Naturwissenschaftliche Analysen vor- und frühgeschichtlicher Keramik II Universitätsforschungen zur prähistorischen Archäologie Band 216 Aus der Abt. Vor- und Frühgeschichtliche Archäologie der Universität Hamburg 2012 Verlag Dr. Rudolf Habelt GmbH, Bonn Naturwissenschaftliche Analysen vor- und frühgeschichtlicher Keramik II Methoden, Anwendungsbereiche, Auswertungsmöglichkeiten Dritter und vierter internationaler Workshop für junge Wissenschaftlerinnen und Wissenschaftler in Hamburg am . Februar 2010 und 5. Februar 2011 herausgegeben von Britta Ramminger Ole Stilborg 2012 Verlag Dr. Rudolf Habelt GmbH, Bonn ISBN 978-3-7749-3 Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie. Detailliertere bibliografische Daten sind im Internet über <http: //dnb.d-nb.de> abrufbar. Copyright 2012 by Dr. Rudolf Habelt GmbH, Bonn VORWORT DER HERAUSGEBER Die Reihe „Universitätsforschungen zur prähistorischen Archäologie“ soll einem in der jüngeren Vergangenheit entstandenen Bedürfnis Rechnung tragen, nämlich Examensarbeiten und andere Forschungsleistungen vornehmlich jüngerer Wissenschaftler in die Öffentlichkeit zu tragen. Die etablierten Reihen und Zeitschriften des Faches reichen längst nicht mehr aus, die vorhandenen Manuskripte aufzunehmen. Die Universitäten sind deshalb aufgerufen, Abhilfe zu schaffen. Einige von ihnen haben mit den ihnen zur Verfügung stehenden Mitteln unter zumeist tatkräftigem Handanlegen der Autoren die vorliegende Reihe begründet. Thematisch soll darin die ganze Breite des Faches vom Paläolithikum bis zur Archäologie der Neuzeit ihren Platz finden. Ursprünglich hatten sich fünf Universitätsinstitute in Deutschland zur Herausgabe der Reihe zusammengefunden, der Kreis ist inzwischen größer geworden. Er lädt alle interessierten Professoren und Dozenten ein, als Mitherausgeber tätig zu werden und Arbeiten aus ihrem Bereich der Reihe zukommen zu lassen. Für die einzelnen Bände zeichnen jeweils die Autoren und Institute ihrer Herkunft, die im Titel deutlich gekennzeichnet sind, verantwortlich. Sie erstellen Satz, Umbruch und einen Ausdruck. Bei gleicher Anordnung des Umschlages haben die verschiedenen beteiligten Universitäten jeweils eine spezifische Farbe. Finanzierung und Druck erfolgen entweder durch sie selbst oder durch den Verlag Dr. Rudolf Habelt GmbH, der in jedem Fall den Vertrieb der Bände sichert. Herausgeber sind derzeit: Kurt Alt (Mainz) Peter Breuning (Frankfurt am Main) Philippe Della Casa (Zürich) Manfred K.H. Eggert (Tübingen) Clemens Eibner (Heidelberg) Frank Falkenstein (Würzburg) Ralf Gleser (Münster) Bernhard Hänsel (Berlin) Alfred Haffner (Kiel) Svend Hansen (Berlin) Ole Harck (Kiel) Joachim Henning (Frankfurt am Main) Christian Jeunesse (Strasbourg) Albrecht Jockenhövel (Münster) Tobias L. Kienlin (Bochum) Rüdiger Krause (Frankfurt am Main) Klára Kuzmová (Trnava) Amei Lang (München) Achim Leube (Berlin) Andreas Lippert (Wien) Jens Lüning (Frankfurt am Main) Joseph Maran (Heidelberg) Wilfried Menghin (Berlin) Carola Metzner-Nebelsick (München) Johannes Müller (Kiel) Ulrich Müller (Kiel) Michael Müller-Wille (Kiel) Mária Novotná (Trnava) Bernd Päffgen (München) Diamantis Panagiotopoulos (Heidelberg) Christopher Pare (Mainz) Hermann Parzinger (Berlin) Margarita Primas (Zürich) Britta Ramminger (Hamburg) Sabine Rieckhoff (Leipzig) Wolfram Schier (Berlin) Heiko Steuer (Freiburg im Breisgau) Thomas Stöllner (Bochum) Biba Teržan (Berlin) Andreas Zimmermann (Köln) 179 Pottery from Southwestern Paphlagonia I: Terra Sigillata and Red-Slipped Ware ERGÜN LAFLI, GÜLSEREN KAN ŞAHIN (IZMIR) Summary This paper aims to give an overview of the 273 fragments of Roman fineware recovered during archaeological fieldwork at Paphlagonian Hadrianopolis between 2005 and 2008. So far the study of Hellenistic-Roman ceramics at Hadrianopolis seems to indicate the existence of a local industry that produced fineware between the 1st cent. B.C. and the 7th cent. A.D., imitating well-known types produced at other manufacturing centres in Asia Minor and the Black Sea region. The finds examined in this article are from more than 10 different areas at Hadrianopolis, excavated in 2006-2008. First the typology, fabrics, and chronology of identified wares are discussed. Two major find groups are terra sigillata and red-slip wares, almost all of which are of local origin. Our most important contribution to the archaeology of Paphlagonia will be the clarification of the Early Roman ceramic traditions in southwestern Paphlagonia, since our knowledge on Roman Paphlagonia is very limited and terra sigillata from Hadrianopolis seems to be the only material that will lead to a comprehension of the role of the region in the Roman period. Although the study of the ceramics has far-reaching implications in socio-economic and cultural terms, this paper will focus only on the ceramics as main data and will exclude historic sources on Roman Paphlagonia. Keywords: Hadrianopolis, Paphlagonia, north-central Turkey, terra sigillata, Early Roman, local production, coarse ware, Early Byzantium Zusammenfassung Dieser Beitrag gibt einen Überblick über 273 Fragmente römischer Feinkeramik, die bei den archäologischen Grabungen in Hadrianopolis in Paphlagonien zwischen 2005 and 2008 gefunden wurden. So weit aus der bisherigen Untersuchung der hellenistischen bis kaiserzeitlichen Keramiksorten in Hadrianopolis ersichtlich wird, müssen lokale Produktionsstätten, die vom 1. Jh. v. Chr. bis in das 7. Jh. n. Chr. Feinware herstellten, vorhanden gewesen sein. Diese lokalen Werkstätten ahmten weit verbreitete Typen anderer Produktionszentren in Kleinasien und der Schwarzmeerküste nach. Die hier untersuchte Keramik stammt von mehr als 10 verschiedenen Fundstellen in Hadrianopolis und wurde in den Kampagnen 2006-2008 ergraben. Die Untersuchung widmet sich der Typologie, den Fabrikaten und der Chronologie der identifizierten Waren. Zwei umfangreiche Gruppen bilden die Terra Sigillata und Ware mit rotem Überzug, die nahezu ausschließlich lokaler Erzeugung entstammen. Der wichtigste Fortschritt dieser Arbeit für die Archäologie Paphlagoniens besteht in der Typologisierung der frührömischen Keramiktraditionen im südwestlichen Paphlagonien, zumal unser Wissen über Paphlagonien in der Kaiserzeit noch sehr gering ist und die Terra Sigillata aus Hadrianopolis die einzige Fundgruppe zu sein scheint, die zu einem Verständnis der Rolle Hadrianopolis für die Region führen könnte. Auch wenn aus Keramikstudien weitreichende sozio-ökonomische und kulturelle Implikationen abzuleiten sind, versteht sich dieser Beitrag, vorerst unter Ausklammerung der historischen Quellen zum römischen Paphlagonien, als eine Grundlagenarbeit zur Keramikentwicklung in Paphlagonien. Schlüsselworte: Hadrianopolis, Paphlagonien, nördlich-zentrale Türkei, Terra Sigillata, frührömisch, lokale Herstellung, grobe Ware, frühbyzantinisch 180 Özet Bu makalede Paphlagonia Hadrianopolis’inde 2005 ve 2008 yılları arasındaki arazi çalışmalarında ele geçmiş 273 parça Roma dönemi ince seramiği konu edilmektedir. Hadrianopolis’te yapılan Hellenistik ve Roma dönemine ilişkin seramik çalışmaları İ.Ö. 1. yy.’dan İ.S. 7.’a kadar Anadolu ve Karadeniz’deki başka atölyelerin iyi bilinen formlarını taklit eden yerel bir ince seramik üretimi olduğunu göstermiştir. Bu makalede incelenen parçalar Hadrianopolis’te 2006-2008 yılları arasında kazılan 10 farklı bölgede ele geçirilmişlerdir. Bu incelemede belirlenen belli başlı seramik gruplarının tipoloji, malzeme ve kronolojileri tartışılacaktır. Tanıtılan başlıca iki grup, neredeyse tamamı yerel kökenli olan terra sigillatalarla kırmızı astarlı seramiklerdir. Roma dönemi Paphlagonia’sı ile ilgili bilgilerimiz oldukça kısıtlı olduğu ve bu dönemde bölgenin rolünün anlaşılması için elimizdeki tek veri terra sigillata olduğu için, çalışmamızın Paphlagonia arkeolojisine en önemli katkısı güneybatı Paphlagonia’daki erken Roma seramik geleneklerinin aydınlanması olacaktır. Aslında seramik çalışmaları Eskiçağ’daki sosyo-ekomik ve kültürel bazı bilgileri yansıtabileceklerine rağmen, bu makale sadece seramik konusuna ana veri olarak yer vermiş ve Roma dönemi Paphlagonia’sına ilişkin tarihi verileri şimdilik konu dışı bırakmıştır. Anahtar Kelimeler: Hadrianopolis, Paphlagonia, Orta Anadolu’nun kuzeyi, terra sigillata, Erken Roma, yerel üretim, kaba seramik, Erken Bizans. Ergün Laflı ist Abteilungsleiter im Seminar für mittelalterliche Archäologie an der Dokuz-Eylül-Universität Izmir. Er studierte in Ankara, Tübingen und Köln Klassische sowie Vor- und Frühgeschichtliche Archäologie. Sein Schwerpunkt liegt im Bereich der hellenistischen, römischen und spätantik-byzantinischen Keramikforschung. Gülseren Kan Şahin studierte an der Universität Trakya in Edirne, wo sie ihr Diplom in Klassischer Archäologie erhielt. Seit 2009 ist sie Promotionsstudentin an der Dokuz-Eylül-Universität in Izmir und arbeitet im Rahmen ihrer Doktorarbeit über das Thema dieses Aufsatzes. 