites mentioned in the text. Produced by Toby C. Wilkinson.
376
Cide and its Region from Seljuk to Ottoman Times
13.1 Turks in Cide and the Kastamonu Region: 1071-1462 AD Kastamonu province in the Middle Ages was a famous stronghold of the Turkmen, the nomadic Turks whose livelihood relied on transhumant pastoralism. As early as AD 1073, just two years after the fateful battle of Manzikert which signalled the collapse of Byzantine authority in most of the Anatolian interior, Turkmen near Kastamonu almost captured the emperor Alexios Komnenos (Cahen 1968: 48, 1971; Heywood 1960-2007). However, it is not clear how far Turkmen settlement extended. Both Byzantine and Islamic written sources indicate that the Black Sea littoral of Anatolia remained under Christian rule until the Seljuk capture of Sinop in AD 1214 (see Peacock 2010). Even in the mid-13th century, Theodore II Laskaris, emperor of the Byzantine empire in exile of Nicaea (r. 1254-8), mentions that Cape Karambis (Kerembe Burnu), 34 miles east of Cide, formed the frontier of his realm (Korobeinikov 2010: 216, n. 2). Yet Turkmen may have already started to populate parts of the littoral, probably on a seasonal basis, descending from their high summer pastures in the mountains to the lower lying, warmer coastline in winter. This pattern is well attested from Trebizond (Bryer 1975), although the first evidence for the Turkish western Black Sea littoral comes from the 1270s. Ibn Bibi, the Persian chronicler of Seljuk rule, refers to the ‘Turkmen who live on the coast’ (Turkmanan-i sahil-nishin) of Simre (probably modern Taşköprü), Sinop, Samsun and Bafra (Ibn Bibi 1956: 722). Certainly, in the second half of the 13th century, the frontier seemed to suddenly shift westwards. Writing of events around AD 1267, the Byzantine historian Pachymeres describes the coastal strongholds Kromna (Kurucaşile, located midway between Cide and Amasra), Amastris (Amasra), Tios (Filyos) and Heracleia (Ereğli) as the last fortresses in the region in Greek hands, saved only by their access to the sea, which presumably allowed them to be resupplied (Pachymérès 1984: IV.27). Although Cide does not feature in these 13th century sources, it is probable that its surroundings were likewise inhabited seasonally by these coast-dwelling Turkmen, especially if, as seems logical, we take Pachymeres’ account to be in geographical order from east to west. This would mean that Kromna/Kurucaşile was the eastern most Byzantine settlement on the Black Sea; Cide, therefore, would have lain at the western extremity of Turkish control, and the local Turkmen population most probably were predominantly of the Chepni tribe, for these are known to have settled the Sinop area (Ibn Bibi 1956: 729). However, the account of Husam al-Din Khuyi, discussed below, suggests that as late as 1284 the castle of Gideros slightly to the west of Cide but east of Kromna was controlled by a Christian garrison (Khuyi 2000: 282). The process of conquest may therefore have been slower than Pachymeres implies. It has also been suggested on linguistic evidence that significant elements of the Turkish population of the region may have belonged to the southern Turkic Qipchaq grouping who mainly lived in the south Russian steppe on the other side of the Black Sea (Korobeinikov 2010: 223). Although there was certainly a medieval trade in
Turks in Cide and the Kastamonu Region: 1071-1462 AD
377
Qipchaq slaves routed via the Crimea and Sinop (Peacock 2010: 115-8), it is not clear exactly how or when these Qipchaq elements might have appeared in the population more generally, as presumably most slaves would have been sold on to buyers in central Anatolia and the Mediterranean world. Kütüklü and Tunoğlu (2012: 30) note the existence of several Turkic tribal (cemaat) names derived from places in Kastamonu province scattered among Turkmen across Anatolia, according to Ottoman documents. Among them is the name ‘Cidelü’ (‘of Cide’) which is found in documents relating to Kırşehir and Ergani (near Diyarbakır). If this interpretation is accurate, it would confirm the suggestion here that there was significant Turkmen settlement around Cide. However, the villages of the region must have continued to be populated by Greek-speaking Christians, as is suggested by the prevalence of pre-Turkish place names. We know, however, very little of them. What of the political structures? There is no reason to think that the extension of Seljuk control over Sinop was matched by the presence of much effective Seljuk governance along the coastline to its west. Indeed, the silence of the 13th century Islamic sources over the region is suggestive of its peripherality. It seems likely that beyond the main towns of Kastamonu and Sinop, where Seljuk administrative appointees were based, local Turkmen chiefs held sway (Heywood: 1960-2007; Korobeinikov 2004: 90-92). However, even these Seljuk appointees were themselves often prominent local Turkmen. Initially, the Seljuks had appointed descendants of the Danishmendid Turkmen dynasty to administer the northwestern uj, as the Turkmen-inhabited territories of the Seljuk realm were known (Ibn Bibi 1956: 68). Subsequently, a certain Husam al-Din Chupan was appointed as governor of Sinop after its conquest in AD 1214, and his descendants, known as the Çobanid dynasty, but also probably Turkmen chiefs managed to retain a pre-eminence in the region. His grandson, Muzaffar al-Din Yavlak Arslan b. Alpyürek is mentioned by Ibn Bibi (1956: 741) as sipahdār (military governor) of Kastamonu in the 1280s (see Cahen 1971; Yücel 1988, vol. 1: 34-51; Korobeinikov 2004; his regnal dates are uncertain, from c. 1280 to 1291, see Yücel 1988). However, in the second half of the 13th century another layer of political authority was added, in the form of the Seljuks’ own suzerains the Mongols. A Mongol ally, the Pervane Mu‘in al-Din Suleyman, who was effective ruler of Anatolia in the later 13th century, owned much of the Black Sea littoral around Sinop as his personal estate, and his descendants may have maintained a foothold there (further Korobeinikov 2004: 94-7; Peacock 2010: 104-9). The Turkmen particularly resented Mongol rule, which threatened their control of pasturage, and the region witnessed frequent Turkmen revolts in the late 13th century (Korobeinikov 2004). It is with the Çobanids that we have one unexpected source on the Cide region, the aforementioned account by Husam al-Din Khuyi of the Muslim conquest of the castles of Gideros. Khuyi was a Çobanid chancery official, and he was also an eyewitness to the fighting at Gideros, concerning which he composed an elaborate ‘victory-letter’ (fath-nama) celebrating the conquest of the castle by the Muzaffar al-Din Yavlak Arslan b. Alpyürek in Rajab/October 683 AH/AD 1284 (Khuyi 2000: 282-5). Such letters
378
Cide and its Region from Seljuk to Ottoman Times
were circulated within the ruler’s realm to celebrate his conquest (Khuyi 2000: 285), and sometimes to other rulers as well. The letter, then, is propaganda, and Khuyi is at pains to emphasise the stiff resistance the Muslim forces encountered, and to emphasise the magnitude of Yavlak Arslan’s achievement. The campaign is further legitimised through its characterisation as jihad, and seems to have formed part of a broader conquest of Christian-held territory (Khuyi 2000: 284). Khuyi’s letter describes how the campaign targeted ‘the two castles of Gideros on the coast of the Byzantine territories (az sahil-i fasilyus), opposite one another, connected to the sea’. The impregnability of the two castles is emphasised; the defenders were assisted by men from the western Byzantine lands and from Trebizond (az kishwar-i Lashkari wa taraf-i Tarabzun). The Çobanid army, on the other hand, was assisted by fighters from the Turkmen tribes (qaba'il-i atrak) who were noted for their ferocity and were inflamed by desire to fight ‘the enemies of religion’. The attack was launched on the 4th of Rajab, the assailants using a variety of types of mangonels to destroy buildings inside the castle, while a hail of arrows rained down on the defenders. Nonetheless, the attack lasted a full week. It was not until the 11th of Rajab after the defences had been severely weakened and many of the defenders killed that the final assault could be launched. This was spear-headed by the naphtha-throwers (naffāṭ); the naphtha attack forced many of the remaining defenders to flee, some escaping in boats, many drowning. Eventually the Muslim forces seized control of the castle; the fate of the twin castle at Gideros was identical. The interest of Khuyi’s account comes not just from the exceptional detail that he gives on the battle for Gideros, but also from the fact that he emphasises that as late as AD 1284, the garrison was still connected to the broader Byzantine world, with defenders coming from both western Anatolia and Trebizond to defend this strategic location. The Cide region was thus not as isolated as written sources imply. In addition, Khuyi confirms the suspicion that Turkmen played a leading role in the incorporation of the region into the Muslim domains, although one may imagine that practical considerations such as a desire for unfettered access to pastures may have been at least an equally important motivation as jihad, the factor he emphasises. At any rate, by AD 1284 Cide and its region must have been incorporated into the Çobanid domains. From the beginning of the 14th century, the Çobanids were replaced by another Turkmen dynasty, the Isfandiyardids or Candarids (on whom see Yücel 1988: 53-123). Little is known about Candarid society and economy, and the Cide region’s history is totally obscure in this period. The Egyptian author al-‘Umari tells us that Candarid Kastamonu was famous for its horses, which were exported to Egypt, and also carried out trade with the Crimea and territories in modern Russia and the Ukraine (Yücel 1988: 185-6, 296-7; also Kopraman 1988). Meanwhile, Turkish control gradually encroached over the remaining Byzantine strongholds on the Black Sea, with Heraclea/Ereğli falling in 1360 (Pachymeres 1984: 405, n. 5). The Kastamonu region probably came briefly under Ottoman control around AD 1393-9, but Candarid rule was subsequently restored and endured until the final conquest by the Ottoman
Cide and the Kastamonu Region in the Ottoman Period
379
sultan Mehmet I in 1462, following shortly from his occupation of Genoese-held Amasra in 1460 (for a discussion of Çoban Kalesi, a Genoese fortress on the Cide coast, see chapter 12). Somewhere between 1460 and 1462, then, Cide must have been incorporated into the Ottoman state.
13.2 Cide and the Kastamonu Region in the Ottoman Period The Kastamonu region was given the status of a sancak in the eyalet (province) of Anadolu, and initially Kastamonu seems to have been a place of some importance, assigned as its governors the Ottoman princes Sultan Cem (AD 1469-1474) and Şehzade Mahmud (AD 1481-1504). Areas of the Kastamonu sancak, including villages in the İnebolu and Sinop areas, were allotted as has, i.e. the personal land of the prince from which he drew his income (Yakupoğlu 2010: 324). However, the province’s importance soon declined, as it remained rather isolated from the main trade routes, and its economy seems to have been overwhelmingly agricultural. Rope, linen, silk and woollen cloth were also produced, and wine was made by Christians. From the 16th century onwards, most of Kastamonu was assigned as arpalık (an estate as a revenue source) to officials with the ranks of mirliva and mirmiran who usually governed it in absentia by sending agents (mütesellims) to collect revenue (Tosunoğlu 1984; Ergene 2003: 12-15). From the 16th century we have clearer evidence of Cide’s place in the province. In 916 AH/AD 1510, the first known mosque was built in Cide, the Kara Mustafa Camii (Gökoğlu 1952: 238). Cide appears in the AD 1530 defter of Anadolu eyalet as part of the fiefs (timar) assigned to pay the military (sipahis). It was attached to the kaza (administrative subdivision) of Hoşalay (Doğanyurt) and is described as a village (kariye). The defter gives the following statistics for Cide (438 Numaralı Muhasebe-i Vilayet-i Anadolu Defteri (937/1530), vol. 2: 644-5): 40 hane (households) 8 mücerred (unmarried adult males) 2195 [akçes]1 hasil (total revenue) What sort of population does this indicate? Clearly the critical question is how large a hane could be, and estimates of this range from an average of 3.5 to 7 people (Tosunoğlu 1984: 268; Lowry 2002: 48-52), giving a population roughly in the range of 150 to 300. The total population for the sancak of Kastamonu has been estimated at this date at 137,000 (Tosunoğlu 1984: 268).
1 The akçe was a silver coin which was the main currency unit in the Ottoman empire till the 19th century. For a discussion of its fluctuating value see Pamuk (2000).
380
Cide and its Region from Seljuk to Ottoman Times
As for income, the defter (p. 647) gives a total income for the 66 villages of Hoşalay kaza of some 72,252 akçes: Cide therefore, contributed approximately 3% of the kaza’s tax revenue. Some land in Cide was also alienated as waqf (religious endowments), the income from which amounted to 120 akçes, out of a kaza total of 5711, or just over 2% (ibid, p. 648). Most likely this waqf was to finance the Kara Mustafa Cami (on waqfs in Hoşalay see further Kütüklü and Tunoğlu 2012: 103-4).
Fig. 13.2: Traditional houses at Mencekli Köyü Yukarı Mahalle (CAP-P2/12).
Cide was considerably larger than most other villages mentioned in Hoşalay kaza, many of which had as few as 2 or 3 households, and it contributed a much larger proportion of tax income, most other villages paying under 1000 akçes. Only 19 villages had land endowed as waqf, although Cide’s waqf income is distinctly on the low side. It was, then, a place of modest local importance, and the waqf and the building of the Kara Mustafa mosque suggests the gradual development of Muslim life in the town. The figures for the kaza as a whole suggest that Muslim life was rather undeveloped, with merely one hatib (preacher), two imams (prayer leaders), one congregational mosque (cami, presumably the Kara Mustafa Cami), two smaller mosques (mescid), and two zaviyes (Sufi retreats). These figures are comparable to those for other rural areas of Kastamonu sancak, but in contrast to the towns of Sinop, Kastamonu and Taşköprü where there was a much more substantial infrastructure of Muslim institutions, as one would expect (Tosunoğlu 1984: 232). These figures also
Cide and the Kastamonu Region in the Ottoman Period
381
suggest that the countryside remained in large part Christian. In Hoşalay the revenue decreased slightly over the course of the 16th century, although the population of the kaza generally seems to have risen (Tosunoğlu 1984: 153, 333). This may reflect a decrease in taxes such as the bennak levied on non-Muslims.
Fig. 13.3: Pipe heads and fragments. Produced by Lorraine McEwan.
382
Cide and its Region from Seljuk to Ottoman Times
The 17th and 18th centuries represent a particularly obscure period in our knowledge of Kastamonu sancak. From the late 16th century onwards Kastamonu was ravaged by the Celali rebellions which resulted in widespread brigandage and chaos in the countryside that the authorities struggled to suppress (Kütüklü and Tunoğlu 2012: 52-7). The Ottoman empire was becoming increasingly decentralised, and from 1735 Kastamonu sancak was generally assigned as a lifetime tax farm (malikâne) to high ranking military figures. However, because of the lack of central control, there were often several claimants at once (Ergene 2002: 14). Kastamonu sancak main contribution to the empire, however, was its timber industry, which was essential to the Ottoman navy (Kütüklü and Tunoğlu 2012: 43-9; also section 3.2). Indeed, as early as the 13th century, the Moroccan geographer Ibn Sa‘id had recorded how timber was cut down from the mountains northwest of Amasya for use in the Seljuk shipyard there (Cahen 1968: 49). Sinop’s shipyard (tersane) rose to renewed importance after the Ottoman conquest in the 15th century, and was the empire’s third largest tersane after those at Istanbul and Gallipoli (Ünal 2008: 348). Cide and nearby Gideros to its west feature regularly in archival documents requesting the timber for use in shipbuilding. The wood was used not just in Sinop but also in Istanbul (Ünal 2008: 336, n. 1024). Given the military importance of procuring these vital supplies of wood, the instructions from Istanbul often go into considerable detail as to how and from where they should be obtained (further on the infrastructure supporting shipbuilding see Ünal 2008: 347-402). For instance, an order dated 1728 requests wood and rope to be sent from Cide and Gideros for use in the Istanbul shipyard (BOA C.BH. 88/4228), while in 1782 the people of Cide were ordered to collaborate with those of Azdavay kaza to provide the necessary timber (BOA, C.BH. 105.5051). A document of 1812 records an instruction for buffalo in Hoşalay kaza to be used to transport wood from Gideros and surroundings to the harbour for transport by sea to the tersane (BOA, C.BH.111/5394). Clearly the importance of Cide and Gideros for the naval timber industry lay in the fact that they had ample forests on the mountains behind the town, along with easy access to the sea, and were comparatively close to all three of the main Ottoman shipyards. The government’s interest did not extend far beyond shipbuilding. In 1775, the Kadı of Cide complained of the oppression and misdeeds of an important Hoşalay family, the Derbederoğulları (BOA, AE.SABH.I.73/5075), while a group of people from Cide fled local misgoverance for the Crimea in 1799 (BOA, AE.SSLM.III.359/20580). As in much of the Black Sea region, it seems that power lay in the hands of local notables (âyân). From the 19th century, however, with the institution of the Tanzimat reforms, Istanbul took a rather greater interest in the region. At some point during this period Cide was established as a separate kaza in its own right, but the archival documents indicating when this happened have not yet come to light. The census data for AD 1831 indicates that Cide’s population was some 4455, out of a total of 120,123 for Kastamonu sancak (Kütüklü and Tunoğlu 2012: 42), making it one of the larger kazas: neighbouring
Cide and the Kastamonu Region in the Ottoman Period
A
B
C
D
383
Fig. 13.4: Ottoman-period gravestones from Kumluca (A: CAP-P1/5047 and B: CAP-P1/5049), (C) Çayyakka (CAP-P2/749) and (D) Gideros East (CAP-P2/1827).
Hoşalay had a population of 6586, that of Sinop was 7137, while İnebolu was only 2822. The only kaza in a different league was Kastamonu itself, with a population of 14,861. From the mid-19th century detailed information survives for the social and economic history of Kastamonu province, and with it Cide, in the form of temettuat defters for 1260-61 AH /AD 1844-46 which give details of the name, occupation, lands and incomes of each household in each kaza (Kütüklü and Tunoğlu 2012: 62, 71). This information
384
Cide and its Region from Seljuk to Ottoman Times
has not yet been investigated by researchers, and it is beyond the scope of a short piece such as this to do so. It must, however, represent a major desideratum for future work on 19th century Kastamonu and Cide to concentrate on this resource.
Fig. 13.5: Central pillar of the old bridge over the Devrekani near Çamdibi, Loç valley (CAP-P2/1657).
A further administrative reorganisation took place in AD 1846 when Kastamonu was made into a province (vilayet) in its own right. Cide became one of the eight kazas that comprised the sancak of Kastamonu within the vilayet of the same name. The other kazas in the sancak were Kastamonu town, İnebolu, Safranbolu, Taşköprü, Araç and Daday, while the vilayet also comprised the sancaks of Bolu, Sinop and Kengiri (Çankırı) (Kütüklü and Tunoğlu 2012: 129-30). Hoşalay (Doğanyurt) now became a sub-district (nâhiye) of Cide kaza, suggesting Cide’s growing importance in a regional context. However, it was soon eclipsed by the rise of İnebolu from the mid-19th century, which became the main commercial centre on the Kastamonu coastline, with its own customs department (Gümrük Müdürlüğü) to administer its exports of agricultural products, foodstuffs and timber. İnebolu also received a large volume of imports, especially of textiles and cotton, and did business with the UK, France, Italy, AustriaHungary, Germany, Greece and Russia (Kütüklü and Tunoğlu 2012: 166-7, 351-3). Mining was another industry. Kastamonu had been known for its copper mines to the Ottomans since the 16th century (Kütüklü and Tunoğlu 2012: 174), and the nâhiye of Küre-i Nuhas, the copper mining region, was for a while administratively attached to Cide (BOA, DH.MKT.1578/41). Copper mines near Cide were still being operated into the 18th century (BOA, C.DRB.42/2077), but subsequently most of the references in archival sources that
Cide and the Kastamonu Region in the Ottoman Period
385
have come to light to date are not to copper, but to coal mining. Cide’s coal mining potential started to attract attention from the late 19th century. In 1874, a concession was granted to a certain Sergiz Bey for the coal mines in Bartın and Cide (BOA, I.MMS.49/2124). Subsequently, new mining concessions around Cide were granted to other individuals and companies, such as one at Kurucaşile, which was in the hands of a certain Krikor Sinabyan, an Armenian, until his shares were turned over to a state company, the Ticâriye ve Sınâiye ve Mâliye Osmanlı Anonim Şirketi in 1920 (BOA, MV.220/5). The provincial salnâme (yearbook) for 1902-3 offers the following description of Cide (Kütüklü and Tunoğlu 2012: 151-2, 160): “The Kaza of Cide The town (kasaba) of Cide is to the north west, at a distance of 24 hours and 20 minutes on the coast. To the east it borders İnebolu, to the west Amasra and Safranbolu nahiyes which are attached to Bartın, and to the east Daday kaza. Its climate is pleasant, and because the kaza centre is suitable for every kind of civilization and development, it has been decided that the road being constructed to Daday should extent to Cide. Cide kaza consists of 163 villages, 33 large and small mosques (cami ve mescid), one tekye (Sufi retreat), 4 hans (inns/caravanserais), 18 coffee shops (kahvehâne) one public bath (hamam), 12 ovens, 15 large shops, 7 large barns (berhâne), 9378 haylofts and granaries, 53 primary schools (sibyan mektebi), 110 small shops (dükkân), 8 police stations, 286 mills, 20 water sawmills, 21 factories, 26 drinking water fountains. Cide kaza’s land is estimated to comprise 425,246 dönüm2 of cultivated land, 67,799 dönüm of uncultivated land, and 8082 dönüm of fruit and vegetable gardens. Agricultural products: Wheat, barley, sweetcorn, millet, linen, chestnuts, apples etc. Industrial products and commerce: Things like timber, boxwood, boats and rowing boats are produced, exported and sold.”
To put these figures in comparative perspective: Cide emerges as undeveloped, even by the standards of Kastamonu, itself somewhat remote and backward, as it was distant from the main west-east railway line. Every kaza except Cide and Araç had at least one medrese (school); several, such as İnebolu, Taşköprü and even Tosya had a library, but not Cide, Araç or Daday. While İnebolu’s population was only about 50% bigger than Cide’s, it had 10 hamams as compared to Cide’s one, and 145 mosques compared to Cide’s 33, and 716 shops and cafes compared to Cide’s total of 143 (Kütüklü and Tunoğlu 2012: 167). The population of Cide by 1902-3, as attested by the salnâme, had risen to a total of 41,502, who were almost entirely Muslim: a mere 20 men and 18 women are listed as Rum (Greek-speaking Christians) in the 1902-3 salnâme’s population figures, and none as Armenian. This, however, was reasonably representative of rural areas in Kastamonu sancak: only Kastamonu town, İnebolu and Safranbolu had significant Christian populations of the order of 2-3000 each (around 4.5%, of whom the vast majority were Greek; however, small numbers of Armenians are recorded everywhere
2 1 dönüm = 1000 m2.
386
Cide and its Region from Seljuk to Ottoman Times
Fig. 13.6: Çilekçe Köprüsü (CAP-P2/1371).
but in Cide and Safranbolu). Compared to the other constituent sancaks of the vilayet (Bolu, Çankırı and Sinop), Kastamonu had a much larger proportion of Muslim inhabitants (Kütüklü and Tunoğlu 2012: 166-7, 339-40). Doubtless Cide’s remoteness due to the lack of a paved road was the main factor in its lack of development (also chapter 2). Efforts to build the road were not helped by embezzlement on the part of the officials responsible, for which the kaymakam and others were prosecuted in 1906, and the plan was abandoned, despite the pleas of the people of Cide (BOA, DH.MKT.1051/61, DH.MKT.2765/41). Cide’s relatively remote location also encouraged various forms of lawlessness, especially the Laz tobacco smugglers who openly did business in Cide (BOA, DH.MKT 1432/12, DH.MKT1466.88). Indeed, combatting smuggling was the main reason for the plan to construct a telegraph line from İnebolu to Cide in 1890, although the plan was long postponed for the lack of funds (BOA, DH.MKT.1799/4; DH.MKT.1450/97). Various brigands such as the Laz Tufan Reis were also active in the area (BOA, DH.MKT.267/41, DH.MKT.2629/39). Again there is a certain suspicion that the local authorities may have been complicit, for an investigation was started against the kaymakam Tevfik Efendi for turning a blind eye to some individuals’ possession of illegal weapons (BOA, DH.MKT.2478/92). It is perhaps not surprising that many of the archival documents in Istanbul from this period discuss the building of two prisons (men’s and women’s) in Cide; but the town was also furnished with some new amenities such as a government house (Hükümet Konağı) (BOA, DH.MB.HPS 16/66; DH.MB. HPS/17/62, DH.MKT.988/30).
Cide and the Kastamonu Region in the Ottoman Period
387
Fig. 13.7: İlyasbey water installation (CAP-P15/289).
Cide’s small Christian population was growing at the beginning of the 20th century, although a request to be allowed to build a church was rejected by the authorities in 1906 on the grounds that the place chosen was unsuitable (BOA, DH.SFR.396.110), and permission was only finally granted in 1913 (BOA, İ.AZN.111/25). The new attention Cide’s Christians had brought themselves was to redound to their disadvantage: at the beginning of the First World War, fear that they were in secret treasonous contact with the Russians led to an instruction for Ottoman Christian schools and churches to be searched for secret telegraph lines as was also done in Zonguldak, Ereğli, İnebolu and Sinop (BOA, DH.EUM.6Sb 2/49). Cide was bombarded by Russian ships on several occasions, in 1915, 1917 and 1918 (BOA, DH.EUM.5Şb.80/50; DH.İ.UM.EK107/13; DH.İ.UM 21-2/32). The war affected it in other ways too, with refugees from Russianoccupied Trabzon seeking refuge in Cide (BOA, DH.ŞFR 630/91), and nearly 500 men from Cide were killed in the succession of wars that accompanied the collapse of the Ottoman empire, stretching from the Italian invasion of Libya in 1911 to the establishment of the Turkish republic in 1923 (Kütüklü and Tunoğlu 2012: 272-80).
388
Cide and its Region from Seljuk to Ottoman Times
Fig. 13.8: Traditional stone-built bread oven, Çamdibi (CAP-P2/1640).
Fig. 13.9: Recorded finds locations of Seljuk and Ottoman date in the CAP survey area. Produced by Toby C. Wilkinson.
The Material Remains of the Ottoman Period in Cide
389
13.3 The Material Remains of the Ottoman Period in Cide
(Claudia Glatz and T. Emre Şerifoğlu)
Cide’s architectural and archaeological records for this period are rich and diverse. They include the traditional wooden houses, many still occupied (Fig. 13.2) as well as in varying stages of decay. The systematic recording of the remains of the region’s more recent past was beyond the Cide Archaeological Project’s scope, but it is certainly an important undertaking for the future in a region that has seen a recent decline in population as its younger inhabitants move to Istanbul or abroad (also chapter 2) and those who have remained construct new homes using concrete and bricks often right next to their abandoned traditional houses. Archaeological evidence for the Ottoman period, which we encountered both during intensive fieldwalking and were shown by local residents, includes a small collection of 17th and 18th century pipe heads and fragments from the Cide hinterland (F89), a local resident’s private collection at Irmak (S 146), a garden plot in Kumluca (F43), a location near Sofular (F44) and inland Aybasan (F48, F50) (Fig. 13.3 and 13.9). Ottoman gravestones are another category of evidence. At least two 18th century Ottoman graves were recorded near Kumluca amidst a forest, which included one inscribed slab and one decorated with flower and ribbon motifs (Fig. 13.4). Two further gravestones came from the northern edge of the mostly modern Kumluca cemetery (S166). In terms of their styles, the Kumluca gravestones seem to belong to the end of the 18th and 19th century, and one gravestone, whose inscription dates it to 1805, supports this view. Other Ottoman-period gravestones dating from the 19th and early 20th century were recorded in a cemetery at Çayyakka (S87). The earliest Çayyakka gravestone dates to 1807 and the latest one is from 1911. The cemetery also has a gravestone, which imitates Ottoman examples, but dates to 1931. An Ottoman tombstone from 1830 also stands at the western tip of the eastern Gideros promontory together with an upright cannon. In general terms, the gravestones from Cide, which follow imperial Ottoman styles (Laquer 1993; Güvelioğlu 2008; Sevim 2010), imply that the area became more attached to the Ottoman central authority with the beginning of the 19th century. Archaeological evidence for efforts to make the region more accessible include two substantial, probably late Ottoman, stone-built bridges which were in use until recently. They include a bridge across the Devrekani in the Loç valley, of which only a central oval pillar remains (S111) (Fig. 13.5) and a second, similar but much better preserved bridge in the easternmost part of Cide district near the village of Çilekçe (S103) (Fig. 13.6). The latter bridge is some 40 to 50 metres across and has two large arrow-shaped bridge pillars, as well as two bridgeheads. These probably supported a wooden bridge construction. Also dating to the Ottoman and post-Ottoman period and still partly in use is an aqueduct and mill in the Okçular valley (S105) attesting to the long-term importance of this valley for agricultural production (also chapters 11 and 14). The remains of a
390
Cide and its Region from Seljuk to Ottoman Times
watermill and associated installations were also recorded at İlyasbey (S186) (Fig. 13.7) and similar mills, still currently in use, were encountered across the region, as were traditional stone-built bread-ovens (Fig. 13.8).
