Proceedings of the 9th International Congress on the Archaeology of the Ancient Near East, Volume 3: Reports 3447106158, 9783447106153

Citation preview

9 IC A A N E Proceedings of the 9th International Congress on the Archaeology of the Ancient Near East Volume 3 Reports

Harrassowitz Verlag

Proceedings of the 9th International Congress on the Archaeology of the Ancient Near East Volume 3

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings of the 9th International Congress on the Archaeology of the Ancient Near East 9–13 June 2014, Basel Edited by Rolf A. Stucky, Oskar Kaelin and Hans-Peter Mathys

2016

Harrassowitz Verlag · Wiesbaden

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings of the 9th International Congress on the Archaeology of the Ancient Near East Volume 3 Reports Edited by Oskar Kaelin and Hans-Peter Mathys

2016

Harrassowitz Verlag · Wiesbaden

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Cover illustration: © Gino Caspari, Columbia University.

Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.dnb.de abrufbar. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the internet at http://dnb.dnb.de.

For further information about our publishing program consult our website http://www.harrassowitz-verlag.de © Otto Harrassowitz GmbH & Co. KG, Wiesbaden 2016 This work, including all of its parts, is protected by copyright. Any use beyond the limits of copyright law without the permission of the publisher is forbidden and subject to penalty. This applies particularly to reproductions, translations, microfilms and storage and processing in electronic systems. Printed on permanent/durable paper. Printing and binding: Memminger MedienCentrum AG Printed in Germany ISBN 978-3-447-10615-3 e-ISBN PDF 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9th ICAANE, Basel 2014, Vol. 3

Contents Iraq

11

Giacomo Benati – Camille Lecompte New Light on the Early Archives from Ur: The «Ancient Room» Tablet Hoard

13

Eloisa Casadei Southern Mesopotamian Contexts and a Reevaluation of the Ur III Pottery. Reconsidering a Chronological Periodization

31

Franco D’Agostino – Ali Khadem Ghanim – Licia Romano Abu Tbeirah. Preliminary Report of the 2012–2013 Campaigns

45

Paul Zimansky – Elizabeth Stone Tell Sakhariya and Gaeš

57

Iraq (Kurdistan)

67

Kyra Kaercher A Preliminary Assessment of the Ceramic Sequence of Northeastern Iraqi Kurdistan

69

John MacGinnis – Timothy Matney – Kemalettin Köroğlu Excavations in the Lower Town of Ziyaret Tepe 2012 and 2013

83

Karel Nováček – Miroslav Melčák The Medieval Urban Landscape in Northeastern Mesopotamia (MULINEM): First Two Years of the Project

95

Maria Grazia Masetti-Rouault – Olivier Rouault French Excavations in Qasr Shemamok, Iraqi Kurdistan (2013 and 2014 campaigns): the Assyrian Town and Beyond

107

Tevfik Emre Şerifoğlu – Jesse Casana – Claudia Glatz – Shwkr Muhammed Haydar Initial Results of the Sirwan (Upper Diyala) Regional Project

119

Tim B. B. Skuldbøl – Carlo Colantoni First Results of the Rania Plain Survey. Salvage Operations in the Dokan Dam Inundation Zone

131

Aline Tenu – Christine Kepinski Kunara, a Bronze Age City on the Upper Tanjaro (Iraq)

147

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

6

Contents

Syria

161

Cristina Baccarin Consumption in a Temple? An Interpretation of the Ceramic Repertoire of the Early Bronze Age Temple at Tell Ahmar (North Syria)

163

Anna Gómez Bach Characterizing Red Ware : More than a Single Production at Tell Halula (Syria) at the End of the Halaf Period

177

Guy Bunnens A 3rd Millennium Temple at Tell Ahmar (Syria)

187

Marta D’Andrea New Data from Old Excavations: Preliminary Study of the EB IVB Pottery from Area H at Tell Mardikh/Ebla, Syria

199

Stephanie Döpper The LBA pottery of Area BU in the Royal Palace of Qatna, Syria

217

Frank Hole Historical Processes on the Middle Khabur River, Syria, during the Late Ubaid: Intrusion, Attenuation and Divergence

231

Béatrice Muller Panneaux d’incrustation en coquille de Mari, Ville II: implications des matériaux et des techniques

243

Paola Poli Seals and Sealings from Tell Masaikh–Kar-Assurnasirpal. Some New Results

257

Agnese Vacca New Data on the EB III–IVA1 of North-Western Syria in the light of Old and Recent Excavations at Tell Mardikh/Ebla and Tell Tuqan

269

Turkey

283

Gulan Ayaz The Early Iron Age Jewellery from the Karagündüz Necropolis in Eastern Anatolia and Its Relationship to Southern Caucasia

285

Alice Boccia Paterakis – Sachihiro Omura Gold Cloisonné from the Assyrian Colony Period in Central Anatolia

293

Müge Bulu An Intact Palace Kitchen Context from Middle Bronze Age Alalakh: Organization and Function

301

Nilgün Coşkun Middle and Neo-Assyrian Periods of the Harran Plain in Light of a Survey

315

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Contents

7

Anacleto D’Agostino – Valentina Orsi Researches at Uşaklı Höyük (Central Turkey): Survey, Surface Scraping and First Digging Operations

333

Sevinç Günel A New Centre of Intercultural Relations in Western Anatolia during the Late Bronze Age: Çine-Tepecik

347

Sara Pizzimenti – Federico Zaina The Iron Age at Karkemish between Tradition and Innovation. The Case Study of the Pottery Assemblage from Area C

361

Iran

377

Rasoul Seyedin Boroujeni – Saman Hamzavi Zarghani – Mohsen Zeidi Discovery of Long Term Occupation in the Saimarreh River Valley, Western Iran (Report on the Archaeological Survey)

379

Enzo Cocca – Andrea Genito – Bruno Genito – Giulio Maresca A WebGIS about the Italian Archaeological Activities in Sistan, Iran (60s–70s of the 20th century): Archaeo.Pro.Di.Mu.S.

393

Nicolas Assur Corfù Esotericism at Persepolis – Really?

405

Ata Hasanpur – Zahra Hashemi A Comparative Study of the New Sassanid’s Stuccos from Qela Gowri, Ramavand, Lorestan, Iran

417

Najmeh Hassas An Introduction to the Traditional Arch and Vault from the Elamite (1250 BC. M.) to the Qajar period (1925) in Iran

429

Vito Messina – Jafar Mehr Kian The Religious Complex at Shami. Preliminary Report on the Research of the Iranian-Italian Joint Expedition in Khuzestan at Kal-e Chendar

439

Israel – Lebanon – Levant

449

Eva Katarina Glazer Sedentary and Nomadic Population during Bronze Age Southern Levant: An Example of Cultural Contacts

451

Aaron Greener Analyzing the Late Bronze Age Imported Pottery Distribution in the Southern Levant: Overcoming Methodological Challenges

463

Haskel J. Greenfield – Itzhaq Shai – Aren M. Maeir Understanding Early Bronze Age Urban Patterns from the Perspective of Non-Elite Neighbourhood: The Excavations at Tell es-Safi/Gath, Israel 475

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

8

Contents

Ann E. Killebrew – Jamie Quartermaine Total Archaeology@Tel Akko (The 2013 and 2014 Seasons): Excavation, Survey, Community Outreach and New Approaches to Landscape Archaeology in 3D

491

Florine Marchand L’industrie lithique de Tell ‘Arqa (Plaine du Akkar, Liban Nord)

503

Jordan

515

Don Boyer Aqueducts and Birkets: New Evidence of the Water Management System Servicing Gerasa (Jarash), Jordan

517

Marta D’Andrea Pottery Production at Khirbat Iskandar, Jordan. Preliminary Results of the Technological Study of EB IV Pottery from the Site

533

Sumio Fujii Slab-lined Feline Representations: New Finding at ‘Awja 1, a Late Neolithic Open-air Sanctuary in Southernmost Jordan

549

Lucas Petit Tell Damiyah. A Small Settlement Mound with a Remarkable International Role

561

Andrea Polcaro – Juan Muñiz – Valentín Alvarez The New Spanish-Italian Expedition to the EB I site of Jebel al-Mutawwaq, Middle Wadi az-Zarqa, Jordan: Preliminary Results of the 2012–2013 Campaigns

571

Suzanne Richard Recent Excavations at Khirbat Iskandar, Jordan. The EB III/IV Fortifications.

585

Uzbekistan – Turkmenistan – Afghanistan

599

Joaquín María Córdoba – Muhammed Mamedov L’âge du fer au Dehistan. Nouvelles recherches archéologiques turkmènes et espagnoles sur les sites de Geoktchik Depe et Izat Kuli (Province de Balkan, Turkménistan)

601

Fabiana Raiano Pottery from the joint Uzbek-Italian Archaeological Mission at Kojtepa (Samarkand Area – Uzbekistan)

615

Victor Sarianidi – Nadezhda Dubova Types of Graves at Gonur Depe Bronze Age Site in Turkmenistan

631

Liliya Sataeva – Robert Sataev Wood Using at the Bronze Age Site Gonur-Depe (Ancient Margiana, South Turkmenistan)

643

Judith Thomalsky Afghanistan: Ancient Mining and Metallurgy: Initial Project Stage

647

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Contents

9

Oman

663

Michele Degli Esposti Excavations at the Early Bronze Age Site «ST1» near Bisya (Sultanate of Oman): Notes on the Architecture and Material Culture

665

Stephanie Döpper Expressions of Collective Memory – The Reuse of EBA Tombs in the Necropolis of Bāt, Sultanate of Oman

679

Conrad Schmidt Mobile Pastoralists as Global Players: Excavations at Al-Zebah, Sultanate of Oman

689

Azerbaijan – Georgia – Caucasus-Region

697

Jeyhun Eminli – Emil Iskenderov Archaeological Investigations at Piboz Tapa Necropolis. Lerik, Southern Azerbaijan (Preliminary Report 2012–2013)

699

Seiji Kadowaki – Farhad Guliyev – Yoshihiro Nishiaki Chipped Stone Technology of the Earliest Agricultural Village in the Southern Caucasus: Hacı Elamxanlı Tepe (the Beginning of the 6th Millennium BC)

709

René Kunze Interdisciplinary Studies on the Small Finds from the Late Bronze/Early Iron Age Settlements of Udabno I–III (Eastern Georgia)

723

General Topics

735

Silvia Alaura – Davide Nadali Researching the Archives: A Shared Past for the Future. The GRISSO Project

737

Gino Caspari Ein Inventar der Hügelgräber im Altaigebirge. Kulturgüterschutz durch Fernerkundung

741

Fabrice De Backer La construction d’un char de guerre néo-assyrien

751

Rita Gautschy Astronomical Data and their Potential for Chronological Purposes

763

Krzysztof Hipp Šamšī-Adad V’s Campaigns into the Zagros Revisited

769

Chamsia Sadozaï – David Gandreau Conserver après la fouille: la question des structures en terre crue à partir d’exemples en Asie centrale

781

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Iraq

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 13–30

Giacomo Benati – Camille Lecompte

New Light on the Early Archives from Ur: The «Ancient Room» Tablet Hoard This paper reconsiders a group of cuneiform tablets allegedly excavated in Seal-Impression Strata (SIS) 5–4 at Ur (Iraq). Analysis of unpublished records and texts kept in the British Museum of London revealed fresh information that is instrumental to re-discuss provenance and use of the tablets, analyzed here both as excavated artifacts and as economic accounts.

1. Introduction1 Mesopotamian scholarship has been often characterized by a split between the study of written sources and other material culture remains (Zettler 1996: 81). With this paper we intend to apply an integrated strategy to the reconstruction of bureaucracy in Early Dynastic Mesopotamia. The study case is a group of tablets excavated by C. L. Woolley in the area of the Royal Cemetery at Ur. The context of provenance of the tablets is re-discussed on the basis of unpublished archival information, while the texts are reviewed in light of the new contextual frame and of recent work on heretofore unpublished tablets.2 The first part of the paper deals with the find-place of the tablets, while the second part consists of a brief textual analysis of the accounts. 1 2

G. Benati wrote §§ 2–3, C. Lecompte wrote §§ 4–5; §§ 1, 6 were written together. The archival data used in this paper are contained in the so-called «PG Notes», notes taken by the excavators while digging in the RC Area, and the «index cards», notes written for the objects filed under excavation numbers, now part of the Ur Expedition Original Papers Collection kept in the Middle East Department of the British Museum (London). These have been studied by G. Benati, while the original tablets – also kept in the BM – have been studied by C. Lecompte (cf. Lecompte 2013). The authors intend to thank the Middle East Department of the British Museum, in particular J. Taylor and S. Collins, for permission to study the Ur materials in the museum. We also thank N. Marchetti and G. Marchesi for providing valuable feedback and criticism on the matters discussed in this paper.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

14

Benati – Lecompte

Fig. 1. Map showing the location of the excavation areas that yielded 3rd millennium B.C. material culture at Ur (based on Woolley 1932: pl. 63; Woolley 1956: pl. 1; computer graphics by G. Benati)

2. Re-excavating the «Ancient Room»: A Stratigraphic and Contextual Reconstruction At Ur, most of the artifacts dating from the onset of the Early Dynastic period were retrieved in layers of debris excavated in the so-called Royal Cemetery area (hereafter RC area; figs. 1–2). Here, a sequence of refuse heaps – i.e. Seal-Impression Strata – and building remains, pre-dating the graves of the Royal Cemetery, has been documented.3 According to the excavators, a hoard of tablets was retrieved to the South-East of PG1237 on the floor of a room, i.e. Ancient Room (Woolley 1956: 70–71; cf. also Woolley 1929: 327; fig. 2). The original records kept in the British Museum inform us that these remains were encountered while excavating a «long trench behind PG1237», as demonstrated also by a field photo (Woolley 1934: pl. 70a). These features can be observed on a schematic section

3

On the chronology of the lower SIS layers (8–4) see Zettler 1989, Matthews 1993.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Light on the Early Archives from Ur: The «Ancient Room» Tablet Hoard

15

Fig. 2. Map showing some of the archaeological features excavated within the Royal Cemetery Area discussed in the article (based on Karstens 1987: fig. 2; Nissen 1966: pl. 41; Zimmerman 1998: fig. 53; computer graphics by G. Benati).

drawing published in the final report (cf. Woolley 1956: pl. 72). By taking measures on it, one can conclude that the tablets were retrieved at between 7.50 and 7.80 m above sea level. An unpublished sketch drawing from the BM archives provides us a close-up of the Ancient Room’s stratigraphy (fig. 3). The sketch shows two building phases of a structure consisting of a lower floor level (L.903) bounded by wall stubs, and an upper floor (L.900)

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

16

Benati – Lecompte

Fig. 3. Schematic section showing the stratigraphy of the Ancient Room’s loci (based on the section kept in the BM archive; computer graphics by G. Benati).

connected to a wall stretch (W.908) laid upon the former wall (W.905).4 A layer of ash and pottery sherds (F.901) separated the two phases. The lower building phase is the so-called Ancient Room and the tablets were retrieved on its floor (L.903), lying along the foot of a wall (and partially leaning on it), mixed with ash and crumbled mud-bricks (F.902). It seems clear that the Ancient Room building was destroyed by a conflagration and rebuilt shortly after. In doing so, the debris were levelled and sealed by a fill of trash (F.901) laid out as a preparation for the new structures. According to the PG Notes (cf. n. 2), while digging below the floor of the Ancient Room, ash tips were distinguished all over. The Ancient Room building was set on a layer of charred remains and pottery sherds. This stratigraphic information permits to link the features of the Ancient Room with other contexts excavated within the RC Area (see § 3 below).

4

In order to better define the depositional history of this operation, loci numbers (L. = Locus; F. = Fill; W. = Wall) have been allocated to the stratigraphic features on the basis of the information provided by the excavators.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Light on the Early Archives from Ur: The «Ancient Room» Tablet Hoard

17

Fig. 4. Map showing the reconstruction of SIS 5/4 extension and its relationship with F.1011 as discussed in the text and portrayed in the section of fig. 5 (based on Karstens 1987: fig. 2; Nissen 1966: pl. 41; Zimmerman 1998: fig. 53; computer graphics by G. Benati).

2.1. The Ancient Room Assemblage Turning to the finds from the room, the analysis of the index cards provided two data strands: a set of field numbers allocated to the tablet fragments; contextual and altimetric information. Thanks to this information we were able to draft a complete list of the finds excavated in the room, not otherwise given in the reports (table 1). The object cards also report that the tablets were «within the house remains located to the SE of PG1237, between 7.50 and 7.80 m asl, mixed with rubbish» (excerpt from PG note of item U.12578), corroborating the reconstruction proposed above.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

18

Benati – Lecompte

Fig. 5. Section through Pit W showing the stratigraphy of SIS 7/6, SIS 5/4 and F.1011, and the projected position of the Ancient Room (based on Woolley 1956: pl. 77, Sürenhagen 1999: pl. 61; computer graphics by G. Benati).

Then we checked the field numbers against the Archaic Texts edition (UET 2). This operation revealed that most of the tablets were published by Burrows and assigned to SIS 5–4 as a bulk. Consequently, with the field numbers at hand, we have been able to single out the tablets from the corpus and re-establish their context of provenance. Furthermore, we discovered that, among the texts omitted by Burrows and recently published by C. Lecompte (2013), ten are to be assigned to the Ancient Room assemblage, thus improving the textual information at disposal (fig. 6).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Light on the Early Archives from Ur: The «Ancient Room» Tablet Hoard

19

Fig. 6. Photo of newly published tablet U.12927 (Lecompte 2013: no. 46; http://cdli.ucla.edu/ P449033; photo by C. Lecompte and reproduced by courtesy of the Trustees of the British Museum).

The original records also produced evidence of the presence of two clay sealings on the floor of the Ancient Room. One of these is apparently still unpublished. While the other one was studied by R. Matthews (1993: no. 34), and ascribed to the «City Seals» glyptic horizon (cf. also Matthews 2013, and Pittman 2012). The presence of these artifacts in the Ancient Room provides precious chronological as well as administrative insights that are however beyond the scope of this paper. To sum up, it seems clear that we are dealing with a repository of administrative devices kept in a specialized room, possibly connected to a managing institution within the settlement of Ur.

