Landscape Dynamics and Settlement Patterns in Northern Anatolia during the Roman and Byzantine Period 3515112146, 9783515112147

Citation preview

Landscape Dynamics and Settlement Patterns in Northern Anatolia during the Roman and Byzantine Period Edited by Kristina Winther-Jacobsen and Lâtife Summerer

Alte Geschichte Franz Steiner Verlag

Geographica Historica – 32

Landscape Dynamics and Settlement Patterns in Northern Anatolia during the Roman and Byzantine Period Edited by Kristina Winther-Jacobsen and Lâtife Summerer

geographica historica Begründet von Ernst Kirsten, herausgegeben von Eckart Olshausen und Vera Sauer Band 32

Landscape Dynamics and Settlement Patterns in Northern Anatolia during the Roman and Byzantine Period Edited by Kristina Winther-Jacobsen and Lâtife Summerer

Franz Steiner Verlag

The publication of this volume was supported by the Danish Council for Independent Research as part of the research project ‘Where East meets West’.

Typesetting: Vera Sauer Copy-editing: Gina Coulthard Cartography: Richard Szydlak Bibliografische Information der Deutschen Nationalbibliothek: Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über abrufbar. Dieses Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes ist unzulässig und strafbar. © Franz Steiner Verlag, Stuttgart 2015 Druck: Laupp & Göbel, Nehren Gedruckt auf säurefreiem, alterungsbeständigem Papier. Printed in Germany. ISBN 978-3-515-11214-7 (Print) ISBN 978-3-515-11216-1 (E-Book)

Zum Geleit

Im Rahmen des vom Danish Council for Independent Research geförderten Forschungsprojekts Where East meets West veranstalteten Kristina Winther-Jacobsen und Lâtife Summerer im April 2014 in Amasya/Türkei ein Kolloquium mit dem Titel ›Landscape Dynamics and Settlement Patterns in Northern Anatolia during the Roman and Early Byzantine Period‹. Thematisch schloss sich diese Veranstaltung an das Kolloquium ›Space, Place and Identity in Northern Anatolia‹ an, das 2012 ebenfalls im Rahmen des Projekts Where East meets West unter der Leitung von Tønnes Bekker-Nielsen in Kolding/Dänemark stattgefunden hatte und dessen Akten als Band 29 der Reihe Geographica Historica erschienen sind. Der nun vorliegende Band entspricht nicht nur inhaltlich im engeren Sinn, sondern auch gleichsam ideell ganz der Konzeption der Geographica Historica. Einmal mehr zeigt sich, wie fruchtbar es ist, wenn Vertreter verschiedener Forschungsdisziplinen international zusammenarbeiten. Vor allem aber demonstriert dieser Band, welches Potential speziell im Bereich der Archäologie, damit aber auch der Historischen Geographie die Vernetzung universitärer und musealer, überregional und regional organiserter Forschung birgt. Eckart Olshausen und Vera Sauer

Contents

Illustrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11

Kristina Winther-Jacobsen and Lâtife Summerer Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

17

The dynamics of landscapes: cities and territories Philip Bes Roman-period finds from the Cide region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

23

Owen Doonan Settlement and economic intensification in the late Roman/ early Byzantine hinterland of Sinop . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

43

Peri Johnson How did the landscape of Pompeiopolis become Roman? . . . . . . . . . . . . . . . . . . . . . . . .

61

Kristina Winther-Jacobsen Contextualizing Neoklaudiopolis: a glimpse at settlement dynamics in the city’s hinterland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

83

Peter Bikoulis, Hugh Elton, John Haldon and Jim Newhard Above as below: The application of multiple survey techniques at a Byzantine church at Avkat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Max Ritter The end of late antiquity in Paphlagonia: disurbanisation from a comparative perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 Report Celal Özdemir Preliminary results of rescue excavations at the Sanctuary of Zeus Stratios near Yassıçal, Amasya province . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

8

Contents

The dynamics of mortuary space: necropoleis, graves and grave monuments Pavol Hnila Rural necropoleis and settlement dynamics: Thoughts on Roman and Byzantine graves at Oymaağaç Höyük, Samsun province . . . . . . . . . . . . . . . . . . . . . 147 Vera Sauer Two coins from Oymaağaç Höyük: Appendix to Rural necropoleis and settlement dynamics by P. Hnila . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Julia Koch From funerary doorstones in Pompeiopolis to tracing local identity from the Phrygian Highlands to Inner Paphlagonia . . . . . . . . . . . . . . . . . . . . . . 171 Reports Muzaffer Doğanbaş Preliminary results of rescue excavations in the Roman and Byzantine necropoleis of Amasya (1977–2014) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193 İlkay İvgn Results of a rescue excavation in the village of Zafer, near Tekkeköy, Samsun province . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205

The dynamics of decoration: sculptures and mosaics Baran AydIn, Laura Buccino and Lâtife Summerer Honoured, beheaded and buried: A new deposit of statues from Amastris . . . . . . . . . . 219 Luisa Musso Dynamics of mosaic design in northern Anatolia in the Roman and early Byzantine periods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239 Report Esra Keskin Preliminary results of rescue excavations in 2013 at a Roman mosaic-paved building in the village of Yavru, Amasya province . . . . . . . . . . . . . . . . . . 285

Contents

9

The dynamics of circulation: roads, inscriptions, coins Tønnes Bekker-Nielsen and Rainer M. Czichon Ancient Roads and Bridges of the Vezirköprü District . . . . . . . . . . . . . . . . . . . . . . . . . . . 295 Christian Marek Epigraphy and the provincial organisation of Paphlagonian cities in the Roman Empire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307 Vera Sauer The coinage of Neoklaudiopolis and Pompeiopolis against the background of the minting practices of Pontic and Paphlagonian cities during the Roman Imperial period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329

Indices Ancient personal names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Modern personal names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Places . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

341 343 348 353

Illustrations

Kristina Winther-Jacobsen and Lâtife Summerer, Introduction 1. Group photo taken in Amasya during the seminar, 22nd April, 2014

Philip Bes, Roman-period finds from the Cide region 1. 2. 3. 4. 5.

Roman-period sites in the Black Sea area mentioned in the text Concentrations of predominantly Roman-period ceramic finds in the study area Top part of an early Roman gravestone The bridgehead along the Devrekani at Atköprü Fragments of amphorae from Cide

Owen Doonan, Settlement and economic intensification in the late Roman/ early Byzantine hinterland of Sinop 1. Map of the Sinop promontory showing locations mentioned in the text 2. Topographic map of the Demirci valley showing sample tracts and Roman-period archaeological loci 3. Map of the interior of the Demirci valley and the site L97-101 4. Local ceramics from locus L97-101 5. Survey tracts of locus L99-21 (Karapınar) 6. Ceramics from L99-21 (Karapınar) 7. Kiln from Eldevüz in the Demirci valley 8. Map of Kırkgeçit çayı survey area 9. Survey areas in Çakıroğlu and Erikli quadrats

Peri Johnson, How did the landscape of Pompeiopolis become Roman? 1. Map of central northern Turkey 2. Map of the environs of Pompeiopolis 3. Map of features and places in the environs of Pompeiopolis mentioned in the geomorphological description 4. Map of the cemeteries near Pompeiopolis 5. Map of the Çetmi and Donalar areas 6. Map of the Tepedelik area 7. Manigünez Tepeleri 8. Map of the Alisaray area 9. Map of the Aşağıemerce area and the confluence of the Aşağıçayı with the Gökırmak 10. Harmanyeri and Sivritepe

Kristina Winther-Jacobsen, Contextualizing Neoklaudiopolis: a glimpse at settlement dynamics in the city’s hinterland 1. Map of the sites in the survey area discussed in the text 2. Georesistivity map of Papaz Tarlası 3. Gridded survey map of Papaz Tarlası indicating the subsurface structures and other recorded features 4. Functional categories recorded in transect lines across Papaz Tarlası 5. Selection of pottery from Papaz Tarlası

12

Illustrations

6. Stela from Papaz Tarlası 7. Selection of pottery from the Christian sanctuary on the Doyran-Tepeören road 8. Map of the Tepeören settlement system 9. Selection of pottery from Roman Tepeören 10. Selection of pottery from the Tepeören-Merzifon road settlement

Peter Bikoulis, Hugh Elton, John Haldon and Jim Newhard, Above as below: The application of multiple survey techniques at a Byzantine church at Avkat 1. Location of Beyözü/Avkat and the survey area 2. Inscription, church spolia and architectural pieces 3. Exposed section of wall 4. Densities of total counts of artefacts in the Avkat region 5. Structure extents based on GPR survey 6. Super-intensive survey: illustration of methods 7. Super intensive survey areas 8. Artefacts collected during the super-intensive survey 9. Identified and extracted architectural features 10. Density of roof tile in Area 1

Max Ritter, The end of late antiquity in Paphlagonia: disurbanisation from a comparative perspective 1. Map of the administrative units and cities of Paphlagonia during late antiquity 2. Map of the via militaris to the Armenian frontier via the Amnias valley 3. Kız Kalesi

Celal Özdemir, Preliminary results of rescue excavations at the Sanctuary of Zeus Stratios near Yassıçal, Amasya province 1. Plan of the multi-phased building complex 2. Hellenistic coin of Amisos 3. The flooring of Room B 4. View of Room B showing the different floor levels 5. Bronze votive tablet with a dedication to Zeus Stratios 6. Detail of Rooms C-1 and C-2 7. View of the corridor/Room D 8. View of the paved porch 9. Detail of Room A 10. Bovine skull 11. Remains of the temple foundation southwest of the multi-phased building complex 12. General view to the building complex 13. View of Trench AL-1 showing the substructure of the monumental altar

Pavol Hnila, Rural necropoleis and settlement dynamics: Thoughts on Roman and Byzantine graves at Oymaağaç Höyük, Samsun province 1. 2. 3. 4. 5. 6.

Plan of the necropolis of Oymaağaç Höyük Cist grave locus 7585:010 with multiple successive burials Tile graves loci 7585:022 and 7585:026 Personal adornments Pottery Tiles

Vera Sauer, Two coins from Oymaağaç Höyük: Appendix to Rural necropoleis and settlement dynamics by P. Hnila 1. Oymaağaç Höyük, find 7385:025:002 2. American Numismatic Society no. 1944.100.41429 3. Oymaağaç Höyük, find 7584:004:003

Illustrations 4. Dalaison 2008: no. 36a 5. Oymaağaç Höyük, find 7584:004:003

Julia Koch, From funerary doorstones in Pompeiopolis to tracing local identity from the Phrygian Highlands to Inner Paphlagonia 1a. Brick-built village fountain with spolia of a Roman doorstone monument in Afşar Köy 1b. Detail of the doorstone monument in Afşar Köy 2. Marble doorstone imitating house architecture 3. Door stele with tenon crowned by a scallop shell and an acanthus frieze 4. Marble doorstone featuring a door knocker and a pruning hook 5. Doorstone found near modern Boyabat 6. Fragment of a doorstone found near modern Boyabat 7. The fertile valley of the Amnias river near Boyabat 8. Tumulus near Tepedelik 9. Achaemenid rock-cut tomb near Donalar 10. Doorstone of Doryphoros 11. Doorstone crowned by an egg-and-dart motif with a prothesis scene on top

Muzaffer Doğanbaş, Preliminary results of rescue excavations in the Roman and Byzantine necropoleis of Amasya (1977–2014) 1. Map of the locations of the rescue-excavated graves within the modern city of Amasya 2. The Şamlar quarter necropolis 3. Coin of Domitian from the Şamlar quarter necropolis 4. Grave goods from the Şamlar quarter necropolis 5. Graves excavated in 2010 in the Kurşunlu quarter necropolis 6. Three painted graves rescue excavated in the Şamlar quarter necropolis in 2010 7. Interior of a grave rescue excavated in 2014 in the Kirazlıdere quarter necropolis 8. The Üçler quarter necropolis 9. Detail of a painted grave rescue excavated in 2014 in the Kirazlıdere quarter necropolis 10. The Kirazlıdere quarter necropolis in 2014 11. A triple set of graves rescue excavated in the Kirazlıdere quarter necropolis in 2010 12. Terracotta paved floor of a grave in the Kirazlıdere quarter necropolis 13. Soles of shoes found in a grave in the Kurşunlu quarter necropolis

İlkay İvgn, Results of a rescue excavation in the village of Zafer, near Tekkeköy, Samsun province 1. Map of the location 2. Plan of the burial chamber 3. The entrance to the burial chamber 4. The interior of the tomb 5. Sarcophagus 1 6. The interior of Sarcophagus 1 7. Detail of the interior of Sarcophagus 1 8. The finds 9. Sarcophagus 2 10. The interior of Sarcophagus 2 11. The ancient repair on Sarcophagus 1

Baran AydIn, Laura Buccino and Lâtife Summerer, Honoured, beheaded and buried: A new deposit of statues from Amastris 1. 2. 3. 4. 5.

Topographical setting of the monuments in Amastris The rescue excavation The statues in situ in the deposit Inscription fragment found in the deposit Over-life-size cuirass statue

13

14

Illustrations

6. Over-life-size female draped statue 7. Honorific male statue 8. Honorific male statue

Luisa Musso, Dynamics of mosaic design in northern Anatolia in the Roman and early Byzantine periods 1. Map showing the sites of provenance of the mosaics 2. Samsun Museum: Achilles mosaic 3. Samsun Museum: Achilles mosaic, detail 4. Samsun Museum: Achilles mosaic, detail 5. Samsun, Toraman Tepe, Mosaic Hall, detail 6. Samsun, Toraman Tepe, mosaic 7. Sinop Museum, garden 8. Sinop Museum, garden 9. Çiftlik (Sinop), view of the church 10. Kadıoğlu Köyü, villa, apsidal room 11. Drawing of the mosaic found in Rome, outside Porta Portese, in 1823 12. Pompeiopolis, Zimbıllı Tepe: mosaic illustrating the personification of eternal Time 13. Pompeiopolis, private bathhouse complex: reconstruction of the frigidarium mosaic 14. Pompeiopolis, domus on the slopes of Zimbıllı Tepe, room A 15. Pompeiopolis, domus on the slopes of Zımbıllı Tepe, room A, detail 16. Pompeiopolis, domus on the slopes of Zımbıllı Tepe, mosaic pavement of room E 17. Pompeiopolis, domus on the slopes of Zimbıllı Tepe, mosaic pavement of room E, detail 18. Hadrianopolis, Early Byzantine Church A, nave: detail 19. Hadrianopolis, Early Byzantine Church A, north aisle: mosaic 20. Hadrianopolis, Church B: partial view of the eastern part of the building 21. Hadrianopolis, Church B, north aisle: detail 22. Hadrianopolis, Church B, nave: detail 23. Hadrianopolis, Church B, south aisle: figured panel of the pavement mosaic 24. Nea Paphos (Cyprus), House of Aion, mosaic panel representing the ‘Judgement’ of the Nereids

Esra Keskin, Preliminary results of rescue excavations in 2013 at a Roman mosaic-paved building in the village of Yavru, Amasya province 1. 2. 3. 4. 5. 6. 7. 8.

Location of Yavru, Amasya Plan of the building Room 1, the apple tree mosaic Room 1, detail of mosaic Room 2 Room 3, carpet patterned mosaic Room 3, north section Tile graves

Tønnes Bekker-Nielsen and Rainer M. Czichon, Ancient Roads and Bridges of the Vezirköprü District 1. The Kurt köprüsü 2. Longitudinal section through the bed of the Roman highway just north of Doyran village 3. The city site of Neapolis and its roads 4. Remains of a settlement along the road from Tepeören into the Tavşan range 5. Remains of a pier of the bridge at Yürükçal and paving in the bed of the stream 6. Masons’ marks on the bridge at Yürükçal 7. Section of the ancient road between Yürükçal and Güldere 8. Rock cutting at the crossing of a stream immediately east of Aydoğdu village 9. Remains of the ancient road south of Kocaoğlu 10. The ancient road network of the Phazemonitis

Illustrations

Christian Marek, Epigraphy and the provincial organisation of Paphlagonian cities in the Roman Empire 1. 2. 3. 4.

Imhoof-Blumer 1890: 577f. no. 55, pl. IV no. 16 http://www.cngcoins.com/Coin.aspx?CoinID=127091 Fragment of an honorary inscription from Pompeiopolis Squeeze of an unpublished milestone from Pompeiopolis

Vera Sauer, The coinage of Neoklaudiopolis and Pompeiopolis against the background of the minting practices of Pontic and Paphlagonian cities during the Roman Imperial period 1. Map of the Pontic and Paphlagonian cities 2. Dalaison and Delrieux 2014: no. 20b 3. Dalaison 2010: no. 12a

15

Kristina Winther-Jacobsen and Lâtife Summerer

Introduction

Under the Republic, the integration of conquered lands into the Imperium Romanum generally followed one of two approaches. In the West, existing towns were integrated as self-governing units and new colonies of Roman citizens were founded in strategic positions, while in the Hellenised East, new cities were not founded on any significant scale; instead, existing poleis continued to administer their territories on behalf of their new masters. The inland districts of the conquered Mithradatic kingdom, however, were unlike any that Rome had previously annexed, and unsuitable for either method of integration. The victorious general Pompey the Great made a radical new departure in terms of Roman imperialism by creating an urban network from scratch: seven poleis, spaced so as to control the maximum area of territory with the minimal use of resources, were established. Some of Pompey’s work was undone by Mark Antony (39–31 BC), but the Pompeian structures were reinstated by Augustus. The importance of urbanisation as an instrument of Roman domination is universally acknowledged among modern scholars, and in November 2012 the research project Where East meets West hosted a workshop in Kolding on Space, Place and Identity in Northern Anatolia. The aim of the workshop was to explore the genesis and function of the city and its sanctuaries, and their role in the process of provincialisation. On 24–26 April 2014 thirty people met in Amasya at the beautiful Historical Saraydüzü Barracks, National Struggle Museum and Conference Center under a banner welcoming us to the seminar on Landscape Dynamics and Settlement Patterns in Northern Anatolia during the Roman and Early Byzantine Period (Fig. 1). The aim of this seminar was to view the Roman settlements and sanctuaries, and their trajectories specifically in the context of their territory. Over the course of three days researchers from all over the world, all working in northern Anatolia met to exchange ideas and compare results. The 26 papers explored the current state of knowledge on settlement patterns in central northern Anatolia from both archaeological and historical points of view. While some papers presented the results of recent field research, others surveyed little known material ripe for new interpretations. The main principle behind the choice of venue in Amasya was to bring together researchers from both universities and museums: an approach not undertaken previously in Turkey. The aim was to prompt new dynamics in the archaeological exploration of northern Anatolia by encouraging and launching new communications and collaborations between scholars involved in on-going research projects and museum staff concerned with emergency excavations; the event offered the means to share knowledge, data and results that were otherwise difficult to access. With the breaching of the language barrier, the conference was fruitful for all the participants.

18

Kristina Winther-Jacobsen and Lâtife Summerer

Fig. 1: Group photo taken in Amasya during the seminar, 22nd April, 2014. Reflecting the nature of the seminar, this volume consists of a combination of articles; some focus on specific locations, others explore a specific region and some consider particular classes of material culture. Alongside these analytical studies, there are preliminary reports of rescue excavations. The articles are arranged in four thematic sections, as detailed below.

1. The dynamics of landscapes: cities and territories The first three papers of the section on the dynamics of landscapes are closely related to one another and present new results of surveys in neighbouring regions. They are concerned with the identification of settlement patterns and the tracing of the dynamics of spatial organisation. Philip Bes presents the Roman finds from the Cide region revealed by the Cide Archaeological Project. Although the ancient landscapes have today been greatly transformed by natural and cultural processes, the carefully mapped find clusters allow the

Introduction

19

identification of sites which are tentatively associated with ancient place names known from the written sources. Owen Doonan demonstrates the divergent histories of contrasting landscapes in the hinterland of Sinop. Some areas remained isolated for millennia while others became deeply integrated in the olive oil trade networks of the Black Sea region and beyond, especially during the Roman period. Meanwhile, Peri Johnson presents some of the results of survey in the district of Taşköprü. She argues that the prosperity and density of the settlements evidenced in the region correlate with the introduction of Roman urban institutions to Pompeiopolis and that the diffusion of monuments in the landscape reveals a need to establish a connection with the past. The article by Kristina Winther-Jacobsen is the first of four articles which focus on the territory of Neoklaudiopolis.1 Based on fieldwork conducted in 2013, she presents the current evidence for the settlement dynamics of the city’s territory during the Roman and early Byzantine periods, although one nucleus of shifting settlements appears to date back to the Bronze Age. As the author notes, a more detailed analysis of the territories of the modern-day villages of the region would surely reveal more evidence for this nucleated, shifting settlement pattern. The contribution by Peter Bikoulis, Hugh Elton, John Haldon and Jim Newhard addresses an important methodological issue: the correlation between surface and subsurface finds based on the super-intensive survey of a church site at Avkat in 2009. In contrast to the evidence from Papaz Tarlası presented by Winther-Jacobsen, this study shows only a weak positive correlation between the surface distribution of artefacts and subsurface structures. Max Ritter examines the history of the Paphlagonian cities in late antiquity and argues that the alterations to the urban design reflect economic activities and that this should be understood as correlated with the general prosperity of the hinterland. Another example of late antique alterations to the city’s layout is examined in the article by Baran AydIn, Laura Buccino and Lâtife Summerer.2 In the final article in this section, Celal Özdemr presents a preliminary report on the results of rescue excavations at the important Sanctuary of Zeus Stratios in the territory of Amaseia. This eagerly awaited report offers new impetus to the dynamic of the discussion on the character of this monumental hilltop sanctuary in northern Anatolia as well as that on the role of ritual space as common ground amongst the political territories of the region.3

2. The dynamics of mortuary space: necropoleis, graves and grave monuments In the second section, on the dynamics of mortuary space, Pavol Hnila presents his thoughts on the Roman and Byzantine graves at Oymaağaç Höyük and places this rural necropolis and its population within the settlement dynamics of the territory of 1

See also Tønnes Bekker-Nielsen and Rainer Czichon, Pavol Hnila (and Vera Sauer) and Vera Sauer below. 2 See below. 3 C. G. Williamson, ‘Power, politics and panoramas: viewing the sacred landscape of Zeus Stratios near Amaseia’, in T. Bekker-Nielsen (ed.), Space, Place and Identity in Northern Anatolia (Geographica Historica 29). Stuttgart, 175–88.

20

Kristina Winther-Jacobsen and Lâtife Summerer

Neoklaudiopolis. This careful study of the necropolis provides evidence of long-term continuity – possibly over a millennium – of a rather small community. In an appendix, Vera Sauer offers the numismatic basis for the dating of two of the graves. Rather than necropoleis, Julia Koch focuses on individual funerary monuments and presents new evidence from Pompeiopolis on Phrygian doorstones and other monument types from Paphlagonia. Koch argues that identity was the dynamic behind the design and decoration of certain forms of funerary monuments in Paphlagonia. Meanwhile, based on the results of rescue excavations undertaken by the Amasya Museum between 1977 and 2014, the report of Muzaffer Doğanbaş considers the distribution of the necropoleis of Amaseia during the Roman and Byzantine periods and provides a glimpse of the burial customs of the ancient city by presenting evidence of the grave types and offerings. The final contribution to this section, by İlkay İvgn, reports on the results of the 2013 rescue excavation of a rock-cut tomb with tube-shaped clay sarcophagi in the village of Zafer in the territory of Amisos. The interpretation of this tomb is viewed against the occurrence of similar tombs and Hellenistic sites recorded in the territory of Amisos, mainly during the course of rescue excavations. İvgn notes that a large-scale research project is required in order to contextualise the results of the rescue excavations.

3. The dynamics of decoration: sculptures and mosaics The first article in the third section, on the dynamics of decoration, is by Laura Buccino and Lâtife Summerer in cooperation with the director of the Amasra Museum, Baran AydIn, and considers changes in the use of urban space in Amastris in light of the discovery of a cache of buried statues. The material, excavated in 1993, illustrates an interesting example of changing intra-city dynamics during the Roman and early Byzantine periods, when the city contracted.4 In the following article, Luisa Musso examines five Roman and early Byzantine mosaic designs from northern Anatolia. She concludes that, in general terms, most of the mosaic compositions studied differ markedly from the repertoire of western Asia Minor and that the northern Anatolian mosaic pavements tend to be more closely connected with the eastern territories of the Empire. In the final article of this section Esra Keskn documents the presence of a rural Roman villa decorated with mosaic floors in the territory of Amaseia. This is a preliminary report on the results of rescue excavations conducted in the village of Yavru.

4. The dynamics of circulation: roads, inscriptions and coins The first contribution to the section on the dynamics of circulation, by Tønnes BekkerNielsen and Rainer Czichon, updates our current understanding of the ancient road network in the Phazemonitis linking the pre-Roman and Roman road networks. It also presents evidence for two roads linking Neoklaudiopolis to the commercial road identified by John Arthur Ruskin Munro. Christian Marek then revisits the question of 4

See also Ritter above.

Introduction

21

provincial koina and argues, in opposition to the communis opinio, that Paphlagonia did not belong to the koinon of Pontus. He concludes, on the basis of epigraphic evidence including the well-known imperial oath inscription from Vezirköprü and recently discovered inscriptions from Pompeiopolis, that there existed an independent koinon of the province of Paphlagonia. Finally, Vera Sauer discusses the coinage of Neoklaudiopolis and Pompeiopolis against the background of the minting practices of other Pontic and Paphlagonian cities. She argues that the minting strategies of these cities was simultaneously both individual and conventional.

5. Acknowledgements The seminar was sponsored by the Danish Council for Independent Research as part of the research project Where East meets West (University of Southern Denmark), and we wish to thank the Project Director, Tønnes Bekker-Nielsen, for his support. The beautiful venue, the Saraydüzü Barracks was made available by the Turkish Ministry of Culture and Tourism, General Directorate for Cultural Heritage and Museums, to whom we are very grateful. The seminar would not have been possible without the support of the Ministry and General Directorate. We owe a particular debt of gratitude to İsmail Hakkı Göztaș, President of the Chamber of Architects of Amasya, who supported the seminar from the moment of its initiation and, together with his wife Firdevs Öztas, played an instrumental role in its successful outcome. Thanks are due also to the entire staff of the Amasya Museum, especially its Director, Celal Özdemir, for their commitment to the organisation of the conference. We also wish to thank the Municipality of Amasya for its hospitality. Finally, we wish to thank all the contributors as well as Richard Szydlak (cartography), Gina Coulthard (English language revision) and Vera Sauer (typesetting). Without their tireless efforts, this volume could not have been produced in such a timely manner. Kristina Winther-Jacobsen University of Copenhagen Saxo-Institute Karen Blixens vej 4, 2300 Copenhagen S, Denmark [email protected] Lâtife Summerer Ludwig-Maximilians-Universität Institut für Klassische Archäologie Katharina-von-Bora-Straße 10, 80333 Munich, Germany [email protected]

Philip Bes

Roman-period finds from the Cide region

1. Introduction The Black Sea basin and northern Asia Minor in particular have long attracted scholarly attention, albeit on a fairly limited scale; in more recent years, however, the scope has been steadily widening.1 From a Mediterranean perspective, Turkey’s northern coast may be perceived as peripheral, yet culturally it has much in common with the Black Sea basin and inland Asia Minor, and, on a broader scale, also with the Aegean and Mediterranean;2 this was also the case during the Roman period.3 This paper focuses on the Roman-period data – ceramic and architectural – that has been gathered by the Cide Archaeological Project (CAP hereafter) and which serves to illuminate both spatial and chronological patterns of occupation as well as economic life.4 The scattered and restricted nature of the collected evidence precludes a detailed analysis;5 nonetheless, 14 find clusters offer indications that the Cide region was a relatively wellsettled area during the (late) Roman period. The CAP survey area is located about two-thirds along the coast travelling from Constantinople towards Sinope. Ceramic data from these latter two locations together with newly-published evidence from Pompeiopolis, Hadrianopolis and the Paphlagonia survey have served to highlight the late Roman occupational character of the wider region (Fig. 1),6 and the small quantity of datable ceramics from the CAP study area (211 fragments) – further complemented by scattered architectural finds and ancient sources – also

1 2 3 4 5 6

Knipowitsch 1929; Asheri and Hoepfner 1972; Marek 1993; Belke 1996; Gabrielsen and Lund 2007; Kassab Tezgör 2010; Tsetskhladze 2012. Braund 2005: 115, 117. For a brief historical background, see Bes forthcoming (with bibliography). Düring and Glatz forthcoming. Düring and Glatz forthcoming: esp. chapter 4.1–3. Matthews and Glatz 2009; Domżalski 2011; Zhuravlev 2011; Lafli and Kan Şahn 2012a; 2012b; Lafli and Christof 2012. Hellenistic and Roman amphorae from Sinope were widely – if at times thinly – distributed throughout the (central and eastern) Mediterranean: Lund 2007; Pieri 2007: 8–9; Kassab Tezgör 2010; Reynolds 2010; De Boer 2013. Constantinople remains relatively poorly known: Hayes 1992; Bardill and Hayes 2002; Jobst 2005. Results from the Yenikapı and Theodosian Harbour excavations will add significantly to our understanding of Roman-period ceramic trends.

24

Philip Bes

Fig. 1: Roman-period sites in the Black Sea area mentioned in the text (Toby Wilkinson, ©CAP). signals activity and occupation during the Roman period (Fig. 2; also see Table 2 below).7 This paper aims to combine and discuss these three sources of evidence in order to contextualise the results of the CAP survey.

2. The ceramic finds Due to the restricted time in the field, only cursory observations could be made. A common methodological approach was followed; this was directed towards determining: (1) fabric (with the naked eye); (2) fragment (rim, handle, etc.); (3) decoration/surface treatment (e.g. slipped); (4) shape/type (e.g. Late Roman Amphora 2); (5) (primary) function;8 and (6) chronological identification. Three (functional) groups were thus distinguished (Table 1). 2.1 Tablewares Tablewares comprise vessels (open and closed) used for the serving and consumption of food and beverages, and which are also commonly characterised by the presence of slip and/or certain styles of decoration. At the same time, however, we should allow for a fair degree of variability concerning the quantity and variety of vessels in use as tableware – as 7

Hellenistic: ca 325–25/1 BC; early Roman: ca 1–200; middle Roman: ca 200–400; late Roman: ca 400–650/700. Architectural fragments were studied from drawings and photographs. 8 Peña 2007.

Roman-period finds from the Cide region

25

Fig. 2: Concentrations of predominantly Roman-period ceramic finds in the study area; the darker the shading, the lower the relief (Toby Wilkinson, ©CAP). well as the materials these were made of – by a single family or household, as this was likely dependent on, for instance, socio-economic position, where one lived, etc. Amongst a total of 18 fragments, of most precise chronological significance are 11 sherds of Phocaean Red Slip Ware (PRSW).9 Originally associated exclusively with ancient Phokaia in western Turkey, we now know that several other locations in the same region produced similar shapes, and this has gradually prompted a return to the use of the former, more neutral term Late Roman C (LRC). Eight fragments are attributed to the Phokaia workshops proper, spanning the period between the early fifth and early seventh centuries.10 Though obviously too small for a quantitative analysis, the fact that only form 3F occurs more than once is in line with more general (Mediterranean) trends. It is noteworthy that no other class of (late) Roman tableware was recognised. The absence of Pontic Sigillata and Pontic Red Slip Wares, otherwise well attested at Hadrianopolis, Pompeiopolis, Neoklaudiopolis and Sinope (see below), is particularly noteworthy.

9 10

Hayes 1972: 323–70; Vaag 2005; Bes 2015. Hayes forms 3C (n=1), 3E (n=1), 3E–F (n=1), 3F (n=3), 3F/10A (n=1) and 10A (n=1). Three fragments are classified as Çandarlı Ware/Phocaean LRC: cf. Hayes 1972: 317, 369. Archaeometric analyses now indicate that it is not feasible to distinguish confidently between Çandarlı Ware produced during the first three centuries at Çandarlı (ancient Pitane), near Pergamon, and late Hellenistic to early Roman products from Pergamon proper (Pergamenische Sigillata): Japp 2009; Mommsen and Japp 2009; Schneider and Japp 2009.

26

Philip Bes

2.2 Amphorae Hellenistic to late Roman amphorae – vessels primarily intended for storing but above all transporting (agricultural) produce – are the best represented category in the CAP assemblage (n=180), yet the assemblage only modestly reflects this period’s wide variety in amphorae morphology and provenance; most types and fabrics are attested in single or a few occurrences only. The exception is a group of 95 fragments with a presumed local (=Cide) provenance (see below). Among the sample are Aegean classes such as Rhodian and Chian amphorae, the latter known from shipwrecks around Sinope;11 Hellenistic to early Roman Knidian amphorae are marginally more common. Their fabric identifies three sherds as having originated in western Turkey: a distinctive profile probably belongs to the so-called Nikandros group from (the region of) Ephesos and resembles closely examples dating to the third century BC.12 From the same area comes an amphora toe possibly from a middle to late Roman Agora M273/Samos Cistern Type,13 as well as a small rim of possible Roman date. Also from the Aegean come 13 sherds of Late Roman Amphora 2 (LRA 2), a type manufactured at a number of locations yet still poorly understood.14 Three fragments originate from the Argolid, and the horizontal combing on two other fragments tentatively suggests a date prior to the mid sixth century. If LRA 2 was indeed used (predominantly) for the transport of olive oil, its presence in the Cide region may reflect the Black Sea basin’s ‘enormous thirst’ for olive oil.15 Seven fragments bear a strong resemblance to both the Kapitän II and Zeest 80/Knossos Type 39 amphorae. The origin (or origins) of these types, which share macroscopic characteristics,16 remains poorly understood: a (northern) Black Sea origin has been postulated,17 yet a Chian origin has also been proposed for the Kapitän II.18 Further, four fragments appear to be of central North African origin. They are presumably middle to late Roman in date and all come from the wider area around Gideros; one fragment is tentatively assigned to the late Roman period.19 Amphorae from Sinope and its environs – tentatively associated with the transport of wine, (olive?) oil and non-liquid contents such as fish products – are represented by 21 fragments.20 The identification is based on both colour and the presence of black volcanic inclusions (mostly pyroxene). Caution is, however, required with regard to the presence/ absence of pyroxenes, in particular regarding the pâte rosée.21 Only one fragment could be 11 12 13 14 15 16 17 18 19 20 21

Kassab Tezgör et al. 2003: 172–3, nos 7–9. Lawall 2004: 180, figs 3, 4 left. Robinson 1959: 109–10, pl. 29; Reynolds 2010: 97; Pieri 2005: 132–7; on shipwreck finds around Sinope, see Kassab Tezgör et al. 2003: 181–3, nos 26–7. Karagiorgou 2001; Reynolds 2004: 231–2; Pieri 2005: 85–93; Slane and Sanders 2005: 286–7; Reynolds 2010: 95–7. Braund 2005: 122. At Tanagra and Hyettos in Boeotia, central Greece (personal observation). Reynolds 2010: 90. Opaiţ and Paraschiv 2013: 319–20. Bonifay 2004: 9–44, 89–153; Franco 2012. Demr 2007; Lund 2007; Kassab Tezgör 2010: 121–2, 127, 133–4, 137. Erten et al. 2004: 105; also Kassab Tezgör 2010: 121, 123, 134–7.

Roman-period finds from the Cide region

27

typologically identified: in pâte blanche/claire, it belongs to type D Snp III, datable no later than the second half of the sixth century.22 2.3 Cooking wares and other forms Eleven fragments have been confidently classified as cooking wares; the scarcity of reference points other than fabric allows the dating of nine only generally to the Hellenistic to late Roman period. One fragment is thought to come from northern Turkey, given the presence of black volcanic inclusions. One handle fragment contains abundant silvery micaceous inclusions, for which a source in western Turkey can be proposed.23 One fragment of likely Roman date is the rim of a (small) pithos, a large(r) vessel type principally used for storing dry or liquid foodstuffs. Though our understanding of the production, use, life-span and socio-economic context of pithoi is relatively limited, some work has been carried out.24

3. Discussion The limited quantity of identified Roman pottery precludes any detailed quantitative analyses or studies of surface scatters. Some interesting observations can nevertheless be made concerning the functional and chronological dimensions of the pottery and, combined with the architectural finds, their spatial distribution within the CAP survey region. 3.1 Provenance and proportions A relatively significant proportion of the pottery comes from sources that lie well beyond northwestern Asia Minor, such as amphorae from the Aegean and central North Africa, and tablewares from western Turkey. From less distant sources come Sinope-region amphorae and, in particular, a group of narrow-necked amphorae (see below). On the whole, only a small proportion does not originate from a Pontic or Aegean source. The absence of certain well-known classes, however, is noteworthy. No fragments of Late Roman Amphora 1 (LRA 1) were identified; this is a type that was manufactured predominantly in Cilicia and Cyprus and which is attested far and wide at late Roman Mediterranean sites and beyond.25 Tableware products that were supposedly manufactured in or around Hadrianopolis have not been identified,26 nor has African Red Slip Ware (ARSW) from Tunisia.27 Above all, the absence of Pontic Sigillata and Pontic Red Slip Wares is unexpected, 22 23 24 25 26 27

Kassab Tezgör 2010: 135–7. The pâte blanche/claire need not have originated from Sinopean workshops alone: Erten et al. 2004: 104–5; Reynolds 2010: 96. Slane and Sanders 2005: 255–6. De Paepe and Vermeulen 1999; Devos et al. 1999; De Paepe et al. 2001; for a useful synthesis, see Giannopoulou 2010. Decker 2001. LRA 1 possibly (partly) circulated in conjunction with LRA 2: Karagiorgou 2001. Lafli and Kan Şahn 2012a: esp. 51–3. Hayes 1972: 13–299; Bonifay 2004, 45–65, 154–210. ARSW has been attested at, e.g., Hadrianopolis (Lafli and Kan Şahn 2012b) and Sinope (Firat 2010).

28

Philip Bes

Fig. 3: Top part of an early Roman (?) gravestone reading ΓΛΑΥΚΟΣ ΜΕΝΕΦΡΟΝΟΣ ΚΡωMΝΕΙΤΕΣ (‘Glaukos, son of Menephron, from Kromna’) (©CAP). since both are commonly attested at Sinope, Neoklaudiopolis and Pompeiopolis.28 That only Phocaean LRC has been identified – at Sinope, for instance, a more varied tableware repertoire is observed – may be related to the region’s closer proximity to Constantinople. Phocaean LRC is well documented at Constantinople,29 the point of transition for seaborne traffic between the Black Sea and the Aegean, and from whence goods were brought to Asia Minor’s northern shores.30 3.2 Spatial and chronological distribution The majority of the datable pottery collected belongs to the middle Roman period and, above all, the late Roman period; this is a predominance that is well documented elsewhere, for example in rural Greece.31 Geomorphological processes in the Cide region have altered the landscape considerably since antiquity – particularly the coastal and near-coastal zones, where Roman occupation is thought to have been predominantly concentrated – and this (partly) explains the scarcity of Hellenistic and Roman surface finds.32 In spite of this, 14 28 29 30 31 32

Firat 2010; Winther-Jacobsen and Bekker-Nielsen forthcoming; and Domżalski 2011 respectively. Hayes 1972: 418; 1992: 5, 7. Belke 1996: 135–7. Pettegrew 2007. Düring and Glatz forthcoming: esp. 2.3.

Roman-period finds from the Cide region

29

clusters with Roman features or finds can be distinguished: CAP’s site type-classification allows to define some as groups (a collective of find concentrations and/or otherwise identifiable (architectural) traces of (past) occupation/activity), in other cases they are features or spot sites, and in many cases they are chance discoveries and spolia that are of interest.33 Most of these were not significant centres of past human activity, and, since the artefact concentrations are mostly small to very small, this makes it not only difficult to investigate the associations between certain finds, but also, more importantly, to conclude whether these clusters can be referred to as habitation sites or if they represent other activity zones, whether or not they were permanently occupied and where specific agricultural, artisanal or other economic activities were carried out, such as olive-oil production, wine making or pottery production (Fig. 2; Table 2). These 14 clusters are discussed more comprehensively elsewhere34 – here, Table 2 serves as a summary – yet a few significant finds may be singled out. Firstly, an ex situ early Roman(?) gravestone reads ΓΛΑΥΚΟΣ ΜΕΝΕΦΡΟΝΟΣ ΚΡωMΝΕΙΤΕΣ (Fig. 3), and Kromneites could well signify that Glaukos, son of Menephron, was a native of Kromna, thought to be located at present-day Tekkeönü, some 27km west of Cide.35 Secondly, rescue excavations by the Museum of Kastamonu have revealed two cist graves at Türbe Tepe Üste (Cide), datable to the Roman period through the grave goods.36 Significant finds that were not associated with any of these clusters include an early or middle Roman Ionic-Attic column base (an isolated find in the village of Ovacık), a fortification at Kazallı Kalesi, ca 6km east of Cide that contains (a) Roman-period structure(s) which was/were reused in the middle or late Byzantine period.37 Finally, the spectacular and well-preserved bridgehead at Atköprü, dramatically perched on the western cliff alongside the Devrekani (Fig. 4).38

4. Interpretation of the data The ceramic finds and other remains, sometimes architecturally monumental, collected and mapped during three years of survey palely reflect a now largely obscured – due to human and natural factors – Roman cultural and religious landscape that, so it appears, was relatively intensively occupied and populated.39 Despite these circumstances, some observations may nonetheless be made. First and foremost, most of the clusters are coastal or near-coastal. Only two (11 and 12) are situated inland, ca 15km and 14km from the present coastline respectively. Whilst the coverage by CAP was inevitably selective (non-coastal data clusters are obviously also of note, such as Okçular, for instance, which could very well have functioned as a refuge in times of unrest) this general trend adds weight to the hypothesis that Roman occupation/ 33 34 35 36 37 38 39

Glatz et al. forthcoming: esp. table 4.2. Bes forthcoming. Marek 1993: 17, 185; Belke 1996: 241–2. Arrian gives a distance of 150 stadia between Kromna and Egilan/Aigialos/Cide, which corresponds to some 30km: Arrian, Periplus 14. Bleda Düring, personal communication. Cassis forthcoming. Marek 1993: 10, 92. Roueché 2000: 579.

30

Philip Bes

Fig. 4: The bridgehead along the Devrekani at Atköprü. Note the wall in the lower right of the picture, which is part of the bridgehead on the eastern side (©CAP). habitation of the Cide region was largely a coastal affair. Secondly, the majority of the typologically, chronologically or otherwise identifiable pottery fragments belong to the (middle to) late Roman period; to what degree the (middle to) late Roman clusters were occupied contemporaneously remains unanswerable. The pottery by and large reflects exchange trends with west-east (largely Aegean; e.g. amphorae, Phocaean LRC) and east-west (e.g. Sinopean amphorae) directions. Most, if not all, of the non-Black Sea pottery presumably flowed into the Pontic basin via Constantinople, and the actual transport of goods, particularly international exchange, took place by ship.40 Why then only Phocaean LRC is identified remains unclear; that red slip tablewares present a more mixed picture at Sinope might well mirror that city’s economic and redistributive role, which surely promoted an important link with Constantinople. Even if only selected parts of the coastal zone could be surveyed, with an intermediate distance of ca 2km (except between clusters 2–3, where it is ca 6km, and the outlier clusters 13 and 14), these clusters do occur with some regularity. One or more of these were surely good anchorages, also for larger ships, that offered sufficient opportunities for ‘harbourhopping’. In addition to the (partly conjectural) coastal road, Stephen Mitchell rightly 40

Belke 1996: 127; Braund 2005: 115.

Roman-period finds from the Cide region

31

points out that the (largely unknown) non-coastal road network played a crucial role between and within inland regions as well as connecting to the coast;41 the bridgehead at Atköprü attests to this.42 Finally, whereas the scarcity of finds hampers a detailed discussion of site function and settlement hierarchy (cities, villages, estates, etc.), the fact that part of the pottery is non-regional but also functionally varied elucidates the occupational and consumptive character of these settlements, thus indicating that the region actively, albeit largely indirectly, took part in exchange patterns.43 There are some archaeological and literary indications, however, that suggest the region was not merely an inactive consumer of extra-regional goods, but, in fact, held some trump cards that allowed it to take part in production and consumption networks, also during the fairly peaceful centuries of the late Roman period.44 At the same time, this latter period was all but static: the rise of Christianity and monasticism, the changing nature and roles of city and countryside, and also external threats – even if Honorias-Paphlagonia basically remained unaffected – were some of the broader aspects that directed society. As a matter of fact, the Cide region shows signs of socio-cultural and occupational, and perhaps also economic continuity following the disruptive seventh century, even if that continuity was subject to further change.45 Several classical authors refer to northern Asia Minor as a source for quality timber, which perhaps was intended predominantly for shipbuilding.46 It is plausible to think that this industry also prompted ancillary industries, such as the manufacture and upkeep of nets and sails.47 We remain uninformed, however, as to whether this timber was used only locally or was also transported to, for example, Constantinople and other locations around the Pontic basin. We should suppose that these coastal communities practised some level of local shipbuilding and maintenance, if only for fishing activities and transporting goods to and from nearby settlements. In the event that the timber was also destined for further afield, and why should it not be, we may wonder whether in that context shipbuilding was partly decentralised along the northern shores of Asia Minor or whether shipbuilding and related industries were more centralised, in Constantinople for example; it is difficult to conceive that the capital, a city that saw the continuous coming and going of ships of all sorts and sizes, did not have an extensive and highly organised infrastructure for the construction, repair and maintenance of ships. With the exploitation of timber for shipbuilding may have come other commercial exploitation of the resource, perhaps for instance the working of the wood of the Buxus sempervirens L. (boxwood). Its wood is ideally suited for the manufacture of tools and other objects – its use in shipbuilding comes to mind, in the form of tools for instance – and Kytoros and Amastris are specifically mentioned as sources of quality boxwood.48 41 42 43 44 45 46

Belke 1996: 127–30; Foss 2000; Matthews et al. 2009: 187–9, figs 6.25–7. Mitchell 1993: 245–6. Belke 1996: 138–51. Roueché 2000. Cassis forthcoming. Broughton 1975: 836–7; Meiggs 1982: 47, 372, 393; Belke 1996: 139; Bittner 1998: 122–3; Braund 2005: 122. 47 Mitchell 2005: 102–3. 48 Among others Theophrastos; Pliny the Elder; Strabo; for which also see Meiggs 1982: 282; Marek 1993: 100; Belke 1996: 140, esp. n. 12, 245–6; Bittner 1998: 125; Braund 2005: 124.

32

Philip Bes

Fig. 5: Fragments of (narrow-necked) amphorae from Cide (cluster 10b), probably locally manufactured (©CAP). Slightly more tangible is the, albeit restricted, archaeological evidence for the local manufacture of amphorae. Noteworthy in this context are about 95 fragments of what appear to be narrow-necked amphorae, which were collected in particular at clusters 8–10 – the wider area of Cide – though similar fragments show up in a thin diffused scatter throughout the survey area (Fig. 5). All the fragments inspected share a homogeneous fabric, which is 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 and calcitic and/or limestone bits; pores/voids are not infrequent. Morphologically, the handles often have pronounced ridges and can be made rather crudely. The few rims that were seen are lightly thickened yet otherwise plain; body sherds are by and large ribbed. Of particular interest is a production waster, found where more were noted in the field.49 The combination of the quantity, the attested wasters and the homogeneous fabric character amongst the fragments of this relatively sizeable group are reason enough to postulate a local origin. Even if the claim must remain tentative until, perhaps, one day, it can be substantiated through solid evidence (geophysical anomalies, pottery workshops, kilns, etc.) acquired by geophysical prospection and/or archaeological excavation, the proposed scenario suggests that at one or more locations in the wider Cide area there were workshops manufacturing amphorae. As far as the morphology of the fragments allows for any firm statements, interestingly, some fragments nevertheless strongly suggest that they formed 49

Claudia Glatz, personal communication.

Roman-period finds from the Cide region

33

(smaller) amphorae with a high and narrow neck; such traits are not at all unfamiliar: both Herakleia Pontike and Sinope are known to have manufactured similar amphorae. Thus the amphorae thought to have been produced at or near Cide belong to a broader, regional phenomenon of amphorae manufacture along the northwestern coast of Asia Minor.50 Both Herakleian and Sinopean amphorae are attested far and wide,51 yet any statement about the distribution of the amphorae potentially from the area of Cide is impossible. It is not unthinkable that its dispersal – if this did occur – tapped into existing exchange systems, for example those that, among others, accommodated the distribution of Sinopean and other (regional) amphorae around the Black Sea, into the Aegean and around the Mediterranean as a whole.52 The content of these amphorae, regrettably, is another vexing matter, and one that remains unanswerable at present. Although CAP did come across clues, tentative though they are, about ancient agricultural activities, no direct relation to the amphorae has been established. The morphology of the amphorae, if their identification as narrow-necked vessels is correct, does, however, offer a clue. The average neck/rim diameter of the Cide examples, as well as that of its ‘family members’, ranges around 3–4cm, which suggests that the vessels were unlikely filled with anything less than fluid, let alone a more or less solid produce. Rather, liquid matters such as olive oil or, more likely, wine are obviously more suitable candidates.53 In addition, looking to the Tabula Peutingeriana as well as remarks by ancient authors, some of the clusters may be (very) conjecturally identified with some of the ancient settlements along the coast (Table 3). First, cluster 3 is perhaps to be associated with ancient Kytoros – likely echoed in the modern name Gideros – that was equipped with a (natural) harbour.54 It is worth mentioning that the distance between Cythero and Egilan as indicated on the Tabula Peutingeriana is 8 (Roman) miles (=ca 11.8km), which roughly equates to the distance between Gideros (=Kytoros) and Cide (=Egilan/Aigialos), as well as to the 60 stadia (=12km) mentioned by Arrian, who meaningfully also mentions a ‘mooring for ships’.55 Secondly, in relation to cluster 10, Klaus Belke has put forward the idea that ancient Aigialos might be situated near today’s Cide, perhaps at Karaağaç Limanı west of Cide.56 As a matter of fact, Apollonios Rhodios and Strabo both use the name Aigialoi for a stretch of the shore, 100 or more stadia long.57 Strabo further mentions a village of the same name that is ‘ten schoeni distant from Amastris’, a distance that corresponds to some 55km, which in turn approaches the distance between modern Amasra and Cide.58 Thirdly and last, the scarce finds at cluster 14 may tentatively be connected to ancient Thymaina, located some 17–18km east of Cide; it is alternatively known as Thymena, Teuthrania or 50 51 52 53 54

55 56 57 58

Kassab Tezgör 2010: pls 8.4, 22, (related to) type D Snp I. Pieri 2007. Karagiorgou 2001; Swan 2004; 2007; Braund 2005: 120, 127–30; Salmeri 2005: 197. Opaiţ 2007: 101. Marek 1993: 17–8; Belke 1996: 245–6; Foss 2000; Kytoros is not to be confused with Kotyora, to which the Kytoros mentioned by Strabo refers: Strabo 12.3.17; Arrian, Periplus 117; Ercyas 2007: 1196. Arrian, Periplus 14; http://www.tabula-peutingeriana.de/tp/tpx.html, accessed 12 January 2015. Belke 1996: 158; Foss 2000; Marek 1993: 185–7. Arrian, Periplus 112. Strabo 12.3.8, 10.

34

Philip Bes

Timle – the latter name is given to the two Byzantine kales.59 Arrian offers a distance of 90 stadia, about 18km.60 The schematic representation in Table 3, however, makes it clear that trying to associate the distances given in one or more of the ancient sources with archaeological and/or modern settlements does not lead to unambiguous results; this particularly applies to the left half of Table 3. This is not unexpected, as ancient measures could vary; see, for instance, the different distances between Amastris and Aigialos. Also, the given or suggested location of an ancient site can be doubtful or misinterpreted. One should also not overlook the fact that the topography, more often than not, will have determined the trajectory and thus its length.

5. Conclusion Though the archaeological evidence collected by CAP is too scarce to reconstruct in detail the Roman-period settlement pattern for this region, it does help, however, in forwarding some tentative thoughts with regard to the localisation/identification of ancient sites along the coast (Table 3). The combined available data, archaeological and literary, do indeed provide significant clues for an active and possibly relatively well-populated Roman cultural landscape – some architectural remains suggest that at times the urban landscape must have housed some monumental structures. That at least the ceramic evidence is most clear for the late Roman period – with continuation into the Byzantine period – need not result from survey and geomorphological circumstances alone; the region’s relative proximity to the empire’s capital possibly augmented already existing economic activities, for instance concerning the exploitation of wood/timber and grain, and plausibly initiated or advanced the manufacture of amphorae in the area of Cide proper. Evidence for this not only comes from the coastal parts alone; this implies that the rurally settled landscape (Okçular, Gideros/Abdulkadir and Çamdibi)61 was able to tap into the pool of imported goods as well. The continuing settlement occupation in the region of Cide following the late Roman period62 – even if its character changed to comprise a landscape with a more rural character that included churches – does not seem to conform readily to that observed for Paphlagonia, where the centuries after the seventh century are (partly) characterised by ‘fortified hilltop sites’;63 the region of Cide, nonetheless, continued to be a suitable place for human occupation, as it had been for many previous millennia.64 Philip Bes p/a Mariastraat 69 2595 GM Den Haag, The Netherlands [email protected] 59 60 61 62 63 64

Marek 1993: 84; Belke 1996: 274–5; Foss 2000. Arrian, Periplus 14. Glatz and Düring forthcoming a. Cassis forthcoming; Glatz and Düring forthcoming a. Matthews et al. 2009: 190–9, figs 6.29, 6.35. Glatz and Düring forthcoming b.

Roman-period finds from the Cide region

35

Bibliography Adam, J.-P. 1984. La construction romaine. Matériaux et techniques. Paris. Asheri, D. and W. Hoepfner 1972. Über die Frühgeschichte von Herakleia Pontike: Topographische Forschungen (Forschungen an der Nordküste Kleinasiens 1). Vienna. Bardill, J. and J. W. Hayes 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. Paphlagonien und Honorias (Tabula Imperii Byzantini 9). Vienna. 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 (Roman and Late Antique Mediterranean Pottery 6). Oxford. Bes, P. M. forthcoming. ‘Chapter X. The Cide Region in the Roman Period’ in Düring and Glatz (eds). Bittner, A. 1998. Gesellschaft und Wirtschaft in Herakleia Pontike. Eine Polis zwischen Tyrannis und Selbstverwaltung (Asia Minor Studien 30). Bonn. Bonifay, M. 2004. Études sur la céramique romaine tardive d’Afrique (British Archaeological Reports International Series 1301). Oxford. Braund, D. 2005. ‘Across the Black Sea. Patterns of Maritime Exchange on the Northern Periphery of Roman Asia Minor’ in Mitchell and Katsari (eds), 115–38. Broughton, T. R. S. 1975 [reprint]. ‘Part 4: Roman Asia’ in T. Frank (ed.), An Economic Survey of Ancient Rome. New York, 499–918. Cassis, M. forthcoming. ‘Chapter XI. The Cide and Şenpazar Region in the Byzantine Period’ in Düring and Glatz (eds). 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 (British Archaeological Reports International Series 2517). Oxford, 109–14. De Paepe, P. and F. Vermeulen 1999. ‘Étude microscopique et chimique de pithoi, matériaux de construction et autres terres cuites antiques de Pessinonte (Turquie)’, Anatolia Antiqua 7, 115–27. De Paepe, P., G. Devos and F. Vermeulen 2001. ‘Les pithoi de Pessinonte (Anatolie centrale). Données analytiques complémentaires (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, 69–86. Demr, 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: Proceedings of the International, Interdisciplinary Conference, Istanbul, 14–16 October 2004 (British Institute at Ankara Monograph 42). London, 57–64. Devos, G., P. De Paepe and F. Vermeulen 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) (Schriften des Zentrums für Archäologie und Kulturgeschichte des Schwarzmeerraumes 21). Langenweißbach, 163–78. Düring, B. S. and C. G. Glatz forthcoming. ‘Chapter II. Remote Landscapes in Flux – The Cide and Şenpazar Region’ in Düring and Glatz (eds). Düring, B. S. and C. G. Glatz (eds) forthcoming. Kinetic Landscapes: The Cide Archaeological Project 2009–2011: surveying the Western Turkish Black Sea Region. Warsaw. Eiring, J. and J. Lund (eds) 2004. Transport Amphorae and Trade in the Eastern Mediterranean. Acts of the International Colloquium at the Danish Institute at Athens, September 26–29, 2002, Athens (Monographs of the Danish Institute at Athens 5). Athens.

36

Philip Bes

Ercyas, B. D. 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 (British Archaeological Reports International Series 1675). Oxford, 1195–206. Erten, H. N., D. Kassab Tezgör, I. R. Türkmen and A. Zararsiz 2004. ‘The Typology and Trade of the Amphorae of Sinope. Archaeological Study and Scientific Analyses’ in Eiring and Lund (eds), 103–15. Firat, 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 (Varia Anatolica 22). Paris, 187–95. Foss, C. 2000. ‘Map 86 Paphlagonia’ in R. J. A. Talbert (ed.), Barrington Atlas of the Greek and Roman World. Princeton. Franco, P. 2012. African Amphorae from Portus (Unpublished PhD, University of Southampton). Southampton. Gabrielsen, V. and J. Lund (eds) 2007. The Black Sea in Antiquity. Regional and Interregional Economic Exchanges (Black Sea Studies 6). Aarhus. Giannopoulou, M. 2010. Pithoi: Technology and History of Storage Vessels through the Ages (British Archaeological Reports International Series 2140). Oxford. Glatz, C. G. and B. S. Düring forthcoming a. ‘Chapter XIV. A Tale of Three Landscapes: Okçular, Abdulkadir, and Çamdibi in Long-Term Perspective’ in Düring and Glatz (eds). Glatz, C. G. and B. S. Düring forthcoming b. ‘Chapter XV. Conclusions’ in Düring and Glatz (eds). Glatz, C. G., B. S. Düring and T. Wilkinson forthcoming. ‘Chapter IV. Developing and Adaptive Field Methodology for challenging Landscapes’ in Düring and Glatz (eds). Hayes, J. W. 1972. Late Roman Pottery. London. Hayes, J. W. 1992. Excavations at Saraçhane in Istanbul 2. The Pottery. Princeton. 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.), Colloque international pour l’étude de la mosaïque antique et médiévale 9, Rome, Italy 5 Novembre 2001 (Collection de l’École française de Rome 352) (La mosaïque gréco-romaine 9.2). Rome, 1083–101. Karagiorgou, 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, 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 (Varia Anatolica 22). Paris, 121–40. Kassab Tezgör, D., S. Lemaître and D. Pieri 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. Lafli, E. and E. Christof 2012. Hadrianopolis 1. Inschriften aus Paphlagonia (British Archaeological Reports International Series 2366). Oxford. Lafli, E. and G. Kan Şahn 2012a. ‘Terra Sigillata and Red-Slipped Ware from Hadrianopolis in Southwestern Paphlagonia’, Anatolia Antiqua 20, 45–120. Lafli, E. and G. Kan Şahn 2012b. ‘Pottery from Southwestern Paphlagonia 1: Terra Sigillata and Red-Slipped Ware’ in B. Ramminger and O. Stilborg (eds), Naturwissenschaftliche Analysen vorund frühgeschichtlicher Keramik 2. Methoden, Anwendungsbereiche, Auswertungsmöglichkeiten (Universitätsforschungen zur prähistorischen Archäologie 216). Bonn, 179–216. Lawall, M. L. 2004. ‘Archaeological Context and Aegean Amphora Chronologies: A Case Study of Hellenistic Ephesos’ in Eiring and Lund (eds), 171–88.

Roman-period finds from the Cide region

37

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 (Black Sea Studies 2). Aarhus, 161–9. Marek, C. 1993. Stadt, Ära und Territorium in Pontus-Bithynia und Nord-Galatia. Tübingen. Matthews, R. and C. G. Glatz (eds) 2009. At Empire’s Edge: Project Paphlagonia. Regional Survey in North-Central Turkey (British Institute at Ankara Monograph 44). London. Matthews, R., M. Metcalfe and D. Cottica 2009. ‘Landscapes with Figures: Paphlagonia through the Hellenistic, Roman and Byzantine Periods, 330 BC–AD 1453’ in Matthews and Glatz (eds), 173–226. Meiggs, R. 1982. Trees and Timber in the Ancient Mediterranean World. Oxford. Mitchell, S. 1993. The Rise of the Church (Anatolia. Land, Men, and Gods in Asia Minor 2). Oxford. Mitchell, S. 2005. ‘Olive Cultivation in the Economy of Roman Asia Minor’ in Mitchell and Katsari (eds), 83–113. Mitchell, S. and C. Katsari (eds) 2005. Patterns in the Economy of Roman Asia Minor. Swansea Mommsen, H. and S. Japp 2009. ‘Neutronenaktivierungsanalyse von 161 Keramikproben aus Pergamon und Fundorten der Region’, Istanbuler Mitteilungen 59, 269–86. Opaiţ, 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 (Black Sea Studies 6). Aarhus, 101–21. Opaiţ, A. and D. Paraschiv 2013. ‘On the Wine, Olive Oil and Fish Supply of the Countryside in Roman Dobroudja (1st–3rd Centuries AD)’ in L. Buzoianu, P. Dupont and V. Lungu (eds), Production and Trade of Amphorae in the Black Sea III. Actes de la Table Ronde internationale de Constanţa, 6–10 octobre 2009 (Pontica XLVI Supplementum II). Constanţa, 317–33. Peña, J. T. 2007. Roman Pottery in the Archaeological Record. Cambridge. 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 (Bibliothèque Archéologique et Historique 174). Paris. Pieri, D. 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. Actes du 2e colloque international sur la Syrie antique (Tours, 12–13 juin 2003) (Topoi Supplément 8). Lyon, 297–327. 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, 224–69. Reynolds, P. 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. 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 (British Archaeological Reports International Series 2185 I). Oxford, 89–114. Robinson, H. S. 1959. Pottery of the Roman Period. Chronology (The Athenian Agora 5). Princeton. Roueché, C. 2000. ‘Asia Minor and Cyprus’ in A. Cameron, B. Ward-Perkins and M. Whitby (eds), The Cambridge Ancient History XIV. Late Antiquity: Empire and Successors, A.D. 425–600. Cambridge, 571–87. Salmeri, G. 2005. ‘Central Power Intervention and the Economy of the Provinces in the Roman Empire’ in Mitchell and Katsari (eds), 187–206. Schneider, G. and S. Japp 2009. ‘Röntgenfluoreszenzanalysen von 115 Keramikproben aus Pergamon, Çandarlı, Elaia und Atarneus (Türkei)’, Istanbuler Mitteilungen 59, 287–306. Slane, K. W. and G. D. R. Sanders 2005. ‘Corinth: Late Roman Horizons’, Hesperia 74.2, 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 Eiring and Lund (eds), 371–82. Swan, V. G. 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 (Proceedings of the British Academy 141). Oxford, 251–80.

38

Philip Bes

Tsetskhladze, G. R. (ed.) 2012. The Black Sea, Paphlagonia, Pontus and Phrygia in Antiquity. Aspects of Archaeology and Ancient History (British Archaeological Reports International Series 2432). Oxford. 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. Acts from a Ph.D.-Seminar for Young Scholars, Sandbjerg Manorhouse, 12–15 February 1998 (Halicarnassian Studies 3). Odense, 132–8. Winther-Jacobsen, K. and T. Bekker-Nielsen forthcoming. ‘A Late Roman building complex in the Papaz Tarlası, Vezirköprü (ancient Neoklaudiopolis, northern Asia Minor)’, Proceedings of the Danish Institute at Athens 8. 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) (Schriften des Zentrums für Archäologie und Kulturgeschichte des Schwarzmeerraumes 21). Langenweißbach, 149–62.

Table 1 Absolute and relative counts of the identified pottery, by functional category, type and/or fabric ©Philip Bes/CAP Functional Category Tablewares

Amphorae

Cooking Wares ‘Other’ Total

Category

n

Phocaean LRC Phocaean LRC/Çandarlı Ware Jugs Other Local/Regional amphorae Sinopean pâte rosée Sinopean pâte rouge/colorée Sinopean pâte blanche/claire Sinopean LRA 2 Central North African Aegean (Rhodian, Knidian, Chian, Dressel 2–4) Agora M273/Samos Cistern Type Ephesian Nikandros Group Aegean/Black Sea? Kapitän II (and related) Zeest 80/Knossos Type 39 Unidentified LR Aegean Micaceous Ware Northern Turkey? Unidentified Amphora/Cooking vessel Small pithos

8 3 3 4 95 4 10 5 2 13 4 11 2 1 1 5 2 25 1 1 9 1 1

Subtotal

%

Chronology (ca) LR

18

8.53

R–LR HELL.–LR CLASS.–LR (M)R–LR

100–300 300–500 500–650 100–650 LR

180

85.31

R–LR HELL.–LR R–LR CLASS.–MHELL. R–LR R–LR MR–LR HELL.–(E)BYZ. LR

11

5.21 R–LR HELL.–LR

0.47 LR–EBYZ. 2 0.47 R–LR 211 100.00

39

Roman-period finds from the Cide region

Table 2 Background data for the 14 clusters ©Philip Bes/CAP Clus- Location CAP ter Site No(s) 1 Coastal, 100, Gökçe102, kale 104, 106, 108, 176, 178 2

3a

3b

3c

4

Coastal, west of and around Yenice Coastal, around Bay of Gideros

Coastal, southeast of Gideros Coastal, east of Gideros

168, 170

12, 21, 59, 123

17

16, 19, 66

Near– coastal, ca. 3.5km southeast of Kalafat

CAP Unit No(s) 3231, 3233

PotFinds tery Count 5 Amphorae: Sinopean pâte colorée and claire; MR–LR central North African (?); Zeest 80/Knossos Type 39; 16–17 disturbed cist graves (R–LR?), more undisturbed; (lever?) press installation; limestone quarry; ancient road pavement; building remains 1108–9 3 HELL. handle; Phocaean LRC; MR–LR central North African amphora; 2 MR–LR (?) cist graves; Roman (?) building blocks (including ashlar) – ? Knidian amphora; R–LR roof tiles; Roman (?) pottery fragments; LRA 2; MR–LR central North African amphora; Sinopean amphorae in pâte claire and undet. fabric; Phocaean LRC, Hayes form 3F; rectangular basin, base or top moulding (podium?), door threshold (dowel, pivot and bolt holes), ashlar with dowel hole, largely buried column fragment, cist grave – ? R–LR (?) pottery fragments

7, 253, 276, 289, 298, 381, 461 2213

8

2

Ancient Site (?) –

Bibliographic References –

–

–

Belke 1996: 50, 245–6; Kytoros, Foss Cythero, 2000; Cytorus, Tab.Peut. Cytorum section 8; (Gideros?) http://en. wikipedia. org/wiki/ Cytorus R–LR roof tiles; Knidian amphora; Phocaean LRC; Sinopean, Aegean and central North African amphorae; chamosorion; BYZ. monastic complex at Çadır İni may originate in the LR–EBYZ. period (Cassis forthcoming) – – Two pottery fragments: 1× HELL.?

40

Philip Bes

Clus- Location CAP ter Site No(s) 5 Coastal, 73, ca. 800m 160 east of Kalafat

24, 61, 64, 77

6

Nearcoastal, Okçular Valley south and southeast of Okçular Kale

7

Coastal, 109 west of Kumluca

8

– Nearcoastal, ca. 2km east/ southeast of Irmakköy 1, 140, 142, 148, 150, Coastal, 152 between Irmakköy and 2–3, Cide 67, 70–1, 127, 129, 208

9a

9b

CAP Unit No(s) 1067, 1075, 1088

–

Bibliographic References –

–

–

–

–

–

–

R–LR (?) roof tiles; 2 large (grain?) – mortars?/press weights?; ancient (?) basin with clamps (reused?); marble Ionic capital; 3 large dressed blocks: one is possibly a top moulding, others parts of a door frame?; clamp and lifting holes visible 63(+?) Sinopean pâte rose (?), pâte colorée – 3135, (5×); Kapitän II handle; LRA 2 3142–3, 3146–8 (2×); Phocaean LRC; blocks (R–LR, BYZ.?); inscribed gravestone (ER?); R–LR (?) columns (2×); possible door lintel; 2 dressed blocks

–

44, 89, 147, 174, 176, 185, 187, 189, 593, 673, 679, 683, 685, 2239 2104, 2185, 2189

PotFinds tery Count 8 HELL. Rhodian (?) handle; Phocaean LRC, Hayes form 3F; MR–LR jug handle; LRA 2 (wavy combing); Black Sea amphora (related to Kapitän II/Zeest 80/ Knossos Type 39); EastMed amphora?; large (ER?) spolium, possibly upper moulding of a podium; worked stones alongside nearby stream 29 Echinus bowl (?); Sinopean pâte rose, pâte colorée (2×), pâte claire Type D Snp III (6th century?); Phocaean LRC; LRA 2 (3×); MR–LR central North African amphora; column (base?); column base; two ancient cistern mouths, reused as well heads

6

7 3193, 3199, 3205, 3210, 3216–7, 3225 –

?

LRA 2; Sinopean pâte colorée; Phocaean LRC, Hayes form 3E; amphorae related to Kapitän II/ Zeest 80/Knossos Type 39 (2×) No diagnostic pottery and architecture; possibly the small fortification at Hıdır Kalesi (cf. Belke 1996: 216)

Ancient Site (?)

–

41

Roman-period finds from the Cide region Clus- Location CAP ter Site No(s) 114 10a Nearcoastal, west of Cide

CAP Unit No(s) 814, 842, 846, 852, 856, 858 –

PotFinds Ancient tery Site (?) Count 31 Knidian amphora; Phocaean LRC, Hayes forms 3E–F, 10A; Sinopean pâte colorée; local (LR?) amphorae

6

?

10b

Coastal, Cide

196, 206, 502, 519

11

Inland, Loç Valley, between Çamdibi and Şenköy Inland, 14–5km southwest of Aybasan Coastal, ca. 9km east of Cide; around Kuşçu

88, 90, 120, 122, 126, 134

2086, 2108, 3049, 3086

–

12 707, 2127, 2140, 3097–8

12

13

14

10–1, – 89

Coastal, 110 ca. 17km east of Cide; around Uğurlu

–

?

2

Bibliographic References

Belke 1996: 158; Aigialos?, Foss 2000; Egilan? Local (LR?) amphorae; 2 cist Tab.Peut. graves (excavated); blocks of R–LR section 8 public building; column; Roman milestone; Zeus sanctuary ca. 1km southwest of Cide (Belke 1996: 158; Marek 1993: 98–9, n. 685) – – LRA 2 (2×); R–LR/BYZ. square, brick-lined well; pressing stone (?); fluted column/statue fragment (?); stone basin with outlet and carved rosette (reused sarcophagus?); brick/tile grave? HELL. cooking ware; MR–LR – Agora M273/Samos Cistern Type

–

R–LR roof tiles; ceiling cassette

–

– (perhaps funerary monument); stone gutter/imbrex or cancelli pier with slit for chancel screen (though see Adam 1984: 281, fig. 601); several ashlars (modern reuse) of a monumental/public building; moulded fragments (doorpost?); 2 small columns? (tapering: not columns, or recut?) Knidian amphora?; LR–EBYZ. Thymaifragment; possible R–LR wall na, Timle, (Belke 1996: 274–5) Thymena, Teuthrania?

Belke 1996: 274–5; Foss 2000

42

Philip Bes

Table 3 Tentative identification of CAP find clusters with ancient sites indicated on the Tabula Peutingeriana and/or mentioned in ancient sources ©Philip Bes/CAP Ancient Site (?)

Kytoros/ Kytoron/ Cytorus/ Cytorum

Kromna

Amastris

CroLocation on AmaCythero men Tab.Peut. sia? Distance on XI VII Tab. Peut. in mp(1) Distance ≈ ≈ in km 16.3 10.4 Distance in Arr. Peripl. in stadia(2)

Cluster Distance as the crow flies in km (1) (2) (3)

Karambis

Egilan

–

Carambas

VIII

XXVIII

≈ 11.8 60 ≈12km

41.4 210 ≈42km(3)

150 ≈30km

Distance in Strab. in schoeni (4) Modern Location

Aigialos

Thymaina/ Tymle/ Thymena/ Teuthrania

90 ≈18km

10 (≈55km) Hisarköy/ Tekkeönü? –

Amasra – ≈ 25

≈ 16

Gideros

In/ around Cide(5)

In/around Uğurlu

3a–c

10a–b

14

≈ 11.5

≈ 17

In/around Doğanyurt?; Kerembe Burnu? – ≈ 25

1 Roman mile (mp) ≈ 1.479m 1 stadium ≈ 200m Belke 1996 and Foss 2000, however, give the distance between Egilan/Aigialos/Cide and Karambis as ca 34km, which roughly corresponds to modern Ilyasbey and the adjacent landmark at Kerembe Burnu. (4) 1 schoenus ≈ 5.5km. Since ancient measures could vary, no reliable figure can be given. (5) The name Cide is also thought to have derived from Kytoros.

Owen Doonan

Settlement and economic intensification in the late Roman/ early Byzantine hinterland of Sinop

1. Introduction Sinop: nexus of Black Sea trade from the seventh century BC to the 19th century AD Sinop, ancient Sinope, the central port of the southern Black Sea coast, functioned as one of the principal focal points of trade and military power in the region from the foundation of the Ionian colony in the seventh century BC to the onset of the Crimean War in the mid 19th century AD. The centrality of Sinop has made the town and its hinterland highly sensitive to shifts in the economic and military trends of the Black Sea region. The establishment of Constantinople as the seat of economic and political power of the late Roman/early Byzantine period (fourth to seventh century) led to the most dramatic expansion of settlement in the hinterland of Sinope prior to the 20th century. An equally sudden collapse in rural settlement followed in the eighth century. This article examines evidence for settlement and industry on the Sinop promontory with particular focus on two valleys that apparently experienced dramatic early Byzantine expansions, the Demirci valley and the Kırkgeçit çayı valley.

2. Sinop Regional Archaeological Project: overview The Sinop Regional Archaeological Project (SRAP) aims to establish a fundamental record of settlement and human-ecological relations on the Sinop promontory from the early Holocene through to the Ottoman period. The primary method of investigation is systematic pedestrian archaeological survey of sample areas dispersed around the promontory. The survey area in total encompasses approximately 500km2 and is remarkably diverse in land-forms, ecology and occupational history. In pursuit of these aims, SRAP has conducted systematic archaeological survey of approximately 1,000 survey tracts organised into 40 sample quadrats. Each tract consists of a single agricultural field ranging in area from 200–1,000m2; each quadrat is a 1–2km2 area within which we survey as many recently ploughed and planted agricultural fields as we can locate. When possible, we survey only fields with sufficient visibility to record a reliable

44

Owen Doonan

Fig. 1: Map of the Sinop promontory showing locations mentioned in the text. Areas of systematic survey are indicated in white highlighting (base image Google Earth with additions). positive or negative result.1 Tracts and quadrats were mapped by hand onto 1:25,000 topographic maps provided by the Turkish government during the 1996–1999, 2003 and 2006 survey seasons, but since 2010 all mapping has been carried out on georectified Quickbird satellite images and organised into a project GIS; efforts are still continuing to integrate the results of the 1996–1999 surveys into the project GIS. The 1996 field season was an initial opportunistic survey and during the 1997 field season in the Demirci valley we did not collect quantitative information from surveyed tracts. Archaeological finds have been counted, weighed, photographed digitally and recorded by tract since the 1998 field season.

1

Doonan 2004a.

Settlement and economic intensification in the hinterland of Sinop

45

Fig. 2: Topographic map of the Demirci valley showing sample tracts and Roman-period archaeological loci. Loci where kiln-related debris was recorded are marked with a K (base map A. Gantos, ©SRAP archives) 2.1 Demirci valley The Demirci valley opens onto the east-facing shoreline of the Sinop promontory about 15km south of Sinop port. The valley runs about 5km east-west from the north-south ridge that defines the eastern side of the Karasu river valley and the present coast (Fig. 1). A ridge runs the entire length of the valley (maximum elevation 156m a.s.l.) and defines its northern side. The southern side of the valley is more broken up, defined by a range of low hills and side valleys oriented southwest-northeast (Fig. 2). Along the shore a pair of ridges running parallel to the coast close the flanks of the seaward exit of the valley. Soils on the slopes are composed of recently weathered bedrock, composed of mixed marine deposits.2 The valley floor is covered by recent alluvial deposits on top of older clays and bay sediments.3 The shoreline deposits are unstable on top of the underlying sediments, causing landslips which, aided by the strong longshore currents, have resulted in a deeply eroded coastline. It is clear that the late Roman coastline is currently beneath the sea. The modern shore of the valley was an important industrial zone in late Roman times, specifically the rise overlooking the beach from the north which was excavated by Dominique Kassab Tezgör in collaboration with the Sinop Museum;4 in four seasons of fieldwork nearly a dozen amphora kilns were documented. Our team did not survey in the immediate area around the excavation site, but a dense scatter of pottery extends about 2 3 4

For a detailed discussion of the geomorphology of the coast, see Kuzucuoğlu 2010. Doonan 2004b: ch. 2. Kassab Tezgör 2010a.

46

Owen Doonan

Fig. 3: Map of the interior of the Demirci valley and site L97-101 (base map A. Gantos, ©SRAP archives).

Fig. 4: Local ceramics from locus L97-101 in the interior of the Demirci valley. Assemblage is dominated by storage vessels and tiles, kiln debris and wasters (©SRAP archives). 1km along the shoreline south of the kiln site and more than half a kilometre inland.5 The seaward extent of the site and the nature of the port facilities are unknown, since they lie beneath the sea. Fragments of late Roman mosaics and monolithic columns have been reported by members of the Sinop Historical Society, and these are possibly remnants of a villa in the area. Brief summaries will be offered here of the survey results for the inland (Uzungürgen), central (Demirci, Eldevüz) and coastal (Yalı, Keçioğlu) areas of the Demirci valley. 5

Garlan and Kassab Tezgör 1996; Doonan 2004b.

Settlement and economic intensification in the hinterland of Sinop

47

The western end of the Demirci valley was investigated in 54 tracts in the village of Uzungürgen (Fig. 3). These tracts were covered by an almost continuous off-site scatter of abraded late Roman/early Byzantine sherds (mostly amphorae and pithoi) and tile fragments which may have been dispersed by ancient and modern ploughing and manuring. Several extensive, dense scatters were recorded in this part of the valley, of which locus L97101 will serve as an example. Locus L97-101 was the site of a settlement set on a terrace jutting east into the valley and facing directly along the axis of the valley to the sea. The primary scatter covers an area measuring approximately 200m × 200m, or 4ha. The dominant ceramics at the site include storage vessels (pithoi and amphorae) and domestic wares in the buff-pink, black-sand tempered ware typical of local Sinopean fabrics; there are few or no intrusive wares (Fig. 4). To the southeast, a scatter of roof tiles without ceramics in a well-defined patch of black organic soil may indicate an outbuilding at the edge of the settlement. A dense concentration of burnt soil, pithos wasters, roof-tile wasters, slag and vitrified kiln-wall fragments to the south of the main scatter mark a ceramic production area (L97-101D). About 200m further to the south, on the other side of a small ravine, we recorded a stone sarcophagus and a number of roof tile-line graves. Several similar scatters were recorded along the western slopes of the valley, suggesting a series of similar installations. Sparse finds of Roman red-slipped wares demonstrate occupation in this part of the valley from the fourth to the sixth century.6 Approximately 3km to the north the large and complex site of Karapınar (L99-21) offers an instructive contrast to L97-101. The site consists of an exceptionally dense scatter of ceramics spread over an 8ha terrace overlooking the Karasu river valley (Fig. 5). Concentrations of fine wares near the centre of the locus suggest a domestic residential function (Fig. 6); to the north a concentration of burning and burnt bricks indicates a kiln. At the eastern edge of the terrace traces of burials were documented, including tile graves, glass beads and human bones. The finds suggest a date in the fourth century for this locus, which was most likely the site of an agricultural villa.7 In contrast, L97-101 and the loci around it showed a much greater concentration of industrial debris (roof tiles, pithoi, amphora) without much domestic pottery. If Karapınar is an example of an independent villa complex, the inner Demirci valley may provide an example of an industrial area with limited settlement dependent on a coastal center. The central part of the Demirci valley shows greater chronological diversity and complexity in terms of its material assemblages. The middle zone of the valley includes the villages of Eldevüz and Demirci on the northern ridge. The team systematically surveyed 79 tracts in the central part of the valley, documenting 24 loci that range from the Early Bronze Age to the Ottoman period, with loci of the late Roman period predominating. All along the northern ridge evidence was found for ceramic production dating to the Roman period (Fig. 2): slag, vitrified kiln bricks and wasters of pithoi, tiles, amphorae and domestic wares. In one instance, a large kiln was found in good condition, and had recently been dug out for use as a cistern (Fig. 7). The method of construction of this kiln is generally similar to that 6 7

Doonan and Domżalski forthcoming. Doonan and Domżalski forthcoming.

48

Owen Doonan

Fig. 5: Survey tracts of locus L99-21 (Karapınar; ©SRAP archives).

Fig. 6: Ceramics from L99-21 (Karapınar). Assemblage highly diverse, including fine table wares, glass, a terracotta figurine base, food processing (e.g. mortaria), storage vessels (amphora, pithos) (©SRAP archives).

Settlement and economic intensification in the hinterland of Sinop

49

Fig. 7: Kiln from Eldevüz in the Demirci valley (©SRAP archives). of the examples excavated on the coast at Demirci.8 The round outer walls were constructed using broken flat roof tiles bonded together with mud or clay. The preserved diameter of the kiln from the central part of the Demirci valley is ca 5m, with a central pier supporting the kiln floor at about 1m in height. Instead of the twin central piers seen in the examples from the Demirci coastal site, this kiln appears to have a semi-continuous, circular pier running parallel with the outer wall of the kiln. The large size and peculiar shape of the kiln may suggest a functionally-specific design, perhaps for firing pithoi. Several medium-sized to extensive (>1ha) scatters of large-form storage vessels, including pithoi and amphorae in local fabrics, were recorded in the central part of the valley, especially on the south-facing slopes of the northern ridge.9 Locus 97-090 is an extensive scatter of large-form storage vessels measuring ca 200m × 120m (2.4ha). On the western side of the locus, pithos, tile and amphora wasters and burnt kiln bricks indicate an industrial area. Locus 97-096, 500m to the east, is a smaller (ca 1ha) scatter of pithoi and tile that presents a similar ceramic assemblage. All of these sites conform more to the pattern observed in the inner Demirci valley than to the Karapınar model, suggesting that much of this valley may have been organized under the control of a large estate based at the coast rather than as a series of independent rural villas. More than a dozen large limestone ashlars and architectural elements, all found out of context, were noted in Demirci village. Although well-cut stones of this nature might be transported far from their place of origin to be used for building and decoration (most often to towns like Sinop and Gerze), the unusual concentration in the central valley may 8 9

Kassab Tezgör 2010. Loci 79-080, 97-081, 97-083, 97-090, 97-096. Site areas do not include scatters of eroded or ploughed materials beyond the estimated area of primary deposit.

50

Owen Doonan

point to one or more monumental buildings either here or (more likely) on the nearby coast.10 A large pressing block was also documented in Demirci village, which expands their find spots into the valley and adds to the group of pressing blocks excavated by Kassab Tezgör’ s team at the coastal site.11 The southern flanks of the Demirci valley’s coastal outlet to the shore were surveyed at Keçioğlu and along the Yalı köy ridge.12 The survey of 37 tracts in the coastal zone yielded data demonstrating an extensive continuous area of settlement on the southern flank of the valley mouth from the Hellenistic to the early Byzantine period.13 Like the sites in the Demirci village, Keçioğlu was settled in the fourth to third century BC. The extent of Hellenistic, Roman imperial and late Roman settlements at Keçioğlu is difficult to estimate because all the traces of Hellenistic settlement are mixed together with the material remains of the more extensive Roman settlement phase. Roman settlement was established along the top and along the northern slopes of the Yalı köy ridge. Local blacksand-tempered wares predominate, mixed with roof tiles. A series of small loci along the north-facing slope at Keçioğlu may represent isolated installations or farms on the outskirts of the Demirci coastal site. An extensive (ca 6ha) Hellenistic and Roman settlement (Loci L97-114, L97-115, L97116) was documented on the middle elevations of the northwest-facing slope of the ridge, tucked out of view of the sea. The ceramic assemblage comprises varied forms, including dense scatters of fine wares and large storage vessels. The survey documented the highest proportion of imported fine wares in the valley at locus L97-115. A variety of amphorae,14 mortaria, rouletted fine wares and other ceramic types secure the presumed Hellenistic foundation of the settlement and its continuation into late Roman times (Fig. 8). The Roman assemblage shows fewer imported and fine table wares, but glass and relatively fine, red-brown painted wares are present. The imported and fine wares documented at Keçioğlu reflect the similar ceramic pattern of the Demirci and Eldevüz quadrats, but mark a striking contrast to the simpler Roman assemblages encountered in the inland parts of the valley. In sum, the Roman and early Byzantine periods appear to have witnessed the most intensive occupation and exploitation of the valley, particularly of the inland areas. In Roman times the Demirci agricultural landscape may have resembled that of Hellenistic and Roman North Africa.15 The town and environs of Leptiminus (on the Tunisian coast) offers a particularly instructive parallel to the industrial-agricultural landscape observed in the Demirci valley.16 The urban centre of Leptiminus is about the same size as the coastal 10

11 12 13 14 15 16

Similar concentrations have been noted at Çakıroğlu, Erikli in the Kırkgeçit Çay valley and Belören in the foothills to the west of Gerze (see below). All of these appear to be associated with Roman monumental buildings or towns. Kassab Tezgör 2010a: 73–5. Note, Yvon Garlan and Dominique Kassab Tezgör had previously surveyed the coast from Çiftlik to Yalı plajı in search of evidence of amphora kilns. Garlan and Kassab Tezgör 1996. Note that the Demirci coastal site was not surveyed as part of SRAP’s investigation of the valley. Including an amphora bearing the stamp of Kallisthenes Hestiaiou/Nike in a chariot: Conovici 1998: 33 no. 92, ca 265–250 BC). Fentress 2001; Stone et al. 2011. Stone et al. 2011.

Settlement and economic intensification in the hinterland of Sinop

51

site at Demirci (ca 2km2) and the suburban zone there has yielded 14 Roman sites including six farmsteads and two possible villas with high-status architectural features (marble, mosaics, hypocaust tiles and a column base).17 Strabo notes several times that the coast south of Sinope was largely devoted to olive production, which was limited to a small portion of the Black Sea coast.18 Studies of pollen and data from the excavations at the Demirci coastal site show a spike in olive pollen during the fourth and fifth centuries.19 The wide distribution of Sinopean amphorae throughout the Black Sea may be evidence of this region’s predominant position in the production of the important commodity olive oil. At Demirci some Hellenistic fine wares are associated with loci as far as 3km inland, but Roman fine imported wares are limited primarily to nearcoastal areas. At Keçioğlu (L97-114, L97-115, L97-117, which all appear to be part of a larger settlement) by the coast, 14 fragments of red-slip wares have been identified by Krzysztof Domżalski, including African Red Slip, Pontic Red Slip and Late Roman C wares.20 The assemblage from the large locus L97-101 in the interior includes three fragments of imported wares, including Pontic Red Slip and Late Roman C wares. More than a dozen inland loci (L97-79 to L97-89 at Eldevüz, L101 to L103 at Uzungürgen: area ranges from 0.5–5ha) are characterised by heavy scatters of storage wares and limited numbers of ceramics associated with food consumption. The relative absence of Roman fine wares from on-site and off-site scatters in the interior parts of the Demirci valley suggests that, with the exception of L98-21 (Karapınar), the inland loci were farming installations with few residential components and no evidence of elite presence. These farms may have provided agricultural products, particularly olives, which could be packed into amphorae produced at the coastal site and shipped out for consumption in Sinope and beyond. The residential areas in the Demirci valley that show evidence of Roman-period highstatus residences are concentrated on the coast. There is some evidence (well-cut limestone architectural elements in the village of Demirci, evidence of a stone masonry structure at Yalı plajı and a possible submerged structure to the northeast of the amphora production site) to suggest luxury architecture on the coast at Demirci,21 while agricultural installations further inland generally show fewer imports and less substantial buildings. In this case it is reasonable to suggest that the agricultural and ceramic industries were owned or controlled largely by a coastal suburban elite that traded through the port of Sinope. The villa at Karapınar (L99-21) appears to represent something different, perhaps a more or less independent agricultural villa with a significant residential component. Note that this overlooked the Karasu river valley on the interior of the promontory and would not necessarily have much, if any, connection to coastal land holdings. The possibility that a transition in landholding practices from a Roman system of rural free holdings to an Early Byzantine villa-dominated rural production system has recently been presented at the ‘Climate and Society’ conference organized by J. Haldon at Princeton and will be published elsewhere (Doonan 2015). 17 18 19 20 21

Stone et al. 2011: 307. Doonan 2003. Emery-Barbier 2010. Doonan and Domżalski forthcoming. Another example of this kind of structure would be the villa at Kiraztepe (L96-55) about 4km south of Sinop town (Doonan 2004: 101–3).

52

Owen Doonan

2.2 Kırkgeçit çayı valley The Kırkgeçit-Kabalı çayı drainage is the most extensive river system on the Sinop promontory. Both rivers rise in the rugged Pontic coastal mountain range that frustrates communications between the coastal plains and foothills of the Sinop promontory and the important east-west transportation route known as the north Anatolian trunk road that follows the north Anatolian rift valley approximately 100km inland from the coast. The north Anatolian route has been active since at least the third millennium BC when rich deposits of copper and silver in the Pontic mountains were first exploited for export via Assyrian trade networks. Ongoing excavations at several large and monumental sites along this valley testify to the importance of the route of the north Anatolian trunk road from the second millennium onward.22 The ridge overlooking the Kırkgeçit çayı from the north was the most important route connecting the coastal plains of Sinop to the important inland town of Boyabat in Ottoman times,23 and a succession of scholars has identified a cluster of Roman milestones near the modern village of Erikli that suggest a major road since at least the first century.24 The milestones from the Erikli area indicate a distance from Sinope of 36 or 37 Roman miles,25 consistent with the distance a traveller might encounter by following the coast to the outlet of the Kırkgeçit çayı and then the river up onto the ridge. The earliest milestone dates to Flavian times,26 while the majority of the group suggests substantial repair or expansion of the road in the later third and fourth centuries.27 The SRAP team carried out intensive survey in 453 tracts (2.3km2) organised into 23 quadrats along a 10km portion of the Kırkgeçit çayı corridor that connects the coastal town at Çakıroğlu to the highland passes in the Tingiroğlu area. The results can be organised into four zones: (1) coastal; (2) the foothills overlooking the Kırkgeçit çayı; (3) the foothills and highlands overlooking the Sarımsaklı çayı; and (4) the highland pass of the Kırkgeçit çayı (see Fig. 8). In total, 15 Hellenistic/early Roman, 43 late Roman/early Byzantine and four middle to late Byzantine archaeological loci were recorded by the survey. A large coastal settled area was documented in the modern village of Çakıroğlu (Hellenistic to Ottoman occupation).28 In order to determine the urban character of the site, the team carried out geophysical surveys at Çakıroğlu using Ground Penetrating Radar (GPR). Areas associated with surface scatters demonstrate dense concentrations of stone and brick structures consistent with an urban pattern.29 The overall record in the valley above the coastal zone suggests that a pre-existing network of Iron Age to Hellenistic settlements 22 23 24 25 26 27

28 29

See, in particular, the sites of Oymaağaç (Czichon 2009; Czichon et al. 2011) and Pompeiopolis (Summerer et al. 2011). William J. Hamilton recorded detailed notes as he passed along this road in 1836 and stopped near the highland town of Tingiroğlu and the coastal town of Çakıroğlu: Hamilton 1842. Robinson 1905; French 2013: 55–66. French 1988; 2013: milestone 21A (Tıngır 1), pp. 59–60. Robinson 1905: 329; French 2013: milestone 19 (Göllü 1), 55–56. French 2013: the earliest datable milestone from the reign of Vespasian (78), a single example from the reign of Caracalla (213/4), multiple examples from reigns of Diocletian (283 and 293), Constantinus I (324) and Constantius II (340). Doonan et al. 2015. Doonan et al. 2015.

Settlement and economic intensification in the hinterland of Sinop

53

Fig. 8: Map of Kırkgeçit çayı survey area (map after M. Conrad, ©SRAP archives). flanking the river channel established contact with the port of Sinope in the fourth century BC, that a formal Roman road may have been established as early as the first century and that the road was significantly expanded in the fourth century after which settlement in the valley underwent a significant expansion that collapsed around the seventh century.30 Seven survey quadrats in the coastal zone (Çakıroğlu, Altınoğlu, Lala, Çırnık, Sarıyer and Kelemperoğlu) encompassed a total area of 7.2km2, within which the team intensively surveyed 119 tracts (73.4ha). Twenty eight archaeological loci were identified; the majority were associated with what was, most likely, a large late Roman town in the village of Çakıroğlu (Fig. 9a). The valley floor is covered with alluvial deposits and is thus inaccessible for systematic intensive survey. However, evidence thrown up by repeated deep ploughing supplemented by observations made of a nearby building trench in locus L11-28 reveal that at least some of the southern bank of the river was occupied in late Roman times. Ten large and several smaller fragments of fine limestone architectural elements (five capitals, three columns, two bases and numerous smaller fragments) were recovered in the village, and fishermen report finds of mosaics approximately 50m off shore. Limited GPR surveys near the coast and several hundred metres inland suggest that brick and stone masonry structures covered much of the ridge that defines the southern flank of the valley; we propose 30

Doonan et al. 2015.