13.4 Conclusion As indicated at the beginning, much more detailed work is required to establish Cide’s place in regional history. Nonetheless, it seems clear that Cide’s obscurity in the written record reflects its marginal importance. Until the early 20th century and the construction of the paved road, Cide was a remote and impoverished region, sparsely inhabited and to some degree isolated even from the interior heartland of Kastamonu province. Nonetheless, as the evidence of bridge building discussed above suggests, especially from the 19th century, efforts were made to integrate Cide more fully into the road network and hence the regional economy. However, as the continuing occurrence of smuggling and lawlessness suggests, such efforts did not fully bear fruit until republican times. Acknowledgement: The research leading to these results has received funding from the European Research Council under the European Union’s Seventh Framework Programme (FP/2007-2013) / ERC Grant Agreement no. 208476, “The Islamisation of Anatolia, c. 1100-1500”.
Catalogue of Illustrated Artefacts F43 Mouth fragment of a decorated ceramic pipe. Red fabric. Fragment length = 28mm. Outer D= 20mm and 15mm, inner D= 7mm. F44 Small fragment of decorated ceramic pipe. Decorative flower motif. Red fabric. Fragment width = 15mm, wall thickness = 5mm. F48 Small fragment of ceramic pipe head. Buff fabric. Fragment length = 34mm, width = 14mm, wall thickness = 4mm. F50 Small fragment of ceramic pipe head with impressed decoration. Fragment length = 22mm, width = 19mm, wall thickness = 4mm. F89 Fragment of a ceramic pipe head with impressed/stamped decoration. Orange fabric (2.5YR 6/8). Fragment size = 36mm. D= 25mm.
Bibliography archival documents BOA = Başbakanlık Osmanlı Arşivi, Istanbul.
Bibliography
391
published sources Başbakanlık Osmanlı Arşivi 1994. 438 Numaralı Muhasebe-i Vilayet-i Anadolu Defteri (937/1530), vol. II: Bolu, Kastamonu, Kengeri ve Koca-ili Livaları (Dizin ve Tıpkıbasım). Ankara, BOA Yayınları. Bryer, A.M. 1975. Greeks and Türkmens: The Pontic exception. Dumbarton Oaks Papers 29: 113-48. [reprinted in Bryer,A.M. 1980. The Empire of Trebizond and the Pontos. London, Variorum]. Cahen, C. 1968. Ibn Sa‘idsur l’Asie Mineure Seljuqide. Dilve Tarih Coğrafya Fakültesi Tarih Araştırmaları Dergisi 6/x-xi: 41-50. [reprinted in Cahen, C. 1974. Turcobyzantina et Oriens Christianus. London, Variorum]. — 1971. Questions d’histoire de la province de Kastamonou au XIIIe siècle. Selçuklu Araştırmaları Dergisi 3: 145-58. [reprinted in Cahen, C. 1974. Turcobyzantina et Oriens Christianus. London, Variorum]. Ergene, B. 2003. Local Court, Provincial Society and Justice in the Ottoman Empire: Legal Practice and Dispute Resolution in Çankırı and Kastamonu. Leiden, Brill. Gökoğlu, A. 1952. Paphlagonia. Kastamonu, Doğrusöz. Güvelioğlu, İ.G. 2008. Osmanlı Mezar Taşları. Istanbul, Türkiye Anıtlar Derneği Yayınları. Heywood, C. 1960-2007. Kasṭamūnī. Encyclopaedia of Islam. [2nd edition]. Leiden, Brill. Khuyi, Husam al-Din Hasan b. ‘Abd al-Mu’min. 2000. Majmu‘a-yi Athar-i Husam al-Din Khuyi. Tehran, Mirath-i Maktub. Ibn Bibi 1956. El-Evamirü’l-‘Ala’iyye fi ’l-Umûri’-‘Ala’iyye. [facsimile prepared by Adnan Sadık Erzi]. Ankara, Türk Tarih Kurumu. Kopraman, K.Y. 1988. Memluk Kaynaklarına Göre XV. Yüzyılda Kastamonu ve Cevresi. In Türk Tarihinde ve Kültüründe Kastamonu: Tebliğler. Kastamonu, Kastamonu Eğitim Yüksekokulu Koruma Derneği: 17-33. Korobeinikov, D. 2004. The revolt in Kastamonu, c. 1291-1293. Byzantinische Forschungen 28: 87-118. — 2010. How Byzantine were the early Ottomans? Bithynia in c. 1290-1450. In I.V. Zaitsev and S.F. Oroshkova (eds.), Osmanskii Mir i Osmanstika: Sbornik Statei k 100 letiu so dniarozhdeniia S.A. Tveretinovoi. Moscow, Institut Vostokovedeniia RAN: 215-39. Kütüklü, F. and Tunoğlu, K., 2012. Bir Şehrin Hafızası: Osmanlı Arşiv Belgelerinde Kastamonu. Istanbul, Ceren. Laqueur, H.-P. 1993. Osmanische Friedhöfe und Grabsteine in Istanbul. Tübingen, Wasmuth. Lowry, H.W. 2002. Fifteenth-Century Ottoman Realities: Christian Peasant Life on the Aegean Island of Limnos. Istanbul, Eren. Pachymérès, G., 1984. Rélations Historiques. [editor A. Failler, translated by V. Laurent, vol. II: Livres IV-VI]. Paris, Belles Lettres. Pamuk, S. 2000. A Monetary History of the Ottoman Empire. Cambridge, Cambridge University Press. Peacock, A.C.S. 2010. Sinop: A frontier city in Seljuq and Mongol Anatolia. Ancient Civilizations from Scythia to Siberia 16: 103-24. [reprinted in Kassab Tezgör, D. 2012. Sinope: The Results of Fifteen Years of Research. Leiden / Boston, Brill: 103-24]. Sevim, N. 2010. Medeniyetimizin Sessiz Tanıkları (Eyüp Sultan’da Osmanlı Mezar Taşlarıve Edebi Eyüp Sultan’lar). Istanbul, Kitap Dostu. Tosunoğlu, A. 1984. XVI. Yuzyılda Kastamonu Sancağı. Unpublished PhD thesis, Istanbul University. Ünal, M.A. 2008. Osmanlı Devrinde Sinop. XV. Yuzyıldan XVIII. Yuzyıla Sinop Kazası. Isparta, Fakülte Kitabevi. Yakupoğlu, C. 2010. II. Bayezid’inoğlu Şehzade Mahmud’un Hayatıve Faaliyetleri. ZKU Sosyal Bilimler Dergisi 6(12): 319-39. Yücel, Y. 1988. Anadolu Beylikleri Hakkında Araştırmalar. Ankara, Türk Tarih Kurumu.
Claudia Glatz and Bleda S. Düring, with contributions by Marica Cassis
14 A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective At the heart of landscape archaeology lies a fascinating disjunction between how we experience the landscapes that we explore as archaeologists and what we record and ultimately write about them. We experience these landscapes by being in and moving through them. We are moved by stunning vistas and take in new, and - as field seasons progress - more familiar smells, sounds, and events. We experience the physical exhaustion that comes with scaling mountains and the exhilarating feelings of arrival. We become familiar with wind and weather patterns, and before then occasionally the victims of sudden changes in the weather. We associate and create memories of these experiences while interacting with our colleagues and students and with the people we encounter in their fields, gardens and pastures. Our archaeological minds are fascinated by how features and artefacts are situated in particular locales and wonder about the significance of these places in their past social and cultural landscapes. Yet, when we sit at our desks a few years on and write about our archaeological findings, much of this landscape context that made them so interesting in the first place, as well as our own memories of these places are either lost or do not fit easily within the conventions of archaeological data presentation. It is in this context that we should place the interpretive move of survey archaeology from observation and map-making towards attempts to understand and interpret past human existence, by foregrounding how people relate to their landscapes and rework the material residues from previous generations. This is what constitutes the core of the landscape discourse (e.g. Bender 1993) and landscape archaeology (Knapp and Ashmore 1999; Given 2013), both of which have had only limited impact in the Near East in general, and Turkey in particular. In this chapter, we want to investigate the Cide landscapes not as a spatial backdrop to human settlement and activity, but instead as continuously transformed through human interaction with and perception of it. By bringing together the sites and distribution patterns which we recorded in the course of CAP, relevant historical sources and environmental data, we begin to weave a narrative, albeit one with significant gaps, of the long-term development and transformation of three Cide landscapes and the communities that lived, worked and died in them, travelled through them, exerted control over them, and appreciated and harnessed the symbolic significance of particular places within them. While mindful of, and explicit about, the possibilities and restrictions of our source material, in this chapter we want to make room for hypothesis and speculation, analogy, broad-brush comparison and interpretive synthesis in order to paint a credible picture of life in some places and some times in the Cide region that goes beyond the description of artefact types and clusters.
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
393
The three landscapes which we will concentrate on in this chapter are the coastal hinterland of Okçular valley, the harbour of Gideros and its hinterland around Abdulkadir, and the inland Çamdibi/Loç valley (Fig. 14.1). In all three cases, targeted reconnaissance first led us to locally well-known, and in the case of Gideros and Okçular previously published, harbour, fortress and cave sites. Subsequent intensive survey, and at Okçular additional grid-square collections, were employed to gain an in-depth understanding of the landscape contexts of the spectacular cultural and natural features that first drew us to these areas (see chapter 4 for a discussion of field methodology and terminology). We begin this discussion with the valley of Okçular, which is the most intensively explored part of the Cide region, the landscape with the most diverse archaeological finds and the longest and most complete sequence of human occupation in the survey region. Next, we will discuss the region to the west of Okçular, where the Abdulkadir valley connects the natural harbour of Gideros with the coastal hinterland and where a particularly rich surface record allows insights into the region’s Roman and Byzantine past. The final example will take us inland to Çamdibi, where punctuated finds shed light onto aspects of the region’s Prehistoric and Bronze Age communities on the one hand and a Roman and Byzantine rural landscape on the other.
0
5
10 Km
Gideros-Abdulkadir
Okçular Çamdibi / Loç Vadisi
Fig. 14.1: Map showing the CAP survey area and the three landscapes discussed in this chapter (1) Okçular, (2) Abdulkadir-Gideros, and (3) Çamdibi. Produced by Toby C. Wilkinson.
394
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
14.1 Okçular 14.1.1 The Okçular Landscape The limestone outcrop of Okçular Kale (S22) is one of the most iconic natural features of the Cide region (Fig. 14.2). Located on a flat ridge, Okçular Kale stands out from its surrounding landscape of gentle green hills due to the shape and colour of its sharply rising limestone cliffs, which form a natural fortress. Located high above the coastal plain, it commands a view of almost the entire coastline from Cide harbour in the east to Kalafat in the west as well as inland. Moreover, one of the main pre-modern roads winds by Okçular Kale towards the Çamdibi valley (section 14.3), where the remnants of a Roman period bridgehead attest to the importance of this route in the past. The valley of Okçular is heavily eroded today, creating the impression of a badland area of little agriculture potential. The density and diversity of field boundaries which criss-cross the valley, however, suggest that this landscape degradation may be the result of relatively recent neglect and diminishing agricultural activities. A scatter of Late Roman and Byzantine tile and pottery clusters across the valley appear to corroborate this assessment, as do the result of chemical soil analysis (section 16.4). Finally, Okçular Kale encompasses at least one large cave (S24, Fig. 5.11, we were told of others, but only one was located), which appears to have been in repeated use from the Chalcolithic to the Byzantine period. Accessing the cave today is challenging and
Fig. 14.2: A view of Okçular Kale from the interior (S22) (CAP-P1/157).
Okçular
395
involves crawling through a narrow and several metres-long passage, followed by a steep descent onto the current cave floor (Fig. 14.3). Entering Okçular Kale İni, thus, has liminal characteristics that are often associated with caves (e.g. Mlekuž 2012).
Fig. 14.3: A view from the modern cave floor of the narrow, elevated entrance passage of Okçular Kale İni (CAP-P1/400).
Thus, Okçular Kale may be understood as a ‘natural place’ sensu Bradley (2000: 13); an unusual natural feature in the landscape which by virtue of its iconic characteristics becomes imbued with symbolic significance and the focus of human activity. Okçular Kale, however, was not solely attractive at a cognitive level, but for very practical reasons such as its inherent defensibility.
14.1.2 Research at Okçular Okçular Kale and assorted spolia in the nearby modern village have received mentions and cursory descriptions alongside other fortress sites in the Cide region by travellers and epigraphers interested in the Classical and Byzantine past of the region (Kalinka 1933: 48; Gökoğlu 1952: 166; Belke 1996: 255). Prior to CAP, however, no systematic archaeological investigation had been carried out at the kale or its surrounding landscape.
396
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
Okçular valley is the most intensively explored part of the CAP survey area (Fig. 14.4). We worked at Okçular during all three field seasons and many aspects of our field methodology were developed in response to the questions and challenges posed by this landscape and its rich and complex archaeological record.
0
250
500 m
Viewshed Visible from kale
S515
Not visible from kale
Okçular Kale (line of walls)
S64 S22 S24 S61
S516
S514 S77 S81
S513 S80 S83
S512
Cluster – Ceramic cluster
S105 S131
Spot – Isolated find(s) Architectural – Kale (=castle) Cave – Occupied cave Architectural – Building / watermill Architectural – Rock-cut feature / wall Spot – Spoilium
Fig. 14.4: Main archaeological features and clusters in the Okçular valley. Produced by Toby C. Wilkinson.
We were first brought to the imposing limestone outcrop of Okçular Kale by Recai Yılmaz in August 2009 and subsequently explored and recorded as much as possible of the architectural remains on its flat top amidst a dense vegetation cover. The same day we were also shown the entrance of Okçular Kale İni, which extends inside the rock outcrop, but did not enter it until much later in the season. When we finally returned to the cave, equipped with ropes, hardhats and torches on the last day of fieldwork in 2009, joy and surprise came in equal measure when at the very back of the cave we found a rich deposit of Middle Chalcolithic pottery (see section 5.4.1), the first substantial evidence for a Prehistoric presence in the region. The first days of intensive survey in the Aydos valley and the hinterland of Cide town in 2009 were a challenge due to poor ground visibility and low artefact densities on the one hand, and the unsuitability – as it turned out - of the traditional model of Mediterranean-style fieldwalking on the other. Okçular valley, parts of which are affected by severe erosion and thus offer good ground visibility, provided the first
Okçular
397
trial arena for our new survey methodology (see section 4.3.4). In total, we intensively surveyed 119ha of the Okçular valley. Among the most important results of this approach are low density scatters of both Prehistoric and later lithic implements in the fields surrounding the kale and to its east; the likely remnants of Late Roman and Byzantine farmsteads dotted along the eastern side of the valley, a cluster of storage pits or silos; and a scatter of iron oxides or oxy-hydroxides (iron ore) possibly indicating iron production. The fields immediately below the kale and two additional artefact clusters were investigated in more detail using grid-square collection in 2010 and 2011 (G1, G2, G3, G6 and G8) (see section 4.3.5 and Fig. 4.13). In 2010, a geoarchaeological assessment of the Okçular valley was carried out (section 4.6) and in 2011 soil samples were collected from the area of Grid 6 (G6) for geochemical and magnetic analyses (see sections 4.6 and 16.4). Pottery (section 4.4.1.2) and iron ore (section 4.8) were also exported for chemical and petrographic analyses (section 16.3).
14.1.3 The Anatomy of the Okçular Landscape Okçular valley is located about 1km south of the current coastline, from where the terrain rises in several natural terraces to between 200 and 250m above sea level. The valley stretches ca. 3km into the interior, gradually gaining elevation along its northeastern flank and sloping towards the southwest (Fig. 14.5).
Fig. 14.5: View northeast along the coast to Cide, the rock outcrop of Okçular Kale in mid-distance (CAP-P2/665).
398
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
The underlying geology of the Oçkular valley (and adjacent Abdulkadir) constitutes Lower Cretaceous clastic and carbonate rocks framed by Upper JurassicLower Cretaceous Neritic limestone outcrops, which rise sharply above 400m in some cases. A tectonic fault runs along the valley’s southern perimeter in east-west direction (Aksay et al. 2002) (Fig. 14.6).
Fig. 14.6: Limestone ridges defining the southern boundary of Okçular valley (CAP-P2/920).
The valley’s most distinctive geological feature and focus of human activity, settlement and no doubt local cultural memory since the Early Holocene, is the large, lozenge-shaped limestone outcrop of Okçular Kale, which is located at the northern edge of the valley and rises steeply up to 20m above the surrounding fields. Okçular Kale is one of the prime natural fortifications in the region, its modern Turkish name meaning ‘arrowhead castle’, supposedly because large numbers of arrowheads were found there in the past. From its top, one overlooks all of the western extension of the Cide bay and enjoys an almost 180 degree visibility range of the sea, in addition to good visual access to the valley and mountains to the south (Fig. 14.7). A marked topographic zoning characterises the Okçular landscape. Moving from east to west, sharp limestone ridges give way to relatively flat shoulders and a series of gentle, hilly, southeast-facing slopes that are subject to substantial gully and sheet erosion today, particularly in the northern part of the valley. As a
Okçular
399
result of the accumulation of sediments, the valley bottom is relatively flat. This erosion landscape is one of very few areas in the Cide region with sufficiently low vegetation cover and resulting high ground visibility to make intensive fieldwalking a viable strategy of investigation. The same processes, however, also complicate archaeological research due the down-slope displacement and burial of artefacts and ancient landscapes, affecting in particular the earliest periods of human presence (chapters 2 and 4).
0
5
10 Km
Intervisible with Okçular Kale according to viewshed analysis
Okçular Kale
Fig. 14.7: View shed from the top of Okçukar Kale. Produced by Toby C. Wilkinson.
Vegetation and land-use patterns map onto this topographic zoning. Dense marquis covers much of the lower reaches of the limestone outcrops. Hill shoulders and the gentler slopes show signs of long-term farming, which has been completely abandoned in recent decades. This part of the landscape is dissected by a palimpsest of field boundaries made of stone walls, trees, hedges and a few metal fences. Some of these are still in use, others abandoned and in disrepair. Various batches of mixed open forest, some recently planted, border the fields to the south. In addition to copious annual rainfall (see section 2.1.1), the valley is well watered by several springs and many small streams. We also recorded a series of wells amongst the houses and gardens of Okçular Köy, the modern village that surrounds the kale to the south and southwest.
400
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
14.1.4 A Biography of the Okçular Landscape The material culture collected in the Okçular valley reflects about nine millennia of human presence and interaction with this landscape. As far as our material evidence indicates, this is not a continuous presence but rather takes the form of punctuated phases of activity of varying length and intensity. For some periods, especially parts of Prehistory, this presence has left only scant and ephemeral traces, while in others, such as the Roman and Byzantine periods, our record is comparatively rich, providing relatively detailed insights into Okçular’s farming communities and how they connected with, and defended themselves against, neighbouring and more distant regions and powers. The nature of our evidence, in the form of features, surface scatters or combinations of the two, and the ways in which they were found and recorded differs for each period. Certain locations, nevertheless, emerge as places of long-term, or repeatedly re-emerging, cultural significance in the Okçular landscape.
14.1.4.1 Scatters in the Landscape: The Early Holocene The earliest documented phase of human presence in the Okçular valley dates to the later part of the Early Holocene (seventh and sixth millennia BC, chapter 5 for an in-depth discussion). Our data for this period consists exclusively of lithic artefacts, found thinly scattered throughout the Okçular valley, but with a clear concentration in the fields and gardens surrounding the Okçular Kale rock outcrop (Fig. 5.2). The remainder was found scattered on the eastern valley slopes and eroding downhill from hill-shoulders alongside later threshing sledge flints, Late Roman and Byzantine pottery and tiles. The small amounts of Early Holocene material culture from Okçular limit what we can say about the nature of the individuals or communities these artefacts have come to represent and their engagement with the Okçular landscape. Nonetheless, the cluster of artefacts around the kale seems to mark it out as a central place, which probably attracted repeated use or occupation. The concentration of several cores and debitage in this locality points to local tool production, possibly focused on a nearby cave or abris no longer preserved or visible today; locals reported multiple such caves at Okçular Kale, but we were only able to locate one cave entrance. The south facing hill-shoulder beneath Okçular Kale is both sheltered from sea winds and would have afforded a good view over what was probably a densely forested landscape. These forests would have offered plentiful plant and faunal resources, broadly similar to those attested in the Pontic mountains today, such as red deer, roe deer, and wild boars, as well as bear and wolf (IUCN 2001). Chamois and wild goats, surviving at present in the more densely forested eastern parts of the Turkish Black Sea region, were probably also present. Such game resources would have augmented a vegetarian forest-based diet, including mushrooms, berries, hazelnuts and acorns. Hazelnuts and acorns are of particular interest, as they may have formed
Okçular
401
a major component of Prehistoric diet. Both would have been available in relative abundance and could be stored with ease, and this may be one of the reasons why farming appears to have been taken up rather late in this part of Anatolia. At present, it remains unclear whether the Okçular valley formed a permanent or semi-permanent focus for Early Holocene communities and whether agriculture formed part of local subsistence strategies (section 5.3.5). On balance, we appear to be dealing with small and relatively mobile groups. One factor that points in this direction is the diversity of raw materials in the lithic industries, which suggest acquisition through movement or regional exchange networks. At the same time, the available data shows not only that knapping took place locally, but also that more elaborate tools were produced on poor quality and presumably local flint. This suggests a more prolonged, perhaps seasonal, presence.
14.1.4.2 Rock with a View: The Chalcolithic and Early Bronze Age The next episode of human presence at Okçular documented in our survey record is the Middle Chalcolithic (section 5.4.1). The evidence for this period contrasts with the thin surface scatter of Early Holocene lithics in terms of the type, diversity and quantity of recorded artefacts as well as their finds location and the conclusions these allow about Okçular’s Middle Chalcolithic occupants. With the exception of a handful of non-diagnostic surface finds from the fields below Okçular Kale, all our evidence for the Chalcolithic period in this region comes from the cave of Okçular Kale İni. Here, at the back of the cave, we found large quantities of pottery, lithics and obsidian, charcoal, ash and animal bones in a densely packed, stratified midden deposit, for which radiocarbon dates suggest a date in the first centuries of the fifth millennium BC. The nature of the deposit and its location at the very back of the cave and inside a narrow alcove with low ceiling suggest a refuse dump, with rubbish thrown in from the top of the rock outcrop through a now sealed shaft, rather than in situ occupation debris. The densely packed nature of the deposit as well as the large numbers and types of pottery fragments point to a rather extended period of seemingly domestic activities on top of the Okçular rock outcrop. This type of settlement location is attested at other excavated Anatolian sites dating to the fifth millennium BC, such as Güvercinkayası and Kanlıtaş whose inhabitants practised farming and whose material culture shows signs of craft specialisation (Gülçür and Firat 2005; Türkcan 2010). It is plausible that the occupation of Okçular Kale would have been similar in nature. The pottery assemblage is dominated by medium to large jars and simple open bowls. Shapes and finishing techniques resonate in general terms with other Chalcolithic traditions in central and western Anatolia. The assemblage’s formal repertoire as well as the abundant and exclusive use of mineral temper, including locally occurring marble (section 16.3), cast it as the product of a local cultural and technological tradition. Two obsidian artefacts attest to some form of external contact
402
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
and exchange as far south as the Galatian sources, while the lithic assemblage overall corroborates the hypothesis of domestic debris. During at least part of the fifth millennium BC, Okçular valley, thus, appears to have been home to a small Middle Chalcolithic community, residing seasonally or permanently atop Okçular Kale. The nature of this occupation, as far as our evidence suggests, appears to have been domestic in character and is clearly associated with the many strategic advantages the kale offers with respect to safety, visibility as well as the relative proximity of diverse wild and agricultural resources. By the Chalcolithic period, Okçular Kale’s practical advantages and likely symbolic significance, no doubt would have formed part of local long-term knowledge and memory. At the same time, this choice of settlement location ties our Okçular community in with a more widespread trend of Middle Chalcolithic settlement preferences atop elevated and visually striking natural features in western and central Anatolia (section 5.4.1). The idea that the Okçular community was connected with other, more distant groups, finds some support in the obsidian found at the site, which comes from Yağlar, near modern Kızılcahamam (section 4.4.2.2). A smaller collection of pottery fragments whose fabric and surface treatment would suggest dates in the third millennium BC, comes from a second alcove located immediately to the north of where the Middle Chalolithic midden is located (section 6.2.2; Fig. 14.8). The evidence is less unambiguous for the Early Bronze Age, but a
Fig. 14.8: Okçular Kale İni, entrance area to a second alcove with pottery dating to the third millennium BC (S24) (CAP-P2/1056).
Okçular
403
similar, if seemingly less intensive or long-term, process of accumulation of habitation debris from the top of the kale – and thus, continuity in the use and interaction with both the rock outcrop and its surrounding landscape - may be tentatively proposed. Continuity of local cultural traditions are also evident in the Early Bronze Age pottery assemblage, in particular, the continued exclusive use of crushed mineral temper, more extensively documented at Kılıçlı Mağarası, which we will discuss in detail below.
14.1.4.3 Branching Out: The Second Millennium BC Evidence, if still ephemeral, for the later phases of the Bronze Age come from the fields immediately below the south-side of the kale, which we sampled intensively using 5 x 5m grid-squares in 2010 and from a small surface scatter located on a natural hill ca. 300m to the northeast (Fig. 7.5). This natural hill commands a good view of the kale, its surrounding landscape and much of the coastal plain below (Fig. 4.9). The knoll also directly controls the saddle that the modern road connecting Okçular with the coast has to straddle. There is, thus, a fundamental shift in the spatial distribution of our survey data from the kale, or rather its cave, to the wider landscape. Whether this equates with a shift in past behaviour or is the result of our survey strategy, taphonomy and low preservation of Prehistoric surface pottery cannot be resolved at present.
Fig. 14.9: View northeast onto the Kalafat coastal plain from the natural hill where we found a concentration of Late Bronze Age pottery (CAP-P2/1754).
404
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
The second millennium BC pottery from Okçular includes a Middle Bronze Age cooking-pot fragment from just below the kale, and possibly also a highly localised, heavily mineral tempered handmade pottery tradition, some of whose formal characteristics and firing would suggest a later date than traditionally assigned to handmade wares. Diagnostic pieces from the fields below Okçular Kale include the rim-fragment of a small grooved-rim bowl, a large jar with everted rim, and the flat base of a sizable bowl or jar. A concentration of seemingly wheelmade bowls and medium-sized jar fragments with plain surfaces and formal characteristics as well as manufacturing methods typical of Late Bronze Age north-central Anatolia, the heartland of the Hittite empire, also suggest domestic functions. The types of vessels attested at Okçular point to a 15th century BC occupation. This is the time when Hittite textual sources begin to report about what was to become an enduring conflict with a series of groups collectively referred to as the Kaska, who settled in the mountainous northern fringes of Anatolia. With artefact numbers this small, it is difficult to assess the real significance of our finds and what they imply about the nature of the activities at Okçular, and the identity or identities of those who produced and used them. Preliminary petrographic analysis suggests a local origin, and thus, either a south Anatolian, Hittite, presence at the site or intensive and long-term cultural connections with the south that could result in the transfer of technological traditions and familiarity with southern ceramic repertoires. Several scenarios are possible, from a Hittite military stronghold to something more akin to a trading outpost located at the foot of a native settlement on or near Okçular Kale. We may also only speculate about the subsistence strategies of the second millennium BC community or communities at Okçular, but they would have most likely relied on a combination of small-scale agriculture, pastoralism, and hunted and gathered resources of the Black Sea and the region’s extensive forests. The latter remained unaffected by, or more resilient against, anthropogenic deforestation that began to have severe effects on central Anatolian landscapes by the middle of the second millennium BC. More generally, the evidence for the second millennium BC, ephemeral though it may be, suggests that occupation was no longer centred solely on the rock outcrop of Okçular, but on the fields below where the modern village is located, and which was also the main settlement location in the Roman and Byzantine periods (see below).
14.1.4.4 The First Millennium BC Identifying evidence for human presence in the Okçular valley during the first millennium BC has been rather challenging. The Iron Age in particular has proven difficult to pin down across the survey region (chapter 8), and the small number of possible Iron Age sherds from Okçular, both from the cave and surface finds, is not sufficient to propose even a tentative interpretation of human presence in and interaction with this particular landscape.