3. Site Formation Processes in the RC Area In order to better frame the Ancient Room finds, I am going to sketch a relative sequence of depositional events corresponding to the stratigraphy of the layers of debris – SIS and associated fills – excavated within the RC Area (cf. also Marchetti in Marchesi/Marchetti 2011: 52–54). The lowest SIS layer, SIS 8, was distinguished at the bottom of Pit Z (cf. fig. 2). This stratum was covered by two other layers of debris – SIS 7/6 – distinguished in Pits Z and W. A thick stratification of rubbish – SIS 5/4 – then accumulated over SIS 7/6 (fig. 5). SIS 5/4 is described as a deposit formed by tips of lime and decomposed mud-bricks extending over the surface spanning from Pit Z to Pit X. These layers were merged in the course of excavations and treated as a unique stratum (Zettler 1989: 369–372). Evidence from the BM archives however allows us to break the SIS 5/4 dump into discrete dumping episodes (figs. 4–5):

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

20

Benati – Lecompte

1. The tips of SIS 5/4 proper – sharply sloping layers of construction debris, clay sealings, cuneiform tablets and other tools – accumulated rapidly and formed an open dump, distinguished in Pits Z, Y and W (cf. Zimmerman 1998: figs. 56–57). 2. A layer of soot, pottery sherds and administrative tools then accumulated at the foot of SIS 5/4 in Pit W and ran fairly flat for about 25 m towards North-East (F.1011). The layer of soot F.1011, located above the foot of SIS 5/4, provides the virtual linkage between the stratigraphy of the SIS dump and the area of the Ancient Room. This layer was detected between 7.30 m and 7.70 m asl and – as attested by the stratigraphy of Pits D and W (cf. Woolley 1956: pls. 72, 77, 83; fig. 5) – it was covered by ruined mud-brick structures in Pit D. This situation mirrors the Ancient Room stratigraphy. From this evidence, one may assume that layers of incinerated remains were spread in the strip comprised between Pit W and PG1237 and used to level an area for construction works. Well-preserved remains of mud-brick compounds with in situ artifacts were in fact excavated at roughly the same level in other find-spots within the RC area and interpreted by Woolley (1956: 71) as remains of an urban quarter, fostering our assumption (i.e. a house below PG1050, Woolley 1934: 96; cf. figs. 4–5). In this light, we can conclude that multi-room mud-brick buildings were built in the area beside the dump formed by SIS 5/4 rubbish tips and partially covering its foot (therefore post-dating it). Consequently, the debris of SIS 5/4, F.1011 and the Ancient Room (and the urban quarter), may be considered sequential snapshots within a series of operations connected to refuse disposal and a cityscaping episode in early Ur.

4. Textual Analysis of the Tablets from the «Ancient Room» The archaic texts from Ur are the most significant piece of evidence for the first part of the Early Dynastic period in Mesopotamia (Visicato 2000: 13–18; Lecompte 2013: 23). Most of these texts are plain administrative documents of small size related with food rations, fields and personnel (Sallaberger 2010). Having located more accurately the contexts of provenance of the finds studied in this article – the Ancient Room – we can now interconnect archaeological and textual evidence in order to draw some conclusions on the typology of the written records. This can be helpful in the task of better understanding the organization of the administration at Ur. With a total of 198 texts and tablets, the documents uncovered in the Ancient Room represent approximately half of the whole textual corpus of archaic texts from Ur and can be classified as follows:

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Light on the Early Archives from Ur: The «Ancient Room» Tablet Hoard

21

• • • • •

73 tablets (36,9%) deal with foodstuffs (cereals, bread, beer, etc.). 51 tablets (25,7%) deal with cultivated fields. 6 are lists of personal names (3%). 5 deal with livestock (2,5%). The rest deals with containers, textiles, metals, animal skins, timber, reeds, and other commodities (9 texts), and one is a scholastic text (note also the presence of a drawn marking). • 52 texts are still unclear and require further study. Fields and staple products are the main topics of these accounts, in line with early Mesopotamian bookkeeping practices. Nevertheless, significant aspects of this corpus can be underscored taking into account the body of knowledge from the archaic texts. Firstly, most of the texts dealing with fields within the Ur corpus – around 70 – come from the Ancient Room, but seventeen texts, bearing the excavation number U.20081 and coming from another context, also form another important group of similar texts. Given that three of the tablets5 mention the «domain of Nanna» (aša5 Nannax)6, it is possible that the Ancient Room was part of an office specialized in the management of the Nanna temple field estate. Secondly, it appears that this «office» allotted food rations, probably distributed as wages. Accounts related with metal objects, animals and wooden products – albeit only few – may suggest a connection between this office and the distribution of craft goods. Lastly, it is worth mentioning that one school text was uncovered in the Ancient Room, namely the earliest Mesopotamian list of gods (UET 2: no. 105). From this overview, we may point out that the Ancient Room was part of an office connected with the administration of the Nanna sacred household, likely called eš3/AB. The architectural remains may therefore have belonged to the temple or to a building where branches of the administration were located. The association between the remains of the Ancient Room and the AB/eš3 is suggested by two tablets (UET 2: nos. 29 and 95) referring to the še gu7 eš3, «grain consumption/ration of the shrine». Unfortunately, evidence about the administration system of this structure is elusive. Professional titles are encountered in the tablets from the Ancient Room and some of them may be identified as the administrators of the Nanna temple. The subscript of UET 2: no. 108 refers to the official kiĝgal (GAL.UNKEN), traditionally interpreted as «chief of the assembly», but in this case he is the responsible of the transaction involving fields (cf. Charvát 1979: 19; Marchesi in Marchesi/Marchetti 2011: 103 n. 53). This seems to be an important indication on a hierarchy scheme in the management of fields. On 5 6

UET 2: nos. 73, 127, 227. To this add UET 2: no. 358 (U.20081 – 26) which does not come from the Ancient Room but from a trench (Pit X) where debris postdating this locus were possibly discarded. Despite the doubts of Pettinato (1999: 106, n. 53), the reading of the field name as aša5 Nannax is justified by the use of the sign ŠEŠ and not URI3. Cf. also Burrows 1935: 13; Charvát 1979: 17; Steinkeller 1988.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

22

Benati – Lecompte

the other hand, the high amount of foodstuff received by the ensix of Ur in UET 2: no. 88 may also be an evidence of the involvement of this official in a redistributive circuit at Ur, since this quantity is likely to have been distributed under the responsibility of a high-ranking administrator. The ensix however is not mentioned in the final clause of the document and therefore cannot be considered the main official responsible for the whole allotment. Individuals bearing the title saĝĝa/umbisaĝ – possibly high-ranking temple administrators or scribes – are also attested in these texts (Charvát 1979: 18, Sallaberger 2010: 33–34). In particular, one Amar occurs most frequently and all instances related to him are to be found in the Ancient Room corpus (UET 2: nos. 99, 100, 114, 142, 147, 159, 177, 194, 201, 205a). As usual in the texts from Ur, these attestations are ambiguous and a doubt arises on the identity of Amar. The name AMAR.ŠID,7 appears also at Fara and Kiš (cf. Pomponio 1987: 35; Steinkeller 2013: 136, 141). However, since, on the other hand, another saĝĝa is called Ak, the sign set AMAR.ŠID may also be read as Amar saĝĝa/umbisaĝ.8 This individual is attested in records related to the distribution of vessels and containers of barley and seeds for fields, some of them reaching a rather high amount, even though it cannot be converted into modern units due to the uncertainties of the interpretation of the metrological system.9

5. Administration and Institutions in Archaic Ur Turning to the general bureaucratic frame of early Ur, it must be stressed that the organization of the administration is poorly defined since the tablets hardly refer to the offices responsible for the mobilization of resources. Only a few texts give names of main administrators or institutions in the last clauses. Some of the tablets in fact display the use of isolated clauses and sign clusters written apart from the rest of the text, which refer in some cases to the administrators responsible for the transaction recorded. Apart from the kiĝgal, we encounter in similar positions the official engiz – a kind of cook – receiving a number of cakes and maybe responsible for their delivery (UET 2: no. 281), and personal names such as Lugal-PA.SU13(BU)+SIKIL (UET 2: no. 170). While the archaic texts are generally considered as being issued from the administration of the Nanna sanctuary, some scholars hypothesized the existence of other institution levels, such as the king (lugal), the palace (e2-gal), and some sort of city ruler (ensix). Now we take a closer look at these supposed institutional agencies. 7 8

9

Burrows 1935: 30, Personal Name 189, amar-sangu. The sign set AK.ŠID was, however, considered by Burrows (1935: pl. 30: sign 371), and Charvát (1979: 17) to be a term referring to bookkeeping: šid ak. Considering the texts in which it occurs, this term can be confidently identified with a personal name. Only the text UET 2: no. 147 is seemingly a list of personnel.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Light on the Early Archives from Ur: The «Ancient Room» Tablet Hoard

23

Evidence pointing to the existence of an archive from the Nanna temple mainly relies, as mentioned above, on the expression GAN2 Nannax = aša5 Nannax, «field of Nanna». Such clauses prove that a field estate belonged to this shrine and that the other texts uncovered in the aforementioned area and dealing with staples and other products, may also be identified with the administration of the Nanna temple. The term AB/eš3 in the archaic texts – which Burrows (1935: 13) suggested to identify with a «palace or temple» – has already been interpreted by other scholars as the temple of Nanna (Wright 1969: 41; Charvát 1979: 16–17; Steinkeller 1988: 20; Visicato 2000: 16). Sallaberger (2010: 32–33) considered it to be a «great complex, precinct», similarly to the e2-ninnu complex in Girsu which encompassed several sanctuaries. Archaeological evidence confirms the hypothesis that the term AB/ eš3 refers to the precinct of Nanna and that the Ancient Room was likely part of its domain. Charvát (1979) on the basis of the official title kiĝgal, interpreted as «chief of the assembly», and of the lack of a clear kingship, suggested the presence of a consultative body, such as an assembly, in archaic Ur. Textual evidence however suggests only that the kiĝgal is responsible for a large field estate, reflecting the high-ranking position of this manager, or is merely a member of the personnel. The «palace», e2-gal, is mentioned in a few texts. It is worth noting that in a tablet, an administrator bearing the title of nu-banda3 e2-gal «overseer» of the «palace» is attested (UET 2: no. 112; cf. also Sallaberger 2010: 34).10 Other institutional levels of this administration are equally ill-documented. The term ensix (PA.SI) which stands for the later PA.TE.SI = ensi2 , «ruler, governor», is attested in several contexts (cf. Marchesi in Marchesi/Marchetti 2011: 109). The ensix of Ur is surely attested in five documents (UET 2: nos. 86, 88, 99, 177, 22 Lecompte 2013: 63 and Lecompte 2014), whereas two other tablets refer to a PA Uri5,11 which may be a shortened form of PA.SI Uri5. Unfortunately, in none of these texts, the ensix of Ur is in charge of the transaction, but he is merely enlisted among the individuals receiving allotments of containers. However, in three other cases, the ensix receives a very high amount of products (UET 2: nos. 88, 95 and 222). This evidence may suggest his high-ranking position. At any rate, it cannot be determined whether the official designated as PA.SI Uri5 is the ruler of Ur, or another kind of high-ranking official.12

10 We must keep in mind that the term e2-gal is also used in personal names, e.g. Amar-e2-gal. 11 UET 2: nos. 9 and 99, in which, however, the broken term may also be reconstructed as PA.˹SI˺. 12 The title ensi2-gal, «chief steward», is encountered in one tablet (UET 2: no. 35). This official, of

lower rank with respect to the «énsi», seems connected to the upkeeping of water infrastructures (Marchesi in Marchesi/Marchetti 2011: 109) or with the management of gods’ temples, such as Enlil and Nin-girsu.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

24

Benati – Lecompte

On the other hand, the term «king» – Sumerian lugal – exclusively occurs as part of personal names, the only possible exceptions being, on the one hand, the mention of the lugal Lagaš, though uncertain (cf. Marchesi in Marchesi/Marchetti 2011: 100 n. 27) and, on the other hand, an ambiguous expression LUGAL ŠE ŠID (UET 2: no. 162). Since the lugal is not part of the administrative hierarchy gleaned from the archaic texts,13 one may conclude that power was detained by other institutions or that Ur was subordinate to another city (Wright 1969: 41; according to N. Marchetti and G. Marchesi [p.c.], this city might be identified with Uruk). Lastly, some tablets refer to the EN.GAN2 , which may be interpreted as the «ruler domain», similarly to the niĝ2-en-na in late Early Dynastic Lagaš texts, and may point to the existence of field estates controlled either by the Nanna temple, by other institutions, or by a ruler (Burrows 1935: 12; Steinkeller 1988: 20–21; Pettinato 1999: 106).

6. Concluding Remarks The reconstruction outlined above allowed us to re-contextualize an important body of cuneiform artefacts excavated by C. L. Woolley in the RC area. We have proved that at least half of the tablets forming the corpus of the archaic texts were retrieved in primary deposition within architectural context. The finer-grained depositional sequence sketched in this article can be used to clarify the development of bureaucratic as well as economic patterns in Early Dynastic Ur. Overall, the fresh analysis of the texts from the Ancient Room suggests that in this period, managing institutions were equipped with full-fledged bureaucratic offices and repositories of administrative records. The archive of the Ancient Room was seemingly issued by the administrative bodies of the sacred household of Nanna that at this point was a landholding institution capable of controlling field estates, attached personnel, wage labor, distribution of foodstuff, and perhaps craft productions. Further research is however required to bring this situation into clear perspective. From the methodological standpoint, the present paper aimed at establishing a dialogue between disciplines that too often have been kept separate in Mesopotamian scholarly research. If the goal is to adopt a holistic approach in the study of past societies then multiple lines of evidence should be taken into consideration (cf. Zettler 2003). We are confident that

13 Andersson (2012: 245–247) drew some conclusions on the social position of the lugal according to

the onomasticon and to the historical data. Note however that lugal in the onomasticon is in most cases referred to gods and not to rulers. The role of the lugal in the ED I–II society is therefore a speculative matter.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Light on the Early Archives from Ur: The «Ancient Room» Tablet Hoard

25

a problem-oriented approach focusing on Ur social and administrative complexity will contribute to shed more light on the beginning of the Early Dynastic period in Mesopotamia.

Bibliography Andersson, J., 2012. Kingship in the Early Mesopotamian Onomasticon 2800–2200 BCE. Studia Semitica Upsaliensia 28, Uppsala. Burrows, E., 1935. Archaic Texts. Ur Excavations Texts 2, London/Philadelphia. Charvát, P., 1979. Early Ur. Archív Orientální 47, 15–20. Karstens, K., 1987. Typologische Untersuchungen an Gefassen aus altakkadischen Gräbern des Königsfriedhofes in Ur. Ein Beitrag zur modernen archäologischen Methodik. Münchener Vorderasiatische Studien 3, München. Lecompte, C., 2013. Archaic Tablets and Fragments from Ur (ATFU) from L. Woolley’s Excavations at the Royal Cemetery. Nisaba 25), Messina. –

2014. Suggestions and corrections to ATFU – 2. Addenda and errata. Cuneiform Digital Library Notes 2014: 21.

Marchesi, G./Marchetti, N., 2011. The Royal Statuary of Early Dynastic Mesopotamia. Mesopotamian Civilizations 14, Winona Lake. Matthews, R. J., 1993. Cities, Seals and Writing: Archaic Seal Impressions from Jemdet Nasr and Ur Materialien zu den frühen Schriftzeugnissen des Vorderen Orients 2, Berlin. –

2013. “Archaische Stadtsiegel” aus Mesopotamien und Iran. N. Crüsemann et al. (eds.), Uruk. 5000 Jahre Megacity. Berlin, 146–147.

Nissen, H. J., 1966. Zur Datierung des Königsfriedhofes von Ur unter besonderer Berücksichtigung der Stratigraphie der Privatgräber. Beiträge zur ur- und frühgeschichtlichen Archäologie des Mittelmeer-Kulturraumes 3, Bonn. Pettinato, G., 1999. La proprietà fondiaria nella Mesopotamia di 3. millennio dal periodo di Gemdet Nasr alla 3a. dinastia di Ur. H. Klengel/J. M. Renger (eds.), Landwirtschaft im Alten Orient. Ausgewählte Vorträge der XLI. Rencontre Assyriologique Internationale, Berlin, 4.–8. 7. 1994. Berliner Beiträge zum Vorderen Orient 18. Berlin, 99–113. Pittman, H., 2012. Glyptic Art of Konar Sandal South, Observations on the Relative and Absolute Chronology in the Third Millennium BCE. H. Fahimi/K. Alizadeh (eds.), NĀMVARNĀMEH. Papers in Honour of Massoud Azarnoush. Tehran, 79–94. Pomponio, F., 1987. La prosopografia dei testi presargonici di Fara. Studi Semitici 3, Rome. Sallaberger, W., 2010. The City and the Palace at Archaic Ur. K. Saskova et al. (eds.), Shepherds of the Black-headed People: the Royal Office vis-à-vis Godhead in Ancient Mesopotamia. Plzeň, 31–38. Steinkeller, P., 1988. Grundeigentum in Babylonien von Uruk IV zur früdynastischen Periode II. B. Brentjes (ed.), Das Grundeigentum in Mesopotamien, Jahrbuch für Wirtschaftsgeschichte. Berlin, 11–27.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

26 –

Benati – Lecompte 2013. An Archaic “Prisoner Plaque” from Kiš. Revue d’assyriologie et d’archéologie orientale 107, 131–157.

Sürenhagen, D., 1999. Untersuchungen zur relativen Chronologie Babyloniens und angrenzender Gebiete von der ausgehenden ʻUbaidzeit bis zum Beginn der Frühdynastisch II-Zeit.1. Studien zur Chronostratigraphie der südbabylonischen Stadtruinen von Uruk und Ur. Heidelberger Studien zum alten Orient 8, Heidelberg. Visicato, G., 2000. The Power and the Writing. The Early Scribes of Mesopotamia, Bethesda. Woolley, C. L., 1929. Excavations at Ur, 1928–1929. Antiquaries Journal 9, 305–348. –

1932. Excavations at Ur, 1931–1932. Antiquaries Journal 12/4, 355–392.

–

1934. The Royal Cemetery. A Report on the Predynastic and Sargonid Graves Excavated Between 1926 and 1934. Ur Excavations 2, London/Philadelphia.

–

1956. The Early Periods: a Report on the Sites and Objects Prior in Date to the Third Dynasty of Ur Discovered in the Course of Excavations. Ur Excavations 4, London/Philadelphia.

Wright, H. T., 1969. The Administration of Rural Production in Ancient Mesopotamia. Anthropological Papers – University of Michigan, Museum of Anthropology 38, Ann Arbor. Zettler, R. L., 1989. Pottery Profiles Reconstructed from Jar Sealings in the Lower Seal Impression Strata (SIS 8–4) at Ur: New Evidence for Dating. A. Leonard Jr./B. Beyer-Williams (eds.), Essays on Ancient Civilization Presented to Helen J. Kantor. Studies in Ancient Oriental Civilizations 47. Chicago, 369–393. –

1996. Written Documents as Excavated Artifacts and the Holistic Interpretation of the Mesopotamian Archaeological Record. J. S. Cooper/G. M. Schwartz (eds.), The Study of the Ancient Near East in the 21st Century. Winona Lake, 81–101.

–

2003. Reconstructing the World of Ancient Mesopotamia: Divided Beginnings and Holistic History. Journal of the Economic and Social History of the Orient 46, 3–45.

Zimmerman, P. C., 1998. A Critical Reexamination of the Early Dynastic «Royal Tomb» Architecture from Ur. M.A. Thesis, University of Pennsylvania.