181 Fig 1: Map of Paphlagonia. In the Roman and Byzantine periods Paphlagonia was an area on the north-central Black Sea coast of Asia Minor, situated between Bithynia and Pontus, and bordered by Galatia by the eastern prolongation of the Bithynian Olympus. Culturally, it was a contact zone between Greeks in the Black Sea area and the indigenous population of the Central Anatolian plateau. The region is the least well-known area with regard to Hellenistic and Roman ceramics in comparison with other countries located on the Black Sea coasts, namely Bulgaria, Romania, Moldovia, Ukraine, Russia and Georgia. The few Roman pottery studies that have been conducted in the region are not sufficient to draw an accurate picture of the ceramological heritage there (an attempt: LAFLI 2006). Some recent field work has provided results about the Hellenistic and Ro- man ceramic traditions in the region, such as studies at Sinope (FIRAT 2010), Tieion and Pompeiopolis (two recent studies on the pottery finds from the excavations at Pompeiopolis: Early Roman fine wares: ZHURAVLEV 2011; and Late Roman pottery: DOMŻALSKI 2011). In Pompeiopolis numerous “imported wares” were recognized which we did not identify in Hadrianopolis. Thus, Roman pottery from Hadrianopolis reflects a very different picture than the fine wares at Pompeiopolis in Paphlagonia. Excavations and surveys at Aizanoi (ATEŞ 2003), Pessinus (DEVREKER ET AL. 2003), Tavium (LAFLI 2003; WEBERHIDEN 2003), Amorium (HARRISON 1991 and 1992), Gordion (HENRICKSON 1992), Ancyra (KAYA 2005), Çadır Höyük and Kerkenes etc. have brought to light some evidence of local production in the Roman Central Anato- 182 Fig 2: Map of surveyed areas in southwestern Paphlagonia in 2005. lia. By the researches elsewhere in the Black Sea we know that the southern Pontic littoral (i.e. Paphlagonia and its eastern neighbour Pontus) was the major production centre for Pontic sigillata in the Roman times (KNIPOWITSCH 1929). Transport amphorae were produced in Paphlagonian coastal cities, such as Sinope, Heracleia, and Amastris between the 4th cent. B.C. and the 5th cent. A.D. in large quantities. Amisus was also an influential centre for coroplastic production beginning at the latest in the Hellenistic period. Less well known is the hinterland of Paphlagonia, especially its southwestern part where R. J. Matthews of the British Institute of Archaeology at Ankara directed field surveys between 1997 and 2001, producing some HellenisticRoman ceramic evidence. Hadrianopolis is located 3 km west of the modern town of Eskipazar, near Karabük. Geographically it was on the principal western route from the Central Anatolian Plain through the mountains to Bartın and the Black Sea. Thus, it lies basically in Roman southwestern Paphlagonia and on the borderline of the Galatian, Bithynian and Paphlagonian territories (fig. 1). It was a small but important site, controlling this major route and dominating a rich agricultural, especially vinicultural hinterland. It has been determined that the nucleus of the ancient city extended along the Eskipazar-Mengen highway for 8 km east-west and 3 km north-south (fig. 2). The chora of Hadrianopolis is much more extensive in size. At the site it was difficult to locate a classical settlement pattern of a typical Greco-Roman polis, since parts of the city are widely distributed in the area and they are far from each other. 183 Fig 3: Plan of Hadrianopolis in 2008. Southwestern Paphlagonia in the Roman period was a densely populated region. There were two major cities in the inland of Paphlagonia: Hadrianopolis and Gangra (modern Çankırı). Hadrianopolis, however, seems to be more intact than Gangra in its southern neighbourhood. It is the only large-sized settlement of inland Paphlagonia that can be excavated. We know that in the 1st cent. B.C. there was already a site on the location of Hadrianopolis, called Caesareia. Until the era of Hadrian this site was perhaps the regional capital; after Hadrian its name and probably status were changed. We do not know much about the Hadrianic era in Paphlagonia and its impact on ceramic traditions. In 2005 archaeological field surveys were begun at Hadrianopolis and its environs by a team from the Dokuz Eylül University, Izmir. As a result of our 2005 surveys it has been confirmed that Hadrianopolis was indeed located at modern Eskipazar, with finds dating from the 1st cent. B.C. to the 8th cent. A.D. In 2006 and 2007 excavations were carried out, and 2008 was our restoration season on the site. The field activities between 2005 and 2008 (on the field season 2006: LAFLI/ZÄH 2008; season 2007: LAFLI/ZÄH 2009; season 2008: LAFLI 2009a; on the surveys at Kimistene in 2005: LAFLI/CHRISTOF 2011; on all of the field seaons: LAFLI/CHRISTOF 2012; on Middle/Late Byzantine pottery: LAFLI/KAN 184 ŞAHIN 2010; and glass finds: LAFLI 2009b) identified the remains of at least 25 buildings (fig. 3). Among them are two bath buildings of the Late Roman-Early Byzantine periods, two Early Byzantine churches, a domus (formerly called as “villa”), a fortified structure of the Middle Byzantine period, a possible theatron, a vaulted building, a domed building, and some domestic buildings with mosaic floors. In general, these four campaigns have established that Hadrianopolis was a fortified regional centre during the Late Roman and Early Byzantine periods (5th-7th cent. A.D.), when it can easily be defined as a “polis” with civic buildings. It had a fairly large urban population as well as an extensive agrarian rural population in its chora. Most of the visible surface remains in southwestern Paphlagonia belong to this period. Roman and earlier remains seem to consist almost exclusively of inscriptions, rock-cut graves, some cultic monuments, and a small amount of pottery (including sigillata), but no civic buildings. It seems that the city was abandoned during the second quarter of the 8th cent. A.D. Discoveries were also made out in the chora of Hadrianopolis. An area within a radius of 20 km around Hadrianopolis was extensively surveyed and more than 30 archaeological sites with numerous functions were discovered. These sites were höyüks, cemeteries, single graves or grave groups, tumuli, stone quarries, farmyards, villages, sanctuaries, and other remains. From these settlements pottery was the most significant find group in dating and characterising the settlement patterns. In the chora of Hadrianopolis the most visible and largest Hellenistic and Roman site is Kimistene whose settlement status is not definable yet. Description of the Deposits The fineware information from Hadrianopolis is derived from the following areas: Bath A, Domus, Bath B, Apsidial Building, Western Tomb as well as field surveys (cf. table below). Since the excavations have concentrated more on the Early Byzantine period, most of the Late Hellenistic-Roman material was collected from non-stratified contexts that are frequently residual, with a mixture of infills from later periods. In very few excavated areas we have discovered fills dominated by earlier pottery. One example is the Room 9 of Bath A, where a large number of fragmentary sigillata was recovered. Our main criterion for the chronology of deposits was numismatic evidence. The Bath A was a monumental building located in the southern part of the city, ca. 130 m southwest of Early Byzantine Church B, which was excavated in 2003 by the Archaeological Museum of Ereğli. In 2006 and 2007 a great part of the structure was excavated revealing 15 rooms. One can assume that Bath A has five archaeological phases, i.e. A.D. 475-525, 525-575, 575-675, 675725 and 725 (i.e. Arab invasions to Gangra and to Hadrianopolis) to 19th cent. In