54

Owen Doonan

Fig. 9a: Survey areas in Çakıroğlu quadrat. Surveyed tracts are highlighted, areas showing heavy ceramic concentrations are outlined, and point features marked by triangular icons. further geophysical research in this area to demonstrate the urban form and extent of the late Roman town.31 Evidence from the northern flanks of the valley suggests less intensive settlement during late Roman times and also offers important clues about the nature of the transportation infrastructure of the area. A dense concentration of roof tiles was recorded on the ridge overlooking the northern bank of the outlet of the Kırkgeçit çayı (L11-11) together with the remains of an Ottoman mill; this is the place where in 1836 the European traveller William J. Hamilton descended the Lala ridge to the northern bank of the river.32 Previous investigations of the site by the Sinop Museum yielded evidence of a hypocaust system.33 A bath building would not be surprising in this location: on the coastal road at the edge of a large town. Approximately 1km inland a pre-Ottoman stone caravanserai was documented. This seems to be the location where Hamilton crossed the river en route from Bafra to Sinop, and recent analyses suggest evidence of an inland Ottoman road that

31 32 33

Doonan et al. 2014; Doonan et al. 2015. Hamilton 1842: 314. http://www.kulturportali.gov.tr/turkiye/sinop/kulturenvanteri/roma-hamami-kalintisi (accessed January 2015).

Settlement and economic intensification in the hinterland of Sinop

55

Fig. 9b: Survey areas in Erikli (South) quadrat. Surveyed tracts are highlighted, areas showing heavy ceramic concentrations are outlined, and point features marked by triangular icons. Note: icon on left marks location of the pressing block mentioned in the text (maps after M. Conrad, ©SRAP archives). crossed the river at this point.34 It appears that from Roman to late Ottoman times the intersection of coastal and transmontane roads at Çakıroğlu made it a key terrestrial transportation hub on the promontory. From the crossing near the caravanserai to the middle elevations of Yakadibi and Erikli (ca 500m a.s.l.) there appear to be several extensive settled areas interspersed with gaps. Masonry structures, possibly small churches, were identified in thickets adjacent to systematic survey tracts at Kabalı (L11-22) and Kelemperoğlu (L11-17). Locus L11-17 featured a banded-brick masonry structure approximately 20m in length and oriented east-west set adjacent to a 1ha late Roman scatter of ceramics (L11-15). On the opposite bank, loci L11-24 and L11-25 comprise an extensive (ca 8ha) scatter of late Roman ceramics and tile just above the flood zone of the river. The team surveyed ten quadrats (6.25km2) covering 211 systematic tracts (98.5ha) overlooking the river channel from both the northern and southern banks. Thirty-three archaeological loci were recorded in this part of the valley including Iron Age, Hellenistic and late Roman occupations. The settlement pattern in this part of the survey area is distinct from that of the coastal zone, with larger apparently empty spaces between more defined 34

Doonan et al. 2015.

56

Owen Doonan

settlements. The foothills show a significant expansion in settlement during late Roman/ early Byzantine times, particularly in the area of the modern village of Erikli (Fig. 9b). In the early 20th century Robinson recorded a cluster of Roman milestones from Erikli mosque dating mostly to the late third and fourth centuries.35 It appears that following the consolidation and expansion of the road system there was significant population growth. In addition to the series of settlements on both sides of the Kırkgeçit çayı, we recorded Alasökü han (L12-043) on the northern side of the Kabalı river. This is a small, stone-built Ottoman han that is likely to have served a crossroads at the intersection of a north-south crossing of the Kabalı çayı. In this case, there may well have been two main Ottoman routes from Sinop to Boyabat – one inland and one coastal (that then followed the Kırkgeçit çayı upstream) – that came together in the area of the modern village of Erikli. Travelling in July when the Kırkgeçit çayı was easily crossed, Hamilton clearly followed the coastal route that turned upstream at Çakıroğlu. It may have been that in other seasons an inland overland route that crossed the Kabalı river near Alasökü han (L12-043) was preferable. A concentration of Roman loci near Erikli (Erikli south: three loci within 34 tracts (Fig. 9b); Erikli north: six loci within 29 tracts) coincides with finds of two limestone columns that were noted in the garden of the village mosque and a pressing block similar to numerous other examples recorded in the Demirci valley. The ceramic scatters in this area are not as extensive as those in the lower valley, although this area may well have functioned as some kind of village centre. The team recorded on-site and off-site data from seven quadrats in 147 tracts around the villages of Karlı and Sorkum, and identified 11 loci, of which only three are large enough to be considered settlements.36 In the area of Karlı village four quadrats within 115 tracts (35ha) were surveyed intensively during the 2010 field season. The landscape and environment of this area is about the same as that of the foothills of the Kırkgeçit çayı valley with a similar distance and accessibility to the coast; however, the evidence for settlement is much more limited in the Sarımsaklı foothills. Although an increase in settlement density is coincidental with the major demographic and industrial expansion noted in other parts of the Sinop promontory, the Sarımsaklı foothills are much less densely settled than the Demirci or Kırkgeçit çayı valleys. 2.3 Other survey areas: Boztepe, the outer Karasu valley and the west coast The extreme pattern of late Roman/early Byzantine settlement expansion and collapse observed in the Demirci and Kırkgeçit çayı valleys is not uniform across the Sinop promontory. In other areas where our team has carried out systematic surveys (notably Boztepe, the northern outlet of the Karasu valley and the west coast, including Saraycık and Sarıkum) the predominant pattern suggests foundation of small coastal settlements (generally of about 1ha) in the fourth to second century BC and continuity at those settlements through the late Roman/early Byzantine period.37 Isolated farmsteads do appear here and there (in Sarıkum, for example) but there does not seem to be anything like the massive expansion of 35

Robinson 1905: 327–9; for the most comprehensive and up-to-date discussion of this group of milestones, see French 2013: 34–45, 54–66. 36 Doonan and Bauer 2005; Doonan et al. 2015. 37 Doonan et al. 2014.

Settlement and economic intensification in the hinterland of Sinop

57

fourth-century settlement observed along the west coast. It is difficult to establish a pattern in the northwestern part of the promontory because of the heavy forestation; the foundation of two rather substantial settlements (Sarsı L98-10 and Gerna dere L96-46) around the sixth century, however, may point to a brief and comparably late expansion in this area. Gerna dere is an isolated small bay near the dangerous northwestern tip of the promontory where our team documented a church, a possible monastic complex and coastal watch fires in 1996.38 The find of an anonymous follis at the site suggests continued occupation into the middle Byzantine period, perhaps to the twelfth or even thirteenth century. The settlement of Sarsı lies adjacent to the road built in the late third century, according to the milestones from the site of Heladı further west.39 Comparatively numerous finds of imported red-slip wares suggest a date of the fifth to sixth century.40

3. Discussion The contrasting patterns discussed above suggest divergent histories for the different parts of the Sinop promontory. The argument that the east coast was given over to olive production rests on three main bodies of evidence: the survey and environmental data (discussed above), textual references and the amphorae produced in Sinop and exported around the Black Sea and beyond. The tract of land belonging to Sinope and all the mountainous country as far as Bithynia, situated above the sea coast, which has been described, furnishes timber of excellent quality for shipbuilding, and is easily conveyed away. The territory of Sinope produces the maple, and the mountain nut tree, from which wood for tables is cut. The whole country is planted with olive, and cultivation begins a little above the sea coast.41

The primary sources are convincing in their observation that the east coast of the Sinop promontory supported an otherwise very rare olive industry in the Black Sea from late Hellenistic times onward. After the foundation of the great metropolis at Constantinople a consistent demand for olive oil from Black Sea sources may have emerged. Kassab Tezgör has established the extensive distribution of Sinopean amphorae to all parts of the Black Sea and across a significant swath of the Levant, particularly those made of the fabrics she identifies as groups C and D (dating to the fourth to early sixth century and mid sixth century respectively).42 Finds of Kassab Tezgör’s group B (second to fourth centuries) are largely confined to the Black Sea. Thus the distribution suggests a major expansion of Sinope’s market in the fourth or fifth century. Exports to the northern Black Sea seem to have peaked with Group C, while exports to the Bulgarian, Anatolian and Georgian coasts continued with Group D.43 I have argued elsewhere44 that the remarkable collapse in rural settlement in the Sinop hinterland may have been a consequence of the expansion in olive production being 38 39 40 41 42 43 44

Doonan and Smart 2001. French 2013: 34–5. Doonan et al. 2015. Strabo 12.3.12; Falconer (trans.) 1903. Kassab Tezgör 2010b. Kassab Tezgör 2010b: pl. 4. Doonan in Haldon et al. 2014.

58

Owen Doonan

followed by the deterioration of the favourable climatic conditions associated with the Beyşehir Occupation Phase (BOP), a distinctive anthropogenic phenomenon associated in Anatolia with the domination of Mediterranean polyculture. The BOP appears to come to an end in western Asia and the eastern Mediterranean around the mid to late first millennium. The end of the BOP may be correlated with a general deterioration of climate from the ‘Roman warm period’ to a colder and wetter regime.45 Certainly in Sinope, always a marginal ecological microclimate for olive production, even a slight deterioration in climate would have had a significant effect on the olive industry. If, as seems to have been the case along the eastern coast of the Sinop promontory, the hinterland economy was almost exclusively dependent on an intensified olive industry, the collapse of the olive industry would have brought about an equally dramatic collapse in rural settlement. Other parts of the promontory that never underwent this degree of economic intensification experienced neither the explosive growth nor the subsequent decline seen in the Demirci and Kırkgeçit çayı valleys. In many ways, these quieter parts of the Sinop promontory continued the millennia-old isolation that has characterised the region of the Black Sea mountains right into the early years of the Turkish Republic. Owen Doonan California State University Northridge Program in Art History, Art Department 18111 Nordhoff Street, Northridge CA 91330–8300, USA [email protected]

Bibliography Conovici, N. 1998. Histria 8. Les timbres amphoriques 2. Sinope. Paris. Czichon, R. M., J. Klinger, P. Breuer, J. Eerbeek, S. Fox, E. Marinova-Wolff, H. Marquardt, H. von der Osten-Woldenburg, S. Reichmuth, S. Riehl and T. Johannsen 2011. ‘Archäologische Forschungen am Oymaagac Höyük/Nerik(?) in den Jahren 2007–2010’, Mitteilungen der Deutschen Orient-Gesellschaft zu Berlin 143, 169–250. Doonan, O. 2003. ‘Sinope’, in D. Grammenos and E. K. Petropoulos (eds), Ancient Greek Colonies of the Black Sea (Archaeological Institute of Northern Greece 4). Thessalonica, 1379–402. Doonan, O. 2004a. ‘Sampling Sinop: Archaeological Survey in a low visibility environment’, in L. Wandsnider and E. Athanassopoulou (eds), Recent Developments in Mediterranean Survey Archaeology. Philadelphia, 37–54. Doonan, O. 2004b. Sinop Landscapes: Exploring Connection in the Hinterland of a Black Sea Port. Philadelphia. Doonan, O. 2015. ‘Too much of a good thing? Economic success, sustainability and resilience in Pontus, 1st–8th c. CE’ in Climate and Society in Byzantine and Ottoman Anatolia, 300–1900 CE. Towards understanding the impact of climate on complex societies of the pre-industrial era. Princeton University, May 1–3 2015. Doonan, O. forthcoming. ‘Evidence for Roman roads in the territory of ancient Sinope’ (manuscript in preparation). Doonan, O. and A. Bauer 2005. ‘Sinop Province archaeological project: Report on the 2003 field season’, Araştırma Sonuçları Toplantısı 22, 275–84. 45

Haldon et al. 2014. In a subsequent paper I argue that concentration of landholding associated with the rise of a villa-based mode of production may have diminished the resilience of rural settlement and production, contributing to an apparent collapse in rural populations in the seventh and eighth centuries CE (Doonan 2015).

Settlement and economic intensification in the hinterland of Sinop

59

Doonan, O. and K. Domżalski forthcoming. ‘Pontic Red Slip distribution in the Sinop Hinterland’ (manuscript in preparation). Doonan, O. and D. Smart 2001. ‘Gerna dere: Roman and Byzantine settlement along the coast of Sinop’, Talanta 32–33, 17–24. Doonan, O., A. Gantos, F. Hiebert, A. Yaycioğlu and M. Besonen 2001. ‘Systematic Survey of the Karasu River Valley (Sinop province), 1998–99’, TÜBA-AR 4 (Journal of Turkish Academy of Sciences), 113–35. Doonan, O. et al. 2014. ‘Sinop Bolgesel Arkeoloji Projesi 2012 Saha Calismalari’, Araştırma Sonuçları Toplantısı 31, 146–64. Doonan, O. et al. 2015. ‘Sinop Regional Archaeological Project: Report on the 2010–2012 Field Seasons’ in S. Steadman and G. McMahon (eds), The Archaeology of Anatolia:  Current Work (Cambridge Scholars Press). Cambridge, 302–31. Emery-Barbier, A. 2010. ‘Vegetation actuelle et passée de la region de Sinope’, in D. Kassab Tezgör (ed.). Les fouilles et le matériel amphorique de l’atelier de Demirci près de Sinope (Varia Anatolica 22). Paris, 27–40. Fentress, E. 2001. ‘Villas, wine and kilns: the landscape of Jerba in the late Hellenistic period’, Journal of Roman Archaeology 14, 249–68. French, D. 1988. Roman Roads and Milestones of Asia Minor (BAR International Series 392). Oxford. French, D. 2013. Roman Roads and Milestones of Asia Minor 3.4. Pontus and Bithynia. Ankara. Garlan, Y. and D. Kassab Tezgör 1996. ‘Prospection d’ateliers d’amphores et de ceramiques de Sinope’, Anatolia Antiqua 1, 325–34. Haldon, J., N. Roberts, A. Izdebski, D. Fleitmann, M. McCormick, M. Cassis, O. Doonan, W. Eastwood, H. Elton, S. Ladstätter, S. Manning, J. Newhard, K. Nicoll, I. Telelis and E. Xoplaki 2014. ‘The Climate and Environment of Byzantine Anatolia: Integrating Science, History and Archaeology’, Journal of Interdisciplinary History 45.2, 113–61. Hamilton, W. 1842. Researches in Asia Minor, Pontus and Armenia: with some account of their antiquities and geology. London. Kassab Tezgör, D. (ed.) 2010a. Les fouilles et le matériel de l’atelier amphorique de Demirci près de Sinope. Istanbul. Kassab Tezgör, D. 2010b. ‘Le reseau commercial des amphores Sinopéennes entre les IIe–IIIe s. et le VIe s. de notre ère’, in D. Kassab Tezgör and N. Inaishvili (eds), PATABS I: Production and Trade of Amphorae in the Black Sea. Actes de la Table Ronde international de Batoumi et Trabzon, 27– 29 Avril 2006. Istanbul, 167–73. Kuzucuoğlu, C. 2010. ‘Demirci et ses alentours: Dynamique géomorphologique et relations avec l’activité de l’atelier de Demirci’, in D. Kassab Tezgör (ed.), Les fouilles et le matériel amphorique de l’atelier de Demirci près de Sinope (Varia Anatolica 22) Paris, 15–26. Robinson, D. 1905. ‘Greek and Latin Inscriptions from Sinope and Environs’, American Journal of Archaeology 9, 294–333. Stone, D., D. Mattingly and N. Ben Lazreg 2011. Leptiminus (Lamta). Report no. 3: The Field Survey (Journal of Roman Archaeology Supplementary Series 87). Portsmouth, RI. Summerer, L. (ed.) 2011. Pompeiopolis. Eine Zwischenbilanz nach fünf Kampagnen (2006–2010) (ZAKS Reihe 21). Langeweißbach.

Peri Johnson

How did the landscape of Pompeiopolis become Roman?

1. Introduction After its refoundation in 5 BC, the city of Pompeiopolis in Paphlagonia gradually acquired the institutions and monuments of a Roman city.1 By the time of an inscription with a Trajanic list of ephebes, Christian Marek concludes, ‘the epigraphic sources of the city argue for well-established Roman families’.2 Ephebes were typically the youth of landholding families. The established, Roman character of the city in the ephebic list and other epigraphic and historical sources raises the question of how the introduction of Roman urban institutions altered rural landscapes. How Roman were the landscapes lived in and held by these families? Pompeiopolis was founded by the Roman commander Pompey in 64 BC on a ridge along the northern bank of the Gökırmak (Amnias), the northernmost western tributary of the Kızılırmak (Halys) south of the Küre (Pontic) mountain range (Fig. 1). After a period of control by Paphlagonian client princes, Pompeiopolis was refounded in 5 BC and later became the metropolis of Paphlagonia where the provincial senate convened.3 The territory administered by Pompeiopolis encompassed the upland and lowland valleys drained by the Gökırmak and the upper Devrekani river.4 This contribution discusses continuity in the landscapes of Pompeiopolis between the Hellenistic and Roman periods, later changes in Roman settlements, and the relationship of these settlements to burial tumuli and rock-cut tombs in their vicinity. Through the placing of prominent burial monuments, the inhabitants claimed the land they held as their own: both the land they lived in and that over which they held rights of use for cultivation, pastoralism, etc.5 Through this discussion, the paper proposes some tentative conclusions about Roman landholding and the processes of becoming Roman in the landscape of Pompeiopolis.

1 2 3 4 5

The historical and epigraphic interpretations are drawn from the work of Christian Marek (1993: 63–73), epigraphist of the Pompeiopolis in Paphlagonia Project. Marek 1993: 65: ‘Zusammengenommen sprechen die epigraphischen Zeugnisse der Stadt für alteingesessene römische Familien’. Marek 1993: 73–82. Marek 1993: 65–8. For a summary of the questions raised by the concepts of landowning and landholding, see Mientjes and Pluciennik 2013.

62

Peri Johnson

How did the landscape of Pompeiopolis become Roman?

Fig. 2: Map of the environs of Pompeiopolis (author’s map).

Fig. 1: Map of central northern Turkey (author’s map).

63

64

Peri Johnson

Beginning in 2008, two years after the excavations in the city, the Pompeiopolis in Paphlagonia Project6 initiated a survey in the environs of the city, an area approximately coextensive with the district of Taşköprü, the contemporary town on the facing bank of the Gökırmak (Fig. 2). This article draws on the published reports of earlier surveys, the analysis and groundtruthing of images granted by the GeoEye Foundation to the Pompeiopolis in Paphlagonia Project and the preliminary results of the project’s own survey.7 Consequently, the paper does not cover the full extent of Pompeiopolis’s territory but is limited to the central Gökırmak valley covered by the project’s survey. Because the survey has yet to have sufficient members to expand beyond extensive survey techniques and remote sensing, the evidence for the extent and development of Hellenistic (the third quarter of the fourth century to the first century BC) and Roman (the first to the seventh century AD) settlements that can only be identified through intensive fieldwalking is not yet available. The results of earlier projects present problems of their own. Most are incompletely published and lack chronological specificity. For example, the Kastamonu Museum director Ahmet Gökoğlu surveyed in the 1940s what are often undatable tumuli and simple rock-cut tombs, and in 1977 Pauline Donceel-Voûte used the chronological category of ‘Hellenistico-Roman’ in the only classically-focused survey of the Gökırmak valley before the Pompeiopolis survey.8 The Hellenistic and Roman categories of the Kastamonu Project in the late 1990s are exemplary in the preliminary reports of a project that was admittedly focused on Chalcolithic through Iron Age settlements, but they do not reflect the full complexity and specificity that survey assemblages can produce.9 Not until the ceramic specialists of the Pompeiopolis excavation had begun to analyse the locally-made fine and coarse wares have we been able to date with precision early and later Roman settlements.10 Despite Pompeiopolis’ foundation following Roman military conquest, evidence for any Roman cadastral land division for the settlement of veterans has yet to emerge for the landscape around Pompeiopolis. In contrast to any implemented cadastral division, the prosperity and density of the Roman landscape appears to have been a development of the previously existing settlement preferences – predominantly guided by the presence of springs and narrow bands of fertile soil in alluvial terraces along the channels of the Gökırmak and its tributaries. Along the alluvial terraces, extensive and prosperous Roman settlements surrounded by cemeteries of cist graves, tumuli and rock-cut tombs have been

6 7

8 9

10

I owe thanks to the Pompeiopolis in Paphlagonia Project for including me in the team and providing me with access to the excavation data as well as to the unpublished results of pottery studies. The most systematic earlier project is the Kastamonu Project, which published seven preliminary reports (Marro et al. 1996; 1998; Kuzucuoğlu et al. 1997; Özdoğan et al. 1997; 1998; 1999; 2000). Publications of the Pompeiopolis survey are Barat et al. 2009; Johnson 2011. Gökoğlu 1952; Donceel-Voûte 1979. I mention the inadequacies of the published survey assemblages in addition to our own project’s deficiencies here not to critique earlier researchers but to qualify my own interpretations and to emphasise their tentative character. Domżalski 2011; Zhuravlev 2011. In 2012 Andrei OpaiŢ joined the excavations to study the coarse wares.

How did the landscape of Pompeiopolis become Roman?

65

identified in the environs of Hellenistic and earlier settlements. In these areas of multiperiod settlement, associating the larger tumuli or undecorated tombs lacking artefacts with the Roman period and particularly Roman landholders is very tenuous without excavation. At the same time, however, evidence for tumuli dating to the Iron Age (ca 1200–330 BC) across the entire Gökırmak valley is limited to the bulldozed Gödel Kayabaşı tumulus and the Honsalar tumulus surveyed by the Kastamonu Project, south of Kastamonu and in the Daday valley to the northwest of Kastamonu respectively (Fig. 1).11 The four large tumuli on a ridge at Alpagut between two southern tributaries of the Gökırmak east of Kastamonu are only dated to the Iron Age on the assumption that larger tumuli are Iron Age in date.12 The numerous unsurveyed tumuli and settlement traces visible in satellite images of the ridge and valley shoulders at Alpagut suggest caution when interpreting the four surveyed tumuli. Two additional dated tumuli from the Daday valley have been published: Esenler is dated to the Hellenistic period and Höyüktepe to the Roman period.13 Among the tumuli, Höyüktepe stands out as a single prominent Roman tumulus in a detached agricultural valley nestled in the mountains between Kastamonu and Daday. Excavations exposed an ashlar barrel-vaulted chamber built around a sarcophagus decorated with bucrania and garlands. The tumulus is a compelling example of the relationship between tumuli and landholding, here of the surrounding valley where Hellenistic and earlier settlements are absent. Additional Hellenistic and Roman tumuli excavated by Giulio Jacopi in 1935 and 1936 to the west and south of Pompeiopolis are discussed below. Despite the inconclusive evidence, based on the published and well-dated tumuli, and the multitude of tumuli near Pompeiopolis, I propose that in the wider environs of the city the majority of tumuli should be dated to the Roman period, a lesser proportion to the Hellenistic period and a few to the Iron Age. After intensive survey many tumuli, particularly the prominent ones, may emerge to be related to Roman landholding. More conclusive evidence for Roman landholding comes from the single probable Roman villa in the Karapürçek-Yeniköy area discussed below. Other categories of Roman settlements such as villages and hamlets have also been surveyed but both the settlements and tumuli will be introduced below not by category but as places in the landscape. Before turning our attention to these places it is necessary to introduce briefly the geomorphology and discuss how it affects settlement and mortuary landscapes.

11

Gödel Kayabaşı: Marro et al. 1996: 275 s.v. C31/6, 284, 286, 290; Özdoğan et al. 1997: 309–10, 321 fig. 6.3–5, 7–8; Özdoğan et al. 1999: 227; Honsalar: Marro et al. 1998: 317, 319 s.v. B30/1; Özdoğan et al. 1999: 220. 12 Alpagut: Gökoğlu 1952: 49; Marro et al. 1996: 273, 275 s.v. C32/4–7, 277 fig. 1. 13 Esenler: Gökoğlu 1952: 48, 55–6 s.v. Tonoz; Marro et al. 1998: 317, 319 s.v. B30/2, 320, fig. 1, 322; Özdoğan et al. 1999: 221, 236, fig. 14; Höyüktepe: Akok 1948: 848, pls 179–80 s.v. Kaypı; Gökoğlu 1952: 48, 52–5 s.v. Kaypı; Marro et al. 1998: 319 s.v. C31/21, 322; Özdoğan et al. 1999: 224. The Kastamonu Museum conducted salvage excavations of a looted Roman tumulus in the mountainous northwest borderlands of Pompeiopolis’ territory near Selmanlı Mahallesi (Ağlı, Kastamonu). An interior wall of the built burial chamber of the tumulus was decorated with a painted relief panel (Bal 2014).

66

Peri Johnson

Fig. 3: Map of features and places in the environs of Pompeiopolis mentioned in the geomorphological description (author’s map).

How did the landscape of Pompeiopolis become Roman?

67

2. Geomorphology of the environs of Pompeiopolis The Gökırmak valley was formed by simultaneous uplift associated with the North Anatolian Fault Zone and fluvial incision (Fig. 3).14 The terraces along the length of the valley, therefore, are caused by the erosion of bedrock (strath terraces) and are often blanketed with variously consolidated loose rock that easily erodes and is not particularly fertile. Not surprisingly, the terraces with bedrock or loose rock on the surface are the preferred location of settlements and tumuli. Although research, particularly the analysis of erosion and alluvial deposition through coring or trenching, has not been conducted to quantify geomorphological changes, the preference for settlement locations on strath terraces adjacent to flood plains filled with alluvial terraces in modern times does suggest that the alluvial terraces are very roughly similar to those of the past.15 As proxies for coring, the Pompeiopolis project has not yet encountered buried features that could be surveyed in the flood plain to quantify deposition. Pompeiopolis itself lies on a limestone and siltstone terrace adjacent to the flood plains of the Gökırmak and the Oluklar and Kuru streams. Gullies that cut through formerly stable land surfaces at the Roman settlement of Dolaman Arkası along the Aşağıçayı near its confluence with the Gökırmak, however, indicate high levels of erosion since the Roman period. It is assumed, therefore, that alluvial deposition and alteration of the alluvial terraces have occurred in the Gökırmak valley since the Roman period, but have not led to degradation that hinders agriculture. In addition to the unknown quantity of alluvial deposition, a further ambiguity lies in the character of the tributaries of the Gökırmak in the environs of Pompeiopolis. Many of the tributaries have braided gravel beds that reflect the slope of mountain runoff and high sediment loads such as the lower Aşağıçayı’s burden of the sediment from Mount Elekdağ’s eroded surfaces. Although the slopes have not changed significantly since the Roman period, it is likely that erosion in the mountains caused by expanding pastoral, logging and mining activities has significantly altered the sediment load of mountain streams, the geomorphology of their beds and how amenable the deposited alluvium is for agriculture. As important as erosion and deposition for the geomorphology of the Gökırmak valley are the tectonics. Active uplift is occurring in the Aygır Tepesi ridge – Zımbıllı Tepe where Pompeiopolis was founded lies at the southern foot of this ridge – as well as the area of Samanlıören. More important is, however, the Ekinveren Fault, an active thrust fault that walls off the Küre mountain range from the valley terraces in the north and determines the course of the Gökırmak. The larger settlements are located south of the fault where the bedrock is sedimentary and the slopes are gentler, whereas mining and logging settlements are located north of the fault line where the bedrock is generally metamorphic and the valleys are deeply incised. From the southern bank of the Gökırmak the terraces gradually rise above low-lying and densely-settled valleys between the Elekdağ in the east and the Çit valley in the west where 14

Although the Gökırmak valley has been the subject of many geological studies, the recent project led by Cengz YIldIrIm on strath terraces is the most relevant to this article (YIldIrIm et al. 2013). 15 The Kastamonu Project collected cores in drained marshes along the Gökırmak river in 1998 but an analysis has not been published (Özdoğan et al. 2000: 45).

68

Peri Johnson

the foothills of the Ilgaz (Olgassys) mountain range extend as far as the confluence of the Karadere with the Gökırmak. South of Tepedelik, the lowland terraces transition to mountain valleys interspersed among forested peaks up to the summits of the Ilgaz mountain range. The mountain valleys were densely settled in the Hellenistic and Roman periods by smaller settlements. The Pompeiopolis and Kastamonu projects have surveyed settlements on the northeastern slope of Mount Saraycık and to the east along the Akkaya river that drains the northern slopes of the Ilgaz mountain range in a long clockwise loop (Fig. 1). The boundary of both projects is an escarpment caused by the principal branch of the North Anatolian Fault Zone, of which the Ekinveren is a secondary branch, that falls from the Ilgaz summits to the Kızılırmak and the Devrez river below. Although passage over the escarpment is arduous, the extensive surveys demonstrate that the upper mountain slopes were settled and the mountain valleys densely so. In conclusion, although Pompeiopolis lies on a limestone terrace overlooking a narrow band of alluvial terraces, as most settlements in the Gökırmak valley lay, other settlements of the Roman period expand across all areas of the river’s broad and gentle central valley and the detached mountain valleys and slopes.

3. Pompeiopolis’ landscapes of the dead The Saraycık Dağı settlement and others such as those around Ömerli at the western foot of Mount Saraycık tend to have cemeteries of prominent tumuli, rock-cut tombs and cist graves (Fig. 2).16 How we interpret these mortuary landscapes and what they suggest about Roman landholding depends on the settlement history of each place and comparison to the cemeteries around Pompeiopolis. For example, Ömerli’s cemeteries do not appear to be contemporaneous, yet they replicate the assemblage at Pompeiopolis on a smaller scale. At Pompeiopolis, tumuli were given prominent locations on the shoulder and slope of the terrace facing the city on the southern bank of the Gökırmak. The nearest probable Hellenistic settlements to Pompeiopolis are Aşıra on a terrace above the northern bank of the Gökırmak, similar to Pompeiopolis but 3km downstream, and Aygır Kalesi, 4km upstream from Pompeiopolis along the Oluklar stream.17 The absence of a Hellenistic settlement in the immediate environs of Pompeiopolis, particularly around the cemeteries on the southern bank, argues for all the tumuli to be contemporaneous with the imperial and late Roman city. Less doubt emerges with respect to the contemporaneity of the rock-cut tombs within the city. To the north of the urban centre and residential areas of Pompeiopolis on Zımbıllı Tepe, Ferhat Çevik and Max Ritter surveyed a rock-cut cemetery of the city on the Sivritepe ridge (Fig. 4).18 The loosely cemented limestone and siltstone bedrock sequence of the ridge is poor for carving and the tombs are eroded and disturbed. Çevik and Ritter were able, however, to document 96 rock-cut tombs ranging from cists with 16

Saraycık Dağı: Johnson 2011: 198–9; Ömerli: Özdoğan et al. 1998: 68, 71–3 (Namaztepe tumuli, the Mercimekli Sırtı Hellenistic cemetery and the Sülüklütepe Roman and Byzantine cemeteries). 17 Only a Hellenistic incurved rim bowl and common grey wares from late in the sequence attest to the Hellenistic period at Aşıra (see further below). The springs near the Paphlagonian columnar rock-cut tomb at Aygır Kalesi are associated with the ceramic surface scatter of a settlement that we have noted but not documented (see also Gökoğlu 1952: 72–3). 18 Çevik and Ritter 2013.

How did the landscape of Pompeiopolis become Roman?

69

Fig. 4: Map of the cemeteries near Pompeiopolis (author’s map). capping stones (chamosoria), arched recesses (arcosolia) and chamber tombs with arcosolia. The larger tombs are carved on the eastern slope of the ridge that rises from the bank of the Gökırmak and the road to Sinope that ran northward along the bank. Returning to the city and crossing the river, other less systematically documented cemeteries rest on the terrace rising from the southern bank of the Gökırmak.19 Most burials are located on a schist outcrop in the terrace that faces Pompeiopolis. Rock-cut cist burials appear to be concentrated on the northwestern ridge of the outcrop, but are only visible where the burials were excavated into the bedrock and have been looted. Looted features located in areas of overlying loosely consolidated rock are difficult to interpret; disturbed areas such as that on the northern shoulder of the ridge may be cist burials excavated into, or tumuli built of, the rock. Through the middle of the cemetery a hollow way traces the course of a 19

The Roman bridge is thought to have been located in the same place as the Ottoman bridge of Taşköprü (‘stone bridge’), although evidence has not yet been found in the bridge itself (Summerer 2011: 6). A river ford lay at Aşağıemerce in the late Roman period.

70

Peri Johnson

local path to villages south of the outcrop that was traversed before the relocation of a timber depot to the eastern half of the outcrop and also possibly in the Roman period. Burials on the outcrop continue with at least four tumuli on the southwestern shoulder of the outcrop. These tumuli have prominent locations at the edge of a precipitous drop to the ravine of an intermittent stream. The burials continue to the west and east of the schist outcrop. On the western shoulder of the ravine are three more prominently located tumuli and an area of rock-cut cist burials in loosely consolidated rock on Kurban Tepesi. A gently sloping valley with a spring and an off-site scatter of Roman ceramics, typical around Pompeiopolis, separates the three tumuli from a ridge composed of sedimentary rock that is part of the terrace rising from the Gökırmak flood plain and the location of the urban blocks of Taşköprü. Two tumuli are located on the eastern shoulder of the ridge and a third, the Pulcular Tumulus, on the crest. Traces of additional tumuli may have been erased during the earlier use of the summit of the ridge as a timber depot. A rock-cut tomb was exposed in the city during construction in the middle of the ridge’s slope. The tomb raises the possibility that a cemetery of rock-cut tombs was carved into the slope that is covered by urban blocks today, but if the rockcut tomb was capped with a tumulus, a second possibility is that the ridge is the location of dispersed prominently located tumuli. At least two tumuli are also located on the eastern edge of the schist outcrop, a third lies in the environs of Hacılar south of the outcrop (and was damaged ten years ago by looting) and the recently looted Bekdemir Tumulus is on the northern slope of the outcrop. Many other tumuli have been documented in the environs of Taşköprü such as around Kırha, on the road to Bekdemirekşi and on Kuruzluk Tepesi east of Yukarıemerce; many more have yet to be groundtruthed. In conclusion, the suburban tumuli and cemeteries of various rock-cut burials reflect the assemblage of burials found near Pompeiopolis, particularly the role of prominent tumuli in the landscape during the Roman period. The tumuli of Pompeiopolis augment the natural topography and indicate a rootedness in the landscape. As families settled in the landscape, they depended on this prominence and rootedness of tumuli to lay claim to the land they lived in. In 1885 Gustav Hirschfeld defined the Paphlagonian tomb as a rock-cut pedimental façade with a distinctively stocky column style.20 Tombs in this style are now dated to the Achaemenid (Late Iron Age, ca 550–330 BC) and Hellenistic periods, and the closest to Pompeiopolis was carved at Aygır Kalesi. In contrast to the Achaemenid and Hellenistic Paphlagonian tomb, by the Roman period prominent Paphlagonian burials took the form of tumuli that allude to a Roman and Paphlagonian identity that is reflected in Pompeiopolis’ status as the metropolis of Paphlagonia. How this transition came about can best be studied in the rural landscapes of Pompeiopolis.

4. The intertwined rural landscapes of the dead and living during the Hellenistic to Roman transition: Çetmi-Donalar Several areas in the rural landscape of Pompeiopolis present a complexity of settlement histories that we do not see in the immediate vicinity of Pompeiopolis. On the terraces between the Gökırmak and the Ekinveren Fault west of the Aygır Tepesi ridge lie settlements 20

Hirschfeld 1885.

How did the landscape of Pompeiopolis become Roman?

71

Fig. 5: Map of the Çetmi and Donalar areas (author’s map). along the linear alluvium of a series of tributary valleys (Fig. 5). Earliest is the high Early Bronze Age (third millennium BC) mound of Çetmi Höyük (Üyük Tepe) with a surrounding lower mound on the eastern edge of the alluvial fields in the Cipcip valley.21 In the second millennium BC, settlement continues only at a secondary Early Bronze Age settlement at Karacaoğlu on a rocky hill nestled between two intermittent streams 2km east of the Cipcip stream.22 Springs in the area of Karacaoğlu ensured repeated settlement of the area later in the Iron Age and the Byzantine period (eighth to twelfth centuries). The primary settlement of the area in the Iron Age is, however, the defensible ridge location of Yüklütepe. The Cipcip stream runs through a ravine along the eastern edge of a bedrock outcrop in the centre of the settlement, whereas the hollow way of the Ottoman and ancient 21

Çetmi Höyük: Donceel-Voûte 1979: 196; Marro et al. 1996: 275 s.v. B33/1, 279; Özdoğan et al. 1997: 309. 22 Karacaoğlu: Jacopi 1937: 12 s.v. Karacom; Donceel-Voûte 1979: 196.

72

Peri Johnson

route to the Devrekani plateau from Taşköprü runs along the western edge of the outcrop. Roughly hewn blocks from presumably Iron Age defensive walls lie around the outcrop where only Iron Age ceramics were collected in 2009.23 Settlement continued in the Hellenistic and Roman periods to the west, north and south of the outcrop as far as half the way to the three or more tumuli at Zelatin Tepe lying 1km distant on the western shoulder of the ridge in an area of loosely consolidated rock from which the tumuli were built. A roughly finished and poorly preserved columnar rock-cut tomb facing the hollow way of the Devrekani route was carved in a limestone escarpment 1km upstream of the outcrop. Although the tomb is probably Hellenistic and not Roman, the extensive scatter of Roman ceramics requires that the possibility of a Roman date for the three tumuli at Zelatin Tepe and an additional three prominent tumuli on the Kızıltepe ridge lying 1km east-northeast should not be dismissed. Fortunately, not all places have multi-period settlements surrounded by undatable tumuli similar to Yüklütepe. The Kayaaltı settlement at Bademci is located in an area to the west of the Kızıltepe ridge that was newly settled in the Roman period. The Roman and Byzantine Kayaaltı settlement lies 1.8km distant on the eastern side of the ridges of Kızıltepe and Karaoğlan Tepesi, both prominent rocky terraces covered with a web of abandoned field walls.24 Although difficult to date, might these abandoned fields have first been cultivated in the Roman period? In Kayaaltı we see our first example of Roman settlement expansion into an agriculturally more marginal area nestled between the rocky terraces, the eroded hills along the Karadere and the Ekinveren Fault. At Kayaaltı we also see a clearer example of the intertwined landscapes of the dead and living. A Roman sculpted rock-cut tomb with opposing lions in a pediment within a simple frame carved high in the limestone outcrop running along a fault gazes down on the settlement and is near an associated cemetery. The tomb is prominent, inaccessible and singular in the landscape around Bademci. Can we identify landholding through the proxy of the tomb in this place? Although modest, the tomb indicates an exclusive right to carve in its cliff location. The nearby cemetery indicates no such exclusive rights, only a social recognition that the buried were settled in, and belonged to, the area. Although archaeological evidence is always inconclusive and interpretations tentative, landholding covers a broad range of rights of use negotiated through social relations, in contrast to landownership, and Bademci does indicate a social difference between those interred in the rock-cut tomb and those interred in the cist burials. As the landscape filled up with low-lying agricultural settlements in the Hellenistic and Roman periods, of which Bademci is a later example, settlement returned to the fields around Çetmi Höyük but not the mound itself. Burials associated with the Çetmi settlement are prominent tumuli on hills to the south: one tumulus on Kuştepesi and three on Hasankahya ridge. These dispersed tumuli contrast to the cemetery composed of a cluster of at least ten tumuli ranging in height from 3m to 8m at Kırktepeler, on the shoulder above 23

The Yüklütepe outcrop appears the same in the CORONA satellite photograph from 1965; therefore, the later road construction on the eastern edge of the outcrop was probably not associated with the leveling of the summit of the outcrop that would have changed the ceramic distribution. 24 Bademci: Jacopi 1937: 13, pl. 11.36; Gökoğlu 1952: 74–5; Donceel-Voûte 1979: 196; Marro et al. 1996: 275 s.v. B33/4 Kayaaltı; Özdoğan et al. 1997: 311.

How did the landscape of Pompeiopolis become Roman?

73

the alluvial terraces of the Gökırmak 3km south of Çetmi Höyük. In 1935 Jacopi excavated two of the tumuli and on his return in 1936 he assembled the artefacts of an additional four tumuli looted in his absence.25 The tumuli span the fourth to the first century BC and possibly later. They have burial assemblages of moderate wealth, including two gold-leaf diadems and a core-formed amphoriskos dating to the second or first century BC. Although these tumuli do show the difference between singular tumuli located in prominent locations and groups of smaller tumuli, they also only demonstrate the tenuous nature of arguments about unexcavated tumuli and the difficulty of separating the Hellenistic from the Roman at settlements that offer no evidence of rupture comparable to the historical references to conflict in the transition from the Pontic Kingdom to the Roman Empire. Fortunately, the fortified outcrops over the hills to the northeast of Bademci at Donalar do demonstrate the transition from forts for elite residence to a Roman settlement that was in all probability agricultural and possibly also metallurgical, if the slag surveyed is Roman. The Donalar settlement consists of two paired forts on a limestone ridge defining the boundary between the mountains and the valley and cut here by the Karadere.26 Dating to the Late Iron Age and Hellenistic period, the western fort was built with a stepped ashlar wall that rose from the valley floor and climbed the sloping valley face of the ridge. At Donalar only wall footings survive with the blocks long removed. A precipitous drop protected the mountain face of the fort. Fine and common Hellenistic ceramics cover the slope on both sides of the wall and the foot of the slope. A stepped tunnel with cuttings for a door is the only other feature in the walled area. The absence of Roman ceramics from the western outcrop demonstrates the abandonment of the fort in the Roman period and the eventual dismantling of the fortifying wall. The same fate befell the larger eastern fort protected on the east by both an artificial trench and a wall that also ran along the accessible sections of the valley face of the outcrop. The sides of the outcrop facing the mountains and stream are even more precipitous than at the western fort and wall footings are absent. Near the stream, the Kalekapı sculpted Paphlagonian tomb dated to the Achaemenid period is carved in a chiselled vertical limestone face. The only ceramics collected within the fort are a few eroded Late Iron Age or Hellenistic grey wares. The slope on the valley face of the outcrop is, however, graced with an abundance of Roman ceramics. Despite continued settlement on the eastern bank of the Karadere, the ceramic distribution supports the conclusion that the walls of the fort were dismantled in the Roman period. A spiralling stepped tunnel on the mountain face of the outcrop is an altered phreatic tube and dates to the period when the fort was in use and not later. The ritual niche and lower stepped tunnel on the eastern ridge are accessible not from within the walls but from the slope and may have been in use throughout the Roman period. At Donalar we see the abandonment of the most prominent known Hellenistic forts in the environs of Pompeiopolis. Similar to Yüklütepe, the place is not abandoned, however, as the Roman settlement shifts to the slopes below the eastern fort. These shifts at 25

Kırktepeler: Jacopi 1937: 6, 10, pls 5.14, 8.27–9, 9.30–1; 1938: 5–8, pls 2.3–5, 3.6–10, 4.11; Gökoğlu 1952: 49. 26 Donalar: Fourcade 1811: 39–41; Leonhard 1915: 82–3, 129, 132–4, 235–6, 246–57, pls 5, 19, 23–5; Gökoğlu 1952: 65–71, 124; Bittel and Naumann 1965; Gall 1966: 13–57, 116–22, pls 2–3; Donceel-Voûte 1979: 196 s.v. Donalar-Kapıkaya; Barat et al. 2009: 400–4; Summerer and Kienlin 2010.

74

Peri Johnson

Yüklütepe and Donalar, the return to Çetmi Höyük and the expansion into the Bademci area participate in the rising density of settlements during the Roman period, but this is an increase that develops in response to existing places of settlement in the landscape.

5. The prosperous Roman countryside: Karapürçek-Yeniköy and Tepedelik The areas of greatest density of Roman settlement, however, lie closest to the city: both north of the Aygır Tepesi ridge and south of the southern cemeteries of the city. North of the ridge a series of intermittent streams with braided gravel beds carry sediment from the mountains to the Gökırmak north of the Sivritepe cemetery. The gentle slopes around these streams and the numerous springs encouraged settlement expansion into the area that began in the Hellenistic period and flourished in the Roman. Ören, on a ridge near Karapürçek, is the largest settlement in the area. The ridge to the southwest is an extension of the Roman settlement and is the location of the probable Roman villa near Zelo surveyed by the Pompeiopolis project in 2008. The identification of the place as a villa derives from the assemblage of architectural blocks, roof tiles, mosaic tesserae, glass, and fine and common ceramics. In contrast, prosperous Roman villages are identified by their mortar and rubble walls, roof tiles and abundant common ceramics. The architectural traces are much fewer or absent at settlements identified as hamlets. In the late Roman period, the settlement that developed the most is downstream at Yeniköy, where surface ceramics are distributed along a low rise at the edge of the flood plain. Previously, the end of the settlement was dated to the medieval period (approximately the nineth century), but recent analyses by Krzysztof Domżalski have brought the date of the end of the settlement down to the seventh century.27 The assemblage of surface artefacts, particularly the abundance of common ceramics and the absence of roof tiles except at the southern end of the settlement, contrasts with the Zelo assemblage and suggests an extensive agricultural settlement built of wood and mudbrick. At the southern end rises a low mound created by the collapse of a mortar and rubble building roofed with tiles and a short stretch of an apsidal wall, the architectural traces of a church. The late Roman period remains a period of prosperity in the rural landscape, and settlement density continues to be high, but, considering the surface artefacts at Yeniköy, we can propose a movement away from investment in mortar and rubble domestic architecture in rural settlements. A second area of high settlement density lies to the south of Taşköprü around the Tepedelik neighbourhoods of Çevik, Tavukçuoğlu, Aşağı Tepedelik and Yukarı Tepedelik (Fig. 6). In 1936, Jacopi excavated three Roman tumuli located on ridges composed of terraces covered with loosely consolidated rock to the east and west of the Kuru stream.28 The first tumulus (Jacopi’s Tumulus D) covered a simple cist tomb built of unworked slabs and furnished with simple burial artefacts including a jug, a small pot and a spindle whorl. The tomb was excavated in the middle of an Early Bronze Age mound that is now 27

Johnson 2011: 199–200. Intensive survey is required in order to characterise the early Roman period at Yeniköy. 28 Tepedelik: Jacopi 1938: 37–40, pls 34.145, 35.146–9, 36.150–3, 37.154–5.

How did the landscape of Pompeiopolis become Roman?

75

Fig. 6: Map of the Tepedelik area (author’s map). known as Tınar Tepe. A slight scatter of common Roman and later pottery was surveyed on the continuation of the ridge north of a saddle and on the slope below. Jacopi’s two other tumuli are the twin mounds İkitepe located further south near the neighbourhood of Aşağı Tepedelik; the first (C) had a similar burial assemblage to that associated with Tumulus D, whereas the second (A) included one bronze, one silver and six glass vessels among other artefacts. All together these prominent tumuli mark the burials of inhabitants of quite variable wealth near their homes and fields. The similar location and appearance but differing tomb assemblages of the İkitepe tumuli indicate the social and familial relations behind groups of tumuli. Burial in a tumulus, therefore, is not just a sign of wealth and privilege but indicates membership in social and familial groups that determine who is buried where and with what burial artefacts. Many other tumuli are located in this valley. Beginning in the north is a group of three on a terrace above the stream at Çevik, which are now looted or built over, and one tumulus and a surrounding cemetery damaged by reforestation at Derinyar. Unfortunately, the

76

Peri Johnson

Fig. 7: Manigünez Tepeleri from the ascent to the pass over Salavantepe (author’s photo). Roman (and Hellenistic) settlements that accompany these tumuli and burials have not been surveyed. Many tumuli also line the southward route to the mountain valleys that begins its first ascent from İkitepe up to Salavantepe, a pass that defines the boundary between the low-lying Gökırmak valley terraces and the mountain valleys to the south. A parallel ridge to the route, the Manigünez Tepeleri, is dotted with a line of approximately 12 tumuli (Fig. 7) and at least four tumuli are located in prominent places on the east and west of the pass over Salavantepe. Although the landscape was secure and densely settled in the Roman period, the tumuli in the Tepedelik area, most of which are probably early Roman with a few dating to the Hellenistic period, testify to the importance of established and reliable routes between the valley terraces, through mountain valleys and over mountain passes. In the late Roman period, the landscape populated with tumuli and settlements underwent a shift as mortuary practices changed, insecurity increased and routes needed to be defended. One of the most important long-term routes connected the Taşköprü area to the Kızılırmak through the Köçekli plateau.

6. The fortifying of late Roman settlements: Alisaray and Aşağıemerce The Köçekli plateau lies between the Elekdağ and Saraycık mountains, and routes to the Kızılırmak traversed the plateau or joined the Kastamonu route to the Devrez valley upstream of the confluence of the Devrez river with the Kızılırmak by following routes south of Salavantepe. The western entrance to the Köçekli plateau is guarded by a massive schist outcrop, Tepekaya (Fig. 8). From the summit of Tepekaya the area of Aygır Kalesi is visible, 20km distant, as well as a few hills along the route to Taşköprü such Kaletepe near Belençayır and Dikmen Tepesi east of the Tepedelik area. Extensive dry-boulder walls likely dating to the second millennium BC, the first period when imperial (Hittite) control of the

How did the landscape of Pompeiopolis become Roman?

77

Fig. 8: Map of the Alisaray area (author’s map). Gökırmak valley was important, cover the summit and upper slopes of Tepekaya.29 In the Roman and later periods the summit was occupied by a tower or fort but the construction of a water depot has disturbed the deposits, while bringing to the surface uneroded Roman ceramics. Between Tepekaya and the Tepedelik area, the route first descends steeply through forested slopes before passing through the Alisaray valley. Although the valley is not visible from Tepekaya, two fortified outcrops protected the route in the second millennium BC. In the south is a circular Hittite fort on the outcrop on the northern side of Abdalhasan; in the north is the outcrop on the western shoulder of the Aşağıçayı, Kaletepesi.30 Whereas the latter was also occupied in the Late Iron Age and Hellenistic periods, both have a late Roman presence. Kaletepesi was fortified by a mortar and rubble defensive wall dated mostly by ceramics from two looting pits, whereas Abdalhasan has disturbed late Roman deposits that are difficult to characterise at the centre of the summit. Kaletepesi demonstrates that the late Roman landscape experienced a return to the defendable locations 29

Tepekaya: Kuzucuoğlu et al. 1997: 277 s.v. C34/4, 287–8, 293, 297, 304; Özdoğan et al. 1998: 66, 74, 92 figs 7.10–11, 95 fig. 1; Özdoğan et al. 2000: 43–4. 30 Abdalhasan: Barat et al. 2009: 405–6; Kaletepesi: Gökoğlu 1952: 175–6; Johnson 2011: 197–8.

78

Peri Johnson

Fig. 9: Map of the Aşağıemerce area and the confluence of the Aşağıçayı with the Gökırmak (author’s map). occupied in the second and first millennium BC. The largest Roman settlement in the Alisaray valley is located under the low-lying Kese, a village on the eastern bank of the Aşağıçayı. The walls of a late Roman church were excavated by the Kastamonu Museum and walls of a bath building are visible in the river bed. The ridge rising from the western bank of the river is the location of several looted tumuli and disturbed rock-cut cist burials are found on a low ridge above the river in the village of Alisaray. The prosperous Roman village at Kese and the cemeteries of tumuli and rock-cut cist burials are characteristic of Roman settlements that come to be defended by the reoccupation of forts in the late Roman period. A second cluster of settlements identified during analysis of the GeoEye images at a river ford at the confluence of the Aşağıçayı with the Gökırmak, and just 3km downstream of Pompeiopolis, experienced a similar settlement trajectory to that identified in the Alisaray valley: Harmanyeri/Sivritepe, Aşıra, Dolaman Tepesi and Dolaman Arkası (Fig. 9). Located in the acute angle between the river and its tributary, Harmanyeri is composed of artificial terraces built of drystone masonry retaining walls around a ridge covered with the loosely consolidated rock found at many locations within the environs of Pompeiopolis, but here the rock is particularly friable (Fig. 10). The use of the terraces for winnowing wheat has prevented the disturbance of their surfaces. The few sherds visible in a road cut indicate occupation in the second millennium BC and the late Roman period. A separate mound of

How did the landscape of Pompeiopolis become Roman?

79

Fig. 10: Harmanyeri and Sivritepe from Dolaman Tepesi (author’s photo). building collapse composed of roof tiles, mortar, building tiles, rubble and plaster is a separate and much-looted entity on top of the terraced part of the ridge (Fig. 9, L8). A higher and unterraced part of the ridge, Sivritepe, to the south is the location of a late Chalcolithic and Early Bronze Age settlement where at least one much-looted tumulus was built of the prehistoric deposits, similar to the tumulus excavated at Tinar Tepe. Other probable Roman and later burials rest on the western shoulder of the ridge. Determining the extent of settlement on the Harmanyeri-Sivritepe ridge and the wall traces of any particular period from the second millennium BC onwards is difficult. However, the second millennium and late Roman periods are best represented. On the northern bank of the Gökırmak is a second terraced ridge, Aşıra, where unambiguous second millennium, Early Iron Age (ca 1200–950 BC), Hellenistic and late Roman ceramics have emerged from recent looting pits. The visible stretches of retaining walls on Aşıra are built of a mortar and rubble construction that dates to the latter period. The third side of the confluence is defended by the high promontory of Dolaman Tepesi, which was fortified in the Early Iron Age with a drystone schist rubble wall 35m long. Although the brown handmade wares clearly date to the Early Iron Age, a few wheelmade dark grey wares indicate continued occupation into the beginning of the Middle Iron Age (ca 950–550 BC). The early Roman settlements are dispersed at Dolaman Arkası, low-lying terraces along the northern bank of the Aşağıçayı that begin at the southeastern foot of Dolaman Tepesi. The first settlement is a sloping terrace covered with prosperous buildings built of rubble and mortar walls which are visible in looting pits and cut by gullies eroding through lenses of tile roof collapse. The cemetery associated with Dolaman Arkası lies on an outcrop that rises at the eastern edge of the first settlement and in the ravine to the north. A second settlement is located to the east of the road to Çarıklı, a route to the Byzantine fortress of Kızkalesi. At the confluence of the Aşağıçayı in the early Roman period, the higher terraces and promontories were abandoned for more accessible locations along the river banks, whereas the higher terraces were reoccupied once the defence of the landscape became

80

Peri Johnson

necessary in the late Roman period. By the ninth century, however, the defence of the river ford and the wider Taşköprü area required the construction of the fortress on the more elevated and defendable promontory of İncebel Tepesi where the Byzantine fortress of Kızkalesi was built.31

7. Conclusion To conclude, let us return to the question posed in the title of this paper. How did the landscape of Pompeiopolis become Roman? Half of the answer to this question is familiar to researchers studying neighbouring regions of the empire: through the prosperity and density of the settlements in the landscape. Each place contributed differently to the rising density. Places settled in the Hellenistic period continued to be settled, although ruptures invisible in surveyed assemblages must be assumed. Places long abandoned, such as Çetmi Höyük, were resettled. Settlements expanded into more marginal places such as Bademci and Yeniköy, and shifted to more low-lying river banks such as Kese and Dolaman Arkası. Suburban valleys experienced a particularly high increase in settlement. Simultaneously, the landscape filled up with the burial monuments of the inhabitants as they negotiated their place in the landscape. This answer is not as defined as Roman imperial identities fostered by imperial institutions and urban social relations such as the ephebic list but rather a more recursive identity that both cultivated, and was cultivated by, the pax romana and its lingering effects in the Anatolian peninsula. The greater security and prosperity of Roman rural landscapes differed from Hellenistic settlement expansion and was a part of Roman imperial identity and not a condition of its emergence. The other half of the answer is that the landscape became more Roman by becoming more Paphlagonian. On a basic level, becoming Paphlagonian can be the preference for tumuli, understood as attachment to local strategies of disposing of the dead. On a more abstract level, the saturation of the landscape with monuments implies the planting of roots and the attachment of the inhabitants to the places where they lived and the resting places of their dead. The understanding of landholding that emerges from the survey evidence around Pompeiopolis, therefore, is not directly relevant to the urban privilege revealed by epigraphy. Rather, we understand more about the processes through which landscapes come to belong to inhabitants during a period of prosperity and security. Simultaneous with the rising insecurity of the late Roman period, settlement returns to defendable places occupied in the Hellenistic period, the Late Iron Age and the second millennium BC. Living in the landscape in the late Roman period began to require defence of the land and not the population of it with monuments. Peri Johnson University of Illinois at Chicago Department of Art History 929 W. Harrison, Chicago, IL 60607, USA [email protected] 31

See Ritter this volume.

How did the landscape of Pompeiopolis become Roman?

81

Bibliography Akok, M. 1948. ‘Samsun İli Havza İlçesinin Lerdüğe köyünde bulunan tümülüsler’, Belleten 12, 835–54. Bal, N. 2014. ‘Selmanlı Tümülüsü Kurtarma Kazısı’, Müze Çalışmaları ve Kurtarma Kazıları Sempozyumu 2013 22, 337–48. Barat, C., J. Dalaison, F. Delrieux, J.-F. Pichonneau and B. Remy 2009. ‘Prospection et inventaire archéologiques dans la vallée du Gökırmak. Premier rapport préliminaire: région de Taşköprü’, Anatolia Antiqua 17, 393–409. Bittel, K., and R. Naumann 1965. ‘Notizen über einige paphlagonische Felsgräber’, Anadolu Araştırmaları 2.1–2, 71–84. Çevik, F., and M. Ritter 2013. ‘Pompeiopolis nekropolü yüzey araştırması (Paphlagonya)’, Araştırma Sonuçları Toplantısı 2012 30, 169–82. Domżalski, K. 2011. ‘Late Roman pottery from Pompeiopolis’, in Summerer (ed.), 163–78. Donceel-Voûte, P. 1979. ‘Gökırmak valley, 1977 (Recent archaeological research in Turkey)’, Anatolian Studies 29, 196–7. [Fourcade, P.-T.] 1811. ‘Mémoire sur Pompeïopolis ou Tasch-Kouprou, avec quelques remarques sur Tovata ou Voyavat’, Annales des voyages, de la géographie et de l’histoire 14, 30–58. Gall, H. von 1966. Die paphlagonischen Felsgräber: eine Studie zur kleinasiatischen Kunstgeschichte. Tübingen. Gökoğlu, A. 1952. Paphlagonia (Paflagonya): gayri menkul eski eserleri ve arkeolojisi. Kastamonu. Hirschfeld, G. 1885. Paphlagonische Felsengräber: ein Beitrag zur Kunstgeschichte Kleinasiens. Berlin. Jacopi, G. 1937. ‘Missione archeologica italiana in Anatolia relazione sulla prima campagna esplorativa’, Bollettino del Reale Istituto di Archeologia e Storia dell’Arte 7, 3–26, pls 1–30. Jacopi, G. 1938. ‘Esplorazioni e studi in Paflagonia e Cappadocia’, Bollettino del Reale Istituto di Archeologia e Storia dell’Arte 8, 3–43, pls 1–38. Johnson, P. 2011. ‘Topographies of urbanization: survey in and around Pompeiopolis’, in Summerer (ed.), 195–205. Kuzucuoğlu, C., C. Marro, A. Özdoğan and A. Tibet 1997. ‘Prospection archéologique francoturque dans la région de Kastamonu (Mer Noire). Deuxième rapport préliminaire’, Anatolia Antiqua 5, 275–306. Leonhard, R. 1915. Paphlagonia: Reisen und Forschungen im nördlichen Kleinasien. Berlin. Marek, C. 1993. Stadt, Ära und Territorium in Pontus-Bithynia und Nord-Galatia. Tübingen. Marro, C., A. Özdoğan and A. Tibet 1996. ‘Prospection archéologique franco-turque dans la région de Kastamonu (Mer Noire). Premier rapport préliminaire’, Anatolia Antiqua 4, 273–90. Marro, C., A. Özdoğan and A. Tibet 1998. ‘Prospection archéologique franco-turque dans la région de Kastamonu (Mer Noire). Troisième rapport préliminaire’, Anatolia Antiqua 6, 317–335. Mientjes, A. C. and M. Pluciennik 2013. ‘Land ownership and rights of use on land in the south Italian countryside: ethnoarchaeological and historical perspectives’, in M. Relaki and D. Catapodi (eds), An Archaeology of Land Ownership. New York, 229–59. Özdoğan, A., C. Marro and A. Tibet 1997. ‘Kastamonu yüzey araştırması (1995)’, Araştırma Sonuçları Toplantısı 1996 14.2, 303–30. Özdoğan, A., C. Marro and A. Tibet 1999. ‘Kastamonu yüzey araştırması 1997 yılı çalışmaları’, Araştırma Sonuçları Toplantısı 1998 16.2, 219–44. Özdoğan, A., C. Marro, A. Tibet and C. Kuzucuoğlu 1998. ‘Kastamonu yüzey araştırması 1996 yılı çalışmaları’, Araştırma Sonuçları Toplantısı 1997 15.2, 63–104. Özdoğan, A., C. Marro, A. Tibet and C. Kuzucuoğlu 2000. ‘Kastamonu yüzey araştırması 1998 yılı çalışmaları’, Araştırma Sonuçları Toplantısı 1999 17.2, 41–56. Summerer, L. 2011. ‘Pompeiopolis. Eine Zwischenbilanz aus der Metropole Paphlagoniens nach fünf Kampagnen (2006–2010)’ in Summerer (ed.), 3–15. Summerer, L. (ed.) 2011. Pompeiopolis I: eine Zwischenbilanz aus der Metropole Paphlagoniens nach fünf Kampagnen (2006–2010). Langenweißbach. Summerer, L., and A. von Kienlin 2010. ‘Achaemenid Impact in Paphlagonia: rupestral tombs in the Amnias Valley’, in J. Nieling and E. Rehm (eds), Achaemenid impact in the Black Sea; Communication of Powers. Aarhus, 195–221.

82

Peri Johnson

YIldIrIm, C., T. F. Schildgen, H. Echtler, D. Melnick, B. Bookhagen, A. Çner, S. Niedermann, S. Merchel, M. Martschini, P. Steier and M. R. Strecker 2013. ‘Tectonic implications of fluvial incision and pediment deformation at the northern margin of the Central Anatolian Plateau based on multiple cosmogenic nuclides’, Tectonics 32, 1–14. Zhuravlev, D. 2011. ‘Early Roman Fine Ware from Pompeiopolis’, in Summerer (ed.), 149–62.

Kristina Winther-Jacobsen

Contextualising Neoklaudiopolis: a glimpse at settlement dynamics in the city’s hinterland

1. The Where East meets West Project The Where East meets West Project (hereafter WEmW) investigates the Pompeian model of settlements in northern Anatolia and its trajectory from different material and historical perspectives focusing on one of its cities, Neoklaudiopolis, the modern-day Vezirköprü.1 The city is located approximately 100km south of the Black Sea coast on the eastern side of the Halys river (Fig. 1). Tønnes Bekker-Nielsen has recently published an article in Orbis Terrarum on ‘350 years of research on Neoklaudiopolis’ which discusses previous research in the area by numismatists, 19th-century travellers and historical geographers as well as survey archaeologists, and, more specifically, by the Nerik Project.2 Previous research has focused on the rich collection of inscriptions or the Bronze Age, and WEmW is attempting to broaden the focus by fleshing out the settlement system associated with the road infrastructure of the Roman to early Byzantine period. During the extensive survey undertaken in order to map the Roman road network around Neoklaudiopolis we identified elements of the settlement system which resulted in the production of a map covered in dots, some indicating spolia and others noting settlement sites identified as such by find scatters and architecture. The aim of this article is to discuss the survey evidence for a settlement hierarchy, mainly the results of three weeks of work in 2013. The study did not cover a large area systematically or intensively, but it has provided us with a set of case studies. Due to the extensive nature but limited amount of data from the Neoklaudiopolis area, I will explore the settlement dynamics by tying our qualitative data into the quantitatively-documented settlement patterns observed by the Project Paphlagonia survey.

1

WEmW is directed by Tønnes Bekker-Nielsen, University of Southern Denmark. The archaeological fieldwork is directed by Kristina Winther-Jacobsen, University of Copenhagen. 2 See Bekker-Nielsen and Czichon this volume. The Nerik Project (http://www.nerik.de/ survey/orte.php) is directed by Jörg Klinger, Free University of Berlin, and the field director is Rainer Czichon, University of Uşak.

84

Kristina Winther-Jacobsen

Fig. 1: Map of the sites in the survey area discussed in the text (cartography by R. Szydlak).

2. Project Paphlagonia The Project Paphlagonia survey (1997–2001), directed by Roger Matthews, was an ambitious survey which assessed an area of nearly 8,500km2 in north-central Anatolia by a combination of extensive and intensive survey.3 The project experienced the chronological problems which all survey projects recognise:4 (1) it is difficult to differentiate between the Roman and Byzantine periods; (2) the Byzantine period is most richly attested by coarse wares which are difficult to date precisely; (3) frequent occurrences of tiles are even more difficult to assign to specific periods. As noted by Owen Doonan in his review of the publication, one of the interesting observations to come out of the project was that the intensive survey did not really change any of the main conclusions that emerged from the extensive survey.5 As already observed by Burcu Ercyas, Project Paphlagonia confirmed that the rural settlement pattern of the Roman period differed dramatically from that of preceding periods, but was comparable with that of the Ottoman period.6 The early Roman phase appears to be the best attested period when there was a steady increase in the density and intensity of settlements and the exploitation of agricultural resources.7 Urbanisation emerged, but not at the expense of rural settlements, and this resulted in the first clear 3 4 5 6 7

Matthews and Glatz 2009a. Matthews et al. 2009: 173–4; see also Pettegrew 2001; 2002. Doonan 2012: 119–20; Matthews and Glatz 2009a: 16. Ercyas 2006: 61; Matthews et al. 2009: 189. Matthews et al. 2009: 173.

Contextualising Neoklaudiopolis

85

hierarchy of settlements ranging from tiny hamlets/farmsteads through to villages and towns (although even the largest towns were relatively modest and the majority of sites were small).8 Finally, a number of well-constructed rubble-built fortifications, 40–100m long, are characterised by being built in drystone technique with a narrow entrance, regularly-spaced semicircular bastions and no inner structures. Usually there are no finds and the structures are dated to the eighth to the twelfth century for historical reasons.9

3. The comparative analysis The comparative analysis includes only sites identified by scatters of finds, not isolated finds of inscriptions, architectural fragments or spolia often encountered in village fountains.10 The four areas explored are: the Papaz Tarlası martyrium-complex, the Christian sanctuary on the Doyran-Tepeören road, the Tepeören settlement system and the Tepeören-Merzifon road settlement (Fig. 1). 3.1 The Papaz Tarlası martyrium-complex The field known as the Papaz Tarlası is located on a plateau just north of the river from modern Vezirköprü, early Roman Neoklaudiopolis.11 The site was formally identified in 2010 when geoelectric survey and a grab sample of finds collected under the auspices of the Nerik Project and Rainer Czichon revealed a well-preserved structure. This site will be discussed in some detail because it is central to the understanding of the chronology of other sites in the region. The dense 0.4ha scatter of pottery and roof tiles makes this a category 4 site of farmstead level according to the Project Paphlagonia typology; however, geoelectrical survey has revealed the plan of a different kind of structure (Fig. 2). The plan records a cruciform structure of ca 21m × 21m oriented east-west and attached to a square structure of ca 42m × 42m with a hexagonal structure in the middle of about 10m × 10m. Two small rectangular structures are located to the northeast – the so-called northeastern complex. No roadshaped anomalies appear; the complex is, however, well aligned with the route of the modern road, which most likely existed in Roman times, connecting the town Neoklaudiopolis with settlements at Oymaağaç and Adatepe. Remains of the southeastern corner of the foundation of the square structure, constructed of field stones bound with mortar, were exposed by erosion in the slope to the south. The interpretation of the complex associated with the cruciform structure at Papaz Tarlası is based on the plan. Cruciform buildings attached to a square structure with a central construction are well-known across the Christian world and are identified as Christian sanctuaries with a peristyle courtyard (the so-called atrium) with a fountain at the centre. 8 9 10

Matthews et al. 2009: 189. Matthews et al. 2009: 196. Olshausen and Biller 1984; see also the site catalogue of Project Paphlagonia under the category ‘Stones, not in situ’: Matthews and Glatz 2009a: 251–60. 11 A detailed report on the site in English has been submitted to the Proceedings of the Danish Institute at Athens: Winther-Jacobsen and Bekker-Nielsen forthcoming.

86

Kristina Winther-Jacobsen

Fig. 2: Georesistivity map of Papaz Tarlası (map by Harald van der Osten-Woldenburg). Although Christian sanctuaries in the shape of a free-standing Greek cross have wellknown predecessors, they do not appear to have been very common. The closest parallels in terms of both plan and scale come from Chersonesos in Crimea where several have been excavated.12 They can sometimes be reconstructed as having been built over a tomb, in one case the building even contained a reliquary shrine and, consequently, the Crimean structures are interpreted as martyria. The dates of the different cruciform structures in Chersonesos range from the early fifth to the tenth century. The concept of placing a domed tomb at the centre of a cross originates with the Church of the Holy Apostles in Constantinople, but the Latin cross is by far the most common form.13 The only genuine parallels for the free-standing Greek cross shape in Asia Minor known to me are two chapels in Binbirkilise in Cappadocia: Karadağ no. 12 and Karadağ-Mahalaç.14 Due to their size and subordinated association with churches, both these parallels are interpreted as chapels, and, according to the inscription associated with Karadağ-Mahalaç, the chapel was built in the memory of a person named Leo. The combination of the cruciform church and peristyle courtyard with a fountain in the middle finds more parallels, for example the so-called atrium in front of the church of St Peter in Rome.15 Much closer to Neoklaudiopolis, 12

The churches are published in Russian, but have been discussed by Romančuk 2005: 83–6, figs 18.11–15, 24, 27; a conference paper by Liudmila Khrushkova from 2006, previously published on the internet, is no longer available (website last accessed June 2013). 13 Grabar 1972: 151–62; Krautheimer 1986: 172, 175–7; see also Winther-Jacobsen and Bekker-Nielsen forthcoming for a detailed analysis. 14 Ramsay and Bell 1909/2012: 122–5, 249, 556–7. 15 Krautheimer 1986: figs 21–22.

Contextualising Neoklaudiopolis

87

Burcu Ercyas has excavated a hexagonal basin associated with a very complicated hydraulic system near Komana Pontica of a diameter similar to the basin at Papaz Tarlası.16 Ercyas and Ahmet Çnc have identified other examples of hexagonal basins in atria, but they tend to be somewhat smaller, such as that at the episcopal precinct in Kourion on Cyprus.17 Apart from the foundations in situ, we recorded several architectural fragments of grey limestone, which most likely come from the martyrium-complex, as well as numerous roof tiles and some floor tiles/bricks (bipedalis). The small size of the fragmented column found in the field and the date of the other finds make it likely it supported an arched entablature. In order to get a better understanding of the chronology and activities, the Papaz Tarlası was surveyed systematically and intensively in 2013. The field is one geomorphological unit; it was divided into a grid of uniform 10m × 10m blocks. The strategy consisted of counting and recording finds from 10% of the surface; this was achieved by the total count/ recording of all finds in 1m transects, with fieldwalkers spaced at 9 m intervals. Total recording equates to noting anything from the size of a thumbnail upwards. Additionally, a random sample was recorded from squares between and from around the grid at the edge of the field. Figure 3 illustrates the correlation between the surface and sub-surface finds. The map shows the survey grid overlaying the field and the geoelectrical map; structures are indicated in red. Finds were recorded from all the units: the numbers in black indicate the highest densities, while the numbers in blue indicate the lowest densities. The numbers in pink indicate the highest average weights of sherds recorded. The total recording revealed a close correlation between average weight and density, suggesting that new fragments are being ploughed by each agricultural episode, and the distribution of finds showed a close correlation with the structures detected subsurface. The highest concentrations of finds (up to 1.4kg per square metre) were recorded over the cruciform structure and in the southeastern corner of the atrium, but high concentrations were also recorded over the northeastern complex and further northeast of this, leading us to conclude there are probably additional structures in that area. There are a few exceptions of high average weight recorded away from the subsurface structures (indicated on Fig. 3 in purple). These can be explained by the occurrence of a single very large sherd, a type of find which behaves differently on the surface as it gets caught very easily in agricultural equipment.18 The ratio of architectural ceramics to other types of ceramics/pottery is 271kg to 16kg, or 17:1, and the high densities and high average weights are very much the product of the presence of tiles. The style of roof evidenced by these tiles is the so-called Sicilian, constructed with flat pan tiles and open V-shaped cover tiles.19 Although different types of tiles appear differentiated by time, workshop and fabric, they all share a high degree of simplicity with almost no refinements recorded. 16 17 18 19

Ercyas and Çnc 2010. Megaw 2007: fig. 1. Z; Ercyas and Çnc 2010: 293. Baker 1978; Dunnel 1990: 592. For example Adam 1994: fig. 498.

88

Kristina Winther-Jacobsen

Fig. 3: Gridded survey map of Papaz Tarlası indicating the subsurface structures and other recorded features (author’s image). Figure 4 is a distribution map of the non-architectural ceramics recorded in the transect lines, sorted into a functional typology. They are consistent with domestic types: table wares, cooking wares, kitchen wares, storage vessels and lamps. Kitchen wares are not indicated on the plan as they appeared in all the transect lines. The ratio of architectural ceramics to other types of ceramics is of course not constant across the field, but a particularly interesting variation is observed in the cluster overlying the northeastern complex. Here the ratio is only 4:1 as a consequence of 3.59kg of pithos fragments recorded in transect line 150/080; this is a type of kitchen ware rarely recorded in other parts of the field and its presence here suggests a combined domestic-agricultural production function for the northeastern complex. The random recording of finds from the squares between the transect lines was aimed specifically at diagnostic and rare fragments (Fig. 5). Although these data are not quantifiable, they support the suggestion of a domestic function for the northeastern complex: The majority of pithos sherds and all the fragments of spouted basins came from the northeastern part of the field where the amalgamation of tiles, pithoi, kitchen wares, cooking wares and table wares, and the addition of a single iron slag suggests a combination of domestic and productive activities for the northeastern complex.20 However, domestic pottery is also associated with the martyrium-complex and it is very tempting to interpret the 20

Although the iron slag cannot be dated and may be intrusive.

Contextualising Neoklaudiopolis

89

Fig. 4: Functional categories recorded in transect lines across Papaz Tarlası (author’s image). entire complex as a roadside martyrium-church associated with small-scale habitation. Based on the high proportion of red-slipped table wares alone, the site differs from sites classified as farmsteads or hamlets according to the size of the scatter. However, according to written sources, some of the rituals in early Christianity were closely associated with meals.21 The complex does not appear to share any morphological similarities with the monasteries identified by the Paphlagonia survey, and it is also different from the Binbirkilise complexes.22 On current evidence, the Papaz Tarlası complex seems rather supraregional in style, drawing its inspiration directly from Constantinople. For chronological purposes, the table wares and coins are especially significant. The table wares are predominantly red slipped, although a few are non-decorated. Kryzstof Domżalski has kindly looked at the material and identified most of the fragments as Pontic Red Slip. The earliest coin, a late fourth- to early fifth-century silver coin, is contemporary with the earlier Pontic Red Slip forms.23 However, the predominant Pontic Red Slip form is form 7, which dates to the second half of the fifth century. Interestingly, this 21

For early Christian rituals associated with meals, see, for instance, Klinghardt and Taussig 2012. 22 Ramsay and Bell 1909/2012; Matthews et al. 2009, 194. 23 Vera Sauer will publish the coins in the Proceedings of the Danish Institute at Athens.

90

Kristina Winther-Jacobsen

Fig. 5: Selection of pottery from Papaz Tarlası (square 140–150/090–100) (author’s photo). form is quite rare at Pompeiopolis. The table ware chronology suggests the period of consumption was mainly in the second half of the fifth into the sixth century, but, since it is not currently possible to identify the table wares associated with later periods, this conclusion remains highly speculative. A few fragments of ceramics postdating the Late Antique period were recorded (e.g. a stamped tenth-century amphora handle, fig. 5), but none of the pottery can be dated securely to the seventh/eighth century, which is date of the latest coin recorded in 2013. Possibly associated with an earlier phase of occupation of the area are a few tiny abraded fragments of red- and black-slipped table wares which seem to predate the majority of finds. Also found in Papaz Tarlası was a fragment of a grave stela with an irregular round hole in the back which may have been reused as a threshold (Fig. 6). Some remains of mortar with small pebbles on the back indicate that it was fitted into some sort of architectural context. On the front of the block is preserved the top of the double-framed main panel and the lower part of the double-framed pediment flanked by acroteria. At the centre of the pediment is a rosette with curved pointy leafs. The acroteria appear to consist of at least three leaves pointing downwards and ending in three spirals.

Contextualising Neoklaudiopolis

91

Fig. 6: Stela from Papaz Tarlası (author’s photo). In 1988 Eckart Olshausen photographed a stela with a similar frame and rosette 18km southeast of Vezirköprü in Kocaoğlu.24 This stela is inscribed with the era of the city, dating it to the year AD 186/7. A stela from Pompeiopolis in the Kastamonu Museum may have been produced by the same craftsman.25 Although the pediment is decorated with a pine cone resting on acanthus leaves, the proportions and details of the double frame and acroteria are the same and quite distinctive. Six kilometres south of Vezirköprü in Karkucak, a fourth stela in the same style, decorated with a pine cone resting on acanthus leaves, was photographed by Olshausen in 1990.26 The stylistic and decorative similarities suggest the four stelae may have been made by the same craftsman. Consequently, the stela with the 24

Vera Sauer, personal communication: for which I am very grateful. I also wish to thank Eckart Olshausen for allowing me to mention this unpublished material. The stela will be published in the volume of inscriptions from Neoklaudiopolis which is in preparation for the series Inschriften griechischer Städte aus Kleinasien. 25 Marek 1993: cat. Pompeiopolis 38; see also Jacopi 1936: 5. 26 See above, n. 24.

92

Kristina Winther-Jacobsen

era indicated provide an approximate terminus post quem for the structure in Papaz Tarlası in which it was reused. The stelae create a material link between Neoklaudiopolis and its hinterland and Pompeiopolis. Of course, one could argue that the connection is not the provenance of the artefacts, but the craftsman, as itinerant artists are known from various forms of evidence to have existed.27 Consequently, it is necessary to evaluate how secure the find contexts are. Although almost never recorded in situ, the diffusion of architectural fragments, grave stelae and inscriptions creates the impression a diffused settlement pattern. This appears to be part of the evidence for the stability of the hinterland in the urbanisation process of Paphlagonia, even if this is not discussed in the final report of Project Paphlagonia. The Project Paphlagonia site category ‘stones, not in situ’ appears in the catalogue and on the site distribution map (for instance site PS-003a).28 Even large artefacts have long been known to have been mobile, especially so since motorisation but also earlier. The passion for reusing spolia in public structures such as fountains or mosques is evidenced in almost every village around Vezirköprü (e.g. at Incesu, Tepeören, Boğazkoru and Geriz), and many private houses and gardens also include spolia. So-called ‘pretty’ (inscribed and/or decorated stone) pieces would be expected to have a high level of mobility simply because they are attractive. An example of the correlation between mobility, motorisation and attractive pieces is a grave stela built into the garden wall of a house in the village of Yağınözü. When asked about it, the owner of the house explained how he saw the stela in a garage parking lot next to the cemetery in Neoklaudiopolis approximately 6km away and was permitted by the owners of the garage to take it away. By the entrance door to his house were two additional spolia – two capitals of unknown provenance – and the question is where they came from. Do the spolia in the villages come from the actual village? Do they come from the fields of the villagers or even further away? Clearly, more studies of the rural processes of the use of spolia are needed. The parallels for the free-standing cruciform shape may be used as evidence to indicate that the Papaz Tarlası complex is a martyrium-church. Martyria are often located at necropoleis and it is tempting to identify the reused grave stela and the few tiny fragments of earlier Roman table wares as evidence for a roadside necropolis at this location during the early Roman period, but there is no secure evidence to confirm this at present. 3.2 The Christian sanctuary on the Doyran-Tepeören road We now move gradually south from Neoklaudiopolis. Six kilometres away, on a ridge east of Çamlıca village crossed by the Roman road running between Doyran and Tepeören, a site was identified based on a scatter of pottery and tile as well as two large architectural fragments associated with a small mound; a structure of at least 7m × 12m is indicated. The two large building fragments observed are (1) a large block (fragment of an architrave?) measuring 2.2m × 0.77m × 0.83m and moulded on two sides; and (2) a slab with a central recessed panel measuring 1.42m × 0.9m × 0.27m. The small finds consist mainly of roof tiles, kitchen wares and cooking wares similar to the types identified at Papaz Tarlası, as well 27

See, for instance, Trimble 2011: 121. Travelling workshops are a known phenomenon in architectural studies (e.g. the mausoleum workshop of southern Gaul: Kleiner 1977). 28 Matthews and Glatz 2009a: 251–60, fig. 6.29.

Contextualising Neoklaudiopolis

93

Fig. 7: Selection of pottery from the Christian sanctuary on the Doyran-Tepeören road (author’s photo). as a few red-slipped table wares, the form and origin of which could not be identified (Fig. 7). However, at this site we also recorded tiles with plastic flow-directors, which were not observed at Papaz Tarlası.29 Pavol Hnila has identified tiles with plastic flow-directors at Nerik; he has identified parallels for this stylistic feature at Chersonesos dated to the tenth to thirteenth century.30 Consequently, this feature appears to provide us with a genuine chronological distinction. Tiles with plastic flow-directors also appear at Zindankaya, 12km southwest of Vezirköprü, and at early Byzantine Kızkalesi near Pompeiopolis.31 At Zindankaya the tiles with plastic flow-directors are dated from the Late Iron Age to the Hellenistic period and identified as revetment plates. The evidence for dating the fragments before the Roman period is not discussed in the site report; the site is dated tentatively from the Iron Age to the Hellenistic period.32 The presence of tile with plastic flow-directors suggests some occupation in the Byzantine period, and clearly the chronology of the Zindankaya needs revising.

29

Of course, arguing from absence is never secure, even if we did collect 271kg of tile fragments at Papaz Tarlası. 30 Romančuk 2005: fig. 43; see Hnila this volume. 31 For Zindakaya, see Dönmez 2007: figs 11–12; on Kızkalesi, I gratefully thank Max Ritter, personal communication. 32 These dates are repeated by Johnson 2010: 386, cat. no. I.18.

94

Kristina Winther-Jacobsen

Fig. 8: Map of the Tepeören settlement system (cartography by R. Szydlak). Today there is a fountain (çeşme) at Çamlıca, and its source may also have been the attraction in ancient times. Due to the combination of the chronology, the monumental architectural fragments and domestic ceramics, the site can be identified as a small rural church associated with habitation. Rural Christian sanctuaries are a well-known phenomenon in the late Roman period in the eastern Mediterranean.33 Project Paphlagonia associates such sites with small monastic communities.34 3.3 The Tepeören settlement system Southeast of the village of Tepeören WEmW identified a cluster of settlements of different periods (Fig. 8). Based on a dense distribution of tile and pottery (Fig. 9) as well as a single column base a Roman site was identified on a wide plateau along the northern side of the river less than 2km southeast of the village of Tepeören. The wide range of pottery recorded at the site dates to the Roman period: late Roman fragments such as a base of a Pontic Red Slip form 3 ware vessel and early Roman Pontic Sigillata. The tiles are mostly very simple Corinthian style pan tiles, but fragments with plastic flow-directors have also been recorded here. The large number of tiles and amphora handles observed at the western end of the area suggests the settlement was located here where there is also a modern fountain. At 3ha Roman Tepeören qualifies as a category 5 site, defined as a village, in the Project Paphlagonia classification system. Several Paphlagonian settlements of the Roman to Byzantine period 33 34

For example Steinsapir 2005. Matthews et al. 2009: 194.

Contextualising Neoklaudiopolis

95

Fig. 9: Selection of pottery from Roman Tepeören (author’s photo). identified by Project Paphlagonia fall under this category: for example PS-185, which, from the distribution map, appears to be closely associated with PS-184, a necropolis.35 The southern side of Roman Tepeören is cut by the steep river valley and erosion has exposed bones in the southeastern corner of the site. In this area an unusually large proportion of the surface finds were table wares, which supports the identification of this area as a possible necropolis, with the table wares (as well as unguentaria) having been used as typical grave goods. Less than 500m east of the Roman site is a fortified hill measuring approximately 125m × 50m, or ca 0.6ha. The hilltop is completely covered by trees and scrubs as well as occasional robber trenches, and it is difficult to comprehend the layout of the architecture. The circumference wall traceable on the northern and eastern sides appears to be built with drystone masonry at the bottom, whilst the less well-built upper parts are constructed of smaller fieldstones; there seems to be at least one semicircular protrusion on the northern side. There are also remains of fieldstone walls abutting the circumference walls, including some mortared walls on the eastern side. 35

Matthews and Glatz 2009a: 251–60, sites PS-136, 147–8, 185, 227, fig. 6.29.

96

Kristina Winther-Jacobsen

In the Project Paphlagonia survey area Byzantine fortified hilltops ranged in width from 40m to 70m, but out of the five catalogued three were under 1ha.36 For historical reasons, Matthews and his team date such fortifications to the eighth to twelfth century, when the Byzantine Empire was under attack,37 and we suggest a similar date for the kale east of Tepeören based also on the occurrence of Corinthian-style pan tiles among otherwise Old Hittite pottery (see below). However, this kale is clearly different from the fortified hilltops with no inner structures and (usually) no finds identified by the Project Paphlagonia survey. The Byzantine fortification was clearly not the first settlement on the hill. On the southern side linear stone foundations and pottery dated by Rainer Czichon 38 to the Old Hittite period are visible in the slope. This settlement appears to have extended approximately 130m further east than the fortified area. Although this flat area was covered by thick vegetation, animal burrowing had brought sherds to the surface. Furthermore, the south-facing slope southeast of the hill was covered with a dense scatter of some Old Hittite and (mainly) Iron Age sherds, as well as a few later roof-tile fragments. This settlement appears to cover at least 1.5ha. If the provisional dates are correct, the Roman settlement may be interpreted as the successor to the Old Hittite and Iron Age settlements on and below the hill and the predecessor of modern Tepeören. Although modern Tepeören is located on the Roman road, the scarcity of spolia suggests the modern village postdates the Byzantine period. Indeed, the Tepeören settlement system comprises habitations across a small area, located never more than 2km apart, and covering a timespan of several thousand years. The Old Hittite occupation on the hilltop, the possible extension of the settlement towards the south during the Iron Age, the move to a larger plateau during the Roman period, and the possible use of the hilltop as fortification during the Byzantine period provide evidence of the changing needs and conditions of the lives of the inhabitants of the area. 3.4 The Tepeören-Merzifon road settlement Piles of tile, brick and fieldstones, as well as linear double-faced walls associated with earthworks were recorded 3.5km southwest of Roman Tepeören, across the river and up the mountain on the Tavşan ridge facing north. At 2ha the site also belongs to Project Pahlagonia’s category 5, defined as a village (similar to Roman Tepeören), and it appears to be unfortified. The site is difficult to date on the basis of the finds. The roof tiles belong to the same general type as the tiles from Papaz Tarlası; they are very simple with no refinements recorded. The site falls under the category of sites encountered by the Project Paphlagonia survey which are characterised by the frequent occurrence of tiles and are difficult to assign to specific periods.39 The site is located on the Roman road between Tepeören and Merzifon. Today there is a fountain at the site, and water may also have been part of the attraction in ancient times. 36 37 38 39

Matthews et al. 2009: 196. See above, n. 9. Rainer Czichon, personal communication: for which I am very grateful. Matthews et al. 2009: 173.

Contextualising Neoklaudiopolis

97

Fig. 10: Selection of pottery from the Tepeören-Merzifon road settlement (author’s photo).

4. Conclusion Apart from the table wares, the pottery from Neoklaudiopolis and its hinterland seems to be exclusively local, even if the style is regional. So far we have identified very little in the form of imported foodstuffs. Although the three stelae found on the western side of the Halys were not found in situ, the fact that all three come from the area around Neoklaudiopolis suggests a strong connection to this area. Consequently, the distribution of grave stelae of similar design provides evidence of cultural connections between the western side and eastern sides of the Halys, although the stelae could, of course, have been produced by an itinerant craftsman. According to the study of the Roman infrastructure, Neoklaudiopolis lies at the centre of the regional settlement system.40 Hopefully our planned fieldwork at Neoklaudiopolis in 2015 will help us to establish material connections between the city and its hinterland.

40

See Bekker-Nielsen and Czichon this volume.

98

Kristina Winther-Jacobsen

Compared to the results of the Project Paphlagonia survey, at first glance there seems to be a chronological difference. In Paphlagonia the early Roman period is the best attested period, whereas the late Roman period seems to be dominant around Neoklaudiopolis. This may, however, be a phenomenon produced by our very small sample size. Similar to Paphlagonia, urbanisation is new on the western side of the Halys, and the wide distribution of spolia recorded by, among others, Olshausen may indicate that urbanisation did not happen at the expense of rural settlements; however, the spolia are, of course, never found in situ and it is difficult to determine how far they have travelled from their primary use. The results of a single fieldseason have allowed us just a glimpse of the settlement system, but the hierarchic classes of sites – Christian sanctuaries with a domestic element, villagesize settlements and a fortification – are consistent with the types of Roman and early Byzantine sites identified by the Project Paphlagonia survey. The settlement classes missing are farmsteads/hamlets and towns; however pre-Roman settlements, such as, for instance, Oymaağaç and Adatepe, cannot be expected to correlate with the Roman infrastructure and our methodology is not sensitive to this type of site. In the case of Tepeören, which is on the Roman road, the modern village appears to postdate the Byzantine period, but other modern villages may have been occupied since the Roman period. The discussion of settlement patterns in Paphlagonia by Roger Matthews and Claudia Glatz focuses on the quantitative aspect, giving evidence to the many new sites recorded in the Hellenistic to Roman, Roman and Roman to Byzantine periods, as well as to discontinuity.41 Unfortunately, the diachronic recording of sites at a spatial level is not detailed enough to conclude whether or not the Tepeören settlement system may be typical of the wider region. Kristina Winther-Jacobsen University of Copenhagen Saxo-Institute Karen Blixens vej 4, 2300 Copenhagen S, Denmark [email protected]

Bibliography Adam, J.-P. 1994. Roman Building Materials and Techniques. Bloomington/Indianapolis 1994. Baker, C. M. 1978. ‘The size-effect: an explanation of variability in surface assemblage content’, American Antiquity 43, 288–93. Bekker-Nielsen, T. 2013. ‘350 years of research on Neoklaudiopolis (Vezirköprü)’, Orbis Terrarum 11, 3–31. Dönmez, Ş. 2007. ‘Zindankaya: the largest known Iron Age settlement in the Central Black Sea region, Turkey’, in: P. Kitzler (ed.), Third International Congress on Black Sea Antiquities in Prague, 11–18 September 2005 (Eirene 43), 143–60. Doonan, O. 2012. ‘Surveying landscapes: some thoughts on the state of survey archaeology in Anatolia’, Backdirt: Annual Review of the Cotsen Institute of Archaeology 2013, 118–23. Dunnel, R. C. 1990. ‘Artifact size and lateral displacement under tillage: comments on the Odell & Cowan experiment’, American Antiquity 55, 592–4. Ercyas, D. B. 2006. Wealth, Aristocracy and Royal Propaganda under the Hellenistic Kingdom of the Mithridatids in the Central Black Sea Region of Turkey (Colloquia Pontica 12). Leiden. 41

Matthews and Glatz 2009b: 241–2, fig. 8.3.

Contextualising Neoklaudiopolis

99

Ercyas, B. and A. Çnc 2010. ‘The hexagonal basin at Komana: a preliminary architectural study’, Middle East Technical University Journal of Field Archaeology 27.1, 281–96. Grabar, A. 1972. Martyrium, Recherches sur le culte des reliques et l’art premier chrétien I: Architecture. Paris (1946, repr. 1972), 152–61. Jacopi, G. 1936. Dalla Paphlagonia alla Commagene. Rome. Khrushkova, L. G. 2006. ‘Tauric Chersonesus (Crimea) in the 4th–5th centuries: suburban martyria’ Paper presented at the 21st International Congress of Byzantine Studies, London. Translation by Mr. Kesha Gelbak. Kleiner, F. 1977. ‘Artists in the Roman world. An itinerant workshop in Augustan Gaul’, Mélanges de l’École française de Rome. Antiquité 89, 661–96. Klinghardt, M. and H. Taussig (eds) 2012. Mahl und religiöse Identität im frühen Christentum; Meals and Religious Identity in Early Christianity (TANZ 56). Tübingen. Krautheimer, R. 1986. Early Christian and Byzantine Architecture (4th edition with S. Ćurčić). Yale. Marek, C. 1993. Stadt, Ära und Territorium in Pontus-Bithynia und Nord-Galatia (Istanbuler Forschungen Beiheft 39). Tübingen. Matthews, R. and C. Glatz (eds) 2009a. At Empires’ Edge. Project Paphlagonia Regional Survey in North-Central Turkey. London. Matthews, R. and C. Glatz 2009b. ‘People and place in Paphlagonia: trends and patterns in settlement through time’, in Matthews and Glatz (eds), 239–49. Matthews, R., Metcalf, M. and D. Cottica 2009. ‘Landscapes with figures: Paphlagonia through the Hellenistic, Roman and Byzantine periods, 300 BC–AD 1453’, in Matthews and Glatz (eds), 175–226. Megaw, A. H. S. 2007. Kourion: Excavations in the Episcopal Precinct. Washington D.C. Olshausen, E. and J. Biller 1984. Historisch-geographische Aspekte der Geschichte des Pontischen und Armenischen Reiches 1, Untersuchungen zur historischen Geographie von Pontos unter den Mithradatiden (Tübinger Atlas des Vorderen Orients, Beiheft B 29.1). Wiesbaden. Pettegrew, D. K. 2001. ‘Chasing the Classical farmstead: assessing the formation and signature of rural settlement in Greek landscape archaeology’, Journal of Mediterranean Archaeology 14.2, 189– 209. Pettegrew, D. K. 2002. ‘Counting and colouring Classical farms: a response to Osborne, Foxhall and Bintliff et al.’, Journal of Mediterranean Archaeology 15.2., 267–73. Ramsay, W. M. and G. L. Bell 1909/2012. The Thousand and One Churches. Cambridge. Romančuk, A. I. 2005. Studien zur Geschichte und Archäologie des byzantinischen Cherson (Colloquia Pontica 11). Leiden. Steinsapir, A. I. 2005. Rural Sanctuaries in Roman Syria: The Creation of a Sacred Landscape (British Archaeological Report International Series). Oxford. Trimble, J. 2011. Women and Visual Replication in Roman Imperial Art and Culture. Cambridge. Winther-Jacobsen, K. and T. Bekker-Nielsen forthcoming. ‘A Late Roman building complex in the Papaz Tarlası, Vezirköprü (ancient Neoklaudiopolis, northern Asia Minor)’, Proceedings of the Danish Institute at Athens 8.