>> Table 1. Catalog of the tablet fragments from the floor of the Ancient Room (BM = British Museum big number; IM = Iraq Museum catalog number; 37-07-0XX = Penn Museum catalog number).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Light on the Early Archives from Ur: The «Ancient Room» Tablet Hoard

27

Excavation number

Museum number

References

Excavation number

Museum number

References

U.12576

IM 049807

UET 2: no. 69

U.12811

BM 128968

UET 2: no. 79

U.12578

37-07-013

UET 2: no. 70

U.12812

IM 049725

UET 2: no. 83

U.12579

IM

UET 2: no. 27

U.12813

IM 049736

UET 2: no. 85

U.12580

BM 128905

UET 2: no. 71

U.12814

37-07-020

UET 2: no. 84

U.12581+ U.12586+ U.12587+ U.12810+ U.12818a

IM

UET 2: no. 112

U.12815

IM 049686

UET 2: no. 82

U.12582

BM 128928

UET 2: no. 68a

U.12816

37-07-018

UET 2: no. 81a

U.12583

IM 049835

UET 2: no. 22

U.12817a

IM 049733

UET 2: no. 89

U.12584

37-07-026

UET 2: no. 120

U.12819

BM 128896

UET 2: no. 87

U.12585

-

UET 2: no. 72

U.12822

-

UET 2: no. 88

U.12588

BM 128918

UET 2: no. 31

U.12823

37-07-022

UET 2: no. 92

U.12589

BM 128893

UET 2: no. 24

U.12824

IM 052227

UET 2: no. 93

U.12590

IM 049802

UET 2: no. 373

U.12825

BM 128909

UET 2: no. 86

U.12591

37-07-014

UET 2: no. 73

U.12826

IM 052230

UET 2: no. 95

U.12592

-

UET 2: no. 29

U.12827

BM

Nisaba 25: no. 37

U.12593

IM 049808

UET 2: no. 25

U.12828

IM 049789

UET 2: no. 98

U.12594

IM 049735

UET 2: no. 17

U.12829

BM 128923

UET 2: no. 94

U.12595

BM 128935

UET 2: no. 21

U.12830

IM 049759

UET 2: no. 96

U.12597

IM 049734

UET 2: no. 30

U.12831

37-07-024

UET 2: no. 100

U.12598

IM 049763

UET 2: no. 23

U.12832

37-07-023

UET 2: no. 99

U.12599

BM 128906

UET 2: no. 28

U.12833

BM 128922

UET 2: no. 97

U.12801

IM 049737

UET 2: no. 36

U.12834

IM 049717

UET 2: no. 101

U.12802

IM 049673

UET 2: no. 32

U.12835

IM 049793

UET 2: no. 102

U.12803

37-07-007

UET 2: no. 33

U.12836

IM 049718

UET 2: no. 103

U.12804

IM 049731

UET 2: no. 34

U.12837

IM 049770

UET 2: no. 105

U.12805

37-07-015

UET 2: no. 74

U.12838

IM 049676

UET 2: no. 104

U.12806

BM 128945

UET 2: no. 75

U.12839

BM 128958

UET 2: no. 106

U.12807

37-07-016

UET 2: no. 76

U.12840

BM

Nisaba 25: no. 38

U.12808

BM 128948

UET 2: no. 77

U.12809

37-07-017

UET 2: no. 78

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

28

Benati – Lecompte

Excavation number

Museum number

References

Excavation number

Museum number

References

U.12841+ U.12889b

IM 049834

UET 2: no. 108+153

U.12870 U 12870bis

BM 128952

UET 2: no. 139 Lecompte 2014: no. 39

U.12842

IM 049794

UET 2: no. 109

U.12871

IM 049687

UET 2: no. 129

U.12843

BM 128903

UET 2: no. 116

U.12872

37-07-027

UET 2: no. 121

U.12844

BM 128937

UET 2: no. 115

U.12873

IM 049728

UET 2: no. 135

U.12845

IM 049694

UET 2: no. 111

U.12874

37-07-033

UET 2: no. 136

U.12846

BM 128917

UET 2: no. 118

U.12875

BM 128951

UET 2: no. 149

U.12847

37-07-025

UET 2: no. 107

U.12876

37-07-034

UET 2: no. 141

U.12848

IM 049730

UET 2: no. 117

U.12877

BM 128965

UET 2: no. 134

U.12849

BM 128933

UET 2: no. 110

U.12878

37-07-039

UET 2: no. 150

U.12850

BM 128902

UET 2: no. 122

U.12879

-

UET 2: no. 144

U.12851

37-07-028

UET 2: no. 123

U.12880

37-07-032

UET 2: no. 133

U.12852

IM 049785

UET 2: no. 119

U.12881

37-07-037

UET 2: no. 146

U.12853

IM 049761

UET 2: no. 128

U.12882

37-07-038

UET 2: no. 148

U.12854

BM 128915

UET 2: no. 114

U.12883

37-07-036

UET 2: no. 145

U.12855

37-07-031

UET 2: no. 130

U.12884

BM 128959128960

UET 2: no. 151

U.12856

37-07-029

UET 2: no. 124

U.12885

37-07-040

UET 2: no. 152

U.12857

37-07-030

UET 2: no. 126

U.12886

BM

Nisaba 25: no. 40

U.12858

IM 049829

UET 2: no. 132

U.12887

IM 049756

UET 2: no. 156

U.12859

BM 128989

UET 2: no. 125

U.12888(b?)

BM

Nisaba 25: 41

U.12860

-

UET 2: no. 135bis

U.12888a

BM 128941

UET 2: no. 154

U.12861

IM 049803

UET 2: no. 131 + UET 2: 192

U.12890

IM 049680

UET 2: no. 162

U.12862

IM 049753

UET 2: no. 137

U.12891

IM 049780

UET 2: no. 163

U.12863

-

UET 2: no. 127

U.12892

IM 049765

UET 2: no. 159

U.12864

IM 049755

UET 2: no. 140

U.12893

37-07-042

UET 2: no. 158a

U.12865

-

UET 2: no. 136bis

U.12893

BM 128955

Unpublished? Labelled on CDLI as UET 2, 158c

U.12866

37-07-035

UET 2: no. 142

U.12893b

BM 128956

UET 2: no. 158b

U.12867

IM 052228

UET 2: no. 143

U.12894

BM 128914

UET 2: no. 161

U.12868

IM 049766

UET 2: no. 138

U.12895

IM 049839

UET 2: no. 164

U.12869

IM 049813

UET 2: no. 147

U.12896

BM 128992

UET 2: no. 157

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Light on the Early Archives from Ur: The «Ancient Room» Tablet Hoard

29

Excavation number

Museum number

References

Excavation number

Museum number

References

U.12897

BM 128987

UET 2: no. 165

U.12920

IM 049801

UET 2: no. 180

U.12898

IM 052231

UET 2: no. 160

U.12921

IM 049674

UET 2: no. 179

U.12899

BM 128898

UET 2: no. 167

U.12922

BM

Nisaba 25: no. 42

U.12900

IM 052229

UET 2: no. 166

U.12924

BM

Nisaba 25: no. 43

U.12901

IM 049760

UET 2: no. 168

U.12925

BM

Nisaba 25: no. 44

U.12902

37-07-043

Unpublished? Labelled on CDLI as UET 2, 169b

U.12926

BM

Nisaba 25: no. 45

U.12902a+ U.12911b

37-07-050

UET 2: no. 169a + UET 2: no. 182b

U.12927

BM

Nisaba 25: no. 46

U.12902b

37-07-046

UET 2: no. 174

U.12928

37-07-053

UET 2: no. 194

U.12903a

BM 128949

UET 2: no. 171a

U.12929

IM 049822

UET 2: no. 190

U.12903b

BM 128950

UET 2: no. 171b

U.12930

IM 049828

UET 2: no. 192 + UET 2, 131

U.12904

IM 049757

UET 2: no. 175

U.12931

IM 049696

UET 2: no. 195

U.12905

37-07-044

UET 2: no. 170

U.12932

BM 128936

UET 2: no. 191

U.12906

37-07-048

UET 2: no. 177

U.12933

IM 049782, 2

UET 2: no. 193a

U.12907

IM 049754

UET 2: no. 178

U.12933

IM 049782, 5

UET 2: no. 193b

U.12908

37-07-047

UET 2: no. 176

U.12933

IM 049782, 4

UET 2: no. 193c

U.12909

37-07-045

UET 2: no. 172

U.12933

IM 049782, 1

UET 2: no. 193d

U.12910

BM 128988

UET 2: no. 173

U.12934

BM 128983

UET 2: no. 196a

U.12911a

37-07-049

UET 2: no. 182a

U.12934

BM 128986

UET 2: no. 196b

U.12912

BM 128908

UET 2: no. 181

U.12934

BM 128985

UET 2: no. 196c

U.12913

BM 128897

UET 2: no. 183

U.12934

BM 128984

Unpublished? Labelled on CDLI as UET 2, 196d

U.12914

BM 128894

UET 2: no. 184

U.12935

BM 128957

UET 2: no. 197

U.12915

IM 049682

UET 2: no. 186

U.12936+ U.12944

IM 049779

UET 2: no. 199

U.12916

BM 128895

UET 2: no. 185

U.12937+ U.12961a

IM 049713

UET 2: no. 201

U.12917

37-07-051

UET 2: no. 187

U.12938+ U.12951b + U.12964

IM 049781

UET 2: no. 226

U.12918

BM 128921

UET 2: no. 188

U.12939

IM 049800

UET 2: no. 202

U.12919

37-07-052

UET 2: no. 189

U.12940

IM 049685

UET 2: no. 203

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

30

Benati – Lecompte

Excavation number

Museum number

References

Excavation number

Museum number

References

U.12941°

37-07-055

UET 2: no. 205a

U.12960a

IM 049719

UET 2: no. 227

U.12941b

37-07-056

UET 2: no. 205b

U.12960b

-

UET 2: no. 211

U.12942

37-07-057

UET 2: no. 206

U.12960c

37-07-061

UET 2: no. 220a

U.12943

37-07-054

UET 2: no. 204

U.12960d

37-07-062

UET 2: no. 220b

U.12946

BM 128942

UET 2: no. 207

U.12961b+ U.12963

37-07-064

UET 2: no. 229

U.12947

IM 049767

UET 2: no. 198

U.12961c

IM 049827

UET 2: no. 228

U.12948

IM 049724

UET 2: no. 208

U.12962

IM 049791

UET 2: no. 230

U.12949

IM 049683

UET 2: no. 209

U.12965

IM 049714

UET 2: no. 17

U.12950+ U.12953a

IM 049784

UET 2: no. 235

U.12966

IM 052226

UET 2: no. 231

U.12951a

BM 128954

UET 2: no. 210

U.12967

BM 128892

UET 2: no. 225

U.12952a

BM 128938

UET 2: no. 213

U.12968

37-07-063

UET 2: no. 221

U.12952b

BM 128939

UET 2: no. 214

U.12969

IM 049812

UET 2: no. 224

U.12953b

BM

Nisaba 25: 47

U.12954

IM 049804

UET 2: no. 212

U.12955a

37-07-058

UET 2: no. 217

U.12955b

IM 049839

UET 2: no. 200

U.12956a

IM 049811

UET 2: no. 216

U.12956b

BM 128920

UET 2: no. 215

U.12957

IM 049805

UET 2: no. 222

U.12958

IM 049716

UET 2: no. 223

U.12959a

37-07-059

UET 2: no. 218

U.12959b

37-07-060

UET 2: no. 219

Giacomo Benati, University of Bologna, [email protected]. Camille Lecompte, CNRS – Nanterre, [email protected].

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9th ICAANE, Basel 2014, Vol. 3, 31–43

Eloisa Casadei

Southern Mesopotamian Contexts and a Reevaluation of the Ur III Pottery. Reconsidering a Chronological Periodization The transition between the end of the Akkadian and the beginning of the Isin-Larsa period still constitutes an open archaeological problem. Focusing on correlations between different selected sequences, the present work proposes preliminary remarks for a ceramic periodization on the late 3rd millennium BC southern Mesopotamia. In the last few years, a reconsideration of the much-published problem of the period between 3rd and 2nd millennium BC generated new significant interpretations and theories, pointing to a concomitance of different anthropic and climatic factors.1 The similarity of the stratigraphic sequences across Greater Mesopotamia has suggested in many instances a straight connection between different aspects of the daily and political life. The impact of the Akkadian economic and political control in a wider region than the southern alluvial plain, and the spread of the trade connection controlled by the kings of the 3rd Dynasty of Ur (hence Ur

1

In 2012, H. Weiss proposed a workshop related to the end of the 3rd millennium BC in the Khabur region. He presented his theory in seven points: starting from 2300 BC northern Mesopotamia was under the economic control of the southern Akkadian empire. But at ca. 2200 BC, an abrupt climate change caused the aridification of the whole near eastern region. This event persisted ca. 300 years and caused a reduction in sedentary settlement assessment and an arrest in the cereal agro-production. Even the Akkadian territory of the northern Mesopotamia and the adjacent nonimperialized regions underwent the same situation, and the reduction in agro-revenues caused the Akkadian imperial collapse (Weiss 2013: 2–4). On the contrary, G. Schwartz, analysing a wider region from the Alluvial Plain to the Upper Tigris and Upper Khabur regions, asserts there was no single, abrupt catastrophe, and that some regions were more seriously affected than others. In his view, major socio-political and economic changes characterized the transition between Early and Middle Bronze age. In some cases, archaeologists recognized significant changes in the urban assets; but in others, a marked continuity points to a gradual and slow process of transformation (Schwartz 2012: 255–256).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

32

Casadei

III), could generate a chain of interconnections throughout Southern Mesopotamia.2 Nevertheless, though Southern Mesopotamia should represent the focus of many political events, the lack of reliable archaeological data impacts the construction of a sound periodization that links the southern evidence with the other regions. In fact, while the high quantity of historical data had driven the historians in the reconstruction of the chain of events from the decline of the Akkadian dynasty to the collapse of the Ur III kingdom, the archaeological contexts related to this period appear scattered on the territory, and sometime affected by issues with past excavations. Pointing to the transition between the Late/Post-Akkadian Period and the end of the Ur III period, only selected southern stratigraphic sequences can offer a good overview of the gradual changes in the pottery production. Furthermore, information could be gained about how long the old and new characters maintain continuity (McMahon 2012: 25). The present paper intends to detect those morphological elements characterizing the end of the 3rd millennium BC in Southern Mesopotamia, trying to define their chronological trend.3 For this purpose, four major sites have been chosen, according to the availability of pottery repertoire, the nature of the post-Akkadian contexts, and the state of publications.

1. Selected Contexts 1.1 Area TB and the WF Sounding at Nippur The accurate excavation carried out at Nuffar/Nippur allowed the reconstruction of one of the most reliable stratigraphic sequences for Southern Mesopotamia. In area WF, it is possible to follow the gradual transition of a domestic unit, from the Akkadian to the Ur III period. It consists of a 10 x 10 m square, 12 m deep, representing a sequence of 19 levels from ED III to the 1st millennium BC (McMahon 2006: 6). Originally, the area was dedicated to an open space for production activities, probably involving different families. In level XIV, the area underwent a radical change, becoming part of a single unique building, and the plan remained almost the same until level VII, when the area was abandoned. Shortly after, a new occupational phase (level VI) should date to the transition between Ur III/Isin-Larsa, while levels X–VII are ascribed by the excavator to the Ur III period on the base of ceramic parallels with area WC-3, on the fortification line on the south-west

2 3

For an historical overview see Orsi 2010; Schmidt 2005; Sallaberger 1999. The present paper was presented during the poster session of the 9th ICAANE 2014, and it is a preliminary presentation of the MA dissertation on the Ur III phases in Southern Mesopotamia and a redefinition on the pottery assemblage (Casadei in progress).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Southern Mesopotamian Contexts and a Reevaluation of the Ur III Pottery

33

end of the site, securely dated by stratigraphical consideration. In area TB in the so-called Tablet Hill, an Akkadian domestic quarter occupied the lowest levels of the sounding (level XIII–XI). After a period of abandonment (level X) a great complex probably devoted to administrative functions was in use during the Ur III period (IX–V), and relatively abruptly disused at the beginning of the Isin-Larsa period, when a new domestic quarter was built (McCown/Haines 1967; Stone 1987).

1.2 The Northern Quarter at Tell Asmar/Eshnunna At Tell Asmar/Eshnunna, starting from the Akkadian level (Level V) the northern domestic quarter suffers a gradual abandonment. Pertaining to all the occupational phases, House II or the Arch House (Hill 1967: 154–156) may offer another good context for a better defined ceramic sequence. Delougaz attributed level V to the Akkadian period, level IV to a late Akkadian phase, level III to Ur III and level II–I to Isin-Larsa. Gibson proposed a better defined subdivision on the basis of pottery parallels with other excavation, proposing a very late date in the Akkadian period for the later part of level IV (IV a) (Gibson 1982: 531). The Northern Palace was destroyed by fire, as shown by the presence of a thick ash level called the Grey Layer (Hill 1967: 195). After this event, new structures seem to be domestic in character and they present two occupational sub-phases. According to Delougaz, they are dated to the Akkadian and Isin-Larsa periods, respectively. Gibson, on the contrary, preferred to attribute the earlier level to the Ur III period, being the second phase of the palace dated to a late Akkadian phase. The chronological relationship between the domestic quarter and the palace area is still a matter of discussion (see Gibson 1972).

1.3 Pit 5197 and Pit 6G65 at Abu Salabikh The two post-Early Dynastic contexts from Abu Salabikh represent what probably was the last occupational phase of the settlement, as the correlated structures were seriously damaged (being immediately under the surface) (Postgate/Moon 1984). This site has not returned a proper stratigraphic sequence related to the period under discussion. Regardless, the quantity of material documented and its clear post-Early Dynastic date make this material quite useful.

1.4 The Eastern Area of the Royal Cemetery at Ur Finally, at Tell al-Muqayyar/Ur, only the south-eastern sector of the Royal Cemetery produced a sequence from the late Akkadian to the latest part of the Isin-Larsa period. In fact, some of the «Sargonid» and «Second Dynastic» graves were damaged during the construction of the Royal Mausoleums, while different Ur III cylinder seals confirm a later date in the Akkadian period for most of the tombs (Woolley 1934; Buchanan 1954; Moorey 1984). In fact, dated to a period between the last part of the Akkadian dynasty and the rising of the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

34

Casadei

Ur III kingdom, a series of rich built-tombs are attested in Ur, following the traditions of the funerary customs of ED III–Akkadian period. In particular, PG.1847 was partly cut by the construction of the south-west wall of Shulgi’s mausoleum, and the presence of the cylinder seals representing worshippers with only three figures points to a Neo-Sumerian horizon that fits well in a transitional phase between the two dynasties, as stated both by stratigraphy and pottery repertoire. The Mausoleum complex represents the only context securely dated to the Ur III period, and the only one from which a good amount of pottery related to this period was collected (Woolley 1974). In fact, during the supremacy of the Amorite reigns of Isin, Larsa and Babylon, the whole town was reconstructed after the great destruction by the Elamites (Woolley 1974; 1976).