this latest phase Bath A must had been converted into a barn. It was the most important find spot for ceramics and other small finds (glass, bronzes) in our study. The pottery assemblages from the rubble layers of the Bath A contain a great quantity of sherds of Late RomanEarly Byzantine cooking pots as well as some terra sigillata. The largest assemblage is Late Roman-Early Byzantine coarse ware. The deposit ranges from 1st cent. A.D. to 19th cent. A.D. Finds after the 8th cent., however, are very sparse and singular. The assemblages 185 contain very few (almost 10 % of whole assamblages) sherds earlier than 5th cent. A.D. Almost all the sherds of the 1st to 5th cent. A.D. originate from the construction level of the Bath A. It is very possible that the layers are not well-sealed and that there is some stratigraphical disturbance within them. The Domus provided us numerous examples of a specific red-slipped bowl form, with or without slip. Based on numismatic evidence this deposit is typically an early 6th-cent. A.D. deposit, although it contained some earlier and later sherds. It must had been in use until the end of the 7th cent. A.D. Other excavated areas in Hadrianopolis with sigillata finds are as follows: Western and Monumental Tombs (finds from the 1st to 3rd cent. A.D.; with some later sherds) as well as Church B and Absidial Building (finds from the 6th to 8th cent. A.D.; with some earlier sherds). In the west of the township of Eskipazar, near the village of Deresemail and the hamlet of Değirmenbaşı there is a ridge called Asartepe with four hilltops, the ancient site of Kimistene, identified by Strabo (12.3.41; for Kimistene: LAFLI/CHRISTOF 2011). On the northern slope of its Acropolis, on an artificial terrace, lie the remains of a rural sanctuary from the Roman times. All surface remains, the retaining walls of the temple terrace, the rectangular foundations of the temple and of its cella as well as the monumental access staircase leading to the terrace were documented during the 2005 field survey. The stylistic classification of the architectural blocks belonging to the temple reveals at least two phases of the temple construction, one in the late 2nd/early 3rd cent. A.D., and a second, perhaps only a repair phase, during the 3rd cent. A.D. Field surveys at Asartepe covered three further areas or “peaks” along the mountain ridge in addition to the Acropolis: Necropolis 1, Necropolis 2, and the Cistern. Most of the material was collected from the Acropolis. The site clearly shows occupation over a long period, with particularly important phases in the Hellenistic and Roman periods. Surface pottery ranges from the Late Iron Age into the Middle/Late Byzantine period. Although most of the pottery seems to have been locally produced, there are a few imports present. The large amount of pottery confirms the importance of the site, already known from architectural, sculptural and epigraphic finds. A further surveyed area is Gölatarkası, a site ca. 3 km east of Hadrianopolis, between Hadrianopolis and Kimistene. This site yielded some Late Roman red-slipped sherds, especially sherds of 4th-6th cent. A.D. A further site with earlier sigillata (i.e. 1st2nd cent. A.D.) is Boncuklar - a cemetery site -, ca. 10 km southeast of Hadrianopolis. Some inscriptions from the 3rd cent. A.D. were found here. Also to be mentioned is a burial site with rock-cut tombs and a cistern, called Kepez, ca. 12 km southeast of Hadrianopolis where we discovered an homogenous Hellenistic pottery assemblage. Almost no Roman pottery was discovered at this site. 186 Bath A Surface Finds from Hadrianopolis Boncuklar Domus Elsewhere at the Acropolis of Kimistene Acropolis of Kimistene, southern slope Acropolis of Kimistene, eastern slope Acropolis of Kimistene, northern slope Acropolis of Kimistene, western slope Necropolis of Kimistene Göletarkası Western Tomb Monumental Tomb Church B Bath B Absidial Building Kepez Other find spots 113 26 25 22 20 25 7 3 2 3 5 5 3 3 7 1 1 2 Table: Terra Sigillata and red-slipped ware sherds from each place. Pottery Finds During the four seasons of field research between 2005 and 2008 a wide range of ceramic materials was collected from both field surveys and excavations; a total of ca. 1,500 fragments was examined. In this article survey and excavation finds will be examined together. The pottery finds from Hadrianopolis do not enable us to make assumptions on the population of this landscape. The main ceramic groups of Hadrianopolis during the Late Hellenistic, Roman, and Late Roman-Early Byzantine periods are fineware, coarse ware (cooking ware, painted kitchen ware, storage vessels etc.), very few transport amphorae, oil lamps, unguentaria, pithoi, miscellenous terracotta objects, tiles, and bricks as well as water pipes. Generally the surface and stratified collection of local and imported ceramics dating to these periods (1st cent. B.C. to early 8th cent. A.D.) represents an isolated typology with a scanty number of wares and forms with less quality. Through the quantification of the material and examination of clay there is strong evidence to suggest that the great majority of these vessels of all periods were manufactured in southwestern Paphlagonia. Local origin of the material, however, was not proved by mineralogical analysis of the clay. The majority of the pottery found in Hadrianopolis belongs to Late Roman-Early Byzantine coarse wares that include pottery utilised for domestic storage, all phases of food preparation and service, containers and other functions. Fineware of the period between the 5th and 8th cent. A.D. is less in number, perhaps 187 because of the consumption of glass vessels. Generally, the forms of the coarse wares from Hadrianopolis tend to be very conservative and restricted to a few certain types. Kitchen pottery, for instance, is represented by casseroles, pots, frying pans, troughs, basins, pitchers, jugs, juglets and trefoil-jugs with a very limited and slow morphological evolution (such observations were also made in Gordion: HENRICKSON 1992). A small number of very unique local vessels feature a variety of decorations, such as red-brown to brown painting on the flat tondos of the open shapes (fig. 4a-b). Very few transport amphorae, perhaps of local origin as well, were recorded at Hadrianopolis which is puzzling considering the prolific viticulture of the region. In our pottery study we have put all the sherds earlier than the Roman period into a “PreRoman” classification, forming twelve main groups from pre-Iron Age, Iron Age, Hellenistic (on Iron Age and Hellenistic pottery from southwestern Paphlagonia: LAFLI/KAN ŞAHIN in print b) and Late Hellenistic periods. Definable wares of these four periods are as follows: - Pre-Iron Age sherds (from the höyüks in the area), - Iron Age grey ware, - Iron Age painted ware, - Iron Age coarse ware, - Pontic black glazed scyphos fragment of northern Black Sea origin (4th cent. B. C.), - Hellenistic painted ware, - Hellenistic relief ware, - Hellenistic burnished ware, - Red-painted Kepez group, - Late Hellenistic-Early Roman grey ware, - Late Hellenistic-Early Roman brown-slipped ware, - Hellenistic coarse ware. Most of these earlier, so-called pre-Roman groups were found in Kimistene. The groups of Roman finewares in southwestern Paphlagonia are as follows: - Late Hellenistic-Early Roman grey ware, - Late Hellenistic-Early Roman brown-slipped ware (a local alternating Late Hellenistic group of mostly closed vessels for burials, imitating earlier “Greek” forms), - Terra sigillata, - Red-slipped ware, - Non-slipped fineware, - Others. In Hadrianopolis the earliest Roman fineware is represented by grey ware, dating from 1st cent. B.C. to 2nd cent. A.D. (i.e., for the proto-sigillata phase), which was almost exclusively locally produced, following the Late Iron Age (so-called “Phrygian”) tradition in production technics, i.e. manufacture and firing, and clearly also in the development of types. S. Mitchell believed there was no Hellenistic phase of Hadrianopolis (MITCHELL 1993, 93). But we have few Late Hellenistic sherds from Hadrianopolis. It seems that the Late Hellenistic and pre-Hadrianic periods at Hadrianopolis were characterized by grey ware as well as brown-slipped ware, and there are extremely few sherds in