Okçular
405
Slightly better represented is the Hellenistic and, due to blurred chronological and cultural boundaries, the early part of the Roman period. The fields around Okçular Kale and the hill-slopes to the southeast produced a small number of Hellenistic finds, including a possible Classical-Hellenistic echinus bowl. A thin scatter of fine table ware sherds and several fragments of imported amphorae that date to the Hellenistic to Early Roman periods surrounds Okçular Kale (chapters 9 and 10). These finds tie Okçular’s inhabitants to regional production centres and distribution networks around Sinop and the northeast Mediterranean. A possible Hellenistic and several Roman tile fragments point to a settlement near the Okçular Kale rock outcrop. The Hellenistic to Early Roman-period community at Okçular, much like later Roman and Byzantine occupants (see below), appear to have tapped into the agricultural resources of this upland valley while retaining connections with regional and supraregional networks of exchange and cultural traditions. The imported amphorae would have contained products such as wine, (olive) oil and non-liquid cargo, such as fishproducts, that were widely consumed throughout the Hellenistic and Roman worlds. Located about 100-150m to the northeast of Okcular Kale and associated with a scatter of pottery dating from the Hellenistic to early Byzantine period, we found a concentration of ferruginous materials (U593). Analysis by SEM-EDAX of polished sections of two fragments shown here revealed iron oxides and/or oxy-hydroxides which would have had the potential of serving as high-grade iron ore. However the presence of arsenic would have resulted in metallic arsenical iron requiring expertise in both its manufacture and use (section 4.8; Fig. 14.10).
Fig. 14.10: Examples of iron ore from U593 (CAP-P1/4650).
406
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
The limestone ridge framing Okçular valley to the northeast, through which the Devrekanı has deeply incised itself, also bears iron and nickel deposits (Aksay et al. 2002). No extraction sites, however, were recorded during the survey. The location of the iron oxides/oxy-hydroxides a raised locale, close to where the terrain falls away steeply towards the coastal plain and where consistent wind would have swept smoke from any furnaces upwards, is ideally placed for a smelting operation. The prominent location would have also made this activity visible from afar - both the valley and the coast - while close by a distinctive smell from the furnaces would have accompanied this activity. Although located at a safe distance from Okçular Kale and the ceramic and tile clusters that seem to designate the location of the Hellenistic, Roman and Byzantine settlement, this hypothetical production site would have been clearly associated with, and controlled by, the kale.
14.1.4.5 A Rural Landscape: The Roman and Byzantine Periods The richest and most diverse data for the past of the Okçular landscape date to the Roman and Byzantine periods. Again, the boundaries of what may be referred to as Roman and Byzantine and what these imply with regards to the cultural and political identities of those producing and using material culture we assign such labels to are necessarily fuzzy today and probably would have been in the past, especially in a place as remote as Cide. With this in mind, we can begin to trace the development of an agricultural landscape, which was tied in complex ways to the imperial fortunes of Rome and Byzantium. As part of empire-wide developments, the Cide region oscillates between phases of particularity and connectivity with the wider world, but remains occupied throughout periods of general crisis, political change and re-organisation, in part due to its marginal position. The meta-narratives for the Cide region during these periods (chapters 10 and 11), include several major settlement transformations. In the absence of both specific textual sources referring to the Cide region in these periods and archaeological excavations, the precise transformations taking place as a result of the area’s incorporation into the Roman world are difficult to gauge. Our survey results, however, suggest a concentration of Roman period settlement and activity primarily along the coast, but also in the immediate coastal hinterland such as in the Okçular valley. As a result of Roman imperial crisis, environmental stress, as well as the demise of maritime trade networks in the sixth/seventh centuries AD, we notice a shift away from urban centres and maritime trade, to which Cide would have contributed shiptimber, boxwood, and possibly also locally produced amphorae and their as yet unknown contents (chapter 10), to an economy increasingly oriented towards the coastal hinterland and its agricultural resources. Following this process of ruralisation, the later Early and Middle Byzantine period sees increasingly the establishment of decentralised power-bases in the hands
Okçular
407
of land-holding elites across large parts of the Byzantine sphere in the ninth and tenth centuries. Also during this period, evidence points to a revival of Black Sea maritime trade, through ceramic imports and a renewed investment in coastal strongholds to secure these connections (section 11.3.3). Further unrest ensued as the region passed from Byzantine rule to the empires of Trebizond, Nicaea and then the Seljuk and Ottoman (chapter 13), while Genoese merchants retained strategic possessions along the Black Sea coast, including at least one in Cide (chapter 12). The majority of our data for the Roman period falls into the Mid-to-Late Roman phases. Table and imported red-slipped wares as well as a range of amphorae types appear to suggest a continuous and prosperous occupation from the Hellenistic and Early Roman periods in the surroundings of Okçular Kale (Fig. 14.11). The cave of Okçular Kale İni (S24), following its use as a rubbish dump in the Chalcolithic and possibly also Early Bronze Age, shows signs of re-use in the Roman period. Roman pottery from the cave consists of primarily cooking wares, suggesting occupation of some sort or storage. Amphora fragments and large numbers of tiles cluster along the eastern edge of the valley. The most intensively investigated concentration is grid G6 (S513), which attests to a primarily Roman-period occupation and suggest the existence of a series of small, dispersed farms along the valley’s eastern hillshoulders.
0
. .. .... . . . . . .
. ... .
.
.. . .. . . .. . . . .
. .. . . .. . .. . . .. . . . .............. . . . . . . .
250
500 m
Pottery distribution (Roman) . .
.
.
Pottery distribution (Byzantine)
. .
Tile distribution (all periods) .
. ... . . . . . .
.. . . . . . .. .. . . . . . .. . . . . . . . . . .
. . . .
.
. . . .
. . .
. .
. . .
Fig. 14.11: Distribution of Roman and Byzantine period surface clusters and features in the Okçular valley. Produced by Toby C. Wilkinson.
408
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
The identification of the G6/S513 cluster as a habitation site rather than cemetery or special purpose site is supported by the results of chemical and magnetic analyses of 48 soil samples taken from across the ceramic cluster, which yielded magnesium and phosphate concentrations consistent with general habitation (section 16.4). Thus, it would seem that recognisable settlement expanded from the long-term focus of human activity and settlement at Okçular Kale and its more immediate surroundings into its agricultural hinterland during the Roman period. The artefact clusters along the eastern valley edge, if we have correctly identified them as farms, would have overlooked their southwest-ward facing fields that gently sloped towards the valley bottom. The farmers of this period would have also begun to divide this part of the Okçular landscape into individual fields, resulting over time in the palimpsest of crisscrossing field boundaries we can observe on the ground as well as from satellite imagery (Fig. 14.12).
Fig. 14.12: Okçular field-boundaries (CAP-P2/917).
At the same time, our hypothetical farmsteads would have had good visual access to neighbouring farms, which would have also been inter-audible. Each would have maintained a visual connection with Okçular Kale, where we propose the main settlement of the valley was located, whose remnants today include a series of spolia and ceramic concentrations (section 11.3.2.1). The date we can assign to two of these architectural features depends on their identification, either as Hellenistic or Roman purpose-made cistern mouths (chapter
Okçular
409
10), or as Late Roman to early Byzantine re-worked column bases (chapter 11). If we follow the latter interpretation, we may tentatively postulate the existence of at least one monumental structure in the vicinity of Okçular Kale, most likely a Late Roman or Early Byzantine basilical church. Such a church, hypothetical in the absence of in situ architectural remains and the possibility of spolia being brought to Okçular from elsewhere, would most likely have been located on the hill-shoulder at the foot of the rock outcrop of Okçular Kale. From there, it would have overlooked the valley and at the same time be prominently visible as well as potentially audible from across the valley by those working the fields and travelling from the interior. Judging from the scatter of Byzantine pottery in the fields and gardens on the southern hills sloping away from the kale, occupation continued from the Roman into Byzantine period, most likely in the form of a small village. The presence of glazed fine-wares, such as Sgraffito ware, on and near the kale attest to the re-establishment of external trade connections, and the relative wealth of at least some Okçular households, which were able to import dining equipment. It also demonstrates a desire on the part of this community to participate in the use and display of this pottery as part of the projection and negotiation of particular social and cultural identities. This settlement would have had an excellent view across the valley. Perched above was Okçular Kale, which was fortified at key points where the rock outcrop could be scaled. During the Byzantine period, the kale in all probability served as a refuge site rather than an inhabited fortress; a model we see repeated at other Byzantine-period fortified sites in the Cide region (section 11.3.3). The first architectural features atop the rock outcrop, which we can date with some confidence, belong to the Early to Middle Byzantine period, but earlier, non-mortared walls may date as early as the Iron Age. Although extensive and rather impressive in places, the remains of the Byzantine mortared fortification walls - as far as could be reconstructed amidst dense vegetation - do not form a continuous perimeter wall. It is possible, that the use of the cave below in this period is connected to the function of the fortification: as a hiding place for emergency supplies, although few storage containers were found inside the cave. The cave of Okçular Kale İni shows signs of use again in the Middle to Late Byzantine period, where, as in the preceding Roman period, cooking equipment in the form of pots and jugs predominates alongside several pieces of glazed table ware. These suggest some form of occupation, possibly as a refuge or hermit cave. A concern with secure storage is also evident in what appear to be roughly contemporary installations along the southern edge of the modern village. This includes a large, almost circular structure (S80), up to 4m in diameter and made of mortared field-stones atop a rocky spur overlooking the flat valley bottom, where we found three additional pits (e.g. S83). Their precise function remains unclear, but a use as water reservoirs or grain storage pits, seems most convincing. The storage bins at the bottom of the rock outcrop are clearly visible from the locations of the farmsteads across the valley.
410
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
Ceramic evidence in the form of both pottery and tile suggest continuity in occupation at these farmsteads throughout the Byzantine and very likely also the Islamic period. Less dense than the pottery scatters and further down-slope, we recorded several concentrations of honey-coloured flints, whose date and function is as yet unclear, but they may have been part of threshing sledges, which were employed throughout the Byzantine world (Decker 2008: 399). To the south of G6/S513 and S512, we found architectural remains, pottery and tile fragments belonging to a further Byzantine structure, very likely another farmstead (S131), or judging by the large numbers of tiles in the vicinity, perhaps a separate village located near a small stream. Here, we also discovered the remnants of a raised channel whose most recent form probably dates to the Ottoman period (S105) (Fig. 14.13), but which includes seemingly older water conduits. Dense clusters of tiles were
Fig. 14.13: Stone-carved water-conduit associated with S105 (CAP-P2/1410).
Okçular
411
also recorded in the very south of the valley, near the saddle at Karabudakar that leads towards the interior, suggesting the presence of yet more small farmsteads. Thus, in contrast to preceding periods where small communities clustered close to Okçular Kale and capitalised on its strategic and practical advantages, the wider Okçular landscape underwent a process of agricultural colonisation, and with it no doubt an intensification of production during the Late Roman to Byzantine periods. At the same time, the village at Okçular takes on the role of a central settlement, complete with church, fortified stronghold and large-scale storage facilities, and as such continues a by now millennia long tradition of Okçular as a place of central social and cultural significance. Imported fine-wares and examples of good-quality stone carving suggest that this Okçular community flourished for some centuries.
14.1.4.6 Beylik and Ottoman Periods at Okçular With the exception of traditional village houses, some no doubt first constructed in the Ottoman period or earlier, a glazed Ottoman sherd from G6/S153 and a small aqueduct and mill in the southern part of the valley, we found no surface material readily datable to the most recent periods of Okçular’s past. This is one of the more surprising results of CAP. At Okçular and throughout the survey region, a relatively abundant Roman and Byzantine surface record stands in stark contrast to a very limited number of post-Byzantine finds. This was especially puzzling, since we expected hard-fired and glazed ceramics to both preserve well on the surface and to be highly visible during walkover survey. It is also clear that a large number of villages were inhabited until the very recent past. It now appears that the paucity of Beylik and Ottoman period pottery may relate to two developments. First, the dense scatter of settlements in the Cide region seems to be a relatively recent development, the result of a substantial population increase during the 19th century when refugees from regions lost by the Ottoman empire settled in the area. The supply of agricultural produce to Istanbul and Zonguldak provided economic opportunities for Cide’s growing number of village communities (section 2.2.2). A scarcely settled region between ca. 1500 and 1800 AD is, therefore, a plausible explanation for the absence of material datable to this period. Second, it seems that in the Beylik and Ottoman periods, settlement changed from dispersed farmsteads to the more nucleated pattern still visible today. As a result, the Beylik and Ottoman period predecessors of modern villages are buried underneath more recent houses or incorporated into modern architecture. It is also possible that wooden and other organic containers replaced pottery during this period, much like ceramic roof tiles could have been superseded by wooden roof shingles. Both would have dramatically reduced the archaeological footprint of these communities. The modern village of Okçular is quite sizable, although the number of its permanent residents has dwindled dramatically in recent decades as the young migrate to Istanbul, Ankara and abroad (section 2.2.2). The village consists of two loose clusters
412
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
of houses: one surrounding Okçular Kale in two, roughly parallel semi-circles to the south and southwest; and a second string of houses near the southeastern tip of the valley. Until recently, much of the local economy was based on agricultural production; today production satisfies household needs. A dramatic recent reduction in investment and care for the fields surrounding the village, in all likelihood, is also the source of the severe soil erosion we have already discussed. Untended run-off channels are the most likely sources of the deep-cut erosion gullies affecting Okçular’s former fields. To sum up, after the population peak of the Middle Byzantine period, in which the Okçular valley was densely settled and farmed, followed a phase of population decrease and a reduction of agricultural activities. This is corroborated by textual sources, which suggest an increasing nomadic presence in the region from the 11th century AD (section 13.1). A 19th century AD demographic peak and agricultural intensification gradually waned during the last decades as a result of outward migration and agricultural disinvestment. The landscape of Okçular clearly bears the marks and scars of this long series of social, economic, and cultural transformations.
14.2 Abdulkadir-Gideros 14.2.1 The Abdulkadir-Gideros Landscape The central focus of human activity in the second set of landscapes that we turn to now is the bay of Gideros. Much like Okçular Kale, Gideros is a distinctive ‘natural place’ of stunning beauty as well as many practical and strategic advantages (Fig. 14.14).
Fig. 14.14: Gideros bay (CAP-P2/1801).
Abdulkadir-Gideros
413
The bay is nearly circular in shape and is flanked on either side by two elongated limestone promontories. Behind it, the Pontic mountains rise steeply in a funnellike manner. Today, two hamlets are perched on either side of the bay, including two restaurants serving hamsı, Black Sea sardines. Unsurprisingly, Gideros bay features prominently in the tourist brochures of the region as a place that blends history and natural splendour in perfect harmony. Gideros is one of the few natural harbours along the rocky shores of the Turkish western Black Sea region and as a result acquired status and significance as a major harbour and settlement throughout history. Gideros is repeatedly mentioned in Greek and later sources and gives its name to the modern town and district of Cide (chapter 3). To the modern observer at least, this seems somewhat at odds, however, with its rather modest size (also Tsetskhladze 2007: 177). With space limited in the bay area, food and other resources required by Gideros’ Hellenistic, Roman and Byzantine communities, whose repeated mention in textual sources may indicate a fairly substantial settlement, would have had to be either procured through maritime trade or produced elsewhere. The nearest and most accessible landscape, suitable both for agricultural production and settlement in Gideros’ immediate surroundings is the gently sloping valley of Abdulkadir, which is connected to the coast by a low saddle and sheltered from direct view. It seems logical, therefore, to treat these two landscapes as interconnected. A church complex, Çadır İni Kilise (S21), which is located on the saddle that connects Gideros and Abdulkadir valley, would seem to support this assumption.
14.2.2 Research in the Abdulkadir-Gideros Landscape The gentle valley stretching southwest of the village of Abdulkadir is located some 2.5km west of Okçular and formed another focus of intensive and targeted survey (Fig. 14.15). The natural harbour of Gideros was investigated during a number of visits, starting in 2009. The protected anchorage of Gideros is one of the few securely identified places in the Cide region due to its mention as Kyteros in a range of GrecoRoman sources (Belke 1996: 245-246; also chapters 9 to 11, this volume). Gideros and the remnants of fortifications on both its western and eastern promontory are recorded in early travellers’ accounts (Kalinka 1933) and surveys of north Anatolia’s Classical and Byzantine past (e.g. Belke 1996). The first feature investigated in the Abdulkadir valley was the cave of Ballıcı İni (S17) on the side of a large limestone outcrop which forms the southwestern boundary of the valley. As was the case with Okçular Kale, Ballıcı İni too was first shown to us by Recai Yılmaz early in our first field season, alongside the cave of Çadır İni (S20) and the associated Late Roman-Byzantine monastic complex and church of Çadir İni Kilise (S21) at the northwestern perimeter of the valley. We intensively surveyed a large proportion of the Abdulkadir valley early in the 2010 season.
414
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
0
250
500 m
S162
Abdulkadir
S16S68 S66
Gideros harbour S65
S121 S123
S504
S503
S12
S69
S76 S19
S18
S78
S505
S20S21
S60 S59
Cluster – Ceramic cluster Spot – Isolated find(s) Architectural – Kale (=castle) Cave – Occupied cave
S17 S72
Cave – Natural Architectural – Rock-cut feature / wall Architectural – Church/chapel
S30 S520
S63
Spot – Spoilia cluster
Fig. 14.15: Map of Abdulkadir and Gideros showing the location of features and clusters. Produced by Toby C. Wilkinson.
14.2.3 The Anatomy of the Abdulkadir-Gideros Landscape Abdulkadir valley is roughly triangular in shape, stretching inland from the harbour of Gideros in an easterly and northeasterly direction. The valley is ca. 100ha in size and slopes gently northwards towards the coast from the higher-ground (ca. 150m above sea level) where the modern village of Abdulkadir is located. The Abdulkadir and Okçular valleys share a similar macro-geological make-up of limestone outcrops framing a gently sloping valley made up of sedimentary rocks (Aksay et al. 2002). In the southwestern edge of Abdulkadir valley an outcrop of marble is currently quarried from Dösementarla Tepe (Fig. 14.16). The Abdulkadir landscape is also affected by similar processes and patterns of erosion and sedimentation as those at play at Okçular. These are both advantageous for the implementation of intensive field survey techniques due the comparatively high ground visibility that is afforded by erosion horizons, as well as detrimental to the survival of evidence from especially the earliest phases of human presence and activity (section 2.3). Today, the northern hill-slopes are almost entirely denuded of vegetation, except for grass and bushes, while the higher-elevation limestone ridges are covered in dense maquis-type shrub. More so than at Okçular, erosion from surrounding hill-slopes has created a flat valley bottom with deep sedimentary pockets, which is marshy in places and today is used to graze horses and water buffalos (Fig. 14.17). Unlike at Okçular,
Abdulkadir-Gideros
415
Fig. 14.16: View east towards Abdulkadir village, modern marble quarry visible in the southwest corner (CAP-P2/655).
Fig. 14.17: Buffalos and horses grazing in the marshy valley bottom at Abdulkadir valley (CAP-P2/553).
416
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
many field-boundaries in the form of planted trees and bushes appear to be still in use today. The southern edge of the valley is defined by the Dösementarla Tepe range, which rises steeply from the valley bottom and is covered in dense leafy shrub, giving way to mixed forests at higher altitudes. The Abdulkadir valley system is well watered by small streams, especially along its south side and has several springs. During intensive survey we also recorded a stone-lined well near the western valley exit (S65). Today, the valley is accessed via three main routes. A dirt track leads east from Gideros and the main Bartın-İnebolu coastal road through the valley ascending towards the modern village. About mid-way, a second track branches off to the south, providing access to the modern marble quarry and the cave of Ballıcı İni, which are located in the more secluded, southern part of this landscape. From Abdulkadir village, a small, winding road leads eastwards towards İsaköy and Okçular, while another road branches off to the south and provides access to the interior. The coastal plain can be reached via a short but relatively steep descend towards Kalafat village.
14.2.4 A Biography of the Abdulkadir-Gideros Landscape As is the case for neighbouring Okçular, we have evidence for about nine millennia of human presence in, and interaction with, the Abdulkadir-Gideros landscape. Unlike at Okçular, however, Prehistoric surface data is much more limited in Abdulkadir-Gideros, and the Bronze Age seems entirely absent. By contrast, Roman and Byzantine surface data is comparatively abundant. There are good reasons to assume that the developments of Okçular and Abdulkadir were broadly similar and that many of the differences are simply the result of taphonomic factors. At the same time, we surveyed both landscapes with similar intensity and identical methods, which could imply that some of these differences are the result of past behaviour rather than solely artefacts of preservation and recovery.
14.2.4.1 The Early Holocene and Chalcolithic Although much more limited, our understanding of the Early Holocene and Chalcolithic in Abdulkadir-Gideros is comparable to that of Okçular. Our earliest evidence for human presence in the Abdulkadir region dates to the seventh or sixth millennium BC and takes the form of a thin scatter of lithic artefacts made of a variety of raw materials and ranging from cores, flakes and blades to a retouched point (Fig. 5.3). Like in Okçular, there is some evidence for in situ knapping, in this case south of Abdulkadir village, where a core and a core trimming element were found. The tools found in the Abdulkadir valley are broadly comparable to those of Okçular. In both valleys we have elaborately retouched points, probably for spears. However, apart from such objects, most plausibly linked to hunting, we have various objects with more domestic functions, such as scrapers, and a burin. Fascinating is a backed knife with a high gloss that suggests that cereal-like plants or reeds were cut with it, although this
Abdulkadir-Gideros
417
must remain hypothetical until further investigations on the microwear can be carried out. It is surprising that we were unable to identify any Palaeolithic or Early Holocene finds in the spacious and accessible cave of Ballıcı İni (S17), which would have provided an ideal shelter. Conceivably such evidence may now lie buried in deep cave deposits. By contrast, however, we did find Chalcolithic sherds in this and other caves in the area. In fact, evidence for the Chalcolithic derives exclusively from cave sites, including ceramics from Ballıcı İni (S17) (Figs. 14.18-19), Çadır İni (S20; Fig. 11.52), and Gideros Mağara I (S59), located mid-way on a steep limestone rock-face to the southwest of Gideros harbour (chapter 11). In each of these caves only a few sherds were found,
Fig. 14.18: Plan of Ballıcı İni cave (S17). Produced by Victor Klinkenberg and Michele Massa.
418
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
usually undiagnostic, and their dating therefore remains tentative. If correct, however, it would suggest that the Okçular community was not isolated, but part of a more densely inhabited landscape in which caves served as shelters for at least some of these groups. Some of these caves may have been used by shepherds, as they were throughout history (Düring 2011), although in the case at hand we lack any evidence to support such a model.
14.2.4.2 Elusive Millennia: The Bronze and Iron Ages From the end of the seventh century BC, Ionian colonists founded a series of new settlements along the Turkish Black Sea coast, preferring locations with good, natural harbours (section 8.1). Gideros would seem a prime candidate and numerous ancient sources do suggest the existence of a settlement at Gideros from the Iron Age (section 3.1). As a result, we expected to find evidence for a substantial Iron Age presence at Abdulkadir-Gideros. Contrary to our expectations, however, our data for the Iron Age in these landscapes are scant (section 8.4.1). A possible grey-ware sherd comes from the vicinity of Abdulkadir village, and some of the earliest fortification walls at Gideros Kale West (S12), which were constructed using large cut blocks of stone and no mortar (Fig. 8.6), may date to the earlier part of the first millennium BC. Likewise, we have no data at all at present for the preceding two millennia of the Bronze Age. This contrasts markedly with Okçular valley, and would seem best explained as the result of taphonomy and visibility rather than true absences.
14.2.4.3 A Harbour and its Hinterland – Hellenistic to Byzantine Periods Evidence for the Hellenistic period is somewhat more abundant than for the Bronze and Iron Ages. A possible pithos rim and amphora base may indicate the presence of a settlement in the hinterland of Abdulkadir, while a strong Hellenistic to Romanperiod presence is attested at Gideros harbour. As elsewhere in the survey region, it is the Roman and Byzantine periods which represent a watershed with regards to the visibility of settlement and other activities in the archaeological record. Although still modest in comparison to other parts of Turkey, the density of surface finds increases dramatically from the Roman period. The harbour of Gideros, which is protected by two limestone promontories (S12, S121), both of which were fortified, formed the long-term focus of activity in this part of the Cide region from at least the Hellenistic period up to the early modern era. From the architectural remains it would seem that it was most extensively fortified and used during the Byzantine period, but two recognisable phases in the construction of the western fortress suggest a Roman or Late Roman predecessor: Roman period masonry is characterised by large, well-faced stones mortared into a smooth-faced structure.
Abdulkadir-Gideros
419
With the exception of the remnants of a tower on the western edge of Gideros West (S12), cisterns already described by Belke (1996), and some smaller, rectangular structures, dense vegetation at the top of the kale made it very difficult to discern a clear picture of its architectural layout. A series of rock-hewn features attest to the importance of ships and maritime connections for both the western and eastern fortress, even if they are impossible to date. These include now-submerged flattened landings (Fig. 14.19), possible holes to tie boats as well as a flight of rock-cut stairs (Fig. 14.20) leading from the waterfront to the top of the kale. Nineteenth century sources mention an islet in front of Gideros harbor, which is no longer visible today. This would suggest that the current sea levels are higher than in the 19th century and possibly during the Roman and Byzantine periods (section 2.3).
Fig. 14.19: Submerged landings at Gideros East (S121) (CAP-P2/1841).
Hellenistic to Early Roman Knidian amphora fragments, Roman-period roofing tiles, imports from Sinop and the use of Phocaean table wares, suggest the presence of both a well-to-do and well-connected Hellenistic to Roman-period community at Gideros. In the Roman period, a coastal road may have also facilitated land-based movement (section 10.3.3) and connected Gideros with contemporary centres in the Cide coastal plain to the east, and communities at Yenice, Gökçekale, where a large Roman-period cemetery was located, and further afield at Amasra and beyond.
420
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
Trade and its protection was no doubt central to the activities of those living at Gideros, whose daily needs were met in part through imported products. However, we can assume that its inhabitants also exploited the local marine resources, as is still the case today.
Fig. 14.20: Rock-cut stairs at Gideros West (S12) (CAP-P2/1794).
Abdulkadir-Gideros
421
As mentioned above, there is little arable land in the immediate vicinity of Gideros and the nearest agricultural land is located in the Abdulkadir valley. Although sparse overall, the types of Hellenistic to Roman period pottery from Abdulkadir echoes that of Gideros. The finds from intensive survey include a Knidian amphora fragment, Late Roman Phocaean table ware and amphorae from Sinop, the Aegean and north-Africa (section 10.3.2, location 3). A thin scatter of Roman period tiles suggests small-scale habitation in the valley. A rock-cut grave (S66) on the northern slopes may be dated to the Mid- to Late Roman period and provides a glimpse of the ritual and religious aspects of this landscape (also chapter 15). The tomb is cut into a large boulder, which is visible from the valley bottom and the low natural rises where Roman and Byzantine tile and ceramics cluster (Fig. 14.21).
Fig. 14.21: Rock-cut grave (S66) in boulder northwest of Abdulkadir village. Concentrations of tile and pottery were found on the natural hill to the south (CAP-P2/603).
422
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
The majority of surface finds from Abdulkadir valley date to the Byzantine period and, with one exception consist of domestic and table wares as well as tile-fragments, which are scattered throughout the valley. This suggests a widespread but extensive pattern of settlement in the form of dispersed farmsteads or, in the case of a few denser clusters, small villages, which appear to have been occupied throughout the Byzantine period. Several clusters are associated with possible Roman-period predecessors, indicating a degree of continuity in rural life against a backdrop of increased settlement density. A concentration of Early to Middle Byzantine pottery fragments (S63) was recorded in the southern-most part of the valley with a spot-find near the crossing of a small stream and just below the cave entrance of Ballici İni. Two further prominent pottery and tile clusters, which we identified during intensive survey, are located on the lower northern slopes in the centre of the valley and just above the flat valley bottom (S405, S505). These provide the clearest evidence for continuity of occupation during the Byzantine period. The steep rock-outcrop of Kurçalı Tepesi (S16/S68) delimits the northern edge of Abdulkadir valley. Here, possible wall stubs and Byzantine tile are scattered on the surface. This location provides a good, general view of the sea and along the coastline, albeit not of Gideros harbour or the western portion of the Cide coastal plain. The site’s function is unclear, but a lookout position seems plausible (Fig. 14.22).
Fig. 14.22: Kurçalı Tepesi (S16) (CAP-P2/657).