2. Pottery Analysis For this paper, five forms have been chosen as case studies, representing the most characteristic shapes of the southern repertoire during the last quarter of the 3rd millennium BC. They will be analysed in order to note their chronological significance in the southern pottery periodization.4

2.1 Carinated Bowls with Molding Rims (fig. 1) They consist in shallow bowls with straight or slightly concave sides, flat or disc bases.5 The vertical rims are molded in different ways, from shallow everted to emphasize carinated. Their beginning in the last quarter of the 3rd millennium BC is rather undefined. 4

5

Parallels will be made with the materials from the Jazirah and and the Khabur region, and in particular with Tell Brak/Nagar and Tell Mozan/Urkish. These two sites took on a particular role in the political and economic relation with the kings of Akkad and the kings of Ur respectively. The ceramic resemblance of the repertoire with those of the south should help to define, with some respect, the ceramic evolution at the end of the 3rd millennium BC, although direct correlation between such distant regions have to be considered with caution (for a chronological periodization related to the Syrian Jazira, see Lebeau 2011). The first site, Tell Brak, returned a transitional phase (phase N) post-dating the Akkadian period but which could not be placed in a proper EJZ5/Ur III horizon (Oates et al. 2001). During this period the occupational assessment changed, reducing in size and becoming a smaller urban center (Colantoni 2012, 59). Tell Mozan is the only site with a well preserved EJZ 5 repertoire, coming from the House of Pusham in area C (Pfälzner 2012; Schmidt 2013). ARCANE type 117 (Rova 2011: pl. 21: 10–15), attested only during EJZ 5 and not earlier. This form is present at Tell Brak in area SS, level 2 and 1 (phase N, post-akkadian period), but at Tell Mozan (Type 1.04 in Schmidt 2014) it appears only in EJZ 5 contexts (area C2, phases 7–6b, and area AA stratum 4). In particular, in area C2 it is attested for the first time in level C7 at 9, 5 %, and it represents the most popular bowl shape. During the succeeding phases, it progressively goes out of use, being replaced by a bowl with swollen rim (type 1.02).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Southern Mesopotamian Contexts and a Reevaluation of the Ur III Pottery

Fig. 1. Carinated bowls.

Fig. 2. Pear-shaped jars with double ridged rim.

Fig. 3. Pear-shaped jars with triple ridged rim.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

35

36

Casadei

The earliest contexts in the Southern Mesopotamia are dated to the late Akkadian or the very beginning of the Neo-Sumerian period, namely from level XI in the WF sounding and in level IX in area TB at Nippur. Here in level IX to level VII, it is attested only by fragments, and from level VI on its frequency increases. At Eshnunna, it appears for the first time in phase Va of the Private Quarter, a little earlier than in the South. At Ur, the earliest context is represented by the «Sargonid Grave» 659 of the Royal Cemetery. They are also attested in the post-Akkadian deposits from Abu Salabikh. It is ubiquitous in Ur III contexts in the south, and it seems to be well attested in a wider territory.6 According to Armstrong and Gasche (2014, type 20 E1), this type is no longer predominant after the 18th century, as no examples were found in WB IV at Nippur (Gibson 1975; Gibson et al. 1978), and so representing a typical production of the Ur III/Isin-Larsa period (Armstrong/Gasche 2014: 26).

2.2 Pear-shaped Jars with Ridged Rims (fig. 2–3) These characteristic small jars have a straight or flourished neck, an emphasized shoulder and a pointed base.7 It is possible to identify two subtypes, one with a double and one with a multiple ridged rim. The first type (fig. 2) is a typical form of the end of the 3rd millennium BC, appearing from the Akkadian to the Neo-Sumerian period. According to Armstrong and Gasche, the earlier examples have a lower and less pronounced shoulder (Armstrong/Gasche 2013: 45, n. 192). At Ur, it is attested in the graves PG.1850 and PG.1851, belonging to the «Second Dynasty» group of the cemetery. In the Nippur sounding in area WF, it appears for the first time in level XII (late Akkadian) and continues until level VIII (Ur III).8 The second sub-type, characterized by a triple or multiple ridged rim (fig. 3), seems to appear almost simultaneously in all the southern sites.9 At Ur it is present both in the «Sargonid Grave» 1039, in the so-called Shulgi Mausoleum, and in a funerary context dated to the beginning of the 2nd millennium BC, showing a longer life than its corresponding type. In the northern region it seems less common than in the South and only one vessel is attested until now, coming from the layer below the Ituria Palace at Eshnunna, and so pre-dating the

6

7 8 9

As stated out by McMahon, being a typical southern shape, one could expect it not to overlap with the Ur III reign or area of influence. According to her, their presence in a wider area proves a sort of persistence in shared fashion (McMahon 2012: 32). Armstrong and Gasche type 125–130. The two sub-types are sometimes considered together, in spite of their clear shape difference. In the Jazirah, it is attested at Tell Brak, and in area C2 at Mozan, but only in level C7 (EJZ 5, Ur III). According to Schmidt (2013, K 480, type 34.101), only at 0, 1%. Unfortunately, McCown and Haines did not distinguish this form from another one, with simple rounded rim, attested in other contexts quite later than the triple ridge rim (McCown/Haines 1967, type 14A).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Southern Mesopotamian Contexts and a Reevaluation of the Ur III Pottery

37

Fig. 4. Multiple ridged rim craters.

Shu-Sin reign (Delougaz 1952, B.645.540a). The continuation of this form later in the Isin/ Larsa period is not attested.10

2.3 Multiple Ridged Rim Craters (fig. 4) This type is represented by large open vessels, quite deep, with vertical or slightly everted rim modeled with a series of ridges on the exterior side.11 Usually, a combed decoration on the shoulder is characterized by a wavy motive bounded by two horizontal bands. In other cases, only horizontal bands are present on the upper part of the body.12 In Southern Mesopotamia, many sherds were found in post-ED layers at Abu Salabikh (pit 5179 and area 6G65) and in Nippur (area TB level VII and area WF level VI). From Ur, two specimens

10 In the Jazirah, it is attested at Brak only in one specimen and it seems not presented at all at Mozan.

However, it seems possible that a single sherd could belong to this type (Schmidt 2013, K124).

11 Armstrong/Gasche (2014: 69, type 265 A3) pointed out a progressively simplification of the rim

molding, particularly visible at Ur. They also highlight the presence of a hole at the base.

12 This form is well attested at Tell Mozan from area C2 (stratum C7) and area A (stratum 4b), both

dating to EJZ5. They seem to be rare in this early phase, but their percentage growth progressively in the following phases (Schmidt 2013, type 40). It is less common at Tell Brak where only rims of this kind were found in layers related to phase N.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

38

Casadei

come from Amar Sin’s Mausoleum, but many different kinds of basin are used as coffins in the IL period, showing a particular function of this form, even though the type of rim changed slightly. The excavation at Eshnunna returned one feature from the layer below Ibiqadad II palace (beginning of the 2nd millennium BC). Noteworthy is the fact that, until now, this particular type of combed decoration is completely absent in earlier southern contexts, and it seems to be characteristic of the MB I Syrian repertoire (Pruss 2007: fig. 4).

2.4 Triangular Rim Jars with Ridged Shoulder (fig. 5) This particular type of jar presents an emphasized shoulder, often with a series of horizontal ridges, and it terminates with a short, everted neck and a triangular rim.13 A sort of short pedestal is attached at the pointed bottom in a second phase of creation, a very common technique in all the 3rd and 2nd millennium BC pottery in southern Mesopotamia.14 In Southern Mesopotamia, it is characteristic of the late Akkadian period, as demonstrated by its presence in level XI of area TB at Nippur, in level IVa of the Private Quarter, and from the «Sargonid Grave» 653 from the Royal Cemetry of Ur. Nevertheless, it seems to continue throughout the first part of the Ur III period at least, being attested from level X to level VI in the WF sounding at Nippur, in the second phase of the Northern Palace at Eshnunna and in the Amar Sin’s Mausoleum.

2.5 Collared Jars with Everted Rim and Droop Below It (fig. 6) These jars have a globular shape with straight, vertical or slightly everted narrow neck, and short shoulder decorated with horizontal or wavy combing. The rim is externally expanded and is characterized by a droop on the lower side. This type is well attested in both the northern and the southern region.15 At Nippur, it is attested in level VIII in area TB and it spread from level VIII to level VI in the WF sounding. An example identical to those from TB was found above the Bilalama phase of the Palace of Eshnunna, dated to the first phase of the IL period (Delougaz 1967: C.656.340). At Abu Salabikh, it is characteristic of the

13 A variant with simple shoulder is defined ARCANE type 104 and it could be interpreted as an ante-

cedent of our type even in the south (as happened at Nippur, the Diyala and Ur).

14 This jar is rather frequent at Tell Brak, where it appears in area SS (level 2), area FS (level 2a), and

area DH (level 1b), mostly dated to the post-Akkadian period (phase N). On the contrary, it does not seem to be present at all at Tell Mozan. According to McMahon, its introduction from south to north during the Akkadian period might visually materialize southern management of agricultural products. Its persistence during the post-Akkadian period should be considered difficult to explain (McMahon 2012: 31). 15 At Tell Mozan, in area A (stratum 4b) and in area C2 (stratum C7), both related with EJZ5. A specimen comes from area SS level 2 at Tell Brak, dated to phase N. Another example (Oates et al. 2001: fig. 425.801) shows an identical shape but its surface is characterized by a high grooving, coming from FS level 1/2 (phase N).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Southern Mesopotamian Contexts and a Reevaluation of the Ur III Pottery

Fig. 5. Triangular rim jars with corrugated shoulder.

Fig. 6. Collared jars with everted rim and droop below.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

39

40

Casadei

post-ED layers in area 6G65 and in pit 5179. From Ur, a specimen comes from the «Sargonid Grave» 666 and another from the Shulgi’s Mausoleum.

3. Conclusions As is shown in the graphs, some remarks can be made about the types analyzed (fig. 7). Carinated bowls are a typical post-Akkadian shape, attested for the first time at the very beginning of the Ur III period (but pre-dating the rise of the dynasty), and it remained in use until the first half of the Isin/Larsa period. This type is represented in different regional variants, as shown by the earlier specimens at Nippur. Its absence from the levels of the 2nd millennium domestic quarters at Ur might point to a late phase of these contexts during the Larsa period, at least later than the Isin/Larsa levels IV–I of area TB at Nippur. The pear shaped small jar with ridged rim and a single ridge behind it is attested for the first time at a very late date in the Akkadian period, probably shortly before the appearance of the carinated bowl. Until now, it seems not to continue after the first half of the Ur III period, showing a very short period of attestation. On the contrary, its counterpart with a triple or multiple ridged rim appears almost at the same time, but in lower percentage. After half of the Ur III period, when the type with a double ridge disappeared, it became more popular and remained attested until the beginning of the 2nd millennium BC, like the carinated bowl. The multiple ridged rim crater is probably part of a wide pottery tradition that would become very popular during the Middle Bronze Age I. The southern Mesopotamian subtype with multiple ridged rim might represent a local variation, already attested at the end of the 3rd millennium BC, but gradually evolved in a simpler type, closer to the western morphology. The storage jar with corrugated shoulder also is representative of a ceramic period beginning shortly before the Ur III dynasty and dropping out at some point before the end of the period, being absent in early 2nd millennium BC contexts. Finally, the globular jar with the droop below the rim constitutes a particular form not always recognized in the old typology. However, the peculiarity of the rim could be taken as a marker of the Ur III period, but this hypothesis can be confirmed only by further excavations. The work presented here should be considered as a preliminary investigation of the most representative contexts and materials related the Neo-Sumerian period. The appearance and disappearance of the pottery types from one phase to the next do not match the historical (and conventional) phases used in southern Mesopotamia. The next step of the research should be a detail analysis on the stratigraphy of past and future excavations, in order to develop a better-defined ceramic periodization.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Southern Mesopotamian Contexts and a Reevaluation of the Ur III Pottery

Fig. 7. Chronological span of the types (light gray: less than 10%).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

41

42

Casadei

Bibliography Armstrong, J./Gasche, H., 2014. Mesopotamian Pottery. A guide to the Babylonian tradition in the Second Millennium B.C. Chicago. Buchanan, B., 1954. The Date of the so-called Second Dynasty graves of the royal cemetery at Ur. Journal of the American Oriental Society 74, 147–153. Colantoni, C., 2012. Touching the void. The Post-Akkadian viewed from Tell Brak. Weiss 2012, 45–64. Delougaz, P., 1952. Pottery from the Diyala region. Oriental Institute Publications 63. Chicago. Dohmann-Pfälzner, H./Pfälzner, P., 2002. Ausgrabungen der Deutschen Orient-Gesellschaft in der zentralen Oberstadt von Tall Mozan/Urkeš. Bericht über die in Kooperation mit dem IIMAS durchgeführte Kampagne 2001. Mitteilungen der Deutschen Orient-Gesellschaft 134, 149–192. Gibson, M. G., 1975. Excavations at Nippur. Eleventh season. Oriental Institute Communication 22, Chicago. –

1982. A Re-Evaluation of the Akkad Period in the Diyala Region on the basis of recent Excavations at Nippur and in the Hamrin. American Journal of Archaeology 86, 531–538.

Gibson, M. G., et al., 1978. Excavations at Nippur. Twelfth season. Oriental Institute Communication 23. Chicago. Hill, H. D., 1967. Tell Asmar. The Private house Area. P. Delougaz et al., Private Houses and Graves in the Diyala Region, Chicago, 143–266. Laneri, N./Pfälzner, P./Valentini, S. (eds.), 2012. Looking North: The Socio-Economic Dynamics of the Northern Mesopotamian and Anatolian Regions during the Late Third and Early Second Millennium BC. Studien zur Urbanisierung Nordmesopotamiens D1. Wiesbaden. Lebeau, M. (ed.), 2011. ARCANE: Associated Regional Chronologies for the Ancient Near East and the Eastern Mediterranean, Vol. 1: Jazirah. Brepols. McCown D.E/Haines, R. C., 1967. Nippur I: Temple of Enlil, Scribal Quarter, and Soundings. Oriental Institute Publications 78. Chicago. McMahon, A., 2006. Nippur V: The Early Dynastic to Akkadian Transition. The area WF sounding at Nippur. Oriental Institute Publications 129. Chicago. –

2012. Post-Akkadian ceramic assemblages of the central Upper Khabur. Weiss 2012, 25–43.

Moorey, P. R. S., 1984. Where Did They Bury the Kings of the IIIrd Dynasty of Ur? Iraq 46, 1–18. Oates, D./Oates, J./McDonalds, H., 2001. Excavations at Tell Brak, Vol. 2: Nagar in the Third Millennium BC. British School of Archaeology in Iraq and Cambridge. London. Orsi, V., 2010. Crisi e Rigenerazione nella valle dell’Alto Khabur (Siria). La produzione ceramica nel passaggio dal Bronzo Antico al Bronzo Medio. Firenze.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Southern Mesopotamian Contexts and a Reevaluation of the Ur III Pottery

43

Pfälzner, P., 2012. The Question of Desurbanisation versus Reurbanisation of the Syrian Jazirah in the Late Third and Early Second Millennium BC. Laneri et al. 2012, 51–80. Postgate, N./Moon, J., 1984. Late Third Millennium pottery from Abu Essalabikh. Sumer 39, 69–79. Pruss, A., 2007. Comb-incised pottery in Syria and Mesopotamia and its relevance for chronology. P. Matthiae et al., From Relative Chronology to Absolute Chronology: the Second Millennium BC in Syria-Palestine. Proceedings of the International Colloquium (Rome, 29th November – 1st December 2001), Roma, 473–542. Rova, E., 2011. Ceramic. Lebeau 2011: 49–127. Sallaberger, W., 1999. Ur III-Zeit. W. Sallaberger/A. Westenholz, Mesopotamien. Akkade-Zeit und Ur IIIZeit. Annäherungen 3, OBO 161/3. Fribourg/Göttingen. Schmidt, C., 2005. Überregionale Austauschsysteme und Fernhandelswaren in der Ur III-Zeit. Baghdader Mitteilungen 36, 7–151. –

2013. Die Keramik der Früh-Ğazīra-V- bis Alt-Ğazīra II-Zeit. Studien zur Urbanisierung Nordmesopotamiens A4. Wiesbaden.

Schwartz, G. M., 2012. Northern Exposures: Third to Second Millennium BC Transformations in Upper Mesopotamia. Laneri et al. 2012, 256–263. Stone, E., 1987. Nippur Neighbourhoods. Studies in Ancient Oriental Civilization 44. Chicago. Weiss, H. (ed.), 2012. Seven Generations since the Fall of Akkad. Studia Chaburensia 3. Wiesbaden. Woolley, L. C., 1934. Ur Excavations I. The Royal Cemetery. London. –

1974. Ur Excavations VI. The Buildings of the Third Dynasty, London/Philadelphia.

Woolley L. C./Mallowan, M., 1976. Ur Excavations VII. The Old Babylonian Period. London.

Eloisa Casadei, Sapienza Università di Roma.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 45–55

Franco D’Agostino – Ali Khadem Ghanim – Licia Romano

Abu Tbeirah. Preliminary Report of the 2012–2013 Campaigns Acknowledgments Our work in Southern Iraq is possible thanks to the invaluable help of many friends who backed us from the very beginning. Beyond our good friend Abdulamir Al Hamdani, who was the first to introduce us to Abu Tbeirah, Ali Kadhem, Wussal Jasim, Wasan Isa, and all the members of the State Board of Antiquities and Heritage of DhiQar supported us in every possible way with enthusiasm and competence, and the same is true of the archaeo-logical authorities of Baghdad, the former Minister Liwaa Sumaisem, Ahmed Kamil, Rasheed Qais, Abbas Husseini and the entire staff of the Ministry for Tourism and Archaeology. The friendship of HE the Governor of Dhi Qar, Dr Yahya Al-Nasri, is a honor for the Mission, and His sincere interest in the Cultural Heritage is making possible a restart of the scientific activities and of a harmonious touristic development of the region. Both Amir Doshi, Journalist and man of culture of Nasiriyah, and Jasim Al Asadi, Director of Nature Iraq, are close friends whose knowledge of the cultural and natural heritage of the area is a true inspiration for us. Daher Al-Bakka’, who is in charge of our security, is the friend who makes our everyday life easy and comfortable. Finally, we want to mention here Amir, Haider, Nghamesh, Takleef and all the wonderful workers who dig with us in Abu Tbeirah.

1. Introduction1 Abu Tbeirah is located 16 km NE of Ur and 6 km south of Nasiriyah. The Tell is 43 ha wide and is crossed by an old channel running SE-NW and by a pipe-line running from south to north (fig. 1), installed on the site after the ‘60s (as it can be shown by the Satellite CORONA 1

§§1 and 4 are common work of the authors; Franco D’Agostino is author of §2 up to §2.2 and Licia Romano of §2.3 and §3.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

46

D’Agostino – Ghanim – Romano

imagery, taken in 1968). The pipe-line is not the only anthropic activity that damaged the tell: the NW part is now completely missing due to the exploitation of the soil for industrial purposes, and in the NE part several pits have been excavated by the Iraqi army to host antiaircraft gun emplacements. Nevertheless, no traces of intentional looting have been detected on the site. Three excavation campaigns have been carried out and three are the areas so far excavated; a survey of the site was the first activity.2 A stratigraphic trench has been excavated using one of the army pits: the pottery discovered spans from the Akkadian (perhaps Ur III) period to the ED III.

2. Area 1 – SE Most of the activities of the Mission have been focused on the south-eastern area, chosen from the very beginning on the base of the satellite imagery,3 showing a huge building under the surface. The salty crust has been removed for an area of 750 sqm, highlighting several graves immediately under the surface, belonging to the last phase of occupation of the area.