Hadrianopolis earlier than the 1st cent. B.C. Original “Phrygian” grey ware of Late Iron Age has been found at the nearby settlement of Kimistene, the nature of which is not known in detail. Since Kimistene and other nearby settlements in Hadrianopolis’s chora also yielded Bronze and Iron Age pottery, it is surprizing to observe the physical similarity in appearance between Late Hellenistic-Early Roman grey ware in Hadrianopolis and 188 earlier grey wares in its chora. Also brownslipped ware is a close relative of local sigillata in terms of fabric, shapes, slip technics applied etc. In our studies, 273 Roman fineware sherds (terra sigillata and red-slipped wares) in total were examined. The number of sigillata sherds in the entire ceramic material from Hadrianopolis is less than 20 % of the total. Our study is based on fragmentary pieces, since no intact sigillata vessel was found. No grave pottery has been examined in context yet. Their preservation is mostly fine, perhaps because they were never in use as daily vessels. a Early Roman Imported Finewares The inland location of Hadrianopolis in southwestern Paphlagonia means that the quantity of imported pottery is very limited (ca. 2 % of the whole amount of Roman finewares). These are likely to include Eastern Sigillata A of the 1st cent. B.C. (pl. 1, nos. 1-4) and African red-slipped or Pontic red-slipped of the 3rd cent. A.D. (pl. 12, no. 64). ESA is described by J. W. Hayes, P. Kenrick, and K. W. Slane (HAYES 1985, 9-10; KENRICK 1985, fabric 223; SLANE 1997, fabric 267-70) supporting a source in northern Syria (SLANE 1997, 272). The colour of the fabric of the Hadrianopolis sherds belongs to the yellower end of the spectrum, with red-brown to red slip, normally associated with the pre-Augustan period (1st cent. B.C.). Pontic sigil- b Fig. 4a-b: An Early Byzantine painted dish from the Domus. 189 lata is also described by Hayes (HAYES 1985, 92-96, pls. XXII-XXIII ), and only one sherd in southwestern Paphlagonia seems to be of Pontic origin. All five imported sherds were imitated by the local potters. Local Sigillata and Red-Slipped Ware At Hadrianopolis there is not a rich variety of pottery of Roman finewares: between the 1st cent. B.C. and 2nd cent. A.D. we have a terra sigillata tradition, in the 3rd and 4th cent. A.D. red-slipped ware, and after 4th cent. A.D. redslipped forms continued to be produced until the end of the 7th cent. A.D., but in unslipped versions. The majority of the local sigillata were represented as very fragmentary small sherds, ranging between 1 and 7 cm (in average 2 cm) and making them difficult to compare to the standart corpora. The majority of terra sigillata and red-slipped ware from the Early Roman to the Late Roman periods occur in a single fabric likely to be a local or regional production. Taking into account their form, slip and fabric one can easily assume that the local sigillata from Hadrianopolis can be classified according to the common shapes of Eastern and Pontic sigillata. Where these sigillata are produced or how they came to the city has not yet been explained. The later red-slipped ware in Hadrianopolis was documented to a lesser extent; in the surrounding area of Hadrianopolis this ware emerges more frequently. In Hadrianopolis it seems that coarse wares do occur in greater amounts first in Late Roman period. Fabric of the Local Sigillata at Hadrianopolis In various aspects the fabric of the Hadrianopolis sigillata is similar to the local sigillata of Ancyra as well as of Tavium, all of which still remain sparsely known by the ceramic surveyors. Local terra sigillata from southwestern Paphlagonia is distinctive and a product of the fast wheel, which has left wheel marks on the surface. The wall thickness of the terra sigillata varies between 0.2 and 0.6 cm and red-slipped ware between 0.5 and 0.9 cm. Generally, vessels have simple contours. Red-slipped vessels are, however, consisting of large forms with thick walls of coarse fabric and tend to have a poor finish. The firing temperature of Hadrianopolis sigillata is always very high. Their hardness degree at Mohs’ chart is “3”. The clay is fine and average quality in character [fine clay with intensive light red (2.5YR 6/8 -6/6), reddish yellow (5YR 6/6-7/6) and red (10R 5/6-5/8, 2.5YR 5/8)]. The rest of the terra sigillata samples are light red (2.5YR 7/6,10R 6/6-6/8) and red (2.5YR 5/6). They are rarely pink (7.5YR 7/4), red (2.5YR 5/6) and light reddish brown (5YR 6/4). In eight samples there are colour differences because of their uneven firing. As for inclusions, calcite, mica, and sand were applied; grog was rarely used. These inclusions are mostly fine grained and were used together; but in some samples none of them were observed. The average dimension of inclusions is 0.2-0.4 mm. The density of inclusions is average. The clay is well-sieved. In some samples the surface is slightly porous. Identified slip colours differ according to the typology of vessels and date of the product. Their colour resembles the clay colour. Slip- 190 ping technics are mostly the same; it differs only in some sherds. Most popular slip colours on the exterior are red (10R 5/8-4/8-5/6, 2.5YR 5/6-5/8); rarely red (2.5YR 4/6) and light red (2.5YR 6/8). In the interior red (10R 5/8, 2.5YR 5/8) is the most frequent colour. In the unslipped areas it is light red (2.5YR 7/6) and pink (5YR 7/4, 7.5YR 7/4). In open vessels the slip was applied more elaborately. More often the slip was spaced out erraticly, and the slip of interior and exterior sides differs from each other. In most cases sigillata has a look of bichrome painting and contains less metallic sheen. In some cases exterior surface was slipped only in limited areas, in rim parts or in the middle part of the vessel. Exterior surfaces of bases in both open and closed vessels were left unslipped. In some cases only rim parts were slipped. In later red-slipped wares the slip is mostly washed out and calsified. Forms Few forms were produced all of which belong to the most common ones in the Roman Mediterranean. The forms are influced by many of the giant fineware industries throughout the Mediterranean and the Black Sea. In the absence of kiln finds or secure provenience for the terra sigillata found at Hadrianopolis as well as of parallels from nearby sites, it is unwise to define our fabrics or to date them too closely. There is almost no unusal shapes. The forms of the terra sigillata are grouped first into “open” and “closed” categories, of which open and simple forms are more common than closed and complicated ones (more than 60 % of the shapes are open). The main forms of terra sigillata from Hadrianopolis are cups, bowls, dishes, plates and juglets. Large dishes and plates were not frequently consumed in Hadrianopolis. It seems that the sigillata from Hadrianopolis was concentrated on thin-walled vessels and shapes smaller than 20 cm, probably because they were perceived as alternatives to contemporary glass and metal vessels of which we have very few Early Roman finds. No deep bowls or wide forms, such as containers, were discovered. Some of the forms (cups, bowls and juglets) are most common in Hadrianopolis and they reflect a “Paphlagonian-Bithynian” character in their style. Some local pre-Roman forms continued to exist into Roman times in sigillata form. The survival of bowls with simple incurved rim characteristic of the Hellenistic period in a red-slipped form is asthonishing and could be classified as a further evidence for the continuation of pre-Roman pottery manufacture tradition during the Roman era. It is difficult to reconstruct the whole body form only through the rim sherds. Also base, handle and body fragments are difficult to assign to any existing forms, but at least we were able to classify them according to “open” and “closed” shapes and most likely handles should be associated with jugs, since it seems to be the main closed shape among the sigillata. Decoration The finds are mostly without any stamped or painted decoration except for their exterior slip. Stamped decoration is only found on 2 % of the whole assemblage. The most common decoration on the sigillata are impressed geometric designs. These are flutings, ripples, notches, stamps, bands, dotted lines as well as combed, punctuated or