Peter Bikoulis, Hugh Elton, John Haldon and Jim Newhard

Above as below: The application of multiple survey techniques at a Byzantine church at Avkat

1. The Avkat Archaeological Project: background and rationale The Avkat Archaeological Project (AAP) has carried out historical and archaeological investigation at the modern village of Beyözü and its surrounding area in the central part of the Mecitözü ilçe, Çorum province (Fig. 1).1 Apart from a brief visit in the 1960s by David French (Director of the British Institute at Ankara) and İsmet Ediz (Director of the Çorum Museum), no detailed archaeological investigation took place in Beyözü before the AAP began work in 2007. Beyözü, earlier known as Avkat, is often identified as ancient Euchaita, though Euchaita has also been located at Elvançelebi, Çorum and Mecitözü. Euchaita acquired a reputation from the fourth century as the centre of the cult of St Theodore Tiro (‘the Recruit’) who was martyred under Maximinus (308–13). Two centuries later, Euchaita was made a city by Anastasius I (491–518), who gave it a wall and a bishop.2 Regardless of the final determination of Beyözü as Euchaita, the AAP sought to investigate a late Roman and middle Byzantine site of some note. However, if the argument that Beyözü/Avkat is in fact ancient Euchaita is correct, this presents a rare opportunity to examine archaeologically the changes that occurred when a late Roman town became a city. A similarly dated transformation from Roman village to city in the early sixth century is known from Dara in Mesopotamia and comparison with that settlement helps situate Beyözü/Avkat within its wider context of Late Antique urbanism.3

2. Extensive survey within the village Extensive survey showed that the village contained a very large volume of spolia, but no significant imported material (Fig. 2). Almost all of the datable inscriptions were from the fifth century or later. The survey also identified 18 early Byzantine tombstones, none of which gives information beyond names and occupations, such as hegoumenos, deacon or keroularios (candlemaker or someone who looked after the lighting in a church or similar 1

Fieldwork was conducted in 2007–2009 under a permit from the Ministry of Culture and Tourism. See Elton et al. 2009; 2012. 2 Mango and Ševčenko 1972. 3 Croke and Crow 1983.

102

Peter Bikoulis, Hugh Elton, John Haldon and Jim Newhard

Fig. 1: Location of Beyözü/Avkat and the survey area (©AAP 2014). estabishment). Unfortunately, no inscriptions mentioning a bishop were found. Among the spolia, the preserved architectural elements suggest a single church with a mixture of palmette and impost capitals; the diameters of the columns vary between 0.35m and 0.44m. Also found were two chancel screen posts and a screen fragment carved from local

Above as below: The application of multiple survey techniques

103

Fig. 2: Inscription, church spolia and architectural pieces (©AAP 2014). limestone. Another block reportedly found in this area was a fragment of a door jamb or lintel with two zones of decoration: one with interlocking semicircles, the other with a vegetal scroll of alternating grape clusters and trefoils.4 Other finds possibly related to a church

4

Cf. Christof and Koiner 2010.

104

Peter Bikoulis, Hugh Elton, John Haldon and Jim Newhard

Fig. 3: Exposed section of wall (F1726) (©AAP 2014). included fragments from a fifth- or sixth-century wineglass lamp, a pithos with an applied cross, three fragments of Proconnesian marble revetment and a piece from a small white marble column (diameter: 0.07m). Finally, the Anastasian inscription now in Mecitözü was described by several locals as having come from Beyözü. As a block 2.15m long with an unfinished back, it was well suited to being placed over a doorway as a lintel or as a block above a lintel.5 These church elements likely came from an area in the southern part of the village. This was identified during early reconnaissance as a possible site for a structure because the current field boundary incorporates part of a substantial wall. This wall had a facing of small limestone blocks over a cemented rubble core and interleaved courses of tiles. Although an accurate wall thickness could not be measured, the exposed section was over 0.5m thick and was over 3.5m in length, suggesting a substantial public building (Fig. 3). Local informants also related rumours of an illegally-excavated mosaic from this area of the village.

3. Intensive field survey The area within the village containing the exposed wall was prioritised for investigation. Reports from local informants also confirmed that this was an area highly likely to contain a feature. In addition, survey was facilitated by the fact that there were open fields on either side of the field boundary that made visibility possible and access easy. Examination of this area commenced in 2007 using a full range of non-distributive techniques that involved

5

Marek (forthcoming) suggests that the Abgar inscription that was in Gürcü in 1899, 35km west of Beyözü (SP 3.211), was originally from Euchaita. This stone has a previously unread line which mentions a bishop John, perhaps the same bishop John as that recorded in Beyözü (SP 3.227).

Above as below: The application of multiple survey techniques

105

Fig. 4: Densities of total counts of artefacts in the Avkat region, as measured by (a) field, (b) observation point and (c) kernel densities using a 15m cell size, 90m search radius and the Epanechnikov kernel function (©AAP 2014). intensive field walking, magnetometry and Ground Penetrating Radar (GPR); it recommenced in 2009 with very high-intensity surface collection (Fig. 1 Area 1, Fig. 4).6 The 6

A full description of the intensive survey method used can be found in Newhard et al. 2013.

106

Peter Bikoulis, Hugh Elton, John Haldon and Jim Newhard

three survey units overlying the exposed wall section (S0016, S1668 and S1669) were defined by modern field boundaries and roads. Within each survey unit, field walkers were spaced at 15m intervals and walked along a compass bearing determined by the orientation of the survey unit. Given the irregular shapes of modern fields and the likelihood that current field boundaries do not necessarily reflect ancient landscape partitioning, using such data from the survey units directly is highly problematic. In an attempt to bypass the limitations of using modern field boundaries to define survey units, field walkers also recorded artefact counts every 15m as they walked, the data being called an observation point. This created a rough 15m grid within each survey unit. Artefact densities could then be analysed at the level of both the survey unit and the observation point once data had been entered into the GIS. Use of the observation-point scale for artefact distributions permitted a finer level of analysis than simply using standard survey units. For example, the method allows us to distinguish between densities at the eastern and western ends of S1669 that might otherwise have been lost when considering data within a single field. In addition to being able to distinguish scatters within survey units, this technique also allows the identification of artefact clusters that cross survey units or are affected by landscape features such as irrigation ditches, roads, residential units, etc.

4. Remote-sensing survey The remote-sensing team, led by Stephen Wilkes and Meg Watters, used magnetometry and GPR to investigate areas with a high potential for feature identification within the area of the survey permit.7 Initial GPR data within the village at the site of the exposed wall and field boundary (W on Fig. 5) suggested the foundation of an east-west oriented structure that was initially interpreted as measuring 13.75m on its long axis and 10m north-south. However, further analysis suggested that this feature was more likely closer to 40m in length by about 20m in width, pointing to the presence of a rather large structure (Fig. 5). This foundation appeared to be curved at the eastern end, reminiscent of an apse. It was concluded that the foundations identified by the GPR were of a possible ecclesiastical structure. Immediately south of the wall in the field boundary was a surface concentration of brick and tile that appeared on the GPR as a high amplitude reflection. Although the magnetometry work covered some of this building, the magnetometry data and interpretations do not correspond to the features identified by the GPR and were of limited value in interpreting the plan of the structure. The high magnetic field area overlapping the eastern end of the structure should probably be interpreted as a concentration of material from the collapsed building, but it is surprising that it does not show up in the GPR data as a high amplitude reflector.

7

Magnetometry data was collected using a FM256 fluxgate gradiometer, while the GPR survey employed a SIR3000 GPR with both 200MHz and 400MHz antennas.

Above as below: The application of multiple survey techniques

107

Fig. 5: Structure extents based on GPR survey: (a) GPR imagery, (b) interpreted structures overlying GPR data (©AAP 2014).

5. The church The church identified by the subsurface survey appears to be similar in size and plan to other churches in central Anatolia. While these images do not present a complete picture, comparison to excavated churches in the region provides some indication of the likely appearance of the church. The three-aisled basilica plan was typical of most Anatolian churches built in the fifth and sixth centuries. Although churches are found in many sizes, there appear to be two main size groupings: small (ca 15m wide × ca 25m long) and medium (ca 20m wide × ca 40m long). At Hadrianopolis (Eskipazar), in Paphlagonia, two small three-aisled basilicas with nartheces have recently been investigated.8 One structure was located in the city itself, and measured 15m × 23.5m (Church B), while the second church, similar in plan, was positioned 2.5km to the east of the city and measured ca 15m × 20m (Church A). This second church had palmette capitals similar to those found at Beyözü and mosaic floors in the nave, aisles and apse. At Çiftlik, a rural site 10km from Sinop, another small three-aisled basilica was found measuring 17m × 28m.9 It was decorated with mosaic floors, marble

8 9

Lafli and Zäh 2008: 686–99. Hill 1995; 1998.

108

Peter Bikoulis, Hugh Elton, John Haldon and Jim Newhard

revetment on the walls and some wall or ceiling mosaics incorporating glass tesserae. This church was decorated with ‘late Ionic type’ column capitals, different from the types found at Beyözü. The Avkat church (ca 20m × 40m) was likely larger than all these structures, with the preserved diameters of the larger capitals suggesting a lower colonnade about 4m tall. It thus appears closer to the medium-sized churches. Examples of these include the late fifthcentury basilica at Amorium which measured 22m × 37m and was decorated with mosaic, marble and frescoes; at the pilgrimage site of Germia, 100km southwest of Ankara, the church of St Michael measured ca 18m × 41m.10 Finally, there were a few much larger churches, like the partially excavated example from Sebastopolis (Sulusaray), 95km to the southeast and close to modern Tokat.11 Here, the three-aisled basilica measured 35m in width. Its length is impossible to determine because of the presence of modern houses over the western end, but this was clearly a much larger church than anything found by the AAP. Like the Germia, Sebastopolis and Amorium churches, the Beyözü church was probably rebuilt during its period of use, which probably continued into the eleventh or twelfth century. When these rebuildings occurred at other sites, the churches were often reduced in size and were frequently limited to a single apse. These rebuilds tended to use many of the original materials from earlier phases, sometimes but not always with new decorative elements.12

6. ‘Super-intensive’ survey: method and results The intensive survey work was followed up using what we termed a ‘super-intensive’ survey in order to focus on areas of particular note. Our methodology was fashioned after a very high-intensity survey technique used by Andrew Poulter in the Balkans.13 Poulter’ s results suggest that surface artefacts such as roof tiles do not move very far from their point of origin, namely Roman villas and other large structures. Therefore, it was decided that a detailed plot of roof-tile distributions at some locations around Beyözü might be of some value in identifying the approximate locations of buried features. There is little archaeological literature on the relationship between surface collection and subsurface remains, despite the great importance that defining such a connection could potentially have for landscape archaeology and the interpretation of artefact distributions.14 For instance, Charles Redman and Patty Jo Watson note in relation to their research at the prehistoric sites of Çayönü and Girikihacıyan in southwestern Turkey that there is a strong correlation between the density of surface remains and subsurface features.15 Following this, Roger Matthews conducted systematic surface collection prior 10 11 12 13 14

Ivison 2010; Niewöhner and Rheidt 2010. Özcan 1991: 266–9; 1992: 171–2; Kohl 2010. Cf. Westphalen 1998. Poulter et al. 1998; Poulter 2007. For a recent study exploring the archaeological signature and relationship between roof collapse, floors and activity areas, see Friesem et al. 2014. 15 Redman and Watson 1970.

Above as below: The application of multiple survey techniques

109

Fig. 6: Super-intensive survey: illustration of methods (©AAP 2014). to the renewed excavations at Çatalhöyük in central Turkey for the purpose of comparing distributions to underlying architectural features.16 Other studies concentrating on the issue of artefact displacement within the plough zone, which is a major concern throughout the Mediterranean, show that there is great potential for the movement of artefacts both horizontally and vertically.17 Moreover, there seems to be a tendency within surface survey for an overrepresentation of larger artefacts, which some have called the ‘size effect’, as smaller objects are frequently missed because they become well sorted within the plough zone soil matrix.18 These studies would seemingly point to the strong possibility that surface remains may not correlate well with subsurface features. As previously noted, the results of the subsurface testing using GPR and magnetometry strongly confirmed the presence of an architectural feature, which presented a unique opportunity to explore this longstanding and important research question. While we were unable to replicate the exact method of Poulter, such as the use of the local ‘pergel’ as a measurement device, the objective of our research was to test the relationship between surface artefacts and subsurface architectural features. Poulter’s method employed three field walkers engaged in click survey, each recording different artefact categories: namely ceramics, brick/tile and architectural fragments. His collection strategy was 16

Matthews 1996. Unfortunately, a report on the latter aspect of the surface collection has yet to appear. 17 Ammerman 1985; Odell and Cowan 1987; Yorston et al. 1990; Dunnel and Simek 1995; Shott 1995. 18 Baker 1978.

110

Peter Bikoulis, Hugh Elton, John Haldon and Jim Newhard

Fig. 7: Super intensive survey areas (©AAP 2014). Area 1 2

Survey unit S0016, S1669 S2553–S2564 S1496 S2565–S2583

Number of Observation Points 40 1202 32 2071

Observation Point Interval (m) 15 1 15 1

Total Survey Unit area (m2) 5,308.44 1,202 8,879.05 2,071

Table 1: Comparison of ‘super-intensive’ survey units and regular intensive survey units. geared towards finding architectural features associated with villas or other large structures owned by landowning elites by focusing on the densities of ceramics and tile fragments. The method used at Avkat for the super-intensive survey was devised so that it would integrate with the AAP’s GIS and relational database structure.19 The methodology initially involved overlaying a 1m x 1m grid in survey units S0016 and S1669, the two survey units immediately overlying the structure identified by the subsurface testing (Fig. 7). Spaces between existing observation points were broken down further into an additional 15 subpoints of 1m each, resulting in each super-intensive survey unit measuring 15m in length 19

See also Newhard et al. 2013.

Above as below: The application of multiple survey techniques

111

Fig. 8: Artefacts collected during the super-intensive survey (©AAP 2014). (Table 1). A single field walker surveyed each of the new super-intensive survey units. The nature of the survey method allowed for a truly 100% coverage of the area within the superintensive observation points.20 Counts for all types of surface material, such as ceramics, roof tile and chipped stone, were recorded for each 1m × 1m square, but only diagnostic materials were collected (Fig. 8). A second location was selected for super-intensive survey in order to allow comparison with the results from the first area, where a subsurface structure was known to be located. The choice of this secondary area was based on the high density of roof tiles in an ‘off-site’ location, which was only identifiable based on an analysis of the counts of the observation points used by the intensive field survey. Area one, the locality above the possible church, produced a high concentration of ceramic material, including roof tiles (Table 2). Of the ten roof-tile fragments, one was over-fired, perhaps suggesting on-site production of the tiles. A building of 40m × 20m could have required as much as five metric tonnes of tile for the roof alone, in addition to any bricks or tiles used elsewhere in the construction. The pottery finds included a few

20

While many surveys state that an area received ‘100% coverage’, the reality is that much less of the ground is effectively surveyed. For discussion of effective coverage and artefact detection, see Banning et al. 2011.

112

Peter Bikoulis, Hugh Elton, John Haldon and Jim Newhard

Fig. 9: Identified and extracted architectural features (©AAP 2014). residual pieces of early Roman sigillata and two Roman oil lamps.21 The majority of the ceramics, however, were plain and cooking wares from the middle Byzantine period (tenth to twelfth century), though there was a quantity of late Roman and medieval (thirteenth to fifteenth centuries) sherds as well. Of particular interest are a rim fragment of Glazed White Ware II imported from Constantinople in the eleventh or twelfth century and a fragment of Unglazed White Ware probably also imported from Constantinople in the twelfth century. There were also some probable fragments of twelfth-century Ganos wine amphorae imported from the Aegean coast. In addition to the ceramic finds, a number of small finds of particular note, possibly relating to the subsurface structure, were recovered. Nearly 100 mosaic tesserae were found, many of which were blue, black and red in colour and averaged about 1cm3. A single

21

The final ceramic analysis was carried out by Joanita Vroom (University of Leuven).

113

Above as below: The application of multiple survey techniques Location

Area 1

Area 2

AAP total

Measure Count Density (count/m2) % of Total Count Density (count/m2) % of Total Count Density (count/m2) % of Total

Pottery 3,713 3.089 19.6 659 0.318 3.5 18,927 0.003 100.0

Artefact class Tile 5,823 4.844 22.0 7,610 3.675 28.7 26,517 0.004 100.0

Tesserae 91 0.076 100.00 0 0 0.0 91 0.000 100.0

Table 2: Comparison of ‘super-intensive’ artefact results from Area 1 and Area 2 to Avkat Archaeological Project totals. gold-glass example (0.125cm3) was found.22 In addition to these individual tesserae, several fragments of tesserae adhering to mortar were also discovered. When the AAP returned to this area in 2009, more local informants described the discovery of a fresco of a bearded man, a cross and snakes from this area. Unfortunately, it appears to have been inadvertently destroyed by being washed. The super-intensive survey also recovered a single Byzantine follis (sixth to ninth century) which supplements four other folles that had been given to the team by villagers in 2007 and were said to have also come from this area. Finally, six fragments of glass were found, three of which may have been bases of late Roman wineglass lamps. The comparison of surface artefact distribution with subsurface anomalies was pursued in a probabilistic fashion. Point patterns of different artefact categories were smoothed via kernel density function.23 The kernel density function assigns values to a raster cell using a ‘bandwidth’ or the algorithm used to sample neighbouring points around that cell. Within archaeology, kernel densities have been used as a predictive measure of intensity.24 Density maps for tile, ceramics, tesserae and all artefacts were then compared to variances in the GPR data to examine whether a discernible pattern or association was present. The anomalies associated with the church structure appeared 0.60m below the surface, but were most visible between 0.8–1.54m below the surface. However, the detection of architectural features by the GPR is based on interpretations of the primary data. Therefore, a direct comparison of the GPR values was not possible. Instead, these potential structures were extracted from the raw GPR data as vector polygons in order to compare them to surface artefact distributions (Fig. 9). It was hypothesised that if there was pronounced surface displacement of artefacts due to post-depositional processes (i.e. ploughing), then there would be a weak association between the compared variables. Based on the results of these comparisons, it appears that there is a minor positive correlation between surface scatters and underlying architectural features (Table 3). The comparison was made between only three artefact classes (ceramics, tiles and tesserae), and does 22 23 24

Cf. Gill 2002; Wypyski 2005. All spatial analysis was carried out in ArcGIS 10.2. See Baxter et al. 1997; Beardah 1999.

114

Peter Bikoulis, Hugh Elton, John Haldon and Jim Newhard Artefact class Tile Ceramics Tesserae Total artefacts

Correlation coefficient 0.25867 0.24543 0.17718 0.26715

Table 3: Correlation between artefact classes and subsurface anomalies in Area 1. not include others such as chipped stone or architectural fragments. Spatial analysis of the distribution of artefacts in Area 1 found that while concentrations can provide a reasonable suggestion of the presence of a subsurface feature, they do not provide a reliable means of locating it (Fig. 10). However, differences between the concentrations, as measured by correlation with potential subsurface features, are very noticeable. While all of the artefacts have weak positive correlations, the distribution of tiles is marginally more correlated with the theorised subsurface structures than the other single artefact classes, including ceramics. This may be explained by the action of tillage, which other studies have found to affect greatly the spread of ubiquitous artefact classes like ceramics, or it may be that ceramic concentrations are not related to structures at all; instead, they may indicate different aspects of human activity in the landscape. Roof tiles do not appear to move as greatly, likely owing to their heavy concentration after collapse. Likewise, tesserae, which would have been used in abundance on floor and wall surfaces, likely concrete heavily in the immediate vicinity of their deposition and are only moved with pronounced activity. An alternative explanation may be that ploughing disturbs underlying deposits of both tiles and tesserae, continuingly bringing new materials to the surface with minimal horizontal displacement. Thus it would seem that Poulter’s method of focusing on roof tiles can be relied upon to provide a coarse means of locating architectural features in the landscape, although further investigations are clearly needed. However, when combined with other methods, such as GPR, their location can be better determined as there is not a direct correlation between what is above and what is below. Additionally, recovered artefacts that are associated with building materials, for example the small lumps of mortar with adhering mosaic tesserae, further support the assumption that a structure of some sort has been identified. The low quantity of small artefacts like tesserae found in the first pass suggests that this artefact class is not easily spotted in regular pedestrian survey. However, it is weakly correlated with the possible subsurface architectural features, and therefore may be a minor indicator of a subsurface feature or structure. Thus, a mixed method approach that combines differing artefact classes has the potential to identify subsurface features better than simply relying on ceramic concentrations.

7. Conclusion The combination of surface survey techniques used to investigate this area within the village of Beyözü was highly effective, allowing the confident identification of the feature located by the GPR and intensive surveys as a church. A richer level of detail was added using the super-intensive method, which also allowed for the testing of a long-standing

Above as below: The application of multiple survey techniques

115

Fig. 10: Density of roof tile in Area 1, as measured by (a) observation point and (b) kernel densities (©AAP 2014). assumption in archaeological survey concerning the relationship between surface scatters and subsurface architectural features. The results of spatial analysis of roof tile and other artefact classes in Area 1 showed that there was a weak positive correlation between the

116

Peter Bikoulis, Hugh Elton, John Haldon and Jim Newhard

surface distribution of artefacts and the subsurface presence of building foundations. The results from the AAP super-intensive survey suggest that surface concentrations may be able to suggest the possibility that a structure resides beneath the surface, but they cannot be relied upon to provide the exact location. While the foundation appears to be of a late Roman church dating to the fifth or sixth century, the ceramics collected by the highintensity method revealed a different picture. There appears to be continued occupation of the site from the early Roman period through to the Ottoman and modern periods. Like the coins, the majority of the ceramics are dated to the middle Byzantine period. Nonetheless, these various techniques are unable to elucidate the details of the building’s architectural phases; these can only be more thoroughly investigated through excavation. The authors would like to thank the British Institute at Ankara, College of Charleston, Princeton University and Trent University for various types of support, both financial and non-financial. The work of the various team members, especially that of Zach Knox, was critical to the project’s success. Two collaborative research grants from the Trent University Archaeology Centre (TUARC) to Hugh Elton and Peter Bikoulis supported fieldwork in 2008 and 2009, and are gratefully acknowledged. Peter Bikoulis would like to thank Paul Duffy (Toronto) for discussions about the topics considered in this paper. Peter Bikoulis University of Toronto Department of Anthropology 19 Russell Street, Toronto, Ontario M5S 2S2, Canada [email protected] Hugh Elton Trent University Department of Ancient History and Classics 1600 West Bank Drive, Peterborough, Ontario, K9J 7B8, Canada [email protected] John Haldon Princeton University Department of History 129 Dickinson Hall, Princeton NJ 08544-1017, USA [email protected] Jim Newhard College of Charleston Department of Classics 66 George Street, Charleston, SC 29424, USA [email protected]

Bibliography Ammerman, A. 1985. ‘Plow-zone experiments in Calabria, Italy’, Journal of Field Archaeology 12.1, 33– 40. Baker, C. M. 1978. ‘The Size Effect: An Explanation of Variability in Surface Artifact Assemblage’, American Antiquity 43.2, 288–93. Banning, E. B., A. L. Hawkins and S. T. Stewart 2011. ‘Sweep widths and the detection of artifacts in archaeological survey’, Journal of Archaeological Science 38.12, 3447–58. Baxter, M. J., C. C. Beardah and R. V. S. Wright 1997. ‘Some Archaeological Applications of Kernel Density Estimates’, Journal of Archaeological Science 24.4, 347–54. Beardah, C. C. 1999. ‘Uses of Multivariate Kernel Density Estimates in Archaeology in the Age of the Internet’ in L. Dingwall, S. Exon, V. Gaffney, S. Laflin and M. Leusen (eds), CAA97.

Above as below: The application of multiple survey techniques

117

Computer Applications and Quantitative Methods in Archaeology. Proceedings of the 25th Anniversary. Oxford, 107.5–12. Christof, E. and G. Koiner 2010. ‘Ein kaiserzeitlicher Rankenfries und früh- bis mittelbyzantinische liturgische Ausstattungsteile aus Tavium’, Istanbuler Mitteilungen 60, 339–72. Croke, B. and J. Crow 1983. ‘Procopius and Dara’, Journal of Roman Studies 73, 143–59. Dunnel, R. and J. Simek 1995. ‘Artifact size and plowzone processes’, Journal of Field Archaeology 22.3, 305–19. Elton, H., J. Haldon, J. Newhard and S. Lockwood 2009. ‘Avkat Archaeological Project 2007– 2008’, Araştırma Sonuçları Toplantısı 27.3, 29–51. Elton, H., J. Haldon, J. Newhard and P. Bikoulis 2012. ‘Avkat Arkeoloji Projesi, 2007–2009’, Çorum Kazı ve Araştırmalar Sempozyum 2, 203–17. Friesem, D. E., G. Tsartsidou, P. Karkanas and R. Shahack-Gross 2014. ‘Where are the roofs? A geo-ethnoarchaeological study of mud brick structures and their collapse processes, focusing on the identification of roofs’, Archaeological and Anthropological Sciences 6.1, 73–92. Gill, M. A. V. 2002. Amorium Reports, Finds I: The Glass (1987–1997) (British Archaeological Report International Series 1070). Oxford. Hill, S. 1995. ‘The First Season of Rescue Excavation at Çiftlik (Sinop)’, Anatolian Studies 45, 219–32. Hill, S. 1998. ‘Rescue excavations at Çiftlik (Sinop)’ in R. J. Matthews (ed.), Ancient Anatolia. Fifty Years’ Work by the British Institute of Archaeology at Ankara. London, 285–301. Ivison, E. 2010. ‘Kirche und Religiöses Leben im byzantinischen Amorium’ in F. Daim and J. Drauschke (eds), Byzanz – das Römerreich im Mittelalter 2. Mainz, 309–43. Kohl, M. 2010. ‘Sulusaray/Sebastopolis-Heracleopolis: pour une reprise des recherches’, Anatolia Antiqua 18, 89–96. Lafli, E. and A. Zäh 2008. ‘Archäologische Forschungen im byzantinischen Hadrianupolis in Paphlagonien’, Byzantinische Zeitschrift 102, 681–713. Mango, C. and I. Ševčenko 1972. ‘Three Inscriptions of the Reign of Anastasius I and Constantine V’, Byzantinische Zeitschrift 65, 379–93. Marek, C. forthcoming. ‘Epigraphische Forschungen in Bithynien, Galatien, Paphlagonien und Pontos’, Chiron 45. Matthews, R. J. 1996. ‘Systematic Surface Collection’ in I. Hodder (ed.), On the surface: Çatalhöyük 1993–95. London, 73–7. Newhard, J. M. L., N. S. Levine, A. D. Phebus, S. Craft and J. D. Littlefield 2013. ‘A geoinformatic approach to the collection of archaeological survey data’, Cartography and Geographic Information Science 40.1, 3–17. Niewöhner, P. and K. Rheidt 2010. ‘Die Michaelskirche in Germia (Galatien, Türkei)’, Archäologischer Anzeiger 2010.1, 137–60. Odell, G. and F. Cowan 1987. ‘Estimating Tillage Effects on Artifact Distributions’, American Antiquity 52.3, 456–84. Özcan, B. 1991. ‘Sulusaray-Sebastapolis Antik Kenti’, Müze Kurtarma Kazıları Semineri 1, 261–307. Özcan, B. 1992. ‘Sulusaray-1990 Kurtarma Kazısı’, Müze Kurtarma Kazıları Semineri 2, 167–200. Poulter, A. G. 2007. ‘Site-Specific Field Survey: the Methodology’, Proceedings of the British Academy 141, 583–95. Poulter A. G., M. Beckmann and P. Strange 1998. ‘Field Survey at Louloudies: A New Late Roman Fortification in Pieria’, The Annual of the British School at Athens 93, 463–511. Redman, C. L. and P. J. Watson 1970. ‘Systematic, intensive surface collection’, American Antiquity 35.3, 279–91. Shott, M. J. 1995. ‘Reliability of archaeological records on cultivated surfaces: a Michigan case study’, Journal of Field Archaeology 22.4, 475–90. Westphalen, S. 1998. ‘Die Basilika von Priene. Architektur und liturgische Ausstattung’, Istanbuler Mitteilungen 48, 279–340. Wypyski, M. T. 2005. ‘Technical Analysis of Glass Mosaic Tesserae from Amorium’, Dumbarton Oaks Papers 59, 183–92. Yorston, R., V. Gaffney and P. Reynolds 1990. ‘Simulation of artefact movement due to cultivation’, Journal of Archaeological Science 17.1, 67–83.

Max Ritter

The end of late antiquity in Paphlagonia: disurbanisation from a comparative perspective

1. Introduction There have been many attempts, past and present, to formulate a history of Anatolian cities in late antiquity, and their fate and demise during the sixth and seventh centuries.1 As Ine Jacobs puts it, ‘our knowledge about the Late Antique Anatolian city is fairly advanced; there are still many uncertainties – because of the focus of early excavators on the classical city; the focus on large urban centres like Ephesos, Pergamon, Aphrodisias; and the focus on the west and south of Anatolia’.2 Due to this distorted focus it is not feasible to deduce from the better-known cities of Asia Minor to our particular region; regional evidence is needed, especially because settlement patterns and the degree of urbanisation in Paphlagonia have always been different to those of the Aegean shore. The relatively late foundation dates of the cities in the interior of Anatolia also had an impact on the degree of urbanisation during late antiquity, and both the density and size of the cities in Paphlagonia were inferior compared to those of western Asia Minor (Fig. 1). For analytical convenience, the region of Paphlagonia is equated here approximately with the Late Antique provinces of Honorias, Helenopontus and Paphlagonia itself. Inner Paphlagonia has been characterised as an ‘eternal periphery’, although it belonged to the core of the empire in late antiquity. It held relevance mostly as a transit point for the army between the urban centres of the west and the frontier in Armenia. Paphlagonia not being a theatre of war until 600 adds to the importance of research concerning the cities in the area, because we can discount the influence of conflict, which is hardly measurable. In most regions of the empire the effects of war were apparently the motor of development, whether they were the negative effects of incursions or the positive effects of economic stimulation bought about by a military presence.3 Furthermore, Paphlagonia is well-suited for comparative settlement studies because it is considered to be averagely urbanised during both the Late Antique and Byzantine periods. 1

For a summary of research history, see Saradi 2006; also important is Chavarria and Lewit 2004: 3–51. 2 Jacobs 2011. 3 An instructive example for these phenomena are the sites in Moesia Inferior collected by Dunn 1994.

120

Max Ritter

Fig. 1: Map of the administrative units and cities of Paphlagonia during late antiquity (fourth to seventh century) (cartography by R. Szydlak). A short overview of recent projects in the region may help to compare the Late Antique data. My intention here is to give a general sense of the state of current research as it relates to questions discussed in this paper, in order to evaluate hypotheses discussed in Late Antique archaeology in general.

2. Project Paphlagonia Project Paphlagonia was largely conducted in the vilayet of Çankırı and combined both extensive and intensive survey methods.4 The project director, Roger Matthews, stipulates that the proportion of imported pottery is very small in the survey region, which indicates that inner Paphlagonia was not connected to the Mediterranean trading system; even Sinopean and Heraclian amphorae are very rare. William Anderson has recently interpreted this as evidence of a general state of autarky in the Paphlagonian interior during late antiquity.5 Unfortunately, these surveys have not contributed very much to the understanding of Late Antique cities in Paphlagonia, since the survey region does not contain urban centres (Çankırı/Gangra itself was not surveyed, of course, due to continued habitation) and Byzantine reference material was not consulted or available at the time. Information about Paphlagonia’s regional pottery production (which accounts for almost all the Late Antique pottery found) is still sparse due to the lack of stratified pottery from defined 4 5

Matthews et al. 2009: 190–7. Anderson 2011: 116–9.

The end of late antiquity in Paphlagonia

121

contexts which is required for the production of a body of comparative material, and which is desperately needed for use in intensive surveys.6

3. The Cide Archaeological Project The Cide Archaeological Project, directed by Bleda S. Düring, Claudia Glatz and Tevfk E. Şerfoğlu, was originally envisaged for the Devrez valley, but for bureaucratic reasons ended up, rather by chance, on the coast between Kromna and Kap Karambis, including the inland district of Şenpazar.7 Similar to Project Paphlagonia, it was a diachronic survey comprising all periods up to the Ottoman period. Unfortunately, next to no diagnostic pottery (e.g. neither Pontic Sigillata nor Pontic Red Slip Ware) or architecture was found that could be assigned to late antiquity or the Dark Ages/late seventh–ninth centuries.8 Although the survey area includes the ancient settlements of Kytoros and Aigialos (today’s Cide), the scarcity of finds makes it impossible to identify well-defined urban and non-urban zones, or, furthermore, to delineate a picture of these settlements in late antiquity. As a result, neither this project nor Project Paphlagonia are able to provide the material necessary to address questions of changes in Late Antique urbanisation; rather, they draw a vague picture of a remote and rural Paphlagonia in both the interior and at the coast.

4. Hadrianopolis Hadrianopolis was one of the minor cities in the province of Honorias, but, to our advantage, it was surveyed and partially excavated in the 2000s, and field research is still ongoing.9 Hadrianopolis was founded late (as Kaisareis Proseilemmeneitai in the beginning of the first century),10 had a very modest start and a very shadowy development during the High Imperial period/first–fourth centuries; it only rose to some significance in late antiquity, when it was walled, roughly encompassing 20–40ha (unfortunately, only few traces of the walls remain, and not enough to offer a more precise estimate of area). Two mediumsized, but well-furnished churches have been excavated; one is extramural (basilica A: second half of the fifth century with mosaics of the sixth century).11 The supposed episcopal basilica B is apparently part of a large ecclesiastical complex which was detected by geo-radar in 2013 and also contains a mid-Byzantine cross-in-square church.12 Although the seventh-century Vita of St Alypios the Stylite places his column in the necropolis outside the city,13 it now seems possible to locate the supposed pilgrimage complex of his cult in the area of basilica B. In any case, neither basilica A nor B can be connected to the foundation of his veneration, because both were built in the fifth/sixth century. The excavations 6 7 8 9 10 11 12 13

As Anderson comments, with so few diagnostic sherds found during the survey work, interpretation is highly biased in every regard: Anderson 2011: 99–102. Düring and Glatz 2014. Most of the surveyed fortifications, especially those with cribwork, belong rather to the twelft– fourteenth century. See also Bes this volume. Lafli and Zäh 2008; Kilavuz and Çelkbaş 2013. Marek 1993: 118–121. Kilavuz and Çelkbaş 2013: 163–4. Ina Eichner, personal communication. Vita Alypii stylitae Hadrianopolitani prior (BHG 64).

122

Max Ritter

at basilica B revealed Glazed White Ware I, demonstrating continued occupation of the site until at least the eighth century.14 Furthermore, the city contains two excavated baths, one of them altered to a building with an economic function (storehouse?) in the sixth century. A prominent feature of the city is its well-populated and fertile hinterland with many hamlets, which flourished in late antiquity.15 The finds show a clear tendency: the High Imperial period has not left many traces in the archaeological record and imported pottery is very scarce as a whole.16 In the case of Hadrianopolis, these facts point to a rise of the city in terms of size and endowment only in the Late Antique period. The coin finds, which are about to be published,17 reflect a pattern encountered in other cities of Paphlagonia: the sixth century does not provide much; the beginning of the seventh is overrepresented, seeing a decrease after Constans II, and the latest coin in Hadrianopolis is a miliaresion of Leo III found in basilica A. Since no finds dating from the later eighth to the second half of the tenth century have been recorded, it is impossible to say what happened in Hadrianopolis during this period, but a mid-Byzantine recovery did take place in the later tenth century when parts of the town were reoccupied and the Kimistene hill nearby was used as a watchpost.18

5. Tios/Teios The polis and bishopric of Tios/Teios was also located in Honorias, and has been subject to excavation since 2006.19 Fieldwork has focused on the acropolis of the city, where a midByzantine church within a fortification wall with several towers has been uncovered.20 Clear signs of Late Antique activity have not yet been recorded, but research continues and will hopefully reveal more. A droungarios Theodore is testified by a lead seal, and this might indicate that the city’s harbour (now mostly underwater) served as a minor naval base in the eighth century.21 The location of the city at the mouth of the Billaios certainly promoted its distributional role to the interior. Although the Late Antique profile of the city is not clear, a tenth/eleventh-century recovery similar to that of Hadrianopolis took place on the acropolis and stretched into the early fourteenth century as the city remained a Byzantine outpost until that time.22

14 15 16 17 18 19 20 21 22

Lafli and Kan Şahn 2012. Lafli and Zäh 2008. Lafli and Kan Şahn 2012. Lafli et al. forthcoming. Lafli and Christof 2011. Atasoy 2008; Sönmez and Öztürk 2008. Anderson 2009. Nesbitt and Oikonomides 2009: 24, no. 10.3. Belke 1996: 276; Kortanoğlu 2013.

The end of late antiquity in Paphlagonia

123

6. Amastris Although Amastris is certainly one of the major cities of Paphlagonia, a thorough excavation has not been undertaken yet.23 Apart from data recovered from rescue excavations, only parts of the Late Antique city walls are known, mentioned by earlier travellers.24 According to the written sources, the city and its harbour were inferior to Sinope in the High Imperial period, and the city rose to prominence only in the Dark Ages25 when the Black Sea became a theatre of war again in the eighth century and a military wharf with a lighthouse was established there, overseen by a katepano.26 The city was strongly walled (in the eighth and early ninth centuries),27 encompassing around 10ha, but a major proportion of the city’s population lived outside the walls, as related by the Vita of St George of Amastris.28 Furthermore, two prominent churches of the ninth century are still standing (Fâtih camii and Kilise mescidi).29 The urban continuity of Amastris during the Dark Ages can be deduced from the imperial investment made because of its military significance.30

7. Pompeiopolis Pompeiopolis was definitely the second-most important city of the province of Paphlagonia in late antiquity. Its Late Antique walls have left their profile in the landscape and encompass an area of around 32ha. To put this figure into perspective: in Late Antique terms it is appropriate to consider a settlement a city at about 10ha;31 Amorion in Phrygia was 50ha in size (built at the end of fifth century), but had a superior significance as a military base.32 Based on the area of the city, we can assume a population within the walls of Pompeiopolis of about 5,000 to 8,000 inhabitants.33 To date, a large third-century domus has been revealed, the Late Antique period of which is most prominent in the record with many phases and room alterations.34 Two churches of the sixth century are also known, but they can only be roughly located. The finds reflect the tendencies already noted in relation to Hadrianopolis. The proportion of imported table ware from the Mediterranean is low, but the sixth century provides us with a slightly larger amount (e.g. African Red Slip Ware). There are amphorae, but, surprisingly, almost none from Heraclea and Sinope; instead Levantine types make up the lion’s share. Obviously, the well-to-do-citizens of the city consumed Levantine wine well into the seventh century. Some Glazed White Ware I testify to the city being inhabited at that time. 23 24 25 26 27 28 29 30 31 32 33 34

Eyce 1965; Marek 1989. Crow and Hill 1995. Sharp 2010: 53–87. Ahrweiler 1966: 110. Crow and Hill 1995. Vita Georgii Amastridensis (BHG 668): caps 24–5 (Vasilievsky 39): recording an Arab attack of 797. Belke 1996: 168; Sharp 2010: 114. Sharp 2010: 87 (fortification), 103 (churches). Blanton 2000: 64–86. Ivison 2007; Lightfoot 2012. Population density of approximately 100–200 inhabitants per hectare: Lo Cascio 2000. Musso et al. 2011.

124

Max Ritter

Fig. 2: Map of the via militaris to the Armenian frontier via the Amnias valley with Pompeiopolis, Kız Kalesi, and the visible road sections highlighted (cartography by R. Szydlak). The same observation can be made regarding the coin finds. The fourth century is welldocumented, while the fifth, especially its second half, is underrepresented. But this pattern is common across the empire and is founded in the minting policy and problems of preservation.35 As for many Anatolian sites, there is evidence of some sixth-century activity, but a good portion of the coins belongs to the beginning of the seventh century. Contrary to the situation in Hadrianopolis, the middle and the end of the seventh century are also wellrepresented. Apparently, owing to its location on the via militaris from Nicomedia to Satala (and later to the aplēkton Dazimon), the city was a military centre against the Arab advance until the beginning of the eighth century (Fig. 2).36 Sometime in the eighth century, the city was obviously abandoned, as was Hadrianopolis. It is as yet unknown where the population went, but a shift to Kız Kalesi (Fig. 3), as suggested by James Crow 37 and Klaus Belke,38 can be excluded, because this major fortification 6km east of Pompeiopolis was built only at the beginning of the ninth century39 and, according to its size, served as a 35 36 37 38 39

Guest 2012. Avramea 2002: 74–7. Bryer and Winfield 1985: 73 n. 39; Crow 2009: 30. Belke 1996: 261. Kız Kalesi has two phases, and the first compares perfectly to Kütahya castle period 1 (dating to the first half of the ninth century) due to its use of Christian spoils and the way it integrates towers in its defence: Foss 1985. Furthermore, there are strong similarities to the ‘decorative spolia’ phase of Amastris’ fortifications which can be dated approximately to the same period: cf. Sharp 2010: 87 who is more convincing on this point than Crow and Hill 1995, who prefer a mid-eighthcentury date.

125

The end of late antiquity in Paphlagonia

Fig. 3: Kız Kalesi (author’s photo). garrison point on the odos tōn Armeniakon for a corps of around 300 men, which is the size of a Byzantine bandon. It was not a true settlement, and is definitely too late to be considered a settlement substitute for Pompeiopolis. It is therefore a still open question as to what exactly happened in the Amnias valley at the beginning of the eighth century when Pompeiopolis was abandoned step by step and also later when it witnessed no tenth/ eleventh-century recovery, as is usually observed at Late Antique urban centres in Asia Minor.

8. Other cities of the region: Gangra, Sinope, Neoklaudiopolis and Sora For the remaining urban centres of Paphlagonia only a very little information regarding late antiquity is currently available. Gangra-Germanikopolis, as the provincial capital of Paphlagonia, and Ionopolis have been inhabited continuously and thus fieldwork is impossible; not even the city walls have left a trace. The Late Antique walls of Sinope and the architectural remains collected in the local museum testify to a flourishing city during late antiquity and in the middle Byzantine period, but a precise analysis of its urban landscape

126

Max Ritter

during this period was not yet been undertaken.40 At Çiftlik, a little south of Sinop, a fifthcentury basilica was excavated in the 1990s41 and at Gerna Dere to the west of Sinop a Late Antique church has been identified.42 The survey and geophysical investigations of a Danish research project, Where East meets West, have revealed a significant cross-shaped extramural martyrium at Neoklaudiopolis with a lot of surface finds of Late Antique pottery (fifth/sixth century).43 Of the other two poleis mentioned in Hierokles’ Synekdēmos,44 only Sora is plausibly located, near Karabük, while the identification of Dadybra is highly controversial.45 It is impossible to say anything regarding their Late Antique significance, yet, but their elevation to polis status and bishoprics in the period prior to the Council of Chalcedon (sometime in the early fifth century) may be interpreted as a sign of the gradual urbanisation of former villages in order, apparently, to integrate their respective regions into the sphere of Church organisation and control.

9. Development of the cities and the watershed of the sixth century Across all parts of Anatolia, major building activity occurred at the beginning of the Late Antique period between 350 and 400; mostly, city-walls,46 churches and, later, martyria/ pilgrimage churches were built during this period. In Paphlagonia, we can refer to the city walls of Hadrianopolis, Amastris and Pompeiopolis, the churches in Gerna Dere and Çiftlik,47 and the martyrium in Neoklaudiopolis to demonstrate that this was indeed the pattern in Paphlagonia. These new walls do not signify a reduction of the city or its population; on the contrary, earlier fortification projects were envisaged with a projected larger population in mind, while Late Antique walls were built in order to achieve maximum defensive efficiency;48 they are also a sign of vitality, of course. For the second half of the fifth century, we do not yet have evidence of defined building activities in the public sphere.49 The Justinianic programme was very arbitrary in Anatolia: some cities were well-endowed, like Ephesos, Germia, Mokissos, others saw no investment at all, and it seems that Pompeiopolis and the other Paphlagonian cities belong to the second group, as only churches are currently testified for this period. In the second half of the sixth century in particular, the cityscape quickly deteriorated: kilns were built in former public areas; the street grid was progressively abandoned; and habitation density increased.50 Although these changes altered the layout of the Late Antique cities, they also 40 41 42 43 44 45 46 47 48 49 50

Cursory remarks can be found in Doonan 2004; Redford 2011. Hill 1998. Doonan and Smart 2000/2001. Bekker-Nielsen 2013; Winther-Jacobsen this volume. Hierokles, Synekdemos 694.3–696.3. This is not the place to discuss this problem in full. From my point of view, identification with Devrek is incorrect (Belke 1996: 186), while Safranbolu and İskilip are also unsuitable candidates. Niewöhner 2007: 249–52. See also Bikoulis et al. this volume. Kirilov 2007: 3–20. Jacobs postulates there was no deliberate policy of urban renewal at all: Jacobs 2012: 150. Kennedy 1985.

The end of late antiquity in Paphlagonia

127

indicate economic activity.51 Furthermore, it is not appropriate to measure the wealth of a city by the monumentality of its public buildings.52 After the mid sixth century, only minor repairs were undertaken on the most important buildings, and no new building projects were initiated. The only exception to this overall picture is the St Alypios complex at Hadrianopolis, which relates to an important saint who died in the first quarter of the seventh century. As far as Paphlagonia and many other Anatolian regions are concerned (e.g. Lycia, Milesia and Lycaonia),53 late antiquity was a flourishing period for the countryside – many hamlets developed in peaceful Anatolia, to an extent unseen in the High Imperial period.54 Therefore, it is now necessary to focus investigations on the hinterland, since its share of the prosperity of the cities was larger than ever before. In the sixth century, trade volumes were shrinking and trade was regionalised. As Helen Saradi stresses: ‘The network of trade routes and markets was slowly falling apart. This had an impact on all the regions of the empire, since they were interdependent and their interdependency rested on trade’.55 This process affected the coastal regions much more, and we can see that the low-level prosperity which we have in interior Paphlagonia during late antiquity was not shaken in the sixth century. Inner Paphlagonia was neither previously nor later an integral part of the Mediterranean economic system; for Sinope on the coast, however, Owen Doonan conclusively draws the opposite conclusion.56 In sum, the overall situation in Paphlagonia is much the same as for elsewhere in Anatolia, with settlement levels peaking in the early sixth century, followed by site contraction, regression and abandonment. At the same time, there are signs of gradual change rather than sudden collapse (it is, therefore, fruitless to look for disasters in the written sources in order to find a momentum of urban collapse).57 Only a few selected sites were able to stand their ground, such as Amastris. These settlements were definitely not destroyed by warfare, but what exactly happened to Pompeiopolis and Hadrianopolis still remains a mystery, as it does for Sagalassos.58 From our current point of view, three possibilities come to mind: firstly, there was a process of abandonment in favour of smaller upland settlements; secondly, there was state-forced relocation of the population in order to save them from enslavement; and, lastly, there was transurbanisation to a site yet to be determined. For instance, Kolossai’s population went to Chonai and that of Prymnessos went to Akroinos (both in Phrygia); the inhabitants of Phaustinopolis moved to Loulon

51 52 53 54 55 56 57 58

Mundell Mango 2006: 90; Jacobs 2009. Mundell Mango 2011: 239–40. For Lycia, see Kolb et al. 1991; for Milesia, see Lohmann 1999; for Lycaonia, see Baird 2004. Saradi 2006: 33; Dunn stresses the importance of the rural-urban interdependency in late antiquity: Dunn 1994: 71. Saradi 2006: 43. Doonan 2004. This is a method which Clive Foss made popular in Anatolian archaeology, e.g. in Foss 1975. According to the excavators, a severe decline of the city can be reconstructed already in the middle of the sixth century, and next to nothing can be said regarding the period thereafter: Vanhaverbeke et al. 2007.

128

Max Ritter

(Lycaonia); eventually Arta replaced Ambrakia (Epirus); and I find it feasible that Komana Pontika was given up in favour of Dazimon-Dokeia/Tokat (Pontus).59 The archaeological evidence collected so far in Paphlagonia exhibits roughly the same pattern found in other parts of Anatolia, especially in the west and south. Aside from this general picture, differences are apparent already in the sixth century due to a lack of imperial or aristocratic investment, resulting in a faster deterioration of the cityscapes and long-distance-trade. As it seems now, the turning point of urbanity in Paphlagonia is the later part of the century, sometime before the onset of the Persian onslaught. Of course, the sixth century was a difficult period for the empire in general.60 Pestilence struck every 15 years after its first outbreak in 542,61 and climate change occurred especially after 536, when the eruption of a still unknown tropical volcano darkened the sky for more than a year, leading to crop failures until the late 540s.62 But these events cannot have caused a structural change in urban culture across the whole empire, because their effects must have varied from one region to another and, secondly, populations usually recover from disasters after a few decades.63 To illustrate the problem for the region we can consider Sinope’s olive cultivation, which could no longer be sustained under the new climatic conditions;64 but this can only have had a lasting effect on this city’s hinterland and not across the entire interior of Paphlagonia where economic conditions, like crop cultivation and distribution habits, had always been very different (and remained stable until the eleventh century).65 In other words, a structural reason needs to be identified, one that affected almost every city in the empire and which, in turn, is the only way to explain the structural changes in the settlement patterns.66 Hopefully fieldwork in Paphlagonia, a region well-suited to finding appropriate answers (due to its position at the centre of the empire, being averagely urbanised and unaffected by the influences of conflict), will help to tackle this question in the future. Properly stratified excavations in urban centres in combination with extensive surveys in the hinterland will be especially fruitful in this respect, as surveys produce material which can be analysed and checked against stratified assemblages of the same region, be it pottery, tiles or architectural sculpture.

59

60 61 62 63 64 65 66

Interestingly, Dokeia-Dazimon was not a bishopric, while Komana always remained one; the same parallel is to be found for Kastamon and Pompeiopolis; for Komana’s role, cf. Ercyas and Sökmen 2010. Meier 2003. Stathakopoulos 2007: 101–5. Steffensen 2008. Saradi 2006: 40. Doonan 2004. Izdebski 2012: 63–4. Concepts like that of ‘collateral damage’ (Cassis 2014: 307) afflicting those cities far away from war and turmoil are not convincing.

The end of late antiquity in Paphlagonia

129

10. The Sasanian problem As a final remark, I would like to point to a particular issue concerning the influence of the Persians in Paphlagonia at the beginning of the seventh century. This is of particular interest as we cannot resort to written source material for the most part and have to turn to archaeological evidence instead in order to understand the impact the war had in northern Anatolia.67 Clive Foss’ theory that the ‘end of antiquity in Anatolia’68 came with the Persian onslaught has not been supported by most Byzantinists. Nonetheless, the numismatic data from Paphlagonia may be helpful in re-evaluating this thesis. At our disposal we have assemblages collected during fieldwork at Pompeiopolis and Hadrianopolis; as well as the museum collections of Amasra,69 Amasya70 and Sinop,71 each of which stem from their respective region (vilayet). Although there are quite a few differences in their chronological patterns, all the assemblages feature one common trait: a lacuna of coins minted in the decade from 616 to 627. In other words, during this decade the coin supply to the region was most likely disrupted. According to the written sources the Sasanian forces entered eastern Anatolia in 607/8 when they conquered Theodosiopolis; they pillaged Kaisareia in 611 and Melitene in 613. A small force was able to reach the Bosphorus in 616 and the important military base of Ankyra was conquered and destroyed in 620 or 622.72 Most importantly, according to the sources, the Sasanians set up camp somewhere in the Pontus region during the winter of 621/2.73 Only after the failure of the Sasanian siege of Constantinople in 626, was Emperor Heraclius able to turn the tide. The numismatic gap indicates that the Persians were masters of the region during the decade from 616 to 627, which affected a rupture between Paphlagonia and the capital. Interestingly enough, despite this observation, when we look at the archaeological record, the war apparently had no effect at all on urbanisation in Paphlagonia. Both Hadrianopolis and Pompeiopolis continued to be inhabited on the same scale as previously, and may even have profited from an increased importance and a supposed influx of refugees arriving from the insecure regions of the empire during the seventh century. Although the evidence for Amastris and Sinope is less developed, it is clear that their rise in the eighth century, which is testified by their elaborate fortifications and the written sources, must also have a foundation in the preceding period. In any case, the process of desurbanisation began in the sixth century in Paphlagonia due to a multitude of structural reasons, and the Sasanians did not play a major role in that. Only after the Arab advances reached the region at the very beginning of the eighth century

67 68 69 70 71 72 73

For a recent and important attempt to do this, see Drauschke 2013. Foss 1975. Ireland 1996. Ireland 2000. Casey 2010. Summarised in Howard-Johnston 1999; see also Polacek 2008. Georgii Pisidae Expeditio Persica 2.256–60 (Pertusi 109); Foss 1975: 724–6.

130

Max Ritter

were the urban centres abandoned (in the case of Hadrianopolis, Pompeiopolis) or received imperial protection in the form of strong fortifications (Amastris, Sinope). In a way, Late Antique urbanitas ended only at this time in this core part of the empire. I would like to thank Christian Marek, Lâtife Summerer, Kristina Winther-Jacobsen and Ina Eichner for sharing their knowledge and recent research results with me, and for encouraging my own attempts to make sense of the Paphlagonian archaeological data. Max Ritter Johannes-Gutenberg-Universität Mainz Historisches Seminar Jakob-Welder-Weg 18, 55099 Mainz. Germany [email protected]

Bibliography Sources Georgii Pisidae Expeditio Persica / A. Pertusi (ed./trans.), Giorgio di Pisidia, poemi 1: panegirici epici (Studia patristica et byzantina 7). Ettal 1959. Hieroclis Synecdemus / E. Honigmann (ed.), Le synekdèmos d’Hiéroklès et l’opuscule géographique de Georges de Chypre (Corpus Bruxellense Historiae Byzantinae 1). Brussels 1939. Ignatii diaconi Vita Georgii Amastridensis (BHG 668) / V. G. Vasilievsky, Zitie Sv. Georgija Amastridskago, in: Trudy V. G. Vasilevskago 3. Petrograd 1915, 1–71. Vita Alypii stylitae Hadrianopolitani prior (BHG 64) / H. Delehaye (ed.), Les saints stylites (Subsidia hagiographica 14). Brussels 1923 (repr. 1989), 148–69.

Literature Ahrweiler, H. 1966. Byzance et la mer. La marine de guerre – la politique et les institutions maritimes de Byzance aux VIIe–XVe siècles (Bibliothèque byzantine: études 5). Paris. Anderson, W. 2009. ‘Late Byzantine Occupation of the Castle at Tios’, Anatolia Antiqua 17, 265–77. Anderson, W. 2011. Paphlagonia in Late Antiquity: Landscape and Material Culture in North-Central Anatolia (PhD Dissertation). Melbourne. Atasoy, S. 2008. ‘Zonguldak-Filyos (Tios/Tieion/Tion/Tianos/Tieum) kurtarma kazısı’, in İ. Delemen, S. Çokay-Kepçe, A. Özdzbay and Ö. Turak (eds), Euergetes. Prof. Dr. Haluk Abbasoğlu’na 65. Yaş Armağanı. Antalya, 91–7. Avramea, A. 2002. ‘Land and Sea Communications, Fourth–Fifteenth Centuries’, in A. E. Laiou (ed.), The Economic History of Byzantium: From the Seventh through the Fifteenth Century 1. Washington D.C., 57–90. Baird, D. 2004. ‘Settlement Expansion on the Konya Plain, Anatolia 5th–7th Centuries AD’, in W. Bowden, L. Lavan and C. Machado (eds), Recent Research on the Late Antique Countryside (Late Antique Archaeology 2). Leiden, 219–46. Bekker-Nielsen, T. 2013. ‘Neapolis-Neoklaudiopolis: A Roman City in Northern Anatolia’, Der Anschnitt. Beiheft 25, 203–14. Belke, K. 1996. Paphlagonien und Honōrias (Tabula Imperii Byzantini 9). Wien. Blanton, R. E. 2000. Hellenistic, Roman and Byzantine Settlement Patterns of the Coast Lands of Western Rough Cilicia (British Archaeological Report International Series 879). Oxford. Bryer, A. A. M. and D. Winfield 1985. Byzantine Monuments and Topography of the Pontos (Dumbarton Oaks Studies 20). Washington D.C. Casey, J. 2010. Sinope: A Catalogue of the Greek, Roman and Byzantine Coins in Sinop Museum (Turkey) and Related Historical and Numismatic Studies (Royal Numismatic Society Special Publications 44). London.

The end of late antiquity in Paphlagonia

131

Cassis, M. 2014. ‘The Cide and Şenpazar Region in the Byzantine Period’, in B. S. Düring and C. Glatz (eds), Kinetic Landscapes: The Cide Archaeological Project 2009–2011. Surveying the Western Turkish Black Sea Region. Warsaw, 301–40. Chavarria, A. and T. Lewit 2004. ‘Archaeological Research on the Late Antique Countryside: A Bibliographical Essay’, in W. Bowden, L. Lavan and C. Machado (eds), Recent Research on the Late Antique Countryside (Late Antique Archaeology 2). Leiden, 3–51. Crow, J. 2009. ‘Byzantine Castles or Fortified Places in Paphlagonia and Pontus’, in T. Vorderstrasse and J. Roodenberg (eds), Archaeology of the Countryside in Medieval Anatolia (Uitgaven van het Nederlands Instituut voor het Nabije Oosten te Leiden 113). Leiden, 25–43. Crow, J. and S. J. Hill 1995. ‘The Byzantine Fortifications of Amastris in Paphlagonia’, Anatolian Studies 45, 251–65. Doonan, O. P. 2004. Sinop Landscapes: Exploring Connections in a Black Sea Hinterland. Philadelphia. Doonan, O. P. and D. Smart 2000/01. ‘Gerna Dere, Roman and Byantine Settlement in Sinop Province, Turkey’, Talanta 32/33, 17–26. Drauschke, J. 2013. ‘Bemerkungen zu den Auswirkungen der Perser- und Arabereinfälle des 7. Jh. in Kleinasien’ in O. Heinrich-Tamaska (ed.), Rauben – Plündern – Morden. Nachweis von Zerstörung und kriegerischer Gewalt im archäologischen Befund. Hamburg, 117–59. Dunn, A. 1994. ‘The Transition from Polis to Kastron in the Balkans (III–VII cc.): General and Regional Perspectives’, Byzantine and Modern Greek Studies 18, 60–80. Düring, B. S. and C. Glatz 2014 (eds). Kinetic Landscapes: The Cide Archaeological Project 2009– 2011. Surveying the Western Turkish Black Sea Region. Warsaw. Ercyas, D. B. and E. Sökmen 2010. ‘An Overview of Byzantine Period Settlements around Comana Pontica in North-Central Turkey’, Byzantine and Modern Greek Studies 34.2, 119–41. Eyce, S. 1965. Küçük Amasra tarihi ve eski eserleri kılavuzu. Ankara. Foss, C. 1975. ‘The Persians in Asia Minor and the End of Antiquity’, The English Historical Review 90, 721–47. Foss, C. 1985. Kütahya (Survey of Medieval Castles of Anatolia 1, British Archaeological Report International Series 261). Oxford. Guest, P. 2012. ‘The Production, Supply and Use of Late Roman and Early Byzantine Copper Coinage in the Eastern Empire’, Numismatic Chronicle 72, 105–31. Hill, S. J. 1998. ‘Rescue Excavation at Çiftlik (Sinop)’, in R. Matthews (ed.), Ancient Anatolia. Fifty Years’ Work by the British Institute of Archaeology at Ankara. London, 285–300. Howard-Johnston, J. D. 1999. ‘Heraclius’ Persian Campaigns and the Revival of the East Roman Empire, 622–630’, War in history 6, 1–44. Ireland, S. 1996. ‘The Ancient Coins in Amasra Museum’, in R. Ashton (ed.), Studies in Ancient Coinage from Turkey (British Institute of Archaeology at Ankara 17). London, 115–37. Ireland, S. 2000. Greek, Roman, and Byzantine Coins in the Museum at Amasya (Ancient Amaseia), Turkey (British Institute of Archaeology at Ankara 27). London. Ivison, E. A. 2007. ‘Amorium in the Byzantine Dark Ages (Seventh to Ninth Centuries)’, in J. Henning (ed.), Post-Roman Towns. Trade and Settlement in Europe and Byzantium 2 (MillenniumStudien 5.2). Berlin/New York, 25–60. Izdebski, A. 2012. ‘The Changing Landscapes of Byzantine Northern Anatolia’, Archaeologia Bulgarica 16, 47–66. Jacobs, I. 2009, ‘Encroachment in the Eastern Mediterranean between the Fourth and the Seventh Century AD’, Ancient Society 39, 203–44. Jacobs, I. 2011. ‘The Archaeology of Late Antique Anatolia’, in Lexicon of the Springer-Verlag, Berlin. Kennedy, H. 1985. ‘From Polis to Madina: Urban Change in Late Antique and Early Islamic Syria’, Past and Present 106, 3–27. Kilavuz, B. N. and E. Çelkbaş 2013. ‘Paflagonia Hadrianoupolis’i’, Tarih Kültür ve Sanat Araştırmaları Dergisi 2, 159–214. Kirilov, C. 2007. ‘The Reduction of the Fortified City Area in Late Antiquity: Some Reflections on the End of the ‘Antique City’ in the Lands of the Eastern Roman Empire’, in J. Henning (ed.), PostRoman Towns, Trade and Settlement in Europe and Byzantium 2 (Millennium-Studien 5.2). Berlin/ New York, 3–24.

132

Max Ritter

Kolb, F., B. Kupke, M. Miller and M. Zimmermann 1991. ‘Spätantike und byzantinische Besiedlung auf dem Gebiet der lykischen Polis Kyaneai’, Klio 73, 563–85. Kortanoğlu, E. 2013. ‘Tieion/Tios Akropolis’inde Ortaya Çıkartılmış Anıtsal Bir Yapının Kalıntıları Üzerine Gözlemler’, Colloquium Anatolicum 12, 211–38. Lafli, E. and E. Christof 2011. ‘Der kaiserzeitliche Tempel von Asartepe/Kimistene in der Chora des paphlagonischen Hadrianopolis – Ergebnisse der Prospektion von 2005’, Istanbuler Mitteilungen 61, 233–85. Lafli, E. and G. Kan Şahn 2012. ‘Terra Sigillata and Red-Slipped Ware from Hadrianopolis in Southwestern Paphlagonia’, Anatolia Antiqua 20, 45–120. Lafli, E. and A. Zäh 2008. ‘Archäologische Forschungen im byzantinischen Hadrianoupolis in Paphlagonien’, Byzantinische Zeitschrift 101, 681–713. Lafli, E., C. S. Lightfoot and M. Ritter forthcoming. ‘Byzantine Coins from Hadrianoupolis in Paphlagonia’, Byzantine and Modern Greek Studies 39. Lightfoot, C. S. 2012. ‘Business as Usual? Archaeological Evidence for Byzantine Commercial Enterprise at Amorium in the Seventh to Eleventh Centuries’, in C. Morrisson (ed.), Trade and Markets in Byzantium. Washington D.C., 177–91. Lo Cascio, E. 2000. ‘The Population of Roman Italy in Town and Country’, in J. L. Bintliff and K. Sbonias (eds), Reconstructing Past Population Trends in Mediterranean Europe (3000 BC – AD 1800). Oxford, 161–71. Lohmann, H. 1999. ‘Survey in der Chora von Milet. Vorbericht über die Kampagnen der Jahre 1996 und 1997’, Archäologischer Anzeiger, 439–73. Marek, C. 1989. ‘Amastris. Geschichte, Topographie, archäologische Reste’, Istanbuler Mitteilungen 39, 373–89. Marek, C. 1993. Stadt, Ära und Territorium in Pontus-Bithynia und Nord-Galatia (Istanbuler Forschungen Beiheft 39). Tübingen. Matthews, R., M. Metcalf and C. Cottica 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 (British Institute at Ankara 44). London, 173–226. Meier, M. 2003. Das andere Zeitalter Justinians: Kontingenzerfahrung und Kontingenzbewältigung im 6. Jahrhundert n.Chr. Göttingen. Mundell Mango, M. 2006. ‘Action in the Trenches: A Call for a more Dynamic Archaeology of Early Byzantium’, in E. M. Jeffreys (ed.), Proceedings of the 21st International Congress of Byzantine Studies 1. London, 83–98. Mundell Mango, M. 2011. ‘Monumentality versus Economic Vitality: Was a Balance Struck in the Late Antique City?’, in I. Iliev (ed.), Proceedings of the 22nd International Congress of Byzantine Studies 1. Sofia, 239–62. Musso, L., G. Bertoletto, M. Brizzi and B. Westwood 2011. ‘L’edificio abitativo alle pendici orientali dello Zımbıllı Tepe’, in L. Summerer (ed.), Pompeiopolis I. Eine Zwischenbilanz nach fünf Kampagnen (2006–2010). Langenweißbach, 75–120. Nesbitt, J. W. and N. A. Oikonomides 2009. Catalogue of Byzantine Seals at Dumbarton Oaks and in the Fogg Museum of Art. Washington D.C. Niewöhner, P. 2007. ‘Byzantinische Stadtmauern in Anatolien. Vom Statussymbol zum Bollwerk gegen die Araber’, in J. Lorenzen, F. Pirson, P. I. Schröder and U. Wulff-Rheidt (eds), Aktuelle Forschungen zur Konstruktion, Funktion und Semantik antiker Stadtbefestigungen. Istanbul, 239– 60. Polacek, J. 2008. ‘Heraclius and the Persians in 622’, Bizantinistica 10, 105–24. Redford, S. 2011. ‘Sinop in the Summer of 1215: The Beginning of Anatolian Seljuk Architecture’, in D. Kassab Tezgör (ed.), Sinope: un état de la question après quinze ans de travaux: Sinop 2009. Leiden, 125–49. Saradi, H. G. 2006. The Byzantine City in the Sixth Century. Literary Images and Historical Reality. Athens. Sharp, R. St. 2010. The Outside Image: A Comparative Study of External Architectural Display on Middle Byzantine Structures on the Black Sea Littoral (unpublished Ph.D. thesis). Birmingham.

The end of late antiquity in Paphlagonia

133

Sönmez, İ. F. and B. Öztürk 2008. ‘Batı Karadeniz’de Bir Antik Kent Kazısı: Tios (Filyos)’, Arkeoloji Sanat 127, 133–46. Stathakopoulos, D. 2007. ‘Crime and Punishment: The Plague in the Byzantine Empire, 541–749’, in L. K. Little (ed.) Plague and the End of Antiquity. The Pandemic of 541–750. Cambridge, Mass., 99–118. Steffensen, J. P. 2008. ‘High-Resolution Greenland Ice Core Data Show Abrupt Climate Change Happens in Few Years’, Science 321 no. 5889, 680–4. Vanhaverbeke, H., F. Martens and M. Waelkens 2007. ‘Another View on Late Antiquity: Sagalassos (SW Anatolia), its Suburbium and its Countryside in Late Antiquity’, Proceedings of the British Academy 141, 611–48.

Celal Özdemir

Preliminary results of rescue excavations at the Sanctuary of Zeus Stratios near Yassıçal, Amasya province

1. Introduction The Sanctuary of Zeus Stratios is located near Yassıçal, formerly called Ebemi, on top of the so-called Büyük Evliya Çalı (Great Saint Tomb) Hill which is 2.2km as the crow flies southwest of Yassıçal and 1,312m a.s.l. The village of Yassıçal is located on a plateau 1,050m a.s.l. and approximately 7.5km as the crow flies east of Amasya. The ancient road coming from Komana, Neokaisareia and Eupatoria through the Kelkit valley to the city of Amaseia passes northwest of the temenos.1 The Sanctuary of Zeus Stratios is surrounded by a temenos wall encircling an area of 197m in diameter; the wall is 90–100cm thick, preserved to a height of up to 30cm and built of drystone masonry using the natural rock of the hill.

2. The multi-phased building complex Rescue excavation was carried out at the sanctuary from August to October 2006 under the direction of the author with a team comprising staff from the museum and 23 workmen.2 At the northeast end of the temenos a multi-phase building complex of five rooms was excavated (Fig. 1). Before excavation commenced, the area was surveyed and a grid was established. Five trenches were planned in order to understand what had previously been unearthed and destroyed during illegal excavations undertaken in the north and northwest areas of the sanctuary. During sieving of the spoil left by the illegal diggers, a large number of nondiagnostic pottery sherds, two complete terracotta lamps and a Hellenistic coin (Fig. 2) as well as a Roman coin were recovered. In order to investigate the walls unearthed during the illegal excavations, Trench 1 (10m × 10m) was opened and four additional trenches (each 10m × 5m) were laid closely parallel to each other. Excavations were carried out in all five trenches simultaneously. The walls appeared severely damaged on the surface of the trenches but were better preserved at a depth of 25–30cm. The walls were constructed of rubble bound with a sand and limestone 1 2

French 1985: 144–5. Permission and financial support were granted by the General Directorate for Cultural Heritage and Museums of the Ministry of Culture and Tourism. This is a first, very preliminary report of the excavation carried out in 2006. A project including scientific study of the excavated structures and materials is currently being conducted under the auspices of the author.

136

Celal Özdemir

Fig. 1: Plan of the multiphased building complex (C. Özdemir, ©Amasya Museum). mortar after a level of 30cm and were plastered with a 2cm thick layer of sand and limestone mortar. The walls belong to a monumental structure consisting of five rooms and a porch. The walls unearthed in Trench 1 constitute Room A (Fig. 1), which is the largest room in the complex and measures 19.15m × 8.65m. From the soil deposit within Room A, a ceramic lid incised with the Greek word ΔΙΟC probably a genitive Zeus and a bronze coin were found. The walls in Trench 2 constitute Room B (Figs 1, 3–4), where a small altar was found 40cm from the surface. The altar has a rectangular prismal form and is inscribed with a dedication in Greek in six lines. Additionally, a silver coin and a bronze votive tablet (Fig. 5) were also found. Northeast of Trench 2, the walls unearthed in Trench 5 constitute a room separated into two parts and referred to as C1 and C2 (Figs 1 and 6). Between Rooms A and B, a narrow room aligned in a north-south direction was detected. This corridor-shaped room is referred to as Room D (Figs 1 and 7). All the roms are of rectangular plan. The excavations continued within the rooms, and a fill layer composed of dense mortar, plaster fragments and rubble from limestone ashlar blocks was excavated. At the lower level of this fill layer, close to the floor level, rectangular terracotta tiles were recovered. Some of these roof tiles were stamped with the name of the workshop. Based on the plaster fragments, mortar and collapsed material recovered, the author interprets the roof of Rooms A, B and D as vaulted with a tile-covered exterior. Well-preserved floors were detected at a depth of 134cm below the surface in Room A and at a depth of 78cm in Rooms B and D. The floors of C1 and C2 were recovered at a depth of 226cm from the surface in a rather damaged state. At the south end, a porch measuring 3.2m × 22m was located in front of Rooms A–D. The entrances of Rooms A, B and D opened on to this porch, which was paved with massive stone slabs (Figs 1 and 8).

Preliminary results of rescue excavations at the Sanctuary of Zeus Stratios

137

Fig. 2: Hellenistic coin of Amisos (C. Özdemir, ©Amasya Museum). The long, narrow corridor (Room D) has a single door opening out to the porch which also provided access to Room A via two doors in the east wall at either end of the corridor (Fig. 1). The excavation in Room D continued down to the floor level (Fig. 7), and below the layer of the collapsed roof, marble altar fragments, a marble architectural element decorated with acanthus, figural terracottas and two ceramic jugs were recovered. Rooms C1 and C2 constitute the western face of the building (Fig. 6). The two rooms were separated by a wall and there are traces of a connecting door or other access to the porch. Room C1 measures 5.91m × 2.83m; Room C2 measures 4.71m × 2.7m. The floors of these rooms were completely destroyed. Their interior walls were covered with a 2cm-thick layer of sand and limestone plaster. The landscape drops away towards the north, and, as a consequence, the northern halves of Rooms A, B and D had collapsed. During the cleaning of the debris, a coarse stone wall was unearthed; it starts in the middle of the eastern wall of Room A, passes by Room D and ends at the western wall of Room B. In both Rooms A and B two different floor levels were distinguished representing two different phases (Figs 3–4 and 9). Room B measures 6.48m × 11.5m. It has a double-door entrance opening out to the porch (Fig. 1). A bronze coin, terracotta and lamp fragments were revealed in this room at a depth of 50cm below the surface. An 11cm-thick flooring – a mixture of coarse tile fragments and sand and limestone mortar – was found at a depth of 78cm from the surface (Figs 3–4). This flooring sits on a a layer of rock fragments. After the removal of the collapsed parts of the floor in Room A, a second floor covering, constructed of mixed soil and natural red clay, was found ca 15–20cm below the upper one (Fig. 4). This second floor level did not continue up to the northern walls in Rooms A, B and D, but stopped in the middles of the rooms, thus indicating that the original rooms were smaller and were enlarged in a later building phase. The foundation of the enlarged room (Rooms A, B, and D) consisted of a burnt, black deposit of soil with ocassional large rocks up to 173cm deep and created a level surface across the northern and higher southern parts of the room. In the foundation layer of Room B, a bronze coin was found at a depth of 1m and a complete terracotta lamp at 135cm. Four bull skulls were recovered in the pebble preparation layer of the lower floor at a depth of 140cm (Fig. 10). Two Roman city coins were also found at a depth of 1.50–1.55m and another city coin was recovered from the foundation layer.

138

Celal Özdemir

Fig. 3: The flooring of Room B (C. Özdemir, ©Amasya Museum).

Fig. 4: View of Room B showing the different floor levels (C. Özdemir, ©Amasya Museum).

Preliminary results of rescue excavations at the Sanctuary of Zeus Stratios

139

Fig. 5: Bronze votive tablet with a dedication to Zeus Stratios found in Room B (C. Özdemir, ©Amasya Museum). Excavations were also conducted in the porch fronting the building (Figs 1, 11–12). Column bases found here suggest that the porch was lined with Ionic columns. In the porch, a floor paved with irregular stone slabs was observed (Fig. 8). In order to investigate this structure further, a second trench was opened here, and a large inscribed altar was found discarded in a pit at a depth of 3.2m. The four lines of inscription suggest that the altar was a dedication. South of the multi-phased structure and the area in front of the porch where the inscribed altar was unearthed the landscape drops away to the west. The 13m difference in elevation is compensated for by a terrace wall. Excavations conducted in front of the porch recovered a square water basin (1m × 1m and 90cm high). Since no water source was found within the temenos, this stone basin was possibly used for animal sacrifice. In the excavations conducted outside the porch, west of Rooms C1 and C2, a structural foundation consisting of large ashlar blocks was revealed (Fig. 11). The fragments of marble sculpture (depicting humans and eagles), small stone altar fragments with and without inscriptions, as well as the bronze coins and Ionic column bases found in this trench suggest that the structure was a temple.

3. Excavations at the site of the altar Three trenches were opened on top of the hill where the monumental altar itself was located. Trench AL-1 was opened in the western corner of the hill, Trench AL-2 in the eastern corner (Fig. 13) and Trench AL-3 immediately west of AL-2. In all of these trenches fragments of figural sculpture were found close to the surface. In L-shaped Trench AL-1 large ashlar blocks were revealed at a depth of 1.7m below the surface level; these were quite dispersed and randomly thrown on top of one another. However, a structure was discernible from the unspoilt stone row which extended through the whole trench (3.8m × 1.57m wide × 0.5m high). Limestone ashlar blocks dovetailed together using iron clamps were preserved in two courses. The bedrock was cut and prepared in order to built the foundation for the masonry structure. The foundation was raised by three courses of well-cut ashlars.

140

Celal Özdemir

Fig. 6: Detail of Rooms C-1 and C-2 (C. Özdemir, ©Amasya Museum).

Fig. 7: View of the corridor/ Room D (C. Özdemir, ©Amasya Museum).

Preliminary results of rescue excavations at the Sanctuary of Zeus Stratios

Fig. 8: View of the paved porch (C. Özdemir, ©Amasya Museum).

Fig. 9: Detail of Room A showing the walls from the first (Hellenistic?) building phase (C. Özdemir, ©Amasya Museum).

141

142

Celal Özdemir

Fig. 10: Bovine skull found in the fill of the lower floor of Room B (C. Özdemir, ©Amasya Museum). Ashlars of various sizes (0.45m × 0.45m, 0.2m × 0.4m, 0.46m × 0.73cm) were recorded at a depth of 1.7m below the surface in Trenches AL-2 and AL-3. The blocks were dispersed and had been removed from their original settings by illegal excavations. Some nondiagnostic Hellenistic and Roman pottery fragments as well as a few heavily-corroded coins were found in this context. It seems that this was the substructure of the monumental altar which rose to a height of 2.5m from the bedrock, given that there were 3–4 rows of 0.45 by 0.45m large blocks, the four external sides contained a stone and earth filling. It seems that the substructure for the altar was built with ashlars measuring 0.45m × 0.45m, while the superstructure of the altar was built with smaller blocks measuring 0.2m × 0.4m. From these observations it can be concluded that a monumental altar was built on the flat summit of the hill in drystone masonry utilising local stone from the hill itself.

4. The Sanctuary of Zeus Stratios The depictions on the reverse of the Roman-period bronze city coins of Amaseia found in the foundation level of the monumental altar as well as at other sites show a stone built, rectangular structure representing an altar, on some examples with addition of a quadriga and an eagle. It is now evident that these coin representations relate to the hill-top sanctury at Yassıçal. On the other hand, the monumental altar on the Yassıçal hill was surrounded with inscribed stones recording the names of villages and districts of the region as known from the publications of Franz Cumont and David French.3 These inscriptions were arranged in a crescent formation at precise intervals. A new example of such an inscription was recovered in situ during the excavation. It is suggested that people from these villages and districts visited the Sanctuary of Zeus Stratios in order to worship and sacrifice, as well as to participate in oath-taking ceremonies. Possibly, each region or village gathered at their own inscribed stone in order to execute the rituals of the sanctuary.

3

Cumont 1906: 51–2, 172–82; French 1986: 71–83.

Preliminary results of rescue excavations at the Sanctuary of Zeus Stratios

Fig. 11: Remains of the temple foundation southwest of the multi-phased building complex (C. Özdemir, ©Amasya Museum).

Fig. 12: General view to the building complex from the southeast (C. Özdemir, ©Amasya Museum).

143

144

Celal Özdemir

Fig. 13: View of Trench AL-1 showing the substructure of the monumental altar (C. Özdemir, ©Amasya Museum). The establishment of the Zeus Stratios, Zeus the Warrior, altar may have been associated with a military cult in the city of Amaseia, probably introduced when the city became the first capital city of the Pontic/Mithridatic kingdom. Of the 33 coins found during the rescue excavations only a few belong to the Hellenistic period. One of them evidently dates to the reign of the Pontic king Mithradates Eupator, 120–63 BC. On the obverse the head of the young Ares with a Corinthian helmet is depicted. On the reverse is a sheathed sword and the inscription AMIΣOY (Fig. 2). Although it was not found in a closed context, it provides new evidence that the sanctuary may have been active as early as the Hellenistic period. In Room B, a second-century BC coin was recovered.

5. Conclusions The rescue excavations conducted by the Amasya Museum in 2006 provide the first excavated data from the sanctuary on top of the hill outside the village of Yassıçal. On the basis of the coin finds it can be established that the sanctuary was in use as early as the Hellenistic period and the Roman coins recovered in the area suggest that it continued in use into the 4th century. The remains of the structure unearthed on the summit of the hill at the centre of the enclosed sacred area confirm that the Amaseian coin representations of the monumental altar of the Sanctuary of Zeus Stratios relate to the sanctuary at Yassıçal. The excavation at the northeastern end of the temenos unearthed a multi-phased building complex accessible by a columned and paved porch. The function of the multi-phased structure is not well understood. The courses of unearthed walls of different periods show

Preliminary results of rescue excavations at the Sanctuary of Zeus Stratios

145

that the rooms were initially smaller and were enlarged following artificial levelling of the sloping terrain. Based on the various materials found within this structure it is assumed that it served to house priests or to receive offerings made at the sanctuary. The structure unearthed to the south of this building is probably the foundation of a temple. Celal Özdemir Amasya Museum Atatürk Caddesi, Amasya Merkez, Amasya 05100, Turkey [email protected]

Bibliography Cumont, F. and E. 1906. Voyage d’exploration archéologique dans le Pont et la petite-Arménie (Studia Pontica 2). Brussels. French, D. H. 1985. ‘Roma Yolları ve Mil Taşları 1984’, III. Araştırma Sonuçları Toplantısı, 143–54. French, D. H. 1986. ‘Recent Epigraphic Research in Pontus’, Epigraphica Anatolica 8, 71–83.

Pavol Hnila

Rural necropoleis and settlement dynamics: Thoughts on Roman and Byzantine graves at Oymaağaç Höyük, Samsun province

1. Introduction The site of Oymaağaç Höyük is located in the province of Samsun, 7km to the north of the town of Vezirköprü, at the west end of the village of Oymaağaç Köyü.1 It is a small mound (ca 2.5ha) rising 10–20m above its surroundings. The site is almost certainly to be identified with the ancient city of Nerik – an important cultic centre known from Hittite cuneiform texts.2 The mound has, however, multiple settlement levels ranging from at least the Early Bronze Age to the Late Iron Age. Two or three centuries after the desertion of the latest Iron Age settlement, the site started to be used as a necropolis (see the chronological section below). The necropolis spread over the mound and covered an area of more than 2,500m2 (Fig. 1). The exact limits of the necropolis cannot be ascertained because of considerable erosion at the edges of the mound. Furthermore, deep ploughing of the fields before excavation of the site began increased the natural erosion process and heavily damaged many graves; sometimes the evidence of a grave is limited to just a few scattered bones. In fact, the intense agricultural activities of the recent past have eradicated all the graves that once must have existed on top of the mound, as well as an unknown number of graves along the mound perimeter in areas most exposed to erosion. Although the excavation of the necropolis is not yet completed, the excavated graves already permit some interesting considerations that are presented as preliminary results in this paper. Kathryn Marklein and Sherry Fox continue to work on the anthropological analysis of the human bones. A detailed catalogue of the graves and finds will be published later.

1

The Oymaağaç Project is directed by Jörg Klinger (Free University of Berlin) and Rainer Czichon (University of Uşak), and is supported by the Ministry of Culture and Tourism of the Republic of Turkey, the Deutsche Forschungsgemeinschaft, the Gerda Henkel Stiftung and Tepe Knauf Ankara. For more information, see the homepage of the project: www.nerik.de. 2 See Czichon et al. 2011; Czichon 2013.

148

Pavol Hnila

Fig. 1: Plan of the necropolis of Oymaağaç Höyük (with grave types) (plan by author).

2. Grave types So far, 53 graves have been uncovered. Typologically, we can sort them into cist graves, tile graves, stone-lined graves and pit graves. The cist graves are the most elaborate type of grave encountered at Oymaağaç, and so far twelve of them have been excavated. All the cist graves are rectangular in shape. Usually, they are built of large stone slabs combined with smaller stones and joined with lime mortar (Fig. 2). The floors are left untreated, the walls may be plastered with lime. No covering slabs have been found; yet, given the frequent reopenings of the cist graves (a habit which is discussed below), we must assume that some sort of reclosable covering existed, and the lack of evidence should be viewed as a consequence of the poor state of preservation of the graves at the site.3 Comparable cist graves from other sites were usually covered with several stone slabs placed tightly next to each other.4 Tile graves account for only nine examples and can be subdivided into two variants (for the typology of the tiles themselves, see below): (1) graves with two rows of roof tiles forming a gabled roof over the body and two additional roof tiles closing the gable ends at the 3

Similar cist graves cut into the rock at Dökme Tepesi (72km southeast of Oymaağaç Höyük) were better preserved and their walls had recesses for covering plates: see Czichon et al. 2006: 173. 4 Examples of coverings of stone slabs are known from, for example, Pessinus (Krsmanovic and Anderson 2012: fig. 9), Gordion (Goldman 2007: 301) and Troy (Böhlendorf 1998: 264–6). At Kayalıpınar, remains of thick wooden planks were found used as coverings above cist graves made of mud-brick; the published report does not, however, indicate any evidence for reopenings: Zeiler 2006: 241.

Rural necropoleis and settlement dynamics

149

Fig. 2: Cist grave locus 7585:010 with multiple successive burials from Oymaağaç Höyük necropolis (author’s photo). head and the feet (tombs alla cappuccina – Fig. 3); (2) graves in the shape of a rectangular cist consisting of upright roof tiles. All but two of the excavated examples belong to variant (1).5 Thirty graves are pit graves, and these are, in fact, the most numerous type of graves. Pit graves do not have any particular construction feature, other than being pits dug into the earth. The remaining two graves fall into a category which I term ‘stone-lined graves’. The upper part of the grave is roughly delineated with stones, tiles and mud-bricks. This type is likely underrepresented since erosion could have deprived other graves of their uppermost linings and made them appear as simple pit graves. The tile graves and stone-lined graves seem to concentrate in the eastern part of the excavated necropolis; other grave types are distributed without any recognisable pattern. At Oymaağaç, the graves almost never overlap. Only two graves, with slightly different orientations, seem to have been placed unusually close to each other; yet, the area was too disturbed to understand whether or not one grave actually cut the other.6 In another instance, it seems that an earlier grave was disturbed during the construction of a cist grave; the long bones and the skull from the disturbed burial were collected, accompanied by an intact jug, covered with tiles and reburied in the construction fill of the cist grave, leaning against the outer (!) side of the stone cist.7 Secondary burials with missing bones may also 5 6 7

The two exceptions are loci 7585:174 and 7585:042. Grave loci 7685:015 and 7685:018. Grave locus 7585:174. We do not know whether the tiles belonged to the original burial or to the reburial, but they cannot stratigraphically postdate the cist grave. These tiles are of the first type described below in the chronological section and their proposed Roman/early Byzantine date fits well with the Roman dates indicated for the Oymaağaç cist graves.

150

Pavol Hnila

Fig. 3: Tile graves loci 7585:022 and 7585:026 from Oymaağaç Höyük necropolis (photo by H. Marquardt). be present among the multiple burials in some of the cist graves – a possibility that is being checked by the ongoing anthropological analysis. All the other graves respect each other, and we may infer from this that their positions were visibly marked. Furthermore, the reuse of graves with multiple successive burials is possible only if we assume that the graves were marked on the surface and could be easily relocated and reopened.8 Yet, it should be noted that no such markers have been found at the site. A funerary stele made of grey limestone and postdating the late second century was found among the spolia uncovered during the construction of the new village mosque ca 800m to the east of Oymaağaç Höyük,9 but it is unclear from where it originates; it might have come from Oymaağaç Höyük, but it might, equally, have been transported to the mosque from the Roman city of Neoklaudiopolis (7km to the south) or even from some unknown rural necropolis in the vicinity. It is possible, of course, that some grave markers were originally made of wood and have vanished completely or that graves were marked just by stone covering slabs.10

3. Matters of orientation All graves contained inhumations, no signs for cremation were found. The inhumations are aligned roughly east-west, with the head always placed to the west. This is the common burial custom in Christian communities and, although less rigorously applied, it was a 8

Rife 2012 (169) argues for the contrary on the basis that no markers were found above the cist graves at Isthmia. 9 Bekker-Nielsen and Høgel 2013: 154–5. 10 For covering slabs as markings, see Wittwer-Backofen 1987: 382.

Rural necropoleis and settlement dynamics

151

widespread custom also in the early Roman period.11 This orientation practice is usually explained in terms of an ideological or religious necessity for the body to face east, in the direction of the rising sun,12 and it is sometimes explicitly connected to the doctrine of the Resurrection.13 The orientation axis of the Oymaağaç graves varies between 244° and 313°; this is a considerable range and it is thus evident that the graves were not aligned with precision towards the west (270°). However, since most of the graves fit well within the limits set by the positional fluctuation of the rising sun on the horizon throughout the year, perhaps the alignment of each burial was determined according to the position of the rising sun on the day of interment. Nonetheless, there are five outliers for which there is no day of the year, neither in the present nor in the past, on which the sun can rise with the azimuth angle implied by their orientation.14 The motives behind these divergent orientations remain obscure and the possible range of explanations is broad – the outliers may have been oriented, for example, according to some feature in the graveyard or in the village or in the landscape or according to a celestial body. The dead were all buried in supine extended positions, with the arms usually placed over the abdomen. There are some variations: arms may be folded or crossed or each might be placed over the chest and pelvis respectively. Finally, it should be noted that there are no signs that would unequivocally classify the graves as Muslim burials.15

4. Mass burials and the case for a catastrophic event The minimum number of interred individuals differs considerably among grave types. Tile graves and stone-lined graves were employed for discrete individual burials (with a single exception, where two individuals were buried inside the same tile grave). Pit graves usually contain single burials as well, but there are several notable exceptions containing multiple skeletons. The cist graves, on the other hand, always include numerous bodies (up to 28 individuals). 11 12 13 14

Rife 2012: 182. Rife 2012: 108–9, cat. nos 16–7. Kwiatkowska 2009: note 24. The changes in sunrise azimuth angles over centuries and millennia are negligible for the purpose of funerary ritual analysis, ranging in tenths and hundredths of a degree. The maximum and minimum azimuths of sunrise are defined for Oymaağaç as 57° and 121°, meaning that all bodies oriented between 239° and 303° face the rising sun at some period during the year (all azimuth calculations made with NOAA Solar Calculator 2014). The orientation angles of the five outlying graves are 307°, 309°, 311°, 312° and 313°. 15 Kwiatkowska 2009 (132–4) distinguishes between Muslim and late Roman/Byzantine graves at Çatalhöyük. Since both Christians and Muslims buried their dead with the same east-west orientation and without grave goods, she considered as being Muslim all pit graves, with or without niches, with the entire body or at least the head facing south, towards Mecca. At Oymaağaç, only grave locus 7685:051 contained a skeleton with the usual east-west alignment with the head placed to the west but facing south. However, the hands of this skeleton were folded over the belly; this is a standard position encountered numerous times across the Oymaağaç cemetery and is not found among the positions of Muslim burials documented in Kwiatkowska 2009. The position of the head in this particular grave is likely a result of postdepositional processes – a chance slip after the disintegration of the soft parts – rather than a deliberate act during burial.

152

Pavol Hnila

Cist graves are well known from a number of Roman and Byzantine sites as tombs designed for multiple successive burials. And this seems to be the standard scenario at Oymaağaç Höyük, as indicated by frequent disarticulated skeletons (Fig. 2) and possibly also by occasional traces of burning found in the chest areas of superimposed individuals. These traces of burning on bones, combined with remains of charcoal, can best be explained as evidence for the custom of burning incense in the graves.16 The aromatic smoke and also the perceived purificatory effects of the incense would be highly desirable during the burial ritual, since the reopening of a grave containing bodies in various stages of decomposition was definitely not a pleasant experience.17 Yet, at least two cist graves at Oymaağaç Höyük have revealed veritable mass burials; both contained several articulated skeletons from individuals who had died ‘within a relatively close time frame’ – as concluded by the anthropologists Sherry Fox and Kathryn Marklein.18 It is expected that additional cist graves will be classified as mass burials as a consequence of the ongoing anthropological analysis. This is presumably also the case for the tightly-placed multiple skeletons found in some pit graves. The latter make more sense as mass burials interred during a one-off event since any reopening of a pit grave would inevitably lead to interference with previous burials and to intense disarticulation of skeletons. The existence of several mass graves instead of a single larger one is particularly striking. What might be the reason for this? We currently have no reason to believe that the mass graves each date to a different period. On the other hand, there are anthropological clues to indicate that cist graves served for individual families, and they contained individuals of both sexes and all ages.19 The cist graves are sparsely distributed, and do not disturb each other or other existing tombs. We should also note that the construction of a cist from large stone slabs represents a considerable labour investment. Such an investment would be well within the means of wealthier individual families during normal times, but it is highly

16

As concluded for the Byzantine cemetery in Mani, southern Peloponnese: Poulou-Papadimitriou et al. 2012: 393. 17 The written account of St Gregory of Nysa from the fourth century documents how horrified he was by viewing the corpses of his parents when his sister was added to the family tomb: see Rife 2012: 170. 18 Grave loci 7384:009 and 7484:021. Fox and Marklein 2013; 2014: 203; Marklein 2014. Various conflicting definitions of mass graves exist, based on the quantity of bodies, the articulation of the bodies, their positions or a particular motive; for an overview, see Jessee and Skinner 2005; Sprague 2005. 19 The conventional analysis of the skeletal material demonstrates that in one grave there was a much higher percentage of individuals with Os tibiale externum anomaly than in other graves and in the current population. Since the OTE anomaly has a genetic component, its concentration in the given grave points to a familial relation between the deceased: see Fox and Marklein 2013; 2014. Family use is implied not only for the cist graves but also for the tile graves. In one instance two individuals were buried in a single tile grave (7685:007). In other instances tile graves were almost ‘attached’ to each other and formed pairs or even clusters (loci 7585:023 + 7585:027 and loci 7685:006 + 7685:007 + 7685:008), despite the fact that the graves in each group could have been located nearby with more space between them. This practice of ‘additive graves’ is attested also for medieval Europe: Nowotny 2011: 452.