2.1 The Cemeterial Area (fig. 2) The seven graves excavated so far show a singular variety in the inhumation procedure. Grave 6 and Grave 1, excavated one near the other, could belong to a same kin group on the basis of the big distance between the area occupied by these two graves and the other ones. Moreover, inside Grave 6, over the feet of the female first occupant a big group of commingled bones has been found. The study revealed that these bones belonged to a male adult and moreover the two bodies showed a common malformation of the humerus (a small hole that does not show any evidence during the life, due to a genetic characteristic or to physical stress), proving that both the individuals belonged to the same lineage or, less probably, made the same work.4 Thus, it can be argued that the body of the man was exhumed after the death of the woman, and his bones have been recomposed over her feet after her inhumation (first the head and over it the other bones).

2 3 4

For a preliminary report of the survey and of the excavations see: D’Agostino/Romano 2014; D’Agostino et al. 2011 and 2013. We want to acknowledge here the kindness of Prof. Elizabeth Stone (Stony Brook University), who put at our disposal the satellite imagery of the site in her possession. See Mary Anne Tafuri’s analysis of the skeletal remains in D’Agostino et al. 2013: 90 fig. 25.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Abu Tbeirah. Preliminary Report of the 2012-2013 Campaigns

Fig. 1. Abu Tbeirah Satellite imagery (September 14th 2014 copyright e-Geos).

Fig. 2. Area 1 SE - Plan of the Cemeterial Area (2013).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

47

48

D’Agostino – Ghanim – Romano

Fig. 3. Area 1 SE – view of Building A (2013).

This is not the only case of a secondary deposition in the cemetery. Grave 3 in fact was a small pit containing part of a skull and a tooth of a male adult, accompanied by a small equipment made up of a jar, 2 bowls and 3 beakers. Beside Grave 11, a simple pit hosting the body of a perinatal individual, another Grave (16) has been partially excavated during the last season: the adult male individual was accompanied by a rich pottery equipment, though apparently partially looted in antiquity. Grave 15, instead, is an inhumation in sarcophagus robbed and partially destroyed in antiquity but preserving part of the lid still in place. Only the legs of the individual were preserved though in bad conditions due to the weight of the same lid of the sarcophagus. The coffin was surrounded by a huge number of jars and bowls. This is not the only sarcophagus of the cemetery: another one (Grave 2) was completely destroyed by ancient looters while two more have been identified but not yet excavated. The fact that the walls of the big building detected through the satellite imagery were probably only partially visible when the graves were excavated is demonstrated by the fact that some of them cut the preexisting structure.

2.2 Building A – Phase 1 (figs. 3–4) This building, brought to light during the field activities, is a mud-brick structure made up by several rooms arranged around a big court of ca. 80 sqm. A row of rooms runs parallel to the court on the SW, and at least other four perpendicular rooms lay on the east side. Deep rainfall gullies run on the NE corner of the building, eroding the upper part of the structure that now is hardly visible immediately under the salt crust. The excavation of the building led to the identification of two phases of use, distinguished by a different internal circulation pattern and by the use of two distinguishable pavements. The first phase of Building A has been investigated in 6 rooms. Only one line of mudbricks was preserved, though heavily eroded: only 30 cm of soil separated the upper part of the wall from the superficial salt crust. The dimensions of the dark yellowish brown mudbrick seem to have been of 30/35 cm × 15/17 cm, although the measuring were not accurate due to the bad state of preservation. The walls were erected over pre-existing structures, using a stratum of clay to even out and stabilize the base of the new walls.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Abu Tbeirah. Preliminary Report of the 2012-2013 Campaigns

49

Fig. 4. Area 1 SE – Plan of Building A – phase 1 (2013).

Inside Room 1 there was a beaten earth floor with a fireplace in the middle. Along the NE wall a pottery dish with burnt fish bones has been found. In the courtyard the first phase was indicated by a Tannur surrounded by an ashy area. 3 beakers and a shell have been found inside the Tannur and some scraples (AbT.12.135; 143), a pestel (AbT.12.136), pebbles (AbT.12.139) and a bronze fragment (AbT.12.137) have been collected in the surrounding area together with several fragments of bowls, beakers and burnt animal bones.5

5

Mostly pigs, ungulates, goats and shells.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

50

D’Agostino – Ghanim – Romano

While Phase 1 of Room 3 was almost empty, 6 inside Room 4 the first phase was signaled by the presence of 5 so-called foundation bowls.7 Under the pavement a grave (Grave 12) of a female adult of 20–30 years was discovered. The equipment was made up of several stone pestles (AbT.13.19, 36, 38, 44), a fragment of an alabaster vessel (AbT.13.37) and a bivalve shell (AbT.13.39 – Unio Tigridis) with cosmetic paste inside a small reed basket near the left hip of the woman. The pottery assemblage deposed near the body consisted of 3 beakers, 8 conical bowls, fragments of a stemmed dish, a small jar with globular body and a spouted jar. Room 5 was probably exposed for a longer period to water erosion or was damaged in a worse way due to the passage over it of a gully created by the rainfall. Over the pavement an alabaster vase and a limestone one (AbT.13.74–75) have been discovered in pieces spread all over the room. Under the pavement a grave of a child of ca. 6 years has been discovered, completely covered by pure clay. His equipment consisted of a jar, two beakers and two conical bowls. Remains of a reed-mat have been discovered near the body. Room 6 first phase ground surface was in bad condition like the one in Room 5; only two foundation bowls have been found along the NW wall. Outside the building, a big jar decorated with wavy incisions has been recovered in association with a Tannur. Here a big grave has been found with three different inhumations of one subadult and two children. At least 261 drinking vessels in fragments have been discovered in the three graves (more than 200 only in Grave 5); they are clearly related to the funerary banquet.

2.3 Building A – Phase 2 (fig. 5) The second phase of the building has been investigated in Room 1, Room 3 and Room 4. The circulation system in these two rooms was different compared to the later Phase 1: in Room 1 only the door along the NW wall was present. The construction technique foresaw the utilization of dark yellowish brown mud-bricks, of 30 × 15 × 15 cm (the dimensions of the mud-bricks were, however, irregular also in the same wall). The adobes were put in rows, with alternation of two stretchers and one header. Clay was used as mortar. Room 1 Phase 2 was signaled by a reed-mat covering the pavement and preserved especially along the walls. The reed-mat was pierced by a hole whose walls were covered by the same vegetal weave: this clearly indicates that the hole was made after the reed-mat was placed over the pavement and, as some picture from the marshes can show, it hosted probably one of the poles used for holding the reed cover of the building. It is worth noting that 6 7

Only a fragment of a limestone vessel with indentation has been recovered (AbT.12.167). They were located along the SW wall (in the central part and southern part), along the NE wall (in its southern part) and the NW wall.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Abu Tbeirah. Preliminary Report of the 2012-2013 Campaigns

51

Fig. 5. Area 1 SE – Plan of Building A – phase 2 (2013).

the poles used in the Marshland constructions have the bottom tapered (Young 1977: 115, 158): a similar situation can be supposed for the Abu Tbeirah evidence. In the middle of the room there was a fireplace with some postholes around it, probably belonging to a structure which has a parallel from Abu Salabikh (Postgate/Moorey 1976: 143). Under the pavement a grave of a child of 5–6 years has been found (Grave 7). The equipment was composed by a plain rim jar and a small one with band rim and a hole near the base. Inside Room 3 over the ground surface a copper-alloy chisel (AbT.12.164) has been found together with a pebble and a base of a limestone vessel (AbT.12.162–163). An angular small Tannur was located in the NE corner of the room. Under the pavement three graves have been found.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

52

D’Agostino – Ghanim – Romano

Fig. 6. Area 2 NE - Plan of the latest activities in the Area (2013).

Grave 8 was cut by a pit excavated from the pavement above it. Only the lower part of the body of a 3 years old child was recovered with a trumpet base jar and another small double ridged rim jar as equipment. Inside Grave 10 an infant of 6–9 months was buried with fragments of conical bowls and bones of sheep and small ungulates. Grave 9, instead, contained not only the body of a 6 years old child, accompanied by a small reed basket and a spouted vessel, but also remains of two 6 months babies, deposed near the right hip of the child and near his left shoulder. Inside Room 4 the pavement was realized with clay and fragments of mud-bricks. Also, some conical bowls have been discovered along the walls. Near the door towards Room 3 two reed baskets have been found still in place. The pavement was partially dismantled in the middle of the room, showing pre-existing structures under it, apparently delimiting at least two rooms filled with pure sand. Fragments of plano-convex mud-bricks have been found in the debris of the building.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Abu Tbeirah. Preliminary Report of the 2012-2013 Campaigns

53

Fig. 7. Area 2 NE - Donkey grave.

3. Area 2 – NE The second area excavated is located in the NE and has been chosen, in agreement with the local SBAH authorities, because of the discovery of equipment of several graves eroding out from the surface. The excavations revealed a complex system of pits and drainages and at least three graves belonging to the last phase of occupation of this part of the site and dated to the Akkadian period on the basis of the pottery (fig. 6). In Grave 100 an important assemblage was found consisting of numerous jars and conical bowls, three bronze vessels (AbT.12.32; 36; 38), a stone vessel (AbT.12.35), a dagger (AbT.12.31), a so-called «toilet-set» (AbT.12.34), and three precious carnelian beads (AbT.12.33a–c) from the Indus Valley. Grave 101 contains the inhumation of a ca. 2 years old child, buried inside a reed cradle with a unique plain rim jar. This procedure of burying a child in his cradle is still in use among the Ma’dan of the Marshes (Jasim al-Assadi pers. comm.). Inside Grave 102 the body of an adult, whose sex cannot be determined, has been found with few pottery vessels and a shell with cosmetic paste inside as his equipment. Among the pottery vessels a dark gray burnished jar with four vertical handles has been found, which can be dated to the Akkadian period (Gibson 1972: 271).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

54

D’Agostino – Ghanim – Romano

Fig. 8. Area 2 NE – Plan of Building B and Building C (2013).

Among the anthropic activities for the drainage discovered in the middle of the excavated area, a pit clearly distinguishes itself, not only for the pottery recovered inside (in particular a boat clay model (AbT.13.18) and a jar with a stylized bull-head depicted on it), but also for the discovery of a complete skeleton of a dog. Probably the pit represents a grave destroyed by later activities on the spot. Immediately under the salt crust that characterizes Abu Tbeirah surface, a complete skeleton of a donkey (fig. 7), in poor state of preservation, was recovered from a pit. The animal was resting on its left side with tightly flexed limbs (especially the front ones), but the head was positioned in an unnatural way, on the right shoulder as if the neck had been forcedly broken or cut. Also in this case, a similar situation is to be found in Abu Salabikh (Postgate/Moorey 1976: 156, Pl. XXIVa). The morphology of the teeth indicates asinine features rather than hemione or horse ones. The animal was probably a male about 5–5.5 years old.8 The graves and the drainage system insisted on a house quarter: two buildings and a narrow street between them have been found and only partially excavated (fig. 8). Of Building B a courtyard and a small room have been highlighted so far. In the courtyard there was a row of three Tannur; the walls surrounding the courtyard were literally «melted» and the pavement was completely cleared by the flowing of water. This part of the NE area is in fact crossed by several gullies excavated by seasonal rainfalls. Among the few 8

The skeletal remains of the donkey have been analyzed by Francesca Alhaique. DNA analysis are now in progress in collaboration with Univesità degli Studi della Tuscia.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Abu Tbeirah. Preliminary Report of the 2012-2013 Campaigns

55

objects and pottery found, a diorite door-socket (AbT.13.58) and fragments of a stone vessel (AbT.13.61) have been discovered. Apparently the building lies on a previous structure. Only one room of Building C has been excavated in this campaign: the building located in the highest part of the area was not affected by the flowing of waters and some pottery was found in situ on the pavement of the room. Also in this case, lower structures were visible under the building.

4. Conclusions In its last phases Abu Tbeirah seems to have witnessed a reduction and an impoverishment of the settlement. The buildings so far discovered were poor and edified with walls made up by badly shaped bricks and covered with reed roofing, like in the contemporary buildings of the Ma’dan of the Iraqi marshes. In this phase, the settlement seems to have been shifted in the NE part that shows a longer continuity of life, as demonstrated by the pottery discovered belonging at least to the Akkadian period. Further excavations will surely contribute to the understanding of the fascinating and complicate period of transition from the ED to the Akkadian period also in southern Mesopotamia.

Bibliography D’Agostino, F./Romano, L., 2014. Excavation at Abu Tbeirah, Southern Iraq. P. Bieliński et al. (eds.), Proceedings of the 8th ICAANE 30rd April–4th May 2012, University of Warsaw. Wiesbaden, 51–65. D’Agostino, F., et al., 2011. Abu Tbeirah. Preliminary Report of the First Campaign (January–March 2012). Rivista di Studi Orientali 84, 17–34. –

2013. Abu Tbeirah. Preliminary Report of the Second Campaign (October–December 2012). Rivista di Studi Orientali 86, 69–92.

Gibson, Mc., 1972. Umm el-Jīr, a Town in Akkad Source. Journal of Near Eastern Studies 31, 237–294. Postgate, J.N./Moorey, P.R.S., 1976. Excavations at Abu Salabikh, 1975. Iraq 38, 133–169.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 57–66

Paul Zimansky – Elizabeth Stone

Tell Sakhariya and Gaeš Much of Tell Sakhariya, 7 km east of Ur, consists of clean, packed clay. For the Ur III and Old Babylonian periods surface collection and soundings unearthed fragments of royal inscriptions, but no domestic structures. The geographical location and character of the remains, which was seasonally settled in the Kassite period, suggest identification of this site with Gaeš, focal point of Ur’s Akiti festival. At the end of the 3rd millennium BCE, when southern Mesopotamia was ruled from Ur, the city festivals of the moon god, Nanna/Suen, became, in essence, festivals for the kingdom as a whole. The most important of these were the great Akiti festivals, which took place twice a year: for eleven days at the beginning of the 7th month in association with the season of planting, and for seven days at the beginning in the 1st month, that is to say the new year’s festival, which coincided with the harvest (Sallaberger 1993: 170–171). Other Mesopotamian cities celebrated the new year with other gods in other eras–most conspicuously the Neo-Babylonian new year’s festival for Marduk–but the Ur festivals are the ones for which we have the most detailed early documentation. A key aspect of both the planting and the harvesting Akiti festivals was the journey of the god by boat from the city of Ur through agricultural fields to the Temple of Nanna, the Karzida, located at a site called Gaeš. That Gaeš was overwhelmingly ceremonial is suggested by the fact that it was sometimes simply called Karzida, equating the temple with the place itself (Falkenstein 1959: 150). Cuneiform evidence (Edzard/Farber 1974: 50–51) suggests something of what one might expect to find at the site, but not without ambiguities. Karzida means «good harbor or quay» and a canal connected it to Ur, which was less than a day away by barge. The temple was probably not a single structure but rather a complex, which included the residence of the En-priestess of Nanna and a palace at which the king resided when performing ceremonies. There was also an Akiti House or temple. It is unclear whether or not this was part of the Nanna Temple, but associated with it was a storehouse for royal offerings, a courtyard, and a field for growing barley. A cattle pen at Gaeš is also mentioned (Cohen 1993: 407–408). Although the site certainly existed in some form earlier, major building projects

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

58

Zimansky – Stone

Fig 1. Satellite photo marking locations of Ur, Tell Sakhariya, and Carroué’s sketch location of Gaeš (geo-referenced in relation to major sites).

were undertaken there by the Ur III kings Shulgi and Amar-Sin. It continued to be important ceremonially in the Isin-Larsa period, despite the diminished status of Ur itself, and it disappears from the historical record in the Old Babylonian period. That Gaeš is near Ur has never been disputed, but a specific tell is harder to pin down. François Carroué, discussing Sumerian geography relevant to the Iturungal watercourse, published a map tentatively locating Gaeš a little more than six kilometers east northeast of Ur (Carroué 1993: 63). No ancient written source gives any specific information on the direction or distance of Gaeš from Ur, so let us chalk it up to coincidence rather than authority, that Tell Sakhariya, a site that the authors investigated in 2011 and 2012, lies almost precisely at the geo-referenced location of Gaeš in the Carroué sketch (fig. 1). In this communication, we present our archaeological findings and suggest, although certainly cannot prove, that the two locations may indeed be one and the same. We did not set out to find Gaeš. Had we even thought about it, the abundance of Kassite archaeological material at Sakhariya would have discouraged any such identification. We were, in fact, looking for something that Gaeš definitely is not: an ordinary village of the Old Babylonian period. At the risk of digressing, let us outline the objectives of the project that brought us to this place and describe what we recovered archaeologically before returning to the subject of Gaeš. There is a disparity between the abundance of village sites in Mesopotamian texts and dearth of excavated small sites, and very little understanding of how smaller rural settlements relate to urban centers in terms of material culture. Before Saddam’s invasion of

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Tell Sakhariya and Gaeš

59

Kuwait, we had worked at Mashkan-shapir, a major city (Stone/Zimansky 2004). When it became possible to excavate in southern Iraq in 2011 after a hiatus of more than twenty years, the idea of having a look at a smaller site next to Ur immediately suggested itself as a project that could be executed in a field season or two. Our preference was for a site dating to the early 2nd millennium both because the textual record for that period is relatively rich and large habitation areas previously excavated at Ur could be used for comparative purposes. High-resolution satellite imagery of the area had become available in the last decade to help us select a site, and the project would have the added benefit of enabling us to check the relationship between surface traces of architecture visible in these and what actually lay under the ground. Initially, we focused on the small mound of Madthuriya, seven km to the northwest of Ur, but on a visit there in early July 1011 were confounded to discover that it was millennia younger than we had expected. Almost on a whim, we then paid a fleeting visit to Tell Abu Ba’arura Shimal, a site Henry Wright had noted as Kassite in his survey of the early 1960s (Wright 1981: 339, cat. no. 20). On this occasion we found that some of the pottery around the edges of the site looked Old Babylonian, although there was indeed classic Kassite pottery on summit. On a second visit a few days later, during which we were able to explore the topography of the site more fully, Abdulamir al-Dafar who was assisting us on in our search, discovered a large inscribed clay nail datable to Rim-Sin year 15, or 1807 BCE by the convention of the Middle Chronology (fig. 5c). Satellite imagery, at which we had a close look after we returned the United States in September, revealed traces of a square, fort-like building, but we could not identify what Wright had described as a «recent qal’a ruin». We assumed that an excavation at the site would encounter a late Kassite occupation, below which we would find levels dating to the early 2nd millennium. Since other sites in southern Iraq were abandoned in the period between these phases, we weren’t expecting to find anything from the middle of the 2nd millennium. With virtually no bureaucratic problems, we assembled a team1 and began a five-week season on December 15, 2011. Before we began, we thought it appropriate to change the name of the tell from «Abu Ba’arura» (father of sheep droppings) to «Sakhariya» to impart more dignity to our project. Our approach to Tell Sakhariya was an updated version of what we did at the urban site of Mashkan-shapir twenty-five years ago. We began with a surface survey, sketching features and collecting objects and diagnostic ceramics. These confirmed our initial impression that the site dated to both Old Babylonian and Kassite periods. Little patterning was seen in the 1

The expedition consisted of Stone and Zimansky (directors), Irene Winter, Stephanie Rost, John McGuinnes, Demetrios Brellas, and Abdulamir al-Dafar. In the field our Iraqi representatives were Saleem Khalaf (Director of Research for the State Board of Antiquities and Heritage in Baghdad), Ali Khadhem (Director of Antiquities for Dhi Qar Governorate), and Wasa Abdulsahib of the Nasiriyah Museum. Funding for this project was generously supplied by the National Geographic Society.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

60

Zimansky – Stone

Fig. 2. Contour plan of site with distribution of objects and location of trenches.

scatter artifacts, which were generally plebeian (fig. 2). One thing was striking however: the surface of this tiny site yielded four inscribed brick fragments (fig. 5f–j), more than we found in our quite rich survey of the 65 hectare surface of Mashkan-shapir. These inscriptions were too incomplete to assign to a specific ruler, but collectively affirmed erstwhile royal interest in the site. We conducted a magnetic field gradient survey in the hope of identifying sub-surface architecture, as has been done so successfully at Uruk. In this, however, we had no success whatsoever, for reasons that became clear once we started digging. We used the satellite imagery to identify architectural traces as we laid out trenches for soundings in areas where the imagery suggested the presence of structures. We were not optimistic that this architecture would in fact appear in excavation, since we understood that in many cases the visible lines are simply salinized remains of very shallow ghost walls. We also added a few five-byfive meter trenches near the edges of the site to get a feeling for the extent of the tell.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Tell Sakhariya and Gaeš

61

Fig. 3. Boom photo of kiln and burial.