rouletted decorations, 191 circles and other motives arranged horizontally, vertically, centered or in panells. The occasional decorated rim ledge associated with African red-slipped ware on round or rectangular bowls and dishes were also collected. In contrast to Ancyra none of these local vessels bears potters stamps. Chronology The chronology of southwestern Paphlagonian terra sigillata is problematic since we can only deal with very fragmentary material with weak stratification. It seems reasonable to assume that the sigillata at Hadrianopolis is concentrated to the 1st and 2nd cent. A.D. Cups and bowls should have been used as grave gifts of the 1st cent. A.D. (an instance: A grave group at the British Institute of Archaeology at Ankara, collected by J. Mellaart from Mengen-Bartın in southwestern Paphlagonia - figs. 5a-c). After the 3rd cent. A.D. the local red-slipped tradition began, continuing until the late 7th cent. A.D. Sigillata from Hadrianopolis is the evidence of a local red-slipped ware industry in southwestern Paphlagonia, northwestern Galatia, and eastern Bithynia. Evidence for manufacturing sites, manufactures, kiln sites or kiln wasters in or outside our study area, however, has not yet been discovered. Nor has overfired fineware been found so far. All of the sigillata is made of similar local clay, but it has not yet been possible to perform laboratory analyses of the composition. The peak of this quality ware of Roman type seems to fall in the 1st cent. A.D., a period characterised by important historical changes on the northcentral Anatolian plateau and among the population of the region. Production of smooth and shiny surfaced sigillata continued without any break until the 3rd cent. A.D. Later pieces (4th and 5th cent.) do not have the same shiny slip, and therefore should be attributed rather to “red-slipped ware”. Examples of this type of local sigillata are also known to the authors as unprovenienced examples in local museums in north-central Turkey (Museums of Çankırı, Kastamonu, Amasra, and Ereğli). In particular, Roman pottery collections at the Museum of Çankırı derive mostly from burials in this region (figs. 6a-b). We would like to construe the model of local production and distribution of sigillata from Hadrianopolis as “self production and consumption in a limited area”. Limited and isolated typologies and lacking evidence of Fig. 5a-c: Fragments of an Early Roman local sigillata juglet found by J. Mellaart in MengenBartın. 192 archaeological finds one can suppose that shortly after the establishment of Roman political presence in the region, the inhabitants of Roman Paphlagonia began to use sigillata at least for their burials and abandoned their former tradition of Phrygian grey ware at the latest in the 2nd cent. A.D. Form Catalogue Terra Sigillata (pls. 1-11) –Imported Sherds (pl. 1) Fig. 6a-b: Roman sigillata bowls at the Museum of Çankırı; fig. a - inv. no. 145-9.4.72; fig. b - inv. no. 236-73.4.2. import ware are further indications in favour of this type of local economy model. Because of the long distance to the coast and the commercially poor situation, this area reflects a reserved and closed economic system in Roman Asia Minor. Through these local products one receives the impression that the Roman tableware industry in Paphlagonia, in a relatively short period of time (from the 1st cent. B.C.), developed parallel to other manufacturies in western and southern Asia Minor. Perhaps the presence of a large population in southwestern Paphlagonia at the beginning of the 1st cent. A.D. from various parts of the Roman Empire resulted in an active ceramic industry. Through Only four sherds were recorded; three rim fragments as well as a body sherd that is probably belonging to one of the rim sherds. All of these fragments originate from Kimistene. Very few imported sherds were found at Hadrianopolis. Thus one can question the role of Kimistene in terms of imported material. Perhaps its religious status has an impact on the materials’ routes. All of these sherds are ESA products, and not Pontic. Thus, one can assume that Late Hellenistic southwestern Paphlagonia had relationships to the south. ESA Bowl = Hayes 2008, Form 51 (pl. 1/no. 3) Half of a globular body with an incurved rim was recorded. On the interior side a mottled slip. Pink fabric. This fragment was found in Kimistene, at its Acropolis. Typologically similar bowls were classified by Hayes into two groups. At Benghazi similar shapes have a diameter, ranging between 9.8 and 13.5 cm. Our sample has a dia. of 13.4 cm. Dated by Hayes to A.D. 70-120 (HAYES 1985, 37, form 50, pl. VI, nos. 19-20); a closely similar date is given by Kenrick, based on the finds 193 Cup (pl. 2) from Benghazi (KENRICK 1985, 239, fig. 43, 337.1). ESA Bowl = Hayes 2008, Form 12 (pl. 1/no. 1) One example was found in the field surveys at Hadrianopolis. Part of a small bowl with short flat and everted rim. Exterior slip mottled. Pink fabric. In Tel Anafa most of the pieces have a narrow straight rim and are found in ROM 1B stratum (SLANE 1997, 303). Similar fragments at Benghazi were found in the levels of the middle Augustan period and dated to the early 1st cent. B.C. (KENRICK 1985, 418-419). Hayes dated them from 50 B.C. to the beginning of the 1st cent. A.D. (HAYES 2008, 25; HAYES 1985, 19, form 12, pl. II, no. 10). ESA Dish, Hayes 2008, Form 4 (pl. 1/no. 2) Two dish fragments belong to Hayes Form 4. The profile of their rims are convex. In this form rosettes, rulet and palmets are very common on their tondos. Their slip is red in colour and non-porous. Fine clay with very few inclusions. Hayes identified them as “small plate” and dated them to the 1st cent. B.C. (HAYES 2008, 24, form 4, fig. 2, no. 32; HAYES 1985, 16, form 4A, pl. I, no. 9). Local Products – Open Forms (pls. 2-8) As mentioned above this group consists of cups, bowls and dishes. There are furthermore some body and base fragments. Only one fragment belongs to this cup form: a thin-walled cup with everted rim of a possibly half globular shape in the manner of the brown slip tradition of Late Hellenistic period. Shiny brown-reddish slip from the rim to the middle of the exterior body. The rest of the body left unslipped. At Tel Anafa, some smaller samples were dated to the first quarter of the 1st cent. B.C. (SLANE 1997, 299). Bowls (pls. 3-5) Bowl Form 1 (pl. 3) Nine fragments belong to this group. This form (especially nos. 6 and 9-11) is actually a derivation of the most common hemispherical Hellenistic cup form (GUNNEWEG ET AL. 1983, 96, fig. 21.4) as well as of the ESA 1 2 3 Pl. 1: Nos. 1-3 - Imported terra sigillata sherds; no. 1 - ESA bowl fragment; no. 2 - ESA dish fragment; no. 3 - ESA bowl fragment. 194 form 25 in Tel Anafa (SLANE 1997, 309, pl. 18, FW 190). It is interesting to find its original ESA shape with local imitations in southwestern Paphlagonia. No. 13 has a concave profile with an incurved rim. Usually, it has a simple rim and ring base. Its low wall curves gently and continuously to end in a plain, upright rim flattened on the top. In intact forms its base is thickened on the outside. Similar shapes at Pergamon were dated to the mid 1st cent. B.C. and beginning of 2nd cent. A.D. (MEYER-SCHLICHTMANN 1988, 135, pl. 17, 266. T4). This form is also present at Pompeiopolis (ZHURAVLEV 2011, 12, fig. 2, 5-7; dated to 2nd and 3rd cent. A.D .). The colour ranges between reddish brown and red. The non-porous slip ranges between reddish brown and red. Diameters are 13 to 17.6 cm. Bowl Form 2 (pls. 4-5) This bowl form is also an imitation of ESA Samaria Form 10. Nos. 15, 16, 18, 22, 25 and 31 have a flat or beaded rim and curving wall. Nos. 24 and 30 have a more vertical wall different from the others. Nos. 24 and 14 have a beaded rim, and nos. 16 and 20 have an everted rim. Nos. 18, 19, 22, 26, and 29 have a round or globular body form with inward sloping mostly s-profiled wall and thin rim. In some samples the curving wall ends in a thickened rim flattened on the top. This is actually an earlier form which was used in central Anatolia since the Phrygians; it seems to be very popular in Early Roman southwestern Paphlagonia. Among later samples from the Athenian Agora there is a similar one, identified by Hayes (HAYES 2008, 30, fig. 7, nos. 180 and 184). The rim diameters differ between 26.4 and 11.2 cm. The ware colour is light reddish brown and light brown. Dishes (pl. 6) Dish Form 1 (pl. 6/nos. 32-35) A shallow dish with everted rim. The most distinctive typological feature of this dish form is its angled and raised rim. Ca. 2 or 3 cm below the rim there is a further angle which gives the wall an s-profile. Thus, this form resembles Hayes 1985, 29, form 33, pl. V, no. 7. The diameters differ between 16.6 and 17.4 cm. Early Italian sigillata forms are dated to the beginning of the 1st cent. A.D. and the later Italian series are dated by Hayes to A.D. 70-150 (HAYES 2008, 41). Similar forms from Belbek IV Necropolis in the southwestern Crimea were identified by Zhuravlev as “Pontic sigillata” and dated to the last quarter of the 1st or the beginning of the 2nd cent. A.D. (ZHURAVLEV 2009, 36, fig. 4, form 2.2, no. 25). Dish Form 2 (pl. 6/nos. 36-38) A dish with a wide, everted and decorated rim. Smaller as well as larger forms are re- 4 Pl. 2: No. 4 – Local sigillata, a cup fragment. 