Rural necropoleis and settlement dynamics

153

improbable that any rural community would venture to build elaborate mass graves at the time of a catastrophic event. Finally, in some instances mass burials seem to be combined with multiple successive burials in the same tomb.20 A rather complex preliminary picture emerges. The cist graves were established as family tombs and were intended for continued use over several generations. Yet, at some point during their existence there was a catastrophic event that led to the sudden death of many members of the community. The catastrophe likely happened as a single event or over a short time span, and the bodies were mostly interred according to their family affiliations in already existing graves. The family tombs of the necropolis would have been at different stages of their use-lifes: some would have been half-full with multiple successive burials, some may have been newly built and almost empty, while others would have already fallen out of use. Some individuals were collectively buried in simple pit graves, whether family or community related. Whatever happened, the community survived this catastrophe and continued to bury their dead in the usual ways afterwards. For this reason, some cist tombs contain successive burials superimposed over mass burials and some cist tombs, devoid of mass burials, may have been established after the catastrophe. The complete sequence of the interments and their dating is difficult to reconstruct, because of the commingling of remains and the extreme scarcity of datable finds. Perhaps a combination of C14 and DNA analysis might yield results. The nature of the catastrophe remains obscured. No traces of violent death were found on the bones.21 An epidemic or a famine look to be the most likely options. Epidemiologic analysis of the teeth, which is currently under way at the University of Tübingen, may elucidate this matter. As discussed below, the mass graves date generally to the period between the second and fourth century. During this period of time, several large-scale plagues and famines are historically attested, including the well-known Antonine plague of 165–80.22

5. The chronology of the necropolis Stratigraphic observations are of little help for dating the graves with precision: they can only be summoned to conclude that the necropolis is later than the Achaemenid period pits dating to the fourth century BC and earlier than the modern topsoil.23 This very rough estimate, spanning over more than 2,300 years, can be rendered more precise with the help of finds and C14 dates.

20

For example, the tightly packed skeletons in anatomical order in tomb locus 7384:009 were superposed by an additional skeleton with disarticulated bones. The last skeleton added to the tomb seems to be again in anatomical order. 21 I thank Kathryn Marklein for this information. 22 See below, ns 37–38. 23 A ceramic Achaemenid bowl with a straight flaring rim and an omphalos was found in one of the pits (find 7587:005:006). According to Toteva 2007 (120 with further references), straight rims occur on developed versions of Achaemenid bowls and date to the fourth century BC. The Iron Age pottery of Oymaağaç Höyük is being studied by Mehmet Ali YIlmaz: see YIlmaz 2014.

154

Pavol Hnila

Fig. 4: Personal adornments from Oymaağaç Höyük necropolis. (A) Gold earring 7586:032:003. (B) Bronze earring 7385:018:020. (C) Bronze earring or hair coil 7483:048:005 (photos by H. Marquardt). Finds, however, are rare; only one third of the graves yielded artefacts such as personal adornments (Fig. 4A–C), clothing components, coins, pottery and tiles. Personal adornments and clothing components are extremely simple (rings, undecorated loop-shaped earrings, finger rings, arm rings, hair coils made of bronze and iron) and all can be broadly identified as Roman/Byzantine.24 The same unsatisfactorily vague date also applies to the only gold find – an earring made of thin wire with a simple hitch clasp (Fig. 4A).25 The earliest datable finds from the graves – one pottery sherd and one coin – date to the late Hellenistic period. The small pottery fragment was found in a cist grave and belonged to a mould-made bowl with relief decoration of elongated leaves filled with feathering.26 Parallels for this motif date mostly to the second and the first half of the first century BC.27 Presumably to roughly the same time – that is, the reign of Mithradates VI (120–63 BC) – dates a worn bronze coin with a turreted female head on the obverse and possibly a standing figure on the reverse that was found in the loose fill a few centimeters above the left foot of a skeleton in one of the pit graves.28 The coin and the mould-made pottery fragment 24 25

26 27

28

Simple loop-shaped earrings were fashionable even in the eleventh-twelfth century: Berti 2012: 196. Find 7586:032:003. Earrings with simple hitch clasps have been found in fifth- to sixth-century contexts at Peloponnesos: Rife 2012: 189. Sixteen loop-shaped wire ‘earrings with a hoop end to which the other bent end was hooked’ (including two examples made of gold) were found in the Byzantine necropolis on the agora of Iasos and date from the first half of the seventh century to the eleventh-twelfth century: Berti 2012: 192, n. 12. An earring with a hook and clasp has been found in a middle Byzantine grave at Boğazköy, and dates no later than the eleventh century: Böhlendorf-Arslan 2012: 364, fig. 13.13. Find 7484:020:009. Similar bowls from Ephesos have been compared to examples from Delos where they are dated between 166 and 69 BC: Mitsopoulos-Leon 1991: 67, 71, pl. 79. D12. The same decoration (the same fabric and motifs) also occurs on Ephesian relief lamps belonging to the Delos V group dating generally from the second quarter of the second century to the first quarter of the first century BC, with most of the Ephesos examples dating to the first quarter of the first century BC: Mitsopoulos-Leon 2007: 80, cat. no. L93. Find 7385:025:002. See Sauer in appendix to this paper, coin (A).

Rural necropoleis and settlement dynamics

155

Fig. 5: Pottery from Oymaağaç Höyük necropolis. (A) Jug 7584:010:004. (B) Jug 7585:174:002. (C) Mug 7484:021:014 (photos by H. Marquardt). were both found in fills, thus their presence in these particular graves might be unintentional. Yet, since no comparable finds were recovered from the topsoil and no other activity is documented on the site during the time between the settlement of the Achaemenid period and the subsequent necropolis, it seems probable that these rare finds are related to burial practices. They may have either been deposited in the graves on purpose during interment or be ‘coincidental’, having found their way to these particular graves from scattered remains left lying around after earlier burial activities in the immediate vicinity.29 Either way, they are the only testimonies for a possible late Hellenistic beginning of the necropolis. Datable finds cluster around the second century. Particularly significant in terms of chronological precision is an extremely worn coin with a clearly legible countermark with the Greek letters rho and lambda, found near the head of a single burial in a pit grave.30 All known coins with this countermark were minted in Amasya in 112/3, whereas the countermark itself can be a code for the year 127/8.31 Other second-century finds from Oymaağaç include two complete jugs; one was found in a pit grave32 (Fig. 5A) and the other accompanied a reburial interred secondarily during the construction of a cist grave (Fig. 5B).33 Both jugs belong to the Pontic sigillata production and their best parallels date to 150–200 and 125–75 respectively.34 A third nearly complete vessel, a mug in red semicoarse ware 29 30 31 32 33 34

For the general definition of artefacts from burial deposits as incidental, coincidental or accidental see Rife 2012, 23. Find 7584:004:003. See Sauer in appendix to this paper, coin (B). Find 7584:010:004 (found together with fragments of another jug of the same type: find 7584:010:003). Find 7585:174:002. I wish to thank Kristina Winther-Jacobsen for this identification and for pointing out the parallels in Zhuravlev 2010, form 11.4 (pl. 49.387), form 20.2 (pl. 55.439, pl. XXV:439).

156

Pavol Hnila

decorated with incised double horizontal lines, has not yet been satisfactorily dated (Fig. 5C).35 Furthermore, the lower range of the AMS C14 date from a mass burial in a cist grave falls into the second century, although the upper range falls in the fourth century: 130–264 cal. and 277–331 cal. (2σ, 95.4% confidence interval).36 According to this calibrated C14 date, there is the same statistical probability that the catastrophe that resulted in mass graves was caused by the well-known Antonine plague of 165–180 as by the much less well-known famine and epidemics raging in the eastern part of the empire almost 150 years later, during the winter of 312/3.37 It is worth noting that the statistical probability is significantly lower for another devastating event: the so-called Cyprian plague of ca 250–75.38 Of course, other events may have prompted the catastrophe, perhaps even forgotten local causes undocumented in the written records. The only other bone dated by AMS C14 was taken from a tile grave alla cappuccina and belongs to the Byzantine period: 665–822 cal. and 842–60 cal. (2σ, 95.4% confidence interval).39 The C14 date indicating a date for this tile grave in the Byzantine period is important, since the Oymaağaç tile graves have not yielded any finds except of the roof tiles themselves. The roof tiles are mostly well preserved. They are roughly rectangular tegulae with varying dimensions and typologies; no imbrices were found.40 At least three distinct types can be recognised: (1) tegulae with lateral lower cutaways in the straight flanges (Fig. 6A); (2) tegulae with lateral lower cutaways bordered by offset flanges, forming a ‘step’ (Fig. 6B); (3) tegulae with upper cutaways only (Fig. 6C). The tiles of the second and third types frequently have V-shaped or ribbon-shaped flow-directors applied on the tile surface as plastic ridges. The first type was found only in tile graves without gabled roofs. The second and the third types occur together, and they are the only types used in the tile graves with gabled roofs. The second type seems to be chronologically distinct from the first type. The first type has good parallels among the Roman and early Byzantine tiles of Pompeiopolis in Paphlagonia41 and of Kadykovka and Chersonesos in the Crimea (second/third to seventh century);42 the Chersonesos examples of the second type date no earlier than the ninth to 35

36 37 38 39 40 41 42

Find 7484:021:014; see Czichon et al. 2011: fig. 9. Interestingly, the best parallels to date for the shape come from Crete and date from the late Hellenistic period (Eiring 2001: 92–3, pl. 94h) to early Byzantine period (Poulou-Papadimitriou et al. 2012: fig. 7.4). Their fabrics and decorations, understandably, vary considerably. I wish to thank Andrea Berlin for bringing the Knossian Hellenistic examples to my attention. Find 7384:009:013 dated in Arizona as AA95784 (1795±33 14C yrs BP, δ13C=-19.3‰). The dates were calibrated with OxCal v.4.2.2 using atmospheric data from Reimer et al. 2009. For the 312/3 event, see Stathakopoulos 2004: 179–82. For the Cyprian plague, see Groß-Albenhausen 2005. The statistical probabilities refer to the distribution of histogram peaks by the calibrated C14 date. Find 7685:051:007 dated in Arizona as AA95783 (1267±32 14C yrs BP, δ13C=-18.6‰). The dates were calibrated with OxCal v.4.2.2 using atmospheric data from Reimer et al. 2009. Examples of dimensions: 32cm × 48cm (find 7585:021:001:001), 31cm × 44cm (find 7585:021:001:003), 38cm × 48cm (find 7685:007:005). I wish to thank Mariusz Gwiazda for sharing the unpublished results of his work on the roof tiles of Pompeiopolis with me. See Filippenko 1998, fig. 2:3 for Kadykovka example dating to the second/third century and Romančuk 2005: 132, fig. 41.2 for Chersonesos examples dating to sixth to seventh century.

Rural necropoleis and settlement dynamics

157

Fig. 6: Tiles from Oymaağaç Höyük necropolis. (A) Tegula of type 1 (7585:173:001). (B) Tegula of type 2 (7585:021:001:003). (C) Tegula of type 3 (7585:021:001:001) (author’s photo). the thirteenth century.43 The third type is of a more general Roman/Byzantine shape typology, only its applied plastic ridges appear to be distinctive for Byzantine examples and to be missing on the Roman ones. The combination of all the currently available dating evidence suggests the following chronological differentiation of grave types at Oymaağaç: the cist graves are likely Roman, the tile graves alla cappucina are Byzantine, other tile graves are Roman, the stone-lined graves and pit graves can date to either period. The Byzantine tiles and the C14 date pointing to the eighth and ninth centuries represent the current terminus post quem for the end of the necropolis. It is not clear how long thereafter burials continued to take place. The lack of later datable finds and the lack of Muslim burials might imply that burial activity on the site ceased before the late eleventh century, when the Seljuks started to settle in Anatolia. In conclusion, by a cautious estimation, the necropolis was in use for at least 700 years (the second to the ninth century). However, taking into account all the dating evidence, the necropolis was probably used over the course of 1,000 years or even slightly longer (the first century BC to the ninth/tenth century AD).

6. The community of the living Given the longue durée of the necropolis, the lack of stratigraphic suprapositioning among the graves is quite surprising. The rigorous avoidance of disturbances not only implies the aboveground visibility of earlier graves, but also a certain respect for the previously buried members of the community. Both scenarios can best be explained by the assumption that the necropolis was used continuously throughout the entire duration of its use-life, without any major interruption. 43

Romančuk 2005: 133–45, Abb. 43, 45–9.

158

Pavol Hnila

The community that used the necropolis probably lived not far from the höyük, most likely in the territory of the modern-day village. So far, however, no traces of settlement(s) contemporary with the necropolis have been found nearby.44 As noted above in the introductory section about the necropolis, there can be hardly any doubt that more graves than those recovered by excavation were once present both on the top of the mound and along the current perimeter of the necropolis. Given the ratio between the excavated and the unexcavated areas of the necropolis, we should assume that at least twice as many graves as those excavated originally existed. Thus, a conservative estimate of the total numbers of burials is 450 individuals, and likely more.45 With the estimated number of buried individuals and the approximate duration of the necropolis at hand, it is possible to calculate the average size of the living population that used the necropolis, provided the life expectancy of a newborn is known. The life expectancy of a newborn calculated as an average from the grave stelae inscriptions from the northern Anatolian Roman province of Pontus and Bithynia is 36.93 years.46 This is higher than the average of 25 to 33 years determined by anthropological analysis in the rare cases where the human remains from Roman and Byzantine necropoleis in Anatolia have been studied and published by physical anthropologists.47 Until an anthropological evaluation of the life expectancy specific for the Oymaağaç community can be calculated, it seems most reasonable to operate with the figure of 30 years (an approximate mean of the most extreme values proposed above). According Evžen Neustupný’s equation for calculating population size, the number of dead individuals needs to be multiplied by the life expectancy of a newborn and the result divided by the duration of the necropolis.48 If we consider the minimum duration of the necropolis, this equates to 450 × 30 ÷ 700 = 19.28 individuals. If we consider also the dating evidence from the grave fills (supporting a longer duration of use for the necropolis), the population would diminish to 13.5 individuals (450 × 30 ÷ 1,000). Of course, this is the average size of the population at any time. There were surely considerable fluctuations, as indicated by the fact that most of the skeletons were found in cist graves that seem to date to the Roman period. If so, this would further decrease the size of the Byzantine population.

44 45

46

47

48

Some of the Roman spolia found in the village mosque may have originated from this settlement (see note 9). This figure is calculated from the conjectural limits reconstructed from the excavated areas. If any evidence for further graves were to be discovered on the outer perimeter of the mound, the number of individuals could rise considerably. Marek 2003: 128. Christian Marek recommends treating this mean value with caution. The sample is chronologically widely spread and is only representative for that fraction of the population wealthy enough to commission stone stelae for their deceased. He also points to methodological issues and inconsistencies with year counts. Life expectancy was 24.6 years for those buried in the Roman-Byzantine necropolis of Elaiussa Sebaste in Cilicia (Paine and Vargiu 2010: 283) and 33 years for those buried in the Byzantine necropolis at Boğazköy/Hattusa (Wittwer-Backofen 1987: 385). Neustupný 1983: 9, 20 (equation 32b).

Rural necropoleis and settlement dynamics

159

Every figure of the Neustupný equation applied to Oymaağaç is an estimate, and, in addition, we also do not know what portion of the living population may have migrated, so the outcomes need to be treated with caution. The point, though, is not to establish the absolute size of the population but rather to emphasise the fact that the community using the necropolis consisted, on average, of only a handful of families. This conclusion does not change, even if the estimate of the total number of burials is tripled. In terms of finds density, it can be observed that the three graves that yielded the most grave goods were located close to the top of the mound.49 All three graves were cist graves with multiple burials, and the higher frequency and the more valuable character of the finds (jewellery, including a gold earring), combined with a prominent position in the necropolis, all imply an elevated social status for the families who buried their dead there. Nonetheless, the finds from the urban necropoleis of Neoklaudiopolis and other provincial centres in Pontus tend to be substantially more numerous and precious.50 In conclusion, it seems that Oymaağaç Höyük necropolis was used by a small rural community throughout its history. It was not, however, the only Roman and Byzantine funerary place in the territory of the modern village. Typologically similar cist graves cut into the rock were recently discovered by villagers on a hill ca 1.5km to the east of Oymaağaç Höyük.51 Furthermore, Byzantine columns and other stray finds in and near the modern village necropolis may point to the existence of a church and of further graves ca 2km southeast of Oymaağaç Höyük.52 Should the rock-cut cist graves east of the village prove to be contemporary with the Roman cist graves at Oymaağaç Höyük, there are two possible scenarios: (1) the necropoleis belonged to two different communities living in the same rural settlement; (2) we are dealing with two of many separate smaller hamlets dispersed across the hinterland of the Roman and Byzantine provincial city of Neoklaudiopolis/ Andrapa (modern-day Vezirköprü).53 The Byzantine finds from the modern village graveyard, on the other hand, include glazed pottery postdating the necropolis at Oymaağaç Höyük, and these finds may represent the same community following a shift to a new funerary area over time or a new community that recolonised the area after a settlement hiatus.

49 50

Locus 7585:010, locus groups LGR 0019 and LGR 0020. Czichon et al. 2006 (173–5) mention rich glass finds, coins and rings from graves under presentday densely built Cumhuriyet Mahallesi, Vezirköprü. For the location of other Neoklaudiopolis cemeteries, see Bekker-Nielsen 2013a: 211. Precious finds have also been recoverd from Niksar/ Neokaisareia (Özcan 1991), Samsun/Amisos (Ünan 2013: 398–401) and Amasya/Amaseia (Takaz 1975; Yüce 1995; Yüce and Özdemr 2000; Özdemr and Doğanbaş 2011). 51 Personal communication, Rainer Czichon. 52 Czichon et al. 2006: 171. 53 For survey finds indicating cemeteries, tumuli and hamlets dating to the Roman and Byzantine periods, see Czichon et al. 2006: 171–5; for epigraphic evidence indicating Roman settlements, see Olshausen and Biller 1984, s.v. Avdan and s.v. Kızılcaören; for identification of Neapolis/ Neoklaudiopolis, see Bekker-Nielsen 2013a; 2013b. Other inscriptions from the chora of Neoklaudiopolis are currently being prepared for publication (Eckart Olshausen/Vera Sauer, personal communication).

160

Pavol Hnila

7. Implications for the settlement pattern The study of the necropolis at Oymaağaç Höyük has several significant implications for the understanding of the settlement pattern of the region, in terms of both content and method. I shall conclude this preliminary study by enumerating the most important among them. 1. The earliest possible date for the beginning of burial activity at Oymaağaç Höyük either slightly predates the founding of the neighbouring city of Neapolis (later Neoklaudiopolis) by Pompey the Great or is roughly contemporary with it. 2. The necropolis (and, necessarily, also the still unidentified corresponding settlement) demonstrates a considerable level of continuity over 700–1,000 years. The necropolis activity ceases in the middle Byzantine period. 3. The community using the necropolis was small and its units were likely family-based. There were fluctuations in population size, and the population seems to have diminished drastically in Byzantine times. 4. There may have been multiple communities using multiple necropoleis in the territory of one and the same ancient settlement. 5. There are small differences between the burials in the necropolis (types of graves, their orientation, the presence or absence of grave goods) which may reasonably be related to social differences within the community. 6. Given the simple grave types and the scarcity of finds, the rural population using the necropolis seems to have been of lower economic status (or, at least, not particularly prone to manifestations of conspicuous consumption in funerary ritual) when compared with the city populations of the same period. 7. The mass graves imply a non-violent catastrophic event in ancient Oymaağaç, most likely an epidemic or a famine, between the second and the fourth century. Unless the catastrophe was fairly local, mass graves should also be present in other contemporary rural and urban necropoleis in Pontus. Given the widespread use of multi-generational family tombs, only a combination of the utmost care during excavation and the anthropological expertise is going to distinguish between the multiple burials and mass burials within the same tomb. 8. The presence of tiles is often interpreted by survey projects as an indication of settlements. Yet, tiles were also used as building material for graves from the Archaic period up to Ottoman times, and tile graves were very common, especially in the Roman and Byzantine periods.54 The presence of tiles can thus lead to an interpretation bias and to overrepresentation of settlements in survey distribution charts. At Oymaağaç Höyük, there was no proof of a necropolis before the actual excavation began – no funerary stelae and no bones, only pottery, stones and tiles were observed during the preceding survey.55 As a consequence, special care should be taken when analysing settlement patterns, density and dynamics. In particular, if two or more contemporary sites are noted as being very close to each other during micro-regional 54

Graves covered with tiles dating to the Ottoman period are commonly found at Pergamon: Rheidt 1991: 225. 55 Local farmers probably did notice human bones after ploughing, but did not inform archaeologists.

Rural necropoleis and settlement dynamics

161

surveys, there is an augmented probability that at least one of them was actually a necropolis. 9. Tiles are often difficult to date and without a proper methodology they can easily end up undated or falsely dated in survey and excavation reports.56 At Oymaağaç, the tile graves are not accompanied by any grave goods. In an archaeological study based on survey alone, all the tile fragments would have been almost inevitably linked to the latest datable surface finds, and thus have ended up with a middle Roman date. Yet, according to the C14 sample of the bones and according to the tile parallels, a ninthcentury date for most of the Oymaağaç tile graves seems more appropriate. I personally suspect that a number of Roman and Byzantine necropoleis might have escaped recognition during surveys for very similar reasons – due to the interpretative merging of an earlier settlement with an unrecognised later necropolis. For example, at the site of Zindankaya (15km southwest of Oymaağaç), the tiles published in the survey report look identical to the Oymaağaç middle to late Byzantine examples, but, because of other surface finds, they were interpreted as the remains of a wooden temple and dated to the Late Iron Age to Hellenistic period.57 Since the Zindankaya site has only been surveyed and not excavated, it is not possibile to check the respective interpretations and dates; we are left to operate with only assumptions and analogies. The potential of rural necropoleis across Anatolia to aid in the study of landscape dynamics and settlement patterns is underestimated. One of the reasons for this is that the rural necropoleis of the Roman and Byzantine periods only rarely receive the attention they deserve. Usually, they remain either in the shadow of better-studied urban necropoleis or they receive collateral treatment by projects focused on earlier periods. Only very rarely do the treatments include demographic data.58 Many Stone Age and Bronze Age sites in Anatolia excavated in recent years – for example Troy,59 Boğazköy,60 Çatalhöyük,61 Kayalıpınar,62 Ilıpınar63 and Barçın,64 to name just a few – have also included Roman and/or Byzantine rural necropoleis. It is to be expected that Roman and Byzantine rural necropoleis will continue to appear on prehistoric sites excavated in the future, so it is evident that here lies a huge opportunity. It is imperative to stipulate the need for the equal treatment of all periods – in terms of carefully excavating, processing, studying and swiftly publishing all the relevant materials – and to develop cooperation between projects with methodologies calibrated for rural necropoleis – both surveys and excavations. Especially the physical 56

57 58 59 60 61 62 63 64

For the problems associated with dating tiles recorded during surveys in Anatolia, see also Matthews et al. 2009: 173. For an innovative dated typology of Romano-British tiles, see Warry 2006a; 2006b. Dönmez 2007: 146–7, figs 10–12. For an overview of Byzantine cemeteries, see Kollig et al. 2010. Böhlendorf 1998. Kühne 1969. Kwiatkowska 2009. Zeiler 2006; Salzmann 2009. Roodenberg 2009: 154–7. Roodenberg 2009: 157–8.

162

Pavol Hnila

anthropology needs to be better integrated and play a more substantial role in the future research projects. I offer thanks to the editors of this volume for their invitation to contribute, and to Kristina WintherJacobsen in particular for her patience and support. I also sincerely thank Rainer Czichon for his encouragement to present these first thoughts and Jörg Klinger for permitting me to publish the Oymaağaç material. My warmest thanks go to Kathryn Marklein who has enthusiastically discussed the anthropological evidence with me. Pavol Hnila Freie Universität Berlin Institut für Altorientalistik Fabeckstraße 23–25, 14195 Berlin, Germany [email protected]

Bibliography Bekker-Nielsen, T. 2013a. ‘Neapolis-Neoklaudiopolis: a Roman city in northern Anatolia’, in Ü. YalçIn (ed.), Anatolian Metal VI (Der Anschnitt, Beiheft 25). Bochum, 203–14. Bekker-Nielsen, T. 2013b. ‘350 Years of Research on Neoklaudiopolis (Vezirköprü)’, Orbis Terrarum 11, 3–31. Bekker-Nielsen, T. and C. Høgel 2013. ‘Three Epitaphs from the Vezirköprü Region’, Epigraphica Anatolica 45, 153–60. Berti, F. 2012. ‘Grave goods from the Necropolis in the Agora of Iasos’, in B. Böhlendorf-Arslan and A. Ricci (eds), Byzantine small finds in archaeological contexts (Byzas 15). Istanbul, 187–211. Böhlendorf, B. 1998. ‘Byzantinisches Gräberfeld in Troia’, Studia Troica 8, 263–73. Böhlendorf-Arslan, B. 2012. ‘Das bewegliche Inventar eines mittelbyzantinischen Dorfes: Kleinfunde aus Boğazköy’, in B. Böhlendorf-Arslan and A. Ricci (eds), Byzantine small finds in archaeological contexts (Byzas 15). Istanbul, 351–68. Czichon, R. M. 2013. ‘Oymaağaç Höyük/Nerik (?)’, in M. Doğan-Alparslan and M. Alparslan (eds), Hititler: Bir Anadolu İmperatorluğu; Hittites: An Anatolian Empire. Istanbul, 298–309. Czichon, R. M., M. Flender, J. Klinger, V. von Seckendorf and H. Kürschner 2006. ‘Interdisziplinäre Geländebegehung im Gebiet von Oymaağaç-Vezirköprü/Provinz Samsun’, Mitteilungen der Deutschen Orient-Gesellschaft zu Berlin 138, 157–76. Czichon, R. M., J. Klinger, P. Breuer, J. Eerbeek, S. Fox, E. Marinova-Wolff, H. Marquardt, H. von der Osten-Woldenburg, S. Reichmuth, S. Riehl and T. Johannsen 2011. ‘Archäologische Forschungen am Oymaağaç Höyük/Nerik (?) in den Jahren 2007–2010’, Mitteilungen der Deutschen Orient-Gesellschaft zu Berlin 143, 169–250. Dönmez, Ş. 2007. ‘Zindankaya – the largest known Iron Age settlement in the Central Black Sea region, Turkey’, in: P. Kitzler (ed.), Third International Congress on Black Sea Antiquities in Prague, 11–18 September 2005 (Eirene 43), 143–60. Eiring, L. J. 2001. ‘The Hellenistic Period’, in J. N. Coldstream, L. J. Eiring and G. Forster (eds), Knossos pottery handbook: Greek and Roman. London, 91–135. Filippenko, A. A. 1998. ‘Stroitelnyye keramicheskiye materialy s punktov dislokaciyi rimskikh voysk v Kadykovke i na vysote Kazatskoy’ (=Building ceramic materials from the points of dislocation of Roman army at Kadykovka and Kazatskaya Hill), in M. I. Zolotarev (ed.), Khersonesskiy sbornik 9. Sbornik nauchnykh statey. Sevastopol, 110–117. Fox, S. C. and K. Marklein 2013. ‘Foot for Thought? Contextualization of os tibiale externum from two Roman period mass graves at Oymaağaç Höyük, Turkey (Abstracts – AAPA Presentations)’, American Journal of Physical Anthropology 150 (S56), 125. Fox, S. C. and K. Marklein 2014. ‘Primary and secondary burials with commingled remains from archaeological contexts in Cyprus, Greece, and Turkey’, in A. J. Osterholtz, K. M. Baustian and D. L. Martin (eds), Commingled and disarticulated human remains: working toward improved theory, method, and data. New York, 193–211.

Rural necropoleis and settlement dynamics

163

Goldman, A. L. 2007. ‘The Roman-period cemeteries at Gordion in Galatia’, Journal of Roman Archaeology 20, 299–320. Groß-Albenhausen, K. 2005. ‘Seuchen im 3. Jh. – ein methodisches Problem’, in M. Meier (ed.), Pest. Die Geschichte eines Menschheitstraumas. Stuttgart, 78–85. Jessee, E. and M. Skinner 2005. ‘A typology of mass grave and mass grave-related sites’, Forensic Science International 152.1, 55–9. Kollig, I. C., M. J. J. Jacinto Fragata and K. W. Alt 2010. ‘Anthropologische Forschungen zum byzantinischen Reich: ein Stiefkind der Wissenschaft?’, in F. Daim and J. Drauschke (eds), Byzanz – das Römerreich im Mittelalter. Teil 1. Welt der Ideen, Welt der Dinge. Mainz, 479–91. Krsmanovic, D. and W. Anderson 2012. ‘Paths of the dead: interpreting funerary practice at Roman-period Pessinus, central Anatolia’, Melbourne Historical Journal 40.2 (The Amphora Issue), 58–87. Kühne, H. 1969. ‘Die Bestattungen der hellenistischen bis spätkaiserzeitlichen Periode’, in K. Bittel (ed.), Funde aus den Grabungen 1967 und 1968 (Boğazköy IV). Berlin, 35–45. Kwiatkowska, M. 2009. ‘Byzantine and Muslim Cemeteries at Çatalhöyük – An outline’, in T. Vorderstrasse and J. Roodenberg (eds), Archaeology of the countryside in medieval Anatolia (PIHANS 113). Leiden, 129–38. Marek, C. 2003. Pontus et Bithynia: die römischen Provinzen im Norden Kleinasiens. Mainz. Marklein, K. E. 2014. ‘Runs in the Family: Osteoarthritis among Biologically Related Individuals in Roman Turkey (Abstracts – AAPA Presentations)’, American Journal of Physical Anthropology 153 (S58), 177. Matthews, R., M. Metcalfe and D. Cottica 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, 173– 226. Mitsopoulos-Leon, V. 1991. Forschungen in Ephesos 9.2:2. Die Basilika am Staatsmarkt in Ephesos. Kleinfunde 1: Keramik hellenistischer und römischer Zeit. Wien. Mitsopoulos-Leon, V. 2007. ‘Die Lampen’, in V. Mitsopoulos-Leon and C. Lang-Auinger (eds), Forschungen in Ephesos 9.2:3. Die Basilika am Staatsmarkt in Ephesos 2: Funde der klassischen bis römischen Zeit. Wien, 64–113. Neustupný, E. 1983. ‘The demography of prehistoric cemeteries’, Památky archeologické 74, 7–34. NOAA Solar Calculator 2014. ‘NOAA Solar Calculator (U.S. Department of Commerce | National Oceanic and Atmospheric Administration Earth System Research Laboratory | Global Monitoring Division)’, ESRL Global Monitoring Division – GRAD Group, http://www.esrl.noaa.gov/gmd/grad/ solcalc/ (accessed on March 25th 2014). Nowotny, E. 2011. ‘Mehrfachgräber im Gräberfeld von Thunau, Obere Holzwiese’, Archeologické Rozhledy 63, 443–65. Olshausen, E. and J. Biller 1984. Historisch-geographische Aspekte der Geschichte des Pontischen und Armenische Reiches. 1, Untersuchungen zur historischen Geographie von Pontos unter den Mithradatiden. (Tübinger Atlas des Vorderen Orients, Beihefte B 29.1). Wiesbaden. Özcan, B. 1991. ‘Fouilles de la nécropole de Niksar (1982–1987), avec un appendice sur les trouvailles monétaires par B. Özcan et B. Rémy’, in B. Rémy (ed.), Pontica I. Recherches sur l’Histoire du Pont dans l’Antiquité (Varia Anatolica 15). Saint-Étienne, 39–59. Özdemr, C. and M. Doğanbaş 2011. ‘Amasya Merkez Şamlar mahallesi 1079 ada 10 parselde yapılan kurtarma kazısı’, 19. Müze Çalışmaları ve Kurtarma Kazıları Sempozyumu, 149–58. Paine, R. R. and R. Vargiu 2010. ‘Anthropological studies’, in E. Equini Schneider (ed.), Elaiussa Sebaste III. L’agora romana. Istanbul, 259–84. Poulou-Papadimitriou, N., E. Tzavella and J. Ott 2012. ‘Burial Practices in Byzantine Greece: archaeological evidence and methodological problems for its interpretation’, in M. Salamon, M. Wołoszyn, A. Musin and P. Špehar (eds), Rome, Constantinople and newly-converted Europe: archaeological and historical evidence (U Źródeł Europy Środkowo-Wschodniej 1). Warszawa, 377– 428. Reimer, P. J., M. G. L Baillie, E. Bard, A. Bayliss, J. W. Beck, P. G. Blackwell, C. Bronk Ramsey, C. E. Buck, G. S. Burr, R. L. Edwards, M. Friedrich, P. M. Grootes, T. P. Guilderson, I. Hajdas, T, J. Heaton, A. G. Hogg, K. A. Hughen, K. F. Kaiser, B. Kromer,

164

Pavol Hnila

F. G. McCormac, S. W. Manning, R. W. Reimer, D. A. Richards, J. R. Southon, S. Talamo, C. S. M. Turney, J. van der Plicht and C. E. Weyhenmeyer 2009. ‘IntCal09 and Marine09 radiocarbon age calibration curves, 0–50,000 years cal BP’, Radiocarbon 51.4, 1111–50. Rheidt, K. 1991. Altertümer von Pergamon 15.2. Die Stadtgrabung 2: Die Byzantinische Wohnstadt. Berlin. Rife, J. L. 2012. Isthmia 9. The Roman and Byzantine graves and human remains. Princeton. Romančuk, A. I., 2005. Studien zur Geschichte und Archäologie des byzantinischen Cherson (Colloquia Pontica 11). Leiden. Roodenberg, J. 2009. ‘The Byzantine graveyards from Ilıpınar and Barcın in northwest Anatolia’, in T. Vorderstrasse and J. J. Roodenberg (eds), Archaeology of the countryside in medieval Anatolia (PIHANS 113). Leiden, 154–67. Salzmann, C. 2009. ‘Die römisch-frühbyzantinischen Bestattungen von Kayalıpınar’, in A. MüllerKarpe, V. Müller-Karpe, E. Rieken, T. Mühlenbruch, C. Salzmann, M. Zeiler and J. Wangen, ‘Untersuchungen in Kayalıpınar und Umgebung 2006–2009’, Mitteilungen der Deutschen Orient-Gesellschaft 141, 216–21. Sprague, R. 2005. Burial terminology: a guide for researchers. Lanham. Stathakopoulos, D. C. 2004. Famine and pestilence in the Late Roman and Early Byzantine empire: a systematic survey of subsistence crises and epidemics (Birmingham Byzantine and Ottoman monographs 9). Aldershot. Takaz, H. 1975. ‘28.12.1973/26.1.1974 Amasya Kulistepe nekropol kazısı ön raporu’, Türk arkeoloji dergisi 22.1, 109–15. Toteva, G. D. 2007. Local cultures of Late Achaemenid Anatolia (Ph.D. thesis from the University of Minnesota). Ann Arbor. http://search.proquest.com/docview/304840297 (accessed on January 22nd 2015). Ünan, S. 2013. ‘Tunç Çağı’ndan Roma Dönemi’ne Amisos ve Çevresinde Yer Alan Mezar Tipleri ve Ölü Gömme Adetleri Üzerine Genel Bir Değerlendirme. A General Evaluation of the Tomb Types and Burial Customs in and around Amisos from the Bronze Age to the Roman Period’, in Ş. Dönmez (ed.), Güneş Karadeniz’den doğar. Sümer Atasoy’a Armağan Yazılar. Lux ex Ponto Euxino. Studies Presented in Honor of Sümer Atasoy. Ankara, 385–412. Warry, P. 2006a. ‘A dated typology for Roman roof-tiles (tegulae)’, Journal of Roman Archaeology 19, 246–65. Warry, P. 2006b. Tegulae: manufacture, typology and use in Roman Britain (BAR British Series 417). Oxford. Wittwer-Backofen, U. 1987. ‘Anthropologische Untersuchungen des byzantinischen Friedhofs Boğazköy-Hattusa’, IV. Araştırma Sonuçları Toplantısı, Ankara 26–30 Mayıs 1986, 381–99. YIlmaz, M. A. 2014. ‘Contributions to the Early Iron Age Problem in the Central Black Sea Region in Light of Vezirköprü/Oymaağaç Höyük Ceramics’, TÜBA-AR (Türkiye Bilimler Akademisi Arkeoloji Dergisi) 15, 2012, 69–78. Yüce, A. 1995. ‘Amasya Merkez Eski Şamlar mezarlığı 1993 yılı kurtarma kazısı’, V. Müze Kurtarma Kazıları Semineri, 25–28 Nisan 1994 Didim, 1–16. Yüce, A. and C. Özdemr 2000. ‘Amasya ili Merkez ilçe Uygur beldesi kurtarma kazısı’, X. Müze Kurtarma Kazıları Semineri, 26–28 Nisan 1999 Kuşadası, 49–58. Zeiler, M. 2006. ‘Die Bestattungen auf dem Südosthügel von Kayalıpınar’, in A. Müller-Karpe, V. Müller-Karpe, E. Rieken, W. Sommerfeld, G. Wilhelm and M. Zeiler, ‘Untersuchungen in Kayalıpınar 2005’, Mitteilungen der Deutschen Orient-Gesellschaft 138, 236–45. Zhuravlev, D. V. 2010. Krasnolakovaja keramika Jugo-Zapadnogo Kryma I–III vv.n.e.: po materialam pozdneskifskich nekropolej Bel’bekskoj doliny (Materialy po archeologii, istorii i etnografii Tavrii = Materials in archaeology, history and ethnography of Tauria: Supplementum 99). Simferopol.

Vera Sauer

Two coins from Oymaağaç Höyük Appendix to Rural necropoleis and settlement dynamics by P. Hnila

1. Introduction Two coins came to light in the fill of two graves during excavations undertaken by the Oymaağaç Project at Oymaağaç Höyük in 2008 and 2014.1 Both coins are poorly preserved and cannot be identified with certainty. However, there is some evidence that one of the coins (A) was minted under the name of Pharnakeia during the reign of Mithradates VI Eupator (120–63 BC).2 It can be determined with a fairly high degree of probability that the other coin (B) was issued by Amaseia in AD 112/3 and provided with a countermark by the same city in 127/8.3

2. Coin (A): Oymaağaç Höyük, find 7385:025:002 The head of a Tyche with a mural crown can be detected on the obverse of this coin (fig. 1). The traces of the reverse design as can be detected from the image – in particular, a standing figure – are at least partly due to corrosion and cannot be identified with certainty as remnants of the coin decoration or the expected legend. The closest parallel to the Tyche design is to be found on coins of Pharnakeia dated to the reign of Mithradates VI (fig. 2).4 Slightly puzzling, however, is that, in contrast to the parallel examples, such as that from the collection of the American Numismatic Society (i.e. fig. 2), no locks of hair can be detected falling down the neck of Tyche on the coin from Oymaağaç Höyük; this difference cannot be explained convincingly by the varying degrees

1 2

For the find circumstances see Hnila this volume. Eckart Olshausen came to this view already in 2008 on the occasion of a first tentative assessment (Czichon et al. 2011: 196). Julie Dalaison – with similar caution – shares this view and, according to her, it is highly probable that the coin dates to the first century BC even if it cannot be assigned to the coinage of Pharnakeia (personal communication). On the bronze coinage of the reign of Mithradates VI, which in all probability was only pseudo civic, see Højte 2009: 98–9. On Pharnakeia, see Wilson 1960: 248–51; Wojan 2003: 258–67. 3 On Amaseia and its bronze coinage during the Roman period, see Dalaison 2008. 4 American Numismatic Society no. 1944.100.41429 (http://numismatics.org/collection/1944.100. 41429; accessed March 2015); cf. also Rec. gén. 139 no. 5 (supplementary table L nos 17, 18); SNG BM no. 1286.

166

Vera Sauer

Fig. 1: Oymaağaç Höyük, find 7385:025:002 (photo by H. Marquardt).

Fig. 2: American Numismatic Society no. 1944.100. 41429 (photo and ©American Numismatic Society). Obverse: Tyche (head with mural crown). Reverse: Zeus (standing upright, to the left, leaning on a long sceptre with the upraised left hand, base line; SNG BM no. 1286 indicates a crown in the outstretched right hand), legend: ΦΑΡΝΑ-ΚΕΙΑΣ. of preservation between these two coins.5 The coin from Oymaağac Höyük also differs from the parallels in terms of its diameter and weight (16mm; 4.35g).6 Such deviations are present, however, within the range of Mithradatic bronze coins.7

5

However, compare the two coins listed in Rec. gén.: Rec. gén. 139 no. 5 (supplementary table L nos 17, 18). In the case of no. 18 the hair locks are pronounced, whereas in the case of no. 17 only traces can be detected. 6 American Numismatic Society no. 1944.100.41429: 22.4mm, 7.2g; Rec. gén. 139 no. 5 (supplementary table L no. 17): 20mm; SNG BM no. 1286: 20mm, 5.47g. 7 Many examples can be found in Olshausen et al. 2009. Just to give one example: of all the coins recorded here which belong to the type Ares/sword (without monogram) and were issued under the name of Amisos, the smallest has a diameter of 18mm (no. VI 48), the largest 23.2mm (no. VI 106); the lightest coin weighs 4.61g (no. VI 62), the heaviest 9.10g (no. VI 117).

Two coins from Oymaağaç Höyük

167

3. Coin (B): Oymaağaç Höyük, find 7584:004:003 No images or legends are identifiable on this coin. However, clearly identifiable is a deeply impressed rectangular countermark with the letters rho and lambda written as Ρ and λ respectively (fig. 3). Despite its poor state of preservation, the countermark enables us to assign the coin to a mint with a fairly high degree of probability and thereby to date it. To my knowledge, a countermark with this design is demonstrated exclusively for coins issued by Amaseia (fig. 4),8 and, according to the reverse legends of such coins, they were all minted in the 115th year of the city’s era, i.e. 112/3.9 Further, in all probability, coins tagged with the countermark Ρλ can be assigned to one denomination.10 In the emission in question the diameter of the assaria ranges between 19mm and 23mm. The diameter of the coin from Oymaağaç Höyük (19.8mm) fits well within this range. Its weight (4.35g)11 is also appropriate (considering that the coin lost a significant proportion of its original volume due to abrasion and corrosion), given that parallels weigh between 4.38g and 8.61g. If one accepts the explanation of the meaning of the letter sequence Ρλ, which Julie Dalaison plausibly favours,12 the letters must be read as a number. Thus Ρλ should be understood as ‘year 130 according to the city era’. Assuming the coins were tagged in Amaseia,13 the countermark indicates the year 127/8. It is not impossible that the coin from Oymaağaç Höyük has a second countermark. The ‘suspect’ area has been highlighted in fig. 5, but it can be seen better on the coin than on the photograph. This can be taken as a further indication of the correctness of the suggested identification of the coin as, without exception, all the coins countermarked Ρλ which have emerged up to now bear a second countermark. The reading of the second countermark is, however, problematic. Possibly it consists of the Latin letters A and S.14 Vera Sauer University of Southern Denmark Mühlweg 6, 72414 Rangendingen, Germany [email protected] 8

9

10 11 12 13 14

Cf. Julie Dalaison who, on the countermarks on coins issued by Amaseia in general, states: ‘la comparaison avec les autres ateliers de la région n’apporte rien, puisque les types de contremarques présents à Amaseia lui sont spécifiques’ (Dalaison 2008: 152, n. 39). The tendency evident from the coin corpora of other cities in the region (cf. Sauer this volume: n. 1) confirms this evaluation. Notably, analysis of evidence from beyond the region in question by Christopher J. Howgego points in the same direction (Howgego 1985). Dalaison 2008: nos 34a, 34c, 36a. The reverse legend reads ΑΜΑCcωΝ cΤΟΥC cΙΡ. The city era of Amaseia started in 3/2 BC (Leschhorn 1993: 115–24, 466–9). The countermark is most easily identified on no. 36a. Having compared the mark of this coin with that of the coin from Oymaağaç Höyük, I believe the marking of both coins was executed with the same countermarking die. Cf. Dalaison 2008: 26, 33. This is not a mistake: both coins from Oymaağaç Höyük weigh 4.35g. Dalaison 2008: 153. Such an assumption is actually justified: Dalaison 2008: 153 (with references). On this and why the coins might have been provided with these countermarks, see Dalaison 2008: 33, 152–3.

168

Vera Sauer

Fig. 3: Oymaağaç Höyük, find 7584:004:003 (photo by H. Marquardt).

Fig. 4: Countermarked coin issued by Amaseia (after Dalaison 2008: no. 36a; Ireland and Ireland 2000: no. 109).

Fig. 5: Oymaağaç Höyük, find 7584: 004:003 (photo by H. Marquardt).

Two coins from Oymaağaç Höyük

169

Bibliography Abbreviations Rec. gén.: Waddington, W. H., E. Babelon and T. Reinach, Recueil général des monnaies grecques d’Asie Mineure. Paris, fascicles 1.1 (second edition, 1925); 1.2 (1908); 1.3 (1910); 1.4 (1912); reprinted New York 1976. SNG BM: Sylloge Nummorum Graecorum – [Great Britain], 9 British Museum (London), 1: The Black Sea. London 1993.

Literature Czichon, R. M., et al. 2011. ‘Archäologische Forschungen am Oymaagac Höyük/Nerik (?) in den Jahren 2007–2010’, Mitteilungen der Deutschen Orient-Gesellschaft zu Berlin 143, 169–250. Dalaison, J. 2008. L’atelier d’Amaseia du Pont: Recherches historiques et numismatiques (Numismatica Anatolica 2). Bordeaux. Howgego, C. J. 1985. Greek Imperial Countermarks. Studies in the Provincial Coinage of the Roman Empire (Royal Numismatic Society, Special Publication 17). London. Højte, J. M. 2009. ‘The Administrative Organisation of the Pontic Kingdom’, in: J. M. Højte (ed.), Mithridates VI and the Pontic Kingdom (Black Sea studies 9). Aarhus, 95–107. Ireland, S. and H. Ireland 2000. Greek, Roman, and Byzantine Coins in the Museum at Amasya (Ancient Amaseia), Turkey (Royal Numismatic Society, Special Publication 33. British Institute of Archaeology at Ankara, Monograph 27). London. Leschhorn, W. 1993. Antike Ären: Zeitrechnung, Politik und Geschichte im Schwarzmeerraum und in Kleinasien nördlich des Tauros (Historia Einzelschriften 81). Stuttgart. Olshausen, E., unter Mitarbeit von Studierenden. 2009. Bronzemünzen aus der Zeit Mithradates’ VI. im Museum von Samsun (Geographica Historica, Beiheft 1). Stuttgart. Wilson, D. R. 1960. The Historical Geography of Bithynia, Paphlagonia and Pontus in the Greek and Roman Periods. A new Survey with particular Referencce to Surface Remains still visible (Bachelor thesis, Oxford University). Wojan, F. 2003. ‘Kérasonte du Pont sous l’Empire romain: Étude historique et corpus monétaire’, Revue numismatique 159, 257–90.

Julia Koch

From funerary doorstones in Pompeiopolis to tracing local identity from the Phrygian Highlands to Inner Paphlagonia

Tombs, last homes and bulwarks of mortals, more trusted than homes for people, deposits of tears, the sole, remaining, incorruptible possessions of the dead; city of silence, one’s own house, the abiding bed on which the form is laid, interring beauty. Nor does it again recover from sleep but has become naked. What grave is near, and what body does it contain as inhabitant? Detested symbols of triumph over life, routed standards of the dead, stelae of the dead, speak with mute letters words of those who died.1

1. Introduction In the autumn of 1882 Gustav Hirschfeld travelled throughout Inner Paphlagonia collecting new epigraphic evidence in order to plot a small-scale but rather detailed description of northern Anatolia.2 Within the vast territory of Pompeiopolis he succeeded not only in recovering an honorary inscription dedicated by the council and people of the Paphlagonian metropolis but also in recording a tombstone ‘in Thürform mit Giebelabschluss, wie sie aus Phrygien bekannt sind (0.85m hoch, 0.52m breit). Am linken Thürflügel in Relief oben eine Kline, darunter ein Henkelkorb, am rechten ein Schloss, darunter ein Kreis, vielleicht ein Schild’.3 For decades, his results, which were presented to the Berlin Academy of Sciences, were hardly taken into account due to the lack of pictorial documentation.4 Accordingly, Paphlagonia was described until recently as a terra incognita at the maritime edge of the Roman Empire, even though the body of archaeological material illuminating the process of urbanisation in the rural hinterlands of the southern Black Sea coast has been growing ever since the initiation of the Pompeiopolis Project in 2006.5 By virtue of the 1 2 3 4

Funerary epigram of a pantomime found in coastal Roman Paphlagonia: Jonnes 1994: 10 no. 9. Hirschfeld 1888: 863–92. Hirschfeld 1888: 889 no. 62, 890 no. 65. See Lochman 2003: 151: ‘Die Türsteine sind … sowohl geographisch, als auch zeitlich eng begrenzt. Sie sind innerhalb Kleinasiens auf Phrygien und den angrenzenden Teil Westgalatiens beschränkt und stammen ausnahmslos aus der Kaiserzeit, vorwiegend aus dem 2. und frühen 3. Jh. n.Chr. Außerhalb dieser Zeit und Landschaft sind in Kleinasien keine echten Türsteine und Türstelen zu finden’. 5 According to the results of Project Paphlagonia, an archaeological survey conducted between 1997 and 2001, the modern-day Inner Paphlagonian city of Çankırı has revealed only scarce information about the urban setting and lifestyle of the inhabitants of ancient Gangra-Germanikopolis for the entire Imperial period. See Matthews et al. 2009: 182: ‘Archaeologically we know nothing about what must have been an extensive Roman and Byzantine town that grew up on the lower slopes of

172

Julia Koch

preceding Project Paphlagonia (1997–2001) and the Hadrianopolis Survey Project (2005– 2009) the inland part of north-central Turkey has firstly been recognized as a vigorous arena for cultural contact as indicated by grave monuments found in the modern provinces of Çankırı and Karabük.6 Meanwhile systematic surveys of recent years have gathered further archaeological evidence from the Paphlagonian metropolis of Pompeiopolis and its wide rural chora. This territory stretches from Mount Olgassys in the south to Küre Dağları in the north and includes the modern provincial capital city of Kastamonu, the neighbouring towns of Taşköprü and Boyabat to the east and the numerous small villages and farmsteads within the Devrekâni basin – an area that measures around 40km in width and 80km in length.7 Focusing on new evidence for the spreading of the so-called Phrygian doorstones into Inner Paphlagonia, its different sepulchral settings as public expressions of rising civic prestige will be discussed. As testified not only by epigraphic but more frequently by archaeological evidence, ancient economic activities within the Amnias river valley can be detected through a detailed examination of the visual language depicted on doorstone memorials honoring the newly emerging agrarian-oriented elite. For a reconstruction of the ancient Paphlagonian landscape inhabited by the emerging elite, the earliest contemporary travelogues of the 19th century provide the best evidence since Strabo and Pliny the Elder deliver only a brief glimpse into the different political structures and the perception of topography of both inland and coastal Paphlagonia. Considerations of the establishment of a new communication route system during the period of the pax romana provides a backdrop against which to trace a further expansion of cultural exchanges: looking at different types of sepulchral representations within the Paphlagonian inland cities of Hadrianopolis and Pompeiopolis and the nearby Euxine harbor ports during the High Imperial period, I contrast altering identities synchronically. Finally, within the territory of Pompeiopolis close cultural links to the Phrygian highlands will be revealed against the strong cultural influence of the old Greek Pontic harbor cities on the southern Black Sea coast in a diachronic perspective to further illuminate its close Inner Anatolian annexation.

5

the citadel. Its remains are buried still under the modern town but occasional construction work encounters scrap evidence in the forms of columns, tiles and sherds’. The current state of research is summed up by David French: ‘Research into the archaeology and history of Paphlagonia has been neither widespread nor detailed’ (French 2011: 1). Recent results of the Pompeiopolis Project are published in an interim report: Summerer 2011. 6 See Matthews 2007; Matthews and Glatz 2009; LaflI and Christof 2012a; 2012b. 7 The territory of Pompeiopolis has been identified by Marek 1993: 65–8, map 3. On the preliminary results of non-epigraphic survey finds, see Kunnert 2011: 207–11. Epigraphic surveys, associated with the Pompeiopolis Project, have been carried out by Christian Marek within the modern vilayet of Kastamonu, and investigated the Amnias river valley (2008), the valley of the Araç Çay (2009) and Devrekâni (2010). The title Metropolis Paphlagonias is evidenced by several honorary inscriptions dating from 169 to 195 (Marek 1993: cat. Pompeiopolis nos 3–4, 12) as well as on coins (Dalaison and Delrieux 2011: 143, cat. no. 1. pl. 1).

From funerary doorstones in Pompeiopolis to tracing local identity

173

2. Phrygian doorstones in Paphlagonia: new evidence from Pompeiopolis Four new doorstone monuments from the immediate surroundings of the Roman city centre of Pompeiopolis on Zımbıllı Tepe will be presented. These represent three different designs: built doorstone monuments with a multiple-door façade; freestanding monolithic door stelae; funerary doorstones delicately modelled to fit into burial chambers built of stone masonry. The first doorstone monument was found rebuilt into a brick village fountain in Afşar Köy (Fig. 1). Designed with a gabled roof imitating house architecture it strikingly resembles the second marble doorstone monument which was found in the vicinity of Taşköprü (Fig. 2). Crowned by palmette acroteria, both doorstones display a door knocker and a key plate in relief which imitate the essential components of a functioning twowinged door.8 Since the original setting is not recorded at Pompeiopolis, their use may be assumed on the basis of analogies with doorstone monuments from Phrygian Aizanoi which have been convincingly reconstructed by Kai Jes. In Aizanoi the earliest doorstone monuments, dating to the first century, represent gabled built doorstones, as testified by the tomb of Asklepiades II Charax, a prominent citizen who donated the city’s first marble temple.9 In the period of the pax romana more complex built door façades were constructed during the second and third centuries throughout northeastern Phrygia, as attested at Aizanoi but also at Bağlıca and Başören and Galatian Pessinous.10 At Pessinus, for example, an archigallos of Cybele’s cult as well as a highly esteemed paidonomos, who, as a member of the newly emerging urban elite of Roman Galatia, probably gave benefactions to his native city, were commemorated by a built door façade.11 A partly in situ preserved built door façade from Amorium is of particular significance amongst the total of approximately 800 Anatolian doorstones recorded by Marc Waelkens in the 1980s, since most were reused as marble spolia in late Roman necropoleis as well as in early Byzantine forts.12 As indicated by the built door façade from Amorium, this type of doorstone monument was originally placed above ground level and possibly served as a funerary room for inhumations.13 According to archaeological and – more particularly – epigraphic evidence, the most elaborate grave ensembles with door façades may have functioned as heroa and have incorporated sarcophagi within a surrounding enclosure.14 8

9 10 11 12

13 14

Having been exposed to the elements, the Afşar stele is in quite a poor state of preservation. A door knocker and a bronze key plate were found, for example, in Phrygian Aizanoi (Rheidt 2010: 33, fig. 23) and Lopodunum (Künzl 1998: 65, figs 1–2, pl. 1). See Jes 1997: fig. 6; Rheidt 2010: 17. A precise dating of the gabled doorstone monuments of Aizanoi is provided by epigraphic evidence. Drew-Bear and Lochman 1996: figs 3–4; Jes 2001: figs 19–20; Devreker et al. 2003: 121. Strubbe 2005: 84–7 no. 64, 143–4 no. 130. Waelkens 1986; for a sketch and photograph of the Amorium tomb, see Kelp 2013: 76–7, fig. 4.6–7. On the common reuse of doorstones in cist graves of the fourth and fifth centuries, see Devreker et al. 2003: 53, 118–23, fig. 103, cat. nos B8–B9; the reuse of doorstones for the construction of Byzantine fortresses is attested at Pessinus: Devreker et al. 2003: 350, fig. 194. Devreker et al. 2003: 121. During the first half of the second century in Galatian Pessinus, sarcophagi were deposited within enclosure walls made of built doorstone monuments; Strubbe 2005: 122 no. 105: ‘Mageianos son

174

Julia Koch

Fig. 1: (a) Brick-built village fountain with spolia of a Roman doorstone monument in Afşar Köy, Taşköprü Belediyesi, Vilayet Kastamonu in the hinterlands of Pompeiopolis. (b) Detail showing remnants of a door knocker and key plate in relief (photos by B. Maerzke). In addition to these exceptional built peribolos tombs, more modest door stelae were widely distributed within northern Phrygia and western Galatia. As already noted by Marc Waelkens, since a similar construction has been demonstrated in western Asia Minor for the pre-Roman periods, these Roman monolithic freestanding grave monuments may have

14

of Helios prepared for himself and for his wife Amia daughter of Heliodoros and for his mother Mandnis(?) daughter of Alexadros this sarcophagus with the enclosure’; Strubbe 2005: 138 no. 126 (doorstone of white marble): ‘Terentia Seleukis made [this tomb] for Terentios Sekoundos in memory and for herself while alive, and also the sarcophagus that lays inside. Let it be permitted to no-one to bury anyone else in the sarcophagus’. At Phrygian Aizanoi a built grave monument of the Antonine era housing a columnar sarcophagi was excavated in the early 1990s during rescue excavations: Türktüzün 1991; for a reconstruction, see Wulf 1993: fig. 5. The existence of monumental tomb buildings in Roman northern Anatolia is demonstrated by Hirschfeld 1888: 878.

From funerary doorstones in Pompeiopolis to tracing local identity

175

Fig. 2: Marble doorstone imitating house architecture and crowned by palmette acroteria from Taşköprü/Pompeiopolis. On the door wings are a key plate (top left) and a door knocker (top right) (photo by B. Maerzke). been erected at the dromos of tumuli.15 Perhaps having been set up as a ‘resting place while alive knowing the forgetfulness of heirs’,16 some door stelae from Pessinus were originally placed on a base stone or a plinth as a prestigious mark of arête, sometimes with an incised cupula for ritual libations.17 Crowned by a scallop shell and an anthemion frieze, a door stele from Pompeiopolis displays two knockers and a key plate coupled with a single pruning hook in the upper recessed door panel (Fig. 3). Another pruning hook is shown beside a vine on a third rectangular marble doorstone from Pompeiopolis; this last example seems to have been accurately modelled to fit into the complex sepulchral architecture of a burial chamber built of stone masonry (Fig. 4).18 15

Waelkens 1986: 3–4, 7, with evidence not only for door stelae possibly erected in front of Roman tumuli, but also for convex doorstones that were originally integrated into the krepis of Roman tumuli; on the pre-Roman evidence, see Roosevelt 2006: figs 12, 25 (İkiztepe tumulus), fig. 15 (Karaburun II tumulus), fig. 22 (Bin Tepe), fig. 23 (Kuşaklıtepe). 16 Strubbe 2005: 139–41 no. 127. 17 Lambrechts 1969: 136–7, pl. 5, fig. 11, pl. 6, fig. 13 (cupola with a drainage channel on a stone plinth); Devreker et al. 2003: 120–2, figs 48–9; Krsmanovic and Anderson 2012: 76 fig. 11– 12 (libation cupule stones found reused in a Late Antique building). As recorded by contemporary epigraphic evidence, imperial grave stelae served explicitly as markers of virtue: Matthews et al. 2009: 212. 18 There is scarce archaeological material related to Paphlagonian funerary practices. At Pompeiopolis, next to the settlement on Zımbıllı Tepe, the numerous burial chambers of mainly Byzantine rock-cut tombs of the arcosolia type have been mostly looted. During a 2013 survey, Ferhat Çevk and Max Ritter recorded around 100 graves of different types: Çevik and Ritter 2013: 169–82. In 2005 within the chora of Paphlagonian Hadrianopolis two Roman necropoleis were observed during the course of an archaeological field survey conducted by Ergün LaflI from the Dokuz Eylül University of Izmir. Among the rock-cut tombs and grave stelae, LaflI and his team recorded ‘eine tempelähnliche Grabanlage, ein prostyles Bauwerk, dessen sehr gut erhaltene Architekturteile in einer illegalen Grabung freigelegt wurden’: LaflI and Christof 2012b. The Roman

176

Julia Koch

Fig. 3: Door stele with tenon crowned by a scallop shell and an acanthus frieze from Taşköprü/Pompeiopolis. The stepped and recessed door panels feature door knockers, a key plate and a pruning hook (top left) (photo by M. Demirbaş).

3. Fishing, farming and viticulture in the Amnias Valley As Waelkens convincingly demonstrated in the late 1970s, Phrygian doorstones are a highly valuable source for understanding the social and economic history of Roman Anatolia.19 Helping to define both rural areas and the industrial zones of Inner Anatolia, sepulchral iconography reflects the main regional resources and the dominating regional occupations within Roman Phrygia: in the Phrygian highlands, on the one hand, and around Docimeion with its famous imperial marble quarries of Synnada, on the other.20

18

necropoleis recently discovered in Hadrianopolis have not been investigated in detail by an archaeological excavation campaign. 19 Waelkens 1977. 20 Waelkens 1977: 289.

From funerary doorstones in Pompeiopolis to tracing local identity

177

Fig. 4a–b: Marble doorstone featuring a door knocker and a pruning hook beside a vine branch Taşköprü/Pompeiopolis. On the rear is an accurate modelling for the fitting into an architectural grave ensemble (photo by B. Maerzke). On doorstones from Docimeion, for example, we find picks, hammers and regulae as the tools most often used by stone-cutters, and a lecticarius is represented by an ox-cart transporting marble blocks.21 In contrast, scattered evidence for trade and commercial occupations is known from depictions from central Phrygia of stored oil drums, a shoemaking awl with a cobbler’s last and a blacksmith’s tools.22 Across the Mediterranean, doorstones commemorated soldiers of the seventh Roman legion which was stationed first in Galatia and afterwards in Roman military camps around Tilurium, Bigeste and Salona during early Roman times.23 Therefore, in Dalmatia and Liguria the repertoire of decoration is closely restricted to military iconography, as shown by the depictions of a gladius and a pugio, by a bow and arrow and by a vexillum.24 In Anatolia, however, Roman doorstones communicate valuable information about the greatest food-production centres within Phrygia. Fishing in 21

Waelkens 1986: 191–2 no. 471, fig. 53 (regula), 201 no. 501, pl. 72, no. 502, pl. 74 (tools of a lapicida), 196–7 no. 486, pl. 75 (lecticarius). 22 Waelkens 1986: 169–70 no. 422, pl. 66 (stored oil drums), 168 no. 420, pl. 47 (a shoemaker’s tools), 165 no. 411 (a blacksmith’s tools). 23 Stephen Mitchell convincingly reconstructs the stationing of the legio VII in Galatia and its presence in the province until 13 BC. In AD 7 it was transferred to Pannonia and did not return to the east. Concerning the numerous early veterans who retired in Dalmatia, he conclusively states that ‘the vast majority are first-generation legionaries from the native communities of the eastern empire, above all Galatia and Macedonia’: Mitchell 1976: 304. 24 Lochman 2003: pl. 40, fig. 153 (gladius and pugio); Cambi 1987: pl. 46e (bow and arrow). According to Nenad Cambi, another grave stele was found within the garrison of the 11th Roman

178

Julia Koch

Fig. 5: Doorstone found within the eastern territory of Pompeiopolis near modern Boyabat (now held at the Sinop Museum). Pruning hook, in the lower right corner (photo by R. Hessing). the Sangarios river valley is indicated on doorstones from Amorium showing a cast-net, fumigation and a ready-to-eat plate of freshly-prepared Sangarios fish.25 In the upper Tembris valley, the great centres of agricultural production are identified by Phrygian votive and sepulchral monuments depicting specific decorative motifs: the plough – depicted either alone or with a pair of oxen yoked to it; the sickle; the winnowing fan or the stimulus with the spike used by the ploughman to spur his oxen.26 Finally, the presence of viticulture is well testified within the Dorylaion district, in the vicinity of Kütahya, in the neighbouring valley of Altıntaş and on the southeastern slopes of the Phrygian highlands.27 As presented above new evidence for viticultural activities has emerged for the Amnias river valley in the imagery of the Paphlagonian doorstones (Figs 3–4). Used for the pruning of vines, the falx vinitoria was the most distinctive tool of the vine-dresser’s occupation along with the dolabra, a versatile tool used to chop up and dig out old vines.28 The same implements, 24 25 26

27 28

legion in Burnum picturing a quiver with a bow and arrow: Cambi 1987: no. 104; cf. Lochman 2003: pl. 40, fig. 154 (vexillum). Waelkens 1986: no. 517, fig. 56, no. 554, pl. 80, no. 555, pl. 81. Imperial estates occupied vast territories during the period of the later Empire. Today, the modern province of Eskişehir (Dorylaion) is one of the largest farming regions and principal centres of grain production in Turkey: Waelkens 1977: 283–6. Phrygian votive stelae indicating the significance of agriculture are numerous: Drew-Bear et al. 1999. Waelkens 1977: 278–83. White 1967: 93–6 (falx vinitoria), 61–4 (dolabra).

From funerary doorstones in Pompeiopolis to tracing local identity

179

Fig. 6: Fragment of a doorstone found near modern Boyabat featuring a wool basket, a spindle and a distaff (now held at the Sinop Museum) (photo by R. Hessing). testifying to the exploitation of viticulture in the Amnias river valley, are depicted on previously published grave stelae from Pompeiopolis, one of them dated precisely to 240/1 by epigraphic evidence.29 Another doorstone found within the easternmost territory of Pompeiopolis near modern Boyabat may be dated stylistically to around the early third century (Fig. 5):30 On display are again the vine dresser’s tools, but also farm animals like a partridge as well as a donkey and the fumigation of fish. Thus, displayed on these monuments is substantial evidence for the economic importance of fishing, farming and viticulture within the fertile landscape of Roman Paphlagonia. Viticulture has widely disappeared from the persistently rural landscapes of northern Turkey nowadays, but the doorstones of Pompeiopolis – together with new finds of wine presses in western and eastern Paphlagonia – testify not only to the commercial impact of viticulture, but also to the growing social prestige of the newly emerging urban elites which was closely related to Paphlagonian wine production.31 In addition to viticulture, textile production also seems to have been of widespread importance across the Amnias river valley; this is implied again on doorstones, but also on

29

Jacopi 1936: 5 no. 2, pl. 3, fig. 9; Halfmann and Schwertheim 1986: 131 no. 2 (private collection: Turhan Güler, Istanbul; aedicula stele 240/1); Marek 1993: cat. Pompeiopolis no. 36; Marek 2003: 160, fig. 253 (stele from Hayucak near Taşköprü). 30 Cf. French 2011: 11–3, 23, cat. no. 12 for another grave stele from modern Boyabat epigraphically dating to 205/6. Both style and iconography (farm animals, vine branches with a bunch of grapes, a wool basket with spindles and distaffs) closely resemble to the newly published doorstone from modern Boyabat. 31 LaflI 2012: with evidence for the territory of Hadrianopolis in western Paphlagonia; Kahl 1992: with evidence for the territory of Neoklaudiopolis in eastern Paphlagonia. While the Paphlagonian wine presses are found ex situ no precise dating is possible based on the realia findings. However, according to literary sources the Pontic wine production has been very popular since the physician Apollodorus in his pamphlet advising King Ptolemy what wines to drink first of all praised the wine of Nasperkene in Pontus cf. Pliny 14.76; Strabo 12.3.36 speaks of the inhabitants of Pontic Komana as being voluptuous in their mode of life since all their property is planted with vines.

180

Julia Koch

other grave stelae types depicting wool baskets, spindles and distaffs (Fig. 6).32 The modern province of Kastamonu, including Taşköprü, is famous today for its textile production.33

4. Patterns of communication through Paphlagonian landscapes In the autumn of 1838 William Francis Ainsworth, a founding member of the Royal Geographical Society, left London to travel from Constantinople via Heraclea Pontica to Galatian Ancyra.34 After passing over the Bosphorus and through Bithynia he crossed the Parthenius river ‘to enter upon the no less interesting districts of Paphlagonia’.35 From the upland plains of Eflani and Safranbolu his expedition party descended towards the Amnias valley backed by the city of Kastamonu where the principal commercial activity was the wool trade; the inhabitants proudly insisted they produced textiles nearly as good as those of Ancyra.36 They continued their journey along the gardens of the Gök Irmak to its junction with the Kızıl Irmak or Halys … the united streams flow onwards by Tash Köpri and Boi-abad. …37 Beyond … the river stretched to the eastward along a pleasant valley, cultivated, full of villages, plantations, and gardens … It enters upon the valley of Tash Köpri, surrounded by rocks and mountains, but wooded and cultivated in its centre, and abounding in villages … A little more than an hour’s journey brought us to Boi-abad, a small town beautifully situated upon a tributary to the Gök Irmak … The town is rather scattered, which adds to its beauty, and occupies a valley to the east of the rivulet. The bed of the latter is filled with luxuriant gardens, full of fine fruit-trees over-run by vines.38

While Strabo – who was born in Pontic Amaseia shortly after the Roman annexation – tells us little about the fertile valley of the Amnias river running west to east through Inner Paphlagonia, this picturesque travelogue of the 19th century describes the fertile landscape at the northern foot of Mount Olgassys in great detail (Fig. 7):39 Noting numerous winding rivulets, abundant vegetation and villages, Ainsworth’s colourful depiction culminates in a description of the bountiful fruit gardens and vineyards still seen in the region of Boyabat and towards Sinop. A sharp contrast to the fertile Inner Paphlagonian landscape is drawn in 1842 by his compatriot William John Hamilton, a Fellow of the Royal Geological Society of 32

33 34 35 36 37 38 39

While the vast majority of Phrygian doorstones depicting textile tools are dedicated to females and demonstrate the representation of gender roles, a monument mentioning Papios and Manios, a father and son, has been recovered from northern Phrygia: Waelkens 1986: 114–5 no. 270, fig. 37. Nonetheless, even during the 19th century it was mainly the women of Kastamonu who worked in the cotton industry; copper processing, on the other hand, was a male-dominated field: Ainsworth 1839: 249. During the excavation campaign in Taşköprü the team members of the Pompeiopolis Project are housed in facilities of the former jute factory. Ainsworth 1839. Ainsworth 1839: 235. Ainsworth 1839: 240, 248–9. Ainsworth 1839: 250. Ainsworth 1839: 251–4. ‘The Olgassys is a very lofty mountain, and difficult to be passed … The country around, the Blaene, and the Domanitis, through which the river Amnias runs, is sufficiently fertile’ cf. Strabo 12.3.40.

From funerary doorstones in Pompeiopolis to tracing local identity

181

Fig. 7: The fertile valley of the Amnias river (the modern-day Gökırmak) near Boyabat, a section of the Pontic route crossing Inner Paphlagonia from east to west (author’s photo). London, with respect to the Paphlagonian coastal zone on the southern Pontic shores.40 The wild alpine scenery of the region between Boyabat and the sea port of Sinop is described, including a wooded mountain range stretching far to the east and west, and intersected only by deep, narrow valleys with several rivulets and rivers, steep glens and abrupt ravines. Due to its utter solitude, the harbor city of Sinop was characterised as an island that ‘can boast but little intercourse with the interior’; the city’s commercial activities and communications were consequently carried on via the sea.41 Thanks to Strabo, we do catch a brief glimpse of the political divisions at the maritime edge of the Roman Empire during the first century BC. According to his Geography, coastal Paphlagonia extended way along the southern Black Sea coast from the Halys river beyond Sinope and Amastris as far as Heraclea Pontica, while the interior of Paphlagonia reached to the Kızılırmak.42 A precise distinction was highlighted already during the pre-Roman period between the inland and the coastal parts of Paphlagonia since Mithradates Eupator 40 41 42 42

Hamilton 1842: 314–7. Hamilton 1842: 317. Strabo 12.3.26. ‘Mithradates Eupator possessed the maritime part as far as Heracleia, and of the inland country he had the district nearest to Heracleia, some parts of which extended even beyond the Halys … The remainder was subject to chiefs, even after the overthrow of Mithradates’ cf. Strabo 12.3.9.

182

Julia Koch

possessed only the coastal parts while the inhabitants of the interior were not subject to his Hellenistic Pontic Kingdom.43 After the defeat of Mithradates VI Eupator, the victorious Roman general Pompey the Great established a chain of new cities in the Pontic hinterland. They were all located along a grand military road running from Byzantium through Bithynia and the Amnias river valley via Pompeiopolis further east towards the Upper Euphrates frontier before crossing the Halys and passing in turn the newly founded Pontic poleis of Neoklaudiopolis, Magnopolis, Diospolis and Nikopolis.44 As Stephen Mitchell points out, this northern trunk road running from east to west was a massive upgrade for the Anatolian system of communication.45 At the same time, commercial circuit routes connected the Paphlagonian inland city of Pompeiopolis with the Euxine sea ports of Sinope and AbonouteichosIonopolis bypassing the eastern and western Küre Dağları.46 Important transit routes facilitated further access via Gangra-Germanikopolis and Neoklaudiopolis into Inner Anatolia:47 Starting from Pompeiopolis, a direct route, primarily attested by the Tabula Peutingeriana, ran through Gangra bypassing Mount Olgassys via Ilgazdağı Geçidi or, alternatively, Tosya before proceeding part way along the Devrez Çay valley straight towards the Galatian metropolis of Ancyra. Within the Paphlagonian inland cities of Gangra-Germanikopolis, Tosya and Pompeiopolis the tempting results of an enormous upgrade of the Inner Anatolian system of communication are archaeologically most strikingly mirrored in the adaption of the Phrygian (i.e. Galatian) doorstone habit.48 43 44

45 46 47

48

Corresponding to this, Ptolemy (5.4) writes in his Geography of the second century about the Paphlagonians that occupy the coast, on the one hand, and the interior, on the other hand. Milestones found in Galatia and Pontus point to several major construction periods within half a century of Vespasian’s accession (80–82, 97–100 and 119–122). See French 1988: 916–7 concerning the Roman network of Paphlagonia; map 18 lists milestones found west of Pompeiopolis (nos 503–4) and also milestones found near Boyabat in the easternmost territory of ancient Pompeiopolis (nos 903, 85). David French has recently published illustrations of milestones and distribution maps along with a conspectus of designated roads cf. French 2013. Apart from the milestones there are only few material traces of Pontic roads: ‘Only in the most desolate part of the Lycus valley, between Enderes and Niksar, is its dyke once or twice visible near a solitary fragment of a Roman bridge … Lastly there is the magnificent wreck of the Roman bridge on the Halys, which must have spanned the ordinary summer stream in a single arch over one hundred feet wide … Some remnants of pavement between this bridge and Narlu may possibly represent the old roadway. What traces may exist west of the Halys must be left for other travellers to discover’ cf. Munro 1901: 65. Mitchell 1993: 124–36, maps 3, 8–9 show the Anatolian network. See Belke 1996: 118 for route B5 of the Roman to Byzantine network; cf. Belke 1996: 134–5. Gangra-Germanikopolis: the course of this route close to the modern road network is demonstrated, for example, by remnants of a Roman pavement structure near Kisecik south of Ilgaz (Matthews et al. 2009: 189, fig. 6.27) and some milestones along the Gangra-Ankyra route (Belke 1984: 108; French 2012). Neoklaudiopolis: a topographical survey undertaken in 2010 in Samsun province by the Oymaağac-Nerik Project of the Free University in Berlin and the University of Southern Denmark revealed a section through the Roman roadbed near the village of Doyran made of large stones with superimposed layers of smaller stones and gravel (BekkerNielsen 2013: 203–13, figs 4–5; see also Bekker-Nielsen and Czichon this volume). See Waelkens 1986: no. 801, pl. 99 (Tosya); LaflI and Christof 2012a: fig. 37c–d (GangraGermanikopolis). At least one more funerary doorstone now held at the Çankırı museum is unpublished. Therefore I would like to sincerely thank Ursula Kunnert for kindly providing me information and private photo material during her museum visit in August 2014.

From funerary doorstones in Pompeiopolis to tracing local identity

183

5. Local identity in the rural hinterland of the Euxine: being Paphlagonian through territory and time In 6/5 BC the inland cities of Paphlagonia were incorporated into the Roman province of Galatia.49 Prior to this event, parts of the Paphlagonian interior had been presented to the descendants of Pylaimenes by Mark Antony.50 By establishing a koinon after its reintegration into the Roman Empire, the new administrative entity appointed its own provincial assembly headed by Pompeiopolis, the metropolis Paphlagonias.51 Although Pliny the Elder, through his contemporaneous urban Roman perspective, labelled the inhabitants of the first-century Paphlagonian cities as justly Greeks of Dorian, Ionian and Aeolian origin for all the rest are Barbarians,52 a more complex mélange of cultural influences may be traced by sampling distinct types of grave stelae widely distributed within northern Anatolia during the High Imperial period. In western Paphlagonia, including the territory of Caesarea-Hadrianopolis, funerary column stelae were the most commonly employed freestanding grave markers from the mid second century to the mid third century that were set up on the summits of tumuli in order to be seen from afar.53 These white marble or limestone columns are finished with a moulded top which possibly served as a platform for busts or statues of the deceased or for offerings of wine and food.54 A few grave columns were adorned with sculptured imagery reflecting rural life. Again, the deceased are characterised as viticulturists, traders and cloth merchants or epigraphically honoured as farmers. For example, after retiring from his military service a certain Priscus then turned his mind to ‘ploughing his native land, doing everything with them, he instructed the farmers that he had raised to do all those things which once Hesiod had explained in generous detail to farmers, how to reap great 49

50

51 52 53

54

Gangra-Germanikopolis, Neoklaudiopolis and Pompeiopolis date their coins from this event as era. In 4/3 BC the oath of allegiance was sworn by the Galatian Paphlagonians. From the time of Domitian until the early third century Galatian governors were responsible for the construction and maintenance of the Paphlagonian section of the road between Pompeiopolis and Neoklaudiopolis. Further evidence is given by L. Nonius Calpurnius Asprenas Torquatus known as legatus pro praetore provinciae Galateae, Paphlagoniae, Pamphyliae, Pisidiae (Marek 1993: 71–3). In the second century Ptolemy (5.4) wrote of the Olgassys mountain in the Roman province of Galatia. Pliny the Elder speaks of a gens Paphlagonia, quam Pylaemenia aliqui dixerunt, inclusam a tergo Galatia cf. Pliny 6.2. According to Pliny the people who inhabited the Galatian plateau ‘were sprung from the Gauls, and were called Tolistobogi, Voturi, and Ambitui’, while they that occupied the neighbouring parts of Paphlagonia ‘were named Trocmi’ cf. Pliny 5.32. Beyond Paphlagonian Galatia also Pontus and Nikopolis were presented to client kings by Mark Antony cf. Marek 1993: 49–50, 71–72. Marek 1993: 78. Pliny 6.2. Marek 1993: 100–1, cat. Kaisareia/Hadrianopolis nos 25, 26 (253), 27–8, 30 (205), 31 (220), 33, 36, 38, 40 (238), 41–3, 45, 46 (243), 47–69, 70 (223), 73, 75–7, 78 (152), 79, 86 (218), 89, 91. In the territory of neighbouring Pompeiopolis only one funerary column has been recorded: Marek 1993: cat. Pompeiopolis no. 52. An inscribed funerary column from Üçbaş (modern Karabük province) testifies to its position on top of a burial mound: Marek 1993: cat. Kaisareia/Hadrianopolis no. 67. Further epigraphic evidence is provided by an inscribed grave column from the territory of Bithynian Krateia-Flaviopolis (modern Gerede): Kokkinia 1999; see also LaflI and Christof 2012b: 166–9. Coulton 2005: 129.

184

Julia Koch

harvests’.55 Priscus – who had probably risen to the rank of a legionary standard bearer under Trajan or even held equestrian rank – is lavishly praised with analogies to both Hesiod’s Works and Days and Homer’s Iliad: he ‘accomplished all the feats which once Achilles and the son of Priam had achieved’. This type of small columnar tombstone is widely distributed west of Pompeiopolis, and the adoption of old Greek funerary practices can be observed. During the fifth and fourth centuries BC Athenian colonists arrived in Sinope, some of whom were commemorated with inscribed Doric grave columellae occasionally decorated with an elaborate egg-anddart moulding on the echinus of the capital.56 Known as a kioniskos, this was the dominant type of funerary monument in Attica during the Hellenistic period following the adoption of anti-sumptuary laws on burial practices issued by Demetrios of Phaleron in 317/6 BC.57 While the Roman-period Attic grave columellae of the Athenian necropolis at Kerameikos may be seen as a continuation of a Hellenistic Greek funerary tradition, their sudden appearance in the more remote parts of the Anatolian hinterland seems to reveal an Inner Paphlagonian revival of the Pontic columellae set up by the Athenian Sinopeans in the fifth to fourth century BC.58 Turning to the coastal region, the depiction of banquet scenes on sculpted pedimental stelae of the second century from Amastris, Abonouteichos and Sinope reveal again a strong Greek influence.59 In one example, the deceased is commemorated as a reclining symposiast named Damostratos and is accompanied by his wife, Parthenos.60 Additional cultural links to the Greek world are represented by the names of the phyles of Amastris,61 the renaming of Abonouteichos to Ionopolis62 and the preservation of old Greek dialects 55

56

57

58 59

60 61 62

Cf. Marek 1993: cat. Kaisareia/Hadrianopolis nos 41 (falx vinitoria), 86; Matthews et al. 2009: 206, figs 6.75–6: columnar stele of ‘Kyrittas, farmer, good friend, trader and clothes merchant’ (218); Marek 1993: 100–16, cat. Kaisareia/Hadrianopolis no. 79: the stele of the legionary standard bearer Priscus. R. Haensch (2005) suggests that Priscus may even have held equestrian rank. Plut. Pericles 20.1–2; French 1990; 1991; 2004: 14–31 no. 9–49, pls 4–8; Mitchell 2010: 90–1. According to French, the provenance of these columnar gravestones may be assumed to lie in an extramural necropolis 200m west of the city gate within an area named Kumkapı. Since this area is now occupied by two petrol stations, the Sinop bus station and government buildings, no archaeological research can be conducted. Jones 1988; see also Oliver 2000: 65–74, fig. 3.2. In the Athenian Kerameikos a columella belonging to a member of one of the wealthiest Athenian families was found within the peribolos of Lysimachides. Cicero’s On the laws informs us that Demetrios ‘imposed a limit on new tombs: for he did not want anything to be erected above the mound of earth unless it was a small column (columella) of three cubits and no higher or a table or a libation basin’ (2.25.66: nam super terrae tumulum noluit quicquam nisi columellam tribus cubitis ne altiorem aut mensam aut labellum). For the evidence of imperial Greek columellae, see Moock 1998: 53, pls 3, 7–8, 19, 62. Marek 1993: 67, cat. Amastris nos 48 (184), 49, 50 (111/2), 51, 88, 90 (132), pl. 29, fig. 1–3, cat. Abonuteichos/Ionopolis no. 5; see Pfuhl and Möbius 1977, with references, for Tium (no. 1743, pl. 252) and Heraclea Pontica (nos 1744–6, pls 252–3); for further evidence, not only from Heraclea Pontica, but on numerous Bithynian grave stelae from the Hellenistic period onwards, see also Cremer 1992. Marek 1993: cat. Amastris no. 49, pl. 29, fig. 3. Marek 1993: 97–8. Marek 1993: 98. The petition to change the city’s name to Ionopolis was presented to the emperor Marcus Aurelius by the Prophet Alexander, on behalf of the inhabitants of Abonouteichos.

From funerary doorstones in Pompeiopolis to tracing local identity

185

Fig. 8: Tumulus near Tepedelik. In the foreground is the modern road passing Mount Olgassys to the south and lined with numerous tumuli; in the background is the Amnias river valley bordered by the Küre Dağları to the north (author’s photo). not only in Sinope, but also in the Roman colonies across the entire Black Sea region.63 This all-encompassing Greek self-consciousness of the coastal Paphlagonians during the High Imperial period may best be expressed in the refreshing of historical links with their Greek homeland, while Ionian Miletus, for example, laid a new emphasis on its status as mother city in return.64 Within the Paphlagonian hinterland Greek influence can once again be attested in the territory of Pompeiopolis along the upper Amnias river valley.65 While only a single grave column, although not a single pedimental stele in high-relief displaying a Greek symposium scene, was found throughout the vast chora of the Paphlagonian metropolis,66 another type of funeral monument modelled on the Classical Attic grave stelae was widely distributed during the second and third centuries. This type consists of a slender stele carrying an acanthus-decorated panel with slightly concave vertical sides crowned by an elaborate anthemion with a palmette design.67 As stated recently by Stephen Mitchell, 63 64 65 66 67

Mitchell 2010: 97–8. Mitchell 2010: 110. For Attic grave stelae dating to the pre-Roman period, see French 2011: 4. Marek 1993: cat. Pompeiopolis no. 52 (columella). Marek 1993: 67, pl. 37, fig. 1 presents for the first time a provisional conspectus of some northern Anatolian grave types based on the stone collections of the archaeological museums of Amasra and Kastamonu. For a catalogue with illustrations of the Paphlagonian anthemion stelae, see French 2011. A single acanthus stele is dated by epigraphic evidence (Marek 1993: cat. Pompeiopolis

186

Julia Koch

Fig. 9: Achaemenid rock-cut tomb with a gabled roof and imitating house architecture in the Amnias river valley near Donalar (author’s photo). the perception of Athenian (i.e. Attic-Sinopean) anthemion stelae dating back to the fifth and fourth centuries BC points strikingly in turn to a profound Paphlagonian Hellenism which reached right up into the northern Anatolian country during the High Imperial period.68 The prominent epitaph of the homeristes Kyros, who died in Pompeiopolis after he had often adorned the divine Homer with his stage performances, demonstrates that the Greek world of Homer was literally updated at local festivals.69 This is perhaps not surprising given that the cultural topography of Roman Paphlagonia is populated by chains of prehistoric burial mounds that characterised the heroic landscape. Within the territory of Pompeiopolis around 40 tumuli, measuring a maximum 8m in height and 12m in radius, were recorded on the northern bank of the Amnias river in the early 20th century by Giulio Jacopi during the first excavation campaign at località Kırktepeler which

67

no. 30 (159)), while another dated stele of this northern Anatolian style – as it is referred to by David French – appears slightly different due to its depiction of a female figure in the semicircular anthemion; on the shaft in relief are farm animals, a spindle and a vine branch with bunches of grapes: French 2011: 11–3, cat. no. 12 (205/206). 68 French 2004: 38 no. 64, pl. 10, 40 no. 68, pl. 9, 41 no. 71, pl. 11; Mitchell 2010: 109–10. 69 Mitchell 2010: 106; Marek 1993: cat. Pompeiopolis no. 28.

From funerary doorstones in Pompeiopolis to tracing local identity

187

Fig. 10: Doorstone of Doryphoros recorded in the northwestern territory of Pompeiopolis (Aslan Mahalle, Seydiler, Vilayet Kastamonu) in 2010 (photo by C. Marek). subsequently continued in the adjacent necropolis of Tepedelik (Fig. 8).70 In addition to these tumuli, prominent Achaemenid rock-cut tombs representing houses of the dead constituted permanent features of the landscape of Roman Paphlagonia (Fig. 9).71 In the context of the notion of the domus aeterna, the old-fashioned Anatolian burial custom of the former Persian satrapy of Greater Phrygia were revived by Roman door stelae as a pars pro toto.72 The process of tomb transformation from a house of death into pseudo-architecture may best be traced in the Phrygian highlands. Throughout the centuries numerous rock-cut

70

On Kırktepeler (mid sixth century to the late fourth century BC), see Jacopi 1936: 6–11, figs 14, 27–31; 1937: 3–8, figs 1, 4–11. For a map of the distribution of pre-Roman rock-cut tombs and verified or possible Iron Age tumuli in the Amnias valley, see Johnson 2010: fig. 12; for a conspectus of Achaemenid tumuli around Sinope and Amisos, see Johnson 2010: 115–8; for evidence of the territory of Gangra-Germanikopolis in the modern district of Çankırı, see Matthews 2009; for a conspectus of not only Hellenistic but also possibly Roman tumuli in the Amnias valley, see Johnson this volume. 71 Summerer and Kienlin 2010: 195–221; see also Gall 1966. 72 Waelkens 1986: 18–9, 21–31.

188

Julia Koch

Fig. 11: Doorstone crowned by an eggand-dart motif with a prothesis scene on top from Taşköprü/ Pompeiopolis (photo by R. Hessing). tombs were built within the Kümbet valley in order to commemorate the deceased within finely carved burial chambers. Ultimately, the configuration of this rupestral architecture changed to produce monuments with a sculptured false door façade with door knockers and a key plate in relief.73 Roman door stelae in Phrygia, Galatia and Inner Paphlagonia seem to represent the houses of the dead in a highly symbolic visual language that points to the social status, wealth and profession of the deceased. In the Amnias river valley the commercial importance of farming, fishing and viticulture is evident, while the doorstones from Pompeiopolis show a firm cultural linkage to Inner Anatolia. In the Phrygian highlands, the appearance of doorstone monuments seems closely connected to the process of urbanisation as it marks ‘one of the first steps towards an urban setting and lifestyle’.74 Ute Kelp has recently demonstrated that the Phrygian doorstones – as proven for the Paphlagonian specimens – reached their highest density in the less-developed regions of the rural hinterlands where new social groups of modest wealth assumed the doorstone habit during the period of the pax romana.75 In metropolitan Pompeiopolis, however, some hybrid specimens appeared (Figs 10–11), such as a single doorstone monument crowned by 73 74 75

Haspels 1971: fig. 307, 319–22. Kelp 2013: 83–4. Kelp 2013: 84.

From funerary doorstones in Pompeiopolis to tracing local identity

189

a garland that commemorated one Doryphoros. Another unique doorstones from Pompeiopolis with a banquet scene at the top – as initially recorded by Gustav Hirschfeld – and a prothesis on display testifies to the strong concurrent Hellenistic influence within Roman Paphlagonia.

6. Conclusion In the rural hinterland of the Euxine a significant upturn in civic self-representation can be traced by reference to the most prestigious funerary doorstone monuments of Pompeiopolis. Reflecting for the first time the agrarian-oriented society of Inner Paphlagonia in public spaces, doorstones indicate a step in the development of the urban setting of Roman northern Anatolia. During the High Imperial period the establishment of new communication routes led to an enormous upgrade of cultural exchange as natural barriers to communicate were overcome. As a result the adoption of Greek sepulchral monuments widely distributed during the fifth and fourth centuries BC within the apoikies on the southern Black Sea coast is opposed to funerary dynamics closely related to Inner Anatolia. Simultaneously with vibrant regional interactions, a local differentiation can be observed by contrasting distinct types of the Inner Paphlagonian grave stelae from Hadrianopolis, on the one hand, and the neighbouring city of Pompeiopolis, on the other hand. While during the second and third centuries both the grave columellae and the anthemion stelae point to a profound Hellenism right up the Paphlagonian countryside, a revival of Inner Anatolian burial traditions can a priori be traced by the spread of the so-called Phrygian doorstones into northern Anatolia. In the late 20th century Christian Marek raised the key question to which administrative unit the inland cities of Paphlagonia were associated during the pax romana: ‘In wessen Verwaltungsbereich fiel das Land dann unter Augustus? Wurde es von Ankyra oder von Nikomedeia aus verwaltet, gehörte es künftig zu Galatia oder Pontus-Bithynia?’76 Speaking from the ever growing archaeological record, it seems conspicuous that not a single Bithynian Stockwerkstele nor Pontic relief stele were as yet found throughout the vast territory of the Paphlagonian metropolis.77 However, in accordance with Mareks’ historical reconstruction of the political rearrangement of northern Anatolia in the early Roman period firm cultural links between the Inner Paphlagonian urban elite and Galatia as well as the Phrygian highlands can be observed by reference to the funerary doorstone monuments from Pompeiopolis. My attendance in the Amasya conference was funded by a generous travel grant awarded by the Herbert-Lutz-Gedächtnisstiftung through the Institute of Classical Archaeology at Munich University for which I am grateful. I also wish to thank Christian Marek and his team for providing me with photographic material and broad insights into the results of their recent survey campaigns. My sincerest thanks go to Lâtife Summerer for constantly offering me tremendous support during my research at Pompeiopolis and for providing me with access to the Pompeiopolis Project Archive. Julia Koch Universität Bamberg, Archäologie der Römischen Provinzen Am Kranen 14, 96045 Bamberg, Germany [email protected] 76 77

Marek 1993: 72. See Cremer 1992 with evidence for Greco-Roman Stockwerkstele within northern Anatolia.