We never actually found the walls of the satellite photos in our excavations, but the upper, dusty, salt layer was full of late Kassite ceramics. We also encountered a number of late Kassite burials. An infant–probably a neonate–was buried in a typical late Kassite jar with a goblet base of the same date used as its stopper. In what we interpret as an open area within the square structure we found a burial consisting only of a skull and a tibia–we have no idea what happened to the rest of the body. It was, however, accompanied by three late Kassite vessels, so once again its date is secure. Also probably to be dated to the late Kassite are two pottery kilns. The best preserved of these was cut into the ground, with access from the top (fig. 3). Somewhat earlier debris appeared in a series of ashy trash deposits which contained both pot sherds and animal bones. These seem to have been left by transient populations who stayed at the site intermittently in the mid 2nd millennium. This occupational level also contained a few bricks which appear to have been originally created for earlier structures. In no case did we uncover substantial buildings. The ceramics from this occupation changed rapidly as we dug down, and comparisons with examples from northern Babylonia indicate that they span roughly the early to mid Kassite period. It was from these levels that we recovered the majority of our finds at the site in this short field season, including virtually all of our faunal and floral evidence. The object inventory was quite mundane. The lithic items were for the most part grinding stones, cuboids, and the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

62

Zimansky – Stone

like, but also included one well-made balance weight. Reused sherds that had been rounded into discs and pierced in the center appear to have served as spindle whorls. The conical stone whorls in vogue at Mashkan-shapir and other Old Babylonian urban sites were not found here. Sakhariya produced nearly as many chipped stone tools–including a number of cores–as it did fragments of bronze. This ratio is at odds with the much greater abundance of metal that we found at Mashkan-shapir. Sakhariya is unusual for a 2nd millennium Mesopotamian site in the near absence of terracotta figurines. More precious metals included three pieces of gold–all very small–and a group of even smaller silver objects recovered through flotation. Other items of personal adornment were comparatively common, as one would expect in a mobile population. Many beads of various types of stones were found in both survey and excavation, together with cowry shells cut in the same way those used for jewelry today. Floral and faunal remains reinforce the idea that the occasional occupants of Sakhariya had a varied diet. Preliminary identifications of the plant remains by Dr. Michael Charles of Oxford University did include small amounts of barley, but more significantly many wild plants almost certainly brought in with animal dung and Scorpio, the raw material for reed mats and the huts made from them. Sheep are the animals most commonly represented among the site’s faunal remains, which were analyzed by Dr. K. Twiss of Stony Brook University. These are followed by cattle, donkeys or wild onagers, pigs, dogs, gazelles and goats. These species indicate that Sakhariya’s occupants occasionally exploited arid environments as well as wetter zones, and they supplemented their primarily domesticated meat supply with small numbers of hunted animals. All in all, it would appear that Sakhariya, in this phase, was simply a place where nomads cycled through seasonally, taking advantage of the high ground to keep their feet dry. But where did this high ground come from? This was the issue that frustrated our search for an Old Babylonian village and now takes us back to the subject of Gaeš. Sakhariya was no ordinary tell; rather than rising through the gradual accumulation of layers of habitation debris, it appears to have been created by an act of deliberate human construction. Almost everywhere that we dug in the central part of the site, we ran into heaped up clean clay–so hard it was almost impossible to excavate with our picks and in many places lacking any trace of pottery or other artifactual material. It looked very much as if we were digging platforms that had been created to support sacred buildings. We encountered this material at exactly the same elevation in our two initial trenches, even though they were 55 meters apart. In the western trench the clay was completely clean, hard and sterile–not a single sherd was found. In the trench to the east, it was equally clean and hard, but we discovered a baked brick platform within it, associated with some ceramics and animal bones. This, however, was an anomaly. We cored several meters below it and came up with only clean

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Tell Sakhariya and Gaeš

63

Fig. 4. Fish pond with drain and stratigraphy in section.

mud. The edge of the clay mass was encountered a few meters further east within the next trench, but time prevented us from exploring this edge to any depth. Perhaps contemporary was a broad area of clean water-lain silt with many fish bones, some of which represented whole fish apparently trapped between layers of sediment. At the base of this material we found a well-made drain or channel of baked brick whose base was so level we could not figure out which direction water was meant to flow through it (fig. 4). Could this be a fish pond? It would have been quite a feat to keep a large basin full of water at two meters above plain level, especially considering the aridity and high temperatures of Iraqi summers. The inscriptions we found in this season speak in a consistent voice for royal interest despite the fact that none come from their original context. We have already mentioned the clay cone of Rim-Sin, discovered on the surface in our summer visit, and the four fragments of brick inscriptions (fig. 5a). These are not objects that one would expect to find in domestic contexts and are prima facie evidence of royal building activity. We also found a complete Ur-Namma brick inscription describing the construction of a barag–a pedestal or podium– and a garden (fig. 5h). Two clay cones were found in eroded material near the edges of the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

64

Zimansky – Stone

Fig. 5. Inscriptions found at Tell Sakhariya: inscribed pegs (A-C); tablets (D-E); brick inscriptions (F–J). C & H are reproduced at half the scale of the other objects.

mound (fig. 5a–b). Two Sumerian tablets were among the early Kassite materials along with re-used old Babylonian bricks. One of these is very poorly preserved, and doesn’t tell us much. The other is a quite well preserved receipt for copper utensils dated to year 28 of Shulgi (fig. 5d). The reason it is so well preserved is that it came from inside an early Kassite drinking vessel.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Tell Sakhariya and Gaeš

65

In short, the many royal inscriptions, clay platforms–even the putative fish pond–all indicate that Sakhariya was an important place in the Ur III to Isin-Larsa periods. There is simply too much of this material to support any notion that it might all have been brought here from somewhere else, by design or chance. Yet we have almost nothing in the way of living debris for these periods, and nothing to indicate that a substantial population lived here. There are baked bricks of the sizes that one would expect in an Isin-Larsa site, but no Isin-Larsa structures. The model that best explains this is that Sakhariya was once a much taller mound covering a smaller area. After the Isin-Larsa period, the upper parts of the mound eroded away. Indeed, we find the sloping layers of erosion in our smaller trenches around the edges of the site. However disappointing this was to us in our quest for a village of the Isin/Larsa or early Old Babylonian period, it does at least re-introduce the question of Gaeš. It is a peculiar site indeed that consists of lots of clean, packed mud, numerous fragments of royal inscriptions of the Ur III to Isin-Larsa periods and almost no evidence of living debris. Yet isn’t this precisely what would one expect to find at Gaeš? The distance and location are appropriate for the connections with Ur by canal within an easy day’s journey. The clay platforms would thus have supported the important Akiti house and/or the Karzida complex. Given that fish offerings were apparently part of the Akiti festival (Cohen 1993: 412), the pond is not out of place. Most of the physical evidence for structures on top of the platform, however, has been eroded away. The three distinct occupations attested by archaeological evidence would fit well with the history of Gaeš. From Ur III to early Old Babylonian times there was considerable royal building activity as attested by written documents. The ceramics found within the eastern edge of the platform date to the Sealand Period – if we may be so bold as to use this term for the phase that separates the Old Babylonian from the late Kassite period – and are found in association with reused building materials. This occupation must follow the debacle which affected all of southern Mesopotamia following the rise of Rim-Sin II of Larsa and his subsequent defeat by Samsuiluna. Sakhariya served as the summer camp-ground for a marsh-dwelling population during the next centuries as the site eroded and its cultic functions were forgotten, or at least no longer the focus of royal attention. Like other sites in the south, Sakhariya was re-occupied in late Kassite times, but this is only apparent in the architecture of the surface salt layer and in the pottery, kilns and burials that form the latest excavated materials at the site. None of this, of course, is good news for those who would hope to unlock more the mysteries of Ur’s grand Akiti festivals through archaeological excavation. If Tell Sakhariya is indeed Gaeš, the Karzida is gone.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

66

Zimansky – Stone

Bibliography Carroué, F., 1993. Etudes de Géographie et de Topographie Sumériennes III. L’Iturungal et le Sud Sumérien. Acta Sumerologica 15, 11–69. Cohen, M., 1993. The Cultic Calendars of the Ancient Near East. Bethesda, Maryland. Edzard, D. O./Farber, G., 1974. Répertoire géographique des textes cunéiformes, 2: Die Orts- und Gewässernamen der Zeit der 3. Dynastie von Ur. Wiesbaden. Falkenstein, A., 1959. akiti-Fest und akiti-Festhaus. R. von Kienle et al. (eds.), Festschrift Johannes Friedrich. Heidelberg, 147–182. Sallaberger, W., 1993. Der Kultische Kalender der Ur III-Zeit Teil 1. Untersuchungen zur Assyriologie und Vorderasiatischen Archäologie 7/1. Berlin. Stone, E. C./Zimansky, P., 2004. The Anatomy of a Mesopotamian City: Survey and Soundings at Mashkan-shapir. Winona Lake, Indiana. Wright, H., 1981. The Southern Margins of Sumer: Archaeological Survey of the Area of Eridu and Ur. R. McC. Adams, Heartland of Cities: Surveys of Ancient Settlement and Land Use on the Central Floodplain of the Euphrates. Chicago, 323–345.

Paul Zimansky & Elizabeth Stone, Stony Brook University, New York.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Iraq (Kurdistan)

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 69–82

Kyra Kaercher

A Preliminary Assessment of the Ceramic Sequence of Northeastern Iraqi Kurdistan This paper presents a preliminary assessment of the ceramic sequence from the Rowanduz Archaeological Program of Boston University. The ceramic collections from the Kurdistan Regional Government, Soran Directorate of Archaeology were studied and the resulting sequence spans the Neolithic to modern times.

1. Introduction1 Northeastern Iraq has been largely off limits to archaeologists since Robert J. Braidwood’s pioneering Iraq-Jarmo Project in the late 1940s and early 1950s. At the end of the First Gulf War, Coalition Forces established a safe haven in northern Iraq for the largely Kurdish population and by 1992 Iraqi Kurdistan, consisting of Iraq’s Dohuk, Erbil and Sulaymaniyah Provinces, emerged as a semi-autonomous entity, with its capital in Hawler/Erbil (O’Leary 2006: 24). With this autonomous zone, the Ministry of Tourism, under the Ministry of Municipalities, established the Directorate of Antiquities. The KRG General Directorate of Antiquities oversees the provincial directorates of Erbil, Dohuk, and Sulaymaniyah. In 2012 the General Directorate established Soran, up to then under Erbil, as an independent Directorate that includes the subdistricts of Harir, Khalifan, Diyana, Rowanduz, Sidekan, Mergasur, and the district of Choman (fig. 1). Faced with rapid economic and infrastructure 1

I would like to thank the Director General of Antiquities of the KRG, Othman Zendin Abubakr (Mala Awat), the Director of the Soran Directorate, Abdulwahhab Soleiman, and the staff of the directorate for their permission to study the collection. I also gratefully acknowledge the support and help of Dr. Michael Danti, director of RAP, Dr. Richard Zettler, and the entire 2014–2015 RAP Staff. Lastly thanks to the RAP funders, the Vecchiotti Archaeology Fund at Boston University, Joan Aruz, Curator of the Department of Ancient Near Eastern Art at the Metropolitan Museum of Art, and the University of Pennsylvania Museum, without whom we would not have had a field season.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

70

Kaercher

Fig. 1. Map of Iraqi Kurdistan showing its districts and subdistricts. Soran Directorate is outlined. Modified from Danti 2014: fig. 2.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A Preliminary Assessment of the Ceramic Sequence of Northeastern Iraqi Kurdistan

71

Fig. 2. Map illustrating the location of the 16 sites discussed in this paper. Map Credit: Marshall Schurtz.

developments, including oil exploration, the Soran Directorate has been active in both archaeological reconnaissance and salvage excavations. The surveyed pottery is helping to create a typology, and hence a chronology for this little studied area. This paper presents an analysis of the surveyed pottery. The Soran Directorate spans approximately 7, 500 square kilometers, stretching from the high Zagros to the Erbil plain. Diverse settlements, spanning many millennia of occupation, are scattered around this large region. The earliest site in the region dates from 65, 000 to 35, 000 years ago and contains evidence of Neanderthal occupation. Shanidar Cave, located in the northern part of the Soran Directorate, was excavated by Ralph Solecki from 1957–1961 (Solecki 1971). The valleys of the Zagros provided summer pastures for herders and land for settlements as early as the Pre-pottery Neolithic. Patty Jo Watson’s excavations at Banahilk, located near the modern-day town of Soran shed light on the early Halaf occupation in the area (Braidwood 1960: 33–35). Grotto shrines, rock inscriptions, and steles, like the Topzawa and Kel-i Shin stele, give testament to the Bronze and Iron Age occupations in the area (Edmonds 1966). The town of Rowanduz, located south of Soran was established as the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

72 Ceramic Periods Hassuna Halaf

Kaercher Archaeological Periods Neolithic

Northern Ubaid Early Uruk Late Uruk

7000-6000 BCE 6000-5000 BCE 5000-4000 BCE

Chalcolithic

Ninevite 5 Akkadian

Approximate Dates

4000-3000 BCE 3500-3000 BCE 3000-2500 BCE

Early Bronze Age

2500-2000 BCE

Middle Bronze Age

2000-1700 BCE

Late Bronze Age

1700-1300 BCE

Middle Assyrian

Iron I

1300-1000 BCE

Late Assyrian

Iron II

1000-700 BCE

Post Assyrian

Iron Ill

700-550 BCE

Achaemenid

Iron IV

550-350 BCE

Khabur Mitannian

Hellenistic Partho-Roman

Classical Antiquity

Sasano-Byzantine Early Islamic Middle Islamic

0-400 CE 400-700 CE

Islamic Period

Ottoman Recent

350-0 BCE

700-1000 CE 1000-1500 CE 1500-1900 CE

Recent

1900-Present

Fig. 3. Table showing the correlations between Ceramic Periods, Archaeological Periods and their Approximate Dates.

capital of the Sorani Emirate (1399–1835). During the 19th and 20th centuries, this area was occupied by the Russians, Turks, and British, each of whom left their traces on the archaeology (Eppel 2008). The Directorate supervises excavation and surveys sites uncovered by developers. The artifact assemblages in the Directorate are mostly surface collections from disturbed sites spanning the Neolithic to modern times. In the summer of 2013, the Rowanduz Archaeological Program (RAP) received a fiveyear permit from the Department of Antiquities of the Kurdistan Regional Government to survey and excavate in the Soran District. «This area represents one of the most compelling and unknown corners of the Near East, and as such, RAP seeks to revitalize archaeology in this area through the implementation of a long-term, integrative program of multidisciplinary archaeological research projects and cultural heritage management initiatives (Danti

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A Preliminary Assessment of the Ceramic Sequence of Northeastern Iraqi Kurdistan

73

Fig. 4. Graph showing the number of sites per period from the Soran Directorate Collections, showing 35 sites.

2014a: 50)». RAP aims to implement a long-term program of survey and excavation, as well as cultural heritage management initiatives, in the Directorate. As ceramicist of RAP, I studied the Soran Directorate’s ceramic collections that include assemblages from 40 different sites. The assemblages from some sites included only a few sherds, while others had over 100. This paper will present sites with collections of more than five sherds and will more thoroughly analyze those with more than twenty. Figure 2 shows the distribution of the sixteen sites discussed in this paper.

1.1. Time Periods The archaeological material of northeastern Iraq is little known. The time periods and associated dates are from the Saddam Dam Salvage Project, which lasted from 1983–1986 (Simpson 2008: iv). These have been correlated with the seventh edition of the Working Ceramic Typology composed by the directors of the Assyrian Landscapes Research Group (Ur 2013: 1). Figure 3 lists the period names and dates that are utilized in this paper.

1.2. Ceramic Typology The typology for the Rowanduz Archaeological Program is based on the excavations currently being carried out at Gird-i Dasht. Gird-i Dasht’s occupation spans the early 3rd millennium BCE to the present day. The typology is based on fabric color, surface treatment and temper. The majority of the ceramics can be matched to the typology from the Saddam

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

74

Kaercher

Dam Salvage Project and the Working Ceramic Typology (Simpson 2008; Ur 2013). These two typologies were created for the plains of Northern Mesopotamia, however, excavations at Gird-i Dasht have shown a connection with Northwestern Iran. Excavations from the sites of Hasanlu, Dinkha Tepe, Haftavan Tepe, and Godin Tepe provide comparanda for RAP’s typology. A new typology for this region is combining all this information into a usable format, and is being tested and refined as excavation is continuing.

2. Sites In figure 2, many sites seem to be clustered around the cities of Mergasur, Sidekan, Soran, Choman, and Harir. The mapped locations are based on the limited information we have about the actual location of some of these sites. The Directorate records the cities near the sites, but as we have not visited a majority of these sites, we have no GPS points to place on the map. Below is the description of 16 sites, with their general period, the type of site where known, and a brief analysis of the assemblage.

2.1. Hassuna Period (7000 BCE) Gird-i Khiwat2 is the only Hassuna period site in the Directorate’s collections. This surface collection was gathered in 2013 and contains 11 body sherds none of which are decorated. The ceramics are all handmade with chaff and grit temper. They range in color from red to brown. A stone pestle was collected with the assemblage, and with the ceramic’s construction, the site dates to the Hassuna Period (Ur 2013: T1/14).