195 5 6 8 9 7 10 11 12 13 Pl. 3: Nos. 5-13 - Local sigillata, fragments of bowl form 1. presented. The decoration – even on the outcurved rim – consists of dots, waves etc. The ware contains lime and sand inclusions. Dish Form 3 This is represented only by one single sherd with a 3 cm wide and 1 cm thick distinctively everted rim and a sharply descending wall. Mottled surface in the interior. Light red fabric with some tiny sand inclusions. It was found in Bath A, Room 2b. Dish Form 4 (pl. 6/no. 39) Dish form 4 is also represented by a single sherd. It has an everted, bowed rim and a shallow body form with a gently descending wall. Its form is similar to that of the Hellenistic fish plate. The outside of the rim has a mottled black slip decoration. The rest of the body surface is red. Its fabric is light red in colour. Its diameter is 28.8 cm. It was found in Bath A, Room 5. Zhuravlev dates this form to the 1st-3rd cent. A.D. (ZHURAVLEV 2002, 256). 196 15 14 17 16 18 19 20 21 22 23 24 Pl. 4: Nos. 14-24 - Local sigillata, fragments of bowl form 2. 197 25 26 27 28 29 30 31 Pl. 5: Nos. 25-31 - Local sigillata, fragments of bowl form 2. Plate (pl. 7) Three fragments belong to a plate form with a flat base on a ring foot (WILLIAMS 1989, 13, fig. 5, no. 64). Its surface is smoothed. The slip is red in colour all over the surface except for the foot which remained unslipped. A light reddish brown line in the middle may be result of bad firing or secondary firing. These fragments were found in Bath A. The diameter should be ca. 20 cm. On its tondo is a medaillion with some further decoration patterns consisting of seven recognizable letters in the form of a monogram, and in the middle part a cross and three boxes. In the frame of the medaillion there are two more letters. 198 32 33 34 35 37 38 36 39 Pl. 6: Nos. 32-35 - Local sigillata, fragments of dish form 1; nos. 36-38 - dish form 2; no. 39 - dish form 4. 199 Transcription: In the frame: [-----]HΘ[-----] Sherd c (tondo): [----------------] AKΔ†YLOY Ψ Three boxes Ε have everted rims. From the present body of fragments one can assume that these juglets should had a wide lower body. Nos. 46 and 52 are narrowed to the rim. The fabric has a red to reddish yellow colour and is without visible inclusions. The rim diameters differ between 5.0 and 8.4 cm. Similar rim sherds are published in SLANE 1997, 321, pl. 23, FW 250; and KENRICK 1985, 310, pl. 58, B 450, 452. Base Fragments of Open Form (pl. 8) Handle Fragments of Closed Forms (pl. 11) Four fragments belong to this open form. Some base fragments cannot be assigned to any known forms besides obviously belonging to open forms (dishes or bowls?). They are important in reflecting the range of colours applied at the local sigillata of Hadrianopolis. They have a tiny slip on their exterior and interior surfaces, but the foot is unslipped. The fabric colour varies from red to reddish yellow. The base diameter ranges from 4.6 to 10.0 cm. One fragment was found in Kimistene, the others were found at the Bath A and B. Two type of handles have been observed: plain and grooved. Body Fragments of Closed Forms The fabric colour is reddish yellow to red and the exterior slip reddish brown to red. Rim Fragments without any Form Definition None of the 76 fragments can give us any clue to their forms. The fabric colour varies between reddish brown and red. Most of them are thin and both sides slipped. Their slip varies in colour between reddish yellow and red. The paste colour varies between light red and red. Most of them do not have any visible inclusions. A major part of them originate from Kimistene. Closed Forms - Juglet (pls. 9-11) Lamp Fragments in Sigillata Tradition The form is characterised by a rim with slight waist and plump body below. The outer lower edge of the everted rim is thickened with a groove on the exterior surface (nos. 49-50). The neck is conical. Nos. 44 and 48 Only two examples belonged to this form. They are indications of lamp production in local sigillata tradition. Their fabric is of the same colour and their exterior surface has red slip. Body Sherds of Open Forms 200 Pl. 7: No. 40a-c – Local sigillata, three fragments of a plate. a-b c a-b Red-Slipped Wares (pls. 12-22) -– Imports (pl. 12) Local Red-Slipped Wares – Open Forms (pls. 13-21) African Red-Slipped Bowl = BONIFAY 2004, 181, type 55, 180, fig. 96, no. 4 (pl. 12) There are numerous fragments of bowls, and very few dishes and plates which are only represented by base fragments. It seems that the cups and small bowls known in the previous periods were already forgotten. Only one rim sherd could be determined as “African red-slipped ware” (similar to: PAZ/ VARGAS 2011, 90, no. 13, fig. 3.). It is a bowl form with convex wall curving up to a rounded rim thickened on the outside. The exterior and the inside of the rim is slipped and its slip is flaking on all of the surface. The fabric is hard and red with limestone inclusions. Bowls (pls. 13-19) Mostly large bowl forms were made. They have taken the 3rd cent. bowl form of the Phocaean red-slipped ware as a model and they continued to use this form until the 6th-7th cent. A.D., but in unslipped versions. These forms do not change for a long period of time. The potters behind these bowls imitate several workshops, mostly concentrating on later products, such as African and Phocaean 201 Pl. 8: Nos. 41-43 - Local sigillata, base fragments of open forms. 41 42 43 red-slipped wares. Recently J. Poblome and N. Fırat attempted to understand relationships between African, Cypriote, Sagalassian and other eastern red-slipped productions (POBLOME/FIRAT 2011). It seems that the domus has a great variety of 6th and 7th cent. A.D. bowls. Bowl Form 1 (pl. 13) Five fragments belong to this shallow bowl form. Their rims are thickened on the exterior and connected to the body with a concave profile. Their slip colour ranges from reddish brown to red. Their slip is thin and matt. The fabric is reddish yellow, yellowish red, and red with sand, lime and micaceous inclusions. Rim diameters range from 22.0 to 25.2 cm. The form is similar to some Sagalassian shapes of the 4th cent. A.D. (POBLOME 1999, 309, type 1C160, fig. 61, no. 6 and type 1C170, fig. 63, no. 1, 3). In Amorium it has been dated to the late 4th or 5th cent. A.D. (HARRISON 1991, 224, fig. 6, no. 4, 8). Bowl Form 2 (pl. 14) Only one example was found in the domus. This rim fragment has a groove on its top. Its lip is thinned. Red slip covers all of its surface. Its fine fabric is red in colour. Rim diameter is 15.6 cm. Similar forms were noted in Sagalassus during the first half of the 5th cent. A.D. (POBLOME 1999, 306 and 368, fig. 31, no. 5). Bowl Form 3 (pls. 15-17) The largest group is represented by 19 examples with minor typological differences. The first example has a broad flange on the exterior of the rim. The others have vertical rims, generally concave on the outer surface, thickened on the outside and has a rounded lip (HAYES 1972, 329). The wall is flaring, curved or slightly angled. Similar shapes to no. 72 were discovered at Sagalassus, Perge and Hierapolis and dated to the 6th cent. A.D. (POBLOME ET AL. 2001, 126, fig. 4, no. 7), whereas Hayes dates simi- 202 44 45 46 47 48 49 50 51 52 Pl. 9: Nos. 44-52 – Local sigillata, juglet fragments. 