Abdulkadir-Gideros
423
The religious and ritual landscapes of Abdulkadir, Gideros and other parts of the Cide region were transformed during the Late Roman to Early Byzantine period by the construction of Christian churches and monasteries. Evidence for a basilical church at or near Gideros includes several spolia (S123) in the form of a rectangular basin, a column fragment, a chancel screen and other pieces of worked stone. Located on the mountain slope behind Gideros bay we recorded a basilical church (S21), a cave (S20) and associated structures, which point towards a monastic complex. Over time the church was modified and reduced in size to accommodate a grave, possibly for a local holy figure (section 11.3.4), but remained in use throughout the Byzantine period. The church was built partially above Çadır İni cave which was incorporated into the church complex. Ledges, sockets and benches are carved into the cave walls and floor. The association between cave and church is no doubt significant. A practical function such as a cool storage is plausible, but a culturally or spiritually significant role may be equally or simultaneously possible. Our survey of the cave yielded a Prehistoric pottery fragment as well as fragments of human bone, which may point towards the long-term cultural significance of this location. The monastic community would probably have been a landowner in the Abdulkadir valley, whose agricultural produce would have supported the institution and freed the majority of its members from daily subsistence chores, allowing more time to be devoted to religious endeavours. Given the size of the complex, the presence of the cave, and the proximity of Gideros Mağarası, we are confident that this was a pilgrimage centre associated with an unidentified holy figure. The monastic community would have held processions and would have participated in the community life of Gideros. The complex was no doubt visible from Gideros, or at least well-known to its inhabitants and a short ascend away from the kales. The inhabitants of Abdulkadir could have reached it within ca. 20 minutes on foot, a trip they would have made regularly on their way to Gideros harbour. Another church (S116) stood on the Kalafat coastal plain, which again would have been relatively easily accessible as well as visible from the high-ground of Abdulkadir village (Fig. 14.23). Another well-known aspect of the Byzantine religious landscape are hermit caves. These include the cave of Ballıcı İni (S17), where a small cache of Middle Byzantine coarse or cooking ware points to some form of habitation, possibly a hermit. By contrast, there is little doubt about the use of Gideros Mağarası (S59) during the Byzantine period, which appears to have served as the habitation of a significant hermit or monastic figure, and was subsequently turned into a religious commemorative shrine with numerous crosses edged into the cave wall alongside several inscriptions (Figs 11.53-6). The majority of the pottery from the cave as well as some tiles thrown up by illegal digging, points to use throughout the Byzantine period. Gideros Mağarası, while located relatively close to both the Çadır İni monastic complex and Gideros harbour, is, unlike Ballıcı İni, at least today very difficult to
424
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
0
250
500 m
.
. .
.. . .
.
. . . . . . . .. . . . . . . . . .
. .
.
.. . .
.
.
. . .
.
. . . . . .
.. . .
Pottery distribution (Roman) Pottery distribution (Byzantine) Tile distribution (all periods)
. .
.
.. . .
.
.
.. . .
Fig. 14.23: Roman and Byzantine features and surface clusters in the Abdulkadir-Gideros region. Produced by Toby C. Wilkinson.
reach, as one of our field teams found out while struggling uphill through dense forest vegetation. At Gideros, there is evidence for a substantial re-building phase on the western promontory (S12), which on the basis of the architectural style of loose, mortared fieldstones similar to that of Okçular Kale and other castle sites in the region, may be ascribed to the Middle Byzantine revival. Two round towers (Belke 1996: 245-6) were also erected around the same time on the eastern side of the harbour (S121). The pottery from Gideros, mostly collected on the western promontory and adjacent fields also points to a strong Middle to Late Byzantine presence at the site, concomitant with the temporary renewal of Byzantine power in the region and revival of maritime trade in the 12th to 14th centuries. We may thus again imagine Gideros as a bustling harbour engaged in the transhipment of timber, boxwood, and agricultural goods produced in its hinterland. The importance of Gideros is clearly attested by the pitched battle over its control that took place between Muslim and Christian forces in AD 1284 (section 13.1).
14.2.4.4 Beylik and Ottoman Periods at Abdulkadir-Gideros As at Okçular, the relatively abundant data for the Byzantine period is followed by a paucity of evidence for the Beylik and Ottoman periods. Reasons for this would appear to be broadly similar to those already discussed for Okçular. As with Okçular,
Çamdibi
425
many of the houses inhabited today in the Gideros and Abdulkadir landscapes and perhaps also installations such as the piers in Gideros bay may originate in the Ottoman or even earlier periods. Yet, we have no distinctive surface material that we can ascribe with confidence to the post-Byzantine past in these two landscapes. At Gideros East (S121), an Ottoman tombstone from 1830 stands at the promontory’s western tip together with an upright cannon. A stone-built house, now a mosque, differs dramatically from surrounding structures and is reminiscent of traditional Greek architecture (Fig. 14.24).
Fig. 14.24: Stone-built architecture at Gideros East. (CAP-P2/1799).
As at Okçular it is during the most recent building phases at Gideros that we see the incorporation of spolia from monumental structures in the local architecture on the one hand, and in the case of Çadır İni Kilise, the ongoing exploitation of earlier structures as quarries for building material on the other.
14.3 Çamdibi 14.3.1 The Çamdibi Landscape The Çamdibi landscape, locally known also as ‘Loç’, is one of the most dramatic landscapes of the Cide region (Fig. 14.16). The deeply incised valley of the Devrekani river is surrounded by spectacular, steep mountains. Located close to it are naturally
426
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
wondrous places such as the Valla canyon, the world’s second largest canyon, and a huge sinkhole measuring 80m across. In the steep limestone cliffs that surround the Çamdibi valley there are many caves, one of the largest and most impressive is Kılıçlı Mağarası (S93/S88), which has yielded a rich corpus of Prehistoric and later ceramics and other finds (chapters 5, 6 and 7). Unlike the Okçular rock outcrop and Gideros harbour, this cave is not a prominent feature in the landscape: its entrances are hidden by dense forest and require local knowledge to locate, much like Uçakdibi Mağarası (S40) located a few kilometres to the west.
Fig. 14.25: View south across part of the Çamdibi/Loç valley (CAP-P2/1433).
14.3.2 Research at Çamdibi The Camdibi area, and in particular the cave of Kılıçlı Mağarası, were first visited in 2010. Kılıçlı Mağarası (S93/S88) is located at the southern edge of the valley, while the Roman period bridgehead of Atköprü (S95) marks its northern limit. One of the objectives for the Çamdibi intensive survey in 2011 was, therefore, to contextualise the cave and the bridge in their surrounding landscape. Although the vegetation cover in 2011 was dense even by Cide standards, we did manage to gain a much fuller understanding of the region’s occupation history. In particular, Çamdibi appears to have formed a focus for Early Bronze Age interaction and Late Roman settlement.
Çamdibi
427
14.3.3 Anatomy of the Çamdibi Landscape The Çamdibi/Loç valley, is located in the southeastern corner of Cide district on the Koca Dere/Devrekani river (both names are used interchangeably in the area), just after it emerges from the large Valla canyon to the south. It is situated at the end of a road coming from Okçular, and it is not clear whether in the past this road would have continued towards Şenpazar and across the steep mountains that flank the valley. The remoteness of the valley, which becomes all the more pronounced during the winter months when even today heavy snow can cut off communications for several weeks, contrasts with the density of recent occupation. The Çamdibi/Loç valley today is dotted with villages and farmsteads, although most appear to have been recently or partially abandoned. Even so, they provide a good proxy for the agricultural potential of this landscape, which is further borne out by evidence for a rather substantial Roman and Byzantine presence.
14.3.4 Biography of the Çamdibi Landscape Our data covers about eight millennia of occupation in the region, but it is highly uneven across this time span and often lacking altogether for many periods. As at Okçular and in Abdulkadir, information for many periods comes from caves only, and casts some doubt on the reliability of observed surface patterns.
14.3.4.1 Chance Finds and Buried Landscapes: The Early Holocene A chance find of a flint knife (L265) and chisel (F56), dug up as part of a foundation pit for an electricity pole (S132) on a flat-topped natural elevation adjacent to the village mosque, constitute the earliest evidence for human presence in the Çamdibi valley. Both can be tentatively dated to the sixth and fifth millennia BC, although later dates up until the third millennium BC are also possible (section 5.3.3). Other than remind us of the dynamic character of this mountainous region and the buried Prehistoric landscapes that remain inaccessible to surface archaeology, the two finds provide us with little further insights about the nature of Prehistoric presence in this landscape, which would have been densely forested at the time. The chisel most likely would have been used for woodworking but does not show any wear, and neither does the flint knife. One can only speculate how these objects ended up where we found them, whether through loss or deliberate deposition. More than anything else, this finds context tells us how little we still know about the Prehistory of Cide-Şenpazar.
428
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
14.3.4.2 Consuming Caves: The Later Chalcolithic and Bronze Ages Aside from the above chance finds, no surface pottery pre-dating the Roman period was collected during intensive surface survey in the Çamdibi area. This closely parallels surface artefact patterns observed elsewhere in the survey area. Again, similar to the landscapes discussed above, Chalcolithic and Bronze Age assemblages from Çamdibi derive from two connected galleries of a large cave, Kılıçlı Mağarası (S88/S93), which is located at the southern tip of a narrow valley east of, and running parallel with, the course of the Devrekani river. The richness and diversity of the Chalcolithic and Bronze Age pottery assemblages from Kılıçlı Mağarası also match that of Okçular Kale İni, but the two main chronological foci of occupation are later and date to the earlier part of the Late Chalcolithic and the later part of the Early Bronze Age. The cave also yielded smaller quantities of second millennium BC, Iron Age and Early Byzantine materials. The assemblage of Kılıçlı Mağarası also differs from Okçular in that it represents primary occupation rather than domestic refuse. Just as the cave at Okçular, Kılıçlı Mağarası too has been subject to extensive illegal excavations, endangering its rich and unique cultural heritage. Kılıçlı Mağarası has two main galleries. The entrance area of the lower gallery (S88) is easily accessible from the valley, but the sheltered space is small and narrow and only a climb over several steep rock faces will allow access deeper into the cave, which appears to be subject to regular flooding. The pottery found in S88 amidst otherwise clean, fine sand, appears to have been washed down through a narrow passage from a second, higher-lying gallery (S93), possibly the result of a sink-hole, where the majority of the pottery was found. This part of the cave is accessed today via a steep climb up the mountain just to the east of the entrance of S88 and followed by a ca. 6m descent onto the current cave floor (Fig. 6.12). It is possible that in the past there were other, less arduous, entrances to the cave. S93 has a large dome-like forehall, which is very light as well as partly exposed to the elements (Fig. 14.26). A smaller chamber at the back of the cave formed the focus of past activity. Diagnostic fragments of handmade simple bowls, one with white painted decoration on the inside, simple and carinated pots and jars alongside numerous body pieces point to a period of use during the early to mid - Late Chalcolithic period and to cultural connections with western Turkey and the central Anatolian plateau (section 5.4.4). Attested vessel types point towards a use of the cave as a storage location, while small quantities of human bone, a bead and spindle whorl suggest a burial context. The main phase of use of Kılıçlı Mağarası dates to the second half of the third millennium BC. For this period, the cave yielded an extensive and strikingly diverse assemblage of handmade, well-fired pottery, often decorated with plastic bands and varyingly coloured burnished exteriors ranging from purplish red to pale pink and yellow. While some vessels show general connections with the
Çamdibi
429
central Anatolian plateau, several others show strong affinities with ceramic traditions from western Turkey and the Balkans. The nature of this assemblage points to short-term storage as well as consumption and display rather than habitation or temporary sheltering. Its evident use from the Chalcolithic period, the majestic beauty of the upper cave gallery with its enormous dome, and the difficulty and a certain degree of danger that is associated with its access, at least today, no doubt imbued this place with symbolic significance and made it an ideal location for socially charged consumption events. During the later part of the third millennium BC, western and central Anatolian communities underwent a series of developments leading to greater social complexity and differentiation.
Fig. 14.26: Large, dome-like forehall of Kılıçlı Mağarası (S93) (CAP-P2/1017).
430
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
The pottery from Kılıçlı is roughly contemporary with these developments. The vessel types and styles attested in the cave suggest that Çamdibi’s Early Bronze Age inhabitants used material culture and participated in behaviours that echo the feasting practices of west and central Anatolian communities such as those of Troy and Alaca Höyük. Petrographic and chemical analyses point to diverse, but probably regional origins, of the vessels from Kılıçlı (section 16.3). This may be the result of exchange or inter-community interaction at the cave, with participants contributing or bringing along their own, locally produced vessels. In the absence of other indications for Chalcolithic or Bronze Age settlement and activity anywhere in the valley, however, it is difficult to fully understand the cave’s social and symbolic significance and its relationship with surrounding communities and the wider landscape. In visual terms, the cave entrance provides a panoramic view northwards all across the valley (Fig. 14.27) and may also be visible, at least to the trained eye, from certain points in the landscape.
Fig. 14.27: View north across the Çamdibi/Loç valley from the present entrance of Kılıçlı Mağarası (S93) (CAP-P2/1610).
Several pieces from Kılıçlı can be ascribed to the transition from the third to the second millennium BC, while a small number of simple rim pieces made of medium-coarse buff fabrics, suggests some activity in the cave during the second millennium BC. As in other parts of the survey region, the Iron Age and Hellenistic periods are difficult to define in the Çamdibi area. Some tentative Iron Age fabrics may be identified among the Kılıçlı assemblage (chapter 8), but no definitive Hellenistic pottery was collected at Çamdibi, which, given its inland location, is not altogether surprising.
Çamdibi
431
14.3.4.3 The Late Roman to Early Byzantine Period The Çamdibi area was most intensively settled during the Late Roman/Early Byzantine period. This is based on the extensive scatter of roofing tiles that litter much of this landscape, near the modern village but reaching up high on the eastern mountain slopes. Again, this was a landscape settled primarily by dispersed farmsteads and villages. Remnants of walls and coloured plaster were found associated with such a tile cluster at S113, confirming their identification as habitation sites. Roughly contemporary extant structures include a beautifully constructed square well (S90), now sheltered by a garden-shed in the northern part of Çamdibi village, which may point towards a long-term agricultural presence in the region, a rather impressive bridge head, and the remnants of a paved road alongside the Devrekanı river at Atköprü (S95). The original construction of the bridge can tentatively be dated to the Late Hellenistic-Early Roman period (chapter 10; but see chapter 11 for a somewhat later date), which identifies this landscape as an important communication corridor with the interior, most likely forming the main connection between the coast and Kastamonu at the time. Spolia, include a sarcophagus with a drainage hole cut at the bottom (S128), a second large stone basin with a carved decoration on the side (S126), and two pillar fragments (S130), and a possible capital fragment (S122) as well as several large dressed blocks of stone are incorporated into the foundations of modern village houses (S124). All of these suggest the presence of one or more rather large-scale and most likely Late Roman to Early Byzantine period structures in the area. Surface pottery is scant but for the Roman period includes two Aegean LRA 2 fragments. Byzantine coarse pottery is notoriously difficult to date, but remarkable is the absence of Mid-to-Late Byzantine glazed wares from the region. The only fine-ware examples come from the nearby cave of Kılıçlı Mağarası (S88, S93), which contained two fragments of Early to Middle Byzantine White Ware alongside several pieces of domestic pottery, which suggest some form of habitation during this period, most likely a hermit. Further to the south along the Devrekani, we also recorded the ruins of Koca Kale (S154), whose architecture suggests a date in the Late Roman to Middle Byzantine periods (section 11.3.3.9). It was clearly positioned strategically along the river and main communication route inland. Two stone-cut features (S138 and S154), reused as water basins in Hamitli village, were recorded in the vicinity of the fortress. To summarise, the overall picture of Roman and Byzantine occupation of the Çamdibi/Loç valley is one of a relatively peaceful and prosperous, densely settled agricultural landscape, with farmsteads scattered across the valley and one or more monumentalised civic or religious structures.
14.3.4.4 Beylik and Ottoman to Modern Periods at Çamdibi Similar to Okçular and Abdulkadir, we found surprisingly little surface material that could be dated to the period between ca. AD 1500-1900. The importance of the
432
A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective
Çamdibi/Loç valley as a thoroughfare to and from the coast in the Medieval and postMedieval period is highlighted by the remnants of a stone-built bridge, which was in use until recently and is located not far from the current bridge over the Devrekani (S111) (Fig. 13.5). A traditional stone-built bread oven in a little square amidst several village houses in Çamdibi attests to domestic and community life during the last century, as do the many traditional houses, now interspersed with more modern constructions (Fig. 13.8). Some of the stone foundations of these houses incorporate Roman dressed stones, while sarcophagi now serve as water basins.
14.4 Concluding Remarks The comparison of the three Cide landscapes leads to several observations. The first is that settlement through time is broadly comparable across all three areas. In all of them, occupation peaks in the Late Roman and Byzantine periods. Settlement patterns consist of isolated farmsteads and some larger villages integrated in diverse cultural landscapes with a zoning of activities. These include monumental churches, fortified sites near population concentrations but on defensible locations, and hermit caves located in the mountains away from agricultural areas. Following the Byzantine period, however, population levels, or alternatively their archaeological footprint, seem to diminish dramatically and this too we can observe in all three landscapes. Population levels recover in the 19th, only to dwindle once more in the 21st century. A second observation is that although periods pre-dating the Hellenistic and Roman eras are ephemeral in all three landscapes, there are clear differences between them. For example, whereas Early Holocene lithic artefacts were recorded during surface survey both at Abdulkadir and Okçular, no such finds were made in Çamdibi. Here the period is attested by a chance discovery near a recent pit. Data for these periods derives almost exclusively from caves, whose investigation by CAP relied heavily on local knowledge. Many more such caves await future discovery and investigation. By contrast, the surface material dating to the second millennium BC from Okçular is at present unique, and it is difficult to assess whether there are other sites in the area with similar characteristics. In any case, they were not found in Abdulkadir and Çamdibi. Finally, our survey results suggest a strong ‘persistence of place’ in the Cide region. Locales such as the Okçular rock outcrop, Gideros harbour, Ballıcı İni, and Kılıçlı Mağarası, attracted communities over the millennia. In each period these places were given different functions and meanings, but long-term memory and past material culture no doubt was incorporated in each era’s cultural landscapes and constituted its sense of place. To conclude then, we are well aware that we have only begun to scratch the surface of these landscapes with our investigations. To gain a more complete picture, much more research is needed, including excavations. By demonstrating
Bibliography
433
the importance, in these dramatic landscapes, of natural places and their re-use and re-interpretation over the millennia, and by highlighting the most deep-reaching historical developments, we feel that we have made a first step along this long, and, in the case of Cide-Şenpazar, steep and rugged path.
Bibliography Aksay, A. Pehlivan, Ş. Gedik, İ., Bilginer, E. Duru, M. Akbaş, B. and Altun, İ. 2002. Türkiye Jeoloji Haritası. Geological Map of Turkey. Zonguldak. Ankara, Maden Tetkik ve Arama Genel Müdürlügü. Belke, K. 1996. Paphlagonien und Honōrias. Vienna, Österreichische Akademie der Wissenschaften. Bender, B. 1993. Introduction, landscape, meaning and action. In B. Bender, ed., Landscape, Politics and Perspectives. Oxford, Berg: 1-17. Bradley, R. 2000. An Archaeology of Natural Places. London, Routledge. Decker, M. 2008. Agriculture and agricultural technology. In E. Jeffreys, H. Haldon, and R. Cormack, eds., The Oxford Handbook of Byzantine Studies. Oxford, Oxford University Press: 397-406. Düring, B.S. 2011. Millenia in the Middle? Reconsidering the Chalcolithic of Asia Minor. In S. Steadman and G. McMahon, eds., Oxford Handbook of Ancient Anatolia, 8000-323 BC. Oxford, Blackwell: 796-812. Given, M. 2013. Commotion, collaboration, conviviality: Mediterranean survey and the interpretation of landscape. Journal of Mediterranean Archaeology 26: 3-36. Gökoğlu, A. 1952. Paphlagonia. Kastamonu, Doğrusöz. Gülçur, S., and Firat, C. 2005. Spatial analysis of Güvercinkayası, A Middle Chalcolithic hilltop settlement in northwestern Cappadocia, A preliminary report. Anatolia Antiqua 13: 41-52. IUCN. 2001. The 2000 IUCN Red List of Threatened Species. Gland/Cambridge, IUCN. Kalinka, E. 1933. Aus Bithynien und Umgebung. Jahreshefte des Österreichischen Archäologischen Institutes 28: 45-112. Knapp, B., and Asmore, W. 1999. Archaeologies of Landscape: Contemporary Perspectives. Malden, Blackwell. Mlekuž, D. 2011. Notes from the underground: Caves and people in the Mesolithic and Neolithic Karst. In K.A. Bergsvik and R. Skeates, eds., Caves in Context. The Cultural Significance of Caves and Rockshelters in Europe. Oxford, Oxbow: 199-211. Tsetskhladze, G. 2007. Greeks and locals in the southern Black Sea littoral: A re-examination. In G. Herman and I. Shatzman, eds., Greeks between East and West. Essays in Greek Literature and History in Memory of David Asheri. Jerusalem, The Israel Academy of Sciences and Humanities: 160-95. Türkcan, A.U. 2010. Kanlıtaş Höyük ve Civarı İnönü, Eskişehir yüzey araştırması. Araştırma Sonuçları Toplantısı 28/1: 303-28.
Claudia Glatz and Bleda S. Düring
15 Conclusions
In the course of three field seasons between 2009 and 2011, the Cide Archaeological Project recorded features, surface clusters and individual artefacts that we can date from the Early Holocene to the region’s Ottoman and sub-recent past. This extends the known human history of the Turkish western Black Sea region by roughly 10,000 years (Table 15.1). Much of this occupation history had not previously been documented in the region’s challenging landscapes (Düring and Glatz 2010) and confirms two of our main working hypotheses. The first is that the absence of evidence for these periods prior to our investigations was a function of limited research. The second premise is that a combination of intensive and targeted survey would allow us to identify the ephemeral assemblages of the relatively small scale early-to-mid Holocene (Düring 2008) and Bronze Age communities, which we set out to find at the start of the project. Tab. 15.1: Periods identified in the course of CAP. Period
Dates
2009
2010
2011
Beylik/Ottoman Byzantine Roman Hellenistic Iron Age Late Bronze Age Middle Bronze Age Early Bronze Age Late Chalcolithic Middle Chalcolithic Early Holocene
AD 1500-1920 AD 650/700 – 1500 AD 1 – 650/700 330 – 1 BC 1200 – 330 BC 1600 – 1200 BC 2000 – 1600 BC 3000 – 2000 BC 4000 – 3000 BC 5500 – 4000 BC 10,000 – 5500 BC
+ + + + + + +/-
+ + + + + + + + + + +/-
+ + + + + + + + + + +
As we have discussed throughout this volume, not all episodes and eras of the region’s past are equally represented in our survey record and for the most part we can only begin to sketch the outlines of social and economic organisation, cultural traditions and the relationships of local communities with surrounding regions. Much of this is due to the ephemeral nature of the surface record, a problem that affects especially the early phases of human occupation. These assemblages are ephemeral because of the substantial taphonomic processes at work in this mountainous and geologically diverse region. Recovery is further hindered by dense vegetation cover and varying degrees of artefact preservation. The paucity of surface assemblages could also be a product of local communities’ social organisation and cultural practices. These include settlement preferences and construction techniques and materials, and the potential use of containers made from perishable materials. In the case of the local
Conclusions
435
Chalcolithic and Bronze Ages, for instance, settlement practices did not result in the formation of mounded sites which are iconic of these periods elsewhere in Turkey and the wider Near East. This made the identification of sites from these periods challenging. The fact that we found evidence for them at all is due to our hybrid and adaptive survey method. This method combines traditional extensive and site-based explorations with hyper-intensive fieldwalking and secondary sampling strategies. This field methodology not only helped to shed light onto the region’s more distant Prehistoric and Bronze Age past, but also significantly enriched our understanding of later periods by contributing a rural perspective on Roman and Byzantine CideŞenpazar. It highlighted the interconnected nature of different types of communities with the ritual, political and economic landscapes, both local and imperial, of these periods. The question of the existence of local cultural traditions, and interwoven with this, issues of connectivity and particularity, forms a second leitmotif in the chapters of this volume. From a practical point of view, localised material culture traditions are, in the absence of comparable, stratified assemblages almost impossible to date. These difficulties almost inevitably bring with them a skewed, external framework of very generic narratives. For the Neolithic and Chalcolithic these revolve around the emergence of farming and sedentism, for the Early Bronze Age/Middle Bronze Age around the emergence of metallurgy and long distance trade. From the Late Bronze Age onwards, we are dealing with imperial or colonial perspectives, for the Iron Age it is the arrival of the Greeks. Such general narratives are often difficult to surmount and replace - or at least temper – with archaeologically informed local datasets. While it is essential to contextualise our results in the broader developments that occur in Anatolia and the Circumpontic, the real challenge is to reconstruct the local trajectories within this wider framework. For example, we can recognise ‘Hellenistic’, ‘Roman’ and ‘Byzantine’ material culture with relative ease in the surface record, but it is much more difficult to understand what happens in this marginal region in these periods. In other words, we have to ask ourselves - even if answers are bound to remain partial, circumstantial or altogether impossible at this stage of research - what it means to live in Cide-Şenpazar during the Hellenistic, Roman or Byzantine periods in cultural terms and what the presence and consumption of ‘Hittite’, ‘Hellenistic’ or ‘Roman’ material culture implies about local communities or factions within them, their aspirations and connections with regions more readily part of these imperial and colonial worlds. In the previous chapter, we discussed three intensively explored archaeological landscapes. In this final discussion, we want to broaden our focus again to encompass all of our survey area and to highlight some of the central themes of past human presence and activity in Cide and Şenpazar documented in our survey data and to explore the region’s changing relationships with the wider world.
436
Conclusions
15.1 Settlement, Subsistence and Production The earliest evidence for human activity in the Cide region dates to the Early Holocene and takes the form of lithic surface scatters and spot finds from locations in the coastal hinterland of Okçular and Abdulkadir and the inland locations of Çamdibi and Aybasan. In the case of Okçular and Abdulkadir, which we discuss in detail in chapters 5 and 14, the lithic assemblages consist of multifunctional implements for every-day use as well as a few large points, most plausibly interpreted as spear tips. This may suggest the hunting of larger animals, real or symbolic, while the gloss on some cutting tools might suggest the harvesting of cereals or reeds. Also present in these assemblages are small numbers of cores suggesting localised production of lithic implements from a variety of raw materials, none of which, however, appear to be local to this part of Cide-Şenpazar. The character and surface distribution of these lithic artefacts highlights both distinct foci of activity, most notably the area surrounding the Okçular Kale (S22) limestone outcrop, but also movement across and engagement with the wider landscape. Whether or not this surface data represents a more permanent form of habitation or at least a regularly recurring one, however, remains a question for future research. The extensive lithic clusters on the lower slopes of a large and today extensively cultivated mountain at Aybasan (S82, S98), by contrast, present the remnants of a knapping workshop ca. 27km to the east of Okçular in a difficult to access part of Şenpazar district. The diversity and high quality of raw materials present in this assemblage is striking and the site is best interpreted as a knapping station used by groups who were part of extensive exchange networks through which they obtained obsidian from sources in the Galatian massif located in central Anatolia. In contrast to Okçular and Abdulkadir, there is little evidence for subsistence activities or habitation at Aybasan. An isolated chance find of a flint knife and polished chisel at Çamdibi dating most likely to the sixth and fifth millennia BC remains difficult to interpret, but suggests that woodworking was important. The first unequivocal evidence for more permanent settlement in the Cide region comes from Okçular Kale İni (S24). An ashy midden deposit containing pottery fragments, charcoal, bone and some lithics, which we interpret as the accumulation of domestic refuse over an extended, if not necessarily continuous, period of time during the fifth and perhaps again during the third millennium BC from a settlement located at the top of Okçular Kale. Chemical and petrographic analysis of the pottery from the cave points to localised production, while obsidian finds confirm a continued link with sources in the Çankırı area of north-central Anatolia. We also found smaller numbers of Chalcolithic pottery and chipped stone assemblages in cave sites in the Abdulkadir and Gideros area as well as at Derebağ Köy Mağarası in the far eastern corner of the survey region. More substantial numbers of probably Late Chalcolithic material derive from Kılıçlı Mağarası in the Çamdibi area. None of these assemblages, however, can be readily associated with settlement or habitation comparable to Okçular Kale.