190

Julia Koch

Bibliography Ainsworth, W. 1839. ‘Notes on a Journey from Constantinople, by Heraclea, to Angora, in the autumn of 1838’, Journal of the Royal Geographical Society of London 9, 216–76. Bekker-Nielsen, T. 2013. ‘Neapolis-Neoklaudiopolis: A Roman City in Northern Anatolia’, in Ü. Yalçn (ed.), Anatolian Metal 6 (Der Anschnitt Supplement 25). Bochum, 203–13. Belke, K. 1984. Galatien und Lykaonien (Tabula Imperii Byzantini 4). Wien. Belke, K. 1996. Paphlagonien und Honōrias (Tabula Imperii Byzantini 9). Wien. Cambi, N. 1987. ‘Salona und seine Nekropolen’, in H. von Hesberg and P. Zanker (eds), Römische Gräberstraßen. Selbstdarstellung – Status – Standard. Colloquium Munich, 28–30 October 1985. München, 251–79. Çevk, F. and M. Ritter 2013. ‘Pompeiopolis nekropolü yüzey araştırması (Paphlagonya)’, Araştırma Sonuçları Toplantısı 2013, 169–82. Coulton, J. J. 2005. ‘Pedestals as Altars in Roman Asia Minor’, Anatolian Studies 55, 127–57. Cremer, M. 1992. Hellenistisch-römische Grabstelen im nordwestlichen Kleinasien 2. Bithynien (Asia Minor Studien 4). Bonn. Dalaison, J. and F. Delrieux 2011. ‘Les monnaies trouvées à Pompeiopolis durant les campagnes de fouilles 2006–2009’, in Summerer 2011, 141–8. Devreker, J., H. Thoen and F. Vermeulen 2003. Excavations in Pessinus: the so-called Acropolis. From Hellenistic and Roman cemetery to Byzantine castle. Ghent. Drew-Bear, T. and T. Lochman 1996. ‘Grabreliefs aus Amorion, Orkistos und der antiken Siedlung von Bağlica. Zeugen verlorengegangener Grabbauten’, Arkeoloji Dergisi 4, 109–34. Drew-Bear, T., C. M. Thomas and M. YIldIzturan 1999. Phrygian Votive Steles. Ankara. French, D. 1988. Roman Roads and Milestones of Asia Minor 2. An Interim Catalogue of Milestones (BIAA Monographs 9, BAR International Series 392). Oxford. French, D. H. 1990. ‘Sinopean Notes 1’, Epigraphica Anatolica 16, 45–64. French, D. H. 1991. ‘Sinopean Notes 2’, Epigraphica Anatolica 18, 141–55. French, D. H. 2004. The Inscriptions of Sinope (Inschriften griechischer Städte aus Kleinasien 64). Bonn. French, D. 2011. Funerary Stelae from Paphlagonia (BIAA Electronic Monograph 1). Ankara. French, D. H. 2012. Roman Roads and Milestones of Asia Minor 3.2 Galatia (BIAA Electronic Monograph 2). Ankara. French, D. H. 2013. Roman Roads and Milestones of Asia Minor 3.4 Pontus et Bithynia (with Northern Galatia) (BIAA Electronic Monograph 4). Ankara. Gall, H. von 1966. Die paphlagonischen Felsgräber. Eine Studie zur kleinasiatischen Kunstgeschichte (Istanbuler Mitteilungen Supplement 1). Tübingen. Haensch, R. 2005. ‘Rom und Chersonesus Taurica. Die Beziehungen beider Staaten im Lichte der Ehrung des T. Aurelius Calpurnianus Apollonides’, in V. Cojocaru (ed.), Ethnic Contacts and Cultural Exchange North and West of the Black Sea from the Greek Colonization to the Ottoman Conquest. Iasi, 261–3. Halfmann, H. and E. Schwertheim 1986. ‘Neue Inschriften aus Paphlagonien und Bithynien’, Epigraphica Anatolica 8, 129–32. Hamilton, W. J. 1842. Researches in Asia Minor, Pontus and Armenia with some account of their Antiquities and Geology 1. London. Haspels, C. H. E. 1971. The Highlands of Phrygia. Sites and Monuments. Princeton. Hirschfeld, G. 1888. ‘Inschriften aus dem Norden Kleinasiens besonders aus Bithynien und Paphlagonien’, Sitzungsberichte der phil.-hist. Klasse der Preußischen Akademie der Wissenschaften. Berlin, 863–92. Jacopi, G. 1936. Dalla Paflagonia alla Commagene: relazione sulla prima campagna esplorativa. Rome. Jacopi, G. 1937. Esplorazioni e studi in Paflagonia e Cappadocia: relazione sulla seconda campagna esplorativa Agosto – Ottobre 1936. Rome. Jes, K. 1997. ‘Gebaute Türgrabsteine in Aizanoi’, Istanbuler Mitteilungen 47, 231–50. Jes, K. 2001. ‘Türgrabsteine in Aizanoi II: Fassadenmonumente mit Scheintür’, Istanbuler Mitteilungen 51, 279–318.

From funerary doorstones in Pompeiopolis to tracing local identity

191

Johnson, P. 2010. Landscapes of Achaemenid Paphlagonia (PhD Dissertation at the University of Pennsylvania). Philadelphia. Jones, C. P. 1988. ‘A monument from Sinope’, Journal of Hellenistic Studies 108, 193–4. Jonnes, L. 1994. The inscriptions of Heraclea Pontica (Inschriften griechischer Städte aus Kleinasien 47). Bonn. Kahl, G. 1992. ‘Überreste von antiken Keltern im Gebiet von Neoklaudiopolis’, Hohenheimer Themen. Zeitschrift für kulturwissenschaftliche Themen 1, 23–41. Kelp, U. 2013. ‘Grave Monuments and Local Identities in Roman Phrygia’, in P. Thonemann (ed.), Roman Phrygia. Culture and Society. Cambridge, 70–94. Kokkinia, C. 1999. ‘Ein neues Grabepigramm aus Bithynien’, Epigraphica Anatolica 31, 167–70. Krsmanovic, D. and W. Anderson 2012. ‘Paths of the Dead – Interpreting Funerary Practice at Roman-Period Pessinus, Central Anatolia’, Melbourne Historical Journal 40.2, 58–87. Künzl, E. 1998. Das römische Prunkportal von Ladenburg (ALManach 3). Stuttgart. Kunnert, U. 2011. ‘Archäologische Funde und Befunde in den Territorien von Pompeiopolis und Abonuteichos-Ionopolis’, in Summerer 2011, 207–14. LaflI, E. 2012. ‘Archäologische Evidenzen zum Weinanbau im südwestlichen Paphlagonien in römischer und frühbyzantinischer Zeit’, in E. Olshausen and V. Sauer (eds), Die Schätze der Erde – Natürliche Ressourcen in der antiken Welt. Stuttgarter Kolloquium zur Historischen Geographie des Altertums 10, 2008 (Geographica Historica 28). Stuttgart, 261–79. LaflI, E. and E. Christof 2012a. Hadrianopolis I. Inschriften aus Paphlagonia (BAR International Series 2366). Oxford. LaflI, E. and E. Christof 2012b. ‘Drei neu entdeckte Phallossteine aus der Chora von Hadrianopolis’, in: G. R. Tsetskhladze (ed.), The Black Sea, Paphlagonia, Pontus and Phrygia in Antiquity. Aspects of Archaeology and Ancient History (BAR International Series 2432). Oxford, 163–70. Lambrechts, P. 1969. ‘Les fouilles de Pessinonte: La Nécropole’, L’Antiquité Classique 38, 121–46. Lochman, T. 2003. Studien zu kaiserzeitlichen Grab- und Votivreliefs aus Phrygien. Basel. Marek, C. 1993. Stadt, Ära und Territorium in Pontus-Bithynia und Nord-Galatia (Istanbuler Forschungen 39). Tübingen. Marek, C. 2003. Pontus et Bithynia. Die römischen Provinzen im Norden Kleinasiens. Mainz. Matthews, R. 2007. ‘An arena for cultural contact: Paphlagonia (north-central Turkey) through prehistory’, Anatolian Studies 57, 25–34. Matthews, R. 2009. ‘A Dark Age, Grey Ware and Elusive Empires: Paphlagonia through the Iron Age, 1200–330 BC’, in Matthews and Glatz 2009, 149–71. Matthews, R. and C. Glatz (eds) 2009, At Empire’s Edge: Project Paphlagonia. Regional Survey in North-Central Turkey (BIAA Monograph 44). London. Matthews, R., M. Metcalfe and D. Cottica 2009. ‘Landscapes with Figures: Paphlagonia through the Hellenistic, Roman and Byzantine Periods (330 BC–AD 1453)’, in Matthews and Glatz 2009, 173–226. Mitchell, S. 1976. ‘Legio VII and the Garrison of Augustan Galatia’, Classical Quarterly 26, 298–308. Mitchell, S. 1993. Anatolia. Land, Men, and Gods in Asia Minor 1. The Celts and the Impact of Roman Rule. Oxford. Mitchell, S. 2010. ‘The Ionians of Paphlagonia’, in T. Whitmarsh (ed.), Local Knowledge and Microidentities in the Imperial Greek World. Cambridge, 86–110. Moock, D. W. von 1998. Die figürlichen Grabstelen Attikas in der Kaiserzeit. Studien zur Verbreitung, Chronologie, Typologie und Ikonographie. Mainz. Munro, J. A. R. 1901. ‘Roads in Pontus, Royal and Roman’, Journal of Hellenic Studies 21, 52–66. Oliver, G. 2000. ‘Athenian Funerary Monuments: Style, Grandeur, and Cost’, in G. J. Oliver (ed.), The Epigraphy of Death. Studies in the History and Society of Greece and Rome. Liverpool, 59–80. Pfuhl, E. and H. Möbius 1977. Die ostgriechischen Grabreliefs. Mainz. Rheidt, K. 2010. Aizanoi und Anatolien. Neue Entdeckungen zur Geschichte und Archäologie im Hochland des westlichen Kleinasien. Mainz. Roosevelt, C. H. 2006. ‘Symbolic Door Stelae and Graveside Monuments in Western Anatolia’, American Journal of Archaeology 110, 65–91. Strubbe, J. 2005. The Inscriptions of Pessinous (Inschriften griechischer Städte aus Kleinasien 66). Bonn.

192

Julia Koch

Summerer, L. and A. von Kienlin 2010. ‘Achaemenid Impact in Paphlagonia. Rupestral Tombs in the Amnias Valley’, in J. Nieling and E. Rehm (ed.), Achaemenid Impact in the Black Sea. Communication of Powers. Aarhus, 195–221. Summerer, L. (ed.) 2011. Pompeiopolis I. Eine Zwischenbilanz aus der Metropole Paphlagoniens nach fünf Kampagnen (2006–2010) (Schriften des Zentrums für Archäologie und Kulturgeschichte des Schwarzmeerraumes 21). Langenweißbach. Türktüzün, M. 1991. ‘Çavdarhisar (Aizanoi) Roma Devri Nekropolü Kurtarma Kazızı’, Müze Çalişmaları ve Kurtarma Kazıları Sempozyumu 2, 81–94. Waelkens, M. 1977. ‘Phrygian Votive and Tombstones as Sources of the Social and Economic Life in Roman Antiquity’, Ancient Society 8, 277–315. Waelkens, M. 1986. Die kleinasiatischen Türsteine. Typologische und epigraphische Untersuchungen der kleinasiatischen Grabreliefs mit Scheintür. Mainz. White, K. D. 1967. Agricultural Implements of the Roman World. Cambridge. Wulf, U. 1993. ‘Zwei Grabbauten in der Südwestnekropole von Aizanoi’, Archäologischer Anzeiger, 527–41.

Muzaffer Doğanbaş

Preliminary results of rescue excavations in the Roman and Byzantine necropoleis of Amasya (1977–2014)

1. Introduction Evidence of human settlement has been identified at Amasya from the Chalcolithic period onwards, but it was during the Hellenistic period, when the Pontic Kingdom became powerful, that the city rose to prominence and there is evidence of important public improvements and economic stability. The city was the capital of the Pontic Kingdom during this period and kept the title for ca 120 years; this status resulted in a rapid expansion of the city in terms of both urban development and population increase. Amasya, which had been destroyed in the wars between the Pontic kings and the Romans, rapidly flourished under Roman hegemony, especially from the first century AD onwards, and it was an important city throughout the Roman period due largely to its location at the intersection of strategic trade routes. This period of peace and stability under Roman rule brought about population increase and economic growth, similar to that seen in the Hellenistic past, and this in turn contributed to the development of the city into a metropolis. These phenomena – public improvements, population density and economic development – provided the basis for the establishment of large necropolis areas around the urban centre. Five necropoleis have been identified within the modern cityscape to date (Fig. 1). The first necropolis was discovered in the Şamlar quarter during railway works in the 1920s. Built tombs were first mentioned by a local historian, Osman Fevz Olcay, in his book, Amasya Tarihi Zeyli, published in 1937 (Fig. 2). Other groups of graves recue excavated in Kurşunlu, Dere, Üçler quarters recently attest to the existence of further necropoleis which are unknown in the literature until now.

2. Rescue excavations in the Şamlar-İhsaniye quarters The first rescue excavations in the Şamlar-İhsaniye quarters were carried out by the Amasya Museum Directorate following the unearthing of part of a necropolis area during road construction works at Kulistepe in the İhsaniye quarter in 1974.1 The area of Eski Şamlar Mezarlığı had been approved for habitation by the Amasya Municipality and in 1977 graves from the cemetery were removed and the area was opened to development 1

Takaz 1975: 109.

194

Muzaffer Doğanbaş

Fig. 1: Map of the locations of the rescue-excavated graves within the modern city of Amasya (cartography by R. Szydlak). (Fig. 1). In 1986 excavation continued in the area of Eski Şamlar Mezarlığı (Old Şamlar Cemetery). Later, during construction of a children’s playground, a football field and housing, several further graves were unearthed. In 1993 excavation in Eski Şamlar Mezarlığı continued for another 19 days, and a further 21 graves were identified. In three of the five trenches, 13 graves were unearthed consisting of two different types, two cist graves and eleven vaulted chambers. Due to new construction works the Amasya Museum was urged to carry out rescue excavations between April and May 2006. Based on their general architectural structure and grave offerings, the tombs were dated to the late Roman period. The graves were organized in groups of two or three adjacent to each other. Some graves were covered with flat stone slabs, others were roofed with vaults.2 All these graves are dated to the late Roman period.3 In 2001, rescue excavations were conducted at the construction site of the Hüseyin Poroy Elementary School in city block 504, parcel numbers 1, 2, 3, 4 and 28, close to the railway and within the limits of the İhsaniye quarter. The excavations brought to light three Roman vaulted chamber tombs constructed next to each other. In 2006, excavations at the necropolis in the Şamlar quarter continued. Some graves were detected during construction work at Ziya Paşa Elementary School on Elmasiye Street, 2 3

Tektaş 1989: 1715–9. Yüce 1994: 1–16.

Preliminary results of rescue excavations in the necropoleis of Amasya

195

Fig. 2: The Şamlar quarter necropolis (after Olcay 1937: 82).

Fig. 3: Coin of Domitian from the Şamlar quarter necropolis (M. Doğanbaş, ©Amasya Museum). and a total of seven graves were revealed during the course of the ensuing rescue excavations. These vaulted graves are dated to the second and third centuries. Of the five graves detected during rescue excavations in May 2006 at a construction site in city block 1085, parcel number 11 on Elmasiye Street, two were cist graves closed by a flat stone slab and three were vaulted. Based on the evidence of coin finds, these graves can be dated to the early fourth century. The cover stone of a grave was found during construction works in city block 1079, parcel number 10 on Elmasiye Street. The ensuing rescue excavation, carried out by Amasya Museum at the end of August 2009, unearthed two graves which belong to the simple stone cover type. Jewelry, such as earrings and rings, as well as a mirror were found, and these grave goods point to these being female burials. Furthermore, the coins found among the grave goods were helpful in dating the graves. While a silver coin found in grave 1 dates to the period of Domitian (81–96) (Fig. 3), the silver coin found in grave 2 dates to the period of Antoninus Pius (138–61). These ‘oboloi’ were possibly meant as payment for Charon, and if this assumption is correct, the coins are most likely to have been in circulation at the time of burial. Accordingly, grave 1 can be dated to the reign of Domitian or not long after and grave 2 to the reign of Antoninus Pius or not long after.4

4

Doğanbaş and Özdemr 2010: 149–58.

196

Muzaffer Doğanbaş

Fig. 4: Grave goods from the Şamlar quarter necropolis (M. Doğanbaş, ©Amasya Museum). Graves covered with stone slabs were discovered at a construction site in city block 1086, parcel number 9 on Elmasiye Street of the Şamlar quarter; rescue excavations conducted by the Amasya Museum in 2010 recovered five graves. Even though the graves were partly destroyed by the construction work, it was clear that they had not been looted. Three rings, a gold necklace, a cornelian ring stone, three glass unguentaria, two glass bottles and a terracotta unguentarium were among the grave goods (Fig. 4). Based on the evidence of these finds, the graves can be dated to the second century.

3. Rescue excavations in the Kurşunlu quarter Following the discovery in 2000 of architectural remains ca 1m below ground level at a building site in city block 428, parcel number 8, of the Kurşunlu quarter (Fig. 1), construction works were halted for two days of rescue excavation. A row of vaulted graves was discovered (Fig. 11). The main walls of the graves were constructed of rubble and the cradles of the vaults were built with tile bricks. Besides the vaulted graves other grave types such as tile lined graves and simple pit graves covered with stone slabs, were also observed in this area.5 In terms of their general architectural forms, the graves reflect the character of

5

Doğanbaş 2002: 63–72.

Preliminary results of rescue excavations in the necropoleis of Amasya

Fig. 5: Graves excavated in 2010 in the Kurşunlu quarter necropolis (M. Doğanbaş, ©Amasya Museum).

Fig. 6: Three painted graves rescue excavated in the Şamlar quarter necropolis in 2010 (M. Doğanbaş, ©Amasya Museum).

197

198

Muzaffer Doğanbaş

Fig. 7: Interior of a grave rescue excavated in 2014 in the Kirazlıdere quarter necropolis (M. Doğanbaş, ©Amasya Museum). third- and fourth-century examples. A bronze coin of Amaseia from the time of Severus Alexander (222–35) was recovered from one of the graves. After a hiatus of ten years, rescue excavations in the Kurşunlu quarter recommenced in 2010 (Fig. 5). Architectural remains of Roman-period graves were exposed during road enlargement works. Following an inspection of the find-spot, rescue excavations were conducted on the corner of and between Ömür Street and Çelik Street of the central Amasya Kurşunlu quarter. Seven Roman graves were unearthed in four trenches. The graves were constructed of rubble, tile bricks and mortar, and were plastered with lime. Two of the graves were vaulted and the entrances were stone covered. Other graves were arranged with flat stone lids. The interior walls of the graves were decorated with painted garlands (Fig. 6), stylized Medusa heads and grape-eating birds on ivy branches painted on a layer of lime plaster (Fig. 7). However, the humidity inside the graves had caused the plaster to spall off and the graves themselves had been damaged during the initial road construction 30–40 years previously; the latter may also account for the absence of finds. The graves are dated to the second to third century.

4. Rescue excavations in the Dere quarter The discovery of graves, unearthed during construction works in the old open market of the Dere quarter (Fig. 1), resulted in four days of rescue excavations being undertaken by the Amasya Museum in May 2006. In city block 222, parcel numbers 2, 3 and 4, seven Roman graves were excavated; three were vaulted and four were of the cist-grave type. A coin belonging to the period of Antoninus Pius was recovered from grave 7 and suggests a date of burial in the second century. Additionally, it seems that other graves are situated under the main street which provides access to the Dere quarter.

5. Rescue excavations in the Üçler quarter Further graves were discovered and partly destroyed during construction works in city block 1046, parcel number 11, in the Üçler quarter. Rescue excavations undertaken by the Amasya Museum ensued in December 2013 (Fig. 8). At the southern and western edges of parcel number 11, seven graves were recorded, but the graves continue beneath the street

Preliminary results of rescue excavations in the necropoleis of Amasya

199

Fig. 8: The Üçler quarter necropolis (M. Doğanbaş, ©Amasya Museum).

Fig. 9: Detail of a painted grave rescue excavated in 2014 in the Kirazlıdere quarter necropolis (M. Doğanbaş, ©Amasya Museum). adjacent to the excavation area. The graves were situated at a depth of 7–8m below street level. They show characteristics of the late Roman and Byzantine periods, being built of rubble and vaulted with brick tiles; they can be dated to the fourth to fifth century.

6. Rescue excavations in the Kirazlıdere quarter The first rescue excavation in the Kirazlıdere quarter was conducted in 2006 (Fig. 1), when a grave was detected under the street during construction work in city block 1200, parcel number 6 on Terminal Street. Due to ongoing building works, the rescue excavation was limited to one day with the aim of recording the single grave uncovered during construction. It was a cist grave of a rectangular shape, closed with flat stone slabs. Based on the

200

Muzaffer Doğanbaş

Fig. 10: The Kirazlıdere quarter necropolis in 2014 (M. Doğanbaş, ©Amasya Museum).

Fig. 11: A triple set of graves rescue excavated in the Kirazlıdere quarter necropolis in 2010 (M. Doğanbaş, ©Amasya Museum).

Preliminary results of rescue excavations in the necropoleis of Amasya

201

general structure of the architecture and Roman coin finds, the grave can be dated to the third century. When, in spring 2010, a vaulted grave was discovered during construction works in city block 100, parcel number 52, a second rescue excavation was executed in the Kirazlıdere quarter. Thirteen graves were excavated in parcel number 52. Based on the coin finds and the painted cross designs in Grave 5 (Fig. 9), these are datable to the fourth to fifth century. A third rescue excavation in the Kirazlıdere quarter was carried out in 2014 (Fig. 10). This was located in parcel number 65, which is adjacent to the necropolis area identified in parcel number 52. In this area, to the west of parcel number 52, graves with vaults and closed with flat stone slabs were found at a depth of 3.5m below the surface. In total, 22 graves were excavated in five weeks. With only one exception, all the burials in this area were constructed as vaulted chamber tombs.

7. Discussion The rescue excavations carried out under the direction of the Amasya Museum since 1974 have identified several necropoleis of the Roman period: in the Şamlar, İhsaniye, Kurşunlu, Dere and Üçler quarters. A necropolis of the Byzantine period was recorded in the Kirazlıdere quarter. In addition, based on the evidence of grave offerings, the Memi Dede Cemetery within the boundaries of the Gökmedrese quarter has also been identified as a Romanperiod necropolis. Since this cemetery was also used during the Islamic period, further investigations in this area are not possible. Most of the excavated tombs were constructed with rubble and brick tiles using a mixed lime and sand mortar, and they were closed with three or four, in few cases five, flat stone slabs. Some graves are distinguished by their size and vaulted construction. Three-quarters of the upper parts of the tomb chambers were vaulted using brick tiles and mortar; the quarter towards the entrance was left uncovered, but closed with stone slabs (Fig. 11). Generally, the graves were arranged in rows, in groups of two, three or four. Seven Byzantine vaulted chamber tombs were recorded in the Kirazlıdere necropolis. The multiple burials observed in these tombs suggest they were family graves. After the vaulted chamber tombs, the second most common type of funerary architecture in the city is the cist grave covered entirely with flat stones and the third most common type of burial is the tile grave, which appears particularly in the necropoleis of the Kirazlıdere and Dere quarters. These graves were arranged for narrow, single burials and were covered with large tiles. In some examples of this grave type, the floor of the grave was also paved with tiles. A grave excavated in city block 100, parcel number 52, in the Kirazlıdere quarter merits special attention. It was closed with a solid, flat stone lid set over walls constructed of rubble and mortar. Interestingly, two sides of the lid were equipped with metal hoops, which served to lift it and which, in turn, suggests the grave was frequently reopened. Another interesting observation is the five holes in the tiles paving the floor of the grave (Fig. 12). These holes are connected to each other by channels beneath the floor, the function of

202

Muzaffer Doğanbaş

Fig. 12: Terracotta paved floor of a grave with five holes in the Kirazlıdere quarter necropolis (M. Doğanbaş, ©Amasya Museum).

Fig. 13: Soles of shoes found in a grave in the Kurşunlu quarter necropolis (M. Doğanbaş, ©Amasya Museum).

Preliminary results of rescue excavations in the necropoleis of Amasya

203

which is not fully understood. Whether this apparent drainage installation had a ritual or a practical function is not known. At any rate, it is unique in the necropoleis of Amasya. In the western part of this grave was a ›pillow‹ constructed from two tile fragments. Similar built pillows are also attested in the graves of the Kurşunlu quarter necropolis. Inside one of the graves unearthed at the corner of Ömür Street and Çelik Street in the Kurşunlu quarter necropolis area there was a tile-built niche which probably received grave offerings. This niche is, however, also unique in the necropoleis of Amasya. In terms of burial customs, it is observed that the corpses were usually placed with their heads oriented to the west. As is usual for Roman burials, all graves contain offerings; particularly common is the placement of an ‘obolos’ in the mouth of the corpse: payment for Charon. Apart from coins, clay and glass unguentaria, bronze mirrors, rings and necklaces of gold are attested. Although the custom of offering grave goods was generally abandoned during the Christian period, in a few later graves rings and similar objects were found. In some graves in the necropoleis of the Kurşunlu and Kirazlıdere quarters the soles of shoes were discovered (Fig. 13), suggesting that the deceased were fully dressed. Some of the Roman graves were decorated with painted garlands (Fig. 6), gorgoneia, ivy branches and grape-eating birds (Fig. 7). Cross motifs of various sizes were painted with red dye. In a recently excavated grave in the Kirazlıdere necropolis, the capital letters A, face-up E and reverse S were found scattered between the arms of a large cross painted on the inner wall of the tomb chamber (Fig. 9). The author owes thanks to Lâtife Summerer who translated and improved the Turkish version of this report. Muzaffer Doğanbaş Amasya Museum Atatürk Caddesi, Amasya Merkez/Amasya 05100, Turkey [email protected]

Bibliography Doğanbaş, M. 2002. ‘Amasya Merkez Kurşunlu Mahallesi Temel Hafriyatı Kurtarma Kazısı’, 12. Müze Çalışmaları ve Kurtarma Kazıları Sempozyumu. Ankara, 63–72. Doğanbaş, M. and C. Özdemr 2010. ‘Amasya Merkez Şamlar Mahallesi 107 ve 10 Parselde Yapılan Kurtarma Kazısı’, 19. Müze Çalışmaları ve Kurtarma Kazıları Sempozyumu. Ankara, 149–58. Olcay, F. 1937. Amasya Tarihi Zeyli. Amasya. Takaz, H. 1975. ‘Amasya Kulistepe Nekropol Kazısı Ön Raporu’, Türk Arkeoloji Dergisi 22.1, 109. Tektaş, M. 1989. ‘Amasya Merkez Eski Şamlar Mezarlığı 1986 Kurtarma Kazısı Raporu’, Belleten 52, 1716–9. Yüce, A. 1994. ‘Amasya Merkez Eski Şamlar Mezarlığı 1993 Yılı Kurtarma Kazısı’, 5. Müze Kurtarma Kazıları Sempozyumu. Ankara, 1–16.

İlkay İvgin

Results of a rescue excavation in the village of Zafer, near Tekkeköy, Samsun province

1. Introduction A tomb chamber was discovered during road construction works near the village of Zafer in a locality known as Azman, 33km east of Samsun and not far from the well-known prehistoric site of Tekkeköy (Fig. 1). In September 2013, when the Samsun Museum was first notified of the discovery, excavations began immediately.1 The tomb chamber is located on the slope of a high hill in a valley running towards the Black Sea. This small-scale tomb, cut into conglomerate rock, consists of a single chamber which contains neither a niche nor a kline. The plan of the chamber is almost square, but slightly irregular, and its roof is roughly rounded and dome-like. The northern wall is longer than the southern wall (2.1m and 1.7m); the western wall is 2m long and the eastern wall 1.9m. A well-centred opening in the eastern wall is square (60cm × 60cm) and the height of the chamber is 1.45m (Fig. 2). Since the entrance to the tomb was destroyed during the roadworks, it is not known whether or not there was a dromos. However, a semicircular overhang was observed over the entrance, which may point to the existence of a dromos (Fig. 3). Whilst the destruction means it is unclear how the door was closed, based on similar tombs in the region we may assume it was closed with a stone slab,2 although no sign of one is preserved. Two tube-shaped terracotta sarcophagi, an unguentarium and numerous scattered human bones were found in the interior of the tomb chamber. The find situation of the scattered bones does not allow the determination of the total number of individuals concerned (Fig. 4). Nevertheless, the skeletal remains suggest that at least two persons were buried in this tomb. Unfortunately, the character of the finds does not permit the establishment of a chronology precise enough to distinguish individual phases other from the suggestion that the skeletal remains scattered on the floor belong to earlier burials while the terracotta sarcophagi were used in later burials. Individuals associated with the earlier burials were probably interred directly on the floor of the tomb chamber. Amongst the bones found on the floor, there were three ring-shaped bronze earrings, two intact and one 1

Rescue excavation permit number 176675 of the General-Directorate of Cultural Heritage and Museums of the Ministry of Culture and Tourism. I owe thank to Lâtife Summerer who improved this report. 2 Ünan 2010: 50–99.

206

İlkay İvgin

Fig. 1: Map of the location (cartography by R. Szydlak).

Fig. 2: Plan of the burial chamber (plan by U. Akyüz).

Results of a rescue excavation in the village of Zafer

207

Fig. 3: The entrance to the burial chamber (author’s photo). fragmented. For the (assumed) later burials in clay sarcophagi, the corpses were laid in a dorsal position. Apart from the scattered bones, the most important evidence for understanding the later use of the tomb chamber is the extension of the western and northern walls in order to fit the clay sarcophagi (Fig. 2). The two tube-shaped terracotta sarcophagi were placed directly on the floor leaning against each other, immediately to the right of the entrance and against the northern wall with the heads orientated to the east. Both lids were sealed with clay and lime mortar. Sarcophagus 2 was set against the northern wall of the burial chamber, while the longer Sarcophagus 1 was located immediately to the south. Sarcophagus 1 was found broken in two in the middle, while Sarcophagus 2 was intact. An unguentarium with a broken rim was found immediately adjacent to the head end of Sarcophagus 1, leaning against the wall (Fig. 4c).

2. The finds The clay sarcophagi and other finds were removed carefully from the tomb chamber and taken to the Samsun Museum.3 The lids of the sarcophagi, which were sealed with clay and lime mortar, were opened and the human bones inside the broken Sarcophagus 1 were removed for further investigation. A silver ring was still on the left-hand ring finger. Since the skeleton inside Sarcophagus 2 was better preserved than that in Sarcophagus 1, it was preferred to keep it untouched for utilisation in a future museum exhibition. Sarcophagus 2 yielded no grave goods. In total, seven grave goods, including a silver ring, an unguentarium and three bronze earrings were recorded in the chamber. 3

The burial chamber was registered with the Samsun Cultural Heritage Protection Regional Committee. In accordance with the decree of the committee, the entrance to the burial chamber was closed with iron bars and the grave protected.

208

İlkay İvgin

Fig. 4a–b: The interior of the tomb (©Samsun Museum Archives).

Results of a rescue excavation in the village of Zafer

209

Fig. 4c–d: The interior of the tomb (©Samsun Museum Archives). 2.1 Sarcophagus 1 Sarcophagus 1 measures: length 1.85m; diameter 44cm at the wide end and 19cm at the narrow end; wall thickness 1.2cm average. The sarcophagus is made of clay fired light brown. It belongs to the tube-shaped type of sarcophagus, named for its cylindrical form which tapers from head to foot (Fig. 5).4 Lids covered the two openings on the sarcophagus: one at the head end (lid 35cm × 50cm) and one at the foot end (lid 17cm × 18cm). Both lids have a single knob-shaped handle and corresponding pierced protrusions which must have served to tie them together. Apparently, this closure system was not deemed sufficient, and 4

Cf. Richter 2011: 192–7, pl. 258.7–8 (Röhrensarkophage). Yurtseven 2006: 93 calls this the ‘torpeda’ type.

210

İlkay İvgin

Fig. 5: Sarcophagus 1 (author’s photo).

Fig. 6: The interior of Sarcophagus 1 (author’s photo).

Results of a rescue excavation in the village of Zafer

211

Fig. 7: Detail of the interior of Sarcophagus 1 (author’s photo).

Fig. 8a–c: The finds (author’s photo). so the lids were also sealed with clay and lime mortar. Since the opening at the wider end of the sarcophagus was larger than the one at the foot, the corpse must have been placed in the sarcophagus from the upper opening and then pulled into the sarcophagus from the smaller opening at the foot end. Eight parallel decorative relief lines encircle the body of the terracotta sarcophagus at ca 20cm intervals from the foot to below the upper lid. The condition of the sarcophagus is generally fine, even though there are signs of wear and chips on the surface. The sarcophagus had been broken into two parts and fixed already

212

İlkay İvgin

Fig. 9: Sarcophagus 2 (author’s photo).

Fig. 10: The interior of Sarcophagus 2 (author’s photo).

Results of a rescue excavation in the village of Zafer

213

in antiquity; this repair held over the centuries and was broken only during the removal of the sarcophagus from the tomb. Unfortunately, the bones fell from the broken sarcophagus and were collected for later anthropological analyses (Fig. 6). During the investigation and cleaning of the interior of the sarcophagus, imprints of textile were detected where the head of the deceased had been placed. No other textile imprints were found in the interior, and it is not clear whether those near the head originate from a burial robe or a pillow. The burial yielded no grave goods except for the silver ring mentioned above (Fig. 7). The ring measures: diameter 1.9cm; thickness 1.5mm; weight 0.9gr. It is designed simply with overlapping open finials (Fig. 8b) and was found in generally good condition; it was broken only during its recovery. 2.2 Sarcophagus 2 Sarcophagus 2 measures: length 1.7m; diameter 46cm at the wide end and 22cm at the narrow end; wall thickness 2.5cm average (Fig. 9). Like Sarcophagus 1, Sarcophagus 2 too is made of light-brown baked clay. It also belongs to the tube-shaped type. In contrast to Sarcophagus 1, however, both lids are almost square: the one at the wide end measures 45cm × 45cm and that at the narrow end measures 20cm × 21.5cm. There is a handle on only the lid from the narrow end, and there are no pierced protrusions. The lids of the sarcophagus were sealed with clay and lime mortar. Relief lines running lengthwise and across decorate the sarcophagus at ca 30cm intervals, while a cross motif appears at the foot end. This sarcophagus yielded no grave goods (Fig. 10). 2.3 Unguentarium The unguentarium was found standing but leaning against the eastern wall, ca 15cm from the top of Sarcophagus 1. It measures: preserved height 24.5cm; body diameter 9.5cm; rim and base diameters 3cm. The terracotta unguentarium is light red in colour with a white slip. It is well preserved except for some lime deposition, a broken rim and incomplete base (Fig. 8a). 2.4 Earrings Two intact bronze earrings and one half of another were found; they are all poorly preserved (Fig. 8c).

3. The repair to Sarcophagus 1 The ancient repairs to Sarcophagus 1 raise a number of questions. The area of the repair is on the back of the sarcophagus, where it touched the floor, and close to the middle. This suggests that it cracked immediately after firing and that repairs were implemented straight away in order to prevent further cracking in this area. First, with the help of drill,5 22 holes 5

Ulrich 2007: 30–2. The instrument defined here as an antique drill should probably be identified as a bowl drill.

214

İlkay İvgin

Fig. 11: The ancient repair on Sarcophagus 1 (author’s photo). at precise intervals were opened next to each other in the cracked area (Fig. 11); these were then connected to each other with ca 4cm-long lead clasps. Why was such a repair implemented? Could a repaired sarcophagus be sold?6 If so, was it cheaper than an undamaged one and to what extent did the purchase of a repaired sarcophagus point to the socio-economic rank of the deceased? After the sarcophagus was placed in the tomb chamber, the original crack increased over time and eventually caused the sarcophagus to break into two parts.7

6 7

On the social meaning of ancient repairs, see Dooijes and Nieuwenhuyse 2007: 15–20. Restoration of the repaired section was implemented by the author by the application of 50% Paraloid B 72 prepared in acetone to the walls of the fracture. The sections of the sarcophagus were

Results of a rescue excavation in the village of Zafer

215

4. Discussion The Samsun region, the area around the ancient Greek city of Amisos, has been densely settled since prehistoric times because of its favourable location on the Black Sea coast.8 According to the written sources, Milesians or/and Phokaians first founded the city which was settled also by Athenians in the fifth century BC.9 Little is known about the early period of the city due to a lack of archaeological evidence;10 the Hellenistic period of the city, however, is better attested due to grave finds.11 In recent years, numerous rescue excavations undertaken by the Samsun Museum have enhanced the body of archaeological data attesting to Hellenistic and Roman Amisos.12 Furthermore, several Hellenistic sites have been recorded in the surroundings of Amisos: Kurupelit (Karacadoğan Höyük) ca 10km from the city,13 Asarağaç Kalesi 10km east of Samsun,14 Akalan 18km southwest,15 İkiztepe Höyük in the Bafra district of Samsun,16 Asarkale 3km from the village of Asar near Kolay,17 Martı Kale ca.15km from Asarkale in the Bafra district18 and the settlement close to the village of Bengü submerged by the Altınkaya dam.19 The closest Hellenistic settlement of the Zaferköy tomb is Asarağaç Kalesi, but it is unlikely that the individuals buried in the Zaferköy tomb lived in this settlement which is ca 10km away; the existence of an unknown Hellenistic settlement in the closer vicinity of the grave is possible. Graves of the type found at Zaferköy, dating to the Hellenistic period, are known from across the Amisos region.20 They consist, generally, of one or more chambers cut into the conglomerate rock. Mostly oriented to the east and northeast, they sometimes yield rock-cut klinai and niches. The corpse could be a placed in a terracotta sarcophagi or simply on the floor or on a wooden kline. In some cases, the ashes of cremated corpses were put in urns and placed in small niches.21 The Zaferköy tomb conforms to the general rock-cut type, but it contains no rock-cut furniture. The find situation, with both scattered skeletal remains and intact sarcophagi, points to two different burial types in two different phases of use. Little is known about the earlier phase of the grave, and we assume that the corpses were simply placed on the floor. 7 8 9 10 11

12 13 14 15 16 17 18 19 20 21

consolidated by attaching them to each other with fibre glass and epoxy fasteners, a method not entirely dissimilar to the ancient one. Strabo 12.2.14. Strabo 12.2.14. However, black- and red-figure Attic pottery fragments recovered from Kurupelit (Karacadoğan Höyük) in 2009 were the first substantial evidence for the colonisation: Akyüz et al. 2011: 127. In the early 20th century Theodor Macridy unearthed a huge number of Hellenistic graves: Summerer 1999. In 1995, in the course of rescue excavations, the Samsun Museum unearthed a tomb chamber with rich, gold grave goods: Erciyas 2006. Akkaya 1993: 209. Akyüz et al. 2011: 117. Dönmez 2002: 133. Dönmez 2002: 136. AlkIm 1988: 1, 84–5. Blg et al. 2002: 41. Dönmez 2002: 136. Dönmez 2002: 136. They are still unpublished, but mentioned by Richter 2011: 192–7. Ünan 2013: 393–4.

216

İlkay İvgin

Neither the form nor the construction technique of the sarcophagi can be dated with any accuracy. This type of sarcophagus was used over a long period of time from the Hellenistic to the Byzantine period.22 The Hellenistic type is known from examples attested in the Amisos region in Dereler village and quarters of Samsun Semizli, Yeşilkent and Kadıköy.23 The dated examples unearthed during rescue excavations in the Kalınkaya Tumulus are the closest parallels to those from the Amisos region.24 Roman examples of this type of sarcophagus are attested in Yozgat-Akdağmağden,25 the necropolis area of Mersin/Soli Pompeiopolis (first to third century),26 the Tarsus Köylü Garajı tomb (mid first to late second century)27 and Patara (mid fourth century).28 As a result, the crucial evidence allowing us to date the tomb to the Hellenistic period is the architecture and the style of the unguentarium.29 Similar tombs of the Hellenistic period in the Amisos/Samsun region have been recorded in the Cedit and Kalkanca quarters and at Hendek Street of Samsun, Tınaztepe, Dikilitaş I-II, Yeşidere I-II, YeşilkentDemirciyeri, Dereler village I-II and Kozlu village.30 The unguentarium is the only firmly datable find from the Zaferköy tomb. Its particular form with a long cylindrical neck, a round body and a high stemmed base, which is almost of the same dimensions as the neck, is surely Hellenistic, probably second century BC.31 It is possible that the unguentarium is associated with the burials in the sarcophagi, because the use of such unguentaria as grave offering in combination with clay sarcophagi is well attested in Amisos. The breaking of the rim, also observed in other burials, might be seen in relation to burial rituals, as a demonstration of grief, or it might have been done to prevent reuse.32 İlkay İvgin Samsun Museum Samsun Arkeoloji ve Etnografya Müzesi Müdürlüğü Kale Mahalessi 19, Mayıs Bulvarı 5, İlkadım, Samsun 55100, Turkey [email protected]

Bibliography Akkaya, M. 1993. ‘Amisos Antik Kenti Kurtarma Kazısı’, III. Müze Kurtarma Kazıları Semineri, 207– 18. Akyüz, U., M. KolağasIoğlu and O. A. Şrn 2011. ‘Samsun İli, Atakum İlçesi, Çakalca (Karacadoğan Höyük) Kurupelit Kurtarma Kazısı 2009’, 19. Müze Çalışmaları ve Kurtarma Kazıları Sempozyumu, 117–34. AlkIm, U. B., H. AlkIm and Ö. Blg 1988. İkiztepe I. Türk Tarih Kurumu Basımevi. Ankara. 22 23 24 25 26 27 28 29

Tunay 1970: 184. Ünan 2013: 400. Tunay 1970: 184. Taş et al. 2014: 10. YağcI 2003: 36–7. Yurtseven 2006: 93, 101–2. IşIk 2012: 28–9. Unguentaria broken at the rim or base are often observed in the graves around Amisos: Ünan 2013: 398. 30 Ünan 2013: 389. 31 Cvelek 2006: 50. 32 Ünan 2013: 389.

Results of a rescue excavation in the village of Zafer

217

Blg, Ö., S. Atasoy, F. Gökçe and Ş. Dönmez 2002. ‘2001 Yılı Samsun İli Yüzey Araştırması’, 20. Araştırma Sonuçları Toplantısı 2, 41–50. Cvelek, A. 2006. ‘Stratonikeia – Akdağ Nekropolisi’nden Bir Mezar’, Anadolu/Anatolia 30, 47–64. Çlngroğlu, A., Z. M. Göz and G. Polat 2012. Ege Üniversitesi Arkeoloji Kazıları. Izmir. Dönmez, Ş. 2002. ‘Samsun-Asarağaç Kalesi’, Anadolu Araştırmaları 16, 133–44. Dooijes R. and O. P. Nieuwenhuyse 2007. ‘Ancient Repairs: Techniques and Social Meaning’, in: M. Benz and U. Kästner (eds), Konservieren oder Restaurieren, die Restaurierung griechischer Vasen von der Antike bis heute. Munich, 15–20. Ercyas, D. B. 2006. Wealth, Aristocracy, and Royal Propaganda under the Hellenistic Kingdom of Mithridatids in Central Black Sea Region in Turkey (Colloquia Pontica 12). Leiden. IşIk, H. İ. 2012. ‘Patara 2010 Yılı Kazı ve Restorasyon Çalışmaları’, 33. Kazı Sonuçları Toplantısı 3. Ankara, 15–40. İplkçoğlu, B. 2007. Hellen ve Roma Tarihinin Ana Hatları (Arkeoloji ve Sanat Yayınları). Istanbul. Richter, Ch. H. 2011. Parthische Pantoffelsarkophage: Untersuchungen zu einer Sargform Mesopotamiens im Vergleich mit Tonsärgen von Ägypten über den Mittelmeerraum bis Zentralasien. Münster. Saraçoğlu, A. 2011. ‘Hellenistic and Roman Unguentaria from the Necropolis of Tralleis’, Anadolu/ Anatolia 37, 1–42. Summerer, L. 1999. Hellenistische Terrakotten aus Amisos. Ein Beitrag zur Kunstgeschichte des Pontosgebietes (Geographica Historica 13). Stuttgart. Summerer, L. 2007. ‘Greeks and Natives on the Southern Black Sea Coast in Antiquity’, in G. Erkurt and S. Mitchell (eds), The Black Sea: Past, Present and Future. Proceedings of the International, Interdisciplinary Conference, Istanbul, 14–16 October 2004. London, 27–36. Taş, İ., M. Arslan and H. K. Şenyurt 2014. ‘Akdağmağdeni Cam Yüzük Kaşı Mühür’, Haberler 38, 10–2. Tsetskhladze, G. R. 2010. ‘Black Sea Trade: Some Further General Observations’, Anadolu Araştırmaları 19.1, 2006, 197–212. Tunay, M. İ. 1970. ‘Bir Toprak Lahit’, Arkitekt 1970–04 (340), 184. Ulrich, R. B. 2007. Roman Woodworking. Yale. Ünan, S. 2010. Samsun ve Çevresi Mezar Tipleri ve Ölü Gömme Adetleri (MA thesis), Dumlupınar Üniversitesi, Sosyal Bilimler Enstitüsü, Protohistorya ve Önasya Arkeolojisi Ana Bilim Dalı, Kütahya. Ünan, S. 2013. ‘Amisos ve Çevresinde Yer Alan Mezar Tipleri ve Ölü Gömme Adetleri’, in: H. YayIna, Ş. Dönmez and H. YayInlarI (eds), Güneş Karadeniz’den Doğar, Sümer Atasoy’a Armağan Yazılar. Ankara, 387–414. YağcI, R. 2003. ‘Soli-Pompeipolis 2002’, Anadolu Akdenizi, Arkeoloji Haberleri 1, 34–37. Yurtseven, F. 2006. ‘Tarsus Köylü Garajı Mezarı Buluntuları’, Anadolu/Anatolia 31, 91–121.

Baran Aydın, Laura Buccino and Lâtife Summerer

Honoured, beheaded and buried: A new deposit of statues from Amastris

1. Introduction In this article the topic of the Landscape Dynamics conference will be viewed from an urban-centred perspective. Roman cities characteristically contained spaces combining landscape with urban design and sculpture. As in cities of other regions of the Roman Empire, the Pontic cities set up statues of emperors and the urban elite in public spaces and buildings in order to legitimise political power. Such dynamics are demonstrated in the famous request of Arrian that Hadrian send a statue of himself to Trapezous.1 A statue of the emperor already stood in the city – perhaps at the harbour, indicating towards the sea – but, being artistically poor, it had to be replaced, because the emperor’s statue should be a monument for eternity. Within the framework of a forthcoming project on the sculptures housed at the Amasra Museum, this article presents an example from Amastris demonstrating the urban setting of imperial/honorific statues and their fate in late antiquity. Four headless statues were unearthed during rescue excavations conducted by the Amasra Museum in 1993. These were published in two preliminary Turkish reports and interpreted as evidence of ‘pagans hiding their statues from destruction by Christians’.2 Astonishingly, these publications did not attract the attention of scholars interested in the archaeology and history of Amastris.3 They were also overlooked by recent studies on Late Antique deposits of statues.4 Some unauthorised publications of photographs of the statues ignored the find context, not being aware of the excavation reports.5 1

2 3 4 5

Arrian, Periplus 1.3–4: ‘your statue [i.e. the statue of Hadrian] stands in a pleasant pose pointing toward the sea – but neither does it look like you nor is it beautiful. So send a statue in the same pose worthy of receiving your name: the place lends itself completely to a monument for eternity’. See Boatwright 2000: 139–40; Belfiore 2009: 123, 139; Karanastasi 2012–2013: 333–4, nn. 55–6; Cadario 2014: 107, n. 19. AteşoğullarI and Şmşek 1995: 93–111. See, for example, Crow and Hill 1995; Ercyas 2003: 1419–30; Zavagno 2009: 129–51; Sharp 2010: 53–144; Zavagno 2012. Anghel 2011. The over-life-size statue with cuirass is published: Marek 2003: fig. 138; Guldager Bilde et al. 2007–2008: fig. 34; Karanastasi 2012–2013: pl. 8.1. Marek 2003: fig. 137, another statue showing a male figure with himation, is wrongly labelled as ‘togatus’ and is from the same context near the Bedesten.

220

Baran Aydın, Laura Buccino and Lâtife Summerer

Fig. 1: Topographical setting of the monuments in Amastris (Cartography by R. Szydlak after Marek 1989: fig. 2).

Honoured, beheaded and buried

221

Fig. 2: The rescue excavation (©Amasra Museum). The present paper will focus on a discussion of this important deposit of statues, based on the published excavation reports, and discuss its value as testimony of the changing urban landscape of Amastris in late antiquity. It will first examine the age of the statues, then the composition and the date of the deposit, and, finally, its spatial setting within the city. By contextualising this material and its contact within the urban development of Amastris, the article aims to contribute to a better understanding of regional dynamics in northern Anatolia.

2. The rescue excavation and find context A statue was discovered by accident during the excavation of a septic tank in the small industrial area of the Bedesten district of Amasra on 20 October 1993 (Fig. 1). The local museum was immediately notified of the discovery and conducted a rescue excavation. A small trench (4.1m × 2.2m) was opened, and brought to light three more marble sculptures (Fig. 2). The statues were aligned in an east-west direction, three of them laying on their backs and one on face down; they were arranged in two rows placed next to each other (Fig. 3). They were covered by a 10–15cm thick layer of soil and large flat stone slabs which sealed off the deposit.6 Immediately to the south of the deposit a solid wall of large ashlar blocks was preserved, standing to a height of 95cm. Since the aim of the excavation was

6

The sizes of the stone slabs are given as 78cm × 56cm × 23cm and 36cm × 23cm × 12cm: AteşoğullarI and Şmşek 1995: 94.

222

Baran Aydın, Laura Buccino and Lâtife Summerer

Fig. 3: The statues in situ in the deposit (©Amasra Museum).

Fig. 4: Inscription fragment found in the deposit (©Amasra Museum). merely to rescue the statues, the architectural context of this wall was not investigated further. It was simply recorded that it continued in an east-west alignment.7 The excavators report that immediately to the south of the statues that is to say next to the wall, three clay pipes were recorded in situ.8 Together with the statues, some marble fragments were also found. A few of them were immediately identified as broken parts of one of the statues. In 7 8

AteşoğullarI and Şmşek 1995: 94. AteşoğullarI and Şmşek 1995: 94: ‘heykellerin hemen güneyinde; üç adet toprak künk original kullanım pozisyonunda ele geçirilmiştir’.

Honoured, beheaded and buried

223

addition, fragments of an inscription came to light. On the largest fragment is an inscription of two lines, with the Greek letters ΟΡΟΣΙ and ΣΟ OR ΣΘ preserved (Fig. 4); the full inscription is impossible to restore since the preserved letters offer too many possibilities. Whether or not they relate to one of the statues, that is to say whether or not they belong to a statue base, also remains unclear. The layer of soil covering the statues included mortar, broken roof tiles and pottery sherds, while the soil on top of the stone slabs yielded a few mosaic tesserae and four corroded bronze coins of the Late Antique period. The four statues were exhibited in the Amasra Museum until 2014, the museum is currently closed due to major renovation works.

3. The statues 3.1 Over-life-size statue of an emperor wearing a cuirass The over-life-size statue of an emperor wearing a cuirass (inv. 807 3.4 A 93) is not entirely unknown in the literature,9 although it has been neither properly studied nor published (Fig. 5). The torso is preserved to a height of 140cm. It was over life-size when complete, probably standing more than 2m high. The statue is carved from a grained, greyish marble. The missing head was worked separately and added to the body. A large part of the right shoulder and arm is broken off. The upper-left part of the torso, where the fold of the military cloak was fastened, is also missing. The missing raised left arm, which was possibly holding a spear, must have been worked separately, although no sign of a dowel hole is preserved.10 To allow for the insertion of the head, a large, deep hole was roughly hewn and this seems to correspond to the square neckline of the cuirass. The missing left leg of the statue was leaning on a support, which takes the form of an eagle on a rock.11 Part of the left ankle with the laces of the upper part of the footwear (calceus patricius) and a small section of the foot (height 50cm, width 48cm) are preserved on the surviving fragment of the support, which is corresponding to the statue. Judging from the position of the support, the folds of the cloak on the right thigh and the slightly outturned left hip, the weight was carried on the left leg; the right leg was bent at the knee and perhaps set slightly to the side. On the small preserved part of the lower-right arm, a row of the long fringes of the pteryges and the short sleeve of the tunic are visible. The figure wears a muscle cuirass and a military cloak which was draped over the missing right shoulder. The cuirass is decorated with a female figure at the centre who wears a double-girded peplos. She is holding a spear with her right hand, with the tip pointing downwards, and in her left hand she holds a shield. On either side of her is a flying Victory with wings and floating drapery; the Victories are shown in a heraldic pose and hold a crown above the central figure’s head. The predominant position of the central figure is 9

AteşoğullarI and Şmşek 1995: 93–111, figs 3–4, 11–4; AteşoğullarI 1999: 207–18, figs 1– 2; Marek 2003: 93, fig. 138; Guldager Bilde et al. 2007–2008: 156, fig. 34; Karanastasi 2012–2013: 364, no. 17, pl. 8.1; Cadario 2014: 107, 111, n. 19. 10 Colossal statues usually have separately-worked limbs. 11 The support was found close to the torso and remained unnoticed by later publications: Guldager Bilde et al. 2007–2008: 156, fig. 34; Karanastasi 2012–2013.

224

Baran Aydın, Laura Buccino and Lâtife Summerer

Fig. 5: Over-life-size cuirass statue representing, probably, Hadrian (©Amasra Museum). further indicated by her placement on the back of the she-wolf suckling Romulus and Remus. The she-wolf herself appears anatomically odd. While her back turned head with triangular ears evidently convey a wolf, her muscular body, legs, paws and tails rather resemble to a feline. The lion like she-wolf is supported by an acanthus from which, on the right side, a branch of ivy (or a vine?) emerges. The decoration of the cuirass is topped by a gorgoneion which is shown in a schematic style. The shoulder straps have lion-head finials. From the lower edge of the cuirass hangs a single row of pteryges with a semicircular cut. Each pteryx at the front is decorated with a motif (from left to right: heads of the Gorgon, a lion, Zeus Ammon, a satyr and a ram, and a full figure of an eagle); those on the back have simple rosettes. From the left shoulder the paludamentum folds fell down the back. The back of the statue is worked in a cursory manner; apparently it was set up against a wall and the back was not intended to be seen. Due to the pose and costume, the statue is identified as a variation of the well-known Hierapytna type, named after the provenance of the most famous example, otherwise

Honoured, beheaded and buried

225

known as the oriental type.12 According to communis opinio the iconography of this statue type was created during the time of Hadrian, probably by a Greek workshop in the early years of his reign. The decorative programme of the cuirass is complex and has been a matter of discussion in several studies.13 The decorative style of the cuirass is repeated, with variation, on more than 20 statues, most of which originate in the Greek East of the Roman Empire. They differ in terms of the ponderation and the arrangement of the paludamentum.14 Beside the new statue from Amastris, two other fragmentary examples of the oriental type are known from Asia Minor: one comes from Prusias ad Hypium in Bithynia and the other is housed in the Antalya Museum, but is without provenance.15 Although the example from Amastris shares the main features of the oriental type, there are some differences in the decoration of the cuirass: unlike the canonical type, the lower edge of the cuirass is not curved at the centre and the skirt of metal lappets consists only of a single row. The flying Victories converge towards the central figure and do not, as is usual, emerge symmetrically from the acanthus. The tendril growing from the acanthus at the right side seems to be misplaced and detached from the right Victory, while it is omitted from the left side.16 Also remarkable is the lack of symbols, such as the owl and the snake, which normally flank the central figure and which are interpreted by most scholars as representations of Athena/Minerva (or Palladium).17 For the example from Amastris an identification of the central figure is difficult since the aegis is absent. Although the head of the figure is largely damaged, it can be determined from the outlines of the remaining relief that the headgear was probably a Corinthian helmet. The motifs of the pteryges correspond with the examples of the canonical type. They represent decorated metal plaques of the cuirass which generically refer to the military sphere and might also have an apotropaic meaning. Noteworthy is the motif of the eagle which also appears on the support. The supports of other statues attributed to this type show either a kneeling prisoner, sometimes trampled by the foot of the emperor (as in the famous statue from Hierapytna), or a more generic palm trunk.18 The elaboration of the cuirass decoration on the torso from Amastris has been considered as provincial and simplified.19 The style and some of the iconographic features are in fact different from other known examples of this type. The composition is poorly balanced: the Victory on the right side is slightly misplaced as her wing touches the lion on the shoulder strap and the 12

13 14 15 16 17

18 19

Two detailed studies on the Hierapytna type have been published recently, but do not consider the statue from Amastris: Cavalieri and Jusseret 2009; Bergmann 2010. For the latest discussion of this type, see Friedland and Tykot 2012: 53–60, fig. 1; Karanastasi 2012–2013; see also the references in the following note. Cavalieri and Jusseret 2009; Bergmann 2010; Karanastasi 2012–2013; Cadario 2014. Karanastasi 2012–2013 gathers examples attributable to this statue type; two pieces are uncertain (nos 23–4) and another five (nos 25–9) are probably not relevant. Karanastasi 2012–2013: 364, nn. 18–9. On all other examples the tendrils are arranged symmetrically: Bergmann 2010: 259–89. Cavalieri and Jusseret (2009: 377–8) and Bergmann (2010: passim) interpret the figure as Palladion, the archaistic idol; according to Karanastasi (2012–2013: 332–3) it is a representation of the Classical Athena type. However, the presence of the she-wolf in the iconography of Athena would be unusual and requires an explication. Cavalieri and Jusseret 2009; Bergmann 2010; Karanastasi 2012–2013. Karanastasi 2012–2013: 325, 334.

226

Baran Aydın, Laura Buccino and Lâtife Summerer

tendril motif below is asymmetrical. From these observations it can be concluded that the style and perhaps also the marble used can be identified as ‘local’ products. With the current state of research, however, it is difficult to attribute this craftsmanship to Amastris itself or to another city in the region. A more detailed stylistic examination of all statue finds from this city remains the task of future research. In favour of Amastris being a centre of sculptural art in the Imperial period we can refer to Pliny the Younger who praises the city in a letter to Trajan as civitas elegans et ornata.20 In addition, the variation of the characteristic decorative programme of the oriental type, as shown in the central figure and the omission of some details, is to be valued as the creative transformation of the model to a local need, rather dismissed as the result of ignorance or a lack of knowledge of the original model. The eagle on the support could also have local implications, perhaps referring to the cult of Zeus Strategos, the main god of the city, who is depicted on the coins of Amastris.21 Therefore, the evaluation of the sculpture as ‘provincial’, as stated by Pavlina Karanastasi,22 should be regarded with caution, for the differences and variations from the common type could possibly have links to the local civic and cultural environment; the circumstances relating to imperial statue dedications in the Black Sea area were surely different from those at the centre of the empire.23 Since the separately-produced portrait head is missing, the identification of the statue with Hadrian remains unsecure. Although the statue from Amastris corresponds in terms of type, style and technique with the known cuirassed statues of Hadrian, we cannot exclude the possibility that the statue type designed for Hadrian could also have been used to honour one of his successors, as attested in other cases.24 3.2 Over-life-size statue of a woman wearing chiton and himation The over-life-size statue of a woman wearing a sleeved chiton (tunica), himation and perhaps stola (inv. 804 3.1 A 93) is particularly interesting for its high quality (Fig. 6) and because it replicates the statuary type of the so-called Artemisia from the Mausoleum of Halicarnassus. The head, originally worked with the body, is broken in the centre of the neck. The outstretched forearms, which were attached separately, are missing; the same is true for the right foot and a large part of the plinth is also missing. Of the left foot only the toes remain. The statue must have been more than 2m high when complete (the preserved height is 180cm) and corresponds in size to the cuirass statue discussed above. The figure stands on the left leg, while the right leg is flexed and extended. The statue wears a longsleeved chiton fastened with buttons on the shoulders and most probably also a stola. The 20

21 22 23 24

Pliny, Letters 10.98: ‘To the emperor Trajan. The elegant and beautiful city of Amastris, Sir, has among other capital buildings a most noble and extensive piazza (plateia) …’; see also Marek 1989: 377; Mosch 2013: 119. It has also been suggested that ta Amastrianou in Constantinople, described in the Byzantine sources of the eight and eleventh centuries, was adorned with statues which came from Amastris: Mosch 2013: 119–24. However, the location was apparently named after its benefactor who came from Amastris: see below. Mosch 2013: 122. Karanastasi 2012–2013: 333–4. Karanastasi 2012–2013: 338. The same type was used for Antoninus Pius in the Nympheum of Herodes Atticus in Olympia: Karanastasi 2012–2013: 328, 361, no. 9.

Honoured, beheaded and buried

227

Fig. 6: Over-life-size female draped statue (©Amasra Museum). latter is recognisable at the right shoulder as a strap which then goes down but gets lost at the breast level along the outline of the bust. The dense folds in the crook of the left arm look as if they are a kolpos. A himation is draped over the top, forming a mass of bunchedup folds around the waist and tightly enveloping the hips and legs; it is then wrapped around the back, finally falling over the left shoulder. As mentioned above, in terms of ponderation and the arrangement of the drapery, this statue refers to the same model of the late Classical statue of Artemisia of Halicarnassos housed in the British Museum.25 It lacks, however, the drapery detail of Artemisia, whose mantle falls down from the right shoulder, under the right arm and wraps the body beneath

25

For the so-called Artemisia, see Waywell 1978: 103–5, no. 27, pl. 13. Cf. also the veiled statue of a Vestalis in Rome (Antiquario Forense), dated to the early Antonine period, which seems to be dependent on the same model: Kruse 1975: 163–4, 365–6, no. D 70, pl. 62.

228

Baran Aydın, Laura Buccino and Lâtife Summerer

Fig. 7: Honorific male statue (©Amasra Museum). the breast. This styling is often used for Roman statues, especially during the Early Imperial period (e.g. the statues of veiled women, the so-called Orans type).26 A date in the Julio-Claudian period (at any rate not later than end of the first century) is indicated by the possible presence of the stola27 as well as the precise execution of the drapery and the plasticity of modelling. The over-life-size height and refined quality suggest that it was an honorary statue of a female member of the imperial family, as attested frequently for the similar statues of the Orans type.

26

The so-called Orans type is assumed to depend on either a late Classical original or a new Augustan creation. For recent discussion, see Geominy 2004: 300–2; Reinsberg 2005: particularly the comparison with the so-called Artemisia at 209; Valeri 2013: 319–20, no. IX.3 with bibliography. 27 Scholz 1992: 19; Alexandridis 2004: 53–4.

Honoured, beheaded and buried

229

3.3 Life-size statue of a man wearing a pallium The headless life-size statue of a man (inv. 806 3.3 A 93) is deeply draped with a pallium (himation) (Fig. 7). It stands in the so-called ‘arm-sling’ posture which was frequently used for honorary statues in the Greek East28 and is also attested for Hellenistic portrait statues as well as in funerary reliefs from the second century BC onwards.29 The pallium is worn over the chiton, wrapping both arms; the right arm is bent at the elbow and held diagonally across the chest. The left arm is slightly bent at the elbow and held close to the body while the hand carries an attribute, probably a volumen but now lost. The preserved height, 167cm, suggests that the statue was life-size.30 The separately worked head as well as the left hand and the right foot below the ankle are missing. The right leg carries the weight of the body, while the left leg is markedly flexed and positioned slightly diagonally inwards. Massive folds of the mantle fall down from the left arm on to the support which is attached to the leg. The feet are fitted with sandals laced up at the ankle (crepidae). The arc-like arrangement of folds on the body, without a central pleat between the legs, allows the attribution of the statue to the so-called Cyrene type, as classified by Erich Polaschek. This type was often used for honorary portrait statues, especially in Asia Minor and Cyrenaica in the second century.31 The statue is of excellent quality. The folds are differentiated using a drill and the body parts where the garment fits tightly are modelled plastically. The modelling of the right hand is particularly elaborate: the fingers are curled naturalistically and the dorsum of the hand is carefully polished. The decorative style in the rendering of the arcshaped folds at the sides, the deep furrows and net drilling in the drapery, and the fall of the pallium on the left forearm in particular argue in favour of a date around the middle of the second century. 3.4 Life-size statue of a man wearing a pallium The second draped male statue (inv. 805 3.2 A 93) belongs to the same statue type as the previous one and can also be dated to the second century (Fig. 8).32 However, this statue differs considerably from the previous one in terms of ponderation (the left leg is engaged and the right is flexed and free) and the elaboration of the drapery as well as the sculptural technique. The preserved height, 162cm, suggests a life-size statue, equal in height to the previous example. The missing head broke away at the neck. The right shoulder, the lower portions of the legs and the falling folds of the mantle are also broken, while the surface is heavily worn. The figure is shown in a rhetorical posture: his right hand gestures as if in mid-speech and his left hand grasps an attribute, possibly a carved scroll. In comparison to the first mantle statue, the drapery is modest. The garment wrapping the arms and the body shows concentric arc-shaped folds, which are richer around the Vneck of the mantle. The pallium falls from the left wrist in two symmetrical edges, which are deeply carved by drilling. This statue can be dated to the first half of the second century, 28 29 30 31 32

For this statue type, see Polaschek 1969; Smith 1998: 65–6; Filges 2000; Fejfer 2008: 196–7. Dillon 2006: 74. AteşoğullarI and Şmşek 1995: 93–111, figs 3–4, 11, 15–6. Cf. Hallett 2006: 50–151, 172, no. 48, pls 40–2; Lenaghan 2006: 176–7, no. 49, pl. 43. AteşoğullarI and Şmşek 1995: 93–111, figs 3–4, 11, 17; Marek 2003: 47, 93, fig. 70. Cf. for the ponderation the so-called Delos type: Lenaghan 2006: 185–6, no. 58, pl. 54.

230

Baran Aydın, Laura Buccino and Lâtife Summerer

Fig. 8: Honorific male statue (©Amasra Museum). although it was surely produced by a different workshop than that which was responsible for the previous statue.

4. The topographic and architectural setting of the deposit The statue deposit was discovered about 50m east of the so-called Bedesten in the narrow valley of Amasra Deresi, where the Plateia, the broad street decorated with colonnades and mentioned by Pliny the Younger as a particularly elegant feature of the city, was probably located.33 The Bedesten itself is an Early Imperial construction of brick-built walls and opus reticulatum details enclosing an area of 5,000m2. Adolf Hoffmann tentatively associates it with a large market building. Its design, however, hardly corresponds to the typical Roman market buildings. It was possibly a warehouse (horreum) on the arterial road leading 33

Marek 1989: 382. Pliny, Letters 10.98.

Honoured, beheaded and buried

231

towards the south.34 A city gate to the south of the Bedesten was still visible in the 19th century35 and a building at this gate with ashlar blocks was tentatively identified as a monumental tomb.36 The ruins of a ‘temple’ with large columns were observed by Eugène Boré in 1838.37 Based on epigraphic evidence and the remains of supporting vaulted structures, a theatre has been identified at the foot of the hill to the west of the Bedesten.38 The evidence of a funerary inscription testifies that there was a sacred district (ἱερὰ ἄμφοδα) in which inhabitants constituted the administrative Platearchai.39 Luis Robert rightly draws the conclusion that the Plateia must have been situated in this sacred district. It was named ‘hiera amfoda’ because the most important sanctuaries of the city were located there.40 The location to the east of the Bedesten where the statue deposit was discovered has not been reported or mapped as an archaeological find-spot previously, neither in early traveller reports nor recent surveys.41 Whilst another assemblage of sculptures from Amastris which arrived at the Istanbul Archaeology Museums in 1890 is very similar in terms of composition, including a cuirassed torso and a draped female statue, both over-life-size and datable to the second century,42 unfortunately, the find-spot of these statues is not recorded. Although the excavators did not investigate the walls near which the statues were deposited, the large ashlar blocks, visible on the photographs taken during the excavation, indicate a public building. As mentioned above, the excavators observed terracotta water pipes (one inside and another alongside the wall), which are, of course, best explained as parts of a bathhouse. From the archaeological context, it is not possible to determine whether the statues were housed in this building and then toppled from their bases, decapitated and buried or removed to this location from elsewhere. Statue reuse was widespread in late antiquity;43 notably, the streets and squares of Constantinople were full of pagan statues removed from elsewhere.44 Possibly, statues were also removed from Amastris to Constantinople, where a semicircular space in front of the Capitol, richly adorned with statues, was called ta Amastrianou.45 Statues could also have been reused within a single urban area, being moved from one location to another and thus transforming the cityscape in late antiquity. The good state of preservation together with 34 35 36 37 38 39 40 41 42 43 44 45

Hoffmann 1989; Marek 1989: 379, fig. 2. Ainsworth 1842; Diest 1889; Hoffmann 1989: 210; Marek 1993: 93, n. 641. Diest 1889: 70; Marek 1989: 379, 382, fig. 2. Boré 1840: 232; cf. Marek 1989: 382–3. Marek 1993: 94–5; Ercyas 2003: 1422, 1424. Robert 1980: 151–63; cf. Marek 1993: cat. Amastris no. 56. Robert 1980: 156; cf. Marek 1989: 382; 1993: 93. See the results of the fairly recent British survey: Crow and Hill 1995. Mendel 1914: 345–6 no. 1108, 361–88 no. 1127. Leone 2013: 182–7. Lavan 2011a: xxxii; 2011b: passim. According to Mosch (2013: 119–20, n. 66), ta Amastrianou in Constantinople was adorned with statues removed from Amastris. However, according to Kedrenos (1.566) the location was named after its benefactor, who came from Amastris (cf. Berger 1988: 343), and thus the name does not relate to the origin of the statues.

232

Baran Aydın, Laura Buccino and Lâtife Summerer

the enormous weight of the statues under consideration, however, argues for a burial near their primary context. It is unlikely that such large sculptures could have survived longdistance transport without suffering any damage. In addition, although the statues range in date from the Julio-Claudian to the Antonine period, the nature of the deposit, which includes only portraits, two over-life-size and two life-size figures, suggests that they may have been displayed together as a group of imperial/honorific statues.46 The placement of the deposit next to the wall is surely no coincidence. The statues may have been displayed here prior to falling from their bases. The building with water installations can reasonably be interpreted as a thermae or even as a gymnasium-bath complex, and both gymnasia and thermae were appropriate locations for the setting up of imperial statues, as well as for the self-representation of a city’s elite. In late antiquity Amastris shrank dramatically, and the land walls of the city did not enclose the Bedesten area, including the Plateia and the sacred district.47 With the rise of Christianity, the religious values connected with this ‘sacred district’ must have shifted and this extramural location gained other meanings and functions. The statue group itself lost its spatial function, possibly with the abandonment or reorganisation of the building in which it was originally housed. In any case, the display of these statues was no longer appropriate to the use of the space. However, although they were no longer wanted for display, they were neither relocated nor simply discarded, but were buried and sealed by a stone floor.

5. The imperial cult in Amastris Although imperial statues were present in non-cultic contexts, there are some arguments which favour the cultic significance of the over-life-size cuirass statue representing, seemingly, the emperor Hadrian. It is known that many temples and shrines were founded for Hadrian on the occasion of his trip through Asia Minor.48 He probably visited Amastris in 124, and this could have prompted the establishment of his cult with statue dedications associated with Zeus Strategos.49 An inscription from 131 records that an altar with statue of a satyr was dedicated to Hadrian.50 Christian Marek envisages agonic games, held within the framework of the imperial cult, as the background for this dedication.51 The statue of a satyr was probably venerated as a cultic personification of these local games.52 The main argument for the cuirass statue being related to the local cult of Zeus is its support in the form of an eagle. Therefore, we can assume that the cuirass statue of Hadrian was associated with Zeus Strategos, who was, according to the epigraphic evidence, the main deity of Amastris. Based on this, we suggest that the statue was set up against the 46 47 48 49 50 51 52

The date of the imperial statues of the Augusteum of Narona spans from the early first to the end of the second century: Marin 2001: 92. Hill 1989: 13–4; Crow and Hill 1995: 262. Witulski 2011: 153–70. Birley 2000: 156; Mosch 2013: 125, n. 84. SEG 35.1317; Marek 1993: 96, cat. Amastris no. 10, pl. 27.1. Marek 1993: 95. Marek 1993: 96.

Honoured, beheaded and buried

233

wall where the burial was discovered and which was possibly the Kaisersaal of a bathgymnasium complex, and that it was associated with an earlier imperial female statue and two later male honorific statues. Imperial cult shrines such as this are well-known in gymnasium-bath-complexes in Asia Minor.53 However, it is not possible to know for certain whether this supposed emperor cult in Amastris was still active in late antiquity.

6. When and why were the statues buried? As indicated above, the stratigraphic description of the burial is very summary, but the excavators clearly note that the statues were sealed with flat stone slabs, presumably a stone pavement. Four coins found in the layer on top of this stone floor roughly date to the first half of the fifth century and thus provide a terminus ante quem for the burial.54 Although the absolute chronology of the erection, removal, beheading and burial of the statues remains unclear, a date in the late fourth or early fifth century for the burial seems the most feasible interpretation of the archaeological data. Late Antique statue burials are a widespread phenomenon throughout the ancient world. Recent studies note that the statues could be buried for a wide range of reasons and not all find contexts that include statues are deposits, i.e. intentionally buried.55 Statue burials should be distinguished from construction dumps and lime-kiln deposits. In the case of Amastris, the photographs taken at the time of the excavation clearly show that the statues were found arranged next to each other; thus it is unlikely that they fell or were discarded accidently (Fig. 3). The careful placement together with the fact that all the statues were decapitated, either by the removal of the separately-produced heads or by the necks being severed, demonstrates an intentional burial. Apart from the missing heads, the statues do not show signs of intentional damage that would indicate a hostile treatment. The state of preservation of the hands on both of the life-size, draped male statues is remarkable (Figs 7–8). A missing shoulder might have been scattered when the statue was taken down or the head was removed, while the weaker parts, such as feet, outstretched arms and hands with attributes, presumably became detached, when the statues fell from their bases. Although no detailed research has been carried out to date, some of the fragments found in the burial could be attributed to the cuirass statue. If this is indeed the case, it would demonstrate that the statues were fragmented immediately prior to or during their deposition. What happened to the separated heads is a difficult question to answer. Considering the fact that the statues were found densely packed in a small trench (4.1m × 2.2m), it is 53 54

Yegül 1982. Since the available photos of the coins do not show the reverses it is very difficult to identify: one of the coins (in the middle) surely belongs in the period of Valentinian or Theodosius, in the time 350–450. The narrow face reminds Valentinian II, Honorius or Theodosius II. The revers would provide information for a precise identification. The coins at the left below looks like fifth century coins. These coins seem to be an assemblage of the fifth century. We owe thanks Max Ritter for his expertise on this. 55 However, the new studies have shown that most statue caches should not be generally connected with Christians: Ambrogi 2011; Anghel 2011; Caseau 2011: 493–7; Lavan 2011a: xxix–xxxiii.

234

Baran Aydın, Laura Buccino and Lâtife Summerer

possible the burial also included the heads, which have remained undiscovered.56 Another possibility is that they were relocated elsewhere for reuse or discarded.57 As mentioned above, the excavators concluded that the statues had been hidden by pagans to protect them from possible destruction by Christians.58 This interpretation is based on general preconceptions about religious conflict in the city during the early Christian period.59 It is known from written sources that pagans did hide their statues to save them from destruction by Christians.60 However, these were mainly images of pagan gods, while the statues of former emperors were usually respected and redisplayed in public places in late antiquity.61 Although the find context itself provides no concrete evidence for the identification of the depositors, it suggests neither excessive haste under threat of danger nor hostile condemnation.62 As these statues had lost their place of significance and were no longer appropriate functionally, the depositors did not have it in mind to reuse, relocate or mutilate them. Whoever the depositors were, they clearly found it sufficient to de-individualise the statues by removing the portrait heads. Having lost their identity and significance, they were removed from view by being buried under a new floor. These observations raise the question to what extend the particular treatment of the statues was religiously or ideologically motivated. It is possible that Christians beheaded and buried the statues fearing their malefic influence.63 Although such demonisation usually affected statues of the gods, the images of former emperors were treated in this way when they held a religious significance.64 The statues might also have been beheaded and buried by pagans to prevent their reuse when the function of the cult place which housed these statues changed. The simplest explanation for the burial, however, is that it resulted from the abandonment and transformation of the building in which the statues were housed.65 Anna Leone has recently proposed that in the case of most statue deposits, the sculptures were stored for reuse at a later date and then simply cleared away when the space was reorganised.66 56 57

58 59 60 61 62

63 64 65 66

Heads were found close to the decapitated statues in, for example, the Mithreum of London: Anghel 2011: 199. Another unpublished find context with marble sculptures comes from the river Parthenios and includes mostly heads broken from statues. The authors of this article aim to publish these sculptures too. AteşoğullarI and Şmşek 1995: 94. For a recent discussion on the widespread hypothesis of ‘hidden statues’ in late antiquity, see Anghel 2011: 289–90. Caseau 2011: 485–8. Lavan 2011b: 467. On written sources on the hiding of statues and the threat posed by Christians to statues, see Anghel 2011: 265–74; Caseau 2011: 479–85. On statues destroyed by Christians usually being marked as such by signs such as carved crosses, intentionally mutilated limbs and smashed heads, see, for example, the imperial cult temple in Eretria: Schmid 2001: 81–142. On written sources on the demonisation of pagan statues, see Caseau 2011: 479–83. On the danger posed by pagan statues to Christians, see Lavan 2011b: 444 with references. Lavan 2011b: 459. On statue burials as end-use rituals, see Lavan 2011a: xxxi, n. 60. Leone 2013: 122–88.

Honoured, beheaded and buried

235

Applied to the case of Amastris such an interpretation would, however, not explain why the burial was so carefully arranged and sealed. At any rate, whoever the depositors were and whatever their motivation, the burial itself demonstrates an essential shift not only in the significance of these statues but also in the organisation of the urban space. When the fortified city was reduced in late antiquity the new land walls separated from the urban centre an area which had been considered ‘sacred’. The statue group, as well as the building in which it had been housed, no longer had a spatial function and both were cleared away to enable reuse of the area for another purpose.

7. Conclusions A stylistic analysis of four headless statues which were revealed during rescue excavations in Amasra in 1993 demonstrates that they span in date the Julio-Claudian to the Antonine period. Their arrangement and the coherent composition of the deposit suggest that they were displayed as a group in a public building close by, possibly in a gymnasium-bath complex. The lack of heads prevents positive identification, but the over-life-size cuirass statue may well depict Hadrian, as statues of this type were very common under this emperor in the eastern part of the empire. The statue could have been set up on the occasion of Hadrian’s visit to Amastris, probably in 124.67 The equally over-life-size statue of a draped woman dating to the Julio-Claudian period probably represents an empress of this time, while both the life-size draped male statues probably portray two nobles of the city from the second century. The topographical setting of the find-spot in the ‘sacred district’ suggests that the presumed statue of Hadrian was connected with the imperial cult associated with the local cult of Zeus Strategos. If the identification of the building as a gymnasium-bath complex is correct, the cuirass statue can also be associated with imperial honours celebrated in agonic games. The building seems to have been abandoned in late antiquity when changes in the religious and political structures must have been reflected in city design and architectural features. Amastris shrank and the new land walls of the city separated the ‘sacred district’ from the urban centre. The space, now outside the fortified city, was reused for another purpose, as indicated by the burial of the statues of the preceding building under a stone floor. Although the excavated evidence is limited, this statue deposit from Amastris provides a glimpse into the transformation of the urban landscape in late antiquity. Since the archaeological potential of the find context has seemingly not yet been exhausted, future excavations in this area may modify these preliminary conclusions.

67

Birley 2000: 156–7; Mosch 2013: 125, n. 84. Dedications for Hadrian in Amastris are attested also by inscriptions: Marek 1993: 96, cat. Amastris nos 9–10, pl. 27.1; SEG 35.1317 (dedication of a satyr and an altar in 131: see above and n. 50).

236

Baran Aydın, Laura Buccino and Lâtife Summerer Baran Aydın Amasra Museum Kum Mahallesi, Çamlık Sokak 4, Amasra Merkez/Bartın, Turkey [email protected] Laura Buccino Università degli Studi di Firenze Scuola di Specializzazione in Beni archeologici Via Laura 48, 50121 Firenze, Italy [email protected] Lâtife Summerer Ludwig-Maximilians-Universität München Institut für Klassische Archäologie Katharina-von-Bora-Straße 10, 80333 München, Germany [email protected]

Bibliography Abbreviations SEG: Supplementum Epigraphicum Graecum. J. C. Gieben, Amsterdam 1923–.

Literature Ainsworth, W. F. 1842. Travels and Researches in Asia Minor I. London. Alexandridis, A. 2004. Die Frauen des römischen Kaiserhauses. Eine Untersuchung ihrer bildlichen Darstellung von Livia bis Iulia Domna. Mainz am Rhein. Ambrogi, A. 2011. ‘Sugli occultamenti antichi di statue. Le testimonianze archeologiche a Roma’, Mitteilungen des Deutschen Archäologischen Instituts. Römische Abteilung 117, 511–66. Anghel, S. 2011. Burying the Gods: Depositing Statues in Late Antiquity (PhD Dissertation at the Columbia University). Ann Arbor. AteşoğullarI, S. 1999. ‘Amasra Müzesi’nden Zırhlı bir İmparator Torsosu’, Anadolu Medeniyetleri Müzesi 1998 Yıllığı, 207–18. AteşoğullarI, S. and Ö. Şmşek 1995. ‘Amasra ilçesi Küçüksanayi sitesi inşaasında bulunan Roma dönemi heykellerini kurtarma kazısı’, in: 5. Müze Kurtarma Kazıları Semineri, Didim 25–28 Nisan 1994. Ankara, 93–111. Belfiore, S. 2009. Il Periplo del Ponto Eusino di Arriano e altri testi sul Mar Nero e il Bosforo. Spazio geografico, mito e dominio ai confini dell’Impero Romano. Milano. Berger, A. 1988. Untersuchungen zu den Patria Konstantinopoleos (Ποικιλά Βυζαντινά 8). Bonn. Bergmann, B. 2010. ‘Bar Kochba und das Panhellenion. Die Panzerstatue Hadrians aus Hierapytna/ Kreta (Istanbul, Archäologisches Museum Inv.Nr. 50) und der Panzertorso Inv.Nr. 8097 im Piräusmuseum von Athen’, Istanbuler Mitteilungen 60, 203–89. Birley, A. R. 2000. Hadrian. The Restless Emperor. London. Boatwright, M. 2000. Hadrian and the Cities of the Roman Empire. Princeton. Boré, E. 1840. Correspondance et Mémoires d’un voyageur en Orient. Paris. Cadario, M. 2014. ‘L’immagine militare di Adriano’, in E. Calandra and B. Adembri (eds), Adriano e la Grecia. Villa Adriana tra classicità ed ellenismo. Studi e ricerche. Milano, 106–13. Caseau, B. 2011. ‘Religious Intolerance and Pagan Statuary’, in Lavan and Mulryan (eds), 479–502. Cavalieri, M. and S. Jusseret 2009. ‘Hadrien et la Crète. Le témoignage des statues cuirassées de Gortyne et de Knossos’, Mitteilungen des Deutschen Archäologischen Institut. Athenische Abteilung 124, 357–402. Crow, J. and S. J. Hill 1995. ‘The Byzantine Fortifications of Amastris in Paphlagonia’, Anatolian Studies 45, 251–65. Diest, W. von 1889. ‘Von Pergamon über den Dindymos zum Pontus’, Petermanns Geographische Mitteilungen. Ergänzungsheft 94, 68–71.

Honoured, beheaded and buried

237

Dillon, Sh. 2006. Ancient Greek Portrait Sculpture: Contexts, Subjects, and Styles. Cambridge. Ercyas, B. 2003. ‘Heracleia Pontica – Amastris’, in D. V. Grammenos and E. K. Petropoulos (eds), Ancient Colonies in the Black Sea II. Thessaloniki, 1403–31. Fejfer, J. 2008. Roman Portraits in Context. Berlin and New York. Filges, A. 2000. ‘Himationträger, Palliaten und Togaten: Der männliche Mantel-Normaltypus und seine regionalen Varianten in Rundplastik und Relief ’, in T. Mattern and D. Koro (eds), Munus. Festschrift für Hans Wiegartz. Münster, 95–109. Friedland, E. A. and R. H. Tykot 2012. ‘Quarry Origins, commission, and import of the marble sculptures from the Roman Theatre in Philadelphia/Amman, Jordan’, in A. Gutiérrez GarciaMoreno, M. P. Lapuente Mercadal and I. Rodà de Llanza (eds), Interdisciplinary Studies on Ancient Stone. Proceedings of the IX ASMOSIA Conference in Tarragona. Tarragona, 52–60. Geominy, W. 2004. ‘Die Zeit von 390 bis 360 v.Chr.’, in P. C. Bol (ed.), Die Geschichte der antiken Bildhauerkunst, II. Klassische Plastik. Mainz, 259–302. Guldager Bilde, P. et al. 2007–2008. ‘Archaeology in the Black Sea Region in Classical Antiquity 1993–2007’, Archaeological Reports for 2007–2008 (London) 54, 115–73. Hallett, Ch. 2006. ‘Catalogue no. 48’, in Smith (ed.), 150–1, 172. Hill, St. 1989. ‘Survey Work at Amasra, Zonguldak’, Anatolian Studies 39, 13–4. Hoffmann, A. 1989. ‘Zum “Bedesten” in Amastris. Ein römischer Marktbau?’, Istanbuler Mitteilungen 39, 197–210. Karanastasi, P. 2012–2013. ‘Hadrian in Panzer: Kaiserstatuen zwischen Realpolitik und Philhellenismus’, Jahrbuch des Deutschen Archäologischen Instituts 127–128, 323–91. Kruse, H. J. 1975. Römische weibliche Gewandstatuen des 2. Jh. n.Chr. (PhD Dissertation). Göttingen. Lavan, L. 2011a. ‘The End of the Temples: Towards a New Narrative?’, in Lavan and Mulryan (eds), xv–lxiv. Lavan, L. 2011b. ‘Political Talismans? Residual “Pagan” Statues in Late Antique Public Space’, in Lavan and Mulryan (eds), 439–78. Lavan, L. and M. Mulryan (eds) 2011. The Archaeology of Late Antique ‘Paganism’. Leiden and Boston. Lenaghan, J. and S. Dillon 2006. ‘Catalogue no. 58’, in Smith (ed.), 185–6. Lenaghan, J. and R. R. R. Smith 2006. ‘Catalogue no. 49’, in Smith (ed.), 176–7. Leone, A. 2013. The End of the Pagan City: Religion, Economy, and Urbanism in Late Antique North Africa. Oxford. Marek, Chr. 1989. ‘Amastris. Geschichte, Topographie, archäologische Reste’, Istanbuler Mitteilungen 39, 373–89. Marek, Chr. 1993. Stadt, Ära und Territorium in Pontus-Bithynia und Nord-Galatia. Tübingen. Marek, Chr. 2003. Pontus et Bithynia: die römischen Provinzen im Norden Kleinasien. Mainz am Rhein. Marin, E. 2001. ‘The Temple of the Imperial Cult (Augusteum) at Narona and its Statues: Interim Report’, Journal of Roman Archaeology 14, 81–112. Mendel, G. 1914. Catalogue des Sculptures grecques, romaines et byzantines III. Paris. Mosch, H.-Ch. von 2013. ‘Hadrians “Sandalenlöser”. Der Hermes des Lysipp (?) auf den Münzen von Trapezous, Amastris und Markianopolis’, Jahrbuch für Numismatik und Geldgeschichte 63, 93–149. Polaschek, K. 1969. Untersuchungen zu griechischen Mantelstatuen. Der Himationtypus mit Armschlinge (PhD Dissertation). Berlin. Reinsberg, C. 2005. ‘Demeter, Artemisia und die pietas Augustae: zur spätklassischen Statue der Orans’ in Festschrift for Ramazan Özgan. Istanbul, 297–316. Robert, L. 1980. A travers l’Asie Mineure. Paris. Schmid, St. 2011. ‘Worshipping the Emperor(s): a New Temple of the Imperial Cult at Eretria and the Ancient Destruction of its Statues’, Journal of Roman Archaeology 14, 113–42. Scholz, B. I. 1992. Untersuchungen zur Tracht der römischen Matrona. Köln, Weimar and Wien. Sharp, R. St. 2010. The Outside Image: A Comparative Study of External Architectural Display on Middle Byzantine Structures on the Black Sea Littoral (PhD Dissertation). Birmingham. Smith, R. R. R. 1998. ‘Cultural Choice and Political Identity in Honorific Portrait Statues in the Greek East in the second century A.D.’, Journal of Roman Studies 88, 56–93. Smith, R. R. R. (ed.) 2006. Roman Portrait Statuary from Aphrodisias, Mainz am Rhein.

238

Baran Aydın, Laura Buccino and Lâtife Summerer

Valeri, C. 2013, ‘Catalogue no. IX.3’, in E. La Rocca, C. Parisi Presicce and A. Lo Monaco (eds), Augusto. Catalogue of the Exhibition in Rome, Scuderie del Quirinale, 18 November 2013 – 9 February 2014. Milano, 319–20. Waywell, G. B. 1978. The Free-standing Sculptures of the Mausoleum at Halicarnassus in the British Museum: a Catalogue. London. Witulski, Th. 2011. Kaiserkult in Kleinasien. Die Entwicklung der kultisch-religiösen Kaiserverehrung in der römischen Provinz Asia von Augustus bis Antoninus Pius. Göttigen. 2nd edition. Yegül, F. K. 1982. ‘A Study in Architectural Iconography. Kaisersaal and the Imperial Cult’, The Art Bulletin 64, 7–31. Zavagno, L. 2009. Cities in Transition: Urbanism in Byzantium between Late Antiquity and the Early Middle Ages (BAR International Series 2030). Oxford. Zavagno, L. 2012. ‘Amastris (Paphlagonia): a Study in Byzantine Urban History between Late Antiquity and Early Middle Ages’, in G. R. Tsetskhladze (ed.), The Black Sea, Paphlagonia, Pontus and Phrygia in Antiquity (BAR International Series 2432). Oxford.

Luisa Musso

Dynamics of mosaic design in northern Anatolia in the Roman and early Byzantine periods

1. Introduction Since the 1990s northern Anatolia has moved increasingly to the centre stage of archaeological research, with projects ranging from regional surveys to the reconstruction of the ancient landscape to the organisation and publication of several symposia.1 This volume is itself another example of the growing research interest in this region. This particular article provides a survey of some of the mosaic pavements found in northern Anatolia. The nature of chance finds (Fig. 1), incomplete excavations and a lack of reliable data on archaeological contexts have all led to mosaics being considered, for the most part, simply as ‘patterns and panels’; they have tended to be divorced from their architectural setting and thus largely deprived of the necessary interaction between artefact and context. Without a systematic and comprehensive distribution map of mosaics, correlated with a continuously updated census of finds, our ability to define production centres, to identify regional characteristics and to gauge the intrinsic value of mosaics as an expression of handicraft traditions, as a vector of trade relations and as an indicator of local prosperity is severely handicapped. In the absence of such a map, any attempt to define a chronological sequence of mosaics cannot but lack credibility. There are, however, a number of positive tools and developments for the study of mosaics. We may refer to the annual publications of excavation reports, topographical surveys and museum activities published by the General Directorate for Cultural Heritage and Museums, the planning of a corpus of the mosaics of Turkey,2 the creation of the Center for Mosaic Research at Uludağ University, Bursa, and, not least, the introduction of the Journal of Mosaic Research. The territorial extent of this region prompted the need to establish, for the purposes of this article, a geographic boundary to the east along the course of the Yeşilırmak river (the ancient Iris); significant finds of mosaics of the Late Antique period in the territory of Amasya is presented in the report of Esra Keskin in this volume. It is not my intention to present a place-by-place digest of Pontic ‘mosaic topography’. Instead I will focus on five particular case studies: the coastal sites of Samsun (Amisos), Sinop (Sinope) and Kadıoğlu 1 2

Kassab Tezgör 2013. The coordinators of the project are Mustafa Şahin, President of AIEMA-Türkiye, Werner Jobst and David Parrish: see Kökten 2011; Parrish et al. 2011.

240

Luisa Musso

Fig. 1: Map showing the sites of provenance of the mosaics (cartography by R. Szydlak). Köyü (Tios/Tieion), and the site of Tașköprü (Pompeiopolis) and Eskipazar (Hadrianopolis) of inner Paphlagonia. The criteria I adopted in making this choice include our knowledge of the particular context or, as in the case of the mosaic from Samsun, the value of the mosaic itself in terms of composition, iconography, content and workshop.

2. Amisos A mosaic measuring 6m × 8m was found in 1958 during excavations at Toraman Tepe (Fig. 2). The attribution to a villa3 is possible, if only hypothetical. Two figurative panels are placed in geometric settings. The focus of the entire composition is the panel portraying Achilles with his mother Thetis, both identified by their names in Greek letters (Fig. 3). The setting is minimal: the massive figures stand out from the landscape which ensures a pictorial depth with the sea in the background and a cavern behind Thetis.4 Eight contiguous panels are symmetrically placed round the main panel: four rectangular panels show Nereids riding on the backs of marine monsters, while four square panels at the corners are filled with female busts of the Seasons. The ‘compartmental’ articulation of the composition is well known in both the East and West:5 a similar layout is found in the mosaic of a domus at Trinquetaille (Arles) which displays a personification of Time surrounded by panels with Nereids riding on Tritons and the Seasons (putti).6 The framing system is 3 4 5 6

Şahn 2004; 2005; 2006. Homer, Iliad 18.66, 18.86. Cf. DGMR II: pl. 409b; Levi 1947: 1, 68, 69, fig. 28. Musée d’Arles antique: Sintès 1996: 107–9, no. 87 (V. Blanc-Bijon), late second century; Lancha 1997: 104–5, no. 52, pls B, 38.

Dynamics of mosaic design in northern Anatolia

Fig. 2: Samsun Museum: Achilles mosaic (photo by L. Summerer).

Fig. 3: Samsun Museum: Achilles mosaic, detail (photo by L. Summerer).