203 53 54 55 56 57 58 59 60 61 Pl. 10: Nos. 53-61 - Local sigillata, base fragments of closed forms. 204 lar examples to the second half of the 5th and first half of the 6th cent. A.D. (HAYES 1972, 337-338). In Pompeiopolis, K. Domżalski has determined similar sherds as “Late Roman C/Phocaean Red Slip ware” and dated them to the late 5th-early 6th cent. A.D. (DOMŻALSKI 2011, 10, fig. 11, Form 3). This form is dated on Thasos to the second half of the 5th cent. A.D. (ABADIE-REYNAL 1992, 19, fig. 3-4, nos. 64-75), at Ephesus to the late 5th and early 6th cent. A.D. (GASSNER 1997, 143-144, figs. 558-562) and in Amorium to the 5th cent. A.D. (HARRISON 1992, 216, fig. 5, no. 19). Similar forms were discovered at Çerkeş/Kızıllar (Antoninopolis?), Bozoğlu/ Çerkeş, Yazı/Ilgaz, and Kuzören/Eskipazar (MATTHEWS ET AL. 2009, 218, fig. 6.103, nos. 1-3 from the site PS066; no. 7 from PS074; no. 12 from PS115; and fig. 6.104, nos. 4-5 from PS157). Its slip varies in colour between reddish brown and light red and is applied only in a limited area between the inner surface and the rim part of the bowl. It is fine and matt. Its fabric is reddish yellow to red with sand and lime inclusions. Most of them were found at Domus. Rim diameters vary from 18.0 to 30.8 cm. Bowl Form 4 (pls. 18-19) This form may be a continuation of the tradition of carinated bowls. It is definitely not a homogenous type. Generally its walls have a complex profile. Most have had a large diameter. Sometimes the bowls with straight rim have a groove on the upper exterior similar to nos. 89 and 90 which were found in the Western Tomb. No. 92 has an inverted rim and its lip is cut on the exterior. 62 63 Pl. 11: Nos. 62-63 - Local sigillata, handle fragments of closed forms. Fragments of a bowl form with vertical incurved rim (HAYES 1972, 325, LRC, Form 1A, fig. 65) come from the Agora deposits in Athens and are dated to the late 4th-early 5th cent. by Hayes. Dish (pl. 20) We have only one dish fragment belonging to this form. The everted and rounded rim has a painted dark reddish to brown band on its interior surface. This form was popular in the second half of the 1st cent. A.D. (POBLOME 1999, 308, Variant 1C 132, fig. 56); but in Sagalassus it is common until the 3rd cent. A.D. In Aizanoi it is known in the 3rd-4th cent. A.D. (ATEŞ 2003, 150, Teller 12, pl. 44, no. 246). Base Fragments of Open Forms (pl. 21) These base fragments have a curved floor. No. 101 has a stamped decoration on its floor and a tapering foot resembling Hayes’ Late Roman C, form 1A, dated to the late 4th and early 5th cent. A.D. (HAYES 1972, 327, LRC, Form 1A, fig. 65). 205 64 Pl. 12: No. 64 – Fragment of an African red-slipped bowl. 65 66 67 68 69 Pl. 13: Nos. 65-69 – Local red-slipped ware, fragments of bowl form 1. 70 Pl. 14: No. 70 - Local red-slipped ware, fragment of bowl form 2. 206 71 72 73 74 75 76 Pl. 15: Nos. 71-76 - Local red-slipped ware, fragments of bowl form 3. 207 77 78 79 80 81 Pl. 16: Nos. 77-81 - Local red-slipped ware, fragments of bowl form 3. 208 82 83 84 85 86 87 88 Pl. 17: Nos. 82-88 - Local red-slipped ware, fragments of bowl form 3. 209 89 91 92 93 94 95 Pl. 18: Nos. 89-95 - Local red-slipped ware, fragments of bowl form 4. 210 96 97 98 99 Pl. 19: Nos. 96-99 - Local red-slipped ware, fragments of bowl form 4. 100 Pl. 20: No. 100 – Local red-slipped ware, a dish fragment. 211 101 Pl. 21: No. 101 - Local red-slipped ware, base fragment of an open form. 102 Pl. 22: No. 102 - Local red-slipped ware, fragment of a jug/small pot? Closed Forms (pl. 22) It seems that red-slipped wares in closed forms are very few in numbers, perhaps due to the rising utilization of glass as table ware. The only fragment is a thin walled jug/small pot (?) fragment with a squat neck and bulgy body. Conclusions Most of the presumed local sigillata described above has a hard, homogenous fabric and contains some visible inclusions. Some sherds have technical defects. The typology of the local sigillata and red-slipped ware resembles that of the western and southern Anatolian workshops. In Bath A a large number of very small sigillata sherds was found; they were perhaps deposited in the debris of the Bath. Since we have no evidence of terracotta figurines or lamps from Kimistene’s Acropolis, it is interesting to see how frequent sigillata is. Both of the main survey areas, i.e. Hadrianopolis and Kimistene, are similar with respect to sigillata and red-slipped traditions, 212 but Kimistene has very few later ceramics after the 4th cent. A.D. when the pagan temple at Kimistene was abandoned. The amount of terra sigillata at Hadrianopolis and in its surrounding area is not large. It seems that sigillata in this landscape was associated with local demands, especially for burial gifts. In this case sigillata in southwestern Paphlagonia cannot be interpreted as a “commercial good”. Sigillata was part of the urban pottery tradition in Roman Asia Minor from the 1st cent. A.D. onwards, but the function of sigillata in southwestern Paphlagonia was different: in this part of Asia Minor sigillata was mostly used as burial goods during the 1st cent. B.C. to the 2nd cent. A.D., while later on - after 3rd cent. A.D. - red-slipped vessels began to be used as table ware. Further evidence for the common use of local sigillata as burial gifts has been provided by the surveys of G. Karauğuz in Devrek, Gökçebey, Çaycuma, and Çaydeğirmeni in the Province of Zonguldak (KARAUĞUZ 2006, 327). Also a closed vessel fragment (fig. 2a-c) collected by J. Mellaart in the area and today stored in the Sherd Collection of the British Institute of Archaeology at Ankara shows that local sigillata from Hadrianopolis can be found as far west as Mengen. It is possible that Greek and Roman food culture was not introduced into southwestern Paphlagonia before the beginning of the red-slipped ware tradition and that wood was an important material for the crockery before that. The sigillata finds indicate regional production in southwestern Paphlagonia, northwestern Galatia, and eastern Bithynia. The location of this workshop should be somewhere between Gangra and Hadrianopolis, most probably in the chora of Hadrianopolis, where there was a constant water supply and suitable clay. More work is also required in the study of modern production centres, such as Çömlekçiler Village of Eskipazar, to understand ancient kiln sites. Notes and Acknowledgements All the plates were drawn by G. Kan Şahin, and maps are by E. Laflı. We would like to thank to following colleagues for their help with various aspects of this paper (alphabetically): William Anderson (Melbourne), Eva Christof (Graz), Aslı Cumalıoğlu (Izmir), Sylvia Fünfschilling (Augst), Chris Lightfoot (New York), John Lund (Copenhagen), Arthur Muller (Lille/Athens), Britta Ramminger (Hamburg), Ole Stilborg (Storefors) and Sami Patacı (Istanbul). For this article Munsell Soil Color Charts 2009 has been used. An extensive catalogue of each piece will appear as: Laflı/Kan Şahin in print a. Second article of this series will be published in the next volume of the same workshop. Abbreviations and References ABADIE-REYNAL 1992: C. Abadie-Reynal, Les céramiques fines. In: C. Abadie-Reynal, J.-P. Sodini, La céramique paléochrétienne de Thasos (Aliki, Delkos, fouilles anciennes). École française d’Athènes, Études Thasiennes, XIII (Paris/Athens 1992) 13-35 and figs. 1-14. ATEŞ 2003: G. Ateş, Die rote Feinkeramik von Aizanoi als lokaler Kulturträger (Untersuchungen zum Verhältnis von lokaler roter Glanztonware und importierter Sigillata) (unpubl. diss.; University of Heidelberg 2003). 