Settlement, Subsistence and Production
437
A single site does not constitute a settlement pattern, but general parallels can nonetheless be drawn with the settlement preferences of other, roughly contemporary, Chalcolithic communities in western and central Anatolia, which also settled on elevated and striking natural features (section 5.4.2). The size of the community resident on top of Okçular Kale, the permanent or seasonal character of occupation and the primary subsistence practices of its inhabitants remain questions for future investigation. What we can state with certainty is that the significance of Okçular Kale and its immediate surroundings as a primary focus of human activity and settlement, which is hinted at already in the Early Holocene data and is most clearly articulated by Roman and Byzantine period finds and features, was consolidated for the first time during the Chalcolithic. A smaller number of Early Bronze Age pottery from Okçular Kale İni may point to a second, albeit much less substantial or short-term, episode of occupation at the top of the kale. Our remaining evidence for the Early Bronze Age comes from the cave sites of Kılıçlı Mağarası, and to a lesser extent, Derebağ Köy Mağarası. The morphologies of these caves, their locations in the landscape and accessibility, and in the case of Kılıçlı Mağarası also assemblage composition, do not suggest habitation, but are more plausibly linked with specialised activities (see below). Evidence for activity and probably also some form of occupation dating to the Middle (ca. 2000-1600 BC) and the Late Bronze Age (ca. 1600-1200 BC) derive from Okçular Kale’s immediate vicinity and from the slopes of a small natural hill ca. 300m to its northeast. The pottery compares well with central Anatolian traditions, but appears of local manufacture (section 16.3). In addition to the small number of standard north-central Anatolian-style pottery fragments, a much larger assemblage of a highly localised, and possibly contemporary, ceramic tradition was also collected in the surroundings of Okçular Kale. The formal types and coarse appearance of ‘Okçular ware’ also points towards some form of settlement at the site. Little more can be said, however, on the basis of this limited assemblage about the size and nature of later Bronze Age communities at Okçular and their subsistence practices. With the exception of a small number of pottery fragments from Derebağ Köy and Kılıçlı Mağarası, which we tentatively assign dates either in the early part or more generally the second millennium BC, the Okçular clusters remain an isolated phenomenon. As a result, our second millennium BC finds do not lend themselves to wide-ranging cross-regional comparisons. They are, however, significant in that they constitute the first empirical evidence for an occupation of the Black Sea coast during the Late Bronze Age. This comes in the form of a community that may not necessarily be of Hittite or central Anatolian origin, but one that is consuming and – judging from our chemical and petrographic data - producing standard north-central Anatolian-style pottery comparable both to the material found in regions to the south of the Pontic mountains, such as inner Paphalagonia (Glatz et al. 2009) as well as Hittite centres on the central Anatolian plateau (Schoop 2011). The Iron Ages and to some extent also the subsequent Hellenistic period are among the most elusive phases of the Cide past as regards our surface data.
438
Conclusions
An enigmatic cache of Early Iron Age (ca. 1200-950 BC) pottery from Derebağ Köy Mağarası stands in isolation not only in our survey area but also in neighbouring Black Sea regions, while the material shows general similarities with assemblages in Anatolia but also perhaps further afield in southern Europe (chapter 8). The pottery, which is coarse, handmade and includes some fragments of very large pots, certainly points to domestic activities, storage and consumption, but the cave itself seems illsuited for permanent habitation. Occasional surface finds and pottery from other caves point to a Middle and Late Iron Age presence in the survey region, but a scarcity of diagnostic pieces and the absence of the most readily identifiable hallmarks of the Anatolian Iron Age, Phrygian grey wares, painted pottery as well as burial tumuli and rock-cut tombs, make it difficult to qualify the nature of the local Iron Age. From the little evidence that we have, however, it would seem that it differs somewhat from the materials and patterns observed in neighbouring regions. In Sinop province, rope impressed pottery similar to sites located along the northwest Black Sea coast have been associated with the Early Iron Age (Doonan 2004: 55-9). Others have suggested a complete hiatus of settlement in Sinop and Samsun between ca. 1800 and 800 BC and a subsequent lack of Phrygian grey wares along the coast (Işın 1998: 110; Dönmez 2003: 214; 2005: 68). To the south of our survey area, we have evidence of human presence in inner Paphlagonia in the form of handmade painted pottery typical of the middle part of the Early Iron Age as well as Middle Iron Age painted and grey-wares (Matthews 2009), which are also attested in smaller numbers in southern Kastamonu (Marro et al. 1998). Thus, although not entirely surprising when considered in the context of other coastal surveys, this lack of Iron Age surface finds in Cide is nonetheless puzzling in the light of unambiguous textual references to the region in the first millennium BC. The Iliad makes mention of the west-central Black Sea region, naming Kytoros/ Gideros as well as Aegialus/Cide among other Black Sea settlements from whence Paphlagonians came to the aide of Troy. The same sites are mentioned in later Greek sources as colonial settlements, which would have been established from the seventh century onwards (chapter 3, 8 and 9). Yet there is little Iron Age or later Hellenistic artefactual data that would allow us to identify securely their archaeological remains. At Gideros, several courses of masonry constructed without the use of mortar may point towards Iron Age/Hellenistic beginnings of its fortifications, as does a similarly constructed wall at Okçular Kale. There is a thin scatter of Hellenistic pottery across much of the coastal part of the survey area and in some hinterland locations. The most readily identifiable finds categories are transport amphorae, which attest to the region’s participation in regional and supra-regional exchange networks. Much rarer are roof tiles, jar and pithos fragments and table wares representative of Hellenistic ways of life. It is not until the Mid- to Late Roman period, that we can begin to characterise a general pattern of settlement in the Cide-Şenpazar region. Comparable to other parts of northern Anatolia (e.g. Matthews and Glatz 2009: 247-8) our data shows an increase
Settlement, Subsistence and Production
439
in the quantity of surface material dating to this and the following Byzantine period, which may indicate a rise in population levels. We also see a more diverse range of surface data and evidence for settlement and other specialised activities in different parts of the landscape and throughout the survey region. We found Roman period surface material primarily, but by no means exclusively, in coastal landscapes. Our surface record of ceramic and tile clusters and a thin but persistent scatter of architectural spolia points to a relatively densely settled region, whose inhabitants, at least those who used recognisable Roman material culture, appear focused on urban or quasi-urban settlements. The best example for this is the area in and around the modern town of Cide and its coastal plain. Here, a series of modern construction pits, concentrations of spolia, excavated graves and extensive surface scatters in the hills above the modern town demarcate what is probably a substantial Roman period settlement. Also dotted along the coast in regular intervals are fortified strongholds such as Gideros West (S12), Kazallı Kalesi (S91) and Timle Kale (S110, S112), whose foundations can be dated mostly to the Late Roman period on the basis of architectural features. Fortress locations and orientation focused on the visual control of the coastline suggest a primary concern with seafaring and coastal security. Evidence for rural, agricultural settlement and land-use come from the coastal hinterland at Okçular, Abdulkadir, Çamdibi and Aybasan in the form of tile, and to a lesser extent, pottery clusters representative of small villages or individual farmsteads, which dotted the inland valleys of the region. Spolia, a stone-built well (S90) and the remnants of a stone bridge (S95) in the Çamdibi area inform us about the existence of a well-developed road network connecting these landscapes to the urban, culturally and politically more firmly ‘Roman’ places along the coast. Both Bithynia and Paphlagonia were sources of grain for the Roman military from at least the second and third centuries (Mitchell 1993), whose tax burden was no doubt also shouldered by farmers in the Cide-Şenpazar region. Other important products of Cide-Şenpazar and the western Black Sea region which are repeatedly mentioned in Greek, Roman and later sources include timber, ships fashioned from it, and boxwood for the manufacture of domestic and personal implements (see chapters 3, 9-11). We have no direct archaeological evidence for any of these products or their production. What we do have is possible evidence for local iron production to the north east of Okçular Kale, to be dated between the Hellenistic and Early Byzantine period. Late Roman amphorae were probably locally produced in the western hills above modern Cide. We do not know what they contained, but some appear to be of the small, narrow-necked variety, carrying perhaps more precious cargos than grain or fish products. The remnants of what appears to have been a press installation at Gökçekale (S176) points to wine or olive oil production in the southern portion of the Cide coastline. This tentative pattern of settlement, and to some extent also subsistence and production that we can tease from our Mid-to-Late Roman period data appears to continue in the Early Byzantine period. The region, however, gradually sheds its
440
Conclusions
monumental, quasi-urban and coastal focus as trade and imperial hold over the region waned between the seventh and the ninth centuries AD. Settlement continued and seemingly expanded, however, throughout the Byzantine period in inland locations, where we see an increased investment in agricultural landscapes such as in the Okçular valley. Here, a densely settled cultural landscape takes shape that includes large villages, hamlets, church complexes, hermits’ caves, and refuge forts. This stands in stark contrast with both the archaeological and palynological data for regions located to the south of the Pontic mountains, where surveys have reported a marked decline in rural settlement and agricultural production from the eighth century onwards (Matthews et al. 2009: 190-191; Kealhofer 2005: 148; Erciyas 2006: 59; Izdebski 2012). More secure maritime conditions during the ninth and 11th centuries saw a subsequent return to the coast and the re-occupation and expansion of the fortresses founded during preceding periods. New inland fortifications, such as Akça Kale (S135), may have been constructed to ward off incursions from the south. From the early 13th century, the region passed first to Beylik and subsequently Ottoman control. In AD 1284, a week-long battle took place over the castle of Gideros, in which both Muslims and Christians mustered considerable force, testifying to the continuing strategic importance of the harbour. Following this event, it appears that Cide, like Kastamonu province in general, was settled by nomadic, transhumant Turks. We have no archaeological data to confirm or challenge this scenario. Historical sources suggest a very low population density in the region from the 13th to the 19th century AD, consisting of mainly small agricultural villages and hamlets, and followed by a substantial rise in the rural population in the 19th century (chapter 13). This rise in population thus represents an early modern phenomenon, possibly linked to migration from regions lost to the Ottoman empire or a response to new demands for agricultural produce from the Zonguldak and Istanbul regions in the era of industrialisation. It is possible, that the rise in Muslim populations at this time effectively changed the demographic balance of the area, which even in the 18th and 19th centuries had significant minorities of Greek and Armenian Christians. It is probably no coincidence that much of our Ottoman archaeological record, in the form of tombstones, pipes and bridges, dates from the 18th century. During the Ottoman period, the region’s most sought-after products remained timber and ships in addition to copper and coal (chapter 3).
15.2 12,000 Years of Solitude? - Connectivity and Particularity in the Cide Region Islands and mountainous regions can be considered similar types of landscapes when it comes to questions of connectivity and particularity (Bevan and Conolly 2013: 6 for a recent discussion). This is because the physical characteristics of both discourage
12,000 Years of Solitude? - Connectivity and Particularity in the Cide Region
441
interaction, or put another way, pose particular challenges and provide particular motivations for interaction. No region of Turkey, including Cide, was too remote or physically too difficult to access in the past and so unlikely to have been entirely isolated from surrounding regions and societies. Sporadic contact and its material remnants, however, do not necessarily equate with connectivity and the social and cultural developments we would normally expect to take place within and between communities engaged in frequent and longer-term interactions, whether on an equal or unequal footing. Using archaeological and textual sources we can paint a picture of the Cide region as both remote from and connected with Anatolia and the Black Sea in political, economic as well as socio-cultural terms. The degree to which it veers between one and the other naturally varies both within and between historical periods, as does our ability to recognise this in the archaeological record. Taking a long-term perspective, we might say that Cide’s cultural traditions, as seen through our surface data, tend to have a clearly identifiable and strong local character, particularly during the earliest phases of human presence in the region. The best example for this is the heavily mineral tempered and long-lasting technological tradition that characterises the majority of Chalcolithic and Bronze Age pottery. However, we also find evidence for the region’s participation in wider exchange networks in the form of, for instance, non-local flint and obsidian artefacts. Yet, the use of low-quality flint and the practice of depleting obsidian cores to the extreme suggest the relative scarcity of these resources and, by extension, infrequent exchange of raw materials. From the Bronze Age onwards the use not only of material culture but seemingly also associated social practices, better attested from elsewhere in Turkey and the Black Sea region, becomes more pronounced. This is particularly evident in the Kılıçlı Mağarası assemblage, whose formal types and decorative patterns echo west and central Anatolian pottery traditions and whose functional composition points to practices of communal consumption, possibly of an intercommunity scale as suggested by the diverse geological and chemical signatures of the assemblage (section 16.3). Seafaring is key in facilitating Cide’s inter-regional connectivity. Even if shortdistance maritime travel must have been practiced before then, empirical evidence for seafaring and the maintenance of long-distance colonial and trade networks in the Black Sea dates from the second quarter of the first millennium BC (for a summary see Düring in press). In Cide, imports, such as fine wares and amphorae, become more readily recognisable in the surface record from the second half of the first millennium BC and provide us with direct evidence of the region’s participation in exchange and consumption networks at both regional and much wider scales. The mechanisms through which these products reached Cide, however, why they were sought after and by whom – local communities, colonial expatriates or a combination of the two – remains an open question.
442
Conclusions
In contrast to the general scarcity of evidence for the Iron Age and Hellenistic periods in Cide, Roman presence in the region, in particular along the coast and its hinterland, can be recognised easily enough in our survey record and consists of architectural spolia, imported amphorae and occasional fine-wares. These suggest a deliberate imperial presence and projection of official Roman cultural identity in strategic places. At the same time, a programme of fortress construction points to a preoccupation with the coastline and the maritime routes that tied the region to the rest of the Roman world. Roman and later also Ottoman period investment in overland transport infrastructure in the form of roads and bridges were intended to make the region more accessible to respective armies and tax collectors. What of Cide-Şenpazar’s cultural identity? During the Roman period, urban centres such as Cide would have housed one or more monumental temples. These would have been part of, and have reinforced, a Roman imperial narrative, which only few of Cide’s inhabitants, most likely only those posted there on imperial business, would have felt any true connection with. Cide’s Roman period burial record conveys an alternative perspective on local cultural identity and its constitution. This includes a cemetery of cist graves located on the scenic promontory of Gökçekale (S602) and several more at Yenice (S170) in the western part of the survey region, a rock cut grave on the western slope above Abdulkadir valley (S66), as well as two excavated cist graves at Türbe Tepe Üste in the hills to the southeast of Cide town. Several possible sarcophagi were also found inland in Çamdibi. These graves and associated burial customs are not ‘Roman’ in a cultural sense. Roman burials normally consisted of cremations and Roman law requires cemeteries to be located at the outskirts of settlements. All of the Cide graves suggest inhumation burials and all are associated spatially and visually with contemporary structures or settlements. Preferential grave placement on higher-ground may suggest the adoption of elements of Hellenistic burial custom, no doubt amalgamated with as yet unknown local practices. Religious practice and with it the region’s ritual landscape underwent a major and deep-reaching reconfiguration with the advent of Christianity in the Late Roman and Early Byzantine period. Every village community would have boasted a highly visible, monumental basilical church as its architectural focus, several of which we recorded in situ or as spolia in the course of CAP. An increasing number of CideŞenpazar inhabitants would have participated in Christian rituals on a daily basis in these churches. However, we ought not to forget those who remained pagan up until the seventh or eighth centuries AD, and the opportunities a region as geographically fragmented and remote as Cide could offer to those adhering to older beliefs and practices. With contact often comes conflict, the material culture of which is strongly represented in our survey record. It is probably no coincidence that the earliest well documented settlement in the region is located on top of the rock outcrop of Okçular Kale, which is a natural fortress that can be easily defended if necessary, and from where the surrounding landscape can be kept under surveillance. By the Byzantine
Looking Back and Moving Forward
443
period, Okçular Kale’s role as a place of local defence is again clearly established and underscored by the presence of other large-scale storage facilities in the valley below. The most evident signs of conflict, conquest and defence are the many fortified sites which we have recorded in the Cide-Şenpazar region and which reflect oscillating concerns with the coast and the protection of the agricultural hinterland. Fortress location and period of construction, thus, reflect different sets of social and political structures and their power. Roman and Early Byzantine maritime fortifications are more closely associated with imperial presence and control, whereas Middle Byzantine fortresses are the product of concerted efforts by local communities and their patrons to protect themselves and their agricultural produce. To sum up, the region appears to retain throughout its history a strong local character while engaging in contact and exchange from the Early Holocene. It witnesses several punctuated leaps in connectivity in particular following the advent and expansion of maritime technology and the arrival of external colonial and imperial powers. Imperial agendas and investment will have played an important role in the degree to which local communities were drawn into new relationships and cultural practices. At the same time, the region’s topography holds many opportunities for dissent and subversiveness as attested by the Kaska of the second millennium BC or the tobacco smugglers of the Ottoman period.
15.3 Looking Back and Moving Forward The Cide Archaeological Project is the first archaeological project to explore systematically the archaeology of the Cide and Şenpazar regions of Kastamonu province. In the course of this three-year project, we discovered and documented a large number of previously unknown or cursorily published archaeological sites and contextualised them in their wider landscapes. In this volume, we have presented our research methodologies, primary data, and our interpretations of Cide and Şenpazar’s long-term history, its communities, and their relationships with the wider world. Despite our best efforts, copious amounts of sweat, occasional tears and blood during climbing expeditions and caving adventures, much, much more remains to be discovered about this fascinating region perched on Turkey’s northern edge. We hope that more archaeologists will dare venture into the Pontic mountains in the future.
Bibliography Bevan, A., and Conolly, J. 2013. Mediterranean Islands, Fragile Communities and Persistent Landscapes. Antikythera in Long-Term Perspective. Cambridge, Cambridge University Press. Dönmez, Ş. 2003. The Early Iron Age problem in the central Black Sea region. In B. Fischer, H. Genz, É. Jean, and K. Köroğlu, eds., Identifying Changes: The Transition from Bronze to Iron Ages in Anatolia and its Neighbouring Regions. Istanbul, Ege Yayınları: 213-28.
444
Conclusions
— 2005. Amasya province in the Iron Age. In A. Çilingiroğlu and G. Darbyshire, eds., Anatolian Iron Ages 5: Proceedings of the Fifth Anatolian Iron Ages Colloquium held at Van, 6-10 August 2001. London, British School or Archaeology at Ankara: 65-74. Doonan, O. P. 2004. Sinop Landscapes: Exploring Connection in a Black Sea Hinterland. Philadelphia, University of Pennsylvania Museum of Archaeology and Anthropology. Düring B. S. 2008. The Early Holocene occupation of North-Central Anatolia between 1,000 and 6,000 BC cal: Investigating an archaeological terra incognita. Anatolian Studies 58: 15-46. — in press. Marginal landscapes and connected societies: Assessing connections of the west-central Turkish Black Sea region in the Early to Mid Holocene. In C. Glatz, S. Sherratt, and A. Bauer, eds., Archaeology of the Early Black Sea Region. Connectivity and Remoteness. London, Routledge. Düring, B. S., and Glatz, C. 2010. The Cide Archaeological Project 2009: First results. Anatolia Antiqua 18: 203-13. Erciyas, D. B. 2006. Wealth, Aristocracy and Royal Propaganda under the Hellenistic Kingdom of the Mithradatids in the Central Black Sea Region of Turkey. Leiden, Brill. Glatz, C., Matthews, R., and Schachner, A. 2009. A landscape of conflict and control: Paphlagonia during the second millennium BC. In R. Matthews and C. Glatz, eds., At Empires’ Edge: Project Paphlagonia. Regional Survey in North-Central Turkey. London, British Institute at Ankara: 107-47. Işın, M. A. 1998. Sinop regional field survey. Anatolia Antiqua 6: 95-139. Izdebski, A. 2012. The changing landscapes of Byzantine northern Anatolia. Archaeologia Bulgarica 16 (1): 47-66. Kealhofer, L. 2005. The Gordion regional survey. Settlement and land use. In L. Kealhofer, ed., The Archaeology of Midas and the Phrygians. Recent Work at Gordion. Philadelphia, University of Pennsylvania Press: 137-48. Marro, C., Özdoğan, A., and Tibet, A. 1998. Prospection archéologique Franco-Turque dans la région de Kastamonu (Mer Noire). Troisième rapport préliminaire. Anatolia Antiqua 6: 317-35. Matthews, R. 2009. A dark age, grey ware and elusive empires: Paphlagonia through the Iron Age, 1200-330 BC. In R. Matthews and C. Glatz, eds., At Empire’s Edge: Project Paphlagonia. Regional Survey in North-Central Turkey. London, British Institute at Ankara: 149-72. Matthews, R. and Glatz, C. 2009. People and places in Paphlagonia: Trends and patterns in settlement through time. In R. Matthews and C. Glatz, eds., At Empires’ Edge: Project Paphlagonia. Regional Survey in North-Central Turkey. London, British Institute at Ankara: 239-49. Matthews, R., Metcalfe, M. and Cottica, D. 2009. Landscapes with figures: Paphlagonia through the Hellenistic, Roman, and Byzantine periods, 330 BC-AD 1453. In R. Matthews and C. Glatz, eds., At Empires’ Edge: Project Paphlagonia Regional Survey in North-Central Turkey. London, British Institute at Ankara: 173-226. Mitchell, S. 1993. Anatolia. Land, Men, and Gods in Asia Minor 2. The Rise of the Church. Oxford, Clarendon Press. Schoop, U.-D. 2011. Hittite pottery: A summary. In H. Genz, and D.P. Mielke, eds., Insights into Hittite History and Archaeology. Leuven, Peeters: 241-73.
16 Appendices 16.1 Appendix I. Luminescence Dating Report, Cide – Turkey, NCL-8111 Jakob Wallinga and Alice J. Versendaal Two samples for luminescence dating were submitted by Bleda Düring of Leiden University. Luminescence dating analyses were performed at the Netherlands Centre for Luminescence dating, Delft University of Technology. The goal of the luminescence dating is to determine the time of deposition of beach sands found ~2m below the surface, some 300m from the present coast. An improved chronology of the formation of the Cide coastal plain will be constructed based on the luminescence dates obtained. Optically stimulated luminescence (OSL or optical) dating determines the timing of deposition and burial of sand or silt-sized mineral grains. The method makes use of a tiny light signal emitted by grains of quartz or feldspar. This luminescence signal results from the recombination of trapped charge inside the crystal lattice. The signal is reset (bleached) by sunlight and builds up after deposition and burial due to exposure to natural ionizing radiation. This background radiation results mainly from the radioactive decay of 40K and the Uranium and Thorium decay chains, with an additional contribution from cosmic rays. Luminescence methods are applicable over the age range of a few years up to about 150,000 (quartz) or 500,000 (feldspar) years. The methods can provide a precision of 5% of the age at best (1 sigma). Two quantities are determined for luminescence dating. Measurement of the OSL signal on the purified quartz mineral fraction reveals how much ionizing radiation the sample received since the last bleaching event. This measurement is combined with a measurement of the background radiation level at the sample position. The luminescence age is then obtained by dividing the amount of radiation received (palaeodose, Gy) by the yearly dose (dose rate, Gy/ka): Age (ka) = Palaeodose (Gy) / dose rate (Gy/ka). For additional information on luminescence dating the reader is referred to Aitken (1997), Wallinga et al. (2007) and Wintle (2008).
16.1.1 Methods Dose rate For dose rate estimation a puck is prepared by mixing (ground) sediment with wax. The wax serves to provide a pre-determined geometry and to retain Radon in the sample. Activity concentrations of 40K and several nuclides from the Uranium and Thorium decay chains are measured using high-resolution gamma ray spectrometer. Results are combined with information on burial history, water and organic content
446
Appendices
history, and the grain size fraction used for luminescence measurements to calculate the effective dose rate. We assumed gradual burial of the samples to the present depth below the surface, and estimated water contents to be 10% by weight based on saturated water content of sandy deposits. There were no signs of disequilibrium in the Uranium decay chain. Resulting dose rate values range from 0.98 – 1.00 Gy/ka, which is comparable to other samples from similar deposits. Palaeodose The quartz grain size fraction of 180-212μm was purified by sieving and chemical treatment (HCl, H2O2 en HF and rinsing with HCl). No density separation was used. Based on a number of tests, suitable measurement parameters were selected for use in the Single Aliquot Regenerative (SAR) dose procedure (Murray and Wintle 2003). Within the SAR procedure, the natural luminescence signal is first measured, and then compared to laboratory induced luminescence signals to find what radiation dose is needed to induce an equally bright signal. This value provides the palaeodose. The procedure monitors and corrects for luminescence sensitivity changes during the measurement. The most light-sensitive OSL signal of quartz grains is selected using the ‘Early Background’ approach (Cunningham and Wallinga 2011). To obtain a good estimate of the burial dose, measurements are repeated on at least 41 aliquots per sample. Each of these aliquots consists of a sample disc containing ≈75 grains (2mm diameter sample). To test the SAR procedure and the selected measurement parameters, a laboratory given dose is retrieved with the adopted procedure; the measured dose agreed favourably with the given dose (dose recovery ratio 1.05 ± 0.02, n=18). Palaeodose estimates on the single aliquots were slightly overdispersed, especially for the second sample; i.e. the data points were more scattered than expected based on the measurement uncertainties. This suggests an additional source of scatter, e.g. mixing of grains of different ages or incomplete resetting of the OSL signal in some grains prior to deposition (heterogeneous bleaching). Given the palaeodose distributions obtained and the depositional environment of these samples there are no clear signs of heterogeneous bleaching or bioturbation. In the absence of an explanation for the slight overdispersion, we decided to use the Central Age Model (Galbraith et al. 1999) to determine the burial dose from the palaeodose distribution.
16.1.2 Dating Results For each of the samples, the burial age is determined by dividing the burial dose by the doserate (table 16.1). The errors shown indicate the 1-sigma (68%) uncertainty, including all systematic and random uncertainties in both burial dose and dose rate estimation. Results are graphically shown in radial plots which indicate singlealiquot ages, and the final age estimate. The validity estimate in table 16.1 is based on
Appendices
447
a combination of the palaeodose distribution, the consistency of the ages obtained, and the luminescence properties of the samples. Tab. 16.1: Summary of luminescence dating results. sample No NCL poject ID NCL-8111013 NCL-8111014
equivalent dose rate dose (Gy) (Gy/ka)
CAP2010/ 3.9 ± 0.2 SP02/GS01 CAP2010/ 3.7 ± 0.2 SP02/GS03
age (ka)
(syst.)
(rand.) validity
0.98 ± 0.03 3.94 ± 0.22 0.14
0.17
likely OK
1.00 ± 0.03 3.67 ± 0.20 0.13
0.15
likely OK
16.1.3 Synthesis Luminescence ages were obtained by OSL measurements on purified quartz mineral extracts. The results indicate that the deposits were formed around 3.8ka (ca. 1800 BC). Although the palaeodose data is slightly overdispersed, the consistency of the dates contributes to the confidence in the dating results; we expect that the OSL ages accurately represent the burial age of the sediments.
Bibliography Aitken, M.J. 1997. Luminiscence dating. In R.E. Taylor and M.J. Aitken, eds., Chronometric Dating in Archaeology. New York, Plenum Press: 183-215. Cunningham, A.C., and Wallinga, J. 2010. Selection of integration time-intervals for quartz OSL decay curves. Quaternary Geochronology 5: 657-66. —2011. Expectations of scatter in equivalent-dose distributions when using multi-grain aliquots for OSL dating. Geochronometria 38: 424-31. Galbraith, R.F. 1990. The radial plot – graphical assessment of spread in ages. Nuclear Tracks and Radiation Measurements 17: 207-14. Galbraith, R.F., Roberts, R.G., Laslett, G.M., Yoshida, H., and Olley, J.M. 1999. Optical dating of single and multiple grains of quartz from Jinmium rock shelter, northern Australia. Experimental design and statistical models. Archaeometry 41: 339-64. Murray, A.S., and Wintle, A.G. 2003. The single aliquot regenerative dose protocol: Potential for improvements in reliability. Radiation Measurements 37: 377-81. Roberts, R.G., Galbraith, R.F. Yoshida, H., Laslett, G.M., and Olley, J.M. 2000. Distinguishing dose populations in sediment mixtures: A test of single-grain optical dating procedures using mixtures of laboratory-dosed quartz. Radiation Measurements 32: 459-65. Wallinga, J., Davids, F., Dijkmans J.W.A. 2007. Luminescence dating of Netherlands’ sediments. Netherlands Journal of Geosciences – Geologie en Mijnbouw 86: 179-96. Wintle, A. G. 2008. Fifty years of luminescence dating. Archaeometry 50: 276-312. Wintle, A.G., and Murray, A.S. 2006. A review of quartz optically stimulated luminescence characteristics and their relevance in single-aliquot regeneration dating protocols. Radiation Measurements 41: 369-91.
448
Appendices
16.2 Appendix II. Main Recording Forms of CAP Toby C. Wilkinson
Fig. 16.1A: The CAP site recording form.