2.2. Halaf Period (6000 BCE) Gird-i Banahilk is the only Halaf period site in the Directorate’s collections. It is a low oval mound located on a point that overlooks a tributary of the Balakiyan River, which empties into the Rowanduz Gorge, and was collected in 2012 and 2013. Patty Jo Watson placed four test trenches on the mound in 1954 (Braidwood 1960: 33). Watson states the pottery found in these excavations is part of an «Eastern Halaf Ware» and is closer in form to Tell Arpachiyah than to Tell Halaf (Braidwood 1960: 35). Watson reached virgin soil at four meters, with no evidence of earlier occupations. The Directorate’s collection includes 104 sherds. The 86 decorated sherds are highly painted with brown or red paint on a smoothed surface. Little to no temper is used, and they are well fired. One unusual group appears in the survey. This group of 13 ceramics 2

The normalization of Kurdish names into English is based on Thackston’s book, Sorani Kurdish: A Reference Grammar with Selected Readings (Thackston 2006).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A Preliminary Assessment of the Ceramic Sequence of Northeastern Iraqi Kurdistan

75

Fig. 5. Ceramic illustration. Banahilk nos. 1-12, Dashti Dearyan no.13-14, Ashkawt-i Kafran nos. 15-16, Bokadera nos. 17-27.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

76

Kaercher

has dark brown paint, on a fine red ware. This decoration is not found in typical Halaf collections, so it possibly marks a local variation of the general Halaf assemblage. Mostly geometric patterns are present including diamonds with dots, lattice patterns, chevrons, herringbones, and arcs (fig. 5.1–7). In the excavated ceramics, from 2014, two examples of an ibex appear, and Watson recorded two examples of birds (Braidwood 1960: 38). The forms of these vessels are flared beakers, deep bowls, and straight-sided bowls (fig. 5.8–12). This may denote a dining or food storage function for these decorated ceramics. The high percentage (83%) of painted sherds from survey could be sampling bias, where only decorated sherds were collected or saved. However, Watson noted that 60% of her sherds were painted, and excavations in 2014 carried out by RAP also record a high percentage of painted wares.

2.3. Iron II/Assyrian (1000–700 BCE) Dasht-i Deareyan and Gird-i Tile are two sites that date to the Assyrian Period. Dasht-i Deareyan was collected in 2011 and Gird-i Tile was collected in 2012. The rim sherds at both sites are from both straight-sided beakers (fig. 5.13) and jars that match sherds from Khirbet Qasrij (Curtis 1989: pl. 43, no. 321: pl. 35, no. 196: Ur 2013: T11/8). They are made of a red fabric with grit temper, and the ones from Dasht-i Deareyan have a white slip. The base from Dasht-i Deareyan is similar to a base found at Khirbet Qasrij (Curtis 1989: fig. 43 no. 306); however, instead of a straight line around the base, the addition is wavy. The decorated sherds at both sites contain combed wavy lines located around the shoulders (fig. 5.14).

2.4. Iron III–Achaemenid (700 – 350 BCE) Ashkawt-i Kafran, Bokadera, Ghabrestan-i Topzawa, Gund-i Topzawa, Hanara, and P. Muhammad Ahmed Haji all date to the Iron III–Achaemenid period (700 – 350 BCE). Ashkawt-i Kafran was collected in 2012 (fig. 5.15–16). Bokadera was collected in 2011, 2012, and 2013. Gund-i Topzawa and Ghabristan-i Topzawa are located along the north slope of the Topzawa Çay Valley located northeast of Sidekan. They were discovered during road widening work in 2013. Gund-i Topzawa is a 133 m stretch of cultural deposit, and Ghabrestan-i Topzawa is a stone-built tomb. Hanara was collected in 2012 and contains 110 pieces, with a majority of these sherds being body sherds with no decoration. P. Muhammad Ahmad Haji was collected in 2012 (fig. 6.4–5). The forms of the Iron III period include large jugs and bowls with handles and spouts (fig. 5.21–27). Sherds from Bokadera, Ashkawt-i Kafran, and Gund-i Topzawa date to this period, based on comparanda with the ceramics found in northwestern Iran, especially Hasanlu (Danti 2013: 194, 232–244), Dinkha Tepe IV (Muscarella 1968: fig. 13), Agrab Tepe (Muscarella 1973: fig. 14), and Bastam (Kroll 1979: 84).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A Preliminary Assessment of the Ceramic Sequence of Northeastern Iraqi Kurdistan

77

Fig. 6. Ceramic illustration. Ghabrestan-i Topzawa nos. 1-3, P. Muhammad Ahmed Haji nos. 4-5, Qalaat Lokhan nos. 6-11, Gird-i Dasht nos. 12-21, Paarcha Goza nos. 22-29, Gird-i Khiso no. 30, P. Jameel Daoud Ali nos. 31-33.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

78

Kaercher

The decorations on these ceramics are of a combed incised design, or an applied rope design, both on the shoulders of large jars (fig. 5.16, 18–19). The incised design has parallels with both the Assyrian heartland (Curtis 1989: pl. 43, no.299) and with northern Iran at Godin (Gopnik 2011: Pl. 11 no. 133). The applied rope design (fig. 5.26) is found at Khirbet Qasrij and Qasrij Cliff, as well as ceramics from Hasanlu III (Curtis 1989: fig. 7; Danti 2013: 232–242). Bokadera is the largest Iron III site in the Directorate’s collection with 123 sherds, the majority of which are large jugs and bowls with handles and spouts (fig. 5.17–33). It is a cave site and the vessels were probably used for water or food storage. The location near Choman could point to a cache for travelers using the Haji Omran Pass. This pass links the Diyana plain to northwestern Iran. The forms of the Iron IV/Achaemenid period include sherds from Ashkawt-i Kafran, Ghabrestan-i Topzawa, Harana, and P. Muhammad Ahmed Haji. A spouted jar (fig. 6.1) has parallels with Hasanlu IIIa (Young 1965: fig. 2 no. 5), Pasargade (Ghirshman 1954: pl. XXIV no. 6), and the Spring Cemetery at Persepolis (Schmidt 1957: pl. 89 no. 5). Another spouted jar has close parallels at Agrab Tepe (Muscarella 1973: fig. 15 no. 8), Hasanlu IIIa (Dyson 1965: 205, 212 n. 36), and Susa Achaemenid Village I (Ghirshman 1954: pl. XXIX, GS 2242). Two examples of a goblet base date to the Achaemenid period as based on the Working Typology (Ur 2013: T14/15). The flattened rims are comparable to Godin II, in central Iran (Gopnik 2011: no 123). Ghabrestan-i Topzawa is a stone lined tomb of Achaemenid date (550–350 BCE) that contains ceramics in a wide range of fine and utilitarian wares (fig 6.1–3). This site is the largest Achaemenid collection in the Directorate. Many of the vessels are poorly fired, and when placed in water disintegrate. The tomb vessels could be possible imitations of utilitarian vessels buried with the dead, as they would not function in life. The radiocarbon date near the ceramics is from approximately 300 BCE, placing the first use of the tomb at the end of the Achaemenid period (Danti 2014: 30). At both Gund-i Topzawa and Ghabrestan-i Topzawa, RAP is continuing excavation in order to understand more about the Iron III – Achaemenid Periods in the Sidekan Region.

2.5. Islamic Ashkawt-i Baaramos and Qalaat Lokhan are the two sites in the region that date exclusively to the Islamic period. Ashkawt-i Baaramos was collected in 2011 and the assemblage contained seven sherds, a bracelet, and a bead. The bracelet is of white stone and the bead is red and white glass. Based on the Islamic dates of the bracelet and bead, the site dates to the Middle Islamic Period (1000–1500 CE) (Spaer 1992). The ceramic collection is formed of small pieces of coarse ware, with no rims or bases, and includes two incised pieces. The two decorated pieces have parallel-incised lines, but are too small to be diagnostic.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A Preliminary Assessment of the Ceramic Sequence of Northeastern Iraqi Kurdistan

79

Qalaat Lokhan is located along the road between Soran and Rowanduz. The site was collected in 2012 and 2013. In 2012, the Directorate began to level the site for the construction of a new museum. They hit a stone structure and stopped work. In 2013, the Directorate asked RAP to conduct test excavations at the site. The collections from survey include multiple glazed pieces, as well as a pipe, that point to a date in the Ottoman Period (fig. 6.6–11) (1500–1900 CE). Operation 1, which was placed inside the stone building turned up mostly tablewares such as small cups, bowls, and plates. These ceramics were well made with little to no temper, and generally smoothed. Some of the pieces, mostly the bowls, were glazed in either green or blue. Operation 2, located to the north of the stone building, contained mostly cooking wares. These ceramics are coarser than those found in Operation 1, and have a grit temper. They are mostly brown, some with a black paint design on them (fig. 6.9). The assemblage consists of large pots and jars. Along with the ceramics excavated from Qalaat Lokhan, a coin dating to the Sorani Emirate (1790–1845 CE) appeared.

2.6. Multiple Gird-i Dasht, Gird-i Khiso, Paarcha Goza, and P. Jameel Daoud Ali, span multiple periods. The site of Gird-i Dasht is located on the Diyana Plain and was collected in 2012–2013. It is a truncated conical mound that reaches 20 meters above the surrounding plain. An earthen ramp – modified in modern times – slopes from the southeastern upper corner to the northeastern corner, marking the entrance to the upper fortified settlement of the Islamic era, and may correspond to the gateways of earlier periods. A lower mound stretches about 1.25 HA. Local informants link the final occupation of the mound to the Sorani Emirate (1790–1845 CE) and report that an agricultural suq was situated atop the mound. The Directorate collected 73 sherds in 2011 and 2012. The intensive survey performed by RAP in 2013, brought the total of surveyed ceramics to 425 (fig. 6.12–21). Ten sherds are a buff fabric with a green glaze, also known as Geruz Ware (fig. 6.15) (1256–1335 CE) (Danti 2004). Eight pieces are of a white fabric with little to no temper. Two of these form handles with incised lines, one forms a spout, and the others are body sherds (fig. 6.13, 17, 20). These have parallels with the Erbil Plain Archaeological Survey (EPAS) and date to the Sassanian Period (700 – 1000 CE) (Ur 2013: T16/2). Thirty-four pieces date to the Iron III/Late Assyrian Period (1000–700 BCE). Twelve of them are handles with incised designs, like those found at Bokadera (fig. 6.16). The other twenty-two pieces have wavy and straight lined combed impressions (fig. 6.12, 14), comparable with Khirbet Qasrij (Curtis 1989: pl. 42, No. 229). Five of these incised pieces are made of grey ware, a type found at Hasanlu and other sites in northwestern Iran (Danti 2013: 187–205). There are four pieces of Khabur Ware (fig. 6.21), a buff ware with brown or reddish brown paint in geometric designs, in the collection dating to the Middle Bronze Age (2000 – 1700 BCE) (Danti 2014: fig. 4.2; Ur 2013: T8/1). There are three pieces of chaff-faced buff ware with incised lines that date to the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

80

Kaercher

Early Bronze Age based on comparanda with EPAS (Ur 2013: T4/22). The ceramics show that Gird-i Dasht was inhabited from at least the Early Bronze Age to modern times (2500 BCE – Present). This is probably related to the site’s prominent position in the Diyana Plain, and its ability to control the routes from northwestern Iran to the Erbil Plain. Gird-i Khiso was collected in 2012 and dates from the Iron III to the Middle Islamic period (700 BCE – 1500 CE). The assemblage contains 43 sherds, of which, four are bases, seven are rims, and three are handles. The bases, two flat, one ring, and one rounded (fig. 6.30), all date to the Assyrian period. The two flat bases are made of a red ware, and have parallels in form with Qasrij Cliff (Curtis 1989: fig 7, no. 6). The rounded base is made of a buff ware with grit temper and has parallels with Khirbet Qasrij (Curtis 1989: fig 43, no. 322). The ring base is made of a buff ware with little to no temper and also has parallels with Khirbet Qasrij (Curtis 1989: fig 44, no. 346). The three rims from the Assyrian Period are similar to jars found at Khirbet Qasrij (Curtis 1989: fig. 43, no. 217–219), Hasanlu, Dinkha Tepe, and Godin Tepe. The two that date to the Sassanian Period are made of a green ware, and are similar in form to those found on the Erbil Plain Archaeological Survey (Ur 2013: T15/1). The three handles have incised holes and date to the Iron III period as based on comparanda with Bokadera, Hasanlu III (Danti 2013: 242) and Agrab Tepe (Muscarella 1973: fig. 13). Comb-incised lines and wavy rope decoration are indicative of Iron III wares as well. One sherd has green glaze on the inside, and is diagnostic of Geruz Ware (Danti 2004). Paarcha Goza was collected in 2011 and 2013 and dates from the Late Bronze Age to the Sassanian Period (1600 BCE – 700 CE). Eleven of the 42 sherds from Paarcha Goza were diagnostic, with button bases and carinated bowls (fig. 6.22–29). The bases date to the Late Bronze/Early Iron Age based on the Working Typology (Ur 2013: T10/8). There are two examples of carinated bowls which date to the Assyrian Period based on the Saddam Dam Salvage Project (Curtis 1989: pl. 24). One dish has both incised decoration as well as a raised lip located halfway around the plate (fig. 6.24). This dish dates to the Sassanian/Early Islamic period. Lastly, one sherd has black paint in a pattern that resembles the Ottoman ceramics from Qalaat Lokhan (1750 – 1900 CE). P. Jameel Daoud Ali was collected in 2013 and consists of eight sherds containing two complete profiles, one base, and one handle (fig. 6.31–33). Both profiles represent small bowls or dishes. One is made from a reddish brown ware and is carinated similar to bowls from the Erbil Plain Archaeological Survey (Ur 2013: T11/12). The second bowl has incised circular designs indicative of the Middle to Late Islamic Period (Ur 2013: T16/6). The base is made of a buff fabric with sand temper and dates to the Post-Assyrian period based on comparanda with Hasanlu (Danti 2013: fig. 4.3). The handle is not diagnostic of any period. This site dates from the Post-Assyrian to the Late Islamic Period (1000 BCE – 1900 CE).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A Preliminary Assessment of the Ceramic Sequence of Northeastern Iraqi Kurdistan

81

3. Conclusions The area in the Soran Directorate has been continuously inhabited from at least the 6th millennium BCE. It has been under the control of the Akkadian, Assyrian, Persian, and Ottoman empires as well as smaller regional kingdoms like Musasir and the Sorani Emirate. The distribution of sites over time in Figure 4 shows peaks in occupation during the Late Assyrian and Islamic periods. Interestingly we currently lack evidence of Chalcolithic occupation. Uruk beveled rim bowls, which are indicative of habitation during the 4th millennium, are absent in the area. The absence of this ceramic in the Soran Directorate and the appearance of it on the Erbil plain could point to differences in population or habitation during this period. RAP is continuing excavation in order to define habitation and create a more concrete typology to address broader themes like trade, migration, economics, and political control. The ceramics from the Directorate, however unexcavated, can give us a starting point in creating a chronology for this little studied region.

Bibliography Braidwood, R./Howe, B., 1960. Prehistoric Investigations in Iraqi Kurdistan. University of Chicago Press. Chicago. Curtis, J., 1989. Excavations at Qasrij Cliff and Khirbet Qasrij. British Museum Publications Ltd. London. Danti, M., 2004. The Ilkhanid Heartland: Hasanlu Tepe (Iran) Period I. Hasanlu Excavation Reports II. University Museum Monograph 120. University of Pennsylvania Museum. Philadelphia. –

2013. Hasanlu V: The Late Bronze and Iron I Periods. Hasanlu Excavation Reports III. University

–

2014a. The Rowanduz Archaeological Program 2013. First Report to the Kurdistan Regional

–

2014b. Searching for the Kingdom of Musasir. Expedition 56.3, 26–33.

of Pennsylvania Press. Philadelphia. Government. Unpublished Report. Dyson, R., 1965. Problems of Protohistoric Iran as Seen from Hasanlu. Journal of Near Eastern Studies 24, 193–217. Edmonds, C. J., 1966. Some Ancient Monuments on the Iraqi-Persian Boundary. Iraq 28.2, 159–163. Eppel, M., 2008. The Demise of the Kurdish Emirates: The Impact of Ottoman Reforms and International Relations on Kurdistan during the First Half of the Nineteenth Century. Middle Eastern Studies 44.2, 237–258. Gopnik, H., 2011. Godin Period II Pottery Typology and Catalogue. Unpublished Report. Ghirshman, R., 1954. Village Perse-Achéménide. Mémoires de la Mission Archéologique en Iran XXXVI. Presses Universitaires de France. Paris. Kroll, S., 1979. Die urartäische Keramik aus Bastam. W. Kleiss (ed.). Bastam I. Berlin, 203–220.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

82

Kaercher

Muscarella, M., 1968. Excavations at Dinkha Tepe, 1966. Bulletin of the Metropolitan Museum of Art 27, 187–196. –

1973. Excavations at Agrab Tepe, Iran. Metropolitan Museum Journal 8, 47–76.

O’Leary, B./McGarry, J./Salih, K., 2006. The Future of Kurdistan in Iraq. University of Pennsylvania Press. Philadelphia. Schimdt, E., 1957. Persepolis II. Oriental Institute Publication LXIX. The University of Chicago. Chicago. Simpson, S. J., 2008. Ancient Settlement in the Zammar Region: Excavations by the British Archaeological Expedition to Iraq in the Saddam Dam Salvage Project 1985–86. Vol 2. British Archaeological Reports Ltd. London. Solecki, R., 1971. Shanidar: The First Flower People. Alfred A. Knopf Inc. Ann Arbor. Spaer, M., 1992. The Islamic Bracelets of Palestine: Preliminary Findings. Journal of Glass Studies 34, 44–62. Thackston, W. M., 2006. Sorani Kurdish: A Reference Grammar With Selected Readings. Unpublished Reference. http://www.fas.harvard.edu/~iranian/Sorani/(retrieved May 15th, 2014). Ur, J., 2013. Working Ceramics Typology. 7th ed. Unpublished Report. Young, T. C., 1965. A Comparative Ceramic Chronology for Western Iran, 1500–500B.C. Iran 3, 53–85.

Kyra Kaercher, University of Pennsylvania Museum.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 83–94

John MacGinnis – Timothy Matney – Kemalettin Köroğlu

Excavations in the Lower Town of Ziyaret Tepe 2012 and 2013 with contributions by Mary Shepperson, Kristina Sauer and Daniele Arroyo-Barrantes This paper summarises the results of the 2012 and 2013 seasons of fieldwork at the site of Ziyaret Tepe in southeastern Turkey, the Neo-Assyrian provincial capital of Tušhan. The principal objectives were to investigate a number of features inside the southern and western city walls and earlier phases of the Operation G/R administrative complex. The site of Ziyaret Tepe lies on the upper Tigris in a strategic location which for two hundred years marked the northern border of the Assyrian empire (fig. 1). In earlier submissions we have given summaries of the work at Ziyaret Tepe from the inception of the project in 1997 up until early 2010 (Matney et al. 2012), and of the subsequent discoveries of the summer 2010 and 2011 seasons (MacGinnis et al. 2014). In this submission we now report on the final operations in the lower town, the 2012 and 2013 seasons (fig. 2).

Operation K In the 2012 season we returned to Operation K in order to more fully investigate the area of low status housing discovered in the 2003 and 2004 seasons (Matney/Rainville 2005: 31–35), uncovering a further section to the east. The plan of the Operation K complex now comprises eight rooms (fig. 3). As with the remains uncovered earlier, the newly excavated part of the building also had two occupational phases and it was confirmed that the ground plan in both phases was the same. In addition to Late Assyrian pottery, smallfinds included two fibula fragments, one stone vessel fragment, one stamp seal and two beads. It is notable that this material is of a generally lower quality and quantity than found at

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

84

MacGinnis – Matney – Köroğlu

Fig. 1. The Assyrian empire.