241

242

Luisa Musso

Fig. 4: Samsun Museum: Achilles mosaic, detail (photo by L. Summerer). elaborate: all the panels are surrounded by a wave pattern and linked by a guilloche; on the outside there is a second wave pattern, with inverted colours. The main panel is enclosed on three sides by a pattern based on hexagons; squares are joined to the hexagons, with triangles formed in the reserved areas between them, enriched with geometric ornament, and create the effect of interlocking dodecagons.7 A second figured panel (Fig. 4) is enclosed with a different pattern: rows of adjacent octagons are linked by rectangles alternating with poised squares, with lozenges flanking their sides.8 It shows a scene of sacrifice and faces towards the panel of Achilles and Thetis. A white bull, its rear hooves tied together, is about to be immolated by a victimarius wielding an axe, while the officiant, dressed in a tunic adorned with clavi and a himation with a patera in one hand and a sprig in the other, is in the act of placing offerings on an altar.9 A proposed date in the first half of the third century can be endorsed.10 The main panel of the mosaic relates to the Homeric episode of Thetis inciting her son Achilles, who is consumed by grief at the loss of Patroclus, to arm himself and join the battle.11 The precise moment represented is that of the arming of Achilles, in the presence 7 8 9 10

DGMR I: pl. 205c. DGMR I: pl. 176b (variant). Şahn 2005: 423; 2006: 153 identifies the figure of the officiant as a portrait of the patron. D. Parrish (in Şahin 2005: 424) proposes that the mosaic be dated to the period of Caracalla, who visited the tomb of Achilles in 214: see Grossardt 2006: 2.726; Rutherford 2009: 237–8. 11 Homer, Iliad 19.1–36.

Dynamics of mosaic design in northern Anatolia

243

of his mother who had procured the divine gift of the arms for him: Achilles is accompanied by the arms forged by Hephaestus, those ‘glorious arms, exceeding fair, such as no mortal shoulders ever bore’.12 Achilles holds the spear in his left hand, the shield is at his side and the helmet is placed on a column-shaped support, while the sword remains in its scabbard in Thetis’ hand. The scene of the consignment of the new arms, combined with the ceremony of putting them on, represents the hero’s investiture;13 indeed the ‘bright treasure’ forged by the god is not only the motif around which the life and death of the hero revolve,14 but also the prefiguration of his ‘glorious death’.15 The emotional tension of the scene is emphasised by the contrasting psychological situations of the protagonists: Achilles is shown as a young boy full of bravado, while Thetis is lost in thought, filled with a presentiment of her son’s not too distant end. The heroic nature of Achilles is heightened by his adolescent appearance, his nudity and his posture, raised over a kind of pedestal. The Nereids form part of the retinue of Thetis;16 even if they do not transport the arms,17 as ‘divine escorts’ they participate in the crucial moments of the life of Achilles.18 The connection between this image and the scene of sacrifice is corroborated by their arrangement: one composition faces the other. Much emphasis has been placed on the significance of this mosaic in terms of the diffusion of the cult of Achilles and its ‘local’ colouring in the Black Sea region,19 with its centre at Olbia, the seat of a kind of ‘state cult’ of Achilles heros/theos,20 and this topic requires a short discussion. According to the literary tradition, the cult went back to Thetis’s initiative to transfer the hero’s mortal remains to the island of Leuke.21 In spite of the fascination exerted by the myth and the spell cast by the name of Leuke, there is not the slightest evidence to support the suggestion that this iconography is exclusively ‘Pontic’: the subject is neither the transport of Achilles to the island of Leuke nor the cult of the hero on the island, where the co-protagonist is not Thetis, but Helen. The alleged ‘regional’ dimension of the cult is repudiated by its far wider diffusion from Greece to western Asia Minor, from the Black Sea to Sicily.22 The iconography of the sacrificial scene points to the cult of a hero who turned himself into a god: for whereas heroic cult prescribed sacrifices in pits, divine sacrifices were celebrated on an

12 13 14 15 16 17

18 19 20 21 22

Homer, Iliad 19.10–11. Menichetti 2009a: 146–52. An eloquent image is provided by the cornelian signed by Dioscurides: Vollenweider 1966: 61, 113, pl. 63.1, 3–4; Ghedini 1994: 310. Mavrojannis 1994: 345. Menichetti 2009b: 106. Icard-Gianolio and Zvabados 1992: 786; 2009; Neira 2011: 636–40. In Homer’s account, Thetis goes alone to visit Hephaestus (Iliad 18.503–7) and she alone hands over the arms to her son (Iliad 19.1–13). By contrast, in the central composition of the lost Aeschylean trilogy dedicated to Achilles (Nereids), it is to the Nereids that the honour of handing over the arms to the hero is entrusted: Kossatz-Deissmann 1981: 122. Barringer 1995: 49–54; Mavrojannis 1994: 302–3, 309–10. Şahn 2005: 420–3; 2006: 153, 157–60; Ochal-Czarnowicz 2011: 271–3. Hedreen 1991; Hupe 2006; 2007; Bujskikh 2007. Kossatz-Deissmann 1981: 193–5; Mavrojannis 1994: 303; Barringer 1995: 50–2. Kemp-Lindemann 1975: 242–7; Hommel 1980; Mavrojannis 1994: 313–22.

244

Luisa Musso

altar.23 Enlightening is a passage in the dialogue On Heroes,24 written by Flavius Philostratus in the years of the principate of Severus Alexander, in which Achilles, hero and god, assumes a clear centrality.25 According to the prescription of the Oracle of Dodona, the Thessalians, in a very remote period, were each year obliged to send an embassy by sea to Troy and perform, in keeping with the twofold nature – heroic and divine – of Achilles, rites of a twofold nature. The ritual of the sacrifice was duplicated in two distinct acts,26 culminating in the killing, respectively, of a black bull and a white bull. The immolation of the black bull, consecrated to Achilles the hero, had a markedly chthonian character and took place, after the nocturnal disembarkation, in bothroi sunk into the top of his tumulus; the sacrifice (thysia) of the white bull, in conformity with a ritual fitting for a god,27 took place on the seashore, prior to embarkation at dawn. So in the Heroikos Achilles is indeed the hero after whom the sanctuary on the island of Leuke was named, but to him a transregional cult was dedicated: a cult enhanced by a link to Alexander the Great, who, in his emulation of Achilles and as part of his own self-heroisation, landed at Troy and sacrificed on the tomb of his ‘Homeric counterpart’.28 In the lower left of the mosaic is placed the signature of the mosaicist: Ὀρέντης ἐψηφοθέτησα (‘I, Orentes, laid the tessellatum’). Orentes has been considered a variant of the placename Orontes:29 the derivation of a personal name from that of a river is well known. The name Ὀρέντης is attested at Athens in the Imperial period;30 its diffusion in Rome (in Latin Orontes or Orentes) has been studied by Heikki Solin who has collected nine instances between the Augustan period and the first half of the third century, most of them referring to slaves or freedmen.31 In our case, it has already been inferred from the name that the mosaicist must have come from the region of Antioch.32 Of course, the greatest prudence is needed in suggesting the actual transfer of craftsmen from one region to another. The Antiochene style may,33 however, be recalled both by the composition of the Samsun mosaic and by its system of framing the figurative panel; just such a frame is closely paralleled by the sumptuous image of the hierogamia of Dionysos and Ariadne at Zeugma.34 With the present state of our knowledge, the development of mosaic pavements in Amisos and its territory cannot be followed systematically. Nevertheless, it is worth recalling a number of geometric floor mosaics, datable between the fifth and sixth century, recovered 23 24 25 26 27 28 29 30 31 32 33 34

Mavrojannis 1994: 299–300. Philostratus, On Heroes 52.3–54.1; Mavrojannis 1994: 308; Berenson Maclean and Aitken 2003: 76–81; Grossardt 2006: 1.235–6, 2.719–28; Rutherford 2009: 234–5, 242–3. Philostratus, On Heroes 45.1–47.14; Aitken 2001; Grossardt 2006: 2.721; Rutherford 2009: 238–45. Philostratus, On Heroes 53.8. Philostratus, On Heroes 52.13. Diodorus Siculus 17.17–8; Plutarch, Alexander 15.7–9; Arrian, Anabasis 1.11–2. Donderer 2008: 27, 64–5, A20, pl. 9.2. Athenians Project http://atheniansproject.com/: no.271555 (accessed January 2015). Solin 2003: 699, s.v. Oronta. See Kondoleon in Şahn 2005: 425; Becker and Kondoleon 2005: 38. Cf. Şahn 2005: 424. Campbell 1999; Önal 2009: 80–1, late second/early third century.

Dynamics of mosaic design in northern Anatolia

245

Fig. 5: Samsun, Toraman Tepe, Mosaic Hall, detail (after Aybek and Öz 2012: fig. 5). during the course of rescue excavations conducted by the Samsum Museum in the 1990s in the area of the Samsun Sahara Medical School and Training Center (Fig. 5).35 Other mosaics of identical technique and standard of workmanship have been found in the area of the US military base during construction work at the base (Fig. 6).36 Decorative patterns and centralised compositional schemes (grid pattern of swastika-meander;37 orthogonal pattern of tangent circles;38 ‘starred octagon’39) are in general conventional,40 and the repertoire of forms is standardised.

35 36

37 38 39 40

Akkaya 1993; 1998; Ertuğrul and Atasoy 1998; Aybek and Öz 2012. Images of these mosaics are available on the website of an American soldier, George William Durman, who was stationed at the Samsun base between 1958 and 1960: http://www.merhabausmilitary.com/1DURMANindex.html (accessed January 2015). Cf. DGMR I: pl. 194c; II: pl. 407b. DGMR II: pl. 403a. DGMR II: pl. 373a. In the composition with starred octagon, note the presence of an elaborate version of the so-called ‘silver-plate pattern’: DGMR II: 43; Donceel-Voûte 1994: 206, 210. The repertoire of this pattern, originating from opus sectile decoration, is well attested in Antioch (Levi 1947: 1.285, 423, 424, 436, 468, 2. pls 82a–b, 114a, 123c) and in northern Syrian pavements of the late fourth to mid

246

Luisa Musso

Fig. 6: Samsun, Toraman Tepe, mosaic (after http://www.merhabausmilitary.com/1DURMANindex.html).

3. Sinope Another colony of Milesian foundation, Sinope, has yielded a rich assortment of pavement decorations; a portion of pebble pavement, with a frieze formed by acanthus scrolls, represents the earliest example of this type of composition in Asia Minor.41 A large mosaic pavement with a circular centralised geometric composition, displayed in the Sinop Museum,42 shows the seven busts of the Muses in hexagonal medallions arranged in a circle and identified by their attributes. Set against a dark background, their heads are wreathed with leaves and distinguished by feathers; Melpomene is placed at the centre. The subject of the Muses is in itself conventional;43 the predilection for furnishing homes with themes associated with the paideia is well-known. The compartmental organisation is perfectly suited to this iconography: we can recall the magnificent mosaic of Zeugma, where the Muses, in the canonical number of nine, are each distinguished by their names and not by their attributes,44 and that from Kos, in which the busts of the nine Muses stand out 40

41 42 43 44

fifth century (Donceel-Voûte 1988: 117–9, 123–5; Zaqzuq 1995: 238–40, 250–1, figs 23–4; Zaqzuq and Piccirillo 1999), but was also used in western Asia Minor and at Constantinople: Jobst 2011: 492, 494, figs 26–7, 501. Salzmann 1982: 36, 113, no. 123, pl. 63.1, final decades of the fourth century BC. http://www.pbase.com/osmantanidik/sinop_mosaics (accessed January 2015). Theophilidou 1984; Ternes 2003. Uluslararası Kazı, Araștırma ve Arkeometri Sempozyumu’nun 30. yılı anısına (2008): poster. The mosaic from Baalbek with Calliope encircled by busts of Socrates and the Seven Sages is identical in terms of composition and decorative patterns: Theophilidou 1984: 261–3, no. 4, fig. 9; Donderer 1989: 53–4, no. A2, pl. 2, third century.

Dynamics of mosaic design in northern Anatolia

247

against a black background in circular compartments.45 What is peculiar to the example in Sinop is that it includes the arms of the Muses: this helps to lend animation to their pose. Despite its by no means high quality, the mosaic reveals a pronounced volumetric effect, the pursuit of a style of greater movement and the use of highlights in the faces: features that distinguish this version from the rigidity and repetitiveness of the example from Kos. The latter is dated between the late third and early fourth century, while a date in the third century would seem appropriate for the mosaic in Sinop. The compositional scheme, consisting of a circle enclosing a triaxial pattern as a honeycomb, is paralleled by a large number of examples, concentrated in the western part of the empire and with a maximum diffusion in the second century.46 However, it is the many polychrome mosaics displayed in the garden of the Sinop Museum that deserve particular attention.47 Transferred to concrete panels and awaiting re-assemblage, this formidable collection of geometric mosaics presents an opportunity to trace an evolutionary model of geometric schemes, decorative motifs and ensembles at the regional level.48 It may also be used to elucidate their relation with other geographic areas, and to reconstruct the chronological sequence of a genre of craftsmanship that has not yet been adequately appreciated. This in turn would help overcome, at the regional level, the preponderant reference to the sequence of Antioch mosaics proposed by Doro Levi over 70 years ago.49 These mosaics are also diagnostic in another sense: they attest the spread of compositions with purely geometric schemes – of unified patterns and centralised compositions – after the mid fourth century. This phenomenon has been magisterially analysed by Ernst Kitzinger:50 the ‘taste for the geometric’ was disseminated through a predilection for abstract decoration and polychrome richness, emphasised by the massive use of the ‘rainbow style’, that was developed from the mid fourth century onwards in conformity with the tendency to banish the figurative repertoire from pavements.51 One pavement, dismembered into several panels, offers a version of the widely disseminated orthogonal pattern of tangent octagons, hexagons and crosses (Fig. 7.1–2).52 The geometric grid formed by the interlocking octagons, hexagons and crosses is characterised by the precision of the technique, the sobriety of fillings and the simplicity of the border. An original aspect is the way the filling of one elongated hexagon has been transformed by the mosaicist into a lizard. The octagons are filled with a varied selection of avian fauna, rendered with a bright and varied palette. In the more peripheral octagons appear flower 45 46 47

48 49 50 51 52

Theophilidou 1984: 244–7, no. 18, fig. 2; De Matteis 2004: 138–40, no. 65, pl. 80. Werner 1998: 306–13. For present purposes I will limit myself to reviewing, without consideration of their contexts, the geometric schemes of a limited number of floor mosaics, with a view to gaining an overall picture of mosaic production in the Pontic area. The limitations inherent in a criterion based on the preference accorded to particular ‘motifs’ have been underlined by Clarke 1994 and Poulsen 2012; see also Donceel-Voûte 2001. Balty 2001: 303–5. Kitzinger 1965. Levi 1947: 1.405–7; Balty 1984: 440–51; 1989: 508–11; 1995: 170; Hellenkemper Salies 1991: 325; Scheibelreiter 2008: 99. DGMR I: pl. 180b–c.

248

Luisa Musso

Fig. 7.1–2: Sinop Museum, garden (photo by G. Bertolotto). heads with elongated petals, rendered with subtle gradations of colours: a peculiarity that reinforces the attribution of the mosaic to the third century.53 The same geometric scheme, but with very different aesthetic effects, recurs in an example with a decorative geometric repertoire and motifs in the ‘rainbow style’ that can be dated to the late fourth or early fifth century (Fig. 7.4).54 The polychrome filling of the hexagons with chevrons and the type of ‘shaded rosette’55 recur, within an identical geometric grid, at Antioch, Bath D56 and, in a later version, in the Constantinian Villa.57 The

53

54 55 56

57

Cf. Antioch, House A: Campbell 1988: 25, IVA9, pl. 78, first half of the third century; Metropolis, ‘Bankettsaal’: Scheibelreiter-Gail 2011: 304–5, Kat. 84, 605, figs 384–5, second or early third century. Campbell 1988: 96, 180b; Hellenkemper Salies 1991: 322–3, fig. 5. DGMR II: pl. 255i. Bath D, fourth century (Levi 1947: 1.427, 429, fig. 164, 2. pl. 117; Campbell 1988: 16–7, IVA5, figs 10–11, pls 53, 55, 57–8); martyrium of St Babylas, dated to 387, south wing (Levi 1947: 1.423– 4, 2. pl. 114a). Levi 1947: 1.436, 2. pl. 123c. Similar filling motifs are found in the baptistery of the martyrium of St Babylas, fifth century: Levi 1947: 1.425, 2. pl. 115c; Campbell 1988: 46, 18f, 93, 147a, pl. 135.

Dynamics of mosaic design in northern Anatolia

249

Fig. 7.3: Sinop Museum, garden (photo by L. Buccino). composition based on a pattern of tangent interlooped circles (Fig. 7.3) should belong to the same context:58 a guilloche and a rainbow-cableare interlaced to form circles and squares; the rendering of the shaded rosette is identical to the previous example.59 The identification of at least three rectangular panels in the surviving fragments suggests a composition with continuous decoration.60 A significant parallel can be found in the narthex of the martyrium of Misis (Mopsuestia), Cilicia (early fifth century).61 The fashion for compositions based on a pattern of interlooped circles forming irregular concave octagons62 is testified by a corner fragment in which a guilloche and an undulating ribbon are interlaced, and the colours counterchanged. This kind of decoration enjoyed maximum

58 59 60 61

DGMR I: pl. 69e. Cf. Antioch, Bath C, second half of the fourth century: Campbell 1988: 36, IVA15a, 38, pl. 102. DGMR I: pl. 148g. Budde 1969: 55, ill. 12. The pattern recurs in centralised compositions: Budde 1969: 58, figs 27, 29, ills 134, 140; for the dating, see Hellenkemper Salies 1991: 325–8. The contrast with identical geometric compositions in western Asia Minor can be gauged by comparison with an example from Smyrna: Ersoy and Alatepel 2011: 20, fig. 3, drg. 3, 21–3, figs 5–8; Ersoy and Yolaçan 2012. 62 DGMR I: pl. 235a.

250

Luisa Musso

Fig. 7.4: Sinop Museum, garden (photo by L. Buccino). favour, in the oriental area, between the end of the fourth and the first decades of the following century.63 The Sinopean panel with large rosette (Fig. 8.1) formed part of a composition with a starred octagon; the fragments with compartments linked by a rainbow-cable (Fig. 8.2) belong to the same geometric pattern.64 Identical compositions are encountered at Antioch dating to between the end of the fourth and the fifth century: Yakto Complex, lower level,65 House of the Buffet Supper, upper level,66 and House of Ge and the Seasons, upper level.67 The comparable mosaic pavement of the Synagogue of Apamea dates to 391–2.68 In all these compositions the lozenges enclose the so-called ‘silver-plate pattern’: all present the characteristic ‘buttons’ in the corners of the octagons.69 The decorative motif inside the central clipeus finds an excellent parallel in a mosaic in Apamea.70 On the basis of these comparisons, and with the help of the repertoire from northern Syria, in many cases dated

63 64 65

66 67 68 69

70

Balty 1984: 449–50; 1995: 26–7, 81. DGMR II: pl. 373a. The centralised composition is widespread in the Syrian area: Balty 1984: 450; 1989: 511, 518; Donceel-Voûte 1988: 559, s.v. octogone étoilé; Jobst 2011: 500–1. Rockland (Maine), Wellesley College, Farnsworth Art Museum, second half of the fourth century: Stillwell 1938: 182, no. 41, pl. 29; Lassus 1938b: 110–4, figs 22–8; Levi 1947: 1.281, 422, 2. pl. 111a–b. Levi 1947: 1.311, 313, fig. 128, 2. pl. 125a, fifth century. Levi 1947: 1.347, 2. pl. 82a, fifth century; Balty 1999: 150; 2001: 311, n. 37. See also the composition of a mosaic in Kos, with a bust of Tyche at the centre: Brouscari 1997. Balty 1986a: 8, fig. 9; 1999: 143–4, pl. 53.1. This composition is present in the church of Hama dated to 416: Zaqzuq 1983: pls 12–3, 23–6. On this motif, characteristic of Syrian workshops, see Hellenkemper Salies 1987: 307; 1990b: 616, no. 2; Balty 1999: 143–4, 149; 2001: 309–10, 312, pls 2.1, 3.1–2; Jouejati 2012. In the dated Syrian examples the motif is attested between 370 and 442. Level below the cathedral, third quarter of the fourth century: Balty 1981: 208, fig. 233; DGMR II: pl. 335d. Close in style is the decoration of the clipeus with a bust of Tyche from Bet Shean (Scythopolis), early sixth century: Yeivin 1987–1988: 18–9, photo 9.

Dynamics of mosaic design in northern Anatolia

Fig. 8.1–3: Sinop Museum, garden (photo by L. Buccino).

251

252

Luisa Musso

Fig. 8.4–5: Sinop Museum, garden (photo by L. Buccino). by inscriptions ad annum, the example in Sinop may be dated between the last quarter of the fourth and the first quarter of the following century. The panel with motifs en entrelacs (Fig. 8.3) is also part of the wide repertoire of centralised compositions with composite patterns; it consists of four rainbow-cables that curve to form four ellipses, arranged along the diagonals of the square, each containing a poised square bordered by a guilloche. The taste for interlaced patterns is well attested in the eastern territories, though is not exclusive to this region. An excellent analogy is offered by

Dynamics of mosaic design in northern Anatolia

253

the northern aisle of the already cited church of Misis-Mopsuestia (Cilicia) which dates to the early fifth century.71 The popularity enjoyed by these compositions is confirmed by the exceptional compositions in the church of Hama (416).72 Nor were they short-lived: they continued to be used right down to the sixth century.73 A number of large circular compositions that presumably formed part of the same pavement can be distinguished in a homogeneous ensemble of mosaics. One consists of a ‘wheel’ filled with rows of perspective cubes (Fig. 8.4);74 another ‘wheel’ is filled with a series of concentric circles arranged around a square. A third composition has a chequerboard pattern of orthogonal bands with squares overlapping the intersections, forming cruciform compartments occupied by poised squares (Fig. 8.5);75 a fourth consists of rectangular panels with decoration of lozenges and a double reverse-returned swastikameander enclosing squares and rectangles.76 The compositions are distinguished by decorative motifs in ‘rainbow style’. A date in the fifth century seems plausible. One particular observation emerges from this rapid survey: a clearly defined set of patterns and filling motifs associated with Syrian ateliers was widely used in the region of Sinop in the fifth century.77 At Çiftlik, on the coast at the end of a valley south of Sinop, polychrome mosaics from an early church and an associated building were discovered in the 1990s during the course of rescue excavations.78 Exclusively geometric and rendered with sharply contrasting colours, the church mosaics consist of a repetitive running-pelta pattern in the narthex,79 interlocking circles in both aisles and a multiple-design composition in the nave, the latter consisting of individual panels with circular and square designs, no two of which are identical, arranged within a regular chequerboard framed by a meander border (Fig. 9).80 The proposed date of these mosaics – to the early fifth century, if not to the late fourth81 – seems to be plausible given comparisons with examples of the late fourth century (e.g. Qausīyeh, martyrium of St Babylas, east aisle, 387;82 Kourion, Cyprus, episcopal complex, diakonikon, mosaic below the sixth-century pavement83). From the building forming part of this same 71 72 73 74 75

76 77

78 79 80 81 82 83

Budde 1969: 57–67, figs 27–30, 32, pls 132–42. Zaqzuq 1983; Balty 2001: 310, pl. 5.2. Donceel-Voûte 1995: 91, fig. 4, 93; Donceel-Voûte 1988: 138–45; AssimakopoulouAtzaka 2011: 113–4, pl. 4.24a; Jouejati-Madwar 2005. DGMR II: pl. 338c. DGMR I: pl. 146a (variant); Campbell 1988: 93, 146c. There are close parallels in Antioch dating to the fourth century (Campbell 1988: 10, IVA2e, pls 37–8; 29, IVA4a, pl. 46; 15, IVA5, pls 49, 60; 82, IVA39a, pl. 230) and the fifth century: Campbell 1988: 27–8, IVA10, pls 81–2. DGMR I: pl. 40a; Campbell 1988: 87, 39e. Other patterns used in the mosaics housed in the Sinop Museum, such as the quasi-tangent swastika-peltae with a central knot (DGMR I: pl. 223) or the network of squares formed by the intersection of orthogonal bands (DGMR I: pl. 144b, e), are also frequently encountered in western Anatolia: Scheibelreiter-Gail 2011: 170, 174–5; Jobst 2011: 493–7, 501. Hill 1995a; 1995b; 1996; 1997; TatlIcan 1997. DGMR I: pl. 222e. Hill 1995b: 226, pls 32–3; 1998b: 288, 292, figs 23.3, 43–4; TatlIcan 1997: figs 16–23. Hill 1998b: 292. Fillings in ‘rainbow style’ are absent. Lassus 1938a: 19–20, fig. 18; Levi 1947: 1.285, 424, 2. pl. 113b; Campbell 1988: 45, IVA18d, fig. 133. Megaw 2007: 68, fig. 1.S, 146–8, pl. 1.19e–f; Kondoleon 2007.

254

Luisa Musso

Fig. 9: Çiftlik (Sinop), view of the church from the west (after Hill 1998b: pl. 43, ©British Institute at Ankara, photographic reproduction by R. Hessing). complex comes a pavement that must have decorated a peristyle:84 in view of its precarious state of conservation it was transferred to the Sinop Museum.85 The central composition presents an elaborate orthogonal pattern of intersecting and spaced circles, defined by interlooped bands;86 it is bordered by a wide frame containing a row of swastika-peltae.87

4. Kadıoğlu Köyü The hinterland of Tios/Tieion provides an exceptional figured mosaic; it was only relatively recently unearthed during the course of rescue excavations and so has not yet received the scholarly attention it deserves.88 Its architectural setting is a Roman villa situated not far from the river Billaios (Filyos çayı).89 The emergency excavation of 2008 uncovered in the

84 85 86 87 88

Hill 1995b: fig. 1, pls 29b–30b; 1998b: 289–90; TatlIcan 1997: fig. 25. Hill 1998a. DGMR I: pl. 244d. DGMR I: pl. 223e. Mercan and Göçen 2012; LaflI and Christof 2012: 7, n. 32, 12, fig. 23a–c; The mosaics have been reinterred. 89 Marek 1988: 403–4; Öztürk 2012; 2013.

Dynamics of mosaic design in northern Anatolia

255

Fig. 10.1–2: Kadıoğlu Köyü, villa, apsidal room (after Mercan and Göçen 2012: figs 6–7).

256

Luisa Musso

Fig. 10.3–4: Kadıoğlu Köyü, villa, apsidal room, details (after Mercan and Göçen 2012: figs 7b–c). village Kadıoğlu90 a series of rooms decorated with mosaic pavements, one of them illustrating Lycurgus killing the nymph Ambrosia.91 Another room is an apsidal salon, presumably used as a dining and reception hall, of a type familiar from the residential buildings of Asia Minor. The square pavement of the central space (7.3m × 7.3m) is occupied by a figured composition, while the pavement of the apse, which might have contained a stibadium, is decorated with a chequerboard pattern with the effect of a polychrome guilloche mat (Fig. 10.1–2).92 The central square is surrounded by multiple frames. The innermost border consists of a broad frieze containing a luxuriant stylised acanthus scroll which on each side forms a pair of facing volutes, encircling, in alternation, hunting Erotes and wild animals.93 In the corners the acanthus scrolls converge on four large and intensely expressive heads with human features but with the ears of a bull: the faces are broad and fleshy with wide-open eyes under knitted eyebrows, straight sharply contoured noses and pronounced chins; a kalathos is placed over each head.94 The most characteristic feature of 90 91

District of Çaycuma, province of Zonguldak. Mercan and Göçen 2012: fig. 1; see Bruneau and Vatin 1966; Farnoux 1992: 313–5, 317–8; Grassigli 1995. 92 DGMR I: pl. 140e. 93 A major contribution to this subject remains Toynbee and Ward Perkins 1950. 94 An attribute of Ge in the Apamea mosaic: Balty 1973: 340–3, pl. 18.2. Kalathoi adorn the leafy heads along the acanthus-scrolled frieze of the Bellerophon mosaic at Palmyra: Gawlikowski and Żuchowska 2010: 11, 16, 27–8, figs 1–2, 32a–g, mid third century; Balty 2011a: 81, fig. 4d– e. The motif recurs in the mid third-century mosaic of Shahba-Philippopolis in a composition

Dynamics of mosaic design in northern Anatolia

257

Fig. 10.5: Kadıoğlu Köyü, villa, apsidal room, details (after Mercan and Göçen 2012: fig. 10). these heads is the way they morph into plants, as is apparent by the phytomorphic rendering of their beards and the curls of their heads. Leafy heads of this kind are associated with the themes of vegetation and fertility: they allude to the cyclical and perennial regeneration of nature, and therefore are to be associated with the Dionysiac sphere.95 The acanthus frieze is bordered on the outside by a sequence of 20 rectangular compartments (80cm × 60cm), linked together by a guilloche: they contain scenes from the amphitheatre consisting, in rough alternation, of combats between huntsmen and animals (tigers, leopards and bears: Fig. 10.3–4) and combats between wild beasts alone (a feline in the act of attacking a herbivore; a tiger battling against a leopard), with the exception of the four corner compartments that each display a foliate-skirted creature,96 a motif familiar in the repertoire of grotesques.97 The whole composition is enclosed on the outside by a wave pattern with developed volutes. This is interrupted on the entrance side, which is decorated with a fishing scene projected into a enchanted marine environment: the cobalt sea is thronged with large fish, aquatic birds alight and three boats rock on the surface (Fig. 10.5). 94

similar to that of the Kadıoğlu Köyü mosaic: Balty 1977: 24–5, no. 7; 2011a: 77, 80–1, 85, figs 3c, 7c. 95 Toynbee and Ward-Perkins 1950: 4–6; Jucker 1961: 164–208. Cf. the two female busts emerging from the acanthus-scroll border of the triclinium mosaic in the House of Dionysos at Sepphoris in central Galilee: Talgam and Weiss 2004: 88–94, pls 12–4. Balty 2011a: 81–3 traces the origin of the motif to the heads of Oceanus and to gorgoneia, but this interpretation is undermined by the pronounced phytomorphic component and the lack of specific attributes. 96 The figures are similar to the ‘putti’ morphing into plants in the luxuriant acanthus volutes of some Syrian mosaics (Balty 2011a: 75, fig. 2a–c); in particular the mosaic found at Homs, first half of the third century, in which hunting Erotes alternate with beasts: Abdallah 2011: 11, figs 2–5. 97 Among the many parallels, cf. the friezes with grotesques in the frigidarium of the North Baths at Acholla, Tunis, Bardo Museum, early second century: Picard 1980: 81, 85, figs 8, 13. The motif was widespread in the Vesuvian area.

258

Luisa Musso

Two of the boats are manned by Erotes (one steers the boat while two others are intent on fishing); the boat at the centre is filled with Psychai. The type of centralised composition, characterised by two contiguous figured borders – on the inside, the frieze with the motif of the ‘inhabited scroll’; on the outside, the succession of figured panels – merits closer attention. A good parallel is offered by a mosaic found in Rome, outside Porta Portese, in 1823 (Fig. 11):98 the (lost) central square was surrounded by a wide border frieze populated by huntsmen and wild animals, and dominated at the corners by large heads with phytomorphic beards and hair.99 This in turn was framed on the outside by a border with meander in perspective alternating with rectangular panels including still-lifes of animals destined for the table and compartments with busts of the Seasons. This compositional system is decidedly unusual in the urban context; it seems to derive from an eastern model. A similar syntax is found in the pavement of the Constantinian Villa at Daphne;100 the association of acanthus scrolls and external frieze enclosing hunting scenes recurs in a mosaic from Syedra in the Alanya Museum.101 The use of scenes of combat in the amphitheatre – mainly venationes, but also gladiatorial combats – articulated in continuous friezes round the borders of the main composition or along the sides of a peristyle,102 is very familiar in Italy,103 in Africa Proconsularis104 and in the eastern Mediterranean, from Kos105 to Orthosia (Meander valley)106 and Nea Paphos.107 The formula of the composition, however, is altered in the example at Kadıoğlu Köyü: here the continuous narration is interrupted, the scenes are isolated in a series of separate rectangular panels.108 The seascape populated by Erotes, in the guise of fishermen or boatmen, is widespread in various genres, whether in murals, mosaics or reliefs.109 Psychai can take part in these marine activities, mainly as musical accompaniers in a context with a strong nilotic connotation. A unique version, which shows them naked and swimming, transformed into the chariot of Eros, is that of the mosaic in the House of the Boat of Psyches at Antioch.110 98 99 100 101 102 103 104 105

106 107 108

109 110

Visconti 1825: 659–62, pl. 2: ‘operato a colori con meraviglioso artifizio’; Werner 1994: 1.90–2, K 36, with hypothetical dating between the late first and early second century. Michaelis 1882: 656–77, no. 136: ‘Tritons’; Werner 1994: ‘Faunus’. Stillwell 1938: 197–200, no. 87, pls 62–8; Levi 1947: 1.226–8, fig. 85, 2. pls 52–3; Baratte 1978: 99–106, no. 45, figs 94–114. Tülek 2011a: 641–3, 645, figs 5–6, with hunting Erotes, tentatively dated to the early third century. Kondoleon 1995, 273–90; Werner 1998: 46–7. For example Rome, Vatican Museums, from S. Sabina, late second century: Werner 1998: 43–54. Tripoli, National Museum, from the villa at Zliten: Aurigemma 1926: 131–201. Mosaic of the Judgement of Paris, late second/early third century: De Matteis 2004: 38–42, 48– 50, no. 1, 201–2, pls 2, 8–13 and B; Orpheus mosaic, second half of the third century: De Matteis 2004: 145–7, no. 70, pl. 86.1–2. Scheibelreiter-Gail 2011: 327–30, Kat. 100, 636, fig. 441, 638, figs 453–4. House of Dionysos, peristyle: Kondoleon 1995: 271–314. A sequence of panels with scenes of venatio is shown in the reconstruction of the mosaic pavement of the so-called ‘throne room’, palace of Gamzigrad (Romuliana), first third of the fourth century: Cvetković Tomašević 1994: 148, pl. 71.1. Blanc and Gury 1986: 1002–8, 1046. Levi 1947: 1.176–8, 2. pl. 37b.

Dynamics of mosaic design in northern Anatolia

259

Fig. 11: Drawing of the mosaic found in Rome, outside Porta Portese, in 1823 (after Visconti 1825: pl. 2). A prominent role in the overall iconography of the Kadıoğlu Köyü mosaic is that of the opulent ‘inhabited acanthus scroll’,111 enhanced by the leafy heads.112 The closest parallels

111 112

Levi 1947: 1.490–508; Balty 1977: 56, nos 20–3; 2011a; Dauphin 1987; Ovadiah 2001: 6. Balty 2011a: 77–81, figs 3–5.

260

Luisa Musso

are found in mosaics produced in Antioch,113 Apamea,114 Mariamin (Hama)115 and Homs (Emesa),116 or in versions influenced by craftsmen active in the cities of the Orontes valley: from Cyrrhus117 to Zeugma on the Euphrates,118 and from Edessa in the Osrhoene119 to Shahba-Philippopolis in the province of Arabia120 and, further beyond, in Palestine.121 Mainly set against a black background,122 and dating from the third to the fifth century, these acanthus-scroll borders are diversified in terms of craftsmanship but similar in subject, compositional syntax, taste for exuberant foliations and use of chiaroscuro. The motif of the kalathos as the crowning ornament of the heads in the corners is attested in the luxurious border of the mosaic of the Triumph of Dionysos at Sétif (Algeria), for which relations with eastern Greek workshops have already been conjectured.123 Nor can we fail to recall the border of the Great Palace Mosaic in Constantinople, whose chronology124 remains controversial.125 Potential parallels with compositions in vogue in the Syrian area after the mid fifth century seem to hold good.126 They are corroborated by kinship both with the Great Hunt mosaic from Apamea, Triclinos House,127 and with Antiochene examples of ‘inhabited scrolls’, by affinities in composition (‘figure carpet’) and in technique (fan-shaped arrangement of the background tessellation), and by iconographic consonances with the stunning mosaic pavements of the palatial residence of Şanlıurfa (Edessa), assigned to the age of Justinian:128 here the essential foreignness to the local mosaic tradition leaves little doubt about the assignment of the work to itinerant craftsmen, whose 113

114 115 116 117 118

119 120 121 122 123 124 125 126

127

128

See the two male heads in the border with grape-harvesting Erotes in the Constantinian Villa at Daphne, lower level: Stillwell 1938: 200, no. 88, section 1, 6, pl. 69; Levi 1947: 1.226, 508, 2. pl. 143b–e. The so-called mosaic of Meleager and Atalanta, border with hunting Erotes and masks in the corners, last quarter of the fifth century: Balty 1977: 118–23, nos 54–6; 2011a: 77, fig. 5. Hama Museum, mosaic with concert scene: Zaqzuq and Duchesne-Guillemin 1970: 93–104, pls 1–7; Balty 1977: 100–1, no. 45; 2011a: 77, fig. 3a–b; 2011b, late fourth/early fifth century. See above, n. 96. Abdul-Massih 2009: 298–303, 305; Balty 2011a: 77, 80, fig. 3e, third century. Mosaic of the provinces: Kriseleit 2000: 45–51, no. 13, first half of the third century; House of Zosimos, Synaristosai Mosaic: Abadie-Reynal and Darmon 2003: 88–95, figs 10, 12–6, 18–9, 23; Önal 2009: 48–9, late second/early third century. Karabulut et al. 2011: 56–7, figs 60–1, 65–8, early sixth century. Souweida National Museum, mid third century: see n. 94. The mosaic tradition of the inhabited acanthus scroll continued in the sixth century in religious buildings: Piccirillo 1993: 96–107. Nablus (Samaria), third century: Dauphin 1979; Ovadiah 1987: 129–30, no. 217, pls 149–53, 190–2; Sevilla Sadeh 2005–2006: 438–44. Legacy of opus vermiculatum: Balty 1989: 504. Blanchard-Lemée 2005: 295, fig. 4a, 298–9, fig. 8a, 300–1; Buccino 2013: 212–3, M26. Jobst and Vetters 1992; Jobst et al. 1999; Jobst 2005: 1089–92, 1099, first half of the sixth century. Bardill 2004: 1.134–47; 2006: 12–20, late sixth/first decades of the seventh century. Hellenkemper Salies 1987: 297–308; 1990b: 619–22, no. 5, fig. 121; 1994; 1999: 686. Constantinople seems to have been long affected by the influence of Antiochene workshops: Hellenkemper Salies 1990b: 614–8, nos 1–3; Parrish 2008: 965–6, 969, figs 4–6. Bruxelles, Musée Royaux d’Arts et d’Histoire: Balty 1969; 1977: 104–9, nos 47–9; 1986a: 12–4, figs 26–34; 1995: 23–4, 50, 78–9, 173, pl. 22; 1997: 90–1, fig. 7. The proposed dating to the fifth century conflicts with the date ad annum (539) reported in the mosaic inscription placed in the threshold. Karabulut et al. 2011.

Dynamics of mosaic design in northern Anatolia

261

qualitative level leads us to ascribe it to Antioch,129 or at any rate to the region that had its cultural focal point in the northern Orontes valley.130 The selection of motifs in the Kadıoğlu Köyü mosaic – leafy masks and hunting Erotes, scenes of venatio and fishing – was aimed at emphasising the forces of nature: a savage nature that had to be tamed, through hunting, by Erotes and by man, but at the same time an opulent nature, fruitful in its exuberant display of plant tendrils and foliate masks, peaceful, prosperous and happy in the harmony of sea and sky. The inclusion of Erotes and Psychai is a demonstration of this. A date for the mosaic in the second half of the third century, if not to the early fourth century, seems probable.

5. Pompeiopolis The mosaics uncovered at Pompeiopolis in Paphlagonia, the number of which is constantly increasing thanks to the systematic exploration of the site since 2006,131 enable us to trace the development of local ateliers of mosaic pavements in a period stretching roughly from the mid third to the fifth century. Some of the mosaic finds predate the recent excavation campaigns. Sporadic exploration of Zımbıllı Tepe led to the discovery of a mosaic of considerable iconographic significance. It was partially exposed and then reinterred, and is documented by archival photos only (Fig. 12).132 A polychrome geometric field with continuous decoration133 surrounds a large central square enclosing a circular band containing images of the twelve signs of the Zodiac in separate compartments. A bust of an old man with flowing white locks and beard is portrayed within the circle; female busts of the Seasons are inserted in the triangles between the external margin of the circle and the square. The combination of signs of the Zodiac and Seasons suggests an identification of the bust as the personification of Time in its solar dimension. The zodiacal ellipse, which materialises the solar ecliptic, forms part of the iconography both of Time, often represented in the act of controlling the passage of time, and of the Sun, which may in fact appear within the zodiacal circle, as exemplified for 129

130

131 132 133

The Antiochene origin of the mosaicists is suggested by the emphasis given to the cycle of Achilles, a favourite theme in Antiochene figurative culture, the fashion for personifications (Ktisis) and the imposing figures of the tiger and lion, which recalls the image and layout of the famous Beribboned Lion: Levi 1947: 1.313–5, 2. pl. 70c. The mosaics of Edessa enable us to define better the radius of action of the mosaicists trained in the major centres of northwestern Syria and to estimate what might have been the level of coeval Antiochene production, of which precious little of any note has been preserved for the period after the mid fifth century. The manifest technical, compositional and iconographic affinity with the Great Palace Mosaic in Constantinople (despite its peculiar stylistic character) leaves the imperial complex less ‘isolated’ than it has hitherto been reputed to have been. We cannot but wonder whether the presence of top-level craftsmen of ‘Antiochene’ training, active in Constantinople as in Edessa in the age of Justinian, might not be the result of the contraction of patronage and investments in Antioch itself, following the series of major disasters that struck the city from the second quarter of the sixth century onwards: Foss 1997: 190–5, 259–60. Summerer 2011; Summerer and Kienlin 2013. I am indebted to the generosity of Lâtife Summerer for familiarising me with this mosaic. DGMR I: pl. 186b–c. The motif is attested at Antioch, House of the Boat of Psyches: Stillwell 1938: 184, no. 46, Panel F, pl. 34; it recurs in the composition with Lycurgus and Ambrosia at Kadıoğlu Köyü.

262

Luisa Musso

Fig. 12: Pompeiopolis, Zimbıllı Tepe: mosaic illustrating the personification of eternal Time (©Kastamonu Museum). instance by the fine mosaic recently discovered at Bursa in the Tophane quarter. The Pompeiopolis mosaic may help to eliminate a spurious distinction which has enjoyed some favour in iconographic studies of the image of Time: a distinction predicated on the relative age of the personification. On this basis, two distinct iconographical versions have been elaborated: one that portrays Time in the guise of an old man (Aion), the other as a youth, almost invariably accompanied by the Seasons, endowed with the zodiacal ellipse and with attributes recalling the fertility of the earth (Genius of the Year). The Pompeiopolis mosaic shows, on the contrary, that there is no distinction between an annual time puer and an eternal time senex, one with the zodiacal ellipse, the other without it. Time is, by its very nature, and as the ancient sources repeatedly attest, at once young and old: the personification corresponds not to a single calendar year, but to the ‘cosmic year’ that eternally begins afresh and is eternally renewed in tandem with the passage of the sun.134 Another figured mosaic with a geometric scheme and remarkable for its composition and brilliant polychromy decorated the frigidarium of a small bathhouse complex, presumably belonging to a private residence, situated on the southeastern slopes of Zımbıllı Tepe (Fig. 13).135 The mosaic was excavated in 1971; in 1984 what remained of its geometric composition136 was detached and transferred to the Kastamonu Museum.137 The central panel comprised a Nereid, nude but for her bracelets and anklets, transported by a Triton on the surface of the sea. The main field of the mosaic, surrounded by a border with 134 135 136 137

Musso 2008: 167–9. Musso et al. 2011: 75–6, 100–1, pls 2–3. The figured panel had by this time already been lost. Thanks to the good offices of the Director of the Kastamonu Museum, Nimet Bal, it has been possible to document the panels, and this, in turn, has enabled A. Hartmann to produce a virtual restoration of the original composition.

Dynamics of mosaic design in northern Anatolia

263

Fig. 13: Pompeiopolis, private bathhouse complex: reconstruction of the frigidarium mosaic (A. Hartmann, ©Pompeiopolis Project). perspective meander and compartments containing geometric motifs, was subdivided into four portions by a guilloche and filled with patterns of overlapping scales alternated with perspective cubes. The simplified figural design, the schematic representation of the waves,

264

Luisa Musso

Fig. 14: Pompeiopolis, domus on the slopes of Zimbıllı Tepe, room A (©Kastamonu Museum). the use of perspective meander with interposed panels,138 the rendering of the chevron motif with rows of tesserae in zigzagged colours, not yet dissolved in the tonal range typical of the ‘rainbow style’, all suggest that the mosaic dates from the first half of the fourth century.139 The change of taste registered during the second half of the century, and manifested in the preference for purely geometric compositions, makes a later dating improbable. Excavations conducted by the Kastamonu Museum in 1984140 brought to light part of a residential complex situated close to the bathhouse: only the recent resumption of the excavations is beginning to reveal its extent and articulation.141 It is a domus of considerable size, built on a series of terraces accommodated to the hillside; its various quarters are distributed around a huge peristyle. A spacious reception room in the lower part of the complex, accessible from two entrances, was adorned with a mosaic pavement with a geometric pattern (Fig. 14). It was dismembered and placed in the storerooms of the Kastamonu Museum. The composition was articulated in two ‘carpets’ of different size: the rectangular sector close to the two entrances included a pattern of intersecting circles, while the main field contained a swastika meander pattern surrounded by a sequence of nonadjacent rectangular and square panels, characterised by the sobriety of their geometric 138 139 140 141

Antioch, Bath D: Campbell 1988: 16–7, IVA5h, first quarter of the fourth century. This date is to be preferred to the earlier one I proposed previously: Musso et al. 2011: 76. Yaman 1991. For an interim report, see Musso et al. 2011.

Dynamics of mosaic design in northern Anatolia

265

Fig. 15: Pompeiopolis, domus on the slopes of Zımbıllı Tepe, room A: detail of mosaic inscription (R. Hessing, ©Pompeiopolis Project). ornamentation. A reconstruction of the pavement is hampered by the barely legible photographic documentation accompished at the time of discovery, and by the large gaps in the mosaic texture. A tabula ansata containing a greeting was placed in front of one of the entrances (Fig. 15). Based on a standard formula for good-omen inscriptions – ΕΙCΕΛΘΕ ΕΠΑΓΑΘΩ (‘Enter for your well-being’)142 – it belongs to the sizeable category of pavement mosaic inscriptions with formulaic expressions of welcome, salutation and good wishes.143 A date in the first half of the fourth century is suggested both by the type of border formed by a succession of panels144 and by the absence of the ‘rainbow style’, which came to the fore in the second half of the century. The pièce d’excellence of the domus is a mosaic uncovered in another room situated close to the one containing the inscription (Figs 16–7). The central composition is surrounded by a double frame: on the outside an ivy-leaf tendril and on the inside a twisted ribbon. A female bust in frontal pose, with head adorned with a diadem and surrounded by a nimbus, sits in the central panel, enclosed within a double square, the outer one skewed diagonally by 90°. She holds a fruit with two fingers; a branch rests on her right shoulder. The busts of the Four Seasons are placed in the outer angles formed by the poised square. The fruitbearing connotation of the personification and her connection with the Seasons call to mind the personification of Ananeosis (reparatio/renovatio), linked to the cyclical renewal of the year.145 A coin of Arcadius, struck by the mint of Constantinople (388–92), was found in contact with the mosaic pavement. The compositional system of the mosaic, characterised by the extension of the geometric design over a disproportionately large area, and 142

Brixhe 1992: no. 499. The well-attested formula ΕΠΛΓΑΘΩ is inscribed in pebbles in a mosaic of Nea Paphos dating to the Hellenistic period: Michaelides 1992: 2; it appears at Cyrene, Insula of Jason Magnus, set in a tabula ansata: Venturini 2013: 64, no. 32, pls 33–4, second century; see also 84–5, no. 103, pls 101a–b, 103. The combination with a verb inviting the visitor to cross the threshold is found in the co-called Eustolios Complex at Kourion, early fifth century (ΕΙCΑ[ΓΕ] / ΕΠΑΓΑΘ[Ω] / ΕΥΤΥΧΩC / ΤΩ ΟΙΚ[Ω], ‘Enter to thy good luck and may thy coming bless this house’): Mitford 1971: 352–3, no. 201. The formula ΚΑΛΩC ΗΚΕΙC is found in the threshold of a house under the theatre of Kom el-Dikka, Alexandria, within a tabula ansata: Majcherek 1993: 16, fig. 1, 18–9. 143 Scheibelreiter-Gail 2012: 148, n. 67. 144 Similar borders are found in Antiochene mosaics datable to the third century: Levi 1947: 1.163–4, fig. 62, 2. pl. 102a–e; 1.218–9, 2. pls 48d, 106c; 1.196, fig. 73, 197. See also the mosaic of Lycurgus and Ambrosia at Kadıoğlu Köyü. 145 Levi 1947: 1.320–1, 626, 2. pl. 73b: the bust of Ananeosis is surrounded by a garland of leaves and fruits enriched with miniature busts of the Four Seasons. The personification recurs in a mosaic in Daphne: see n. 147.

266

Luisa Musso

Fig. 16: Pompeiopolis, domus on the slopes of Zımbıllı Tepe, mosaic pavement of room E (M. Brizzi, ©Pompeiopolis Project). the subordination to it of the figural subject, combined with its peculiarities of form and its iconographic features – such as elaborate jewellery – are consonant with figured mosaics of the period of Theodosius II (408–50). The taste, displayed in profane contexts (prevalently but not exclusively domestic), for personifications of abstract ideas (Ktisis, Ananeosis, Tryphe being among the most widespread) and of concepts that express virtues or moral qualities (e.g. Megalopsychia), was developed in the fourth century and reached its peak in the fifth.146 The list of such 146

Downey 1938; Levi 1947: 1.253–6; Campbell 1994; López Monteagudo 1997; Huskinson 2003: 144–5; Leader-Newby 2005.

Dynamics of mosaic design in northern Anatolia

267

Fig. 17: Pompeiopolis, domus on the slopes of Zimbıllı Tepe, mosaic pavement of room E: detail of the central panel (Β. Maerzke, ©Pompeiopolis Project). personifications is long: a great many are found in mosaics covering an extensive area of the eastern Mediterranean with its main centre at Antioch.147 One recently published example, found at Sinop in the Kefevi district, shows the personification of Tryphe in its connotation of opulence and luxury.148 These are, in any case, figures whose connotations were easily adaptable, with minimal changes, to the most varied allegorical abstractions. Sometimes no need was felt to ‘label’ them at all, in conformity with a process that tended to transform them into a vaguer and more adaptable cultural, symbolic and allegorising repertoire. These images would not only have functioned as allegorical expressions of good wishes, but would, during receptions and symposia, have offered cues for the display of erudition or allegorical disquisitions on literary and philosophical topics (reception of the classical paideia). In this way they became an integral part of the intellectualised entertainments that characterised so much of Late Antique society. Such personifications are also a celebration of the master of the house and, consequently, of the house itself: themes such as liberality, prosperity, fecundity, renewal, pleasure and luxury, sublimated in abstract concepts, must

147

More recent discoveries include the personification of [ΑΝΑ]ΝΕ/ΩCΙC from Daphne (Çelk 2008: 30, 32–3, fig. 13), an anonymous personification (Ge?) at Zeugma, House of the Euphrates (Önal 2009: 70–1), and that of Ktisis in the palatial complex at Şanlıurfa (Karabulut et al. 2011: 47–79, figs 53–4, 140, fig. 28). 148 Tülek 2011b.

268

Luisa Musso

Fig. 18: Hadrianopolis, early Byzantine Church A, nave: detail of the mosaic pavement (after PatacI et al. 2012: fig. 4, photographic reproduction by R. Hessing). have complied with the patrons’ mentality.149 In this context we can also place a mosaic epigraph in verses installed at the entrance of the southern rooms of the so-called Eustolios Complex at Kourion, in which Αἰδώς (Reverence), Σωφροσύνη (Temperance) and a third personification, whose name has been lost, tend the ἐξέδρα and ‘the fragrant hall’.150

6. Hadrianopolis For the last group of mosaics we remain within the confines of Paphlagonia. In the early fifth century, Hadrianopolis, situated about 3km west of modern Eskipazar, to the south of Karabük, formed part of the province of Honorias.151 Field surveys and excavation campaigns conducted here in recent years have led to the rediscovery of what must have been a flourishing city, as well as the episcopal see suffragan of Claudiopolis, during the late Roman and early Byzantine period.152 This is testified by the number of important buildings erected between the fifth and sixth century; these include, to cite just some of the main ones: two early Byzantine churches of basilica plan, two bathhouses and a secular building (‘Late Roman Villa’).153 Mosaic decoration was not the exclusive prerogative of Christian churches, but it is to these that we will confine our attention.154

149

150 151 152 153 154

Ὑγία (Health), Ζωή (Life), Χαρά (Joy), Εἰρήνη (Peace), Εὐθυμία (Cheerfulness) and Ἐλπίς (Hope) appear within the laurel wreath of the so-called House of Charidemos in Halikarnassos, fifth century: Isager 1997: 27, fig. 2; Scheibelreiter-Gail 2012: 145–6. Mitford 1971: 354–5, no. 203. Marek 1993: 116–25; Belke 1996: 155–7, s.v. Adrianupolis. A team headed by Ergün Laflı (Dokuz-Eylül-University Izmir) was active at the site between 2006 and 2010. LaflI and Zäh 2008; 2009; LaflI and Christof 2012: 1–13. LaflI and Zäh 2009: 656; PatacI 2011; PatacI et al. 2012; Kilavuz and Çelkbaş 2013.

Dynamics of mosaic design in northern Anatolia

269

Fig. 19: Hadrianopolis, early Byzantine Church A, north aisle: mosaic (after PatacI et al. 2012: fig. 2). The so-called early Byzantine Church A is located some 2.5km east of the city walls. The major composition of the nave’s pavement is a grid pattern of adjacent circles and cushions,155 the latter containing birds, whether isolated or in pairs, overlapped by square compartments containing quadrupeds and a griffin (Fig. 18). Both pavements of the aisles present a pattern based on the motif of the eight-pointed star. A variegated assortment of birds is placed in the compartments of the mosaic in the north aisle (Fig. 19), while quadrupeds and a peacock are inserted in those of the south aisle. The use of birds and animals bears witness to the richness and diversity of creation, and does not demand a merely decorative or an allegorical interpretation.156 In the sanctuary interlaced guilloches and ribbons form a pattern of tangent spindles and circles, which isolate compartments containing vases and birds. A mosaic inscription is placed in front of the bema.157 Geometric patterns, a faunal repertoire and graphic figural design recur in very similar form in the basilica complex of Hagios Stephanos at Kephalos on the island of Kos.158 The decoration of the church at Hadrianopolis testifies to the evolutionary process that the geometric pattern underwent after the mid fifth century, with the introduction of isolated figures, 155 156 157 158

DGMR I: pl. 150f (variant). Donceel-Voûte 1988: 476–8; Hellenkemper Salies 1991: 325. LaflI and Christof 2012: 42–3, nos 8–9. De Matteis 2013: 43–4, no. 26, pl. 29.1; 50, no. 41, pl. 37.2, second half of fifth/first half of sixth century.

270

Luisa Musso

Fig. 20: Hadrianopolis, Church B: partial view of the eastern part of the building (photo by L. Summerer). mainly of animals, accompanied by some scattered plant motifs in rows of compartments. A date in the early sixth century, close to that of the major basilica and baptistery of Hagios Stephanos, seems convincing.159 The mosaic pavements laid in the nave and two aisles of Church B, situated in the centre of the town, can also be dated to the first half of the sixth century (Fig. 20). The apse mosaic contains a pair of peacocks, facing each other on either side of a large labrum, against an arabesque backdrop of vine tendrils and with birds placed in the rinceaux thus formed.160 Below is a mutilated quadruped, one of a symmetrical pair (the second has been lost). The diversified geometric mosaics that decorate the pavements of the nave and aisles161 are characterised by chromatic unity, based on the use of a simple colour scheme, and compositional affinities which stem from the simplicity and rigidity of the patterns. The scheme of the north aisle, consisting of an orthogonal pattern of circles in shaded bands interlooped

159 160

LaflI and Zäh 2008: 689, 691 nn. 27, 34; 2009: 656. Confronted peacocks recur at the entrance to the zone reserved for the clergy; the motif is widespread from the early fifth century: Donceel-Voûte 1988: 16–9, 564, s.v. vase au départ d’une vigne habitée. Cf. the mosaic in the naos of the church of St John the Baptist at Oum Hartaïne, ca 500: Donceel-Voûte 1988: 198–9, fig., 189, pl. h.-t. 11. 161 This diversification, together with the strict organisation of the compositions and the simplicity of the colour combination, supports a date in the first decades of the sixth century: DonceelVoûte 1995: 88–93, 99.

Dynamics of mosaic design in northern Anatolia

271

tangentially162 (Fig. 21), is well documented by examples dating to the late fifth/early sixth century at Antioch,163 in Syria164 and on Cyprus.165 An identical scheme, distinguished by a small knotted circle inscribed in the space between the compartments, can be found in an annex of the church of Jiyé to the north of Sidon.166 An Antiochene source is further revealed by the sumptuous decorative motif of the three-dimensionally folded ribbon that winds along the border of the geometric mosaic of the nave, with cuboids in perspective placed in the interspaces (Fig. 20).167 The figurative motifs of Church B are of great interest: personifications of the Four Rivers of Paradise are arranged in a row in the oblong panel that flanks the eastern margin of the pavement of the nave. Their torsos emerge from the waters of the rivers they each personify, while their heads are turned toward the central axis of the composition. Each is identified by his name spelt out in large, carefully written Greek letters:168 ΓΕ/ΩΝ (Geon) has grey hair and a grey beard (Fig. 22); ΦΙ/CΩΝ (Phison) is a youth with a chubby face; ΤΙΓΡ/ΙC (Tigris) is bearded but still in the flower of youth; while the fourth, ΕΥΦΡΑ/ΤΗC (Euphrates), is youthful. They all wear a kind of Phrygian cap, perhaps to denote them as ‘orientals’.169 Their riverine nature is alluded to by the happy motif of the cornucopia overflowing with water that each of them holds.170 The fish that flit amid the dense patterns of waves that ruffle the surfaces of the rivers and the water spouting in abundance from the cornucopias help to transmit the enchantment of the Garden of Eden. The Four Rivers symbolise the regenerating waters of baptism, but, in this context, the composition seems more like an excerptum of a vision of paradise, linked to the concept of the order of the elements of the natural world. Connate with this vision is an isolated image placed in a panel in the southern aisle: that of a docile white bull slaking its thirst at a river or spring with its own image reflected in the water (Fig. 23).171 The source of both images is, indeed, to be found in visions of the biblical Paradise,172 allusive of the peace and beauty of the divine creation, in markedly symbolic contexts; such ‘paradise mosaics’ are well documented 162 163 164 165 166 167

168 169 170

171

172

DGMR I: pl. 235a (variant). The filling motif consists of rosettes of four adjacent bilobed and shaded petals. House of the Phoenix: Levi 1947: 1.351, 442, 480, 2. pls 83b, 135. Donceel-Voûte 1988: 558. Michaelides 1988: 140, 146–8, figs 62–4. Donceel-Voûte 1988: 358, pl. h.-t. 14, 379, 459, with dating to the 570s. DGMR I: pl. 66a. Antioch, House of the Bird-Rinceau, upper level, first half of the sixth century: Stillwell 1938: 186–7, no. 55, pl. 41; Levi 1947: 1.366, 2. pls 91, 92.2, 181d; Kondoleon 2000: 208–10, no. 96. LaflI and Christof 2012: 45–8, nos 12–5. Cf. the personification of the River Jordan in the Chludov Psalter, ninth century: Diehl 1933: pl. 72.2. The association of the water-gushing cornucopia with river gods is well known; see Daphne, House of Menander: Levi 1947: 1.205, 2. pl. 46c; Tegea (Arcadia), basilica of Thyrsos: Maguire 1987a: 24–8, figs 15–21; Hachlili 2009: 183. See the image of the bull in a panel of the mosaic at Qasr el-Lebia (Cyrenaica), of Justinianic date: Alföldi-Rosenbaum and Ward-Perkins 1980: 127, E5, pl. 13.4; and, especially, that in the Carrand Diptych in Florence, Museo Nazionale del Bargello, portrayed at the source of the Four Rivers of Paradise, late fourth/early fifth century: Maguire 1987b: 365–6, fig. 2. Maguire 1987b: 363–5.

272

Luisa Musso

Fig. 21: Hadrianopolis, Church B, north aisle: detail of the pavement mosaic (photo by L. Summerer).

Fig. 22: Hadrianopolis, Church B, nave: detail of the pavement mosaic (photo by L. Summerer).

Fig. 23: Hadrianopolis, Church B, south aisle: figured panel of the pavement mosaic (after PatacI et al. 2012: fig. 7).

Dynamics of mosaic design in northern Anatolia

273

in Syria,173 Cilicia and Commagene from the fifth century AD.174 The version shown here of the personified Rivers of Paradise,175 their busts emerging from the waves,176 is original and imaginative: it reveals a breach with the corresponding iconography of the classical tradition, in which rivers are portrayed semi-recumbent and accompanied by the attributes of an upturned urn, a marsh reed and a horn of plenty with fruits spilling from it.177 Here, the posture of the right arm bent at the elbow and, especially, the palm of the hand open and turned towards the spectator, in a gesture that expresses wonder or marvel, recurs in the personifications of creatures that inhabit the depth of the sea and whose torsos similarly emerge from the surface of the water:178 for example, the personification of ΠΟΝ/ΤΟC (Pontos) (Fig. 24) in the panel representing the ‘Judgement’ of the Nereids in the mosaic of the House of Aion at Nea Paphos, fourth century,179 or that of ΘΑΛΑΣΣΑ (Thalassa) in the mosaic of the Church of the Apostles at Madaba, AD 578,180 both accompanied by their respective legends. In contrast to the undoubted originality of the iconography, the style of the mosaic is poor: a circumstance that reminds Ergün LaflI and Alexander Zäh of the mosaic pavements of Osrhoene,181 though that observation is insufficient to corroborate the presence at Hadrianopolis of itinerant craftsmen from that region.

7. Conclusion Much progress has been made in terms of our knowledge of mosaics in the Black Sea region since Sheila Campbell stated in an article published over 30 years ago in the American Journal of Archaeology: ‘I know of only one mosaic found in a city on the Black Sea coast’.182 Campbell’s research approach was an ambitious one and, for the period, innovative: to overcome a purely iconographic approach by prioritising the definition of the characteristics of the workshops of mosaicists active throughout the area of present-day Turkey and by drawing attention to hitherto-ignored traditions, i.e. those other than the ‘Antiochene manner’. The question posed by Campbell, and which I repose here, is whether it is possible to maintain the existence of different local identities, of well-defined 173

174 175 176 177

178

179 180 181 182

Church of the Holy Martyrs at Tayyibat al-Imâm, north of Hama, 442: Zaqzuq 1995: 238–40, 249–50, figs 20–2; Farioli Campanati 1999: 173–4; Zaqzuq and Piccirillo 1999: 445, fig. 9; Hachlili 2009: 182–3. Hellenkemper Salies 1990a: 336–42, 344–7. Février 1956; Maguire 1987a; 1999; 2002: 25–7; Djuric 1989. Cf. the mosaic of the baptistry of the church of Mukheitem near Jabaliyah (Gaza), of Justinianic date: Humbert 1999; Hachlili 2009: 180–1, figs 8.1, 8.2c. On the iconography of rivers, see Ostrowski 1991; Huskinson 2005; Trovabene 2005: 129. Maguire 1999 (180, 200–1, pl. 4) illustrates four examples of the sixth century from Madaba and Umm al-Rasas; see also Hachlili 2009: 181–2, fig. VIII-2a–b. The detail of the hand outstretched reappears in the description of Bythos, personification of the marine abysses, in the ekphrastic poem known as Tabula Mundi, composed by John of Gaza (sixth century): 1.303–7 (ed. Friedländer 1912); Tissi 2012: 259. Compare the image of Bythos in a miniature of the Psaltery of Athos, AD 1084: Balty 1986b: 173, no. 3. Daszewski 1985: 32, pl. 10. Piccirillo 1993: 96, fig. 78, 98, fig. 80, 106. LaflI and Zäh 2008: 698, 703–4; 2009: 656; LaflI and Christof 2012: 43, no. 10, 48, fig. 15b–c. Campbell 1979: 287.

274

Luisa Musso

Fig. 24: Nea Paphos (Cyprus), House of Aion, mosaic panel representing the ‘Judgement’ of the Nereids: detail (after Daszewski 1985: pl. 10, photographic reproduction by R. Hessing). ‘regional manners’, over a relatively long time span; in other words, is it possible to maintain over the long term a community of workshop traditions and regional preferences that are primarily expressed in compositions and in a well-defined decorative repertoire and only secondarily in iconographic choices?183 Is it possible to recognise the use of a shared

183

Hellenkemper Salies 1991: 327. On the problematic definition of ‘regional school’, see Bruneau 1966: 715–9.

Dynamics of mosaic design in northern Anatolia

275

‘grammar’ and ‘syntax’ of a kind that may have given rise to a peculiar style, in rather the same way that this can be done for the variegated output that can be unified under the Antiochene/Syrian ‘cartel’? And – I would add – is it possible to define the mutual influences exerted by the various regional schools by using the stylistic affinities in the mosaic output of geographically far-separated regions as chronological criteria? The corpus of material analysed here does not furnish an unequivocal answer, but that is no reason to be content to define the Black Sea region as a ‘melting pot’. One fact does seem to emerge quite clearly: in general terms, most mosaic compositions of the Pontic area differ markedly from the repertoire of western Asia Minor, which it is now possible to evaluate in its diachronic articulation thanks to the exhaustive monograph of Veronika ScheibelreiterGail.184 In tackling the problem of Pontic mosaics, the question arises whether it is possible to identify a common thread, a fil rouge, that may link the workshops of the Pontic area with those of Constantinople, the political influence of which on the region seems to have made itself felt at least from the late fourth century, when the Patriarch of Constantinople gradually began to impose his own jurisdiction on the dioceses of Asia and Pontus. The data collected in the course of my research, far from corroborating this hypothesis, seems to me to converge and point to quite another ‘trail’, one that connects a large portion of the Pontic mosaic pavements with the eastern territories of the empire. At Samsun this is shown by a predilection for imposing mythological themes, by conformity to the Hellenistic compositional tradition185 and by the fashion (anecdotal in character) for labelling figures with their names.186 At Sinop it is shown by some of the geometric compositions.187 At Sinop and at Pompeiopolis – and also at Amasya – it is demonstrated by the vogue for personifications of abstract subjects, portrayed as busts in separate compartments. At Kadıoğlu Köyü it can be seen in the acanthus ‘inhabited-scroll’ border and in the type of composition. In Church B at Hadrianopolis some of the decorative parts and particular iconographies seem indebted to the ‘Antiochene’ or, rather, northern Syrian tradition. In this context, the signature of Orentes on the mosaic in Amisos is telling. The centrality of the Antiochene experience,188 which the current tendency is to downplay,189 and, more generally, that of northern Syria is not, therefore, limited to the southern territories of Anatolia (with Cyprus foremost) and to the eastern regions (Commagene190 with Zeugma 184 185 186 187

Scheibelreiter-Gail 2011. Balty 1995: 162–8. Leader-Newby 2005: 235–40; 2007. Hellenkemper Salies 1987 (307) emphasises the ‘north Syrian’ peculiarity of compositions such as those of the martyrium of Qausīyeh and of the synagogue of Apamea. Characteristic of this craftsmanship is the presence of ‘buttons’ in the corners of the compositions with starred octagon: the motif is attested at Sinop: see the mosaic from Sinop, supra. 188 A grave limitation in evaluating the Late Antique facies of Antioch mosaics is our almost total archaeological ignorance of the major churches of the metropolis: Mayer and Allen 2012. 189 For example Weiss 2011 (941), in relation to mosaic production at Sepphoris (Galilee). 190 The mosaic of the church of Akdeğirmen Höyük, third quarter of the fifth century, is clearly indebted to the Syrian and Syro-Cicilian school of mosaics: Candemr and Wagner 1978: 196– 7, 201–3, 205–10, fig. 1, pls 74–9; Hellenkemper Salies 1990a: 344–7, fig. 68.

276

Luisa Musso

undoubtedly, but also Osrhoene with Edessa): it is also revealed as an inescapable source of influence for the northeastern region of Anatolia191 and for the Pontic area. Luisa Musso Università Roma Tre Dipartimento di Studi Umanistici Piazza della Repubblica 10, Rome 00185, Italy [email protected]

Bibliography Abbreviations DGMR I: C. Balmelle, M. Blanchard-Lemée, J. Christophe, J.-P. Darmon, A.-M. GuimierSorbets, H. Lavagne, R. Prudhomme and H. Stern 1985. Le décor géométrique de la mosaïque romaine I. Répertoire graphique et descriptif des compositions linéaires et isotropes. Paris. DGMR II: C. Balmelle, M. Blanchard-Lemée, J.-P. Darmon, S. Gozlan and M.-P. Raynaud 2002. Le décor géométrique de la mosaïque romaine II. Répertoire graphique et descriptif des décors centrés. Paris. Mosaïque gréco-romaine IV: J.-P. Darmon and A. Rebourg (eds) 1994. La Mosaïque gréco-romaine, IV. IVe Colloque International pour l’étude de la mosaïque antique, Trèves 8–14 août 1984 (Supplément au Bulletin de l’A.I.E.M.A.). Paris. Mosaïque gréco-romaine V: P. Johnson, R. Ling and D. J. Smith (eds) 1994–1995. Fifth International Colloquium on Ancient Mosaics, Bath September 5–12 1987, 2 vols (Journal of Roman Archaeology. Supplementary series 9). Ann Arbor. Mosaïque gréco-romaine VII: M. Ennaïfer and A. Rebourg (eds) 1999. La mosaïque gréco-romaine, VII. VIIe Colloque International pour l’étude de la mosaïque antique. Actes, Tunis 3–7 octobre 1994, 2 vols. Tunis. Mosaïque gréco-romaine IX: H. Morlier, Ch. Bailly, D. Janneteau and M. Tahiri (eds) 2005. La mosaïque gréco-romaine, IX. Actes du IXe Colloque International pour l’étude de la mosaïque antique et médiévale, Rome 5–10 novembre 2001, 2 vols (Collection de l’École française de Rome 352). Rome. Mosaïque gréco-romaine X: V. H. Correia et al. (eds) 2011. O mosaico romano nos centros e nas periferias. Originalidades, influências e identidades. Actas do X Colóquio Internacional da Associação Internacional para o Estudo do Mosaico Antigo (AIEMA), Museo Monográfico de Conimbriga (Portugal), 29 de outubro a 3 de novembro de 2005. Conimbriga. Mosaïque gréco-romaine XI: M. Şahn (ed.) 2011. La mosaïque gréco-romaine XI. XI. Uluslararası Antik Mozaik Sempozyumu / XIth International Colloquium on Ancient Mosaics. Türkiye Mozaikleri ve Antik Dönemden Ortaçag Dünyasina Diger Mozaiklerle Paralel Gelisimi: Mozaiklerin Baslangicindan Geç Bizans Çagina Kadar Ikonografi, Stil ve Teknik Üzerine Sorular / Mosaics of Turkey and Parallel Developments in the Rest of the Ancient and Medieval World: Questions of Iconography, Style and Technique from the Beginnings of the Mosaic until the Late Byzantine Era, Bursa 16–20 Ekim 2009 (Uludag Üniversitesi Mozaik Arastirmalari Merkezi Yayinlari Serisi 1). Istanbul.

Literature Abadie-Reynal C. and J.-P. Darmon 2003. ‘La maison et la mosaïque des Synaristôsai (Les Femmes au déjeuner de Ménandre)’, Zeugma: Interim Report. Rescue Excavations (Packard Humanities Institute), Inscription of Antiochus I, Bronze Statue of Mars, House and Mosaic of the Synaristôsai, and 191

See the mosaic pavement of the church of Altıntepe/Erzincan, which presents in its nave a figure carpet thronged with running animals set against the background of a floral semis: Can 2009; 2011; this is dated to the mid sixth century and attributed to an itinerant workshop from Cilicia or northern Syria. The composition falls into the typically Syrian category of paysage animé libre: Donceel-Voûte 1994: 208–9.

Dynamics of mosaic design in northern Anatolia

277

Recent Work on the Roman Army at Zeugma (Journal of Roman Archaeology. Supplementary series 51). Portsmouth, 79–99. Abdallah, K. 2011. ‘Mosaïque d’Héraclès découverte à Homs (Syrie centrale)’, Mosaïque grécoromaine XI, 1–13. Abdul-Massih, J. 2009. ‘La fortification polygonale et les mosaïques d’une maison romaine à Cyrrhus (Nebi Houri)’, Syria 86, 289–306. Aitken, E. B. 2001. ‘The Cult of Achilles in Philostratus’ Heroikos: A Study in the Relation of Canon and Ritual’, in S. R. Asirvatham, C. O. Pache and J. Watrous (eds), Between Magic and Religion. Interdisciplinary Studies in Ancient Mediterranean Religion and Society. Lanham, Md., 127–35. Akkaya, M. 1993. ‘Amisos Antik Kenti Kurtarma Kazısı ve Mozaik Kaldırma Çalışmaları’, 3. Müze Kurtarma Kazıları Semineri, 27–30 Nisan 1992, Efes. Ankara, 207–18. Akkaya, M. 1998. ‘Amisos Antik Kenti Kurtarma Kazısı ve Mozaik Kaldırma Çalışmaları’, 8. Müze Kurtarma Kazıları Semineri, 7–9 Nisan 1997, Kuşadası. Ankara, 43–51. Alföldi-Rosenbaum, E. and J. B. Ward-Perkins 1980. Justinianic Mosaic Pavements in Cyrenaican Churches (Monografie di archeologia libica 14). Rome. Assimakopoulou-Atzaka, P. 2011. ‘The Iconography of the Mosaic Pavement of the Church of Tell Amarna’, in Ö. Tunca, T. Waliszewski and V. Koniordos (eds), Tell Amarna (Syrie) V. La basilique byzantine et ses mosaïques (Publications de la Mission archéologique de l’Université de Liège en Syrie). Leuven, Paris and Walpole, MA, 101–52. Aurigemma, S. 1926. I mosaici di Zliten (Africa Italiana. Collezione di monografie a cura del Ministero delle Colonie 2). Rome and Milan. Aybek, S. and A. K. Öz 2012. ‘Mosaic Researches at Amisos, 1996’, Journal of Mosaic Research 5, 13–23. Balty, J. 1969. La grande mosaïque de chasse du triclinos (Fouilles d’Apamée de Syrie. Miscellanea 2). Brussels. Balty, J. 1973. ‘Mosaïque de Gê et des Saisons à Apamée’, Syria 50, 311–47. Balty, J. 1977. Mosaïques antiques de Syrie. Brussels. Balty, J.-Ch. 1981, Guide d’Apamée. Brussels. Balty, J. 1984. ‘Les mosaïques de Syrie au Ve siècle et leur répertoire’, Byzantion 54, 437–68 (reprinted in Balty 1995, 87–109). Balty, J. 1986a. Musées Royaux d’Art et d’Histoire. Guide du visiteur. Mosaïques d’Apamée. Brussels. Balty, J. Ch. 1986b. ‘Bythos I’, Lexicon Iconographicum Mythologiae Classicae 3.1, 173–4. Balty, J. 1989. ‘La mosaïque en Syrie’, in J.-M. Dentzer and W. Orthmann (eds), Archéologie et histoire de la Syrie 2. La Syrie de l’époque achéménide à l’avènement de l’Islam (Schriften zur Vorderasiatischen Archäologie 1). Saarbrücken, 491–524 (reprinted in Balty 1995, 59–86). Balty, J. 1995. Mosaïques antiques du Proche-Orient. Chronologie, iconographie, interprétation (Annales littéraires de l’Université de Besançon 551). Paris. Balty, J. 1997. ‘Mosaïque et architecture domestique dans l’Apamée des Ve et VIe siècles’, in S. Isager, and B. Poulsen (eds), Patron and Pavements in Late Antiquity (Halicarnassian studies 2). Odense, 84–110. Balty, J. 1999. ‘Réflexions sur la chronologie: À propos d’un groupe de mosaïques de Syrie du Nord’, Mosaïque gréco-romaine VII, 141–50. Balty, J. 2001. ‘Doro Levi, Antioch Mosaic Pavements: cinquante ans après’, Byzantion 71, 303–24. Balty, J. 2011a. ‘Le rinceau d’acanthe à fond noir dans la mosaïque syrienne: l’exemple de Mariamin’, in H. Meyza and I. Zych (eds), Classica orientalia. Essays presented to Wiktor Andrzej Daszewski on his 75th birthday. Warsaw, 73–88. Balty, J. 2011b. ‘Une chronologie discutée: la mosaïque de Mariamin’, Mosaïque gréco-romaine X, 647– 55. Baratte, Fr. 1978. Catalogue des mosaïques romaines et paléochrétiennes du Musée du Louvre. Paris. Bardill, J. 2004. Brickstamps of Constantinople (Oxford monographs on classical archaeology). Oxford. Bardill, J. 2006. ‘Visualizing the Great Palace of the Byzantine Emperors at Constantinople. Archaeology, Text, and Topography’, in F. A. Bauer (ed.), Visualisierungen von Herrschaft. Frühmittelalterliche Residenzen, Gestalt und Zeremoniell. Internationales Kolloquium, Istanbul 3.–4. Juni 2004 (Byzas. Veröffentlichungen des Deutschen Archäologischen Instituts Istanbul 5). Istanbul, 5–45. Barringer, J. M. 1995. Divine escorts. Nereids in Archaic and Classical Greek Art. Ann Arbor.

278

Luisa Musso

Becker, L. and Ch. Kondoleon 2005. ‘The Atrium House Triclinium. Archaeology, Iconography, Technique and Style’, in L. Becker and Ch. Kondoleon (eds), The Arts of Antioch. Art Historical and Scientific Approaches to Roman Mosaics and a Catalogue of the Worcester Art Museum Antioch Collection. Worcester Art Museum, 17–48. Belke, K. 1996. Paphlagonien und Honōrias (Tabula Imperii Byzantini 9). Wien. Ben Abed, A., M. Demas and Th. Roby (eds) 2008. Lessons Learned. Reflecting on the Theory and Practice of Mosaic Conservation. Proceedings of the 9th ICCM Conference, Hammamet, Tunisia, November 29 – December 3, 2005 (The Getty Conservation Institute). Los Angeles. Berenson Maclean, J. K. and E. B. Aitken (eds) 2003. Flavius Philostratus. On Heroes, with a Preliminary Essay by C. Dué and G. Nagy (Writings from the Greco-Roman world 3). Leiden. Blanc, N. and Fr. Gury 1986. ‘Eros/Amor, Cupido’, Lexicon Iconographicum Mythologiae Classicae 3.1, 952–1049. Blanchard-Lemée, M. 2005. ‘Le triclinium à la mosaïque dionysiaque de Sétif (Algérie)’, Mosaïque gréco-romaine IX.1, 291–301. Brixhe, Cl. 1992. ‘Paphlagonie. Pompeiopolis’, Revue des Études Grecques 105, 521. Brouscari, E. 1997. ‘The Tyche of Cos on a mosaic from a Late Antique house in Cos’, in S. Isager, and B. Poulsen (eds), Patron and Pavements in Late Antiquity (Halicarnassian studies 2). Odense, 65–77. Bruneau, Ph. 1966. ‘Perspectives sur un domaine encore mal exploré de l’art antique. À propos des Actes du Colloque International sur la mosaïque gréco-romaine’, Revue des Études Grecques 79, 704– 26. Bruneau, Ph. and Cl. Vatin 1966. ‘Lycurgue et Ambrosia sur une nouvelle mosaïque de Délos’, Bulletin de correspondance hellénique 90, 391–427. Buccino, L. 2013. Dioniso trionfatore. Percorsi e interpretazione del mito del trionfo indiano nelle fonti e nell’iconografia antiche (Bullettino della Commissione archeologica comunale di Roma. Supplementi 21). Rome. Budde, L. 1969. Antike Mosaiken in Kilikien I. Frühchristliche Mosaiken in Misis-Mopsuhestia (Beiträge zur Kunst des christlichen Ostens 5). Recklinghausen. Bujskikh, S. 2007. ‘Der Achilleus-Kult und die griechische Kolonisation des unteren Bug-Gebiets’, in A. Bresson, A. Ivantchik and J.-L. Ferrary (eds), Une koinè pontique. Cités grecques, sociétés indigènes et empires mondiaux sur le littoral nord de la mer Noire (VIIe s. a.C. – IIIe s. p.C.) (Ausonius éditions. Mémoires 18). Bordeaux, 201–12. Campbell, S. 1979. ‘Roman Mosaic Workshops in Turkey’, American Journal of Archaeology 83, 287– 92. Campbell, S. 1988. The Corpus of Mosaic Pavements in Turkey. The Mosaics of Antioch (Subsidia mediaevalia 15). Toronto. Campbell, S. 1994. ‘Enhanced Images: The Power of Gems’, Mosaïque gréco-romaine IV, 55–9. Campbell, S. 1999. ‘A new Dionysos Mosaic at Zeugma’, Mosaïque gréco-romaine VII.2, 711–2. Can, B. 2009. ‘Erzincan Altıntepe Church with Mosaic’, Journal of Mosaic Research 3, 5–13. Can, B. 2011. ‘Technical, Stylistic, Iconographic Evaluation and Dating of Mosaics of Altıntepe Church’, Mosaïque gréco-romaine XI, 225–34. Candemr, H. and J. Wagner 1978. ‘Christliche Mosaiken in der nördlichen Euphratesia’, in S. Şahn, E. Schwertheim and J. Wagner (eds), Studien zur Religion und Kultur Kleinasiens. Festschrift für Fr. K. Dörner zum 65. Geburtstag am 28. Februar 1976 (Études préliminaires aux religions orientales dans l’Empire romain 66). Leiden, 192–231. Clarke, J. R. 1994. ‘Mosaic workshops at Pompeii and Ostia Antica’, Mosaïque gréco-romaine V.1, 89– 102. Cvetković Tomašević, G. 1994. ‘Mosaïques découvertes à Ulpiana en 1982 et un groupe de mosaïques du IVe siècle. Essai de datation et d’interprétation’, Mosaïque gréco-romaine IV, 145–50. Çelk, Ö. 2008. ‘Yılı Hatay Müzesi Mozaik Kurtarma Kazıları’, in M. Şahn (ed.), IV. Uluslararası Türkiye Mozaik Korpusu Sempozyum Bildirileri ‘Geçmişten Günümüze Mozaik Köprüsü’. The Proceedings of IV. International Mosaic Corpus of Turkey ‘The Mosaic Bridge from Past to Present’, Gaziantep 6–7 Haziran/June 2007. Bursa, 27–33. Daszewski, W. A. 1985. Dionysos der Erlöser. Griechische Mythen im spätantiken Cypern (Trierer Beiträge zur Altertumskunde 2). Mainz.

Dynamics of mosaic design in northern Anatolia

279

Dauphin, C. M. 1979. ‘A Roman Mosaic Pavement from Nablus’ Israel Exploration Journal 29, 11–33. Dauphin, C. 1987. ‘The Development of the “Inhabited Scroll” in Architectural Sculpture and Mosaic Art from Late Imperial Time to the Seventh Century A.D.’, Levant 19, 183–212. De Matteis, L. M. 2004. Mosaici di Cos. Dagli scavi delle missioni italiane e tedesche (1900–1945) (Monografie della Scuola Archeologica di Atene e delle Missioni Italiane in Oriente 17), Athens. De Matteis, L. M. 2013. Mosaici tardoantichi dell’isola di Cos. Scavi italiani 1912–1945 (BAR International Series 2515). Oxford. Diehl, Ch. 1933. La peinture byzantine. Paris. Djuric, D. 1989. ‘Ateni and the Rivers of Paradise in Byzantine Art’, Zograf 20, 22–9. Donceel-Voûte, P. 1988. Les pavements des églises byzantines de Syrie et du Liban. Décor, archéologie et liturgie (Publications d’Histoire de l’Art et d’Archéologie de l’Université Catholique de Louvain 69). Louvain-la-Neuve. Donceel-Voûte, P. 1994. ‘Le Ve siècle dans les mosaïques syriennes’, Mosaïque gréco-romaine IV, 205–10. Donceel-Voûte, P. 1995. ‘Syro-Phoenician mosaics of the 6th century’, Mosaïque gréco-romaine 1995.2, , 88–100. Donceel-Voûte, P. 2001. ‘Rapport de synthèse I: ateliers et écoles de mosaïstes’, in D. Paunier and Ch. Schmidt (eds), La mosaïque gréco-romaine, VIII. Actes du VIIIème Colloque international pour l’étude de la mosaïque antique et médiévale, Lausanne (Suisse) 6–11 octobre 1997 (Cahiers d’archéologie romande 85–86), Lausanne, 2, 514–9. Donderer, M. 1989. Die Mosaizisten der Antike und ihre wirtschaftliche Stellung. Eine Quellenstudie (Erlanger Forschungen. Reihe A, Geisteswissenschaften 48). Erlangen. Donderer, M. 2008. Die Mosaizisten der Antike, II. Epigraphische Quellen – Neufunde und Nachträge (Erlanger Forschungen. Reihe A, Geisteswissenschaften 116). Erlangen. Downey, G. 1938. ‘Personifications of Abstract Ideas in the Antioch Mosaics’, Transactions and Proceedings of the American Philological Association 69, 349–63. Ersoy, A. and S. Alatepel 2011. ‘Smyrna Agorası Mozaikli Yapı Mozaikleri’ Journal of Mosaic Research 4, 17–26. Ersoy, A. and D. Yolaçan 2012. ‘Smyrna Agorası Mozaikli Yapı Mozaik Döşemi Konservasyon Çalışmaları’ Journal of Mosaic Research 5, 71–80. Ertuğrul, Ö. and S. Atasoy 1998. ‘1996 Amisos Kazısı’ 19. Kazı Sonuçları Toplantısı 2, 523–30. Farioli Campanati, R. 1999. ‘Jerusalem and Bethlehem in the Iconography of Church Sanctuary Mosaics’, in M. Piccirillo and E. Alliata (eds), The Madaba Map. Centenary 1897–1997. Travelling through the Byzantine Umayyad Period. Proceedings of the International Conference, Amman 7– 9 April 1997 (Studium biblicum franciscanum. Collectio maior 40). Jerusalem, 172–7. Farnoux, A. 1992. ‘Lykourgos I’, Lexicon Iconographicum Mythologiae Classicae 6.1, 309–19. Février, P.-A. 1956. ‘Les quatre fleuves du Paradis’, Rivista di Archeologia Cristiana 32, 179–99. Foss, Cl. 1997. ‘Syria in Transition, A.D. 550–750: An Archaeological Approach’, Dumbarton Oaks Papers 51, 189–269. Friedländer, P. 1912. Johannes von Gaza und Paulus Silentiarius. Kunstbeschreibungen justinianischer Zeit (Sammlung Wissenschaftlicher Kommentare zu griechischen und römischen Schriftstellern), Leipzig and Berlin 1912. Gawlikowski, M. and M. Żuchowska 2010. ‘La mosaïque de Bellérophon’, in M. Gawlikowski (ed.), Studia Palmyreńskie XI. Warsaw, 9–42. Ghedini, F. 1994. ‘La fortuna del mito di Achille nella propaganda tardo repubblicana ed imperiale’, Latomus 53, 297–316. Grassigli, G. L. 1995. ‘L’iconografia di Licurgo. Iniziazione e trionfo dionisiaco’, Ostraka. Rivista di antichità 4, 229–48. Grossardt, P. 2006. Einführung, Übersetzung und Kommentar zu Heroikos von Flavius Philostrat (Schweizerische Beiträge zur Altertumswissenschaft 33). Basel. Hachlili, R. 2009. Ancient Mosaic Pavements. Themes, Issues, and Trends. Selected Studies, Leiden and Boston. Hedreen, G. 1991. ‘The Cult of Achilles in the Euxine’, Hesperia 60, 313–30. Hellenkemper Salies, G. 1987. ‘Die Datierung der Mosaiken im Großen Palast zu Konstantinopel’, Bonner Jahrbücher 187, 273–308.

280

Luisa Musso

Hellenkemper Salies, G. 1990a. ‘Kommagene – Kilikien – Isaurien’, Reallexikon zur byzantinischen Kunst 4, 319–56. Hellenkemper Salies, G. 1990b. ‘Konstantinopel’, Reallexikon zur byzantinischen Kunst 4, 613–25. Hellenkemper Salies, G. 1991. ‘Kilikische Pavimente. Ein Aspekt künstlerischer Auseinandersetzung zwischen Kilikien und der Levante’, Anatolia Antiqua/Eski Anadolu 1, 317–30. Hellenkemper Salies, G. 1994. ‘Die Datierung der Mosaiken im Großen Palast zu Konstantinopel’, Mosaïque gréco-romaine IV, 185–8. Hellenkemper Salies, G. 1999. ‘À propos des mosaïques du Grand Palais de Constantinople’, Mosaïque gréco-romaine VII, 685–90. Hill, S. 1995a. ‘Çiftlik’, Anatolian Archaeology. British Institute at Ankara Research Report 1, 2–3. Hill, S. 1995b. ‘The First Season of Rescue Excavation at Çiftlik (Sinop)’, Anatolian Studies 45, 219–31. Hill, S. 1996. ‘Çiftlik’, Anatolian Archaeology. British Institute at Ankara Research Report 2, 4–5. Hill, S. 1997. ‘Çiftlik’, Anatolian Archaeology. British Institute at Ankara Research Report 3, 2–3. Hill, S. 1998a. ‘Çiftlik’, Anatolian Archaeology. British Institute at Ankara Research Report 4, 6. Hill, S. 1998b. ‘Rescue Excavations at Çiftlik (Sinop)’, in R. Matthews (ed.), Ancient Anatolia. Fifty Years’ Work by the British Institute of Archaeology at Ankara. London, 285–300. Hommel, H. 1980. Der Gott Achilleus (Sitzungsberichte der Heidelberger Akademie der Wissenschaften. Philosophisch-historische Klasse, Abhandlung 1980, 1). Heidelberg. Humbert, J.-B. 1999. ‘The Rivers of Paradise in the Byzantine Church near Jabaliyah – Gaza’, in M. Piccirillo and E. Alliata (eds), The Madaba Map. Centenary 1897–1997. Travelling through the Byzantine Umayyad Period. Proceedings of the International Conference, Amman 7–9 April 1997 (Studium biblicum franciscanum. Collectio maior 40). Jerusalem, 216–18. Hupe, J. (ed.) 2006. Der Achilleus-Kult im nördlichen Schwarzmeerraum vom Beginn der griechischen Kolonisation bis in die römische Kaiserzeit. Beiträge zur Akkulturationsforschung (Internationale Archäologie 94). Rahden. Hupe, J. 2007. ‘Aspekte des Achilleus Pontarches-Kultes in Olbia’, in A. Bresson, A. Ivantchik and J.-L. Ferrary (eds) 2011. Une koinè pontique. Cités grecques, sociétés indigènes et empires mondiaux sur le littoral nord de la mer Noire (VIIe s. a.C. – IIIe s. p.C.) (Ausonius éditions. Mémoires 18). Bordeaux, 213–23. Huskinson, J. 2003. ‘Surveying the Scene: Antioch Mosaic Pavements as a source of Historical Evidence’, in I. Sandwell and J. Huskinson (eds), Culture and Society in Later Roman Antioch. Papers from a Colloquium, London December 2001. Oxford, 134–52. Huskinson J. M. 2005. ‘Rivers of Roman Antioch’, in E. Stafford and J. Herrin (eds), Personification in the Greek World. From Antiquity to Byzantium (Centre for Hellenic Studies, King’s College, London 7). Aldershot, 247–64. Icard-Gianolio, N. and A.-V. Zvabados 1992. ‘Nereides’, Lexicon Iconographicum Mythologiae Classicae 6.1, 785–824. Icard-Gianolio, N. and A.-V. Zvabados 2009, ‘Nereides’, Lexicon Iconographicum Mythologiae Classicae suppl. 2009, 385–9. Isager, S. 1997. ‘The late Roman villa in Halikarnassos. The inscriptions’, in Isager, S. and B. Poulsen (eds) 1997. Patron and Pavements in Late Antiquity (Halicarnassian studies 2). Odense, 24–9. Jobst, W. 2005. ‘Das Palastmosaik von Konstantinopel. Chronologie und Ikonographie’ Mosaïque gréco-romaine IX.2, 1083–1101. Jobst, W. 2011. ‘Das Mosaikpaviment der frühchristlichen Basilika von Gönen/Germe in Mysien (Hellespont)’, Mosaïque gréco-romaine XI, 483–504. Jobst, W., R. Kastler and V. Scheibelreiter (eds) 1999. Neue Forschungen und Restaurierungen im byzantinischen Kaiserpalast von Istanbul. Akten der Internationalen Fachtagung, Istanbul 1991 (Österreichische Akademie der Wissenschaften. Philosophisch-Historische Klasse. Denkschriften 273). Vienna. Jobst, W. and H. Vetters (eds) 1992. Mosaikenforschung im Kaiserpalast von Konstantinopel. Vorbericht über das Forschungs- und Restaurierungsprojekt am Palastmosaik in den Jahren 1983–1988 (Österreichische Akademie der Wissenschaften. Philosophisch-Historische Klasse, Denkschriften 228). Vienna.