Appendices
Fig. 16.1B: The CAP site recording form.
449
Fig. 16.2: The CAP Unit form.
450 Appendices
Fig. 16.3: The CAP Grid form.
Appendices 451
452
Appendices
16.3 Appendix III. Chemical and Petrographic Study of Pottery from the Cide Archaeological Project Richard Jones This report on the study of pottery from the Cide Archaeological Project, whose aims were set out in section 4.4.1.2 represents work in progress. The samples are listed in table 16.2. Tab. 16.2: Pottery samples analysed chemically and petrographically. ceramic ID
site
site type
village
date
56
24
cave
Okçular
MCHA
170
17
cave
Abdulkadir
CHA
358
surface
Okçular
MBA-LBA? (Okçular ware)
633
surface
Okçular
HEL
750
surface
Okçular
HEL
790
24
cave
Okçular
MCHA
803
24
cave
Okçular
MCHA
812
24
cave
Okçular
MCHA
1088
surface
Okçular
MBA-LBA? (Okçular ware)
2856
surface
Okçular
MBA
2956
surface
Okçular
MBA-LBA? (Okçular ware)
cave
Çamdibi/Kılıçlı
EB II-III
surface
Cide
MROM-LROM
3114
88
3124 3132
surface
Cide
MROM-LROM
3168
88
cave
Çamdibi/Kılıçlı
EB II-III
3186
88
cave
Çamdibi/Kılıçlı
EB II-III
3615
88
cave
Çamdibi/Kılıçlı
EB II-III
3621
88
cave
Çamdibi/Kılıçlı
EB II-III
3644
188
cave
Çamdibi/Kılıçlı
EB II-III
3646
188
cave
Çamdibi/Kılıçlı
EB II-III
3740
88
cave
Çamdibi/Kılıçlı
EB II-III
4023
188
cave
Derebağ
EB II-III?
4027
188
cave
Derebağ
EB II-III?
4032
188
cave
Derebağ
EIA
4169
188
cave
Derebağ
EIA
surface
Okçular
LBA
4185 4221
188
cave
Derebağ
CHA/EB I
4222
188
cave
Derebağ
EIA
4228
188
cave
Derebağ
EIA
Appendices
453
Beginning with the chemical results, the compositions (except for C56) determined by portable XRF are given in table 16.3. Not unexpectedly, they display a range of element concentrations, notably in calcium, manganese and strontium, and this is borne out in the global view of the data obtained by applying principal components analysis (Fig. 16.4a). The fine wares of Roman date, C3124 and C3132, have similar compositions, as do the two examples of Hellenistic fine wares, C633 and C750, but none of these stands clearly apart from the Prehistoric and Bronze Age pottery (Fig. 16.4b). No attempt is made here to consider the sources of these fine wares. There is little evidence of clustering of samples according to date among the Prehistoric and Bronze Age samples. a)
b)
Fig. 16.4: Principal components analysis plot of all samples for all fourteen elements according to (a) sample number and (b) approximate date.
454
Appendices
Tab. 16.3: Chemical compositions of the pottery expressed as weight percent element (Si, Al, Ca, Fe, K) and ppm element (remaining elements). sample
%Si
%Al
%Ca %Fe
%K
Ti
Mn
Ba
Zr
Sr
Rb
Zn
Cu
Cr
C 170
17,8
7,0
14,9
3,0
1,3 2617 1460 895
70
136
40
101
48
119
C 358
19,0
6,7
10,9
3,3
0,6 4721 1034 1867 164
147
31
75
55
119
C 633
27,4
8,6
6,0
4,0
1,2 3862 756
431
122
322
26
74
41
266
C 750
22,2
7,1
9,4
2,6
1,8 2673 827
904
108
295
53
58
45
291
C 790
18,3
5,2
13,9
2,9
0,8 2908 1618 1008
93
117
30
79
52
133
C 803
15,2
4,4
17,1
4,4
0,7 2277 1206 971
80
149
27
82
45
117
C 803
17,3
5,3
15,4
5,4
0,8 2401 1680 1095
79
480
27
98
66
153
C 812
20,9
5,4
11,3
3,9
0,8 3189 1897 1053 100
136
29
89
64
132
C 1088 21,6
7,7
12,5
3,4
1,2 3391 1274 713
128
113
51
106
43
148
C 2856 17,1
6,3
13,8
3,1
0,6 3931 1491 977
151
196
31
60
77
149
C 2956 19,0
7,3
9,6
3,6
0,9 4060 912 1724 140
102
35
79
41
193
C 3114 29,0
7,3
1,4
3,6
0,9 3652 337
515
170
78
41
58
30
157
C 3124 31,0
9,7
0,4
4,0
1,4 3939 393
626
171
93
59
71
44
184
C 3132 30,2 10,1
3,2
4,3
1,4 3866 373
558
190
79
54
61
41
176
C 3168 28,8
8,6
1,9
2,7
2,4 3215 591
481
140
56
62
58
38
122
C 3186 26,1
8,9
1,8
4,5
2,2 3050 1502 922
108 1033
57
72
64
148
C 3615 23,8
8,8
8,0
2,9
1,7 3055 760
613
111
103
55
85
55
170
C 3621 25,7
7,2
2,0
3,2
1,6 3277 600
633
151
83
60
101
76
160
C 3644 27,9
7,7
1,9
3,3
1,7 2711 260
514
129
56
47
65
39
160
C 3646 26,9
8,3
3,5
2,7
1,9 3165 684
603
131
76
66
61
48
152
C 3740 30,4
6,5
1,3
3,5
1,3 3700 446
651
183
87
41
73
31
152
C 4023 24,3
8,2
2,5
4,1
2,1 3158 872
763
97
546
39
80
53
139
C 4027 22,9
9,2
2,3
5,8
2,6 3763 1309 733
104
251
66
81
95
197
C 4032 23,4
7,9
1,8
3,9
1,9 3645 532
688
104
255
33
73
59
190
C 4169 27,1
8,3
1,6
4,2
1,2 3032 1231 607
77
276
26
74
69
104
C 4185 26,4 11,2
1,4
3,0
1,6 4913 401
196
126
50
84
50
164
C 4221 25,5
8,3
1,1
4,2
3,7 4060 574 1225 121
289
51
65
40
136
C 4222 22,8
6,8
6,4
3,4
2,8 3103 966
570
111
185
45
83
42
169
C 4228 23,9
9,5
1,8
4,3
1,7 3222 956 1040
80
212
26
77
71
118
461
Appendices
455
A more encouraging picture emerges in plotting the calcium and strontium contents; setting aside the anomalous Sr contents in C3186 and C4023 (both of EB II-III date), Fig. 16.5a shows contrasting calcium contents, high at Okçular and Abdulkadir and low at Çamdibi/Kılıçlı Mağarası and Derebağ Köy Mağarası. Whereas these two elements are commonly highly correlated, this correlation is clearly broken among the Derebağ samples. Other bi-variate plots, including Ca-Mn, reveal a similar picture. C4185, a surface Late Bronze Age sherd from Okçular, has a composition more typical of Çamdibi owing to its low Ca and Sr, and the calcareous nature of C3186 found at Çamdibi is more typical of Okçular. The most satisfactory outcome of the chemical analysis is the way the three main sites, Okçular, Çamdibi and Derebağ, can be partially discriminated; this could be indicative of multiple potting traditions within the region associated with particular localities, and furthermore, in view of the chronological range of samples within each potting tradition (Fig. 16.5b), production was long lived, continuing from the end of the Chalcolithic to the end of the Bronze Age. Turning to the corresponding petrographic compositions, the following samples from Okçular are characterised in a striking manner by the presence of marble: C56, C170, C358, C790, C803, C1088, C2856 and C2956: these samples encompassing a range of wares and date. In C803 (Fig. 16.6a) the larger fragments of marble are very angular and fresh ranging up to 2mm in length; there is frequent small rounded quartz up to 0.2mm, very occasionally 0.4mm; occasional muscovite mica. The frequency and size of marble fragments varies at Okçular; by comparison with C803 it is much lower in C1088, and size ranges up to 4mm in C2856; the dark rounded inclusions are a feature of this last sample (Fig. 16.6b). Although the coastal region is sedimentary in character, the presence of marble is confirmed by the occurrence of a currentlyexploited marble quarry in the Abdulkadir valley (chapter 14). It is interesting to find a typical example of this fabric in C4222 found at Derebağ, while C4185 found at Okçular has no marble, nor does it relate to any of the material found at Çamdibi and Derebağ. At Çamdibi, the situation is very different. For a start, the matrix is often reddish, but there is little uniformity of composition. To take a few examples, C3168 has distinctive orange isotropic inclusions which are not present in the other examples of EB II-III pottery from this cave, C3186 (Fig. 16.7a), C3621 and C3644. Instead, C3186 is characterised by a wide range of inclusion types in the 1-2mm size range, including plagioclase feldspar forming probable olivine basalt and plagioclase feldspar phenocrysts in a (volcanic) glass. C3644 with a red matrix contains some feldspar and large polycrystalline quartz associated with granite; there is some calcite and marble. In C3644 frequent well-sorted quartz occurs, up to 0.3mm in size, some of it polycrystalline (up to 1mm) deriving from a schist; besides occasional marble, a single very fine-grained crystalline calcite inclusion 2mm long is observed. This calcite inclusion differs from the occasional subrounded calcite, typically 1mm, present in C3646 (Fig. 16.7b) which has very common, small quartz in a fine red matrix. In a similarly fine red matrix with frequent quartz, some of which is of granitic origin, C3740 (Fig. 16.7c) contains a distinctive igneous inclusion, as well as occasional dark isotropic subrounded fragments with quartz phenocrysts and subrounded opaques.
456
Appendices
a)
b)
c)
Fig. 16.5: Ca-Sr plot of samples (except C3186 and C4023) from (a) Okçular (O), Abdulkadir (A), Çamdibi (C) and Derebağ (D), (b) according to date and (c) according to sample number.
Appendices
a)
457
b)
Fig. 16.6: Photomicrographs (x6.3, xpl) of (a) C803 and (b) C2856 from Okçular.
There appears similar diversity among the samples at Derebağ. On the one hand, C4169 shares the feature of feldspar phenocrysts in a (volcanic) glass, but its dark red, coarse matrix contains a distinctive large sedimentary fragment 3.5mm in size (Fig. 16.8a) and there are some fragments which are not vitreous but contain feldspar phenocrysts. On the other, C4027 has a large (2mm) black isotropic subrounded fragment with quartz phenocrysts, quite different from C4169. C4032, which is very quartz rich, some of it of large size, differs from C4169 of the same date.
16.3.1 Discussion The chemical characterisation suggests a classification of three main groups. The one at Okçular represents a distinct potting tradition which, according to the corresponding petrographic data, is demonstrably local. However, the other two chemical groups, which tend to comprise the Çamdibi and Derebağ samples respectively, should not be taken to correspond to either local production at each locality or, in the light of the petrographic data, separate ‘potting traditions’ practised in area around these two caves. The chemical data seems to be giving an over-simplified picture. Instead, the petrographic data is
458
Appendices
pointing to remarkable diversity of composition, supporting the view that Çamdibi and Derebağ were receiving pottery from several different (and including local) sources. Both data sets concur in placing C3186 from Çamdibi as a product of Okçular. But for the rest, it is only possible at present to recognise that this pottery originated from sedimentary, metamorphic and igneous, including volcanic environments. Acknowledgement: Many thanks to Martin Lee for discussion of some of the thin sections. a)
b)
c)
Fig. 16.7: Photomicrographs of (a) C3186, (b) C3646 and (c) C3740 from Çamdibi (magnifications all in the range x6.3-8.5; all ppl).
Appendices
a)
459
b)
Fig. 16.8: Photomicrographs (x6.3; ppl) of (a) C4169 and (b) C4228 from Derebağ Köy Mağarası.
460
Appendices
16.4 Appendix IV. Chemical and Magnetic Analysis of Soils from Grid 6 (G6) at Okçular Richard Jones During the 2011 season of the Cide Archaeological Project, 48 soil samples were taken from grid G6 for laboratory analysis in Glasgow for the purpose of locating areas of anthropogenic activity and determining the extent of their correlation with surface sherd distribution. This investigation was designed to combine traditional soil phosphate determination with magnetic susceptibility, whose level in soil is significantly enhanced by the effects of localised intense heating. Multi-element analysis was also adopted because there is growing recognition that particular human activities can leave distinct chemical patterning in soils (see, for example, the Laconia rural sites project in Greece [Cavanagh et al. 2005: 32-7, 305-10] and the intra-site study at Çatalhöyük [Middleton et al. 2005]). The samples were taken at 10m intervals at locations shown in (Fig. 16.9); samples 49 and 50 are control samples lay outside the survey area. The land encompassing G6 which is gently sloping is now given over to pasture. In recent decades it has not been cultivated and so has been unaffected by the introduction of phosphate-rich fertiliser.
Fig. 16.9: The sample locations in G6 superimposed over the sherd count distribution showing high density in dark shades.
Appendices
461
The weight of soil taken at each location was ca. 0.5kg. On arrival in Glasgow, the soils were dried, and a 100gm subsample was then taken for sieving with a 2mm Endicott sieve and the powder collected in a small plastic bag. For three samples – 15, 19 and 24 – larger subsamples were prepared and analysed separately to check for comparability of analysis results. Samples 1-3 were not analysed.
16.4.1 Analyses 1. Total phosphate (P) concentration (in mg P/100 mg of soil) was determined by the standard molybdenum blue colorimetric method (Bethell and Máté 1989: 21-2). 2. Magnetic susceptibility (MS) was determined with a Bartington MS2 system. Both low and high frequency measurements were made on a known weight of sample, typically between 8 and 10gm. 3. Chemical composition was determined non-destructively by X-ray fluorescence (XRF), using a portable Thermo NitonXL3t instrument operating in the Soils mode calibration. The sample powder was placed in a plastic vial with a Mylar window and analysed in at least two positions; agreement between replicate analyses was better than 5%; this figure also applied to agreement between the 100gm and larger samples in the case of 15, 19 and 24 for elements other than Cu and Pb for which it was c. 20%. There were three soil standards: NIST 2383, 2709a and 2780. Quantitative determinations were made of major elements Fe and Ca, minor elements K and Ti and trace elements Zr, Sr, Rb, Pb, Cu, Zn, Mn, V and Sc.
16.4.2 Results Collectively, P concentrations and MS (low frequency) values are relatively low and exhibit rather small ranges. Loss on ignition which is indicative of the organic content in the soil varies from 3.6 to 12% with a mean of 7.2%. Examination of the distributions of P and MS along linear sample sequences reveals generally good agreement between P and MS, as demonstrated for example in the long sequence, 12-37 (Fig. 16.10), where both parameters reveal enhanced values at 13-14 and 18-19. Taking arbitrarily high P and MS to represent values greater than 140 and 20 units respectively and setting aside the anomalous P concentration in control sample 50, Fig. 16.11 highlights the areas of enhancement. The hotspot at 30, 31, 5, 32 lies to the side of, and smeared away from, the moderate sherd density at F17-G17. At 13, 14, 7, 8, 15 the enhancement is close to the spatially large high density which peaks from L13 to M15. There is a notable absence of P and MS enhancement associated with the sherd densities centered at K3-4 (i.e. 25-26), and the converse effect is encountered at 18-19 and 40.
462
Appendices
Fig. 16.10: Phosphate (blue) and Magnetic Susceptibility (red) values in 12-37 and including the values for 8 which is labelled 14a.
Fig. 16.11: Phosphate and MS enhancements highlighted on the sherd distribution plan in G6.
Appendices
463
Turning to the chemical data obtained by XRF (Table 16.4), there is a striking range in calcium content (0.7-8.7%) (Fig. 16.12), the majority of samples being classed as calcareous which accords well with their light colour. The other elements show narrower concentration ranges, and moreover for the trace metals, Pb, Zn, Cu and V, the absolute concentrations are all low. The following correlations are observed: high for Ca-Sr (R2 correlation coefficient of 0.90), and lower but nevertheless noteworthy for Ti-Pb (R2 0.53), Ti-Zr (R2 0.48) and Rb-K (R2 0.41). Zn and Cu show little or no correlation between each other or with the other elements. In view of the wide calcium range, it was decided to normalise the compositions to that element. Fig. 16.13a, b shows the Zr/Ca, Sr/Ca, Zn/Ca and Cu/Ca in sequence 12-27. The following remarks can be made about each sequence: –– Sequence 4-11: Peak in 7 with a possible smaller one in 4 for Zr-Sr; peaks in 4 and 7-8 for metals Cu, Zn Fe –– Sequence 12-27 (Fig. 16.13): All elements indicating peaks in 14-15, 18-20, 24? and 27 –– Sequence 28-36: All elements indicating peaks in 28, 30 and 34-35 –– Sequence 37-42: All elements indicating peaks in 39 and 40 –– Sequence 43-48: Peak in 47 for iron, and the other elements show an anomalously large peak in 48
Fig. 16.12: The calcium distribution for all samples.
106
209
209
199
199
188
183
164
167
171
168
208
214
180
163
213
212
186
220
189
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
167
179
109
107
120
166
166
96
102
156
180
190
192
123
107
97
121
119
Sr
Zr
Sample
104
111
114
126
129
124
123
137
144
126
137
135
130
136
142
124
130
128
120
Rb
17
22
19
23
25
15
18
26
28
18
18
16
16
19
24
28
23
29
24
Pb
100
98
99
102
116
111
106
115
131
104
111
98
112
111
112
118
116
110
105
Zn
48
29
32
45
43
25
46
38
53
49
51
26
47
44
41
53
38
38
37
Cu
25513
28822
28885
34504
33401
31551
28156
35227
36550
42894
34833
29938
36296
40620
37535
34058
32139
32852
32940
Fe
466
475
480
706
674
403
434
670
710
736
662
634
728
525
467
708
565
72
700
Mn
141
146
172
181
172
164
162
217
221
197
200
166
190
227
220
197
164
162
147
V
3248
3451
4173
4205
4637
3210
3470
4618
4220
3391
3606
3296
3456
3928
4139
4033
3874
3894
3350
Ti
326
381
289
209
256
501
443
180
171
462
418
512
448
314
180
259
280
289
170
Sc
75811
78697
40444
35386
39469
79268
76250
26983
29076
69587
76369
82942
84976
47124
36036
42900
53759
49518
31700
Ca
20384
21101
22677
22558
24759
24493
24460
25725
25262
24767
27231
25891
27449
25899
26382
24015
24371
22940
19000
K
21
22
23
33
30
17
21
29
34
49
18
19
21
18
28
31
29
32
28
P
12
12,4
15,3
25,8
32,7
13,4
11,8
23,8
31,2
24,2
15,6
12,8
13,2
17
17
36,7
18,4
54,3
14,9
MS low freq.
Tab. 16.4 Element concentrations (in ppm), phosphate (P) (mg/100 mg soil) and magnetic susceptibility (MS) determined at low and high frequencies
11,2
12,2
14,1
23,8
31,4
10,7
11,4
22,3
26,3
22,2
13,6
11
12,3
18,5
14,7
31,5
17,7
49,4
17,9
MS high freq.
464 Appendices
Zr
197
200
199
187
230
202
186
198
180
220
219
234
230
215
211
203
230
237
151
Sample
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
223
98
115
143
133
113
91
87
99
112
98
101
143
101
127
172
185
132
134
Sr
117
124
119
122
122
124
133
128
120
120
114
123
149
154
114
128
111
120
120
Rb
18
28
25
23
22
20
27
29
23
32
21
25
18
26
26
17
18
20
21
Pb
106
115
101
107
105
107
115
113
114
114
108
108
102
116
104
104
104
106
104
Zn
53
40
42
46
47
42
40
39
38
47
37
36
37
48
42
44
48
45
36
Cu
28550
32399
31532
30750
31045
30908
33759
32676
30543
31716
29562
31812
43299
37264
28400
31727
30017
29132
29817
Fe
428
626
667
612
586
689
677
837
697
856
552
615
408
487
533
465
440
462
467
Mn
132
203
164
160
161
182
177
194
198
179
204
198
194
241
163
145
169
155
157
V
2615
4580
4190
3800
3962
4237
4510
4557
4183
3964
4082
4190
3897
4483
4040
3542
3402
3787
4128
Ti
427
183
280
317
264
221
123
170
197
283
260
211
235
143
297
427
464
329
317
Sc
86707
24310
37934
56990
50982
41090
19950
19906
39692
47862
48712
27293
48739
24482
50037
70926
92795
52275
57572
Ca
20080
22702
22489
22588
22687
22699
22259
22725
22007
22677
24691
25242
26091
27965
23089
23786
23426
21695
23486
K
14
30
29
23
20
32
29
32
29
32
31
30
24
29
23
21
22
24
21
P
8,7
15,5
19,9
14,1
13,2
21,4
22,2
24,4
22,4
35,9
16,3
19
15,6
14,8
14,5
12,3
11,4
18,3
14,2
MS low freq.
8,9
14,6
20,6
14,6
12
20,4
20,9
22,3
21,6
31
15,4
18,4
13,1
14,8
12,4
10,9
11
17,2
13,9
MS high freq.
Tab. 16.4 Element concentrations (in ppm), phosphate (P) (mg/100 mg soil) and magnetic susceptibility (MS) determined at low and high frequencies
continued
Appendices 465
Zr
167
187
188
195
170
204
280
169
165
Sample
42
43
44
45
46
47
48
49
50
86
123
66
130
197
195
158
152
210
Sr
130
146
123
117
129
121
112
121
137
Rb
27
18
32
19
20
15
19
15
17
Pb
106
113
99
109
110
103
103
104
90
Zn
33
30
40
41
48
40
49
51
38
Cu
35466
28726
32607
29962
30610
29409
30309
28234
22106
Fe
665
324
701
575
440
493
539
406
403
Mn
244
155
216
154
158
149
159
165
179
V
4216
3034
4664
3793
3342
3391
3546
3576
3546
Ti
210
296
84
348
464
482
378
363
364
Sc
38474
47003
7435
58365
78665
81392
71657
68804
85399
Ca
24823
22079
19646
22054
24808
24110
22834
23139
25610
K
34
13
25
22
20
21
26
23
15
P
19,1
10,9
19,2
11,4
11,4
10,1
18,7
10,7
9
MS low freq.
18,8
12,4
16,9
11,5
11,5
10,3
17,7
10
9,9
MS high freq.
Tab. 16.4 Element concentrations (in ppm), phosphate (P) (mg/100 mg soil) and magnetic susceptibility (MS) determined at low and high frequencies
continued
466 Appendices
Appendices
467
a)
b)
Fig. 16.13: (a) Zr/Ca and Sr/Ca ratios and (b) Zn/Ca and Cu/Ca ratios in the sample sequence 12-27.
468
Appendices
Revising Fig. 16.11 to incorporate the chemical data gives Fig. 16.14 with the multi-element marked in yellow. Good agreement between P, MS and multi-element composition appears in the areas: –– 7, 8, 14, 15 –– 18, 19, ?20 –– 35, 36 –– 39, 40 there is little multi-element composition enhancement around 5-32 but it is apparent near the sherd density peak at 25-26.
Fig. 16.14: Locations of the ‘hotspots’: phosphate and MS in black (as from Fig. 16.11); multi-element composition in yellow (dotted yellow for uncertain enhancement).
16.4.3 Discussion The most positive outcome of this investigation is the level of spatial correlation between enhanced levels of three soil parameters, phosphate, magnetic susceptibility and multi-element composition, at several locations. This would strongly suggest these locations reflect areas of anthropogenic activity, the low level of enhancement pointing towards minor or scattered, even intermittent settlement, plausibly
Appendices
469
associated with agricultural activity; the presence in G6 of domestic pottery and building tile helps to corroborate such an interpretation. A specific date range for G6 is difficult to pinpoint exactly because, with the exception of a fragment of Ottoman glazed ware, investigations did not yield any definitively datable fine wares. The majority of ceramic data points to habitation from the Roman period (chapter 10), and there is evidence for continuity in the Late Roman/Early Byzantine period (chapter 11). The higher values of magnetic susceptibility from G6 are consistent with areas of localised burning of waste or from hearths and not with burial; there is no hint from the data of the presence of any craft-based activity such as pottery making. Turning to the respective locations of enhancement and high density of surface sherds, the correspondence is variable, and here the limitation of partial rather than total sampling within G6 has to be acknowledged. In the case of 7, 8, 14 and 15 it is quite good but there is clearly some displacement. At the high sherd density at 25-26 there is only multi-element enhancement and it is discrete at that. The enhancement at 19-20 and 39-40 may link together but in any case there is no associated high sherd density. The long ‘smear’ of high P and MS between 5 and 36 with discrete multi-element enhancement within and adjacent to it is not readily explainable. Nevertheless two points can be made here: first, the terrain is uneven due mainly to the presence of shallow gullies which run parallel to the line of slope, and second an effect of the sloping terrain will be to cause surface artefacts to be more mobile than the top soil. In light of both these factors good correspondence between soil and artefact should not be expected. The observation of low P concentrations encountered in the calcareous soils in G6, which has been noted above, is also recorded in the corresponding results of soil phosphate analysis carried out in the Urfa region of southern Turkey. Here, Wilkinson (1990, 76) found the highly calcareous soils yielded lower phosphate than the redder soils on the terraces. Acknowledgement: Many thanks to Gert Petersen for carrying out the phosphate analyses.
Bibliography Bethell P. and Máté, I. 1989. Soil phosphorus analysis. In J. Henderson, ed., Scientific Analysis in Archaeology. Oxford: Oxford University Committee for Archaeology: 1-29. Cavanagh, W.G., Mee, C., and James, P. 2005. The Laconia Rural Sites Project. British School at Athens Supp. Vol. 36. London. Middleton W.D., Price, T., and Meiggs, D. 2005. Chemical analysis of floor sediments. In I. Hodder, ed., Inhabiting Çatalhöyük: Reports from the 1995-99 Seasons. Cambridge: McDonald Institute/ British Institute at Ankara: 399-412. Wilkinson T.J. 1990. Town and Country in SE Anatolia I. Settlement and Land Use at Kurban Höyük and other Sites in the Lower Karababa Basin. Chicago: Oriental Institute.
List of Figures Muddy boots of CAP team members at the project base after a day in the field. XV Fig. 2.1: Average monthly temperatures and precipitation for Bartın (after Alex 1985: 55). 10 Fig. 2.2: δ18O ‰ VPDB and δ13C ‰ VPDB isotopic measurements from Sofular cave from 14,000 BP (right) up to 2006 (source data: www.ncdc.noaa.gov). 11 Fig. 2.3: Hypothetical reconstruction of the coastline of the Cide-Şenpazar region at -40 metres (ca. 10,000 BC) and -20 metres (ca. 6000 BC) in relation to the modern coastline (from ca. 3000 BC). Based on bathymetric data. Produced by Victor Klinkenberg. 15 Fig. 2.4: Photograph from the 1950s showing pier and water taxis to the steamboat at Cide. From the album of Fikriye Doğan. Provided by Nesrin Şahin and Murat Karasalihoğlu. 19 Fig. 2.5: Sea currents in the Black Sea. Produced by Victor Klinkenberg. 21 Fig. 2.6: Photograph from the 1950s: Cide town and surrounding landscape. Provided by Nesrin Şahin and Murat Karasalihoğlu. 22 Fig. 2.7: Geological makeup of the Cide and Şenpazar districts (from Uğuz et al. 2002 and Aksay et al. 2002) and the distribution of villages and hamlets (stars) in the research area. 24 Fig. 2.8: Photograph from 1939 showing a newly constructed boat from Cide town being hauled to the sea by a group of men and water buffalos. Provided by Nesrin Şahin and Murat Karasalihoğlu. 25 Fig. 2.9: Locations of geoarchaeological investigations in the Cide-Şenpazar region. Produced by Toby C. Wilkinson. 27 Fig. 3.1: Map of the Cide and Şenpazar region showing the main locations of towns and landmarks discussed in this chapter. Produced by Toby C. Wilkinson. 50 Fig. 4.1: Map showing the CAP survey area. Produced by Toby C. Wilkinson. 58 61 Fig. 4.2: Surveying parts of Okçular valley affected by erosion (CAP-P1/388). Fig. 4.3: Map showing areas covered by forest in the survey region. Produced by Toby C. Wilkinson. 61 Fig. 4.4: Map showing areas above and below 15 degrees of slope. Produced by Toby C. Wilkinson. 62 Fig. 4.5: Map showing CAP’s ‘effective’ survey region. Produced by Toby C. Wilkinson. 63 Fig. 4.6: Data-entry and artefact documentation on the top floor of the Yalı Hotel. Photo by Imke van 64 Hagen. Fig. 4.7: Recai Yılmaz talking to Michele Massa in the Gideros İni I cave, which he brought us to in 72 2009. Photo by Renate de Boer. 73 Fig. 4.8: An important invitation to drink çay by locals at Kuşçu (CAP-P2/0033). Fig. 4.9: Diagram showing (a) traditional linear transects, (b) undulating transects and (c) CAP’s 76 meandering approach. Fig. 4.10: Google Earth satellite image of Okçular valley, showing the large-scale erosion on the 78 eastern valley flanks. Fig. 4.11: Map of Okçular valley showing units surveyed in 2009-2011. Produced by Toby C. 79 Wilkinson. Fig. 4.12: Map of the Loç valley showing a combination of arbitrary unit and survey areas defined by 80 field boundaries and other landscape features. Produced by Toby C. Wilkinson. 81 Fig. 4.13: Location of CAP grid-squares in Okçular valley. Produced by Toby C. Wilkinson. Fig. 4.14: Binary diagram of Ba/Zr-La/Th ratios for the Yağlar-Sakaeli/Orta obsidian outcrops and for the three studied artefacts. This diagram shows that one of the artefacts (L197) comes from the Yağlar region, while the two others (L136 and L224) come from the Sakaeli/Orta district. 89 Produced by Bernard Gratuze.