213 BONIFAY 2004: M. Bonifay, Études sur la céramique romaine tardive d’Afrique. British Archaeological Reports, International Series 1301 (Oxford 2004). DEVREKER ET AL. 2003: J. Devreker, H. Thoen, F. Vermeulen, Excavations in Pessinus: the So-called Acropolis. From Hellenistic and Roman Cemetery to Byzantine Castle. Ghent University, Archaeological Reports 1 (Ghent 2003). DOMŻALSKI 2011: K. Domżalski, Late Roman Pottery from Pompeiopolis In: <http://www. klass-archaeologie.uni-muenchen.de/projekte/pompeiopolis/projekt/materialstudien/ spaetroemkeramik.pdf> (15/12/2011). FIRAT 2010: N. Fırat, Les céramiques sigillées mises au jour dans l’atelier de Demirci. In: D. Kassab Tezgör (ed.), Les fouilles et le matérial de l’atelier amphorique de Demirci près de Sinope, Varia Anatolica 22 (İstanbul/Paris 2010) 187-195 and pls. 49-54. GASSNER 1997: V. Gassner, Das Südtor der Tetragonos-Agora, Keramik und Kleinfunde. Österreichische Akademie der Wissenschaften, Forschungen in Ephesos 13: 1/1 (Vienna 1997). GUNNEWEG ET AL. 1983: J. Gunneweg, I. Perlman, J. Yellin, The Provenience, Typology and Chronology of the Eastern Terra Sigillata. The Hebrew University of Jerusalem, QEDEM, Monographs of the Institute of Archaeology 17 (Jerusalem 1983). HAYES 1972: J. W. Hayes, Late Roman Pottery (London 1972). HAYES 1985: J. W. Hayes, Sigillate oriental. In: R. Bianchi Bandinelli, G. Becatti (eds.), Enciclopedia dell’arte antica classica e orientale. Atlante delle forme ceramiche, II, Ceramica fine Romana nell basino Mediteraneo (tardo-ellenismo e primo impero) (Rome 1985) 1-96. HAYES 2008: J. W. Hayes, The Athenian Agora 32: Roman Pottery. Fine-ware Imports (Princeton, N.J. 2008). HARRISON 1991: R. M. Harrison, Amorium Excavations 1990: The Third Preliminary Report. Anatolian Studies 41, 1991, 215-229. HARRISON 1992: R. M. Harrison, Amorium Excavations 1991: The Fourth Preliminary Report. Anatolian Studies 42, 1992, 207222. HENRICKSON 1992: R. C. Henrickson, Politics, Economics and Ceramic Continuity at Gordion in the Late Second and First Millennia B.C. In: W. D. Kingery (ed.), Social and Cultural Context of New Ceramic Technologies (Westerville, OH 1992) 89-176. KAYA 2003: N. Kaya, Ankara Ulus Kazısında Ele Geçen Baskılı Kaplar. In: C. Abadie-Reynal (ed.), Les céramiques en Anatolie aux époques hellénistique et romaine, Actes de la table ronde d’Istanbul, 23-24 mai 1996. Institut français d’études anatoliennes, Georges Dumézil, Varia anatolica 15 (Paris 2003) 111116 and pls. 72-73. KENRICK 1985: P. M. Kenrick, Excavations at Sidi Khrebish Benghazi (Berenike), Vol. III, Part 1: The Fine Pottery. Supplements to the Libya Antiqua V (Tripoli 1985). 214 KNIPOWITSCH 1929: T. P. Knipowitsch, Untersuchungen zur Keramik römischer Zeit aus den griechischen Städten an der Nordküste des Schwarzen Meeres I. Die Keramik römischer Zeit aus Olbia in der Sammlung der Ermitage. Materialien zur römisch-germanischen Keramik IV (Frankfurt 1929). KARAUĞUZ 2006: G. Karauğuz, 2005 Yılı Devrek-Gökçebey (Tefen) Yüzey Araştırması. In: T.C. Kültür ve Turizm Bakanlığı Kültür Varlıkları ve Müzeler Genel Müdürlüğü, 24. Araştırma Sonuçları Toplantısı (Ankara 2006) 327-340. LAFLI 2003: E. Laflı, Neue Erkenntnisse zur lokalen Keramik aus Tavium. In: K. Strobel, C. Gerber, Tavium (Büyüknefes, Provinz Yozgat)-Bericht über die Kampagnen 20002002. Istanbuler Mitteilungen 53, 2003, 153157. LAFLI 2006: E. Laflı, Hellenistic and Roman Ceramic Archaeology on the Southern Black Sea Coast: An Overview of the Turkish Archaeological Literature related to Ceramic Archaeology of Paphlagonia and Pontus. In: B. D. Erciyas, E. Kopraman (eds.), Karadeniz Araştırmaları Sempozyum Bildirileri/Black Sea Studies Symposium Proceedings, 2004 Ankara (Istanbul 2006) 189-202. LAFLI 2009a: E. Laflı, Paphlagonia Hadrianopolis’i Arkeolojik Kazıları ve Onarım Çalışmaları 2008 Yılı Çalışma Raporu, Arkeoloji ve Sanat 31/131, 2009, 39-62. LAFLI 2009b: E. Laflı, Glass from Hadrianopolis (Paphlagonia). In: E. Laflı (ed.), Late Antique/Early Byzantine Glass in the Eastern Mediterranean (Izmir 2009) 161-170. LAFLI/CHRISTOF 2011: E. Laflı, E. Christof, Der kaiserzeitliche Tempel von Asartepe/Kimistene in der Chora des paphlagonischen Hadrianopolis – Ergebnisse der Prospektion von 2005. Istanbuler Mitteilungen 61, 2011, 233-285. LAFLI/CHRISTOF 2012: E. Laflı, E. Christof, Hadrianopolis I: Inschriften aus Paphlagonia, British Archaeological Reports, International Series 2366 (Oxford 2011) (in print). LAFLI/KAN ŞAHÍN 2010: E. Laflı, G. Kan Şahin, Hadrianoupolis ve Çevresinden Geç Ortaçağ Sırlı Seramik Örnekleri. In: K. Pektaş et al. (eds.), Proceedings of the XIIIth Symposium of Medieval and Turkish Period Excavations and Art Historical Researches, 14-16 October 2009 (Istanbul 2010) 427-432. LAFLI/KAN ŞAHÍN in print a: E. Laflı, G. Kan Şahin, Terra Sigillata and Red-Slipped Ware from Hadrianopolis in Southwestern Paphlagonia. Anatolica Antiqua 20, 2012 (in print). LAFLI/KAN ŞAHÍN in print b: E. Laflı, G. Kan Şahin, A Preliminary Report on the Iron Age and Hellenistic Ceramics from Southwestern Paphlagonia. In: S. Drougou et al. (eds.), Η΄ Επιστημονική Συνάντηση για την Ελληνιστική Κεραμική (Ιωάννινα 2009) / H’ Scientific Meeting on Hellenistic Pottery, Ioannina 2009. Proceedings (Athens 2012; in print). LAFLI/ZÄH 2008: E. Laflı, A. Zäh, Archäologische Forschungen im byzantinischen Hadrianupolis in Paphlagonien. Byzantinische Zeitschrift 101, 2008, Heft 2, 681-714 and pls. XIII-XXVI. 215 LAFLI/ZÄH 2009: E. Laflı, A. Zäh, Beiträge zur frühbyzantinischen Profanarchitektur aus Hadrianupolis-Blütezeit unter Kaiser Iustinian I. Byzantinische Zeitschrift 102, 2009, Heft 2, 639-659 and pls. V-XIII. MATTHEWS ET AL. 2009: R. Matthews, M. Metcalfe, D. Cottica, Landscapes with Figures: Paphlagonia through the Hellenistic, Roman and Byzantine Periods, 330 AD-AD 1453. In: R. Matthews, C. Glatz (eds.), At Empires’ Edge: Project Paphlagonia. Survey in North-Central Turkey. British Institute of Archaeology at Ankara, Monograph 44 (London 2009) 182-186. MEYER-SCHLICHTMANN 1988: C. MeyerSchlichtmann, Die pergamenische Sigillata aus der Stadtgrabung von Pergamon. Deutsches Archäologisches Institut, Pergamenische Forschungen 6 (Mainz 1988). MITCHELL 1993: S. Mitchell, Anatolia: Land, Men, and Gods in Asia Minor. Vol. 1: The Celts in Anatolia and the Impact of Roman Rule; Vol. 2: The Rise of the Church (Oxford 1993). PAZ/VARGAS 2011: J. V. Paz, E. G. Vargas, Sigillatas africanas y orientales de mediados del VI d. C. procedentes de los rellenos de colmatación de una cisterna de Hispalis (Sevilla). Los contextos de la Plaza de la Pescadería. In: M. Á. Cau, P. Reynolds, M. Bonifay (eds.), LRFW 1. Late Roman Fine Wares. Solving Problems of Typology and Chronology. A Review of the Evidence, Debate and New Contexts. Roman and Late Antique Mediterranean Pottery 1 (Oxford 2011) 87-97. POBLOME 1999: J. Poblome, Sagalassos Red Slip Ware, Typology and Chronology. Studies in Eastern Mediterranean Archaeology 2 (Brepols 1999). POBLOME ET AL. 2001: J. Poblome, D. Patrick, D. Cottica, F. Nalan, A New Early Byzantine Production Centre in Western Asia Minor. A Petrographical and Geochemical Study of Red Slip Ware from Hierapolis, Perge and Sagalassos. Rei Cretariae Romanae Favtorum, Acta 37, 2001, 119-126. POBLOME/FIRAT 2011: J. Poblome, N. Fırat, Late Roman D. A Matter of Open(ing) or Closed Horizons? In: M. Á. Cau, P. Reynolds, M. Bonifay (eds.), LRFW 1. Late Roman Fine Wares. Solving Problems of Typology and Chronology. A Review of the Evidence, Debate and New Contexts. Roman and Late Antique Mediterranean Pottery 1 (Oxford 2011) 49-55. SLANE 1997: K. W. Slane, The Fine Wares. In: A. Berlin, K. W. Slane, Tel Anafa II, i: The Hellenistic and Roman Pottery. Journal of Roman Archaeology, Supplementary Series 10, II, i (Ann Arbor, MI 1997) 247-406. WEBER-HIDEN 2003: I. Weber-Hiden, Keramik aus hellenistischer bis frühbyzantinischer Zeit aus Tavium/Büyük Nefes: Bemerkungen und Übersicht über das Begehungsmaterial der Kampagnen 1998-2000 aus drei ausgewählten Bereichen des Stadtgebietes. Anatolia Antiqua 11, 2003, 253-322. WILLIAMS 1989: C. Willams, Anemurium. The Roman and Early Byzantine Pottery. Pontifical Institute of Mediaeval Studies, Subsidia Mediaevalia 16 (Toronto, ON 1989). 216 ZHURAVLEV 2009: D. Zhuravlev, Pontic Sigillata Plates with a Vertical Rim from the Belbek IV Necropolis in the South-Western Crimea. Ancient Civilizations from Scythia to Siberia 15, 2009, 25-94. ZHURAVLEV 2011: D. Zhuravlev, Early Roman Fine Ware from Pompeiopolis. In: <http:// www.klass-archaeologie.uni-muenchen.de/ projekte/pompeiopolis/projekt/materialstudien/fruehroemkeramik.pdf> (15/12/2011). Addresses: Doç. Dr. Ergün LAFLI Dokuz Eylül Üniversitesi, Edebiyat Fakültesi, Arkeoloji Bölümü Tınaztepe/Kaynaklar Yerleşkesi, Buca, TR-35160 Izmir TURKEY elafli@yahoo.ca Gülseren KAN ŞAHİN Dokuz Eylül Üniversitesi, Edebiyat Fakültesi, Arkeoloji Bölümü Tınaztepe/Kaynaklar Yerleşkesi, Buca TR-35160 Izmir TURKEY gulserenkan@hotmail.com