Dynamics of mosaic design in northern Anatolia

281

Jouejati, R. 2012. ‘Les mosaïques d’une église apostolique à Tell Aar’, Zeitschrift für Orient-Archäologie, 272–93. Jouejati-Madwar, R. 2005. ‘A Mosaicist’s Workshop in Epiphania (Hama, Syria) at the Beginning of the 5th Century’, Mosaïque gréco-romaine IX.2, 775–84. Jucker, H. 1961. Das Bildnis im Blätterkelch. Geschichte und Bedeutung einer römischen Porträtform (Bibliotheca helvetica romana 3). Olten. Karabulut, H., M. Önal and N. Dervişoğlu 2011. Haleplibahçe Mozaikleri Şanlıurfa/Edessa. Istanbul. Kassab Tezgör, D. 2013. ‘Un état des travaux archéologiques sur la côte sud de la mer Noire’, in Bru, H. and G. Labarre (eds), L’Anatolie des peuples, des cités et des cultures (IIe millénaire av.J.-C. – Ve siècle ap.J.-C.). Colloque international de Besançon, 26–27 novembre 2010. 1. Autour du projet d’Atlas historique et archéologique de l’Asie Mineure antique (Collection Institut des Sciences et Techniques de l’Antiquité). Besançon, 167–74. Kemp-Lindemann, D. 1975. Darstellungen des Achilleus in griechischer und römischer Kunst (Archäologische Studien 3). Bern. Kilavuz, B. N. and E. Çelkbaş 2013. ‘Paphlagonia Hadrianoupolis’i’, Tarih Kültür ve Sanat Araştırmaları Dergisi 2, 159–214. Kitzinger, E. 1965. ‘Stylistic developments in pavement mosaics in the Greek East from the age of Constantine to the age of Justinian’, in G. Picard and H. Stern (eds), La mosaïque gréco-romaine. Colloque Paris 29 Août – 3 Septembre 1963 (Colloques internationaux du Centre national de la recherche scientifique). Paris, 341–52. Kökten, H. 2011. ‘A step forward in the Mosaic Corpus of Turkey: Condition Survey of Mosaics’, Mosaïque gréco-romaine XI, 975–81. Kondoleon, Ch. 1995. Domestic and Divine. Roman Mosaics in the House of Dionysos. Ithaca, N.Y. and London. Kondoleon, Ch. 2000. ‘Mosaics of Antioch’, in Ch. Kondoleon (ed.), Antioch. The lost ancient city. Published on the occasion of the exhibition held at the Worcester Art Museum, Worcester, Mass., Oct. 7, 2000 – Feb. 4, 2001, at the Cleveland Art Museum, Cleveland, Ohio, Mar. 18 – June 3, 2001, and at the Baltimore Museum of Art, Baltimore, Maryland, Sept. 16 – Dec. 30, 2001. Princeton, N.J. and Worcester, MA, 63–77. Kondoleon, Ch. 2007. ‘A Mosaic Floor under the Diakonikon’, in A. H. S. Megaw, Kourion. Excavations in the Episcopal Precinct (Dumbarton Oaks studies 38). Washington D.C., 351–9. Kossatz-Deissmann, A. 1981. ‘Achilleus’, Lexicon Iconographicum Mythologiae Classicae 1.1, 37–200. Kriseleit, I. 2000. Altes Museum, Pergamonmuseum. Antike Mosaiken. Berlin and Mainz. LaflI, E. and E. Christof, 2012. Hadrianopolis I. Inschriften aus Paphlagonia (BAR International Series 2366). Oxford. LaflI, E. and A. Zäh 2008. ‘Archäologische Forschungen im byzantinischen Hadrianoupolis in Paphlagonien’, Byzantinische Zeitschrift 101, 681–713. LaflI, E. and A. Zäh 2009. ‘Beiträge zu frühbyzantinischen Profanarchitektur aus Hadrianupolis – Blütezeit unter Kaiser Justinian I.’, Byzantinische Zeitschrift 102, 639–59. Lancha, J. 1997. Mosaïque et culture dans l’Occident romain (Ier–IVe s.) (Bibliotheca Archaeologica 20). Rome. Lassus, J. 1938a. ‘L’Église cruciforme. Antioche-Kaoussié 12-F’, in R. Stillwell (ed.), Antioch on-theOrontes, II. The Excavations 1933–1936 (Publication of the Committee for the Excavation of Antioch and its Vicinity) Princeton, N.J., 5–44. Lassus, J. 1938b. ‘Une villa de plaisance à Daphné-Yakto’, in R. Stillwell (ed.), Antioch on-theOrontes, II. The Excavations 1933–1936 (Publication of the Committee for the Excavation of Antioch and its Vicinity) Princeton, N.J., 95–147. Leader-Newby, R. 2005. ‘Personifications and paideia in Late Antique mosaics from the Greek East’, in E. Stafford and J. Herrin (eds), Personification in the Greek World. From Antiquity to Byzantium (Centre for Hellenic Studies, King’s College, London 7). Aldershot, 231–46. Leader-Newby, R. 2007. ‘Inscribed Mosaics in the Late Roman Empire: Perspectives from East and West’, in Z. Newby and R. Leader-Newby (eds), Art and Inscriptions in the Ancient World. Cambridge, 179–99. Levi, D. 1947. Antioch Mosaic Pavements, 2 vols. Princeton, N.J.

282

Luisa Musso

López Monteagudo, G. 1997. ‘Personificaciones alegóricas en mosaicos del Oriente y de Hispania: la representación de conceptos abstractos’, in J. M. Blázquez Martínez, A. González Blanco and R. González Fernández (eds), La tradición en la Antigüedad Tardía (Antigüedad y cristianismo 14). Murcia, 335–61. Maguire, H. 1987a. Earth and Ocean. The Terrestrial World in Early Byzantine Art (Monographs on the fine arts 43). University Park and London. Maguire, H. 1987b. ‘Adam and the Animals: Allegory and the Literal Sense in Early Christian Art’, Dumbarton Oaks Papers 41 (Studies on Art and Archaeology in Honor of Ernst Kitzinger), 363–73. Maguire, H. 1999. ‘The Nile and the Rivers of Paradise’, in M. Piccirillo and E. Alliata (eds), The Madaba Map. Centenary 1897–1997. Travelling through the Byzantine Umayyad Period. Proceedings of the International Conference, Amman 7–9 April 1997 (Studium biblicum franciscanum. Collectio maior 40). Jerusalem, 179–84. Maguire, H. 2002. ‘Paradise Withdrawn’, in A. Littlewood, H. Maguire and J. WolschkeBulmahn (eds), Byzantine Garden Culture. Papers presented at a colloquium in November 1996 at Dumbarton Oaks, Washington. Washington, D.C., 23–35. Majcherek, G. 1993. ‘Excavations in Alexandria, 1992–93’, Polish Archaeology in the Mediterranean, 5. Reports 1993, 11–20. Marek, Ch. 1988. ‘Forschungen im Zonguldak 1986’, 5. Araştırma Sonuçları Toplantısı, Ankara 6–10 Nisan 1987. Ankara, 403–7. Marek, Ch. 1993. Stadt, Ära und Territorium in Pontus-Bithynia und Nord-Galatia (Istanbuler Forschungen 39). Tübingen. Mavrojannis, Th. 1994. ‘L’Achilleion nel santuario di Poseidon e Anfitrite a Tenos. Un capitolo di storia della gens giulio-claudia in Oriente’, Ostraka 3, 291–347. Mayer, W. and P. Allen 2012. The Churches of Syrian Antioch (300–638 CE) (Late Antique history and religion 5). Leuven, Paris and Walpole, MA. Megaw, A. H. S. 2007. Kourion. Excavations in the Episcopal Precinct (Dumbarton Oaks studies 38). Washington D.C. Menichetti, M. 2009a. ‘Le armi magiche della guerra e della seduzione. I modelli omerici’, Incidenza dell’Antico. Dialoghi di storia greca 7, 137–57. Menichetti, M. 2009b. ‘Lo scudo e le armi magiche della guerra’, in M. d’Acunto, R. Palmisciano (eds), Lo scudo di Achille nell’Iliade: esperienze ermeneutiche a confronto. Atti della giornata di studi, Napoli 12 Maggio 2008 (Annali dell’Istituto Universitario Orientale di Napoli. Sezione filologicoletteraria 31). Pisa and Rome, 97–110. Mercan, A. and Ü. Göçen 2012. ‘Kadıoğlu Köyü 2010 Yılı Mozaik Kurtarma Kazısı’, in 20. Müze Çalışmaları ve Kurtarma Kazıları Sempozyumu, 25–29 Nisan 2011, Bodrum. Ankara, 135–47. Michaelides, D. 1988. ‘Mosaic pavements from Early Christian Cult Buildings in Cyprus’, in W. A. Daszewski and D. Michaelides (eds), Mosaic Floors in Cyprus (Biblioteca di Felix Ravenna 3). Ravenna, 81–153. Michaelides, D. 1992. Cypriot Mosaics (Picture book 7). Revised edition. Nicosia. Michaelis, A. 1882. Ancient Marbles in Great Britain. Cambridge. Mitford, T. B. 1971. The Inscriptions of Kourion (Memoirs of the American Philosophical Society 83). Philadelphia. Musso, L. 2008. ‘Il Sole è misura del Tempo. L’immagine del Tempo eterno nella sua manifestazione solare’, in M. Medri (ed.), Sentinum 295 a.C. – Sassoferrato 2006. Una città romana tra storia e archeologia. Convegno Internazionale, Sassoferrato 2006 (Studia archaeologica 163). Rome, 151–90. Musso, L., G. Bertolotto, M. Brizzi and B. E. Westwood 2011, ‘L’edificio abitativo alle pendici orientali dello Zımbıllı Tepe. Il sito e i rinvenimenti (1971–2010) – I mosaici dell’edificio abitativo’, in L. Summerer (ed.), Pompeiopolis I. Eine Zwischenbilanz aus der Metropole Paphlagoniens nach fünf Kampagnen (2006–2010) (Schriften des Zentrums für Archäologie und Kulturgeschichte des Schwarzmeerraumes 21). Langenweißbach, 75–120. Neira, L. 2011. ‘The Sea Thiasos of Nereids and Tritons in the Roman Mosaics of Turkey’, Mosaïque gréco-romaine XI, 631–54. Ochal-Czarnowicz, A. 2011. ‘The Cult of Achilles on the Coast of the Black Sea’, in E. PapuciWładyka, M. Vickers, J. Bodzek and D. Braund (eds), Pontika 2008. Recent Research on the

Dynamics of mosaic design in northern Anatolia

283

Northern and Eastern Black Sea in Ancient Times. Proceedings of the International Conference, Kraków 21st–26th April 2008 (BAR International Series 2240). Oxford, 269–74. Ostrowski, J. A. 1991. Personification of Rivers in Greek and Roman art. Warsaw and Krakow. Ovadiah, A. 2001. ‘The “Peopled” Scroll Motif in the Land of Israel in Late Antiquity’, in N. KenaanKedar and A. Ovadiah (eds), The Metamorphosis of Marginal Images: From Antiquity to Present Time. Tel Aviv, 1–10. Ovadiah, R. and A. 1987. Hellenistic, Roman and Early Byzantine Mosaic Pavements in Israel (Biblioteca Archeologica 6). Rome. Önal, M. 2009. Zeugma mosaics: a corpus. Istanbul. Öztürk, B. 2012. ‘Seyahatnamelerde ve Modern Literatürde Tios/Tieion ve Teritoryumu. Tios/Tieion and its Territorium in Historical Itineraries and Modern Literature’, Mediterranean Journal of Humanities 2.1, 161–76 (website: mjh.akdeniz.edu.tr). Öztürk, B. 2013. ‘The History of Tieion/Tios (Eastern Bithynia) in the Light of Inscriptions’, in M. Manoledakis (ed.), Exploring the Hospitable Sea. Proceedings of the International Workshop on the Black Sea in Antiquity, Thessaloniki 21–23 September 2012 (BAR International Series 2498). Oxford, 147–64. Parrish, D., M. Şahn and W. Jobst 2011. ‘The Corpus of Mosaics of Turkey: a new international research project’, Mosaïque gréco-romaine X, 619–27. PatacI, S. 2011. ‘Paphlagonia Hadrianoupolis’i Hamam A ve A Kilisesi Mozaikleri’, Journal of Mosaic Research 4, 27–50. PatacI, S., A. K. Öz and E. LaflI 2012. ‘Paphlagonia Hadrianoupolis’i Mozaik Buluntuları: Ön Değerlendirmeler’, Journal of Mosaic Research 5, 163–72. Picard, G. Ch. 1980. ‘De la Maison d’or de Néron aux thermes d’Acholla. Étude sur les grotesques dans la mosaïque romaine’, Monuments et mémoires. Fondation Eugène Piot 63, 63–104. Piccirillo, M. 1993. The Mosaics of Jordan (American Center of Oriental Research Publications 1). Amman. Piccirillo, M. 2009. ‘Il mosaico della Chiesa dei Santi Martiri nel villaggio di Tayyibat al-Imam – Hama (Siria), 442 d.C.’, in M. C. Lentini (ed.), Mosaici mediterranei. Caltanissetta, 13–31. Poulsen, B. 2012. ‘Identifying mosaic workshops in late antiquity: epigraphic evidence and a case study’, in T. M. Kristensen and B. Poulsen (eds), Ateliers and Artisans in Roman art and archaeology (Journal of Roman Archaeology. Supplementary series 92). Portsmouth, 129–44. Rutherford, I. 2009. ‘Black sails to Achilles: the Thessalian pilgrimage in Philostratus’ Heroicus’, in E. Bowie and J. Elsner (eds), Philostratus (Greek culture in the Roman world). Cambridge, 230–47. Salzmann, D. 1982. Untersuchungen zu den antiken Kieselmosaiken von den Anfängen bis zum Beginn der Tesseratechnik (Archäologische Forschungen 10). Berlin. Scheibelreiter, V. 2008. ‘Ephesische Kirchenmosaiken: eine Sprache der Bilder?’, Mitteilungen zur christlichen Archäologie 14, 71–102. Scheibelreiter-Gail, V. 2011. Die Mosaiken Westkleinasiens. Tessellate des 2. Jahrhunderts v.Chr. bis Anfang des 7. Jahrhunderts n.Chr. (Österreichisches Archäologisches Institut, Sonderschriften 46). Vienna. Scheibelreiter-Gail, V. 2012. ‘Inscriptions in the Late Antique Private House: Some Thoughts about their Function and Distribution’, in S. Birk and B. Poulsen (eds), Patrons and Viewers in Late Antiquity (Aarhus Studies in Mediterranean Antiquity 10). Aarhus, 135–65. Sevilla Sadeh, N. 2005–2006. ‘A Hero in a Girl’s Dress: The Meaning of Achilles’ Image and Marginal Decoration in a Mosaic from Shechem-Nablus’, Assaph. Studies in Art History 10–11 (Studies in Honour of Asher Ovadiah), 431–50. Sintès, Cl. (ed.) 1996. Musée de l’Arles antique. Arles. Solin, H. 2003. Die griechischen Personennamen in Rom. Ein Namenbuch. Second edition (Corpus inscriptionum Latinarum. Auctarium. Series nova 2). Berlin and New York. Stillwell, R. (ed.) 1938. Antioch on-the-Orontes, II. The Excavations 1933–1936 (Publication of the Committee for the Excavation of Antioch and its Vicinity). Princeton, N.J. Summerer, L. (ed.) 2011. Pompeiopolis I. Eine Zwischenbilanz aus der Metropole Paphlagoniens nach fünf Kampagnen (2006–2010) (Schriften des Zentrums für Archäologie und Kulturgeschichte des Schwarzmeerraumes 21). Langenweißbach.

284

Luisa Musso

Summerer, L. and A. von Kienlin 2013. ‘Pompeiopolis. The Metropolis of Paphlagonia’, in H. Bru and G. Labarre (eds), L’Anatolie des peuples, des cités et des cultures (IIe millénaire av.J.-C. – Ve siècle ap.J.-C.). Colloque international de Besançon, 26–27 novembre 2010. 1. Autour du projet d’Atlas historique et archéologique de l’Asie Mineure antique (Collection Institut des Sciences et Techniques de l’Antiquité). Besançon, 115–26. Şahn, D. 2004. Amisos Mozaiği (Ankara, T. C. Kültür ve Turizm Bakanlığı, Kültür Varlıkları ve Müzeler Genel Müdürlüğü). Ankara. Şahn, D. 2005. ‘The Amisos Mosaic of Achilles. Achilles cult in the Black Sea region’, Mosaïque grécoromaine IX.1, 413–26. Şahn, D. 2006. ‘Amisos Mozaiği Işığı Altında Karadeniz’de Akhilleus Kültü. The Iconography of Achilleus-Thetis in the Light of a Mosaic from Amisos’, in B. D. Ercyas and E. Koparal (eds), Karadeniz Araştırmaları Sempozyum Bildirileri. Black Sea Studies. Symposium Proceedings, Ankara 16–17 Nisan 2004 (Yerleşim arkeolojisi serisi. Sempozyum bildirileri 1). Istanbul, 139–60. Talgam, R. and Z. Weiss 2004. The Mosaics of the House of Dionysos at Sepphoris. Excavated by E. M. Meyers, E. Netzer and C. L. Meyers (Qedem 44). Jerusalem. TatlIcan, İ. 1997. ‘Sinop Çiftlik köyü, mozaik kurtarma kazısı’, 7. Müze Kurtarma Kazıları Semineri, 8–10 Nisan 1996, Kuşadası 1996. Ankara, 333–56. Ternes, C. M. 2003. ‘Muses on Mosaics’, in Ch. M. Ternes (ed.), Ancient Roman mosaics. Paths through the Classical Mind. Acta of the Conference held in March 2000 in Luxembourg 2000. Luxembourg, 167–95. Theophilidou, E. 1984. ‘Die Musenmosaiken der römischen Kaiserzeit’, Trierer Zeitschrift 47, 239– 348. Tissi, L. M. 2012. ‘La conversione di Bythos in Giovanni di Gaza’, Prometheus. Rivista di studi classici 38, 255–72. Toynbee, J. M. C. and J. B. Ward-Perkins 1950. ‘Peopled Scrolls: a Hellenistic Motif in Imperial Art’, Papers of the British School at Rome 18 (n.s. 5), 1–43. Trovabene G. 2005. ‘Divinità e personificazioni fluviali nei pavimenti musivi tardo-antichi: aspetti iconografici e variazioni semantiche’, Niš & Byzantium 3, 119–30. Tülek, F. 2011a. ‘Lore of late antique culture in Cilicia: the Syedra mosaics’, Mosaïque gréco-romaine X, 639–45. Tülek, F. 2011b. ‘The Bejewelled Lady of Sinope’, Mosaïque gréco-romaine XI, 921–6. Venturini, F. 2013. I mosaici di Cirene di età ellenistica e romana. Un secolo di scoperte (Monografie di archeologia libica 34). Rome. Visconti, P. 1825. ‘Relazione delli Ritrovamenti di antiche cose seguite in Roma, e ne’ suoi dintorni dal principio dell’anno 1823’, Atti della Pontificia Accademia Romana di Archeologia. Dissertazioni 2, 637–80. Vollenweider, M. L. 1966. Die Steinschneidekunst und ihre Künstler in spätrepublikanischer und augusteischer Zeit. Baden-Baden. Weiss, Z. 2011. ‘Mosaic Art in Ancient Sepphoris: Between East and West’, Mosaïque gréco-romaine XI, 941–51. Werner, K. E. 1994. Mosaiken aus Rom. Polychrome Mosaikpavimente und Emblemata aus Rom und Umgebung. Würzburg. Werner, K. E. 1998. Die Sammlung antiker Mosaiken in den Vatikanischen Museen. Vatican. Yaman, Z. 1991. ‘Kastamonu ili Taşköprü Ilçesi Pompeiopolis (Zımbıllı Tepesi Höyüğü) 1984 Yılı Kurtarma Kazısı’, 1. Müze Kurtarma Kazıları Semineri, 19–20 Nisan 1990, Ankara. Ankara, 63–111. Yeivin, Z. 1987–1988. ‘The Bet Shean Project’, Excavations and Surveys in Israel 6, 7–45. Zaqzuq, A. 1983. ‘Découvertes de mosaïques’, Les Annales Archéologiques Arabes Syriennes 33.2, 141– 78. Zaqzuq, A. 1995. ‘Nuovi mosaici pavimentali nella regione di Hama’, in A. Iacobini and E. Zanini (eds), Arte profana e arte sacra a Bisanzio (Milion 3). Rome, 237–56. Zaqzuq, A. R. and M. Duchesne-Guillemin 1970. ‘La mosaϊque de Mariamin (conservée au Musée de Hama)’ Les Annales Archéologiques Arabes Syriennes 20, 93–125. Zaqzuq, A. and M. Piccirillo 1999. ‘The Mosaic Floor of the Church of the Holy Martyrs at Tayibat al-Imam – Hamah, in Central Syria’, Liber annuus 49, 443–64.

Esra Keskin

Preliminary results of rescue excavations in 2013 at a Roman mosaic-paved building in the village of Yavru, Amasya province

1. Introduction Rescue excavations were undertaken over the course of a month in 2013 at Küpderesi, situated near the village of Yavru within the central district of the province of Amasya (Fig. 1) as a collaboration between the Amasya Museum (The Ministry of Culture and Tourism) and the Hittite University.1 The excavation team worked under the directorship of Celal Özdemir and the scientific guidance of the author. The team consisted of museum archaeologists Muzaffer Doğanbaş, İhsan Polat and Hilal Uçar, art historians Mert Mecek and Alper Atmaca, photographer Şahin Kul and 11 workers. The excavation was prompted by illegal excavations in Küpderesi and aimed to document the structures and mosaics which had been revealed. The pits opened by the treasure hunters were enlarged and turned into regular trenches in which a part of a structure with various rooms and graves were discovered (Fig. 2). This article reports on the rooms of the excavated building, some of which are mosaic-paved, and, briefly, the graves.

2. The rooms The first illegally opened trench (measuring 25m × 25m and 0.5m below ground level) included some architectural remains belonging to a north-south oriented room (7.57m × 4.15m) that was named AYKK-1 (Room 1). This room is divided into two parts. The rectangular part, oriented north-south, is paved with mosaics and has a larger apsidal section (2.6m × 3.38m and 0.47m below ground level) paved with tile (bipedalis). There is also a smaller apse in the eastern wall, which is slightly displaced from the centre. The walls are 82–5cm thick and their preserved height varies between 1.56m and 2.35m. The wall surface is coated with a 2cm-thick layer of hydraulic plaster (Fig. 3). 1

I would like to thank to Celil Öz, the Governor of Amasya, Reha Metin Alkan, Rector of Hitit University, Celal Özdemir, Director of Amasya Museum, and Ahmet Kaya of the Amasya Province Culture and Tourism Directorate for closely following the excavation and for their support in ensuring the quick and safe progress of our work. I also would like to thank all those who participated in the excavation, especially archaeologist Murat Türkmen for his devoted efforts. Special thanks are due to Lâtife Summerer for her improvement of this report.

286

Esra Keskin

Fig. 1: Location of Yavru, Amasya (cartography by R. Szydlak).

Fig. 2: Plan of the building (E. Keskin and T. M. Turan, ©Hitit University).

Preliminary results of rescue excavations in the village of Yavru

287

Fig. 3: Room 1, the apple tree mosaic (Ş. Kul, ©Amasya Museum). A 30cm-thick layer of debris consisting of roof tiles and mortar fragments lay above the mosaic floor; nonetheless, most of the mosaic pavement was recovered in a well-preserved condition, although it had been partly destroyed by the illegal excavations.2 The mosaic is decorated with vegetal, floral and geometric motifs, and animal figures (the Apple Tree Mosaic). The outermost edge of the mosaic is decorated with interlocking motifs of twisted waves and ivy branches, while the main panel consists of a circular emblem, 126cm in diameter (Fig. 4). The outer frame of this central panel is surrounded by a large guilloche motif and interconnected circles (antrolac). The small circles are divided in two equal sections which are decorated with double-strant guilloches and wavy ribbon bands respectively. The space in the corners between the guilloche and the rectangular frame is filled with vegetal motifs. The main emblema (64cm in diameter) shows a naturalistic representation of an apple tree beneath which three partridges are walking. Two of the four apples on the tree are shown amongst the leaves, but two riper and larger ones hang lower. The bichrome rendering of the apples – yellow on one side, red on the other – gives the fruit a naturalistic appearance. The addition of a few white tesserae provides the impression of light reflecting off the apples. The vegetation of the landscape is indicated by herbs shown against the tree trunk. The three partridges, placed on either side of the trunk, face the tree. Red tesserae were used for the beaks and feet of the birds, while their wings and tails consist of black tesserae. 2

Many tesserae found scattered around were collected and the soil from the illegally-excavated pits was sieved to recover more.

288

Esra Keskin

Fig. 4: Room 1, detail of mosaic (Ş. Kul, ©Amasya Museum). The second room was unearthed immediately west of AYKK-1 (Figs 2 and 5). This room, labelled AYKK-2 (Room 2), is rectangular in plan, measures 5.2m × 2.56m and is oriented north-south. The room was accessed via two door openings: one to the south and one to the west. The walls were preserved to a height of up to 90cm. They were constructed of rubble bound with a coarse mortar. Some pottery sherds were recovered during removal of the fill of the room. The floor was laid in lime mortar. The floor level of this room is lower than that of Room 1, which may be a consequence of the sloping terrain. The third room, north of AYKK-1 and referred to as AYKK-3 (Room 3), measures 6.4m × 4.21m. It too has a rectangular plan and extends in a north-south direction (Fig. 2). While the eastern and northern walls were buried, the upper part of the eastern wall (0.85m thick and with a preserved height of 2.2m) is slightly inclined inwards due to exterior pressure. The entrance in the northern wall of the room, framed by two pilasters, lies 20cm above the floor level. The 3.03m-wide opening between the pilasters defines the entrance as monumental (Fig. 6). The lower parts of the pilasters are built of rubble and brick, and measure 69cm by 86.5cm. The western pilaster is preserved to a height of 47.5cm and the eastern one to a height of 104cm. In the upper parts of each of the pilasters, at a height of 50cm, three rows of bricks are inserted. The bricks measure 24cm × 37.5cm and are 4cm thick. A stone threshold with doorposts was found between the pilasters. The floor level between the pilasters is higher than the mosaic-paved room. This higher level continues into Room 2, which is paved with large, square tiles measuring 43cm × 43cm. The western wall of Room 3 is 3.38m long and it has a wide opening situated approximately at the centre. Since the back of the western wall remains buried, it is not known whether or not this leads to another room. The floor of Room 3 is paved with a mosaic measuring 6.4m × 4.21m (the Carpet Patterned Mosaic) (Fig. 7). The composition of the mosaic consists of two rectangular panels

Preliminary results of rescue excavations in the village of Yavru

Fig. 5: Room 2 (Ş. Kul, ©Amasya Museum).

Fig. 6: Room 3, carpet patterned mosaic (Ş. Kul, ©Amasya Museum).

289

290

Esra Keskin

Fig. 7: Room 3, north section (Ş. Kul, ©Amasya Museum). with various geometric patterns surrounded by a border. The border is designed with lozenges and squares decorated with saw-tooth lines and floral swastikas. Both panels are framed with a crested wave motif. The first panel, to the south (2.49m × 2.87m), has a patchwork-like geometric composition comprising 49 squares filled with triangular, semicircular, guilloche chain and saw-tooth line motifs. A single square in the second row from the west displays an ivy branch. The northern panel consists of three sections. The two northernmost are decorated with alternating circular and rectangular motifs. In the second larger panel three squares alternate with three guilloche knots. The southernmost section of the mosaic is a 30cm-high frieze decorated with pelta and triangular and circular patterns.3

3. The graves In the north of trench YK-1, two more trenches measuring 3m × 6m were opened to reveal some tile graves (Fig. 8). The graves were covered with two rows of leaning tiles measuring 30cm × 50cm. They had been largely destroyed due to ploughing. No grave goods were found, although the skeletons were in situ. Since the skeletons were much deteriorated, they were not suitable for anthropological analyses. Thus, they have been reburied. It is, however, clear that this location was used as a necropolis after the mosaic-paved building was abandoned.

3

Er 2006: 306.

Preliminary results of rescue excavations in the village of Yavru

291

Fig. 8: Tile graves (Ş. Kul, ©Amasya Museum).

4. Discussion The rescue excavation in the village of Yavru unearthed three rooms of a building oriented north-south with rooms of rectangular plan. Room 1 seems to have different building phases. The apse to the south has a tile paved floor at a lower level of the main Room. The fact that the apse in the eastern wall is not located centrally suggests that it was added at a later date. How Room 1 was accessed is not clear. Being located in the axis of the southern apse Room 2 could have opened to Room 1. However, there is no door opening preserved between these two rooms and the floor level of the entrance to Room 2 is lower than the level of the floor in Room 1. The function of Room 3 is also poorly understood. Further excavations might elucidate its relationship with the adjacent rooms. Consequently, many questions remain unanswered regarding the chronology and function of the building. However, the large apse recalls the stibadium of a dining room characterised by the semicircular couch.4 Thus, the structure excavated might have been a part of a rural Roman villa. The mosaic floors unearthed in two of the rooms were buried under a thick layer of building debris. Thus, the mosaics are preserved in relatively good condition, though they have lost much of their multi-coloured brightness as a consequence of exposure to lime mortar. The construction technique is opus tesselatum utilizing black, cream, white, green, red and yellow tesserae in vegetal motifs, geometric patterns and animal figures. The apple tree motif of the mosaic in Room 1 is rare in Roman mosaic art. The apple has symbolic meaning in both religious and mythological narratives: the ‘apple of discord’ of the Trojan Wars and the forbidden of the Bible.5 On the other hand, northern Anatolia was renowned 4 5

Dunbabin 1991: 121–48. Generally on the symbolic of apple see: Littlewood 1968: 147–181; Uzun-Öğüş 2010: 147–81.

292

Esra Keskin

for its apple trees in antiquity, as noted by Strabo;6 in fact, Amasya remains the most famous apple producing region of Turkey today. Beside the beauty of the city, 19th-century travelers also mention the unique taste of its apples. These travelers equally refer to the partridges of the region, and thus those represented in mosaic under the apple tree may also be seen as local motifs.7 Animals of the local region are often depicted in Roman mosaic art.8 Thus, the composition with an apple tree and partridges could well have been created in a local workshop. On the other hand, similar compositions utilizing the geometric patterns seen in the Yavru mosaics, especially the interconnected circles, are to be found, for example, in Zeugma dated to the third century.9 Motifs such as waves, guilloche, antrolac, ribbon, pelta and other geometric motifs are generally used across multiple centuries of Roman mosaic art.10 However, similar carpet-like designs in rainbow style occur on the mosaics of the fourth century.11

5. Conclusion The partly-excavated structure revealed during rescue excavation at Yavru is not well understood. However, the monumental scale and the large apse in Room 1 point to a residential building, probably a Roman rural villa, which had at least two construction phases. Close parallels for the mosaics can be found at Zeugma and Antioch, but also in other centres of the Roman East. Based on these parallels, the building can be assigned to the late Roman period. Even though they present common motifs of the period, they also reflect some regional characteristics, which suggests the presence of local workshops. The fact that no finds were recovered, other than a small number of pottery sherds, and that a thick layer of debris covered the structure suggests that it was abandoned after a relatively short period of use. The location was then used as a cemetery. Following the excavation, protective measures were taken at the site. The walls were wrapped in geotextile and their edges protected by sand bags. The better preserved walls in Rooms 1 and 3 were protected with a roof, while the mosaics themselves have been covered with alternating layers of sand and fine net. Esra Keskin Hitit University The Black Sea Archaelogical Application and Research Center (KARM) Çepni Mahallesi Fen Lisesi Caddesi No:58, Çorum 19100, Turkey [email protected]

6 7 8

12.3.15. Tuzcu 2013: 402, 414; Özdemr 2014: 90. For the Hatay Museum mosaics, see http://www.hatayarkeolojimuzesi.gov.tr/HatayMuzeWeb/ flash/main_EN.html (accessed January 2015). For the Great Palace Mosaics in Constantinople, see Jobst et al. 1997. For those of Zeugma, see Ergeç 2007; Dunbabin 2013. For Paphlagonia Hadrianoupolis Church A, see PatacI 2011, figs 6–15. 9 Ergeç 2007: 52–3, 192–3. For the antrolac motif variant in Bodrum, see Özet 2009: fig. 26. 10 For motifs in Roman mosaic art, see Üstüner 2007: 85–108; Dunbabin 1999; Ling 1998; Field 2002. 11 For example in Antioch (Levi 1971: 111a.182, pl. 72a and in the synagogue of Apamea (Balty 1986: figs 5–13). See also the mosaic building in the agora in Smyrna: Ersoy and Alatepel 2011: figs 5–8.

Preliminary results of rescue excavations in the village of Yavru

293

Bibliography Balty, J. 1986. Mosaïques d’Apamée. Brussels. Dunbabin, K. M. D. 1991. ‘Triclinium and Stibadium’, in W. J. Slater (ed.), Dining in a Classical Context. Ann Arbor, 121–48. Dunbabin, K. M. D. 1999. Mosaics of the Greek and Roman World. Cambridge. Dunbabin, K. M. D. 2013. ‘Mosaics’, in W. Aylward (ed.), Excavations at Zeugma I, Los Altos, 149– 67. Er, Y. 2006. Klasik Arkeoloji Sözlüğü. Ankara. Ergeç, R. 2007. Belkıs/Zeugma and Its Mosaics. Istanbul. Ersoy, A. and S. Alatepel 2011. ‘Smyrna Agorası Mozaikli Yapı Mozaikleri’, Journal of Mosaic Research 4, 17–26. Field, R. 2002. Geometric Patterns from Roman Mosaics and How to Draw Them. St. Albans. Jobst, W., B. Erdal and C. Gurtner 1997. İstanbul Büyük Saray Mozaiği. Istanbul. Özdemr, C. 2014. ‘Amasya’ in F. Özdem (ed.), Yar ile Gezdiğim Dağlar. Amasya/Istanbul, 73–99. Özet, A. 2009. ‘Bodrum-Torba Monastery Mosaics’, Journal of Mosaic Research 3, 71–82. Levi, D. 1971. Antioch Mosaic Pavements. Rome. Ling, R. 1998. Ancient Mosaics. Princeton. Littlewood, A. R. 1968. ‘The Symbolism of the Apple in Greek and Roman Literature’, Harvard Studies in Classical Philology 72, 147–81. PatacI, S. 2011. ‘Paphlagonia Hadrianoupolis’ i Hamam A ve A Kilisesi Mozaikleri’, Journal of Mosaic Research 4, 27–50. Tuzcu, A. 2013. İlk Çağlardan Cumhuriyete Seyahatnamelerde Amasya. Amasya. Uzun Öğüş, E. 2010. ‘Fruit Trees on Late Antique Mosaics: From The Worchester Hunt to Jordanian Churches’, Olba 18, 395–419. Üstüner, A. C. 2007. Mozaik Sanatı. Istanbul.

Tønnes Bekker-Nielsen and Rainer M. Czichon

Ancient Roads and Bridges of the Vezirköprü District

1. Introduction In an important paper on Roads in Pontus, Royal and Roman published more than a century ago, J. A. R. Munro described the geography of the Phazemonitis as defined by two main transport axes: the ‘commercial road’ from Zile to Samsun and the ‘trunk road’ from Niksar over the Kılıçarslan pass and through Vezirköprü to Taşköprü.1 According to Munro’s analysis, the city Laodikeia, modern Ladik, was created to guard their point of intersection,2 which Munro located between Ladik and Havza.3 Thus both roads were in existence before the arrival of the Romans and Pompey merely placed his new cities on an already existing highway. Later renovations of this ‘trunk road’ are attested by numerous milestones of the Imperial period.4 This reconstruction, however, raises several problems. First, Ladik is 4km distant from the nearest point of the ‘trunk road’ and more than 10km distant from the road crossing which it was supposedly intended to control. Likewise, the ‘trunk road’ makes an inexplicable 90° bend in Vezirköprü; if, as Munro claimed, the road predates the city, one would expect the ‘trunk road’ to pass straight through the city. Furthermore, the hypothesis fails to provide a plausible context for the structure known as the Kurt köprüsü (‘wolf bridge’), which spans the Istavroz çayı 13km downstream of the Roman bridge at Köprübaşı. The present superstructure with ogive arches (Fig. 1) is clearly post-Roman but the piers incorporate reused blocks which are evidently ancient and may indeed rest upon Roman foundations.5 The Cumont brothers, who visited the site in 1900, assumed the bridge was of Seleucid date and formed part of a ‘chemin de traverse de Vézir-Keupru à Samsoun’.6 Their co-worker Henri Grégoire, on the other hand, attempted to trace a direct Roman road from Vezirköprü to Kavak, but without success.7 Subsequent studies of the region’s road network have followed Munro’s basic premise that the road connecting Pompey’s cities and the road attested by Roman milestones of the Imperial period were one and the same, following a course that took it through Havza, 1 2 3 4 5 6 7

Munro 1901: 53–4. Munro 1901: 60. Munro 1901: 55. Munro 1901: 63. AlkIm 1975: 6; Halfeoğlu, Sert and YIlmaz 2013: 89. Cumont and Cumont 1906: 129. Grégoire 1909: 7–8.

296

Tønnes Bekker-Nielsen and Rainer M. Czichon

Fig. 1: The Kurt köprüsü. Since the photo was taken, the bridge has undergone a heavy-handed restoration (photo by B. Kıvrak). ancient Thermai tôn Phazemonitôn.8 The results of archaeological surveys undertaken in 2010 and 20139 under the aegis of the Nerik project (Free University of Berlin and University of Uşak) and the Where East Meets West project (University of Southern Denmark and University of Copenhagen) respectively, have shown that this hypothesis is seriously flawed and can no longer be upheld. In fact, there were two Roman roads linking Vezirköprü to the ‘commercial road’ and the road network of the Phazemonitis has a longer and more complex history than previously assumed.

2. The road from Doyran to Tepeören In 2010, a vertical section through the Roman highway was exposed by roadworks at the northern approach to Doyran village (Fig. 2).10 This definitely proved what milestone finds had already suggested: that the Roman road coming from Havza entered Vezirköprü from

8 9

Olshausen 1999: 104: ‘West-Ost Binnenroute’. We are grateful to all members of the survey team and especially to Bünyamin Kıvrak for their contribution, as well as to government representatives Müğe Küçük (2010) and Mustafa Kolağasıoğlu (2013). 10 Bekker-Nielsen 2013a: 206–7.

Ancient Roads and Bridges of the Vezirköprü District

297

Fig. 2: Longitudinal section through the bed of the Roman highway just north of Doyran village (photo by T. Bekker-Nielsen). the south along an orientation corresponding to today’s Köprülüler Caddesi, which probably preserves the alignment of the ancient cardo maximus (Fig. 3). This road was studied in further detail in October 2013. It can be followed in the landscape for most of the distance between Doyran and the northern approach to Tepeören village. Since it is oriented nearly directly south, the road from Vezirköprü to Tepeören cannot have been intended to reach the Kılıçarslan pass, which lies east-southeast of Vezirköprü. In 1951, C. A. Burney hypothesised in very general terms that a Bronze Age road had linked the valley of the Amnias (Gökırmak) with Amasya via Vezirköprü and Havza;11 a similar line of reasoning was followed by David Winfield in 1977.12 In a more recent study, J. M. González Salazar and R. L. Castejón describe a possible Hittite route from Alaçam on the plain of Bafra via Oymaağaç and Havza to Amasya.13 Oymaağaç Höyük has been convincingly, though not conclusively, identified as the Hittite centre of Nerik;14 even if this should not be the case, the site was evidently a place of considerable importance in a frontier region of the Hittite Empire, and must have been connected by a road or track southward to the capital at Hattuşa.

11

Burney 1956: 180: ‘The modern road continues via Vezirköprü to Amasya, and probably in ancient times did likewise’. 12 Winfield 1977: 158. 13 Gonzalez Salazar and Castejón 2008: 18. 14 Czichon 2008. For details of the ongoing work at Oymaağaç Höyük, see the project website www. nerik.de.

298

Tønnes Bekker-Nielsen and Rainer M. Czichon

Fig. 3: The city site of Neapolis and its roads (cartography by R. Szydlak).

Ancient Roads and Bridges of the Vezirköprü District

299

Fig. 4: Remains of a settlement along the road from Tepeören into the Tavşan range (photo by T. Bekker-Nielsen). The 2013 survey of the Where East Meets West project identified a possible Old Hittite kale overlooking the İstavroz çayı about 2km southeast of Tepeören village and an extensive late Roman settlement on the plain towards the west.15 South of the İstavroz çayı, remains of a terraced switchback road can be seen ascending the hillside (Fig. 4) and heading into the Tavşan range. On a plateau, at an altitude of ca 1,100m above sea level, remains of an ancient settlement were found along the road-line.16 The road may have continued across the divide of the Tavşan range to the plain of Merzifon.17 Heading southeast from Tepeören another road can be traced, crossing the İstavroz çayı on a multi-span bridge a short distance upstream of the present highway bridge at Köprübaşı and climbing over the Güvenbeli pass to reach the plain around Havza and link up with Munro’s ‘commercial road’.

3. A bridge at Yürükçal In October 2013, a local Where East Meets West project team member, Bünyamin Kıvrak, reported that remains of an ancient structure were to be seen in the bed of the Kuz çayı just below Yürükçal village. Closer inspection revealed the remains of a substantial bridge with at least five piers, one of which had been cut from the living rock. The construction of the 15 16 17

See Winther-Jacobsen this volume. See Winther-Jacobsen this volume. John MacDonald Kinneir, who visited Vezirköprü in 1814, was told that the distance was ‘six hours from Marsawan [Merzifon], and twelve from Amasia’ (Kinneir 1814: 298). The distances by the modern road are 58km and 76km respectively, and, since a rider could hardly be expected to cover nearly 10km per hour (cf. Wilson’s estimate, following note) it would seem that Kinneir’s six hours from Vezirköprü to Merzifon were reckoned via a shortcut across the Tavşan range.

300

Tønnes Bekker-Nielsen and Rainer M. Czichon

Fig. 5: Remains of a pier of the bridge at Yürükçal and paving in the bed of the stream (photo by T. Bekker-Nielsen).

Fig. 6: Masons’ marks on the bridge at Yürükçal (photo by T. BekkerNielsen). other piers (Fig. 5) resembles the bridge at Köprübaşı: a concrete core faced with large stone blocks, some of which bear masons’ marks in the Greek alphabet (Fig. 6). Another common feature is slab paving in the bed of the stream; that which is now visible at Köprübaşı is post-Roman (incorporating reused blocks), but at Yürükçal, the preserved paving appears to be original and in situ. North of the stream, the ancient road can be traced as far as the ridge above Güldere village (Fig. 7). South of the river, it probably followed the same route as the modern road through Çalköy village, entering Vezirköprü from the east. Further studies will be required to establish the purpose of this route and its destination.

Ancient Roads and Bridges of the Vezirköprü District

301

Fig. 7: Section of the ancient road between Yürükçal and Güldere (photo by T. Bekker-Nielsen).

4. Pompey’s road In the course of the October 2013 survey, sections of a road leading east from Vezirköprü were identified at several locations. The section nearest the modern city could not be traced and, like the road from the Yürükçal bridge (above), its course is probably overlain by modern tarmac roads. From a point east of Çakırtaş village, however, the road can be followed through the terrain for a distance of approximately 2km before it disappears in the intensively-cultivated fields. East of Aydoğdu village, the road reappears as an impressive rock-cut stream crossing (Fig. 8); from here, its alignment is preserved as far as the ridge south of Tekkekıran, overlooking the descent to the Kurt köprüsü. Beyond the Kurt köprüsü, a stretch of ancient road is preserved south of Kocaoğlu village, aiming for Ilıca (Fig. 9). Since the area east of Ilıca is outside the area covered by our survey permit, the road was not traced for the remainder of its course to its intersection with the ‘commercial road’. Nonetheless, there can be little doubt that these sections of road, as well as the Kurt köprüsü, form elements of Pompey’s ‘trunk road’ from Pompeiopolis to Nikopolis. Coming from the crossing of the Kızılırmak and approaching Vezirköprü from the west, it passed straight through the site of Neapolis, as one would expect if the road were earlier than or contemporary with the city. When it emerged on the eastern side of the city, it followed a direct east-southeasterly course via the Kurt köprüsü to its point of intersection with the ‘commercial road’, then continued along the northern shore of the Ladik gölü towards the Kılıçarslan pass and the Phanaroia. As it takes a more direct course, the

302

Tønnes Bekker-Nielsen and Rainer M. Czichon

Fig. 8: Rock cutting at the crossing of a stream immediately east of Aydoğdu village (photo by T. Bekker-Nielsen).

Fig. 9: Remains of the ancient road south of Kocaoğlu (photo by T. Bekker-Nielsen).

Ancient Roads and Bridges of the Vezirköprü District

303

Fig. 10: The ancient road network of the Phazemonitis (cartography by R. Szydlak). distance from Vezirköprü to the Ladik gölü by this route is about 5km shorter than via Köprübaşı and Havza; its summit is also at a lower altitude than the 880m of the Güvenbeli pass.18

5. Summary and conclusions The earliest axis of communication that can be identified in the Phazemonitis is the northsouth route leading from the Hittite site at Oymaağaç to the junction at Tepeören where the road divided, one branch continuing into the Tavşan dağları and onwards to the plain of Merzifon, the other branch heading southeast towards the Havza gap and Amasya. These routes, originally perhaps no more than tracks, were eventually upgraded to Roman engineering standards. Also of no doubt is an early date for the road leading northeast from the Havza gap towards Samsun – Munro’ s ‘commercial road’ – following as it does a natural corridor through the landscape (Fig. 10).

18

According to Sir Charles Wilson’s Handbook for Travellers in Asia Minor there were in his time two routes from Vezirköprü to Samsun, one via Havza requiring 22 hours and a ‘direct road’ taking only 20 hours (Wilson 1895: route 2, p. 7), a time saving which corresponds to the shorter distance via the Kurt köprüsü. Although this would seem to vindicate the Cumonts’ conjecture that the route functioned as a ‘chemin de traverse’, Wilson’s information needs to be treated with some caution: cf. Bekker-Nielsen 2013b: 15 n. 72.

304

Tønnes Bekker-Nielsen and Rainer M. Czichon

The east-west ‘trunk road’ came much later, probably as a part of Pompey’s reorganisation of Pontos following the Roman conquest. The primary purposes of this road were military and strategic. Linking two landscapes – the Phazemonitis and the Phanaroia – with broadly similar environmental characteristics and capable of producing more or less the same crops, it was never destined to be an important commercial artery. During the early years of Roman occupation, it linked the five cities which between them controlled the vast sprawling territory of inland Pontus. When that territory was dismembered by Mark Antony and large chunks of it assigned to local dynasts, there was no longer any central authority capable of, or interested in, maintaining the road. By the time Roman rule was reestablished, the pax Romana had eliminated the need for a purely military road. The northern route from Vezirköprü to the ‘commercial road’ via the Kurt köprüsü, having served its purpose for a short time, ceased to be maintained on a regular basis. Instead, Roman governors and road engineers now concentrated their efforts on the road from Havza via Köprübaşı and Tepeören to Vezirköprü, which became the new ‘trunk road’ of the Phazemonitis. Tønnes Bekker-Nielsen University of Southern Denmark Department of History Campusvej 55, 5230 Odense M, Denmark [email protected] Rainer M. Czichon Uşak Üniversitesi Fen Edebiyat Fakültesi, Arkeoloji Bölümü 1 Eylül Yerleşkesi, 64200 Uşak, Turkey [email protected]

Bibliography AlkIm, U. B. 1975. ‘Samsun Bölgesi Çalışmaları 1973’, Türk Arkeoloji Dergisi 22, 5–12. Bekker-Nielsen, T. 2013a. ‘Neapolis-Neoklaudiopolis: A Roman City in Northern Anatolia’, in Ü. Yalcn (ed.), Anatolian Metal VI (Der Anschnitt, Beiheft 25). Bochum, 203–14. Bekker-Nielsen, T. 2013b. ‘350 years of research on Neapolis (Vezirköprü)’ Orbis Terrarum 11, 3– 31. Burney, C. A. 1956. ‘Northern Anatolia before Classical Times’, Anatolian Studies 6, 179–205. Cumont, F. and E. Cumont 1906. Voyage d’exploration archéologique dans le Pont et la Petite Arménie (Studia Pontica 2). Brussels. Czichon, R. M. 2008. ‘Die hethitische Kultur im Mittleren Schwarzmeergebiet unter besonderer Berücksichtigung der Umgebung von Vezirköprü’, in: G. Wilhelm (ed.), Hattusa-Bogazköy: Das Hethiterreich im Spannungsfeld des Alten Orients, 6. Internationales Colloquium der Deutschen OrientGesellschaft, 22.–24. März 2006, Wiesbaden, 265–77. Ercyas, D. B. 2006. Wealth, Aristocracy and Royal Propaganda under the Hellenistic Kingdom of the Mithradatids in the Central Black Sea Region of Turkey (Colloquia Pontica 12). Leiden. González Salazar, J. M. and R. Castejón 2008. ‘Itineraria Anatolica Antiqua: La red viaria de los sectores montañosos septentrionales de Anatolia: las rutas nororientales y el Imperio hitita’, Gerion 26, 7–22. Grégoire, H. 1909. ‘Rapport sur un voyage d’exploration dans le Pont et en Cappadoce’, Bulletin de correspondance hellenique 33, 3–169. Halfeoğlu, F. M., H. Sert and S. YIlmaz 2013. ‘Tarihi Kurt Köprüsü (Mihrapli Köprü, Vezirköprü) Restorasyonu Proje ve Uygulama Çalişma’, METU Journal of the Faculty of Architecture 30, 81–104.

Ancient Roads and Bridges of the Vezirköprü District

305

Kinneir, J. MacDonald 1818. Journey through Asia Minor, Armenia, and Koordistan in the years 1813 and 1814; with remarks on the Marches of Alexander and Retreat of the Ten Thousand. London. Munro, J. A. R. 1901. ‘Roads in Pontus, Royal and Roman’, Journal of Hellenic Studies 21, 52–66. Olshausen, E. 1999. ‘Pontica IV: Das römische Straßennetz in Pontos: Bilanz und Überblick’, Orbis Terrarum 5, 93–113. Wilson, C. 1895. Handbook for Travellers in Asia Minor, Transcaucasia, Persia etc. London. Winfield, D. 1977. ‘The Northern Routes across Anatolia’, Anatolian Studies 27, 151–66.

Christian Marek

Epigraphy and the provincial organisation of Paphlagonian cities in the Roman Empire

The ‘epigraphic habit’ in Paphlagonia is a phenomenon confined almost entirely to inscriptions dating to the Roman Imperial period. The approximately 800–900 inscriptions on stone that are known from the region are distributed amongst the territories of three coastal cities – all of which date back to the Archaic period of Greek colonisation in the Black Sea – and four cities in the interior.1 Of the latter, only Gangra, as the place of residence of a local dynasty, can be considered a noteworthy pre-Roman settlement.2 The sole example of a text that could possibly be dated as early as the Hellenistic period is from its vicinity: a dedication to a local goddess venerated as Hera of Kandara.3 Today, about 60 inscriptions are known from Gangra and about 140 from Pompeiopolis – with two uncertain exceptions ranging from the Trajanic to the Christian period. The diachronic distributions of the inscriptions from Kaisareia-Hadrianopolis and Phazemon-Neapolis are similar, with a palpable climax during the Antonine and Severan periods. An analysis of the various aspects of epigraphic trends has recently been published by Jakob Munk Højte.4 It is unnecessary to repeat his insights here. Instead, I would like to 1

The coastal cities are Amastris, Abonouteichos-Ionopolis, and Sinope. The cities in the interior are Kaisareia-Hadrianopolis, Gangra-Germanikopolis, Pompeiopolis, Phazemon-Neapolis-Neoklaudiopolis. The proposal made in Bekker-Nielsen 2014 to separate the latter city from Paphlagonia is not convincing. The key argument is based on Strabo 12.3.38, where Phazemonitis is said to be part of Πόντος. But Πόντος here – as in the whole of Strabo 12.3 – designates the territory of the former Mithridatic kingdom, to which the northern part of Paphlagonia (including Blaene, Domanitis, Pimolisene and Phazemonitis) belonged. This becomes clear (not only) from 12.3.9: περὶ μὲν δὴ τῶν ἐν τῇ μεσογαίᾳ Παφλαγόνων ἐροῦμεν ὕστερον τῶν μὴ ὑπὸ τῷ Μιθριδάτῃ [i.e. the southern part of Paphlagonia, not part of the kingdom], νῦν δὲ πρόκειται τὴν ὑπ᾿ ἐκείνῳ χώραν, κληθεῖσαν δε Πόντον, διελθεῖν. Starting from coastal Paphlagonia (Amastris, Abonouteichos, Sinope), Strabo’s description proceeds in a circle eastwards along the coast to Amisos, Themiskyra, Sidene, Pharnakeia, Trapezous, then – after a philological excursus (20–27) – to the interior parts of the country southwards to Lesser Armenia and westwards via Phanarhoia, Kabeira, Komana and Zela to Phazemonitis and Amaseia (38), to conclude (40: λοιπὴ δ᾿ ἐστίν) with the northern region of interior Paphlagonia contiguous with Sinope (συναφὴς τῇ Σινωπίδι) that had been part of the ‘Pontic’ kingdom and consequently also part of Pompey’s province (τῆς Ποντικῆς ἐπαρχίας). The previously-announced description of the portion of Paphlagonia which had not been under Mithradates’s rule – τὰ ἐν τῇ μεσογαίᾳ … τὰ μὴ ὑπὸ τῷ Μιθριδάτῃ – is taken up at 12.3.41. 2 Of the Milesian colonies on the Black Sea coast, only Sinope has yielded inscriptions from the Classical period; at least one find of a Hellenistic stone inscription is known from Abonouteichos. The epigraphy of Amastris does not antedate the Julio-Claudian period. 3 Kaygusuz 1983: 60 no. 2. 4 Højte 2006: 15–30.

308

Christian Marek

focus on a particular topic: the ‘provincial organisation of the Paphlagonian cities’. I have settled on this for two reasons: first, to consider the evidence of new documents and, secondly, to deal with relatively recent developments in scholarship that purport as communis opinio what seem to me to be misconceptions. In the first part of what follows I shall concentrate on discussion of a Paphlagonian provincial koinon. The second part deals with new evidence from milestones for the province of Paflagonia in the period of the late Roman Empire. For a different view, the up-to-date treatise by Marco Vitale on eparchia and koinon is hoped to be groundbreaking.5 It furnishes an analysis, both exhaustive and meticulous, of the phenomena in Asia Minor as a whole.

1. A Paphlagonian provincial koinon To start at the beginning: it is quite clear that Pompey confined the part of inland Paphlagonia to be incorporated into the initial province of Pontus et Bithynia to the territories of two cities founded by him – Pompeiopolis and Neapolis. Gangra remained outside and continued to be the residence of dynasts; these were Attalos at the time of Pompey’s presence, Kastor II in 40 BC and Deiotaros Philadelphos a few years later. The extent of the eastern portion of the province called Pontus, to which Paphlagonia belonged, was considerably reduced by Mark Antony. The evolution of new provinces emerged in central and northern Anatolia from the time of Augustus, and none of the provincial formations ever again attributed the inland Paphlagonian cities to the province of Pontus et Bithynia, to which they had initially belonged for a short period of time.6 As far as the formation of provinces is concerned, there is but little uncertainty before the reign of the Severi; continuing controversy, however, obscures a closely-related subject of eminent importance: the internal civic organisation of the so-called provincial koina. The phenomenon in Asia Minor goes back to the Hellenistic period; a koinon certainly existed in Rome’s oldest province on Anatolian soil before the middle of the first century BC7 and another was probably organised in Bithynia not much later. Such a conclusion is supported by the evidence of letters from Brutus adressed to the Bithynians (the authenticity of which is debated)8 and, more reliably, by the wording of a passage in Dio (51.20.7) concerning events in 29 BC when the new princeps allowed the Bithynians to establish a cult for him in Nicomedia. For the Pontic part of the province there is no direct evidence before the Claudian period.9 But it has been widely accepted that an institution corresponding to the Bithynian koinon was created as early as the Pompeian province. A single inscription of Pompeiopolis, if authentic, may confirm this; it honours a Pontarch Apollodorus 5 6

Vitale 2012; see most recently Vitale 2014a. For details, see Marek 1993: 26–88; see also the briefer account in Marek 2003: 36–47; in the context of the provincialisation of Asia Minor, see Marek 2010: 364–8, 384–6, 402–40. 7 Sherk 1969: no. 52 Epistula magistratus Romani ad Miletos aliosque: πρὸς τὸ κοινὸν τῶν Ἑλλήνων γέγραφα. 8 Without conclusive results: Deininger 1966; edition: Torraca 1959. 9 There was a Σεβαστοῦ ἀρχιερεύς in Amastris: Marek 1993: cat. Amastris no. 1c. The argumenta e silentio presented by Dalaison 2014: 127, against the ‘theory of a western Pontic koinon preceding an eastern Pontic koinon’ have no value.

Epigraphy and the provincial organisation of Paphlagonian cities

309

and was reported in 1811 by Pascal Fourcade, but the stone has not been observed since.10 A cornerstone in the gradual Augustan rearrangement was the incorporation into the Roman Empire of the province of Galatia in 25 BC. Together with this, a new provincial koinon emerged in Ankyra. The temple was finished before Augustus died and the oldest priest list inscribed commences in 5/4 BC.11 A few years later, not far away in the north, another initiative of veneration of the princeps is recorded by an inscription of exceptional historic significance. This is the famous imperial oath of Gangra in Paphlagonia, dated to 6 March 3 BC (the stele was found in Vezirköprü).12 It is necessary to go into some detail, particularly of the initial and final paragraphs (lines 1–7 and 36–42), both of which I reproduce here. Ἀπὸ αὐτοκράτορος Καίσ[αρος] θεοῦ υἱοῦ Σεβαστοῦ ὑπατεύ[σαντος τὸ] δωδέκατον ἔτους τρίτου, π[ροτέραι] νωνῶν Μαρτίων, ἐν Γάνγροις EN.Λ[---------- ὅρ-] 5 κος ὁ τελεσθ[εὶς ὑ]πὸ τῶ[ν] κατοικ[ούντων Πα-] φλαγονία[ν καὶ τῶν πραγ]ματευομ[ένων πα-] ρ’ αὐτοῖς Ῥ[ωμαίων. vacat ? ] Following the text of the oath (lines 8–35), the inscription concludes: Κατὰ τὰ αὐτὰ ὤμοσαν καὶ οἱ E[------------] πάντες ἐν τοῖς κατὰ ΤΑCΥΓ[---------Σε-] βαστήοις παρὰ τοῖς βωμοῖ[ς τοῦ Σεβαστοῦ.] Ὁμοίως τε Φαζιμωνεῖται οἱ [τὴν νῦν Νεάπο-] 40 λιν λεγομένην κατοικοῦν[τες ὤμοσαν σύμ-] παντες ἐν Σεβαστήωι παρὰ τ[ῶι βωμῶι τοῦ] Σεβαστοῦ. An eye-catching parallel to the Bithynian example of 26 years earlier is the reference to the participation of both the indigenous population (l. 5–7) (οἱ κατοικοῦντες Παφλαγονίαν; cf. Dio 51.20.7: ξένοι, Ἕλληνάς σφας ἐπικαλέσας [sc. Caesar]) and the Romans residing in the country (οἱ παρ᾿ αὐτοῖς πραγματευόμενοι Ῥωμαῖοι; cf. Dio l.c.: οἱ Ῥωμαῖοι οἱ παρ᾿ αὐτοῖς ἐνοικοῦντες).13 As to where the Paphlagonian oath was actually sworn, the text distinguishes three groups of people at different locations: the first in lines 4–5, the second in line 36 (κατὰ τὰ αὐτὰ ὤμοσαν) and the third in lines 39–40 (ὁμοίως τε ὤμοσαν); the latter two lines are outdented to the left. A closer look at line 4 reveals a lacuna at the right edge after ἐν Γάνγροις and a following sequence of letters ΕΝ.Λ, the paragraph being continued on the next line by ὅρ-]|κος ὁ 10 11 12

Fourcade 1811; Marek 1993: 69–71. Mitchell and French 2012: 138–50. Ed. pr.: Franz Cumont 1900; 1901; SP no. 66; Herrmann 1968: 96–9. A revised edition is presented by Sørensen forthcoming, to whom I owe a debt for a stimulating discussion of the text. A good photograph is printed in Bekker-Nielsen 2014: 67 fig. 3. 13 At this time in Bithynia, the Romans were permitted to venerate only the goddess Roma and the Divus Iulius.

310

Christian Marek

τελεσθ[εὶς] ὑπὸ. Restorations suggested by scholars so far have in common that a further, more specific location is added after ‘in Gangra’.14 But that is not compelling. What can be read as ἐν must not necessarily be interpreted as a preposition but may be part of a compound verb. One may consider as a possibility: Ἐν Γάνγροις ἐν[γ]λ[υφθεὶς ὅρ-]|κος ὁ τελεσθ[εὶς] ὑπὸ τῶ[ν] κατοικ[ούντων Πα-]|φλαγονία[ν, ‘Oath inscribed in Gangra that was sworn by the population of Paphlagonia etc’.15 The precise meaning of the second passage is marred by damage to the stone’s right edge, affecting the ends of lines 36 and 37. If the end of 36 reads οἱ ἐ[ν ---, then οne group or several groups of Paphlagonians are referred to who dwelled or assembled at a location determined by ἐν.... and who swore the oath there at sites called Sebasteia that are particularly distinguished (ἐν τοῖς κατὰ ΤΑCYΓ[--------- Σε-]|βαστήοις). The restorations by Franz Cumont and Theodore Reinach have been widely accepted:16 Κατὰ τὰ αὐτὰ ὤμοσαν καὶ οἱ ἐν [τῆι χώραι]| πάντες ἐν τοῖς κατὰ τὰς ὑ[παρχίας Σε-]|βαστήοις παρὰ τοῖς βωμοῖ[ς τοῦ Σεβαστοῦ], ‘According to the same terms the rural population swore the oath at the altars of Augustus in the Sebasteia of the subdistricts’. To my mind, this suggestion cannot withstand closer examination. The word chora normally refers to a city territory. The meaning would compel us to understand that the rural population of Gangra was mentioned seperately after the citizenry; but the notion in line 5 (‘those who live in Paphlagonia’, κατοικοῦντες Παφλαγονίαν) does not exclude people living in the countryside. Even more problematic is the notion that Sebasteia are temples, sacred precincts (temene) or festivals (heortai). Considering the addition of παρὰ τοῖς βωμοῖ[ς τοῦ Σεβαστοῦ] (‘at the altars of Augustus’), we must understand the word to mean architecturally-designed places of worship. The idea that in the year 3 BC a number of such Sebasteia existed across the territory of Gangra is inconceivable. This sort of institution of the imperial cult in its earliest phase must be looked for exclusively in the urban centres. Finally, the restoration of ὑ[παρχίας, ‘subdistricts’, should be rejected. In Asia Minor, subdivisions of neither the kingdoms of Pergamon, Bithynia, Pontos or Kappadokia nor territories under Roman administration were called ὑπαρχίαι. This is a modern construct. I would therefore suggest a different, quite simple restoration at this point: Κατὰ τὰ αὐτὰ ὤμοσαν καὶ οἱ ἐν [ταῖς πόλεσι]| πάντες ἐν τοῖς κατὰ τὰ σύγ[γραφα Σε]|βαστήοις παρὰ τοῖς βωμοῖ[ς τοῦ Σεβαστοῦ.],17 ‘According to the same terms all those in the cities swore the oath at the altars of Augustus in the Sebasteia as specified in written agreements’. In what follows (lines 39–42), the cities and their Sebasteia are named, where the Paphlagonians swore the oath, in the first instance the Phazemonitai. But was their city the only one apart from Gangra mentioned by name in the inscription, and if so, why? Perhaps the Phazemonitai were selected exempli gratia with regard to the exhibition of a copy 14

Herrmann 1968: 97 n. 26 and 124 lists various restorations: F. Bücheler, F. Cumont (ἐν [κ]ά[στροις]); B. Haussoulier (ἐν [π]α[νηγύρει]); W. Dittenberger, H. Dessau (ἐν [τ]ἀ[γορᾶι]); H. Grégoire (ἐν [Κ]α[ισαρήωι]). 15 One may also consider ἐν[η]λ[ίκων] or ἐν[β]λ[ηθείς]. 16 Line 36: Cumont 1900; 1901; SP no. 66; line 37: Reinach communicated to Cumont according to SP 85. 17 Compare Ramsay 1895–1897: 144, 32: ὁ δῆμος ὁ Μο[σσυνέων καὶ] οἱ εἰς τὸ σύγγρ[αφον στή]λῃ καὶ στεφάν[ῳ ἐτίμησαν]; IG XII 9, 188: τὴν συμμαχί[ην κατὰ τ]ὰ σύνγραφα. A possible restoration would also be: κατὰ τὰς ὑπ[ογραφὰς.

Epigraphy and the provincial organisation of Paphlagonian cities

311

designed for ‘Neapolis’, where – at modern Vezirköprü – the stone was found. It is, however, open to question whether the text of the document ended with line 42 or whether, just below where the surface is now broken, it had further paragraphs mentioning other cities, for example ὁμοίως τε Πομπηιοπολῖται ὤμοσαν ἐν Σεβαστήωι κτλ. The ceremony attested by the document relates to the origin of imperial cult in a newlycreated province – the province of Paphlagonia. It is designated provincia (ἐπαρχία) in epigraphic records of senatorial carreers known as legati pro praetore of Galatia or Galatia/ Cappadocia; I may quote, for instance, Lucius Nonius Calpurnius Asprenas 69/70, governor of provinciae Galateae, Paphlagoniae, Pamphyliae, Pisidiae18 or Lucius Antistius Rusticus 91–94, governor of provinciarum Cappadociae, Galatiae, Ponti, Pisidiae, Paphlagoniae, Armeniae Minoris, Lycaoniae.19 Each of these conglomerations of provinces was subjected to the rule of a single governor whose dominion as a whole is represented by the name at the head of the queue. Thus, Galatia or Cappadocia is distinguished as Statthalterprovinz, a term coined by Marco Vitale.20 What happened in Paphlagonia may be seen as analogous to the introduction of the imperial cult in Ankyra, not more than about 100km away, a few years earlier. The institutional framework for its organisation on a regular basis should – as in Ankyra – be sought in a provincial koinon; but which one is the koinon the Paphlagonian cities – under the rule of the governor in Ankyra – were members of by that time? Until recently, a majority of scholars has adhered to the opinion that the koinon of Pontus et Bithynia, once established under Pompey the Great, persisted unchanged far beyond the breakup of that province and deep into the Imperial period of Anatolia.21 The preeminent reference frequently referred to by scholars is Jürgen Deininger’s book Die Provinziallandtage der Römischen Kaiserzeit (1965). Stephen Mitchell, in a paper published in 2002, traces a ‘Pontic identity’ that was rooted in the awareness of a persistent unity and survived the successive steps of provincial reorganisation: ‘The evidence for the use of the ethnic [‘pontikos’] needs to be combined with the extremely complicated problems concerning the extent and nature of the Pontic koinon, the origin of which can probably be traced back to the creation of Pompeius’ province. It can be used as an argument for the view … that there was only a single Pontic koinon in Asia Minor’.22 This position is generally supported by Xavier Loriot, but modified by the claim that a koinon of Pontus was not established as early as Pompey’s province, indeed not sooner than the Flavian period, but covered, however, the territorial extension of that former Pompeian province and included the cities of Paphlagonia – as concluded by Deininger. Accordingly, Loriot interprets the Pontarch Apollodorus, whose name Pascal Fourcade read in 18 19 20 21 22

AE 1952: 232; Rémy 1989: 142–5 no. 107. AE 1925: 126; Rémy 1989: 194–5 no. 159. Vitale 2012: 1–11, 15. First claimed by Georg Brandis 1896; elaborated by Deininger 1965. Mitchell 2002: 50. In inscriptions from the Imperial period Ποντικός does not exclusively refer to residents of the province in Asia Minor that had first been established by Pompey the Great. Compare IOSPE I2 79: ἐπεὶ Ὀρόντας Ὀλβιοπολείτας Αβαβου υἱὸς ἀνδρὸς οὐ μόνον τᾶς πατρίδος, ἀλλὰ καὶ σύμπαντος τοῦ Ποντικοῦ πρατιστεύσαντος ἔθνεος.

312

Christian Marek

Fig. 1: Coin. Bronze, 30mm. Obverse: Λ Π ΣΕΠ ΓΕΤΑΣ ΚΑΙΣ, bust of Geta, with paludamentum; reverse: above ΚΟΙΝ ΠΟΝ ΝΕΟΚΑΙ ΜΗΤΡΟ (ΝΠ und NN in Ligatur), below ΕΤ ΡΜR (MR in ligature), seated tyche holding cornucopia, flanked by two tychai on her right and three on her left side. Date: Geta 142=205/6.23 (After Imhoof-Blumer 1890: 577f. no. 55, pl. IV no. 16, ©Münzkabinett Winterthur). the inscription of Pompeiopolis, as evidence for membership of that city in an imperial koinon of Pontus.24 If the theory of a single Pontic koinon in imperial Asia Minor embracing cities of different provinces is correct – whether its origin goes back to Pompey or not – membership would apply to as many as 19 poleis. In addition to the four Paphlagonian ones, we can count six coastal cities of Pontus-Bithynia,25 three on the coast of Pontus Polemonianus26 and another six of interior Pontus east of Paphlagonia.27 Such was the size of the territory across which the inhabitants felt a common ‘Pontic’ identity, and as a consequence sent delegates to a federal assembly and celebrated a common festival (koinos agon) in Neokaisareia, today Niksar. The interior part of Pontus was organised as a seperate province Pontus Mediterraneus under Trajan or Hadrian.28 It contained six cities: Amaseia, Zela, Sebastopolis, Komana, Neokaisareia and Sebasteia. The metropolis Neokaisareia struck a series of coins under Septimius Severus and Geta (Fig. 1).29 These coins depict on their reverse six godesses – one seated surrounded by five standing upright. Above the image it is written KOIN(ON) 23

24 25 26 27 28 29

See Vitale 2012: 199 Anm. 1186: ‘(…) die von E. Babelon and Th. Reinach (Rec. gén.) als Zahlzeichen 146 gedeutete Buchstabenfolge ΡΜR – umgerechnet Jahr 209/210 n.Chr. ‒ ist mit guten Gründen als Schreibvariante von RMB (142) zu übernehmen; dazu Weiser 1989, 62 Anm. 53; ausführlich Leschhorn 1993, 137f.’. Loriot 2006a: 531: ‘J’admets donc, à titre d’ hypothèse vraisemblable, l’existence d’un koinon unique, crée à la fin du premier ou au début de second siècle, avec Néocésarée pour centre’. Heraclea, Tieion, Amastris, Abonouteichos/Ionopolis, Sinope and Amisos. Polemonion, Kerasous and Trapezous. Amaseia, Zela, Sebastopolis, Komana, Neokaisareia and Sebasteia. The sources are discussed in Marek 1993: 55–6. Septimius Severus: http://educators.mfa.org/ancient/coin-neocaesarea-bust-septimius-severus104977 (Museum of Fine Arts, Boston, 205/6).

Epigraphy and the provincial organisation of Paphlagonian cities

313

Fig. 2: Coin. Bronze, 22mm. Obvers: ΔΟΜΙΤΙΑΝΟΣ ΚΑΙΣΑΡ ΣΕΒΑΣΤΟΣ, head of Domitian. Reverse: KOI|NON ΠA|ΦΛAΓONI|AC four lines. Date: Domitian, 81–96. Type: RPC II 1629.1 (collection J. S. Wagner). (After http://www.cngcoins.com/Coin.aspx?CoinID= 127091 (CNG Auction 79, lot 580)). PONT(OU) NEOKAI(SAREIA) METRO(POLIS). For almost a century the female figures have

been interpreted by numismatists as tychai representing the six member-cities of a koinon of Pontus,30 but conclusions as to the composition and history of the koinon Pontou have not prevailed. The disturbing evidence of these coins did not go unnoticed by those who emphatically support the theory of a single Pontic koinon. But attempts to rebut it have a sense of desparation. Thus, Barbara Burrell, in her 2004 book Neokoroi. Greek Cities and Roman Emperors, makes the point: But the principles of numismatic interpretation should caution us: in a small space, numbers of figures, like columns on a temple front, may be abbreviated. If there were as many as fifteen in the Pontic koinon, how would you show them all on a coin only a few centimeters in diameter? The coin, therefore, only indicates that there were at least six cities in the koinon, to which Neokaisareia belonged, but does not rule out more.31

Loriot agrees: ‘Quant à déduire des répresentations monétaires le nombre exact des cités membres du koinon, c’est attendre de l’iconographie numismatique une précision qu’elle ne aurait garantir’.32 One wonders what sort of abbreviation the artist who drafted the design had in mind when he – instead of just a single or a couple – plumped for six figures (including one partly covered by the seated figure) and, by chance, matched the number of cities of that particular provincia. Another coin persistently excluded from the discussion of provincial koina was struck under Domitian and bears the embossed inscription: ΚΟΙ|ΝΟΝ ΠΑ|ΦΛΑΓΟΝΙ|ΑC (Fig. 2).33 Long after its publication, it should have been acknowledged by scholarship34 that the koinon attested here must be considered the same as that which was organised by the cities 30 31 32 33 34

As the first, see Chapot 1923: 93–107, esp. 100–1. Burrell 2004: 206. Loriot 2006a: 530. RPC 2: 236 no. 1629. Contra Deininger 1965: 69 n. 2.

314

Christian Marek

of provincia Paphlagonia after the creation of that province 5 BC and which took the initiative to foster the imperial cult and introduce the imperial oath at Gangra two years later. Taking this for granted, one must concede the formation of new koina in northern Anatolia joining and partly replacing the one created under Pompey, and – most importantly – adapting themselves to the new provinces, as in the case of Paphlagonia. Historians are particularly confused by the phenomenon of three koina bearing the same name, κοινὸν Πόντου: one outside Asia Minor consisting of the hexapolis/pentapolis on the western shore of the Black Sea;35 a second comprising the Pontic cities of Pontus-Bithynia; and a third – first attested under Hadrian36 – organised by the poleis of Pontus Mediterraneus. The latter is symbolically described on the coins of its metropolis by the six tychai. Objections may arise concerning the meaning of κοινὸν Παφλαγονίας on the coin issued under Domitian. Deininger dismisses it as a term denoting some obscure regional association rather than a provincial koinon. However, it was Deininger who argued for a close association of the provincial koina in the east with the widely attested chief magistrates who, according to the jurist Modestinus (Dig. 27.1.6.14), exercised Ἔθνους ἱεροσύνη οἷον Ἀσιαρχία, Βιθυναρχία, Καππαδοκαρχία, παρέχει ἀλειτουργησίαν ἐπιτρόπων, τοῦτ᾿ ἔστιν ἕως ἂν ἄρχῃ, ‘some function as highpriest of the ethnic unity such as Asiarchia, Bithynarchia, Cappadocarchia, that granted exemption from liturgies for the period in office’.37 Evidence for a like magistracy in Paphlagonia would prove a provincial koinon. In 2006, when excavations at Pompeiopolis began, fragments of a pillar-shaped monument inscribed on three sides were discovered as stray finds.38 An edition and commentary is currently being prepared for the corpus.39 The inscription on the front (Fig. 3) promulgates the honouring of a distinguished Pompeiopolitan named Markos or Markios who functioned as Παφλαγονιάρχης. As author of the honorary decree, a body is recognised by its title which can be fully restored as:

5

[ἀγα]θ[ῇ τύχῃ] [ἡ ἱε]ρὰ μουσ[ικὴ Σευη-] [ρι]ανὴ Ἀλεξα[νδρι-] ανὴ περιπολιστ[ικὴ] μεγάλη σύνοδο[ς]

Two sides of this block as well as the sides of an appurtenant fragment are inscribed with lists of expenditures undertaken in the course of a public festival including an agon. In consideration of the activity of the world association of competitors, there can be no doubt that the games referred to cannot be classified as local, but certainly must be understood as the provincial agon, the κοινὸς Παφλαγονίας – hitherto unattested. As far as the Imperial period is concerned, for the time it was set up, the new inscription definitely refutes the membership of the Paphlagonian cities in a Pontic koinon. As a compelling conclusion, the dating of Fourcade’ s Pontarch Apollodoros in Pompeiopolis must be kept out of the 35 36 37 38 39

For the literature on this koinon, see Loriot 2006a: 530 n. 55. Macridy 1905: 163 no. 2. Deininger 1965: 41–50; see now Vitale 2014a. I owe thanks to the Pompeiopolis Project for allowing me to publish this inscription. For a preliminary comment on this inscription, see Marek 2011: 191–3, fig. 2a–b.

Epigraphy and the provincial organisation of Paphlagonian cities

315

Fig. 3: Fragment of an honorary inscription from Pompeiopolis (after Marek 2011: 193, pl. 2, fig. 2a). Imperial period; it is most likely assumed to be pre-Augustan, but, in any case, dates prior to the Flavian κοινὸν Παφλαγονίας. The dating of our honorary inscription depends on the title of the world association, composed of a specific sequence of elements. As can be observed by the study of inscriptions and papyri of the later Roman Empire,40 this sequence was gradually complemented by the name of the emperor as an inherent part of the title – at first of the deceased ruler only after consecration but later of the current ruler. Accordingly, in the second line of the inscription the name of Severus Alexander must be restored. Part of his name, however, is written into an erasure clearly visible on the surface of the stone, which can only be interpreted as the damnatio memoriae of his predecessor Elagabalus after 222.41 The koinon of the province of Paphlagonia, therefore, still existed in the late Severan period; it celebrated a provincial agon under the presidency of the Paphlagoniarches. Neither the agon nor the chief magistracy of that koinon are known from any previous source. Until now, we know of Asiarchs, Bithyniarchs, Pontarchs, Galatarchs, Lykiarchs, Pamphyliarchs, Kilikiarchs and Armeniarchs, and, outside Asia Minor, Thrakarchs, Kretarchs, Syriarchs and Phoinikarchs. The close link between koina presided over by these officials and eparchiai in Asia Minor and the Levant is again confirmed by the case of Paphlagonia. 40 41

Frisch 1986. Cf. Marek 2011: 191.

316

Christian Marek

2. New evidence from milestones for the province of Paflagonia I come to the second, briefer part of my discussion. ‘Epigraphic habits’ flourished in Pompeiopolis as in Neoklaudiopolis within about the same span of time in which both cities emitted coins. Soon after the inscription honouring the Paphlagoniarches was set up, provincial coinage in the east ceased altogether42 and provincial administration in Anatolia underwent a fundamental change. The transformations in our region can be traced by the study of epigraphic evidence related to the titles of provincial governors, whom we know of almost exclusively from an abundant series of milestones stretching from the second quarter of the third to the second half of the fourth century. It is significant for the crisis in the East that governors’ names, previously rather rare, show up frequently on these stones.43 For the first time under Severus Alexander Pontus Mediterraneus was no longer ruled by senatorial legates of Cappadocia but by equestrian procurators bearing the official title praesides. The earliest attested is Quintus Faltonius Restitutianus. Various lists have been compiled in the scholarly literature, all of them now outdated.44 At present we know the names of 20 praesides from 78 milestones. Before the epochal changes under the administrations of Diocletian and Constantine starting in about 297, only three of these milestones are from Paphlagonia, all of them from the territory of Neoklaudiopolis: two with the name of Marcus Iunius Valerius Nepotianus (ca 250 under Decius) and one with the name of Aelius Decrianus (ca 251–3). The latter’s title is not preserved on the stone, whereas on a milestone found in the village of Arıca Nepotianus bears a title conspicuously abbreviated as LEG PR P P; this has been interpreted by David French to read legatus praesides provinciae Ponti.45 On the other milestone, found at Ortaklar, the Nepotianus is titled praeses. A distinctive feature of this inscription is that Nepotianus is said to preside over a conglomeration of provinces that can be read as provinciae Galatiae Ponti Λ[....]OI. Pontus must be understood as Pontus Mediterraneus. For the mutilated sequence of letters Λ....OI that follows the name a precise restoration is not readily available. Apart from the find-spot in Paphlagonia the only criterium accessible remains the description of John Arthur Ruskin Munro, the first editor in 1900: ‘Possibly we missed a P at the end of Ponti and might read [P]a[flag]o[niae]’.46 From the period after 297 six Paphlagonian milestones are preserved. These have been found in the territories of Gangra (1), Neoklaudiopolis (3) and Pompeiopolis (3); the latter three are unpublished. Three of them bear the name of Aelius Priscianus, a governor most active in the field of road repairs and construction between 293 and 305. Surprisingly, on stones from Pompeiopolis, Neoklaudiopolis and Gangra – the latter found at Yukarı Yanlar in the immediate vicinity of Çankırı – his title is given as praeses prov(inciae) Pont(i) merely, whereas Paphlagonia, which one might expect here, is absent. This is stressed by 42 43

See Sauer this volume. We know of about 1,100 milestones from Asia Minor, including a few more than 70 from Paphlagonia. 44 French 1986: 81 lists ten between 236 and 305; Loriot 2006b: 424 lists 13 between 279 and 363; see the list below. 45 French 2013: 48 B. 46 Munro 1900: 161 no. 3. This restoration is rejected by Christol and Loriot 1986: 33, but see below.

Epigraphy and the provincial organisation of Paphlagonian cities

317

Loriot as proof of his theory that the province of Paphlagonia had previously ceased to exist and had been subsumed within a considerably enlarged province of Pontus until it later re-emerged in a different shape in the Laterculus Veronensis (compiled between 312 and 324)47 and the catalogue of the participants of the council of Nikaia in 325.48 But the theory is polluted by the evidence of two milestones from the territory of Sinope, dated ca 305–6, that spell out the title of the governor Aurelius Hierax as praeses pr(ovinciae) Ponti Paflag(oniae) and Paflaugon(iae) respectively. This evidence is dismissed as an exception by Mitchell (‘The title … should be seen as no more than a temporary alternative designation of a province that certainly included Paphlagonian territory’),49 and Loriot agrees: ‘Mitchell a très justement souligné la singularité de cette titulature’.50 Many of the governors’ titles on the milestones are extremely abbreviated in ways which do not always allow for certainty of their full unabbreviated form. This, for example, applies particularly to the formula V P PR P P recurring on the stones of Sinope and persistently interpreted by French as v(iro) p(erfectissimo) pr(aeside) p(rovinciae) P(onti).51 His interpretation attaches the territory of Sinope to an equestrian province of Pontus in the later Roman Empire. But which one? Neither the Pontus Mediterraneus nor the Pontus Polemonianus are plausible candidates. According to a well-known and debated passage in the geography of Ptolemy (5.4.1–3), Galatia extended to the coast of the Black Sea from Kytoros in the west to Amisos in the east, which presupposes a preceding detatchment from the province of Pontus-Bithynia between 159 and 169. Scholars refer to that phenomenon as ‘la Galatie maritime’.52 In consideration of the structure of the Statthalterprovinz Galatia, the eparchia affected by that extension must have been Paphlagonia, as correctly pointed out by Vitale.53 For that reason, it is by no means unfounded to consider the P on the milestones of Sinope to represent P(aflagoniae). The theory of Paphlagonia being subsumed within Pontus, however, must be relinquished, since new finds of milestones corroberate its existence before the Verona list. They name two different praesides Paflagoniae within a span of 15 months (1 May 305 to 25 July 306).54 One of the stones, still unpublished,55 is from Pompeiopolis (Fig. 4). It is remarkable that Paflagonia is abbreviated AFL, the initial P missing – just as on the milestone of Nepotianus in Neoklaudiopolis copied by Munro.56 The praeses, Lucilius Crispus, is hitherto unattested in the region, but known from other parts of Asia Minor: in Silifke as praeses provinciae Isauriae, dated to the year immediately following his term of office in Paphlagonia, and in Ankara as a(gens) v(ices) praeff(ectorum) praet(orio).57 47 48 49 50 51 52 53 54

Jones 1954; 1964: 42–3; Barnes 1982: 201–8. Belke 1996: 65. Mitchell 1993: 2.159. Loriot 2006b: 412. French 1981: 153 no. 6; 1986: 71–2. Cumont 1903; see Marek 1993: 84; Vitale 2012: 218–28. Vitale 2012: 228. Second Tetrarchy: Constantius I and Galerius Augusti, and Severus and Maximinus Caesares, see the list below. 55 I wish to thank the Pompeiopolis Excavation Project for allowing me to publish this milestone. 56 See above, n. 46. 57 PIR2 L 382; PLRE I 233; AE 1978: 815.

318

Christian Marek

Fig. 4: Squeeze of an unpublished milestone from Pompeiopolis (©Pompeiopolis Project). A milestone recently published by French adds another testimony to Aurelius Hierax, governor under the Second Tetrarchy.58 Unlike the inscriptions of Sinope and Neoklaudiopolis with his name, this one does not assign two provinces to him – Pontus and Paphlagonia – but Paphlagonia alone. This is all the more remarkable, given that the stone was found in Ezinepazar, a village east of Amasya, i.e. in Pontic territory. An unpublished milestone of Hierax from a village near Zile may possibly have the same title – praeses Pa(flagoniae) – if the letters PA are correctly read. In any case, it becomes evident that with Hierax we do not encounter any singularité. Concurrently, governors on milestones in the west, as in Gangra and Pompeiopolis, are designated praesides Ponti, just as governors in the east, as in Amaseia and perhaps Zela, are designated praesides Paflagoniae. The explanation is simple: the two provinces – Paphlagonia and Pontus – were ruled by one governor, and those drafting the formulae of the milestones did not bother to write out the names of both or abbreviate the name of one province as pars pro toto. Where the stone was set up did not make a difference as to which was given preference. Shortly before 309 a territory (the boundaries of which I shall not discuss here) adopted the new name ‘Pontus of Zeus’ – Diospontus. It occurs on milestones of Amaseia and Neoklaudiopolis,59 but is not used after 325, when Constantine the Great created the province of Helenopontus, named after his mother. In conclusion, the province of Paphlagonia was created under Augustus in 5 BC and expanded to the Black Sea coast under Antoninus Pius or Marcus Aurelius. It was ruled by the legati in Ankyra during the first and second centuries, until, late in the Severan period, 58 59

French 2012b: 146. Loriot 2006b: 413.

Epigraphy and the provincial organisation of Paphlagonian cities

319

equestrian praesides were appointed over both Pontus and Paphlagonia. The essential substructure of a province, according to the Roman formula provinciae, consisted of cities, civitates. Those in Paphlagonia followed the example of their neighbours in Asia, Bithynia, Pontus and Galatia by constituting a permanent provincial koinon for purposes of selfrepresentation and the organisation of imperial cult. The evidence is scarce, but sufficient to prove the existence of the koinon Paphlagoniae most likely from the time of the oath of Gangra, certainly from the rule of Domitian to the rule of Severus Alexander. Whether the koinon was still active under the praesides into the Diocletian and Constantinian epoch we do not know. For modern scholarship, understanding basic administrative structures in northern Anatolia under the principate and the late Empire may depend on the reading of a single letter cut into stone. A meticulous reconstruction of the history of these territories provides the framework for many aspects of civic life. It is to be hoped that agreement upon that framework is reached by international scholarship and concepts that have been clearly proven erroneous be abandoned. I am grateful to Ursula Kunnert, Kristina Winther-Jacobsen, Lâtife Summerer and Marco Vitale for helpful comments. Christian Marek Universität Zürich Historisches Seminar Karl Schmid Strasse 4, 8006 Zürich, Switzerland [email protected]

Bibliography Abbreviations AE: Année épigraphique LPGN: Fraser, P. M. et al. (eds) 1987–. A Lexicon of Greek Personal Names. Oxford. PIR2: Prosopographia Imperii Romani, 2nd edition. PLRE I: Jones, A. H. M, J. R. Martindale and J. Morris 1971. The Prosopography of the Later Roman Empire 1. Cambridge. Rec. gén. I 12: Waddington, W. H., E. Babelon and Th. Reinach 1925. Recueil général des monnaies grecques d’Asie mineure I 12: Pont et Paphlagonie, 2nd ed. Paris. RPC: Burnett, A., M. Amandry and P. P. Rippolès 1992–1999. Roman Provincial Coinage 1–2. London. SP: Anderson J. G. C., F. Cumont and H. Grégoire 1910. Studia Pontica III. Recueil des inscriptions grecques et latines du Pont et de l’Arménie. Bruxelles.

Literature Barnes, T. D. 1982. The New Empire of Diocletian and Constantine. Cambridge (Mass.) and London. Bekker-Nielsen, T. 2014. ‘To Be or not to Be Paphlagonian? A Question of Identity’, in T. BekkerNielsen (ed.), Space, Place and Identity in Northern Anatolia (Geographica Historica 29). Stuttgart, 63–74. Belke, K. 1996. Paphlagonien und Honorias (Tabula Imperii Byzantini 9). Vienna. Brandis, C. G. 1896. ‘Studien zur römischen Verwaltungsgeschichte: I. Wann wurde Pontus et Bithynia kaiserliche Provinz?’ Hermes 31, 161–73. Burrell, B. 2004. Neokoroi. Greek Cities and Roman Emperors. Leiden. Chapot, V. 1923. ‘Frontière nord de la Galatie et les Koina de Pont’, in W. H. Buckler and W. M. Calder (eds), Anatolian Studies Presented to William Mitchell Ramsay. Manchester, 93–107.

320

Christian Marek

Christol, M. and X. Loriot 1986. Le Pontus et ses gouverneurs dans le second tiers du IIIe siècle (Centre Jean Palerne, Mémoirs VII, Recherches épigraphiques: Documents relatifs a l’ histoire des institutions et de l’administration de l’Empire Romain). Saint Étienne, 13–40. Cumont, F. 1900. ‘Inscription grecque de Vézir-Keupru dans l’ancienne Paphlagonie (Asie Mineure)’, Académie des inscriptions et belles-lettres. Comptes rendus des séances de l’année 44, 687–91. Cumont, F. 1901. ‘Un serment de fidélité à l’empereur Auguste’, Revue des études grecques 14, 26–45. Cumont, F. 1903. ‘La Galatie maritime de Ptolemée’, Revue des études grecques 16, 25–7. Dalaison, J. 2014. ‘Civic Pride and Local Identities. The Pontic Cities and their Coinage in the Roman Period’, in T. Bekker-Nielsen (ed.), Space, Place and Identity in Northern Anatolia (Geographica Historica 29). Stuttgart, 125–55. Deininger, J. 1965. Die Provinziallandtage der römischen Kaiserzeit von Augustus bis zum Ende des 3. Jh. n.Chr. Munich. Deininger, J. 1966. ‘Brutus und die Bithynier’, Rheinisches Museum für Philologie 109, 356–72. Fourcade, P. 1811. ‘Mémoire sur Pompeiopolis ou Tasch-Kouprou, avec quelques remarques sur Tovata ou Voyavat’, Annales des Voyages, de la géographie et de l’ histoire 14 (publiées par M. MalteBrun, Paris), 30–58. French, D. 1981. ‘Milestones of Pontus, Galatia, Phrygia, and Lycia’, Zeitschrift für Papyrologie und Epigraphik 43, 149–74. French, D. 1986. ‘Recent Epigraphic Research in Pontus’, Epigraphica Anatolica 8, 71–82. French, D. 2012a. Roman Roads and Milestones of Asia Minor, 3.2. Galatia (BIAA Electronic Monograph). Ankara. French, D. 2012b. Roman Roads and Milestones of Asia Minor, 3.3. Cappadocia (BIAA Electronic Monograph). Ankara. French, D. 2013. Roman Roads and Milestones of Asia Minor 3.4. Pontus et Bithynia (BIAA Electronic Monograph). Ankara. Frisch, P. 1986. Zehn agonistische Papyri. Opladen. Gnoli, T. 2000. ‘Il Ponto e la Bitinia’, in A. M. Braschi and G. Salmeri (eds), Strabone e l’Asia Minore, Incontri Perugini di Storia della Storiographia Antica e sul Mondo Antico 25–28 Maggio 1997 a Perugia. Perugia, 543–64. Herrmann, P. 1968. Der römische Kaisereid. Untersuchungen zu seiner Herkunft und Entwicklung. Göttingen. 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. Århus, 15–30. Imhoof-Blumer, F. 1890. Griechische Münzen. Neue Beiträge und Untersuchungen (Abhandlungen der Bayerischen Akademie der Wissenschaften, Philosophisch-Philologische Klasse 8, Abhandlung 18). Munich. Jones, A. H. M. 1954. ‘The Date and Value of the Verona List’, Journal of Roman Studies 44, 21–29. Jones, A. H. M. 1964. The Later Roman Empire. Oxford. Kaygusuz, İ. 1983. ‘Zwei Inschriften aus Ilgaz (Olgassys) und Kimiatene’, Epigraphica Anatolica 1, 59– 61. Loriot, X. 1974. ‘Le statut administratif du Pont sous le Haut-Empire’, Bulletin de la Société Nationale des Antiquaires de France 109–11. Loriot, X. 1976. ‘Le Pont au IIIe siècle de notre ère’, Bulletin de la Société Nationale des Antiquaires de France, 44–61. Loriot, X. 2006a. ‘Le culte impérial dans le Pont sous le Haut-Empire’, in A. Vigourt et al. (eds), Pouvoir et réligion dans le monde Romain, en hommage à Jean-Pierre Martin. Paris, 521–40. Loriot, X. 2006b. ‘Les gouverneurs du Pont de Gallien à Julien (260–363)’, in S. Demougin et al. (eds), H.-G. Pflaum. Un historien du XXe siècle, Geneva, 399–425. Marek, C. 1993. Stadt, Ära und Territorium in Pontus-Bithynia und Nord-Galatia. Tübingen. Marek, C. 2003. Pontus et Bithynia. Die römischen Provinzen im Norden Kleinasiens. Mainz.