List of Figures
471
Fig. 4.15: Map showing soil sampling locations in grid G6 superimposed over the sherd count 94 distribution showing high density in dark shades. Fig. 5.1: Early Holocene finds locations in the CAP survey area. Produced by Toby C. Wilkinson. 111 Fig. 5.2: The distribution of Prehistoric chipped stone artefacts in the Okçular landscape. Produced 113 by Victor Klinkenberg. Fig. 5.3: The distribution of Prehistoric chipped stone artefacts in the Abdulkadir landscape. 113 Produced by Victor Klinkenberg. Fig. 5.4: Chipped stone cores found in the Okçular and Abudulkadir valleys. Produced by Lorraine 115 McEwan. Fig. 5.5: Chipped stone tools found in the Okçular and Abdulkadir valleys. Produced by Lorraine 116 McEwan. 119 Fig. 5.6: Chipped stone cores found at Aybasan Mahallesi. Produced by Lorraine McEwan. Fig. 5.7: Size distribution of the chipped stone artefacts found at Aybasan Mahallesi. Produced by 120 Michele Massa. 121 Fig. 5.8: Chipped stone tools from Aybasan Mahallesi. Produced by Lorraine McEwan. Fig. 5.9: Flint knife and chisel found at site S132 in the Çamdibi valley. Produced by Lorraine 124 McEwan. Fig. 5.10: C293 found in Unit 3. Handmade, mineral tempered. Produced by Lorraine McEwan. 127 129 Fig. 5.11: Plan of Okçular Kale İni (S24). Produced by Michele Massa and Victor Klinkenberg. Fig. 5.12: View of section through midden deposits at the back end of Okçular Kale İni (S24) containing charcoal, bone, ash, lithics, and pottery in horizontal orientation. In the foreground 130 Imke van Hagen, project member of CAP. Photo by Victor Klinkenberg. Fig. 5.13: Chalcolithic ceramics found at Okçular Kale İni (S24). Produced by Lorraine McEwan. 133 Fig. 5.14: Selected chipped stone artefacts from Okçular Kale İni (S24). Produced by Lorraine 135 McEwan. Fig. 5.15: A selection of Prehistoric sherds from Derebağ Köy Mağarası (S188). Produced by Lorraine 137 McEwan. Fig. 5.16: Chalcolithic ceramics and chipped stone artefacts found at Kılıçlı Mağarası (S88/S93). 139 Produced by Lorraine McEwan. 153 Fig. 6.1: Early Bronze Age sites in the CAP survey area. Produced by Toby C. Wilkinson. Fig. 6.2: Plan of the upper chamber of Kılıçlı Mağarası (S88). Hatched areas are spoil heaps, illicit 154 excavations are marked in black. Produced by Michele Massa and Victor Klinkenberg. Fig. 6.3: Early Bronze Age pottery from Kılıçlı Mağarası. Produced by Lorraine McEwan. 155 157 Fig. 6.4: Early Bronze Age pottery from Kılıçlı Mağarası. Produced by Lorraine McEwan. 158 Fig. 6.5: Early Bronze Age pottery from Kılıçlı Mağarası. Produced by Lorraine McEwan. 160 Fig. 6.6: Early Bronze Age pottery from Kılıçlı Mağarası. Produced by Lorraine McEwan. Fig. 6.7: Early Bronze Age pottery from Kılıçlı Mağarası. Produced by Lorraine McEwan. 161 162 Fig. 6.8: Early Bronze Age pottery from Kılıçlı Mağarası. Produced by Lorraine McEwan. Fig. 6.9: Early Bronze Age pottery from Okçular Kale İni and Derebağ Köy Mağarası. Produced by 164 Lorraine McEwan. Fig. 6.10: Early Bronze Age pottery from Okçular Kale İni and Derebağ Köy Mağarası. Produced by 166 Lorraine McEwan. 170 Fig. 6.11: Kılıçlı Mağarası, lower chamber (S93) (CAP-P2/1001). 170 Fig. 6.12: Downward climb into Kılıçlı Mağarası, upper chamber (S88) (CAP-P2/1010). Fig. 7.1: Map showing the locations of second millennium BC finds. Produced by Toby C. Wilkinson. 185 187 Fig. 7.2: Second millennium BC pottery from the Cide region. Produced by Lorraine McEwan.
472
List of Figures
Fig. 7.3: Second millennium BC pottery from the Cide region. Produced by Lorraine McEwan. 188 Fig. 7.4: Distribution of second millennium BC pottery in the Okçular valley. Produced by Toby C. 189 Wilkinson. Fig. 7.5: View north from Okçular valley showing the locations of grids G1 and G8 (CAP-P2/916). 190 191 Fig. 7.6: Late Bronze Age pottery from the Cide region. Produced by Lorraine McEwan. Fig. 7.7: Late Bronze Age pottery from the Cide region. Produced by Lorraine McEwan. 191 193 Fig. 7.8: ‘Okçular ware’ diagnostic types. Produced by Lorraine McEwan. 194 Fig. 7.9: ‘Okçular ware’ diagnostic types. Produced by Lorraine McEwan. Fig. 8.1: Map of Cide and Şenpazar showing the locations of Early Iron Age finds. Produced by Toby 223 C. Wilkinson 224 Fig. 8.2: Early Iron Age pottery from Cide. Produced by Lorraine McEwan. Fig. 8.3: Early Iron Age cup and handle from Derebağ Köy Mağarası (S188). Produced by Lorraine 225 McEwan. Fig. 8.4: Map of Cide and Şenpazar showing the locations of Middle Iron Age finds. Produced by 226 Toby C. Wilkinson. 226 Fig. 8.5: Middle Iron Age pottery from Cide. Produced by Lorraine McEwan. Fig. 8.6: Map of Cide and Şenpazar showing the locations of Late Iron Age finds. Produced by Toby C. 228 Wilkinson. 228 Fig. 8.7: Late Iron Age pottery from Cide and Şenpazar. Produced by Lorraine McEwan. Fig. 8.8: Wall built of large dry-stone blocks at the base of Gideros Kale West (S12). (CAP-P1/1092). 230 231 Fig. 8.9: Derebağ Köy Mağarası (S188) (CAP-P15/300). Fig. 9.1: Map showing the locations of Hellenistic-period finds. Produced by Toby C. Wilkinson. 250 251 Fig. 9.2: Hellenistic pottery from Cide-Şenpazar. Produced by Lorraine McEwan. Fig. 9.3: Hellenistic pottery from Cide-Şenpazar. Produced by Caner Bakan. 252 253 Fig. 9.4: Possible Hellenistic roof tiles from Cide-Şenpazar. Produced by Lorraine McEwan. Fig. 10.1: Roman-period sites in the Black Sea area mentioned in the text. Produced by Toby C. 261 Wilkinson. 268 Fig. 10.2: A selection of Roman-period sherds. Produced by Lorraine McEwan. Fig. 10.3: Concentrations of predominantly Roman-period ceramic finds in the study area. Produced 273 by Toby C. Wilkinson. Fig. 10.4: Cluster of robbed stone cist graves on the Gökçekale promontory. Part of a larger cemetery. 273 Produced by Lorraine McEwan. Fig. 10.5: Basin of a possible (lever?) press installation, presumably for the production of olive oil, at 274 Gökçekale (S176; location 1) (CAP-P1/5111). Fig. 10.6: Quarry face along the coast, between Gökcekale and Yenice (S182; between locations 1 275 and 2) (CAP-P15/268). 275 Fig. 10.7: One of two cist graves near Yenice (S170; location 2) (CAP-P1/5091). Fig. 10.8: A base or top moulding, perhaps from a podium, from the bay of Gideros (S123; location 276 3a) (CAP-P2/1856). Fig. 10.9: Upper moulding, possibly of a podium, east of Kalafat (S160; location 5) (CAP-P1/5033). 277 278 Fig. 10.10: A (marble?) cistern head, from around Okçular (S61; location 6) (CAP-P1/404). Fig. 10.11: Marble (?) Ionic capital, of Roman date, between Irmakköy and Cide (S150; location 9a) 280 (CAP-P1/4943).
List of Figures
473
Fig. 10.12: Dressed blocks found between Irmakköy and Cide (S152; location 9a) (CAP-P1/4952). On the left a possible top moulding, the other two blocks could be elements of a door frame. 280 Fig. 10.13: Top part of an Early Roman (?) tombstone, reading ΓΛΑΥΚΟΣ ΜΕΝΕΦΡΟΝΟΣ ΚΡωMΝΕΙΤΕΣ 281 (Glaukos, son of Menephron, from Kromna) (S208; location 9b?) (CAP-P15/402). Fig. 10.14: Photograph of a stone basin, with a wheel-like engraving on the short end and an outlet, 283 perhaps a reused sarcophagus (S126; location 11) (CAP-P1/4890). 283 Fig. 10.15: A ceiling cassette fragment from Kuşçu (S89; location 13) (CAP-P2/940). Fig. 10.16: Bridgehead along the Devrekani at Atköprü. Note the wall in the lower part of the picture, 285 which is part of the bridgehead on the eastern side (S95) (CAP-P3/262). Fig. 10.17: Fragments of narrow-necked amphorae from Cide, probably locally manufactured (S196; 287 location 10b) (CAP-P1/5898). 303 Fig. 11.1: Map of Byzantine settlements. Produced by Toby C. Wilkinson. 304 Fig. 11.2: Byzantine sherd density map. Produced by Toby C. Wilkinson. 304 Fig. 11.3: Byzantine tile density map. Produced by Toby C. Wilkinson. Fig. 11.4: Glazed table wares. Produced by Lorraine McEwan. 306 308 Fig. 11.5: Unglazed table wares. Produced by Lorraine McEwan. 310 Fig. 11.6: Kitchen wares. Produced by Lorraine McEwan. 311 Fig. 11.7: Amphorae and tiles. Produced by Lorraine McEwan. Fig. 11.8: A reconstruction of Okçular Kale (S22). Produced by Victor Klinkenberg and Michele Massa. 313 313 Fig. 11.9: Column base (S61) in Okçular village (CAP-P1/404). 314 Fig. 11.10: Column Base (S77) in Okçular village (CAP-P2/699). 315 Fig. 11.11: View of the valley from Okçular Kale (S22) (CAP-P-3/47). 315 Fig. 11.12: Mortared Field Stones, Okçular Kale (S22) (CAP-P1/131). 316 Fig. 11.13: Beam hole, Okçular Kale (S22) (CAP-P1/128). 318 Fig. 11.14: Silo or water feature, Okçular 4 (S80) (CAP-P1/1223). Fig. 11.16: Possible storage pit, Okçular 6 (S83) (CAP-P2/848). 320 321 Fig. 11.17: Map of Okcular tile density. Produced by Toby C. Wilkinson. 323 Fig. 11.18: Carved stone fragment, Çamdibi 12 (S122) (CAP-P1/4885). 323 Fig. 11.19: Stone basin, Çamdibi 2 (S126) (CAP-P1/4890). 326 Fig. 11.20: Map of Byzantine fortifications. Produced by Toby C. Wilkinson. Fig. 11.21: Interior walls of rectangular structure at Kazallı Kalesi (S91). Rows of mortared tile and 327 stone. Possible Late Roman construction (CAP-P1/1507). Fig. 11.22: Exterior of outer wall at Kazallı Kalesi (S91). Several courses of stone interspersed with 327 single rows of tile. Possible Middle Byzantine construction (CAP-P1/1511). Fig. 11.23: External wall at Timle 1 (S110). Well-faced mortared stones. Possible Late Roman 328 construction (CAP-P1/2673). Fig. 11.24: Upper levels at Gideros (S12). Loosely mortared fieldstones. Possible Early Byzantine 329 rebuilding (CAP-P2/1578). Fig. 11.25: External wall at Aydos Kale (S190). Stone courses interspersed with tile courses. Possible 330 Middle Byzantine construction (CAP-P15/313). Fig. 11.26: View of later walls at Kazallı Kalesi (S91). Probable Middle Byzantine construction 330 (CAP-P1/1503). Fig. 11.27: Gideros Doğu Kalesi (S121). Loosely mortared fieldstones. Probable Early Byzantine 332 construction (CAP-P2/1850). Fig. 11.28: Outer walls at Kazallı Kalesi (S91). Possible Early or Middle Byzantine construction 334 (CAP-P2/992). Fig. 11.29: Plan of Timle Kale 1 (S110) and Timle Kale 2 (S112). Produced by Michele Massa and Victor 334 Klinkenberg.
474
List of Figures
335 Fig. 11.30: Internal walls at Timle Kale 1 (S110). Well-faced mortared walls (CAP-P1/2679). Fig. 11.31: Lower external wall at Timle Kale 2. Well-faced stone construction (S112) (CAP-P1/2693). 336 336 Fig. 11.32: Later rebuilding at Timle Kale 2 (S112) (CAP-P1/2691). 337 Fig. 11.33: Plan of Aydos Kale (S190). Produced by Victor Klinkenberg. Fig. 11.34: External wall at Aydos Kale. Stone with sparse tile layers. Probable Middle Byzantine 337 construction (S190) (CAP-P15/319). Fig. 11.35: Interior Herringbone patterning at Aydos Kale (S190). Probable Middle Byzantine 338 construction (CAP-P15/344). Fig. 11.36: External wall at Hıdır Kalesi (S62). Stone and tile courses. Probable Middle or Late 339 Byzantine construction (CAP-P1/409). 339 Fig. 11.37: Detail of construction techniques at Hıdır Kalesi (S62) (CAP-P1/407). 340 Fig. 11.38: Plan of Akça Kale (S135). Produced by Michele Massa. 341 Fig. 11.39: Internal apsidal wall at Akça Kale (S135) (CAP-P15/390). 342 Fig. 11.40: External wall at Koca Kale (S154) (CAP-P1/4978). 342 Fig. 11.41: Detail of mortar at Koca Kale (S154) (CAP-P1/4985). 343 Fig. 11.42: Map of Byzantine religious sites. Produced by Toby C. Wilkinson. 344 Fig. 11.43: In situ column at Kalafat Mezarlık (S116) (CAP-P2/1287). Fig. 11.45: Plan of the church complex of Çadır İni Kilise (S21) and Çadır İni (cave below, S20). 345 Produced by Michele Massa and Victor Klinkenberg. 346 Fig. 11.46: Apse of the Çadır İni Kilise (S21) (CAP-P2/62). 347 Fig. 11.47: In situ chancel screen at Çadır İni Kilise (S21) (CAP-P2/72). 347 Fig. 11.48: Grave in the nave of the Çadır İni Kilise (S21) (CAP-P2/1311). 348 Fig. 11.49: Architectural fragment at Çadır İni Kilise (S21) (CAP-P2/1305). 348 Fig. 11.50: Architectural fragments at Çadır İni Kilise (S21) (CAP-P2/1308). 350 Fig. 11.51: Çadır İni (S20) (CAP-P2/79). 350 Fig. 11.52: Carved benches in Çadır İni (S20) (CAP-P2/84). Fig. 11.53: Plan of Gideros Mağarası I (S59). Produced by Michele Massa and Victor Klinkenberg. 351 351 Fig. 11.54: Crosses in Gideros Mağarası I (S59) (CAP-P1/358). 352 Fig. 11.55: Crosses in Gideros Mağarası I (S59) (CAP-P1/359). 352 Fig. 11.56: Crosses and inscriptions in Gideros Mağarası I (S59) (CAP-P1/365). Fig. 11.57: Plan of Uçakdibi Mağarası (S40). Produced by Michele Massa and Victor Klinkenberg. 353 354 Fig. 11.58: Plan of Ballıcı İni (S17). Produced by Michele Massa and Victor Klinkenberg. Fig. 11.59: Ballıcı İni (S17) (CAP-P2/45). 355 357 Fig. 11.60: Spolia (possible chancel screen) from Kuşçu 2 (S89) (CAP-P2/936). 357 Fig. 11.61: Spolia (basin) from Gideros 2 (S123) (CAP-P2/1855). Fig. 12.1: Map showing the location of Çoban Kalesi (S32) in relation to other kales and 364 contemporary sites. Produced by Toby C. Wilkinson. 365 Fig. 12.2: Schematic plan of Çoban Kalesi (S32). Produced by Victor Klinkenberg. 365 Fig. 12.3: A view of the fortifications of Çoban Kalesi (S32) (CAP-P2/198). 367 Fig. 12.4: Sgraffito sherds from Çoban Kalesi. Produced by Lorraine McEwan. Fig. 12.5: Selected sherds and small finds from Çoban Kalesi (S32). Produced by Lorraine McEwan. 370 Fig. 13.1: Map showing locations and sites mentioned in the text. Produced by Toby C. Wilkinson. 375 380 Fig. 13.2: Traditional houses at Mencekli Köyü Yukarı Mahalle (CAP-P2/12). 381 Fig. 13.3: Pipe heads and fragments. Produced by Lorraine McEwan.
List of Figures
475
Fig. 13.4: Ottoman-period gravestones from Kumluca (A: CAP-P1/5047 and B: CAP-P1/5049), (C) 383 Çayyakka (CAP-P2/749) and (D) Gideros East (CAP-P2/1827). Fig. 13.5: Central pillar of the old bridge over the Devrekani near Çamdibi, Loç valley (CAP-P2/1657). 384 386 Fig. 13.6: Çilekçe Köprüsü (CAP-P2/1371). 387 Fig. 13.7: İlyasbey water installation (CAP-P15/289). 388 Fig. 13.8: Traditional stone-built bread oven, Çamdibi (CAP-P2/1640). Fig. 13.9: Recorded finds locations of Seljuk and Ottoman date in the CAP survey area. Produced by 388 Toby C. Wilkinson. Fig. 14.1: Map showing the CAP survey area and the three landscapes discussed in this chapter (1) 393 Okçular, (2) Abdulkadir-Gideros, and (3) Çamdibi. Produced by Toby C. Wilkinson. Fig. 14.2: A view of Okçular Kale from the interior (S22) (CAP-P1/157). 394 Fig. 14.3: A view from the modern cave floor of the narrow, elevated entrance passage of Okçular 395 Kale İni (CAP-P1/400). Fig. 14.4: Main archaeological features and clusters in the Okçular valley. Produced by Toby C. 396 Wilkinson. Fig. 14.5: View northeast along the coast to Cide, the rock outcrop of Okçular Kale in mid-distance 397 (CAP-P2/665). Fig. 14.6: Limestone ridges defining the southern boundary of Okçular valley (CAP-P2/920). 398 399 Fig. 14.7: View shed from the top of Okçukar Kale. Produced by Toby C. Wilkinson. Fig. 14.8: Okçular Kale İni, entrance area to a second alcove with pottery dating to the third 402 millennium BC (S24) (CAP-P2/1056). Fig. 14.9: View northeast onto the Kalafat coastal plain from the natural hill where we found a 403 concentration of Late Bronze Age pottery (CAP-P2/1754). 405 Fig. 14.10: Examples of iron ore from U593 (CAP-P1/4650). Fig. 14.11: Distribution of Roman and Byzantine period surface clusters and features in the Okçular 407 valley. Produced by Toby C. Wilkinson. 408 Fig. 14.12: Okçular field-boundaries (CAP-P2/917). 410 Fig. 14.13: Stone-carved water-conduit associated with S105 (CAP-P2/1410). 412 Fig. 14.14: Gideros bay (CAP-P2/1801). Fig. 14.15: Map of Abdulkadir and Gideros showing the location of features and clusters. Produced by 414 Toby C. Wilkinson. Fig. 14.16: View east towards Abdulkadir village, modern marble quarry visible in the southwest 415 corner (CAP-P2/655). Fig. 14.17: Buffalos and horses grazing in the marshy valley bottom at Abdulkadir valley 415 (CAP-P2/553). 417 Fig. 14.18: Plan of Ballıcı İni cave (S17). Produced by Victor Klinkenberg and Michele Massa. 419 Fig. 14.19: Submerged landings at Gideros East (S121) (CAP-P2/1841). 420 Fig. 14.20: Rock-cut stairs at Gideros West (S12) (CAP-P2/1794). Fig. 14.21: Rock-cut grave (S66) in boulder northwest of Abdulkadir village. Concentrations of tile 421 and pottery were found on the natural hill to the south (CAP-P2/603). 422 Fig. 14.22: Kurçalı Tepesi (S16) (CAP-P2/657). Fig. 14.23: Roman and Byzantine features and surface clusters in the Abdulkadir-Gideros region. 424 Produced by Toby C. Wilkinson. 425 Fig. 14.24: Stone-built architecture at Gideros East. (CAP-P2/1799). 426 Fig. 14.25: View south across part of the Çamdibi/Loç valley (CAP-P2/1433). 429 Fig. 14.26: Large, dome-like forehall of Kılıçlı Mağarası (S93) (CAP-P2/1017). Fig. 14.27: View north across the Çamdibi/Loç valley from the present entrance of Kılıçlı Mağarası 430 (S93) (CAP-P2/1610).
476
List of Figures
448 Fig. 16.1A: The CAP site recording form. 449 Fig. 16.1B: The CAP site recording form. 450 Fig. 16.2: The CAP Unit form. 451 Fig. 16.3: The CAP Grid form. Fig. 16.4: Principal components analysis plot of all samples for all fourteen elements according to (a) 453 sample number and (b) approximate date. Fig. 16.5: Ca-Sr plot of samples (except C3186 and C4023) from (a) Okçular (O), Abdulkadir (A), Çamdibi (C) and Derebağ (D), (b) according to date and (c) according to sample number. 456 457 Fig. 16.6: Photomicrographs (x6.3, xpl) of (a) C803 and (b) C2856 from Okçular. Fig. 16.7: Photomicrographs of (a) C3186, (b) C3646 and (c) C3740 from Çamdibi (magnifications all 458 in the range x6.3-8.5; all ppl). Fig. 16.8: Photomicrographs (x6.3; ppl) of (a) C4169 and (b) C4228 from Derebağ Köy Mağarası. 459 Fig. 16.9: The sample locations in G6 superimposed over the sherd count distribution showing high 460 density in dark shades. Fig. 16.10: Phosphate (blue) and Magnetic Susceptibility (red) values in 12-37 and including the 462 values for 8 which is labelled 14a. Fig. 16.11: Phosphate and MS enhancements highlighted on the sherd distribution plan in G6. 462 463 Fig. 16.12: The calcium distribution for all samples. Fig. 16.13: (a) Zr/Ca and Sr/Ca ratios and (b) Zn/Ca and Cu/Ca ratios in the sample sequence 12-27. 467 Fig. 16.14: Locations of the ‘hotspots’: phosphate and MS in black (as from Fig. 16.11); multi-element 468 composition in yellow (dotted yellow for uncertain enhancement).
List of Tables Tab. 1.1: Number of field days, size of the team and budget for each of the seasons of the Cide 7 Archaeological Project. Tab. 2.1: Combined description of two profiles in the Cide coastal plain (S67 and S71). Regression 29 phases indicated in italics. Tab. 2.2: Landscape types in the Cide-Şenpazar region, the geological processes that occur in them, 31 and how this may affect archaeology. 48 Tab. 3.1: Economic activities in the Cide region mentioned in travellers’ accounts. Tab. 4.1: Letter codes/abbreviations used in the field, in publication and in the CAP database. 66 67 Tab. 4.2: Classes of sites identified in CAP and their characteristics. 69 Tab. 4.3: Surface collection area-units. 75 Tab. 4.4: List of interview questions. 80 Tab. 4.5: Intensive fieldwalking: number of units and area walked. 82 Tab. 4.6: CAP grid-square collection areas. 84 Tab. 4.7: Abbreviations used for the description of vessel types in the CAP database. 85 Tab. 4.8: Ceramic samples exported for analysis (see table 4.10 for period abbreviations). 87 Tab. 4.9: Main database fields and their explanation for the description of lithic artefacts. 90 Tab. 4.10: Period designations and chronological ranges. Tab. 5.1: Period terminologies used in this chapter compared to traditional terminologies employed 105 in Anatolian archaeology. 131 Tab. 5.2: Five Prehistoric ‘wares’ found at Okçular Kale İni (S24). 133 Tab. 5.3: Radiocarbon dates from Okçular Kale İni (S24). 262 Tab. 10.1: Chronological framework used in this chapter. Tab. 10.2: Absolute counts and relative proportions of the pottery discussed in this chapter, by 264 functional category, and type and/or fabric. 356 Tab. 11.1: Spolia from the Cide Archaeological Project. 434 Tab. 15.1: Periods identified in the course of CAP. 447 Tab. 16.1: Summary of luminescence dating results. 452 Tab. 16.2: Pottery samples analysed chemically and petrographically. Tab. 16.3: Chemical compositions of the pottery expressed as weight percent element (Si, Al, Ca, Fe, 454 K) and ppm element (remaining elements). Tab. 16.4 Element concentrations (in ppm), phosphate (P) (mg/100 mg soil) and magnetic suscep 464 tibility (MS) determined at low and high frequencies
Index A Abant Gölü 10 Abdulkadir 29, 30, 46, 80, 83, 86, 111-114, 116-118, 120, 126, 221, 226, 249, 250, 251, 301, 307, 309, 312, 322, 324, 346, 349, 358, 360, 361, 393, 398, 412-416, 418, 421-425, 431, 436, 439, 442, 452, 455, 456 Abono 49 Achaemenid 221, 229, 246 Achaemenid empire 215 Aegialos 49 Aegialus 37, 38, 298, 438 African Red Slip Ware 271 Ağaçlı 104, 117 agriculture 22-24, 26, 31, 48, 60, 76, 126, 300, 359, 394, 404 Akalan 219 Akarca Sekileri 109 Akça Kale 340, 440 Alaça Höyük 137, 140, 151, 153, 156, 168, 189 Alacalıgöl 134 Alanya 369 al-Din Chupan, Husam 377 al-Din Khuyi, Husam 376, 377 al-Din Yavlak Arslan b. Alpyürek, Muzaffar 377 Alexander the Great 215, 216, 246 Alexios Komnenos 376 Alişar Höyük 56, 140, 151-153, 156, 212, 218, 219, 220 al-Umari 378 Amasra 18, 23, 36, 38, 42, 43, 48, 75, 173, 282, 376, 379, 385, 420 Amastris 38-40, 43, 44, 216, 247-249, 254, 263, 281, 282, 287, 296, 298, 302, 376 Amisos 219 Amorium 346 Amphorae 266, 309 Anaia/Kadı Kalesi 369 Anatolian Plate 12 Anemurium 299, 344 Anitta 195 Annals of Hattusili 196 Apollonius of Rhodes 38, 249, 282 Argonauts 38, 39, 48 Armstrong, Pamela 305 Arrian 39, 44, 276, 279, 284, 298 Arslan, Yavlak 378
Artaxerxes 215 Aşıklı Höyük 1 Atköprü 284, 285, 322, 426, 431 Aybasan 30, 80, 82, 86-88, 96, 110, 112, 117-122, 124-126, 142, 221, 222, 229, 249, 282, 360, 389, 436, 439 Aydos 30, 330, 335, 337, 338 Aydos Kale 22, 30, 49, 76, 328, 329, 335, 338, 340 Azak 368, 369, 371 B Balıbağı 168 Ballıcı İni 83, 111, 127, 309, 324, 349, 354, 355, 413, 416, 417, 423, 432 Baths of Zeuxippus 366 battle of Manzikert 376 Berezan 163 Beycesultan 132, 156, 159 Bijişkyan, P. Minas 44 Black Sea 3, 9-15, 17, 18, 20, 21, 23-25, 31, 36-39, 41, 42, 44, 46, 48-50, 53, 54, 57, 60, 62, 75, 76, 83, 91, 95, 96, 105-107, 109, 125, 141, 151, 152, 156, 159, 163, 167, 169, 171-173, 183-185, 194-204, 212-217, 219-221, 227, 231, 240, 246-249, 252, 254, 260, 261, 263, 264, 267, 269, 271, 277, 279, 286, 288, 289, 294, 297-301, 305, 309, 317, 329, 346, 358, 359, 364, 367, 368, 369, 371, 372, 376, 377, 378, 382, 400, 413, 418, 437, 438, 439, 441 Blaënê 196 Boğazköy 56, 91, 152, 156, 164, 184, 186, 189, 197, 198, 204, 212, 216-221, 223, 230 Bore, Eugene 46 Bosphorus 13, 14, 21, 25, 39, 372 Bourguignon dAnville, Jean Baptiste 43 boxwood 40, 49, 215, 248, 254, 287, 289, 385, 406, 424, 439 Boyabat-Kovuklukaya 151, 168, 169, 172, 195 Boztepe 156, 168 Bronze Age 2, 20, 59, 72, 73, 80, 81, 83, 90, 96, 104, 105, 107, 108, 110, 112, 123, 126, 127, 130, 132, 136-138, 140, 141, 151-153, 155-169, 171-174, 183-186, 189-204, 213, 216, 217, 219, 223, 231, 246, 393, 401-404, 407, 416, 418, 428, 434, 435, 437, 441, 453, 455
Index Bryer and Winfield 297, 312, 325 Büyük Güllücek 108 Büyükkale 186, 189, 218, 223 Büyükkaya 107, 108, 132, 134, 136, 137, 189, 198, 217 Byzantine 3-6, 28-31, 41, 42, 49, 59, 67, 71, 73, 74, 77, 89, 90, 92, 96, 112, 118, 123, 128, 130, 136, 138, 169, 173, 262, 263, 275, 279, 282, 284, 289, 294-305, 307, 309, 312, 314, 317-322, 324-335, 337-339, 341, 343, 344, 346, 349, 355, 358, 359, 360, 366, 367, 369, 371, 372, 376, 378, 393, 394, 395, 397, 400, 404, 405, 406, 407, 409, 410, 411, 412, 413, 416, 418, 421-424, 427, 428, 431, 432, 434, 435, 437, 439, 442, 443, 469 Byzantine fortresses 296, 325 Byzantine White Ware 324, 431 C Çadır Höyük 107, 132, 138, 140, 212, 217, 218, 221, 299, 320 Çadır İni 127, 128, 275, 325, 344, 345, 349, 413, 417, 423, 425 Çadır İni Kilise 344 Caffa 371 Çamdibi 80, 111, 123, 124, 126, 138, 152, 221, 225, 226, 249, 282, 307, 309, 312, 321, 322-324, 341, 353, 358, 360, 361, 384, 388, 393, 394, 425-428, 430, 431, 436, 439, 442, 452, 455-458 Çamlıbel Tarlası 105, 106, 125 Candarids 378 Çandarlı Ware 266 Candaroğlu 41 Cape Carambis 49, 298 Cape Pisello 42, 49 Carambis 38-40, 42-44, 298 Carevgrad Tãrnov 367, 369 castle(s) 3, 28, 41, 44, 49, 50, 67, 71, 72, 92, 312, 364, 371, 376-378, 398, 424, 440 Çatalhöyük 1, 117, 123, 460 Çatalkaya 134 Catullus 40 Čavdar 123 cave(s) 5, 11, 14, 46, 67, 71-74, 78, 83, 88, 92, 109-111, 123, 124, 127, 128, 130, 136-138, 140, 142, 151-153, 156, 159, 163, 167-169, 171-173, 185, 192, 196, 197, 212, 221, 222, 226, 230-232, 317, 322, 324, 345, 346, 349, 393-396, 400, 401, 403, 404, 407,
479
409, 413, 416-418, 422, 423, 426, 428-431, 436-438, 440, 452, 457 Cembalo 369, 371 Çengeltepe 140 Cernavoda 161 Chalcolithic 59, 72, 73, 83, 90, 96, 104-107, 110, 111, 117, 122, 125-128, 130-137, 139-142, 151, 167, 169, 171, 192, 203, 394, 396, 401, 402, 407, 416, 417, 428, 429, 434-437, 441, 455 chemical composition 83, 84, 88, 167 Chepni 376 Cherries 23 Chersonesos 368, 369, 371 Chian amphorae 267 church(es) 28, 44, 74, 92, 260, 282, 297, 312, 343-346, 349, 355, 358, 387, 411, 413, 423, 432, 440, 442 Cide coastal plain 15, 22, 26-29, 92, 97, 111, 420, 422, 445 Cimmerians 213 Circumpontic 13, 31, 104, 250 Claudius Ptolemy 40 climate 9, 10, 11, 23, 105, 385 cluster 23, 28, 30, 66, 67, 81, 87, 93, 97, 111, 200, 201, 273, 324, 325, 397, 400, 407, 408, 421, 431 coal 12, 26, 48, 75, 385, 440 Çobanids 377, 378 Çoban Kalesi 36, 49, 71, 329, 333, 364, 366, 368-372, 379 Constantinople 45, 47, 260, 271, 286, 301, 302, 305, 366, 372 copper 26, 384 Corinth 369 Crimea 21, 39, 44, 75, 232, 250, 369, 372, 377, 378, 382 Cromna 37-40, 44, 247, 249 Curtea de Argeş 367 Cyotorus 39, 40 Cyprus 20, 57, 77, 106, 199, 271 Cytorum 38, 40, 274 Cytorus 37-40, 43, 44, 47, 274 D Daday Çevresi 109 Dahara river 196 Danishmendid 302, 377 Dardanelles 14 De Clavijo, Ruy Gonzale 41 de Marigny, Edouard Taitbout 44 Demircihöyük-Sarıket 156, 168
480
Index
Demircihüyük 122, 123, 126, 132 Denizkonak 49 Derebağ Köy Mağarası 73, 83, 127, 128, 136, 137, 142, 152, 164-167, 171, 185, 186, 221, 222, 225, 230-232, 436-438, 455, 459 de Tournefort, Joseph Pitton 42 Devrekani 30, 123, 168, 172, 196, 284, 285, 384, 389, 425, 428, 431, 432 Devrez 1, 2, 3, 200 Devrez Çay 2, 196 Dikili Tash 161 Diocletian 299 Domanitis 196 Dündartepe 132, 140, 167, 195, 199 dynatoi 301 E Early Bronze Age 151, 153 Early Byzantine 93, 275, 295, 300, 303, 305, 309, 312, 316, 317, 320, 322, 324, 325, 326, 331, 353, 358 Early Eocene 12 Early Holocene 13, 96, 105, 110, 111, 126 earthquakes 12 Eastern Sigillata C 266 echinus 265 echinus bowl 405 ecological zones 5, 6, 9, 59, 70 ecology 9, 10, 31 Egypt 19, 23, 378 empire of Nicaea 296 empire of Trebizond 302, 331 Eneti 37 Epipaleolithic 104, 105 erosion 12, 29-31, 60, 61, 67, 77-79, 82, 95, 106, 109, 111, 112, 118, 168, 396, 398, 412, 414 Erythini 37, 38, 39 Eskiyapar 220 Euchaita/Avkat 296 Evliya Çelebi 42, 47 Eylekderesi 154, 158, 160 Ezero 161 F Fakas 45, 49, 118 Feature(s) 67, 89 field methodology 71, 95, 396 Fikirtepe 1, 91, 125 Filyos culture 172 fish Bonito (palamut), bluefish (lüfer), and anchovy (hamsı) 24, 50, 198, 269, 405, 439
fishing 24, 26, 50, 75, 126, 173 flax 23, 198 flintknapping 87 flood 13, 14 Folteti 161 forest cover 12, 59, 60, 98 Foss and Winfield 325, 328 frontier 2, 196, 202, 288, 376 G Galatia 262 Galatian sources 88, 118, 125, 126, 135, 136, 141, 402 game 400 Gâvurevleri 160, 189, 196, 200 Geduz 42, 49 Genoese 20, 36, 41, 42, 302, 366, 370-372, 379 geoarchaeological survey 27, 30, 59 Geographia 298 Geology 12 Germanikopolis 263 Gerze 200 Ghidrus 43 Gideros 14, 18, 26, 28, 38, 42, 43, 46, 49, 72, 215, 229, 230, 232, 246, 248-250, 252, 274, 276, 298, 325, 328, 329, 331, 332, 344, 349, 351, 352, 356, 357, 376-378, 382, 383, 389, 393, 412-414, 416-426, 432, 436, 438-440 Gideros Mağarası I 127, 128, 349, 423 Gira Petra 371 GIS 54, 55, 63, 65, 67-69, 78, 79, 81, 82, 87 Glaukos 279 Gökçekale promontory 249, 272, 344, 439 Gökırmak 196, 218, 220, 221 Golden Fleece 38, 39, 48 Gölköy Enstitüsü 109 Gordion 212-214, 217, 220 Greek 14, 20, 36, 37, 39, 40, 42-44, 48, 49, 100, 212-216, 218-220, 227, 229, 232, 246-248, 252, 254, 349, 376, 377, 385, 425, 438, 439, 440 Greek colonies 227 Greek colonisation 213 grid-square(s) 63, 68, 70, 81, 82, 93, 96, 110, 127, 204, 393, 397, 403 grooved-rim bowl 404 group 25, 67, 73, 169, 203, 221, 264, 267, 269, 271, 272, 282, 287-289, 302, 307, 309, 368, 369, 382 Güllüavlu 156
Index Güvercinkayası 130, 142, 401 gyres 20 H Hadrianopolis 261, 262, 266, 271, 272 Harmantepe 156 Hasköy Höyüğü 161 Hattian 202 Hattusa 105, 184, 186, 197, 198, 202, 204 hazelnut 23, 400 Heinrich events 11 Hellenisation 214, 254 Hellenistic 5, 26, 83, 90, 215, 246-254, 260-262, 265-267, 270-272, 274-276, 282, 284, 309, 317, 322, 333, 405-407, 413, 418, 419, 421, 430-432, 434, 435, 437-439, 442, 453 Heraclea 38, 48, 219, 231, 249, 263, 276, 288, 302, 376, 378 Heraclius 295 Hermit Caves 343, 349, 423, 432 Hirschfeld, Gustav 294 Histira 163 Hittite 2, 105, 107, 195-204, 208, 213, 216, 435, 437 Hittite empire 2, 202, 213, 217 Hıdır Kalesi 49, 338 Hıdırlı 168, 195 Hoca Çeşme 123 Holocene 2, 10, 11, 12, 13, 14, 19, 59, 90, 96, 104, 105, 110, 111, 118, 122, 126, 127, 136, 141, 142, 398, 400, 401, 416, 417, 427, 432, 434, 436, 437, 443 Homer 37, 38 Hommaire de Hell, Xavier 45 horned handle horizon 134, 141 Horoztepe 153, 168 höyük 96, 167 Hüyükdoruğu 196, 200 hunter-gatherer 142 I Ibn Bibi 376, 377 Ibn Said 382 Ice Age 10, 23 İkiztepe 91, 106-108, 117, 132, 134, 140, 141, 151, 159, 167, 169, 194, 195, 199, 212, 219, 221 Ilgaz mountains 196, 219, 220, 221 Iliad 36, 37, 48, 438 Ilıpınar 1, 123, 125, 138 İlyasbey 18, 38, 49, 63, 325, 356, 387, 390 imperialism 2, 202 İnandıktepe 201
481
İnceburun 109 Incised Sgraffito Ware 368 İnebolu 42, 49, 75, 294, 379, 383-387 intensive survey 2, 6, 55, 57, 60, 65, 66, 68, 70, 76, 94, 96, 97, 106, 123, 204, 312, 320, 393, 416, 421, 422, 426 Intermediate/Cappadocian wares 152, 184 Ionian Black Sea colonies 219 Ionians 214 Ionopolis 43, 49, 248 Iron Age 5, 48, 73, 80, 81, 83, 90, 96, 110, 136, 137, 161, 165, 192, 197, 200, 202, 212-218, 220-232, 234, 236, 249, 253, 371, 404, 409, 418, 428, 430, 434, 435, 438, 442 Isfandiyardids 378 İsfendiyaroğlu 41 Istanbul 17, 18, 21, 22, 23, 25, 26, 42, 46, 75, 128, 305, 366, 372, 382, 386, 389, 411, 440 Izdebski, Adam 295, 296, 300, 301, 302, 316, 317 Izmit-Sapanca depression 14 J Jaupert, P. Amaedee 43 Justinian 263, 299, 326 K Kadarōs 43 Kāhintepe 154 Kaklık Mevkii 154, 159, 168 Kalafat 23, 28-60, 249, 276, 277, 307, 343, 344, 356, 358, 394, 403, 416 Kalafat Mezarlık 343, 358 kale 78, 312, 317, 318, 320, 331, 333, 335, 338, 358, 364, 371, 395, 397, 399, 400-404, 406, 409, 419, 437 Kaletepesi 221 Kalınkaya 168 Kaman-Kalehöyük 172, 217, 218, 219, 220, 221 Kanes 195, 196 Kanlıtaş 130, 142, 401 Kapitän II 267, 271, 277, 279 Kara Agatch 44 Karain 231 Karambis 37-39, 44, 248, 376 Kara Mustafa Camii 379 Karaoğlan 159 Kaska 2, 193, 196, 197, 198, 199, 201, 202, 203, 404, 443 Kastamonu XII, 3, 9, 12, 17, 18, 42, 47, 57, 58, 63, 65, 74, 75, 109, 119, 128, 134, 140, 151, 153, 154, 158-160, 167, 172, 189, 195, 196, 200,
482
Index
201, 212, 218, 286, 301, 302, 375-379, 380, 382, 384, 385, 390, 431, 438, 440, 443 kastra 294 kastron 299 Kavak 199 Kayabaşı 119, 134, 154, 196, 200 Kazallı Kalesi 49, 284, 326, 328, 329, 332, 334, 439 Kaz Gölü 10 Kerçe 42 Kerembe 43-45, 376 Kerkenes Dağ 212, 220 Kidros 43-45, 47 Kilise Tepe 192 kingdom of Nicaea 302, 331 kingdom of Trebizond 296 Kitoron 44 Kitross 44 Kılıçlı Mağarası 83, 123, 127, 128, 138-140, 142, 152, 154, 156, 159, 163, 165, 166, 171-173, 185, 186, 221, 227, 229, 324, 349, 403, 426, 428-432, 436, 437, 441, 455 Kınık 151, 158, 168, 172, 197, 201 Kızılırmak 194, 199, 203, 212, 214, 218-220 Klimax 333 Knidian amphora 267, 275, 421 Koca Dere 123 Koca Kale 329, 340, 431 Koçagöz 153, 168 Komana 296 Komnenian emperors 295 Komnenoi 301, 302, 329 Koumi 45 Köyiçi Tepesi 220 Kromna 39, 216, 248, 279, 376 Kudros 43 Küçükhüyük 159, 168 Küllüoba 151, 156 Kültepe 152, 189, 195, 196, 218 Kurçalı Tepesi 250, 422 Küre mines, mountains 26, 172, 198, 232, 384 Kuşçu 73, 250, 282, 283, 325-357 Kurucaşile 37, 376, 385 Kuşaklı 220 Kuşaklı-Sarissa 198 Kybele 231 Kyteros 37-39, 46, 48, 49, 215, 216, 229, 232, 246-249, 254, 274, 276, 287, 298, 331, 344, 438
L Ladik Gölü 10 Lake District 107 Lake Varna 173 Land of Hatti 196 landscape taphonomy 55, 92 landscape types 5, 59, 60, 70, 86 land-use 55, 60, 74, 79, 142, 399, 439 lapis-lazuli 195 Last Glacial Maximum 14 Late Bronze Age 2, 20, 90, 96, 163, 183-185, 190-204, 213, 216, 217, 219, 223, 231, 246, 403, 434, 435, 437, 455 Late Palaeocene 12 Late Roman 169, 261, 265, 267, 269, 270-272, 275, 289, 295, 300, 307, 312, 320, 322, 324-326, 331, 333, 343, 358, 407, 418, 421, 431, 439, 469 Late Roman Amphora 267, 270, 271, 277, 282, 288, 431 Late Roman C 266 Late Roman micaceous Aegean Ware 270 Late Roman Phocaean 421 Late Roman Phocaean Red Slip Ware 265 Leake, William Martin 43 lithics 30, 63, 67, 69, 81, 86, 87, 91, 104, 108, 110, 130, 140, 142, 401, 436 tools, blades, flakes, debitage, cores 86 Loç 49, 80, 123, 138, 152, 282, 384, 389, 393, 425-427, 431, 432 Lusta 369 Luwian 196 Lydia 214, 215 M Mahmatlar 168 Malyas 49 maquis 9, 414 Maşat Höyük 198, 218, 219, 220 marble 12, 26, 189, 278, 279, 401, 414-416, 455 marine snails 198 maritime trade 20, 21, 199, 406 Marmara region 1, 104, 106, 107, 117, 125, 141 Mediterranean 13, 14, 19, 20, 24, 45, 54-59, 66, 88, 91, 95, 97, 173, 198, 199, 216, 219, 232, 249, 260, 263, 265, 266, 271, 277, 286, 288, 289, 377, 405 Mersin 172 Mesolithic 90, 104, 105, 110, 117 methodology 4, 53, 57, 59, 76, 77, 81, 95, 107, 393, 397, 435
Index Middle Bronze Age 185 migration 18, 217, 412, 440 Miletos 219 mills 23, 47, 67, 385, 390 Mining 384 Ministry of Culture and Tourism XII, 1, 2, 3, 56 Mithridates I 247 Mithridatic Wars 262 Mızrak Mağarası 197, 231 Monochrome Sgraffito Ware 368 Muna 49 Mycenaean 20, 197, 198, 219, 246 N naphtha 378 Natural History 39 Neolithic 1, 13, 87, 90, 104-107, 117, 123, 125, 138, 435 Neolithisation 1 Nerik 200 Nesite 199 Nicomedes V 247 North Anatolian Fault Zone 12 O obsidian 1, 6, 19, 87, 88, 89, 118, 120, 122, 124-126, 135, 136, 141, 145, 146, 401, 402, 436, 441 Okçular Kale İni 78, 124, 127, 128, 131, 132, 134, 140, 142, 152, 165, 221, 222, 227, 230, 395, 396, 401, 407, 409 Okçular 23, 29, 30, 49, 61, 77-83, 85, 87, 88, 91-94, 96, 109, 111-118, 120, 125-136, 140-142, 164-166, 169, 173, 185, 189, 190, 192-194, 200, 201, 203, 204, 221, 222, 227, 229, 230, 249-251, 277, 278, 286, 301, 307, 309, 312-322, 324, 325, 340, 358-360, 389, 393-409, 411-414, 416, 418, 424-428, 431, 432, 436-440, 442, 452, 455-457, 460 Okçular Kale 49, 77, 82, 83, 87, 88, 91, 92, 94, 109, 111, 112, 125, 127, 128, 130-132, 134-136, 140-142, 165, 166, 169, 185, 189, 190, 192, 201, 204, 221, 222, 227, 230, 249, 277, 317, 318, 394-396, 398, 400-402, 404-409, 412, 413, 424, 428, 436-439, 442 Okçular ware 91, 192, 203 Old Assyrian 195, 196 Oluz Höyük 212, 217, 218 Orman Fıdanlığı 108, 117, 122, 123, 126, 134, 136, 137, 140 orography 9
483
Ottoman 25, 26, 41-44, 46-49, 90, 92, 96, 296, 312, 321, 325, 356, 371, 375, 377-379, 382, 383, 387-389, 410, 411, 424, 434, 440, 442, 443, 469 Ottoman empire 41, 387 Ovid 40 Oymaağaç Höyük 199, 200 oysters 198 P Pachymeres 376, 378 Pala 196 Palaeolithic 90, 104, 105, 109, 417 Paphlagonia 1, 17, 37, 39, 212, 213, 215, 216, 221, 246, 247, 261-263, 287, 292, 294-303, 316, 332, 371, 372, 438, 439 Parthenios 37, 38, 39 pāte blanche/claire 269 pāte rosée 269 pāte rouge/colorée 269 Pazarlı Tepe 217, 220 Pendik 1, 125 Periplus 37, 39, 43, 282, 298 Persia Achaemenid empire 216 Pervane Muin al-Din Suleyman 377 Pessinus 270 petrographic 83, 167, 397, 404, 436, 455, 457 Peutinger Atlas 298 Pharnaces 247 Phocaean LRC 271, 274, 275, 277, 279, 281, 286 Phrixus 38, 39 Phrygian grey ware 218, 226, 227 Phrygian kingdom 213 Pleistocene 11, 13, 109 Pliny 39, 40, 44, 263, 287, 298, 343 Polatlı 159 Pompeiopolis 260, 262, 266, 271, 296 Pompey 262 Pomponius Mela 39 Pontic mountains 9, 12, 17, 94, 173, 195, 198, 202, 203, 213, 215, 217, 294, 298, 300, 302, 303, 358, 400, 413, 437, 443 Pontic Red Slip Wares 266 Pontic refugium 23 Pontic Sigillata 266 Pontus 17, 104, 247 portolans 298 Pre- and Protohistoric 31 Prehistoric 80, 81, 83, 88, 92, 96, 104, 106-108, 110, 112, 113, 123, 126, 127, 130, 131, 137,
484
Index
138, 172, 317, 393, 396, 397, 401, 403, 416, 423, 426, 427, 435, 453 Prehistory 1, 88, 104, 106, 128, 138, 400, 427 probabilistic sampling 54, 59 Project Paphlagonia 1, 2, 57, 217, 219, 221, 261 Pseudo-Skylax 37 Pylaemenes 37 Q Qipchaq 376 Quintus Curtius Rufus 215 R Red-Black Burnished Ware 132 Renell, James 43 residue analyses 83 Rhodian amphorae 267 rock-cut grave 421 Roman 3-5, 17, 26, 28-30, 36, 37, 40, 42, 48-50, 59, 67, 73, 74, 89, 90, 93, 96, 112, 123, 169, 173, 202, 210, 212, 247, 250, 253, 260-263, 265-268, 270-277, 279-282, 284-286, 289, 290, 295, 296, 298-300, 305, 307, 309, 312, 318, 320-322, 324-329, 331, 333, 335, 341, 343, 355, 356, 358, 393, 394, 397, 400, 404-409, 411, 413, 416, 419, 421, 422-424, 426-428, 431, 432, 434, 435, 437-439, 442, 443, 453, 469 S Sagalassos 260, 297, 299 Saïd, Edward 41 Sakaeli 1, 88, 89, 125 Sakarya 14, 213 Salhane 161 Salur North 168 Saraçhane 305 seafaring 19, 20, 141, 173, 371, 439, 441 Sea of Marmara 13, 14 Second Millennium BC 90, 403 sedentism 142, 435 Sehzade Mahmud 379 Seljuk 296, 302, 369, 371, 376, 377, 382, 388 Sesamus 37, 38, 39, 216, 247, 248, 249 Seyahatname 42 Sgraffito ware 307, 317, 320, 324, 366, 409 shipbuilding 20, 25, 26, 40, 42, 43, 44, 47, 48, 49, 75, 287, 382 shipwrecks 19, 20, 198 shrimps 195 simas 219 Sinop 9, 13, 17, 21, 38, 39, 42, 48, 57, 75, 108, 109, 151, 153, 156, 163, 167, 168, 172, 173,
174, 186, 195, 196, 199, 212, 214, 215, 216, 218, 219, 223, 247, 248, 249, 260, 263, 266, 267, 269, 271, 272, 279, 286, 288, 296, 298, 300, 376, 377, 379, 380, 382, 383, 384, 386, 387, 405, 419, 438 Sinop Bus Station 200 Sinopean amphorae 269, 275, 281 sinopis 220 Sitagroi 161 siteless survey 54, 58, 65, 76, 95 site preservation 53 sites 1, 2, 6, 14, 30, 43, 49, 54, 56, 57, 59, 62, 63, 65, 66, 67, 68, 71, 73, 74, 81, 83, 91, 93, 95, 96, 97, 104, 105, 106, 108, 109, 117, 122, 125, 126, 127, 128, 132, 134, 137, 140, 141, 151, 154, 155, 159, 161, 167, 168, 169, 172, 173, 183, 184, 185, 192, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 213, 217, 218, 219, 221, 222, 230, 231, 232, 250, 253, 261, 263, 265, 271, 272, 295, 296, 297, 298, 299, 300, 305, 307, 309, 312, 316, 317, 319, 324, 325, 329, 331, 343, 344, 349, 364, 367, 369, 371, 375, 392, 393, 395, 401, 406, 409, 417, 424, 431, 432, 435, 436, 437, 438, 443, 455, 460 slag 95, 172, 197 Slave trading 249 Sofular Mağarası 11 soils 24, 82, 152, 460, 461, 469 soil samples 67, 82, 93, 397, 408, 460 Sokukayası 107, 134 Soreq cave 11 spot 67, 87, 118, 120, 123, 272, 422, 436 stalactites 11, 138 St. Philaretus 301 Strabo 25, 38, 39, 42, 196, 215, 248, 276, 282, 287 Sultan Cem 379 survey permit 2, 4, 5, 28, 58 T Tabula Peutingeriana 276 Tanzimat 382 taphonomic processes 60, 83, 168, 183, 221 targeted survey, reconaissance 1, 6, 54, 62, 63, 68, 71, 73, 77, 97, 106, 364, 378, 393, 413 tekke 47 Tekkeköy 167, 195, 199 temsilçi 56, 72
Index Tepecik- Çiftlik 117 Terme 49 Terra sigillata 265 tersane 25, 382 Texier, Charles 44 Theodore II Laskaris 376 Theophrastus 40, 287 Thessalonica 305 threshing sledge 86, 112, 400 Thymena 39, 43, 49, 248, 284, 298 Tieion 37, 216 Tieium 38 Tiles 309 Time 49 Timle 284, 326, 328, 329, 333, 334, 335, 336, 358, 439 Tios 39, 296, 376 tracts 55, 63, 65, 68, 69, 76, 77 tract-walking 55 trade 19, 20, 21, 26, 42, 75, 195, 198, 199, 215, 216, 218, 220, 246, 247, 248, 249, 254, 298, 299, 301, 302, 305, 326, 358, 359, 372, 376, 378, 379, 406, 409, 413, 424, 435, 440, 441 Trajan 298 transhumant 169, 376, 440 Trialeti 198 Trojan War 36, 48 Troy 37, 91, 134, 151, 152, 155, 156, 160, 163, 172, 186, 438 Tudhaliya I/II 197 Tufan Reis 386 Tuğtepe, Ali Osman 18, 24, 75 Tumanna 196 Turkic tribes 302, 338 Turkmen 376, 377, 378 Turnu-Severin 367 U Uçakdibi Mağarası 349, 426 Ugarit 199 Uğurlu 284 Ulucak 123 units 3, 54, 55, 63, 68, 77, 78, 79, 80, 83, 95, 118, 321, 324, 461 unit-walking 66, 81, 82, 86, 96, 118 Upper Land 196
485
Üyüktepe 196, 200 V Valerius Flaccus 38 Valla 16, 123, 426, 427 Vartitsa 163 vegetation zone 3, 9, 10, 11, 12, 54, 56, 60, 70, 71, 78, 96, 130, 183, 252, 396, 399, 409, 414, 419, 426 Venetian 20, 36, 41, 42, 371 Virgil 40 visibility 4, 31, 53, 54, 55, 58, 60, 66, 77, 78, 79, 83, 96, 105, 106, 168, 252, 325, 343, 398, 399, 402, 414, 418 Vital Cuinet 46 Vroom, Johanna 295 W water buffalos 25, 414 watermills 23 X Xenophon 25, 231 Xerxes 215 Y Yağlar 88, 89, 125, 402 Yalı Hotel 64, 70, 75 Yarıkkaya 107, 108 Yassıkaya 151, 153, 156, 159, 167, 168, 171 Yazıboy 159, 160 Yedikapılı Mağarası 353 Yeniçağa 10 Yenikapı 260 Yılmaz, Recai XIII, 71, 72, 75, 396, 413 Yortan 156, 159 Younger Dryas 10, 14, 19 Yüklütepe 218, 220 Z Zalpa Zalpuwa 194, 195, 196, 199 Zeest 80/Knossos Type 39 267, 269, 271, 277 Zeuxippus 307, 366, 368, 371 Zimnicea 367 Zonguldak 23 Zonguldak Yöresi 109