Epigraphy and the provincial organisation of Paphlagonian cities

321

Marek, C. 2010. Geschichte Kleinasiens in der Antike, 2nd ed. Munich. Marek, C. 2011. ‘Zur Epigraphik von Pompeiopolis: Eine Zwischenbilanz’, in L. Summerer (ed.), Pompeiopolis I. Eine Zwischenbilanz aus der Metropole Paphlagoniens nach fünf Kampagnen (2006– 2010). Langenweißbach, 189–91. Mitchell, S. 1993. Anatolia. Land, Men, and Gods in Asia Minor. Oxford. Mitchell, S. 2002. ‘In Search of the Pontic Community in Antiquity’, in A. K. Bowman et al. (eds), Representations of Empire. Rome and the Mediterranean World (Proceedings of the British Academy 114). Oxford, 35–64. Mitchell, S. and D. French 2012. The Greek and Latin Inscriptions of Ankara (Ancyra) 1. From Augustus to the End of the Third Century AD. München. Munro, J. A. R. 1900. ‘Some Pontic Milestones’, Journal of Hellenic Studies 20, 159–66. Ramsay, W. M. 1895–1897. The Cities and Bishoprics of Phrygia 1, 1–2. Oxford. Rémy, B. 1989. Les carrières sénatoriales dans les provinces romaines d’Anatolie en Haut Empire (31 av.J.-C. – 284 ap.J.-C.). Istanbul. Sherk, R. K. 1969. Roman Documents from the Greek East. Baltimore. Sørensen, S. L. forthcoming. ‘A Re-examination of the Imperial Oath from Vezirköprü’, Philia 1. Summerer, L. (ed.) 2011. Pompeiopolis I. Eine Zwischenbilanz aus der Metropole Paphlagoniens nach fünf Kampagnen (2006–2010). Langenweißbach. Torraca, L. 1959. Marco Giunio Bruto, Epistole Greche. Neapel. Vitale, M. 2012. Eparchie und Koinon in Kleinasien von der ausgehenden Republik bis ins 3. Jh. n.Chr. (Asia Minor Studien 67). Bonn. Vitale, M. 2014a. ‘“Pontic” Communities under Roman Rule. Polis Self-Representation, Provincialisation and the koina “of Pontus”’, in T. Bekker-Nielsen, Space, Place and Identity in Northern Anatolia (Geographica Historica 29). Stuttgart, 49–61. Vitale, M. 2014b. ‘“Priest” – “Eparchy-arch” – “Speaker of the ethnos”: The Areas of Responsibility of the Highest Officials of the Eastern Provincial Imperial Cult’, Mnemosyne (forthcoming 2015: [prepublication: booksandjournals.brillonline.com/content/1568525x/advance]). Weiser, W. 1989. ‘Römische Stadtmünzen aus Bithynia et Pontus, Addenda und Corrigenda zum Recueil Général’, SNR 68, 47–74.

Table Praesides Ponti Paphlagoniae For previous lists see: Loriot 1974: 109–111; 1976: 44–61; Christol and Loriot 1986: 13–40; French 1986: 81; Mitchell 1993: vol. 2, 158–163; Loriot 2006b: 424. 222–235

Severus Alexander

230–235

Quintus Faltonius Restitutianus * Amaseia – Amasya Yüzevler Mahalle (French 1986: 75 no. 2) PROC ET PRAE[S] PROV PONT DEVOTUS NUMIN[I M]AIESTATI [QUE] EORUM

235–236

Claudianus * Amaseia – Hamdi, Çorum (French 2012b: 043=235 o. 236) V E PRAES PROVINC

– Uygur (French 2012b: 145B=235) PRA|ESID PROVINC| κα

322

Christian Marek

235–238

Maximinus

236–238

P. Aelius Vibianus * Amaseia – Boğazkaya (French 2012b: 029C=July 236–April 238) V P PRAESID PROVINC M P κς – Boğazkaya (French 2012b: 29A=July 236–April 238) PR PROVINC PONT vac.

238 238–244

Gordian I, Gordian II, Balbinus, Pupienus Gordian III

238–241

Claudius (?) Marcellus * Sebastopolis – Sulusaray (French 2012b: 061) without title

244–249

Philippus Arabs

247–249

Claudius Aurelius Tiberius * Zela – Çamsaray (French 2012b: 138=July/August 247–September/October 249) V E PRAES P P | vac. | κε

249–251

Decius

250

M. Iunius Valerius Nepotianus * Neoklaudiopolis – Arıca (French 2013: 48B) LEG PR PR P PASS MIL

French: leg(atum) pr(aesidem) p(rovinciae) P(onti) – Ortaklar (French 2013: 56; Christol and Loriot 1986: 33) PRAESIDEM PRO|VINCIAE GALAT|IAE POIITI Λ[…]OI certainly: A[FLAG]ON * Amaseia – Çorum (French 2012b: 037A) [per] VM[.... c. 8 ....] uncertain! 251–253

Trebonianus, Volusianus

251–253

Aelius Decrianus * Neoklaudiopolis – Çayırözü (French 2013: 58) RESTITUERUNT PER AELIUM DICR[

Epigraphy and the provincial organisation of Paphlagonian cities

253 253–260

Aemilianus Valerianus

249–255

C. Iulius Senecio

323

* Ankyra (Mitchell and French 2012: no. 53-4): honorary inscriptions 53: PROC PROV GALAT ITEM VICE PRAESIDIS EIUSD PROV ET PONTI 54: πράξαντα κὲ τὰ τῆς ἡγεμονίας μέρη ἐνδόξως ‘The two provinces specified in 53, Galatia and Pontus, were temporarily united under a single governor for a short period from 250.’ (Christol and Loriot 1986: 36) 253–268

Gallienus

260–272

Invasions of Sassanids, Goths, Palmyreneans

268–270 270 270–275 275–276 276 276–282

Claudius II Gothicus Quintillus Aurelian Tacitus Florianus Probus

279–280

Aelius Quintianus * Sinope – Kerim (French 2013: 15 [1]) V P PR P P

– Çeçe (French 2013: 13A [2]) PR P P

– Chalabde (French 2013: 07A. B) V P PR PR P

French: v(iro) p(erfectissimo) pr(aeside) pr(ovinciae) P(onti) V P PR P P

* Amaseia – Mus. Çorum (French 2012b: 37C) AB AMASIA MIL P ς΄ TRANSITA A SOLO CONPOSITA CURANTE AEL QUINTIANO V P PR P P (hapax) * Zela – Tokat Mus. (French 2012b: 054H) V P PR P P

282–283

Carus

282–283

Claudius Longinus * Sinope – Tıngır (French 2013: 21B. D. E) V P PRAES V P PRAES P

– Çeçe (French 2013: 13A [3]) CURANTE CL LONGI[NO] V P P P vac. ΛN

French: p(raeside) p(rovinciae) P(onti) vac. λν΄

324

Christian Marek

* Neoklaudiopolis – Yenice (French 2013: 32) V P [..] P

*Amaseia – Yerkozlu (French 2012b: 016) V P PRAES P P vac. μ γι΄ – Toklucak (French 2012b: 025) [CL. L]ONG[inus v. p.] PR[ – Çitli (French 2012b: 033B) PRAES vac. λθ΄ * Turhal (French 2012b: 055B=CIL III 1418418) reading and restoration very uncertain * Sebastopolis – Özükavak/Yozgat (French 2012b: 139) title not preserved 283–284 283–285 284–305

Numerianus Carinus Diocletianus, Maximianus

c. 297

New provincial order of Diocletian

293–305

Aurelius Valentinianus * Neoklaudiopolis – Vezirköprü (French 2013: 49D) V P P[....]| S

French: v p p[r pr P]| S

* Amaseia – Çorum Boğazkaya/Bebük (French 2012b: 029D) reading very uncertain 293–305

Aurelius Priscianus (PLRE I, 728) Loriot 2006b, 409–411: Sinope 6, Amaseia 7 (10), Neoklaudiopolis 1, Zela 4, Komana 1, Neocaesarea 3, Gangra 1. * Gangra – Yukarı Yanlar (French 2012a: 28) V P PRAES PRO[..]ONT D N NQ EORUM

* Pompeiopolis – Mus. Kastamonu (unpublished) PRAES vac.? PRO PON[T] * Neoklaudiopolis – Gömlekhisar (French 2013: 47) V P PRAES P P D[.]MQ EORUM

French: praes(es) p(rovinciae) P(onti)

Epigraphy and the provincial organisation of Paphlagonian cities

* Sinope – Sinop Mus. (French 2013: 08A) V P PR PR P

– Çalboğaz, Gürcü Mah., now in the museum (French 2013: 11B) title restored – Tıngır (French 2013: 21A) PR PR P D N M Q. EORUM

– Gerze (French 2013: 14) V P PR PR P

– Göllu (French 2013: 18B [1]) V [P] PR PR P

– Ilıcaköy (French 2013: 44B) V P PR P P

– Bağlıca (French 2013: 44C) V P PR P P

– Çeçe (French 2013: 13A [4]) V P PR PR P

– Kerim (French 2013: 15 [3]) V P PR PR P

* Amaseia – Boğa (French 2012b: 022) V P PRAES P P P D N MQ EORUM referring to PPP French: ‘sic’;

compare below: Hierax, Neoklaudiopolis – Çiğdemlik (French 2012b: 018) V P PRAES P PONT D N MQ EORUM

– Kapıkaya (French 2012b: 147D) name and title restored – Boğazkaya (French 2012b: 029B) [..] PRAES PRO PONTI D N MQ EORUM

– Çitli (French 2012b: 033A. B) V P PRAES PRO PO[N]TI D N MQ EORUM V P PRAES PR PONTI D N MQ EORUM

– Elvancelebi (French 2012b: 035) V P [.....]V PONT D N MQ E[ORUM]

– Akyazı (French 2012b: 143) V P PRAES PR P D N MQ EORUM

– Uygur (French 2012b: 145B) V P PRAES[..]PR P D N MQ EORUM

– Ilıca (French 2012b: 150b) V P PRAES PR P D [n mq] EORUM * Zela – Zile Kale (French 2012b: 056B. C) [..] PRAES PROV PONT [d n] MQ EORUM V P PRAES P[r Pont] D N MQ EO[rum] – Zile Belediye garden (French 2012b: 056D) V P PRAES PONTI D N MQ EORUM

– Güzelbeyli (French 2012b: 140A) V P PRAES PROV PONT D N MQ EORUM

– Yıldıztepe (French 2012b: 141) V P PRAES PROV PONT D N MQ EORUM

325

326

Christian Marek

* Komana – Tokat (French 2012b: 054A) V P PR PR [........]NTI D N M [...]RUM

French: pr(aeses) pr[ovinc Po]nti d n m [q eo]rum * Neokaisareia – Çalkara (French 2012b: 013) PRISCIANUS [....]PR P D N M[q eorum] – Çakırsu (French 2012b: 015) V P PR P[...]NT D N MQ EORUM

– Akça (French 2012b: 012) V P PRAES PR P D N MQ EORUM

293–306 293–311

Constantius I. Galerius

305–306

Aurelius Hierax * Sinope – Museum (French 2013: 08C) V P P[r pr Po]NT PAFLAG(oniae) – Tıngır (French 2013: 21C [3]) V P PR R R ONTI PAFLAUGON

French:

v(ir) p(erfectissimus) pr(aeses) (p)r(ovinciae) R-onti Paflaugon(iae)

* Neoklaudiopolis – Aşağı Narlı (French 2013: 45A) V P PRAES P P P

French: praes(es) p(rovinciae) P(onti) P(aflagoniae); cf. above Priscianus * Amaseia – Ezinepazar (French 2012b: 146=1.5.305–25.7.306) V P RRAES PROV PAFLAG D N (vac.) MQ EORUM * Zela – Derebaşı (unpublished) PR PR PA

305–306

Lucilius Crispus * Pompeiopolis – Kese (Kilise) Köy (ined.) V P PR PR AFL

306–337

Constantinus I

before 310

Ataxius/Ataxus (?) * Djedar, Mauretania: Fragment of a cursus-inscription (AE 2004: 1887) IATAXI PR PROVINCIA PON

Epigraphy and the provincial organisation of Paphlagonian cities

309–310

Flavius Severus * Amaseia – Mecitözü Boğazkaya (French 2012b: 029C [2]) V P P P DIOSP

317–324

Valerius Chrysaorius (cf. PLRE I p. 204) * Neoklaudiopolis – Vezirköprü (French 2013: 49D=CIL III 1418431) V P PRAES PROVINC DIOSPONT

333–335

Flavius Iulius Leontius * Sinope (French 2013: 08 A [2], 08B [2], 08C [2], 21A [4]) PRE PRO ELENOPONT PR PR HELENOPONT PR R [HELE]NOPONTI

title not preserved * Amaseia – Boğazkaya (French 2012b: 029B [2]) V P PRAES PROVINC HELENOP D N MQ EORUM

– Toklucak (French 2012b: 025 [2]) [V] P PRAES PROV HELENOPO[NTI]

– Köse eyüp (French 2012b: 031 [3]; CIL III 1418438) [V] P PRAES PROVINC [H]ELENOP D N MQ vac. EORUM – Çitli (French 2012b: 033A. B [3]) LEONTIUS [---]RAE[--------] N MQ [----] V P PRAES PROV HELENOP DN MQ EORUM

* Zela – Çeltek V [P] PRAES. PR. HELE[NOP]

– Kurupınar (French 2012b: 057; CIL III 1418417) V P PRAES PR HELENOP D N MQ EORUM

– Alayurt (French 2012b: 142) V P PRAES PR HELENOP vac. D N MQ EO[RUM] 337–340 337–350 337–361

Constantinus II Constans I Constantius II

337–340

Flavius Achillius * Sinope – Mus. (French 2013: 08B [3]) V P PRAES PROVINC HELENOP

– Göllü (French 2013: 18A [4], 18B [2]) V[P] PRAES PROVINC HELENOP [----]ELENOP[-----]

– Tıngır (French 2013: 21C [6], 21F) V P PRAES PROVINC HELENOP V P PRAES PR EL[----]

327

328

Christian Marek

* Amisos – Kavak (French 2013: 29, cf. Loriot 2006b: 415) no title * Zela – Çeltek V P PRAES [PR]OV HELENOP DNMEORUM

360–363

Julian

361–363

Flavius Domitius Hilarius * Sinope – Museum (French 2013: 8C [3]) V C PRAESES PROV HELENOPONTI

– Çeçe (French 2013: 12A. 13A [5]) V C PRAESES PROV HELENOPONTI V C PRAESES PROV HELENOPONTI

362–363

Candidianus Basil. Caes. ep. 3; Gregory of Nazianz ep. 10

fourth c.

Iovinus SP no. 26–7

Vera Sauer

The coinage of Neoklaudiopolis and Pompeiopolis against the background of the minting practices of Pontic and Paphlagonian cities during the Roman Imperial period

1. Introduction Until the second half of the third century AD many cities in the eastern parts of the Roman Empire minted coins. This article intends to provide a comparative insight into the habits of minting in the two regions of Pontus and Paphlagonia.1 The focus will be on two cities: the Pontic Neoklaudiopolis2 and the Paphlagonian Pompeiopolis. First, a general introduction to the coinages of both cities is provided. Secondly, the minting practices of Pontic and Paphlagonian cities are compared. In a third part of the article, it is argued that none of the identified practices can be defined as typically Pontic or typically Paphlagonian; rather, the numismatic habits of these cities are characterised by what might be called a conventional individuality.

2. The coinage of Neoklaudiopolis and Pompeiopolis Currently, we know of 60 preserved coins struck by Neoklaudiopolis and 18 struck by Pompeiopolis.3 According to the evidence of these coins, the period of minting at Neoklaudiopolis was the second century, actually the period between 109/104 and 200/1. Pompeiopolis 1

However, in some respects this insight can only be offered in the form of a preliminary look, as comprehensive up-to-date coin corpora are not available for all the cities in question. In relation to the Pontic cities, corpora of Amaseia (Dalaison 2008), Kerasous (Wojan 2003), Komana (Amandry and Rémy 1999), Neokaisareia (Çzmel 2006), Neoklaudiopolis (Dalaison and Delrieux 2014; supplemented by: http://www.aeruginis.de/aeruginis/Sammlung/Detail.php? page=889), Nikopolis (Dalaison 2007), Sebasteia (Schultz 1989; supplemented by: Weiser 1998: 283; RPC online temp. no. 5454–5), Sebastopolis (Amandry and Rémy 1998), Trapezous (Wojan 2006) and Zela (Dalaison et al. 2009) have been established, but the corpus of Amisos is missing. In relation to the Paphlagonian cities, corpora of Pompeiopolis (Dalaison 2010) and Germanikopolis (Bricault and Delrieux 2014) have been accomplished and the corpus of Abonouteichos is in print (Julie Dalaison, Fabrice Delrieux and Marie-Claire Ferriès). In preparation are the corpora for the coins of Amastris (Frédérique Brenier) and Colonia Iulia Felix (Pierre Manisse). 2 The question as to whether Neoklaudiopolis was Pontic or Paphlagonian is complex: BekkerNielsen 2014: 63–74; Dalaison and Delrieux 2014: 161–2; Marek this volume: n 1. 3 Dalaison 2010 no. 15 is, in all probability, not a coin but a medallion: Dalaison 2010: 61. 4 See Sauer forthcoming: appendix 4.

330

Vera Sauer

started minting later in the reign of Antoninus Pius (138–61), and, similarly to Neoklaudiopolis, minted no later than the era of Septimius Severus (193–211). Neither of the cities minted continuously throughout these periods; in fact, minting took place in only a few years. In the case of Neoklaudiopolis eight emissions can be distinguished; in the case of Pompeiopolis three can be identified. The emissions of Neoklaudiopolis date to the reign of Trajan (109/10, 111/2, 112/3), to an unkown year during the reign of Hadrian and to the reigns of Antoninus Pius (155/6, 156/7), Commodus (185/6) and Septimius Severus (200/ 1); those of Pompeiopolis date to unknown years during the reigns of Antoninus Pius, Marcus Aurelius and Septimius Severus. Both cities minted bronze coins – i.e. small change. The denominations of the coins were, in all probability, assarion, diassarion and tetrassarion; Neoklaudiopolis also struck half-assaria.5 In terms of value, a tetrassarion equalled a Roman sestertius, which is a quarter of a denarius. The legends of the coins are in Greek. On the obverse is a portrait of the emperor or of a member of his family, with his/her name and titles. The reverse legend informs us who issued the coin. In the case of Neoklaudiopolis, this was predominately the Neoklaudiopoleitai, i.e. the citizens of Neoklaudiopolis. Sometimes, however, the name of the city itself is given.6 Furthermore, the reverse legends of the coins of Neoklaudiopolis mention the year according to the city era in which the coins were minted.7 In contrast, the year is never given on the coins of Pompeiopolis; the name of the city is always provided,8 never the name of the citizens. In addition, the legend for Pompeiopolis includes the title of metropolis. Occasionally, the legend is even more specific, stating that Pompeiopolis was metropolis Paphlagonias.9 This latter title, also attested on inscriptions, in all probability refers to the rank of the city within the eparchia of Paphlagonia and perhaps also to its rank within the koinon of Paphlagonia.10 The reverse images of the coins of Pompeiopolis, without exception, depict gods:11 Asklepios, Demeter, Dionysos, Herakles, Kybele, Nemesis and the Tyche of the city. The reverse images of the coins of Neoklaudiopolis likewise favour gods: Aphrodite, Asklepios, Athena, Athena Nikephoros, Hygieia, Nike, Serapis, the Tyche of the city, a goddess with the appearance of Artemis Ephesia and a goddess who cannot be identified. Furthermore, the coins of Neoklaudiopolis depict at least one temple. However, the reverse images of the 5 6

7

8 9 10 11

Dalaison 2010: 61; Dalaison and Delrieux 2014: 184. Dalaison and Delrieux 2014: nos 5–7. The names of the citizens and of the city are given in the genitive; there is only one exception where the city name is given in the nominative: Dalaison and Delrieux 2014: no. 5. Only two dies are exempt from this rule: Dalaison and Delrieux 2014 nos 35–6. In the case of four coins, the year cannot be made out: Dalaison and Delrieux 2014: nos 8, 23a, 33; http:// www.aeruginis.de/aeruginis/Sammlung/Detail.php?page=889. The first year of the city era corresponds to the year 6/5 BC – just as in Pompeiopolis, Germanikopolis and Hadrianopolis: Leschhorn 1993: 170–5, 481–4; Marek 1993: 130–1. In the nominative. Dalaison 2010: nos 1, 3–4, 12–4, 16. Cf. Dalaison 2010: 52–3; Sauer forthcoming. As far as we know, Neoklaudiopolis did not have a title. The medallion (cf. n. 3), however, depicts two emperors on horseback crushing an enemy.

The coinage of Neoklaudiopolis and Pompeiopolis

331

coins of this city are not restricted to the cultic sphere; they also refer to the emperor and military issues, such as the depiction of Trajan being crowned by Tyche, of spears and a shield with the title Dakikos of the emperor Trajan written on it or of an eagle between signa militaria with a crown in its beak.12 To a certain degree, it is possible to extrapolate the volume of coinage from both cities: the value of the coinage of Neoklaudiopolis totalled about 60,625 denarii; that of Pompeiopolis about 40,000.13 Although the economic effect of the minting in the actual year it was emitted may have been noticeable, considered over the entire period of coin emissions, the economic effect of the coinage of both cities was obviously not that great. In Neoklaudiopolis the volume of coinage reached its peak during the reign of Septimius Severus and in Pompeiopolis it reached its peak in the era of Marcus Aurelius.14 However, in Neoklaudiopolis the volume of coin production during the reign of Antoninus Pius was not that much smaller than the volume produced during the reign of Septimius Severus. Even though, in the case of Neoklaudiopolis and Pompeiopolis, the exact find-spot of only one single coin is known (a coin of Pompeiopolis which was found in 2006),15 in all probability, we can be fairly certain of the currency pattern due to what is known about the currency patterns of the coinage of other cities in the region (see below). In short, the coins were in currency in the issuing city first and foremost. To a minor degree they might have spread to neighbouring cities as well, but, generally, they did not travel far. Clearly, the coinages of Neoklaudiopolis and Pompeiopolis have much in common, but there are also differences: the number and date of the emissions; the manner of naming the issuing politeia; the mentioning or not of the year of issue in the coin legend; and, last but not least, the preferences with regards to the reverse images. Such dissimilarities should neither be ignored nor overemphasised, as they fall within the range of variation seen in the numismatic habits of various cities – irrespective of them being Pontic or Paphlagonian – as will be demonstrated in the following section.

3. Minting practices of Pontic and Paphlagonian cities When comparing the minting practices of the Pontic and Paphlagonian cities, at first sight many common features can be detected. The striking of coins was an area-wide phenomenon in both regions (Fig. 1); only three of 19 cities did not mint.16 With the exception of Amisos during the reign of Hadrian, the cities minted only bronze coins.17 Apart from Colonia Iulia Felix, which was a colony of Roman citizens settled in the territory of Sinope, Greek was used for the coin legends. The visual language of the coin images did not differ

12 13

14 15 16 17

Dalaison and Delrieux 2014: nos 6–7, resp. no. 1, resp. nos 24–7, 32. Dalaison 2010: 61–3; Dalaison and Delrieux 2014: 184–6. To put these numbers in perspective, the annual pay of a Roman legionnaire in 197 was 600 denarii; thus 60,625 denarii correspond to the pay of 101 legionnaires and 40,000 denarii to the pay of 66 legionnaires. Dalaison 2010: 63–4; cf. Dalaison and Delrieux 2014: tables on 184–5. Dalaison 2010: no. 8b. Hadrianopolis in Paphlagonia and Polemonion and Magnopolis in Pontus. Concerning the silver coinage of Amisos, see Nordbø 1988.

332

Vera Sauer

Fig. 1: Map of the Pontic and Paphlagonian cities. Black dot=cities which did not mint. (Cartography by R. Szydlak). significantly either from city to city (cf. Figs 2 and 3) or from Roman imperial coinage.18 Except for the first emissions of Amastris and Amisos, the same system of denominations was most probably in use in all the cities.19 Except for Colonia Iulia Felix and for Amisos in the third century, the cities struck coins only sporadically (Table). At least for the Pontic inland cities, for Kerasous and Trapezous, and for Pompeiopolis and Germanikopolis,20 it

18

The impression given by Rec. gén. and RPC is that this is true also for those cities for which comprehensive coin corpora have not yet been established. On the Pontic inland cities and on Pompeiopolis and Germanikopolis, see in particular Çzmel 2006: 110–2, 117–30; Dalaison 2007: 230–3; 2008: 149–52, 159–80; 2010: 65–71; Dalaison and Delrieux 2014: 186–91; Sauer forthcoming; on Kerasous and Trapezous, see Wojan 2003; 2006. 19 Dalaison 2008: 33 (Amaseia); Bricault and Delrieux 2014: 61 (Germanikopolis); Wojan 2003: 282 (Kerasous); Amandry and Rémy 1999: 56 (Komana); Dalaison 2010: 61 (Pompeiopolis); Dalaison and Delrieux 2014: 184 (Neoklaudiopolis); Dalaison 2007: 224 (Nikopolis); Amandry and Rémy 1998: 47 (Sebastopolis); Wojan 2006: 218 (Trapezous); Dalaison et al. 2009: 43 (Zela); on Neokaisareia, see Çzmel 2006: 145–53; Dalaison 2008: 35–8. This is presumably true for Sebasteia also, but due to the small number of coins preserved verification is not possible. On Amastris, Amisos and Colonia Iulia Felix, see RPC 1: 30–7, 340, 356, 359–60; RPC 2: 24, 113–5. On the denominations of the silver coinage of Amisos (drachmai, didrachmai and tridrachmai of a local standard), see Nordbø 1988: 170–1, 176–7. 20 Due to the lack of comprehensive up-to-date corpora, no reliable conclusions can be drawn concerning Amastris, Abonouteichos, Colonia Iulia Felix and Amisos.

The coinage of Neoklaudiopolis and Pompeiopolis

333

Fig. 2: Coin of Neoklaudiopolis. Obverse: bust of Commodus. Reverse: Tyche of the city with rudder and cornucopia. Dalaison and Delrieux 2014: no. 20b (after http://www.cngcoins.com/Coin.aspx?CoinID=101262).

Fig. 3: Coin of Pompeiopolis. Obverse: bust of Lucius Verus. Reverse: Tyche of the city with rudder and cornucopia; on top of the cornucopia is an animal, possibly a ram. Dalaison 2010: no. 12a (after http://www. cngcoins.com/Coin.aspx?CoinID=146236). appears that the cities minted coins in relatively small quantities only21 and primarily for local use.22 With the exception of Pompeiopolis, Kerasous and probably Trapezous, the volume of coinage reached its peak during the reign of Septimius Severus, as is generally the

21

The scope of volumes ranges between approximatly 40,000 denarii (Pompeiopolis: Dalaison 2010: 61–3) and 745,000 denarii (Amaseia: Dalaison 2008: 39–41): Sauer forthcoming with further references. With all due caution, it can still be concluded, in light of Rec. gén. and RPC, that the volume of the coinage of Abonouteichos is likely to have fallen within the same range; those of Amastris and, in particularly, those of Amisos and Colonia Iulia Felix, in all probability, were considerably larger. 22 For Inner Pontic cities and for Pompeiopolis and Germanikopolis, see Sauer forthcoming with further references; on Kerasous, see Wojan 2003: 285. In the case of Trapezous there are reports of numerous coins having been found in Kolchis; however, Franck Wojan was not able to verify this claim despite his best efforts: Wojan 2006: 224–5.

334

Vera Sauer

case in the eastern Mediterranean.23 None of the actual triggers for an emission, i.e. the momentum which brought a city to mint, is known.24 Despite all these common features of the minting practices of the Pontic and Paphlagonian cities, plenty of differences can be detected, as the following examples demonstrate.25 There are considerable discrepancies with respect to the dates the cities started minting, the dates they stopped minting and the frequency and the concrete dates of minting (Table).26 The range of starting years falls between 71/70 BC and 208/9 AD.27 The range of years in which last emissions were made falls between 161/2 and 265/6.28 The numbers of emissions vary between two and at least 68.29 Most cities mention the date of issue on every single coin they minted; some cities passed on doing so occasionally.30 Abonouteichos and Pompeiopolis, and Amastris from the time of Domitian, however, desisted from dating without exception. In addition to the common coins which depict the emperor or a member of his family on the obverse, Amastris and Amisos, and, to a far lesser extent, Komana and Neokaisareia struck so-called pseudoautonomous coins which instead depict 23

24 25 26 27

28

29

30

On Pompeiopolis and Neoklaudiopolis, see n. 14; on Amaseia, see Dalaison 2008: 41–2, 183; on Germanikopolis, see Bricault and Delrieux 2014: 51; on Komana, see Amandry and Rémy 1999: 56–8; on Neokaisareia, see Çzmel 2006: 142–4; on Sebastopolis, see Amandry and Rémy 1998: 47–8; on Zela, see Dalaison et al. 2009: 46; on Kerasous and Trapezous, see Sauer forthcoming. In the case of Kerasous the peak falls during the reign of Elagabalus and for Trapezous it falls during the reign of Severus Alexander (although the volume of coinage during the reign of Septimius Severus is only a little smaller). Nikopolis and Sebasteia stopped minting earlier. On the general trend, see Dalaison 2008: 183 n. 17. On the Pontic inland cities and Pompeiopolis, see Sauer forthcoming; on Kerasous and Trapezous, see Wojan 2003; 2006; on Germanikopolis, see Bricault and Delrieux 2014: 71–76. For a broader compilation and a more detailed discussion, see Sauer forthcoming. Of course, these statements are valid only on the basis of published material, which itself depends on preservation. Amastris issued coins from 71/70 BC, Amisos from 61–59 BC, Colonia Iulia Felix from 45 BC, Komana and, in all probability, Amaseia (Dalaison 2008: 24) from the reign of Tiberius, Neokaisareia from the reign of Domitian, Neoklaudiopolis, Nikopolis, Sebasteia, Sebastopolis, Zela, Abonouteichos and Trapezous from the reign of Trajan, Kerasous from the last year of the reign of Hadrian, Pompeiopolis from the reign of Antoninus Pius and Germanikopolis, arguably, from 208/ 9: Bricault and Delrieux 2014: 48, 51. The last emissions of Nikopolis and Sebasteia date to 161/2; those of Komana, Neoklaudiopolis, Zela, Pompeiopolis and Germanikopolis date to the era of Septimius Severus; those of Amaseia and Kerasous date to Severus Alexander; those of Trapezous date to Philip the Arab; those of Abonouteichos date to Trebonianus Gallus; those of Neokaisareia, Sebastopolis, Amastris, Amisos and Colonia Iulia Felix date to Gallienus. Definitely attested are two emissions for Sebasteia and Germanikopolis, three for Pompeiopolis, Nikopolis, Zela and Sebastopolis, seven for Abonouteichos, eight for Neoklaudiopolis, nine for Komana, 11 for Kerasous, 17 for Amaseia and Neokaisareia, 19 for Trapezous, 22 for Amastris, 47 for Amisos and 68 for Colonia Iulia Felix. The table below gives an impression of the variability of the concrete dates of minting. In the case of a few individual years (most obviously 161/2), it must be admitted that several cities decided to mint. Mostly, however, there is variation (observe the years of minting during the reign of Trajan for example). This is true also in later times – even at the very end, during the reign of Gallienus. In rare cases, Amaseia (Dalaison 2008: nos 1–4) and Neoklaudiopolis (Dalaison and Delrieux 2014: no. 35f.); more often, Amisos, Colonia Iulia Felix and Germanikopolis: for references, see Sauer forthcoming. Some emissions may not be dated or even some dies within the same emission.

The coinage of Neoklaudiopolis and Pompeiopolis

335

gods or heroes of the past.31 The subject matter of the reverse images of the coins of all the cities under consideration can be grouped into three major subjects: gods, cult and myth; the emperor and military; and the city itself. The first subject dominates. However, the ratios of the three subjects differ significantly from city to city, and Abonouteichos, Komana, Pompeiopolis, Nikopolis, Sebastopolis and Trapezous only used images of the first category. The diversity of images within each subject category differs significantly. The coins of Komana, without exception, depict Ma (and objects linked to her cult); in contrast, the coins of Amisos depict at least 24 different gods. Some cities focused exclusively on their most important god. In addition to Komana focusing on her goddess Ma, Sebastopolis focused on the main god of the city – Herakles. Other cities did not do so: Zela did not focus on Anaïtis, Neokaisareia did not focus on Men Pharnakou, Abonouteichos did not focus on Glykon and Pompeiopolis, where, according to inscriptions, Zeus-Helios-Serapis was favoured,32 did not, according to the currently available data, depict this god(s) on coins at all. Finally, some cities sit between these two extremes: the coins of Amaseia often but not exclusively refer to Zeus Stratios and those of Trapezous to Mithras. Obviously all these differences have nothing to do with being, or claiming to be (more or less) either Pontic or Paphlagonian. There are both Pontic and Paphlagonian cities which started minting relatively early (Amisos, Komana, Amaseia – Amastris, Colonia Iulia Felix) or late (Kerasous – Pompeiopolis, Germanikopolis), which stopped minting relatively early (Nikopolis, Sebasteia, Komana, Neoklaudiopolis, Zela – Pompeiopolis, Germanikopolis) or late (Amisos, Neokaisareia, Sebastopolis – Amastris, Colonia Iulia Felix), which emitted coins rarely (Sebasteia, Pompeiopolis, Nikopolis, Zela, Sebastopolis – Germanikopolis) or often (Amisos – Colonia Iulia Felix), which struck pseudoautonomous coins (Amisos, Komana, Neokaisareia – Amastris) or did not do so and which chose different topics for the reverse images or desisted from doing so (Komana, Nikopolis, Sebastopolis, Trapezous – Abonouteichos, Pompeiopolis). Furthermore, in specific instances where a feature is exclusively seen in either Pontic or Paphlagonian cities – only the Pontic cities of Komana and Sebastopolis focused on a favourite god, only the Paphlagonian cities of Abonouteichos, Pompeiopolis and Amastris constantly omit the date of issue – not all Pontic/Paphlagonian cities adopt the feature and so break any supposed Pontic/Paphlagonian pattern.

4. The nature of the coinages of Pontic and Paphlagonian cities: conventional individuality In the light of the features noted above, it is argued that there is nothing which in Roman times could be called a Pontic or Paphlagonian manner of minting. It appears that regional affiliation – affiliation to a geographically-defined landscape, to a region with a shared history or to a region united by administration (as an eparchia, a province or a koinon) – was not crucial for the numismatic habits of the cities in Pontus and Paphlagonia. To prevent a misunderstanding, it should be noted that I am not arguing that conditions determined by 31

Amastris: Homer, Dionysos, Helios, Kore, Zeus Strategos and Tyche; Amisos: Antinous, Aphrodite, Apollo, Athena, Dionysos, Hermes, Tyche and Roma; Komana: Ma; Neokaisareia: Zeus: for references, see Sauer forthcoming. 32 Marek 2011: 190.

336

Vera Sauer

regional affiliation did not affect the coinage of the cities at all. Rather, I argue that this was only one factor – and not the most important – within a set of various factors of supraregional and, very importantly, local character which conditioned the numismatic habits of each city, and that this set of factors did not produce coinages which can be called typically Pontic and Paphlagonian. On the contrary, the effect of the interaction of all these factors was that the coinage of each of the Pontic and Paphlagonian cities was simultaneously individual and conventional – as paradoxical as that may initially sound. The coinages range within a ‘framework of the usual’; to a large extent they look uniform (and deeply affected by imperial coinage), but, on the other hand, individual facets can be detected in the coinage of every city, and these individual facets are not marginal.33 The impression given by the evidence is that each city wished, by means of coinage, to strengthen its identity, i.e. its identity as a polis – a polis, however, that was part of the Roman Empire. Being part of a region, howsoever defined, was not that important. Whenever cities did refer to regional matters – for example Pompeiopolis claiming to be metropolis of Paphlagonia – the reasoning behind this is, in a way, self-serving, as the city did not promote the regional entity in question but its individual status, its own prestige, within the entity.34 This article has aimed to present the perspectives opened up by a study of the coinage of the two regions of Paphlagonia and Pontus. Other categories of sources – inscriptions, archaeological artefacts and structures – may offer a different picture of the existence or non-existence of regionally-defined habits.35 Vera Sauer University of Southern Denmark Mühlweg 6, 72414 Rangendingen, Germany [email protected] 33

It should be noted that in the coinage of the Pontic and Paphlagonian cities images are attested which have parallels in neither Roman nor regional coinage: see Sauer forthcoming. This is no surprise, as some of the coins depict unique features such as specific gods, city views and buildings. However, it is remarkable that in some cases cities might have used coin images whose individuality cannot be explained by the singularity of the thing depicted. For instance, on coins of Neoklaudiopolis the depiction of spears and a shield with Dakikos inscribed and the depiction of Tyche crowning the emperor are, as far as they can be judged, without parallels: Dalaison and Delrieux 2014: 188–9. This applies also to the depiction of Marcus Aurelius and Lucius Verus on a medallion issued by Pompeiopolis. The image of an emperor with a lance, on horseback, crushing an enemy is, of course, very common on Roman coinage; the depiction of two emperors doing this side by side is, however, unique: Dalaison 2010: 69–70. Clearly, Pompeiopolis wished to stress the strong collaboration of Marcus Aurelius and Lucius Verus – they are also both depicted on the obverse – in a very striking manner. Furthermore, the success of this collaboration is confirmed explicitly in the legend: Neike Sebaston, ‘the victory of the emperors’. 34 On ‘coinage and identity and prestige’, see Harl 1987; Howgego et al. 2005; Dalaison 2014; Sauer 2014. 35 See also Marek; Musso this volume. Just to mention two examples: in Paphlagonian inscriptions of Roman Imperial times relatively often a strong impetus to prove Greek literacy can be detected (Mitchell 2010: 106–9), an impetus which in Pontic inscriptions, to my knowledge, is far from being so common; in Paphlagonia, amongst other types, grave markers in the shape of an inscribed column were widespread (Marek 2003: 132–7, esp. 137; LaflI and Christof 2010: 166–7; LaflI 2012: 268–9) but they were not so in Pontus (cf. Koch this volume).

The coinage of Neoklaudiopolis and Pompeiopolis

337

Bibliography Abbreviations Rec. gén.: Waddington, W. H., E. Babelon and T. Reinach, Recueil général des monnaies grecques d’Asie Mineure. Paris, fascicles 1.1 (second edition, 1925); 1.2 (1908); 1.3 (1910); 1.4 (1912); reprinted New York 1976. RPC 1: Burnett, A., M. Amandry and R. P. Ripollès, Roman Provincial Coinage 1. London (first ed. 1992). RPC 2: Burnett, A., M. Amandry and R. P. Ripollès, Roman Provincial Coinage 2. London (first ed. 1999). RPC online: http://rpc.ashmus.ox.ac.uk/coins

Literature Amandry, M. and B. Rémy 1998. Pontica II: Les monnaies de l’atelier de Sebastopolis du Pont (Varia Anatolica 10). Paris. Amandry, M. and B. Rémy 1999. Comana du Pont sous l’Empire romain: Étude historique et corpus monétaire. Milan. Bekker-Nielsen, T. 2014. ‘To be or not to be Paphlagonian? A question of identity’, in BekkerNielsen (ed.), 63–74. Bekker-Nielsen, T. (ed.) 2014, Space, Place and Identity in Northern Anatolia (Geographica Historica 29). Stuttgart. Bricault, L. and Delrieux, F. 2014. Gangra-Germanicopolis de Paphlagonie. “Foyer des dieux“. Étude de numismatique et d’ histoire (Numismatica Anatolica 6). Bordeaux. Çzmel, Z. 2006. Le monnayage de Néocésarée et du Koinon du Pont (Glaux: collana di studi e ricerche di numismatica 17). Milan. Dalaison, J. 2007. ‘L’atelier monétaire de Nicopolis en Arménie Mineure’, in J. Dalaison (ed.), Espaces et pouvoirs dans l’antiquité de l’Anatolie à la Gaule: Hommages à B. Rémy. Grenoble, 203–37. Dalaison, J. 2008. L’atelier d’Amaseia du Pont: Recherches historiques et numismatiques (Numismatica Anatolica 2). Bordeaux. Dalaison, J. 2010. ‘L’atelier monétaire de Pompeiopolis en Paphlagonie’, in F. Delrieux and F. Kayser (eds), Des déserts d’Afrique au pays des Allobroges: Hommages offerts à F. Bertrandy. Chambéry, 45–81. Dalaison, J. 2014. ‘Civic pride and local identities: the Pontic cities and their coinage in the Roman period’, in Bekker-Nielsen (ed.), 125–55. Dalaison, J. and F. Delrieux 2014. ‘La cité de Néapolis-Néoclaudiopolis: Histoire et pratiques monétaires’, Anatolia Antiqua 12, 159–98. Dalaison, J., B. Rémy and M. Amandry 2009. Zéla sous l’empire romain: Études historiques et corpus monétaire (Numismatica Anatolica 4). Bordeaux. Harl, K. W. 1987. Civic Coins and Civic Politics in the Roman East A.D. 180–275. Berkeley. Howgego, C., V. Heuchert and A. Burnett (eds) 2005. Coinage and identity in the Roman Provinces. Oxford. Lafli, E. 2012. ‘Archäologische Evidenzen zum Weinanbau im südwestlichen Paphlagonien in römischer und frühbyzantinischer Zeit’ in E. Olshausen and V. Sauer (eds), Die Schätze der Erde – Natürliche Ressourcen in der antiken Welt: Stuttgarter Kolloquium zur Historischen Geographie des Altertums 10, 2008 (Geographica Historica 28). Stuttgart, 261–79. Lafli E. and E. Christof 2012. ‘Drei neu entdeckte Phallossteine aus der Chora von Hadrianopolis’, in: G. R. Tsetskhladze (ed.), The Black Sea, Paphlagonia, Pontus and Phrygia in Antiquity: Aspects of archaeology and ancient history (British Archaeological Report International Series 2432). Oxford, 163–9. Leschhorn, W. 1993. Antike Ären: Zeitrechnung, Politik und Geschichte im Schwarzmeerraum und in Kleinasien nördlich des Tauros (Historia Einzelschriften 81). Stuttgart. Marek, C. 1993. Stadt, Ära und Territorium in Pontus-Bithynia und Nord-Galatia (Istanbuler Forschungen 39). Tübingen.

338

Vera Sauer

Marek, C. 2003. Pontus et Bithynia: Die römischen Provinzen im Norden Kleinasiens. Mainz. Marek, C. 2011. ‘Zur Epigraphik von Pompeiopolis: Eine Zwischenbilanz’, in L. Summerer (ed.), Pompeiopolis I: Eine Zwischenbilanz aus der Metropole Paphlagoniens nach fünf Kampagnen (2006– 2010). Langenweißbach, 49–62. Mitchell, S. 2010. ‘The Ionians of Paphlagonia’, in T. Whitmarsh (ed.), Local knowledge and microidentities in the imperial Greek world (Greek culture in the Roman world). Cambridge, 86–110. Nordbø, J. H. 1988. ‘The Imperial Silver Coinage of Amisus 131/2–137/8 A.D.’, in AAVV, Studies in Ancient History and Numismatics presented to Rudi Thomsen, Aarhus, 166–78. Sauer, V. 2014. ‘Urban space: the evidence of coins’, in Bekker-Nielsen (ed.), 109–24. Sauer, V. forthcoming. ‘Konventionelle Individualität: Zur Münzprägung nordanatolischer Städte in der römischen Kaiserzeit’. Schultz, H.-D. 1989. ‘Megalopolis-Sebasteia’, in G. Le Rider et al. (eds), Kraay-Mørkholm Essays: Numismatic Studies in Memory of C. M. Kraay and O. Mørkholm (Numismatica Lovaniensia 10). Louvain-La-Neuve, 259–66. Weiser, W. 1998. ‘Namen römischer Statthalter auf den Münzen Kleinasiens: Corrigenda und Addenda zu Gerd Stumpfs Münzcorpus’, Zeitschrift für Papyrologie und Epigraphik 123, 275–90. Wojan, F. 2003. ‘Kérasonte du Pont sous l’Empire romain: Étude historique et corpus monétaire’, Revue numismatique 159, 257–90. Wojan, F. 2006. ‘Trapézonte du Pont sous l’Empire romain: Étude historique et corpus monétaire’, Revue numismatique 162, 181–229.

Table Years of issue For Amastris, Amisos, Colonia Iulia Felix and Abonouteichos the data are based on Rec. gén.; Leschhorn 1993; RPC 1; RPC 2; RPC online; Nordbø 1988. Data for the other cities are based on the corpora referred to in n. 1. + *

Coins with legends not mentioning the year of issue. Emissions for which the date of issue, originally given by the legend, can no longer be read on the preserved examples. (1) Most probably these coins were issued in the reign of Tiberius; however, it cannot be ruled out completely that they may in fact have been issued during the reigns of Augustus or Caligula: Dalaison 2008: 24. Italics: Silver coinage of Amisos. List of abbreviations: Ais: Amastris, Aos: Amisos, CIF: Colonia Iulia Felix, Ko: Komana, Aia: Amaseia, Nk: Neokaisareia, Nkp: Neoklaudiopolis, Np: Nikopolis, Sp: Sebastopolis, Tr: Trapezous, Ze: Zela, Se: Sebasteia, Ab: Abonouteichos, Ke: Kerasous, Po: Pompeiopolis, Ge: Germanikopolis. Ais Aos 71/0 61–59 70/69 56 62/1 61–59 59/8 52/1 46/5 40/39 39/8 32/1 29/8

CIF Ko 45 38 35 +

Aia

Nk

Nkp

Np

Sp

Tr

Ze

Se

Ab

Ke

Po

Ge

339

The coinage of Neoklaudiopolis and Pompeiopolis Ais Augustus

Tiberius Caligula Claudius

Nero

Vespasian Domitian + Nerva Trajan

+

Hadrian

Antoninus Pius

+

Marcus + Aurelius et sui, Lucius Verus Com+ modus

Aos 5/4 2/1 9/10 +

CIF Ko 27 26 24 23 11 10 9 7 4 BC 5 AD 6 14 28/9 18 (?) 35/6 19 36/7 37 37/8 38 39/40 46/7 53/4 41 42 52 54 56/7 58 59 68 + 69/70 73 75 77 * (?) 96 96/7 97 97/8 104 * 98/9 108 106/7 117 108/9 113/4 131/2 123 132/3 125 133/4 133 133/4 137 134/5 135/6 136/7 137/8 151/2 138 158/9 140 160/1 143 159

Aia

Nk

Nkp

Np

Sp

Tr

Ze

Se

Ab

Ke

Po

+(1)

78/9 92/3 92/3

112/3 100/1 109/10 104/5 106/7 113/4 113/4 114/5 + 101/2 111/2 112/3 112/3 * 135/6

*

153/4 154/5

155/6 156/7

163/4 162 + +

161/2 161/2 162/3 *

182/3 191/2

186/7 184/5 185/6 187/8 * 188/9

137/8

*

161/2

161/2 163/4 * 183/4 188/9

+

161/2 +

+ 138 141/2 144/5 or 147/8 145/6 151/2 161/2 +

184/5 186/7 *

Ge

340 Ais Septimius + Severus et sui

Caracalla

Macrinus Elagabalus Severus + Alexander Maximi- + nus Thrax

Balbinus Gordia- + nus III Philippus Arabs Traianus Decius Trebonia- + nus Gallus, Volusianus Aemilianus Valerianus, Gallienus Gallienus +

Vera Sauer Aos CIF Ko Aia 209/10 194 205/6 205/6 206/7 * 198 * 207 or 09 208 or 10 210 or 12 + 213/4 212 216/7 or 14 * 213 or 15 214 or 16 215 or 17 217/8 217 or 18 219/20 219 or 20 225/6 221/2 223 226/7 229/30 224 231/2 231/2 232/3 234/5 235 235/6 236/7 * + 237/8 240/1 238 * 239 241 242 245/6 244 + 249/50 249 250/1 +

Nk Nkp Np 204/5 200/1 205/6 * +

Sp Tr 205/6 194/5 198/9 204/5

218/9 219/20 224/5 225/6 226/7

226/7 234/5

241/2

238/9 242/3 243/4

248/9

244/5

255/6 262/3 263/4 264/5 265/6

+

+

+

252/3 254 255 259 260 265

Ab +

Ke

216/7

251/2

254/5 257/8 + +

Ze Se 205/6 206/7 *

263/4

218/9 219/20 225/6

Po +

Ge 208/9 209/10 +

Ancient personal names

Achilles 184, 242–4 Aelius Decrianus 316 Aelius Priscianus 316 Alexander the Great 244 Alexander, the Prophet 184 Alypios the Stylite 121 Ambrosia 256 Amia 174 Anaïtis 335 Anastasius I 101 Antinous 335 Antistius Rusticus, Lucius 311 Antoninus Pius 195, 226, 318, 330–1, 334 Aphrodite 330, 335 Apollodoros 179 Apollodoros, the pontarch 308, 311, 314 Apollon 335 Apollonios Rhodios 33 Arcadius 265 Ariadne 244 Arrian 29, 33–4, 219, 244 Artemis Ephesia 330 Artemisia 226–7 Asklepiades II Charax 173 Asklepios 330 Athena (Minerva) 225, 330, 335 Athena Nikephoros 330 Augustus 17, 189, 309–10, 318, 338 Aurelius Hierax 317–8 Bellerophon 256 Brutus 308 Calliope 246 Caracalla 52, 242 Cassius Dio 308 Charon 203 Cicero 184 Commodus 330 Constans II 122 Constantine I 52, 316 Constantius I 317 Constantius II 52

Damostratos 184 Decius 316 Deiotaros Philadelphos 308 Demeter 330 Demetrios of Phaleron 184 Diocletian 52, 316, 319 Diodorus Siculus 244 Dionysos 244, 330, 335 Dioscurides 243 Domitian 195, 313–4, 319, 334 Doryphoros 187, 189 Elagabalus 315, 334 Faltonius Restitutianus, Quintus 316 Galerius 317 Gallienus 334 Ge 256 George of Amastris 123 Geta 312 Glykon 335 Gregory of Nysa 152 Hadrian 219, 225–6, 232, 235, 312, 331, 334 Helen 243 Heliodoros 174 Helios 335 Helios, father of Mageianos 174 Hephaestus 243 Hera of Kandara 307 Heraclius 129 Herakles 330, 335 Hermes 335 Hesiod 183–4 Hierokles 126 Homer 184, 186, 240, 242–3, 335 Honorius 233 Hygieia 330 Iunius Valerius Nepotianus, Marcus 316–7 John of Gaza 273 Julius Caesar, deified 309

342 Kallisthenes Hestiaiou 50 Kastor II 308 Kybele 330 Kyrittas 184 Kyros 186 Leo III 122 Lucilius Crispus 317 Lucius Verus 336 Lycurgus 256 Lysimachides 184 Ma 335 Mageianos 173 Mandnis 174 Manios 180 Marcus Aurelius 184, 318, 330–1, 336 Mark Antony 17, 183, 304, 308 Markos/Markios 314 Maximinus 101, 317 Mithras 335 Mithradates VI Eupator 144, 154, 181–2 Modestinus 314 Nemesis 330 Nonius Calpunius Asprenas Torquatus, Lucius 183, 311 Oceanus 257 Orentes 244 Papios 180 Parthenos 184 Patroclus 242 Philip the Arab 334 Philostratus 244 Pliny the Elder 31, 172, 179, 183 Pliny the Younger 226, 230 Plutarch 184, 244

Indices Pompey the Great 17, 61, 160, 182, 295, 304, 307– 8, 311, 314 Priam 184 Priscus 184 Ptolemy 182–3, 317 Pylaimenes 183 Remus 224 Roma 309, 335 Romulus 224 Septimius Severus 312, 330–1, 333–4 Serapis 330, 335 Severus II 317 Severus Alexander 198, 244, 315–6, 319, 334 Socrates 246 Strabo 31, 33, 51, 57, 172, 179–81, 215, 292, 307 Terentia Seleukis 174 Terentios Sekoundos 174 Theodore, a drougarios 122 Theodore Tiro 101 Theodosius II 233, 266 Theophrastos 31 Thetis 240, 242–3 Tiberius 334 Trajan 184, 226, 312, 330–1, 334 Trebonianus Gallus 334 Tryphe 267 Tyche 165, 250, 330–1, 335–6 Valentinian II 233 Vespasian 52, 182 Victory (Nike) 50, 223, 225, 330 Zeus 335 Zeus Ammon 224 Zeus Strategos 226, 232, 235, 335 Zeus Stratios 19, 135–145, 335

Modern personal names

Abadie-Reynal, C. 260 Abdallah, K. 257 Abdul-Massih, J. 260 Adam, J.-P. 87 Ahrweiler, H. 123 Ainsworth, W. F. 180, 231 Aitken, E. B. 244 Akkaya, M. 215, 245 Akok, M. 65 Akyüz, U. 215 Alatepeli, S. 249, 292 Alexandridis, A. 228 Alföldi-Rosenbaum, E. 271 Alkım, U. B. 215, 295 Allen, P. 275 Amandry, M. 329, 332, 334 Ambrogi, A. 233 Ammerman, A. 109 Anderson, W. 120–2, 148, 175 Anghel, S. 219, 233–4 Asheri, D. 23 Assimakopoulou-Atzaka, P. 253 Atasoy, S. 122, 245 Ateşoğulları, S. 219, 221–3, 229, 234 Aurigemma, S. 258 Avramea A. 124 Aybek, S. 245 Aydın, B. 19–20, 219–38 Babelon, E. 312 Baird, D. 127 Baker, C. M. 87, 109 Bal, N. 65 Balty, J. 247, 250, 253, 256–7, 259–60, 273, 275, 292 Banning, E. B. 111 Barat, C. 64, 73, 77 Baratte, F. 258 Bardill, J. 23, 35, 260 Barnes, T. B. 317 Barringer, J. M. 243 Bauer, A. 56 Baxter, M. J. 113 Beardah, C. C. 113 Becker, L. 244 Bekker-Nielsen, T. 19–21, 28, 83, 85–6, 97, 126, 150, 159, 182, 295–305, 307, 309, 329

Belfiore, S. 219 Belke, K. 23, 28–31, 33–4, 122–4, 126, 182, 268, 317 Bell, G. L. 86, 89 Berenson-Maclean, J. K. 244 Berger, A. 231 Bergmann, B. 225 Berti, F. 154 Bes, P. 18, 23–42 Bikoulis, P. 19, 101–17, 126 Bilgi, Ö. 215 Biller, J. 85, 159 Birley, A. R. 232, 235 Bittel, K. 73 Bittner, A. 31 Blanc, N. 258 Blanchard-Lemée, M. 260 Blanton, R. E. 123 Boatwright, M. 219 Böhlendorf, B. 148, 161 Böhlendorf-Arslan, B. 154 Bonifay, M. 27 Boré, E. 231 Brandis, C. G. 311 Braund, D. 23, 30–31, 33 Brenier, F. 329 Bricault, L. 329, 332, 334 Brixhe, C. 265 Broughton, T. R. S. 31 Brouscari, E. 250 Bruneau, P. 256, 274 Bryer, A. A. M. 124 Buccino, L. 19–20, 219–38, 260 Bücheler, F. 310 Budde, L. 249, 253 Bujskikh, S. 243 Burney, C. A. 297 Burrell, B. 313 Cadario, M. 219, 223, 225 Cambi, N. 177–8 Campbell, S. 244, 248–9, 253, 264, 266, 273 Can, B. 276 Candemir, H. 275 Caseau, B. 233–4 Casey, J. 129 Cassis, M. 29, 31, 34, 128

344

Indices

Castejón, R. L. 297 Cavalieri, M. 225 Çelik, Ö. 267 Çelikbaş, E. 121, 268 Çevık, F. 68, 175 Chapot, V. 313 Chavarria, A. 119 Christof, E. 23, 103, 122, 172, 175, 182–3, 254, 268–9, 271, 273, 336 Christol, M. 316 Çinici, A. 87 Çizmeli, Z. 329, 332, 334 Clarke, J. R. 247 Conovici, N. 50 Coulthard, G. 21 Coulton, J. J. 183 Cowan, F. 109 Cremer, M. 184, 189 Croke, B. 101 Crow, J. 101, 123–4, 219, 231–2 Cumont, E. 295, 303 Cumont, F. 142, 295, 303, 309–10, 317 Cvetković Tomašević, G. 258 Czichon, R. 19–20, 52, 83, 85, 96–7, 147–8, 156, 159, 165, 182, 295–305 Dalaison, J. 165, 167, 172, 308, 329–40 Darmon, J.-P. 260 Daszewski, W. A. 273–4 Dauphin, C. 259–60 De Boer, J. 23 De Matteis, L. M. 247, 258, 269 De Paepe, P. 27 Decker, M. J. 27 Deininger, J. 308, 311, 313–4 Delrieux, F. 172, 329–34, 336 Dessau, H. 310 Devos, G. 27 Devreker, J. 173, 175 Diehl, C. 271 Diest, W. von 231 Dillon, S. 229 Dittenberger, W. 310 Djuric, D. 273 Doğanbaş, M. 20, 159, 193–203 Domżalski, K. 23, 28, 47, 51, 64, 74, 89 Donceel-Voûte, P. 64, 71–3, 245–7, 250, 253, 269– 71, 276 Donderer, M. 244, 246 Dönmez, Ş. 93, 161, 215 Dooijes, R. 214 Doonan, O. 19, 43–59, 84, 126–8 Downey, G. 266 Drauschke, J. 129 Drew-Bear, T. 173, 178 Duchesne-Guillemin, M. 260

Dunbabin, K. M. D. 291–2 Dunn, A. 119, 127 Dunnel, R. C. 87, 109 Düring, B. S. 23, 28–9, 34, 121 Eichner, I. 121 Eiring, L. J. 156 Elton, H. 19, 101–17 Emery-Barbier, A. 51 Er, Y. 290 Erciyas, D. B. 33, 84, 87, 128, 215, 219, 231 Ergeç, R. 292 Ersoy, A. 249, 292 Erten, H. N. 22 Ertuğrul, Ö 245 Eyice, S. 123 Farioli Campanati, R. 273 Farnoux, A. 256 Fejfer, J. 229 Fentress, E. 50 Ferriès, M.-C. 329 Février, P.-A. 273 Field, R. 292 Filges, A. 229 Filippenko, A. A. 156 Firat, N. 27–8 Foss, C. 31, 33–4, 127, 129, 261 Fourcade, P.-T. 73, 309, 311, 314 Fox, S. 147, 152 French, D. 52, 56–7, 135, 142, 172, 179, 182, 184– 6, 309, 316–8 Friedland, E. A. 225 Friedländer, P. 273 Friesem, D. E. 108 Frisch, P. 315 Gabrielsen, V. 23 Gall, H. von 73, 187 Garlan Y. 46, 50 Gawlikowski, M. 256 Geominy, W. 228 Ghedini, F. 243 Giannopoulou, M. 27 Gill, M. A. V. 113 Glatz, C. 23, 28–9, 32, 34, 92, 95, 98, 121, 172 Göçen, Ü. 254–7 Gökoğlu, A. 64–5, 68, 72–3, 77 Goldman, A. L. 148 González Salazar, J. M. 297 Göztaş, İ. H. 21 Grabar, A. 86 Grassigli, G. L. 256 Grégoire, H. 295, 310 Groß-Albenhausen, K. 156 Grossardt, P. 242, 244 Guest, P. 124

Modern personal names Guldager Bilde, P. 219, 223 Gury, F. 258 Hachlili, R. 271, 273 Haensch, R. 184 Haldon, J. 19, 51, 57–8, 101–17 Halfmann, H. 179 Halifeoğlu, F. M. 295 Hallett, C. 229 Hamilton, W. J. 52, 54, 180 Harl, K. W. 336 Haspels, C. H. E. 188 Haussoulier, B. 310 Hayes, J. W. 23, 25, 27–8 Hedreen, G. 243 Hellenkemper Salies, G. 247–50, 260, 269, 273–5 Herrmann, P. 309–10 Hill, S. 107, 123–4, 126, 219, 231–2, 253–4 Hirschfeld, G. 70, 171, 174, 189 Hnila, P. 19, 93, 147–164, 165 Hoepfner, W. 23 Hoffmann, A. 230–1 Høgel, C. 150 Højte, J. M. 165, 307 Hommel, H. 243 Howard-Johnston, J. D. 129 Howgego, C. J. 167, 336 Humbert, J. B. 273 Hupe, J. 243 Huskinson, J. 266, 273 Icard-Gianolio, N. 243 Imhoof-Blumer, F. 312 Ireland, S. 129 Isager, S. 268 İvgin, İ. 20, 205–17 Ivison, E. 108, 123 Izdebski, A. 128 Jacobs, I. 119, 126–7 Jacopi, G. 65, 71–75, 91, 179, 186–7 Japp, S. 25 Jes, K. 173 Jessee, E. 152 Jobst, W. 246, 253, 260, 292 Johnson, P. 19, 61–82, 93, 187 Jones, A. H. M. 317 Jones, C. P. 184 Jonnes, L. 171 Jouejati, R. 250 Jouejati-Madwar, R. 253 Jucker, H. 257 Jusseret, S. 225 Kahl, G. 179 Kan Şahin, G. 23, 27, 122 Karabulut, H. 260, 267, 281

345

Karagiorgou, O. 27, 33 Karanastasi P. 219, 223, 225–6 Kassab Tezgör, D. 23, 27, 33, 45–6, 49–50, 57, 239 Kaygusuz, İ. 307 Kelp, U. 173, 188 Kemp-Lindemann, D. 243 Kennedy, H. 126 Keskın, E. 20, 239, 285–93 Khrushkova, L. 86 Kienlin, A. von 73, 187, 261 Kılavuz, B. N. 121, 268 Kinneir, J. M. 299 Kirilov, C. 126 Kitzinger, E. 247 Kleiner, F. 92 Klinghardt, M. 89 Knipowitsch, T. 23 Koch, J. 20, 171–192, 336 Kohl, M. 108 Koiner, G. 103 Kokkinia, C. 183 Kökten, H. 239 Kolb, F. 127 Kollig, I. C. 161 Kondoleon, C. 244, 253, 258, 271 Kortanoğlu, E. 122 Kossatz-Deißmann, A. 243 Krautheimer, R. 86 Kriseleit, I. 260 Krsmanovic, D. 148, 175 Kruse, H. J. 227 Kühne, H. 161 Kunnert, U. 172, 182 Künzl, E. 173 Kuzucuoğlu, C. 45, 64, 77 Kwiatkowska, M. 151, 161 Laflı, E. 23, 27, 107, 121–2, 172, 175, 179, 182–3, 254, 268–71, 273, 336 Lambrechts, P. 175 Lancha, J. 240 Lassus, J. 250, 253 Lavan, L. 231, 233–4 Leader-Newby, R. 266, 275 Lenaghan, J. 229 Leone, A. 231, 233–4 Leonhard, R. 73 Leschhorn, W. 167, 312, 330 Levi, D. 247–8, 250, 253, 258–61, 265–6, 271, 292 Lewit, T. 119 Lightfoot, C. S. 123 Ling, R. 292 Littlewood, A. R. 291 Lo Cascio, E. 123 Lochman, T. 171, 173, 177–8 Lohmann, H. 127

346

Indices

López Monteagudo, G. 266 Loriot, X. 311–4, 316–8 Lund, J. 23 Macridy, T. 215, 314 Maguire, H. 271, 273 Majcherek, G. 265 Mango, C. 101 Manisse, P. 329 Marek, C. 20, 23, 29, 31, 33–4, 61, 91, 104, 121, 123, 158, 172, 179, 183–6, 189, 219–20, 223, 226, 229–32, 235, 253, 268, 307–30, 335–6 Marin, E. 232 Marklein, K. 147, 152–3 Marro, C. 64–5, 71–2 Matthews, R. 23, 31, 34, 84–85, 89, 92, 94–6, 98, 108–9, 120, 161, 171–2, 175, 182, 184, 187 Mavrojannis, T. 243–4 Mayer, W. 275 Megaw, A. H. S. 87, 253 Meier, M. 128 Meiggs, R. 31 Mendel, G. 231 Menichetti, M. 243 Mercan, A. 254–7 Michaelides, D. 265, 271 Michaelis, A. 258 Mientjes, A. C. 61 Mitchell, S. 30–1, 177, 182, 184–6, 309, 311, 317, 336 Mitford, T. B. 265, 268 Mitsopoulos-Leon, V. 154 Möbius, H. 184 Mommsen, H. 25 Moock, D. W. von 184 Mosch, H. C. von 226, 231–2, 235 Mundell Mango, M. 127 Munro, J. A. R. 20, 182, 295, 299, 303, 316–7 Musso, L. 20, 123, 239–284, 262, 264, 336 Naumann, R. 73 Neira, L. 243 Nesbitt, J. W. 122 Neustupný, E. 158–9 Newhard, J. 19, 101–17 Nieuwenhuyse, O. P. 214 Niewöhner, P. 108, 126 Nordbø, J. H. 331–2 Nowotny, E. 152 Ochal-Czarnowicz, A. 243 Odell G. 109 Oikonomides, N. A. 122 Olcay, O. F. 193 Oliver, G. 184 Olshausen, E. 85, 91, 98, 159, 165–6, 296 Önal, M. 244, 260, 267

Opaiţ, A. 33, 64 Ovadia, A. 259–60 Ostrowski, J. A. 273 Öz, A. K. 245 Özcan, B. 108, 159 Özdemir, C. 19, 21, 135–45, 159, 195, 292 Özdoğan, A. 64–5, 67–8, 71–2, 77 Özet, A. 292 Öztas, F. 21 Öztürk, B. 122, 254 Paine, R. R. 158 Parrish, D. 239, 242, 260 Pataci, S. 268–9, 272, 292 Peña, J. T. 24 Pettegrew, D. K. 28 Pfuhl, E. 184 Picard, G. C. 257 Piccirillo, M. 246, 260, 273 Pieri, D. 23, 33 Pluciennik, M. 61 Polacek, J. 129 Polaschek, E. 229 Poulsen, B. 247 Poulter, A. 108–9, 114 Poulou-Papadimitriou, N. 152, 156 Ramsay, W. M. 86, 89, 310 Redford, S. 126 Redman, C. 108 Reimer, P. J. 156 Reinach, T. 310, 312 Reinsberg, C. 228 Rémy, B. 311, 329, 332, 334 Reynolds, P. 23, 27 Rheidt, K. 108, 160, 173 Richter, C. H. 209, 215 Rife, J. L. 150–1, 154–5 Ritter, M. 19–20, 68, 80, 93, 119–33, 175 Robert, L. 231 Robinson, D. 52, 56 Romančuk, A. I. 86, 93, 156–7 Roodenberg, J. 161 Roosevelt, C. H. 175 Roueché, C. 29, 31 Rutherford, I. 242, 244 Şahin, D. 240, 242–4 Salmeri, G. 33 Salzmann, C. 161 Salzmann, D. 246 Sanders, G. 27 Saradi, H. G. 119, 127–8 Sauer, V. 19–21, 91, 154–5, 159, 165–9, 316, 329– 40 Scheibelreiter-Gail, V. 247–8, 253, 258, 265, 268, 275

Modern personal names Schmid, S. 234 Schneider, G. 25 Scholz, B. I. 228 Schultz, H.-D. 329 Schwertheim, E. 179 Şerifoğlu, T. E. 121 Sert, H. 295 Ševčenko, I. 101 Sevilla Sadeh, N. 260 Sharp, R. St. 123–4, 219 Sherk, R. K. 308 Shott, 109 Simek, J. 109 Şimşek, Ö. 219, 221–3, 229, 234 Sintès C. 240 Skinner, M. 152 Slane, K. W. 27 Smart, D. 126 Smith, R. R. R. 229 Sökmen, E. 128 Solin, H. 244 Sönmez, I. F. 122 Sørensen, S. L. 309 Sprague, R. 152 Stathakopoulos, D. 128, 156 Steffensen, J. P. 128 Steinsapir, A. I. 94 Stillwell, R. 250, 258, 260–1, 271 Stone, D. 50–1 Stubbe, J. 173–5 Summerer, L. 17–21, 52, 69, 73, 172, 187, 215, 219–238, 261 Swan, V. G. 33 Szydlak, R. 21 Takaz, H. 159, 193 Talgam, R. 257 Tatlıcan, İ. 253–4 Taussig, H. 89 Tektaş, M. 194 Ternes, C. M. 246 Theophilidou, E. 246–7 Tissi, L. M. 273 Torraca, L. 308 Toteva, G. D. 153 Toynbee, J. M. C. 256–7 Trimble, J. 92 Trovabene, G. 273 Tsetskhladze, G. R. 23 Tülek, F. 258, 267 Türktüzün, M. 174 Tuzcu, A. 292 Tykot, R. H. 225 Ulrich, R. B. 213 Ünan, S. 159, 205, 215

347

Üstüner, A. C. 292 Uzun-Öğüş, E. 291 Vaag, L. E. 25 Valeri, C. 228 Vanhaverbeke, H. 127 Vargiu, R. 158 Vatin, C. 256 Venturini, F. 265 Vermeulen, F. 27 Vetters, H. 260 Visconti, P. 258–9 Vitale, M. 308, 311–2, 314, 317 Vollenweider, M. L. 243 Vroom, J. 112 Waelkens, M. 173–8, 180, 182, 187 Wagner, J. 275 Ward Perkins, J. B. 256–7 Ward-Perkins, J. B. 271 Warry, R. 161 Watson, P. J. 108 Waywell, G. B. 227 Weiser, W. 312, 329 Weiss, Z. 257, 275 Werner, K. E. 247, 258 Westphalen, S. 108 White, K. D. 178 Williamson, C. 19 Wilson, C. 299, 303 Wilson, D. R. 165 Winfield, D. 124, 297 Winther-Jacobsen, K. 17–21, 28, 83–99, 126, 299 Wittwer-Backofen, U. 150, 158 Witulski, T. 232 Wojan, F. 165, 329, 332–4 Wulf, U. 174 Wypyski, M. T. 113 Yaman, Z. 264 Yegül, F. K. 233 Yeivin, Z. 250 Yıldırım, C. 65, 67 Yılmaz, M. A. 153 Yılmaz, S. 295 Yolaçan, D. 249 Yorston, R. 109 Yüce, A. 159, 194 Yurtseven, F. 209 Zäh, A. 107, 121–2, 268, 270, 273 Zaqzuq, A. 246, 250, 253, 260, 273 Zavagno, L. 219 Zhuravlev, D. 23, 64, 155 Zieler, M. 148, 161 Zvabados, A.-V. 243 Żuchowska, M. 256

Places

Abdalhasan 77 Abonouteichos (Ionopolis) 182, 184, 307, 312, 329–40 Acholla 257 Adatepe 85, 98 Africa Proconsularis 258 Afşar Köy 173–4 Aigialos (Egilan; Cide) 18, 23–42, 121 Aizanoi 173–4 Akalan 215 Akdeğirmen Höyük 275 Akroinos 127 Alaçam 297 Alanya 258 Alasökü han 56 Alexandria 265 Algeria 260 Alisaray 76, 78 Alpagut 65 Altınoğlu 53 Altıntepe (Erzincan) 276 Amaseia (Amasya) 17, 19–21, 129, 155, 159, 165, 167, 180, 193–203, 239, 275, 285, 292, 297, 299, 303, 307, 312, 318, 329–40 Amasra → Amastris Amastris (Amasra) 20, 31, 33–4, 123–4, 126–7, 129–30, 135–145, 181, 184, 219–38, 307, 312, 329–40 Amasya → Amaseia Ambrakia 128 Amisos (Samsun) 20, 166, 187, 205, 207, 215, 239– 46, 275, 295, 303, 307, 312, 329–40 Amorium 108, 173, 178 Ankyra (Ankara) 108, 129, 180, 182, 189, 309, 311, 317–8 Antalya 225 Antioch 244–5, 247–50, 253, 258, 260–1, 267, 271, 292 Apamea 250, 256, 260, 275, 292 Arcadia 271 Arıca 316 Arta 128 Aşağı Tepedelik 74–5 Aşağıemerce 69, 76 Asar 215 Asarağaç Kalesi 215

Asarkale 215 Aşıra 68, 79 Aslan Mahalle 187 Athens 244 Atköprü 29–31 Avkat → Euchaita Aydoğdu 301–2 Aygır Kalesi 68, 70, 76 Azman 205 Baalbek 246 Bafra 54, 215, 297 Bağlıca 173 Barçın 161 Başören 173 Bedemci 72–4, 80 Bekdemir 70 Bekdemirekşi 70 Belören 50 Bengü 215 Bet Shean → Scythopolis Bigeste 177 Bithynia 158, 180, 182–3, 189, 225, 308–19 Blaene 307 Bodrum → Halicarnassus Boeotia 26 Boğazkoru 92 Boyabat 56, 172, 178–9, 181 Boztepe 56 Burnum 178 Bursa 262 Çakıroğlu 50, 52–6 Çakırtaş 301 Çalköy 300 Çamdibi 34 Çamlıca 92, 94 Cappadocia 86, 310–1, 316 Çarıklı 79 Çayönü 108 Çatalhöyük 109, 151, 161 Çetmi 70–2 Çetmi Höyük (Üyük Tepe) 71–4, 80 Çevik 74–5 Chersonesos (Sevastopol) 86, 93, 156 Chonai 127 Cide → Aigialos

Places Çiftlik 50, 107, 126, 253–4 Cilicia 27, 158, 248, 253, 273, 276 Çırnık 53 Claudiopolis 268 Colonia Iulia Felix 329–40 Commagene 273, 275 Constantinople (Byzantium) 23, 28, 30–1, 57, 86, 89, 112, 180, 182, 226, 231, 246, 260–1, 265, 275, 292 Çorum 101 Crimea 87, 156 Cyprus 27, 251, 271, 275 Cyrenaica (Qasr el-Lebia) 229, 271 Cyrene 265 Cyrrhus 260 Cythero → Kytoros Dalmatia 177 Daphne 258, 260, 265, 267 Dazimon-Dokeia 108, 128 Delos 154 Demirci 46–7, 49–51 Dereler 216 Derinyar 75 Devrek 126 Dikilitaş 216 Diospolis 182 Diospontus 318 Docimeion 176–7 Dodona 244 Dökme Tepesi 148 Dolaman Arkası 67, 78–80 Dolaman Tepesi 78–9 Domanitis 307 Donalar 70–1, 73–4, 186 Dorylaion (Eskişehir) 178 Doyran 85, 92–4, 296 Edessa (Şanlıurfa) 260–1, 267, 276 Egilan → Aigialos Elaiussa Sebaste 158 Eldevüz 46–7, 49, 51 Elvançelebi 101 Emesa (Homs) 257, 260 Enderes 182 Ephesos 26, 126, 154 Epirus 128 Eretria 234 Erikli 50, 52, 55 Esenler 65 Euchaita (Avkat, Boyözü) 19, 101–117 Eupatoria 135 Ezinepazar 318 Florence 271

349

Galatia 173–4, 177, 180, 182–3, 188–9, 309, 311, 317 Galilee 257, 275 Gamzigrad → Romuliana Gangra-Germanikopolis (Çankırı) 120, 125, 171– 2, 182–3, 187, 307–19, 329–40 Geriz 92 Germia 108, 126 Gerna dere 57, 126 Gerze 49–50 Gideros → Kytoros Girikihacıyan 108 Gödel Kayabaşı 65 Gordion 148 Greece 26, 243 Güldere 300–1 Gürcü 104 Hadrianopolis (Kaisareia; Eskipazar) 23, 25–7, 107, 121–4, 126–30, 172, 175, 179, 183–4, 189, 240, 268–73, 275, 307 Halicarnassus 226–7, 268, 292 Hama 250, 253, 260, 273 Hancılar 70 Harmanyeri (Sivritepe) 78–9 Hattusa (Boğazköy) 154, 158, 161, 297 Havza → Thermai tôn Phazemonitôn Helenopontus 119, 318 Heraclea Pontica 33, 180–1, 184, 311 Hierapytna 225 Homs → Emesa Honorias 119, 121–2 Honorias-Paphlagonia 31 Honsalar 65 Höyüktepe 65 Hyettos 26 İkitepe 75–6 İkiztepe 175, 215 Ilgaz 182 Ilıca 301 Ilıpınar 161 İncebel Tepesi 80 Incesu 92 İskilip 126 Isthmia 150 Jabaliyah 273 Kabalı 55 Kabeira 307 Kadioğlu → Tios Kadykovka 156 Kaisareia (Caesarea) → Hadrianopolis Kalekapı 73 Kaletepesi 77 Karaağaç Limanı 33

350

Indices

Karabük 126, 172, 268 Karaburun 175 Karacaoğlu 71 Karadağ 86 Karadağ-Mahalaç 86 Karaoğlan Tepesi 72 Karapınar 47–9, 51 Karapürçek 74 Karkucak 91 Karlı 56 Kastamonu 29, 64–5, 68, 78, 91, 128, 172, 174, 180, 185, 187, 262, 264 Kavak 295 Kayaaltı 72 Kayalıpınar 148, 161 Kazallı Kalesi 29 Keçioğlu 46, 51 Kelemperoğlu 53, 55 Kephalos 269 Kerameikos 184 Kerasous 312, 329–40 Kese 78, 80 Kiraztepe 51 Kırha 70 Kırkgeçit çayı 52–4, 56 Kırktepeler 72–3, 186–7 Kisecik 182 Kızkalesi (Kiz Kalesi) 79–80, 93, 124–5 Kocaoğlu 91, 301–2 Kolay 215 Kolchis 333 Kolossai 127 Köprübaşı 295, 299–300, 303–4 Komana Pontica 128, 135, 179, 312, 329–40 Kos 246, 250, 258, 269 Kourion 253, 265, 268 Kozlu 216 Krateia-Flaviopolis (Gerede) 183 Kromna (Cromna; Hisarköy, Tekkeönü) 28–9, 121 Küpderesi 285–293 Kurban Tepesi 70 Kurt Köprüsü 295, 301, 304 Kurupelit (Karacadoğan Höyük) 215 Kuruzluk Tepesi 70 Kuşaklıtepe 175 Kuştepesi 72 Kütahya 124, 178 Kytoros (Kytoron/Cythero/Cytorus/Cytorum; Gideros, Abdulkadir), 26, 31, 33–4, 121, 317 Ladik → Laodikeia Lala 53 Laodikeia (Ladik) 295, 303 Leptiminus 50 Leuka 243–4

Liguria 177 London 234 Lycaonia (Loulon) 127–8 Lycia 127 Macedonia 177 Madaba 273 Magnopolis 182, 331 Mani 152 Manigünez Tepeleri 76 Mariamin 260 Martı Kale 215 Mecitözü 101, 104 Mercimekli Sırtı 68 Merzifon 85, 96–7, 299, 303 Metropolis 248 Milesia 127 Miletene 129 Miletus 185 Misis → Mopsuestia Moesia Inferior 119 Mokissos 126 Mopsuestia (Misis) 248, 253 Nablus → Samaria Namaztepe 68 Narona 232 Nasperkene 179 Nea Paphos 258, 265, 273–4 Neokaisareia (Niksar) 135, 159, 182, 295, 312, 329–40 Neoklaudiopolis (Phazemon/Neapolis/Andrapa; Vezirköprü) 19–21, 25–6, 28, 83–99, 125–6, 147, 150, 159–60, 179, 182–3, 295–305, 307– 18, 329–40 Nerik 83, 147, 182, 296–7, → Oymaağaç Nikaia 317 Nikomedia 124, 189, 308 Nikopolis 182, 329–40 Niksar → Neokaisareia Okçular 29, 34, Olbia 243 Olympia 226 Ömerli 68 Ören 74 Ortaklar 316 Orthesia 258 Osrhoene 273, 276 Oum Hartaïne 270 Ovacık 29 Oymaağaç 19, 52, 85, 98, 147–169, 182, 297, 303 Palmyra 256 Pannonia 177 Papaz Tarlası 19, 85–93

Places Paphlagonia 19–21, 23, 61, 70, 83–5, 92, 94–7, 107, 119–33, 156, 171–192, 240, 307–28 Patara 216 Peloponnese 152 Pergamon 160, 310 Pessinus 148, 173, 175 Phanaroia 301, 304, 307 Pharnakeia 165, 307 Phaustinopolis 127 Phazemonitis 20, 295, 303–4, 307 Philippopolis (Shahba) 256, 260 Phokaia 25 Phrygia 127, 171, 174, 176–7, 180, 187–9 Pimolisene 307 Polemonion 312, 331 Pompeiopolis (Zımbıllı Tepe, Taşköprü) 19–21, 23, 25–6, 28, 52, 61–82, 90–3, 123–30, 156, 171–80, 182–9, 240, 261–8, 275, 295, 301, 307– 19, 329–40 Pontus 128, 158, 160, 179, 189, 304, 308–19 Prusias ad Hypium 225 Prymnessos 127 Pulcular 70 Rome 86, 227, 258–9 Romuliana (Gamzigrad) 258 Safranbolu 126 Sagalassos 127 Salavantepe 76 Salona 177 Samaria (Nablus) 260 Samsun → Amisos Şanlıurfa → Edessa Saracık 56 Saraycık Dağı 68 Sarıkum 56 Sarıyer 53 Sarsı 57 Satala 124 Scythopolis (Bet Shean) 250 Sebasteia 312, 329–40 Sebastopolis (Sulusaray) 108, 312, 329–40 Selmanlı Mahallesi 65 Sepphoris 257, 275 Setif 260 Shahba → Philippopolis Sicily 243 Sidene 307 Sidon 271 Silifke 317 Sinope (Sinop) 19, 23, 25–8, 30, 33, 43–59, 69, 125–6, 128–30, 178–9, 181–2, 184–5, 187, 239, 246–54, 267, 275, 312, 317–8, → Colonia Iulia Felix Sivrintepe 74

351

Smyrna 248, 292 Soli Pompeiopolis (Mersin) 216 Sora 125–6 Sorkum 56 Sülüklütepe 68 Syedra 258 Synnada 176 Syria 250, 261, 271, 273, 275–6 Tanagra 26 Tarsus 216 Taşköprü → Pompeiopolis Tavukçuoğlu 74 Tayyibat al-Imâm 273 Tegea 271 Tekkekıran 301 Tekkeköy 205 Tekkeönü → Kromna Tepedelik 74, 185, 187 Tepekaya 77 Tepeören 85, 92–8, 297, 299, 303–4 Themiskyra 307 Theodosiopolis 129 Thermai tôn Phazemonitôn (Havza) 295–7, 299, 303 Thymaina (Thymena/Teuthrania/Timle/Tymle; Uğurlu) 33–34 Tilurium 177 Tınar Tepe 75, 79 Tınaztepe 216 Tingiroğlu 52 Tios (Teios/Tieion/Tium; Kadioğlu Köyü) 122, 184, 239, 254–61, 275 Tokat → Dazimon-Dokeia Toraman Tepe 245–6 Tosya 182 Trapezous 219, 307, 312, 329–40 Trinquetaille 240 Tripoli 258 Troy 148, 161 Tunesia 27, 50 Tunis 256 Türbe Tepe Üste 29 Üçbaş 183 Uğurlu → Thymaina Umm al-Rasa 273 Uzungürgen 46–7, 51 Vezirköprü → Neoklaudiopolis Yakadibi 55 Yalı 46 Yalı Plajı 50–1 Yassiçal (Ebemi) 19, 135–45 Yavru 20, 285–293 Yeniköy 74, 80

352 Yeşidere 216 Yeşilkent-Demirciyeri 216 Yozgat-Akdağmağden 216 Yukarı Tepedelik 74 Yukarı Yanlar 316 Yukarıemercer 70 Yüklütepe 71–4 Yürükçal 299–301

Indices Zafer 205–17 Zela (Zile) 295, 307, 312, 318, 329–40 Zelatin Tepe 72 Zelo 74 Zeugma 244, 246, 260, 267, 275, 292 Zile → Zela Zindankaya 93, 161 Zliten 258

Sources

Literary texts Arrian, Anabasis of Alexander 1.11–2 244

14.76

Arrian, Periplus 1.3–4 14

10.98 219 29, 33–4

Plutarch, Alexander

Cassius Dio, Roman History 51.20.7

308

Plutarch, Pericles

Cicero, On the laws 2.25.66 George Kedrenos, History 1.566

15.7–9

231

Hierokles, Synekdemos 694–6

126

Homer, Iliad 18.66 18.86 18.503–7 19.1–36

240 240 243 242–3

Philostratus, On heroes 45–54

244 183 183

226, 230 244 184

Ptolemy, Geography 5.4

George Pisides, Expeditio Persica 2.256–60 129

Pliny the Elder 5.32 6.2

Pliny the Younger, Letters

20.1–2 184

179

182–3, 317

Strabo, Geography 12.2.14 12.3.8 12.3.9 12.3.12 12.3.15 12.3.26 12.3.36 12.3.38 12.3.40 12.3.41

215 33 181, 307 57 292 181 179 307 180 307

Vita Alypii stylitae Hadrianopolitani prior (BHG 64) 121 Vita Georgii Amastridensis (BHG 668) 24–5

123

354

Indices

Inscriptions AE 1925: 126 1952: 232 1978: 815

311 311 317

Bekker-Nielsen and Høgel 2013 2 150 Fourcade 1811 308–9 French 1981 6 (p. 153)

317

French 2004 (IK 64) 9–49 64 68 71

184 186 186 186

French 2011 12

185–6

French 2013 48 B

316

IG XII 9 188

310

IOSPE I2 79

311

Jonnes 1994 9

171

Marek 1993 Abonuteichos/Ionopolis 5 Amastris 1c Amastris 9 Amastris 10 Amastris 48–51 Amastris 88

184 308 235 232, 235 184 184

Amastris 90 Kaisareia/Hadrianopolis 25–8 Kaisareia/Hadrianopolis 30–3 Kaisareia/Hadrianopolis 36 Kaisareia/Hadrianopolis 38 Kaisareia/Hadrianopolis 40–3 Kaisareia/Hadrianopolis 41 Kaisareia/Hadrianopolis 45–70 Kaisareia/Hadrianopolis 73 Kaisareia/Hadrianopolis 75–8 Kaisareia/Hadrianopolis 79 Kaisareia/Hadrianopolis 86 Kaisareia/Hadrianopolis 89 Kaisareia/Hadrianopolis 91 Pompeiopolis 3–4 Pompeiopolis 12 Pompeiopolis 28 Pompeiopolis 30 Pompeiopolis 36 Pompeiopolis 38 Pompeiopolis 52

184 183 183 183 183 183 184 183 183 183 183–4 183–4 183 183 172 172 186 185–6 179 91 183, 185

SEG 35.1317

232, 235

Sherk 1969 52

308

Strubbe 2005 64 105 126 127 130

173 173–4 174 175 173

Studia Pontica 3 66 211 227

309–10 104 104

geographic a historic a

Begründet von Ernst Kirsten, herausgegeben von Eckart Olshausen und Vera Sauer. Die Bände 1–8 sind in den Verlagen Dr. Rudolf Habelt (Bonn) und Adolf M. Hakkert (Amsterdam) erschienen.

Franz Steiner Verlag

9.

ISSN 1381–0472

Gerhard H. Waldherr Erdbeben – Das außergewöhnliche Normale Zur Rezeption seismischer Aktivitäten in literarischen Quellen vom 4. Jahrhundert v. Chr. bis zum 4. Jahrhundert n. Chr. 1997. 271 S., kt. ISBN 978-3-515-07070-6 10. Eckart Olshausen / Holger Sonnabend (Hg.) Naturkatastrophen in der antiken Welt Stuttgarter Kolloquium zur historischen Geographie des Altertums 6, 1996 1998. 485 S. mit zahlr. Abb., kt. ISBN 978-3-515-07252-6 11. Bert Freyberger Südgallien im 1. Jahrhundert v. Chr. Phasen, Konsequenzen und Grenzen römischer Eroberung (125–27/22 v. Chr.) 1999. 320 S. mit 16 Abb., kt. ISBN 978-3-515-07330-1 12. Johannes Engels Augusteische Oikumenegeographie und Universalhistorie im Werk Strabons von Amaseia 1999. 464 S., kt. ISBN 978-3-515-07459-9 13. Lâtife Summerer Hellenistische Terrakotten aus Amisos Ein Beitrag zur Kunstgeschichte des Pontosgebietes 1999. 232 S. und 64 Taf., kt. ISBN 978-3-515-07409-4 14. Stefan Faller Taprobane im Wandel der Zeit Das Śrî-Lankâ-Bild in griechischen und lateinischen Quellen zwischen Alexanderzug und Spätantike 2000. 243 S., kt. ISBN 978-3-515-07471-1 15. Otar Lordkipanidze Phasis The River and City in Colchis 2000. 147 S. und 8 Taf., kt.

ISBN 978-3-515-07271-7 16. Marcus Nenninger Die Römer und der Wald 2001. 268 S. mit 3 Abb., kt. ISBN 978-3-515-07398-1 17. Eckart Olshausen / Holger Sonnabend (Hg.) Zu Wasser und zu Land – Verkehrswege in der antiken Welt Stuttgarter Kolloquium zur historischen Geographie des Altertums 7, 1999 2002. 492 S. mit zahlr. Abb., kt. ISBN 978-3-515-08053-8 18. Maria Francesio L’idea di città in Libanio 2004. 157 S., kt. ISBN 978-3-515-08646-2 19. Frauke Lätsch Insularität und Gesellschaft Untersuchungen zur Auswirkung der Insellage auf die Gesellschaftsentwicklung 2005. 298 S., kt. ISBN 978-3-515-08431-4 20. Jochen Werner Mayer Imus ad villam Studien zur Villeggiatur im stadtrömischen Suburbium in der späten Republik und frühen Kaiserzeit 2005. 266 S., kt. ISBN 978-3-515-08787-2 21. Eckart Olshausen / Holger Sonnabend (Hg.) „Troianer sind wir gewesen“ – Migrationen in der antiken Welt Stuttgarter Kolloquium zur Historischen Geographie des Altertums 8, 2002 2006. 431 S. mit 58 Abb., kt. ISBN 978-3-515-08750-6 22. Jochen Haas Die Umweltkrise des 3. Jahrhundert n. Chr. im Nordwesten des Imperium Romanum Interdisziplinäre Studien zu einem Aspekt der allgemeinen Reichskrise im Bereich der beiden Germaniae sowie der Belgica und der Raetia

23.

24.

25.

26.

2006. 322 S., kt. ISBN 978-3-515-08880-0 Klaus Tausend Verkehrswege der Argolis Rekonstruktion und historische Bedeutung 2006. 226 S. mit 6 Abb., 22 Ktn. und CD-ROM, kt. ISBN 978-3-515-08943-2 Gerhard H. Waldherr / Anselm Smolka (Hg.) Antike Erdbeben im alpinen und zirkumalpinen Raum / Earthquakes in Antiquity in the Alpine and Circum-alpine Region Befunde und Probleme in archäologischer, historischer und seismologischer Sicht / Findings and Problems from an Archaeological, Historical and Seismological Viewpoint. Beiträge des Interdisziplinären Workshops Schloss Hohenkammer vom 14.–15. Mai 2004 2007. 189 S. mit 125 s/w-, 3 Farbabb. und 5 Tab., kt. ISBN 978-3-515-09030-8 Klaus Tausend Im Inneren Germaniens Beziehungen zwischen den germanischen Stämmen vom 1. Jh. v. Chr. bis zum 2. Jh. n. Chr. Mit Beiträgen von Günter Stangl und Sabine Tausend 2009. 282 S. mit 14 Ktn., kt. ISBN 978-3-515-09416-0 Eckart Olshausen / Vera Sauer (Hg.) Die Landschaft und die Religion Stuttgarter Kolloquium zur Historischen Geographie des Altertums 9, 2005

2009. 422 S. mit 94 Abb., kt. ISBN 978-3-515-09422-1 27. Frank Stini Plenum exiliis mare Untersuchungen zum Exil in der römischen Kaiserzeit 2011. 378 S., kt. ISBN 978-3-515-09894-3 28. Eckart Olshausen / Vera Sauer (Hg.) Die Schätze der Erde – Natürliche Ressourcen in der antiken Welt Stuttgarter Kolloquium zur Historischen Geographie des Altertums 10, 2008 2012. 425 S. mit 55 Abb., 1o Tab. und CD-ROM, kt. ISBN 978-3-515-10143-1 29. Tonnes Bekker-Nielsen (ed.) Space, Place and Identity in Northern Anatolia 2014. 271 S. mit 120 s/w-, und 27 Farbabbildungen, kt. ISBN 978-3-515-10748-8 30. Ekaterina Nechaeva Embassies – Negotiations – Gifts Systems of East Roman Diplomacy in Late Antiquity 2014. 306 S., kt. ISBN 978-3-515-10632-0 31. Eckart Olshausen / Vera Sauer (Hg.) Mobilität in den Kulturen der antiken Mittelmeerwelt Stuttgarter Kolloquium zur Historischen Geographie des Altertums 11, 2011 2014. 565 S. mit 26 Abb. und 22 Ktn., kt. ISBN 978-3-515-10883-6

After his victory over Mithradates VI the Roman general Pompey founded a number of cities in order to shape the newly founded Roman province in the inland of the conquered former Pontic kingdom, in the southern Black Sea region. This sparked the beginning of an intense process of urban and rural development peaking in the 2nd century AD and continuing until the Byzantine period, a level of intensity never accomplished in northern Anatolia until today. The reorganization of space through the development of new urban centers affected the whole region

and transformed the territory. This volume reviews current knowledge regarding these new founded Roman cities in relation to their territories, necropoleis and sanctuaries. It consists of 18 articles, which explore dynamics in settlement patterns, architecture, urban and mortuary spaces, monetary circulation and epigraphic habit. Some articles present the results of recent field research, others review little known material ripe for new interpretations, while new archaeological data is provided by the reports of rescue excavations carried out by local museums.

www.steiner-verlag.de Franz Steiner Verlag

ISBN 978-3-515-11214-7