Fig. 2. Ziyaret Tepe excavation areas.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Excavations in the Lower Town of Ziyaret Tepe 2012 and 2013

85

Fig. 3. Plan of Operation K architecture (later phase).

other more substantial domestic residences across Ziyaret Tepe, suggesting that this was an area of lower status housing. Overall the work allowed for the clarification of some previously indistinct details and correction of various earlier assessments. For instance, it is now understood that the courses of mudbrick which we considered to be benches in Room A and Room C (cf. Matney/Rainville 2005: 33–35 and fig. 10) are the walls of the earlier phase. It is also understood that after the destruction of the earlier phase building, the rubble was cleared to foundation level and the earlier phase floors were raised by 15–20 cm. In the later phase, the building was shifted 80–90 cm southwards and was rebuilt using the same plan. There is no evidence of fire or destruction in the Late Assyrian buildings. There is continuity in use through both phases: Room C, excavated in 2013, appears to have been chosen for intramural burial during both phases, while Room A and Room E, excavated in 2003–2004,

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

86

MacGinnis – Matney – Köroğlu

had tannurs in both phases, probably indicating their use as kitchens. Consequently, it can be argued that there is not a long occupational gap between the two phases.

Operation M Operation M was an area first excavated in 2004 when a trench was laid out to groundtruth the presence of an apparent road visible in all the three forms of geophysical survey used at the site, magnetometry, resistivity and ground penetrating radar (Matney/Rainville 2005: 41–42). That excavation did confirm the existence of the street and also evidence for architecture built up alongside. The return to Operation M in 2012 was carried out in order to learn more about the nature of that occupation. In the event the work revealed part of a high status residence with well-built walls up to a metre thick. Cut into the floor of one room was the grave of a man in his early fifties, lying face down. His left hand was resting in dish which must have contained food – a practice already documented in the burials in Operation Q. The man was adorned with beads, earrings, a pendant, three fibulae, an iron dagger and a tiny bronze figurine of the Egyptian deity Bes. Most unusual though is that he was equipped with two cylinder seals (fig. 4), one of which must originally have been hanging from his belt and one of which he must have been holding in his left hand. The first seal, ZT 42149, which is made of white steatite and still has its bronze caps in place (this too we had in Operation Q), depicts a gazelle accompanied by two divine symbols, a star for Ishtar and the crescent of the moon god Sin. The second seal, ZT 37897, is made of faience and shows an archer shooting a snake.

Operation V The architectural remains in Operation V, situated slightly northwest of the Operation Q southern city gate (Matney et al. 2011), were first detected in the 2006 electrical resistivity survey (Matney et al. 2007: 48). In the imagery (fig. 5) a set of two not quite parallel rows of rooms could be identified, measuring approximately 30 by 8 m (external measurements) and separated by an open space, possibly a courtyard. The eastern row of rooms seemed to consist of a series of three rooms running NE-SW, with approximate internal width of 6 m and a total length of at least 30 m. Investigating these features had long been an objective of the project and in 2012 this was at last carried out. Excavation rapidly exposed a pavement of black and white pebbles very close to the surface. The pavement was recovered in its entirety with the exception of an area in the northeastern corner where proximity to the surface had led to its being destroyed by ploughing; there is also damage caused by a number of pits.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Excavations in the Lower Town of Ziyaret Tepe 2012 and 2013

87

Fig. 4. Cylinder seals (a) ZT 42149 and (b) ZT 37897.

With regard to the walls, the south wall was the best preserved, with mud brick lines and mortar joins which could be recognized, showing at least 3 ½ rows of mud bricks abutting the pebble pavement. In order to clarify the width of the eastern and western walls, three slot trenches were excavated whose sections showed that the western wall, which measured up to 2 m wide, was noticeably thicker than the other walls, which were only 1.5 to 1.6 m across. Once the wall lines were ascertained, the collapse outside the room at its eastern and southern façade was removed in anticipation of finding an outer surface. However, what was actually recovered was an older wall, representing Phase Ia, protruding 0.7 m from the eastern wall of the room and clearly different in colour – the mud bricks of Phase Ib are reddish and contain a high proportion of white chalk inclusions, whereas the older wall is made of yellowish bricks; unfortunately, their size could not be determined. The northern wall of the room, separating it from a smaller intermediate room (perhaps a corridor), was the worst preserved, which is not surprising considering the proximity of the deposit to the topsoil in this location; the area was also disturbed by a large pit. The pebble floor itself was built on a platform of mud bricks at least four courses high, on top of which was applied a layer of grayish mud up to 10 cm thick. Into this the pebbles were set vertically side by side, producing a compact, solid and resilient surface. This floor seems

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

88

MacGinnis – Matney – Köroğlu

Fig. 5. Resistivity image from the lower town showing the city wall with the Operation Q city gate at the bottom right and the Operation V architecture above.

to be purely utilitarian in use, in contrast to the checkerboard mosaic pavements of Operation G/R: those pavements were moreover located in courtyards and certainly had a prestige value, whereas the Operation V pavement was located inside a room in a non-residential building. A recess in the northwest part of the pavement probably indicates the presence of a doorway to the open space west of the room. The room measures more than 5 m in width and 13 m long, meaning that for roofing one or more central supports are likely to have been required: this must be the purpose of the two large limestone slabs set into the pavement whose function would otherwise remain obscure (they did not, for example, cover a drain as was initially hypothesised). The limited repertoire of small finds recovered impedes interpretation but a general storage function for the complex seems probable; perhaps, given that the complex appears to extend west beyond an open space and on to a second row of rooms, the possibility should be considered that it was a caravanserai.

Soundings to the East of the Main Operation East of the main excavation area in Operation V two small soundings were excavated in square N780 E890, the southwestern one covering an area of 3 by 3 m, the northwestern one 2 by 2m. Both revealed a cobbled surface, of which the northeastern one is situated 0.7 m

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Excavations in the Lower Town of Ziyaret Tepe 2012 and 2013

89

Fig. 6. Operation W Trench 1 showing the Courtyard 11 pavement with stone feature G-341 and then the lower pavement revealed by excavation of the feature.

deeper than the other. These cobbles, which must be the source of the anomalies visible in the resistivity map, presumably represent a metalled surface in the vicinity of the city gate.

Operation W Operation W is the designation given to renewed investigations into certain features in the administrative complex originally excavated between 2001 and 2010 as Operations G and R. In 2012 two trenches were excavated, comprising a re-investigation of Courtyard 11 (Trench 1) and a re-investigation of Room 10 (Trench 2). In 2013, work was resumed in Operation W with the aim of re-investigating Room 9 (Trench 3).

Trench 1 The first trench was sited in the area of Courtyard 11 (fig. 6). This is the southern of the two checkerboard pavements of Building 2: on the whole this had been neatly laid but there were four areas where the pavement had been cut into and then relaid with much larger and rougher blocks (G-341, G-342, G-343, R-298). Our working theory was that these were graves and in order to test this we decided to excavate one of the features, choosing G-341 for this purpose. The removal of the upper stones of the feature revealed a layer of hard brown clay 30 cm thick, initially devoid of inclusions but in its lower part containing pebbles and some potsherds. This came down onto an irregular pebble/cobble surface which in turn immediately overlay a much more substantial cobble surface which was well laid though not arranged in checkerboard squares in the manner of the upper pavement. This was clearly the pavement of a major earlier phase and not wishing to cut into it we concluded the investigation at this stage. The operation demonstrated that these features were not graves. What they are remains unclear. Perhaps, surprisingly, they are simply repairs, if not overly elegant

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

90

MacGinnis – Matney – Köroğlu

ones. Nevertheless this investigation produced an important result, the demonstration that a substantial earlier phase underlay the Operation G/R complex. This was information which we were able to set to good use straight away in the next trench.

Trench 2 The second trench excavated in Operation W was just to the east, sited in order to come down onto Room 10. This is the larger of the two rooms where tablets were found in the 2002 and 2003 seasons (Parpola 2008). Based on the discovery from the first trench that there was an earlier phase to the building, the aim of Trench 2 was to investigate whether evidence for administrative use could be recovered from this earlier phase. After removing the backfill from the earlier work, we excavated the remaining parts of the floor levels as they had been left in 2002 (north end of the trench) and 2003 (south end of the trench). The floor matrix consisted of red clay with patches of pebbles and cobbles set in, as well as fragments of bitumen and some patches of ash. In this matrix we found a cuneiform tablet, ZT 44030 (fig. 7) and a sealing, ZT 44031 (fig. 8). Removal of these surfaces brought us down onto a subfloor constructed of pink clay, which had a door socket in situ on the western side. This was a very welcome discovery as the location of the door into Room 10 is something which we had not been able to establish in earlier seasons. Removing this packing exposed another surface, which consisted of a layer of broken mud brick overlying a floor, into which a pit for a pithos had been cut, the rim lined by a square of baked half-bricks; the capacity of the pithos was 180 l. All these layers were associated with the later level of the complex. Excavating down further we reached the wall and upper floor level of the expected lower building phase. Below that floor level was a level of clay packing which in turn overlay one more lower floor level. This lowest level, which is on the same elevation as the lower cobble surface found in Trench 1, had a small door socket in situ on the western side and was associated with a large number of pithos fragments, further evidence of the protracted use of this space in an administrative storage capacity.

Trench 3 In 2013 work was resumed in Operation W, this time focusing on Room 9, the smaller archive room of Building 2, with the aim of recovering a more complete picture of the architectural history of the complex. The removal of the backfill from the 2002 excavation produced a spherical black bead (ZT 44090) while the cleaning of the floor as left in 2002 yielded four baked clay tokens (ZT 44095). Below this was found a sub-floor packing consisting of stones and large sherds overlying a band of ashy dirt mixed with clay and containing bones, sherds and brickbats, as well as a baked clay disc (ZT 44109). This band overlay the solid foundation of a floor constructed of mud bricks of tough red clay. Removing this we came down onto the foundations of the walls of the upper phase which consisted

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Excavations in the Lower Town of Ziyaret Tepe 2012 and 2013

Fig. 7. Cuneiform tablet ZT 44030.

91

Fig. 8. Clay sealing ZT 44031.

of three courses of grey brick. These foundations were laid, somewhat irregularly, directly on top of the levelled walls of the underlying phase, indicating that the upper architecture was a rebuilding along the same lines as the lower and with no significant interval of time; in other words, the upper and lower phases together form one level. Salient to this point is that the pithos G-723 was clearly installed as part of the earlier phase, but kept in use on into the later phase as well. The walls of the lower phase defined a space measuring 1.75 x 3.35 m, smaller than the 2.95 x 3.60 m of Room 9 above. A mudbrick platform measuring 80 cm wide and at least 1.50 long was built against the northern wall. A surface running underneath this platform indicated that it was a secondary installation. Below this was a levelling fill of solidly packed broken brick, from which a clay sealing (ZT 44116) and numerous flecks of a red pigment (ZT 44133) were recovered. Underneath all of this was the well preserved wall of a yet earlier phase running north-south across the trench. The material on both the east and the west side of the wall consisted of a very solid mudbrick packing with a trampled surface. No other associated surface was found, and it appears that this wall was preserved as foundations only.

Cylinder Seal A cylinder seal (ZT 44168) was found on the surface east of Operation W.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

92

MacGinnis – Matney – Köroğlu

Tokens Attention is drawn to the publication of the baked clay tokens from Ziyaret Tepe (MacGinnis et al. 2014). Over 460 tokens were found, the great majority from secure Neo-Assyrian contexts and the majority of these from one room in the Operation G/R administrative building. The use of tokens in administration in the 1st millennium BC has not previously been noted, but the finds at Ziyaret Tepe, together with data from other sites such as Tell Tayinat, Tell Sheikh Hamad and Assur, suggests that their use at this time may have been widespread both in the Assyrian empire and perhaps indeed the near east more generally.

Operation Y Operation Y lies at the southwest edge of the lower town, to the west of the gateway in the city wall excavated at Operation Q. The main feature of the architecture in Operation Y is the Neo-Assyrian period lower town city wall. This was exposed at the point where it made a right-angled turn outwards. The wall is very substantial at this point, at least 4.70 m wide. The manner in which the wall makes its ninety-degree turn is unusual. Rather than making a square corner by laying the (35 x 35cm) bricks in straight rows, the corner is constructed as a smooth curve with concentric rows of bricks fanning out to make the turn. An extremely hard-packed clayey deposit running along on the inside of the wall in a strip 2.40 m wide would appear to be a roadway allowing internal communication. A curved bend rather than a sharp corner here would have greatly eased the passage of people, animals and wheeled vehicles. It is interesting to note that a similar gap between the city wall and the close-by architecture was present in the area excavated in Operation K (Matney/Rainville 2005: 33 and fig. 10) and that a free space on the inner side of the defences is likewise visible in the geophysical imagery at Zincirli (Casana/Herrmann 2010: 61, 64–65). The only other element of Neo-Assyrian architecture preserved in Operation Y is the mudbrick wall of a building which runs parallel to the northwest internal face of the city wall in the western part of the trench. The bricks at the western end of this wall are very well preserved and show the wall to be made of two rows of large mudbricks, 40 x 40 cm square, laid in a fine plaster mortar, making a wall 85 cm wide. It seems that only one course of this wall survives, suggesting that the building was deliberately levelled after it fell out of use. The foundation trench for the wall was cut through a deposit of tough reddish clay with frequent white chalky inclusions. This deposit, which is found on both sides of the wall, seems to be a re-deposited natural clay used to prepare a level area for the construction of the building. This levelling layer was probably necessary because the deposit below, which runs under the wall, was a soft ashy fill mixed with mudbrick lumps, pottery and animal bones and generally unsuitable for the foundation of a building. It seems that after the completion

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Excavations in the Lower Town of Ziyaret Tepe 2012 and 2013

93

of the city wall and the roadway the area immediately inside the wall consisted of mixed building rubble and ashy rubbish dumps. This area was then levelled and the ashy deposit capped with the layer of reddish clay before the foundations building were cut into it. At some later stage this building seems to have been destroyed and levelled, leaving only these traces of its foundation. There is no trace of later building activity, so the area was probably left as open space inside the city wall.

Operation Z Operation Z, the last of the archaeological investigations to be undertaken at Ziyaret Tepe, was the designation for a 5 x 5 m trench located in grid square N810 E840, a short distance northeast of Operation Y. The aim of the operation, groundtruthing the existence of a building appearing in the 2006 resistivity plot of the area (Matney et al. 2007: 48, 73 fig. 22), was successfully achieved with the exposure of a wall 1.80 m wide running northeast-southwest across the trench exactly as predicted. A curious facing of stones and potsherds set on edge must stem from a reuse of the wall in the late Assyrian or post-Assyrian period; it had an associated floor with a thin red plaster surface laid on a sub-floor packing of red clay on the southeastern side. The wall itself was made of bricks measuring 38–40 cm square and was very well preserved. There was a red plastered surface on its southeastern side and to the northwest a rough stone pavement containing fragments of plaster, pottery and animal bones. Underneath this all was a zone of alternating bands of broken red brick and layers of ash. A sounding below exposed part of a stone wall foundation and the fragmentary remains of a white plastered floor. With these operations the fieldwork component of the Ziyaret Tepe Archaeological Project, excavation and geophysical survey, is now concluded. It has been a long and productive engagement. In 2014 we conducted an on-site study season in order to complete the recording of ceramics and small finds, conservation of materials in the Diyarbakir Museum (in particular the painted wall plaster from the palace) and analysis of archaeobotanic and archaeozoological datasets (Rosenzweig i. p.; Greenfield 2015; Greenfield/Rosenzweig 2016). The focus of the project now moves to the preparation and publication of final reports as the lasting records of this great project and the work of all the many team members who made it possible.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

94

MacGinnis – Matney – Köroğlu

Bibliography Casana, J. /Herrmann, J., 2010. Settlement history and urban planning at Zincirli Höyük, southern Turkey. Journal of Mediterranean Archaeology 23, 55–80. Greenfield, T. L., 2015. The Palace versus the Home Social Status and Zooarchaeology at Tušḫan (Ziyaret Tepe), a Neo-Assyrian Administrative Provincial Capital in Southeastern Turkey. Journal of Eastern Mediterranean Archaeology and Heritage Studies 3.1, 1–26. Greenfield, T. L./Rosenzweig, M. S., 2016. Assyrian Provincial Life: A Comparison of Botanical and Faunal Remains from Tušhan (Ziyaret Tepe), Southeastern Turkey. R. A. Stucky et al., Proceedings of the 9th International Congress on the Archaeology of the Ancient Near East (9. ICAANE), Vol. 2. Wiesbaden 2016, 305–321. MacGinnis, J. D. A./Matney, T., 2009. Ziyaret Tepe: Digging the Frontier of the Assyrian Empire. Current World Archaeology 37, 30–40. –

2010. Archaeology at the Frontiers: Excavating a Provincial Capital of the Assyrian Empire. Journal of Assyrian Academic Studies 23, 1–19.

MacGinnis, J./Matney, T./Köroğlu, K./Sauer, K./Dosch, J., 2014. Operations in the lower town of Ziyaret Tepe, 2010 and 2011. P. Bieliński et al. (eds.) Proceedings of the 8th International Congress on the Archaeology of the Ancient Near East 30 April–4 May 2012, University of Warsaw. Vol. 2: Excavation and Progress Reports, Posters. Wiesbaden, 34–41. MacGinnis, J./Monroe, D. A./Wicke, W./Matney, T., 2014. Artifacts of Cognition: the use of clay tokens in a Neo-Assyrian provincial administration. Cambridge Archaeological Journal 24, 289–306. Matney, T./Rainville, L., 2005. Archaeological Investigations at Ziyaret Tepe, 2003–4. Anatolica 31, 19–68. Matney, T. et al., 2007. Report on Excavations at Ziyaret Tepe, 2006 Season. Anatolica 33, 23–74. –

2011 Excavations at Ziyaret Tepe, Diyarbakir Province, Turkey, 2009–2010 Seasons. Anatolica 37, 67–114.

Matney, T./Greenfield, T./Köroğlu, K./MacGinnis, J./Proctor, L./Rosenzweig, M./Wicke, D., 2015. Excavations at Ziyaret Tepe, Diyarbakır Province, Turkey, 2011–2014 Seasons. Anatolica 41. Matney, T./Wicke, D./MacGinnis, J./Köroğlu, K., 2012. Uncovering a Provincial Capital of the Assyrian Empire: The Ziyaret Tepe Archaeological Expedition 1997–2100. R. J. Matthews/J. Curtis (eds.), Proceedings of the 7th International Congress on the Archaeology of the Ancient Near East, 12 April –16 April 2010, the British Museum and UCL, London. Volume 3: Fieldwork & Recent Research and Posters. Wiesbaden, 313–324. Parpola, S., 2008. Cuneiform Texts from Ziyaret Tepe (Tushhan), 2002–2003. State Archives of Assyria Bulletin 17, 1–137. Rosenzweig, M., i. p. An environmentality of Neo-Assyrian Agriculture. Journal of Social Anthropology. John MacGinnis, Cambridge University. Timothy Matney, University of Akron. Kemalettin Köroğlu, Marmara University.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9th ICAANE, Basel 2014, Vol. 3, 95–105

Karel Nováček – Miroslav Melčák

The Medieval Urban Landscape in Northeastern Mesopotamia (MULINEM): First Two Years of the Project More than fifteen sites of either confirmed or conjectured urban status existed in the Central Tigris and Adiabene regions between the 6th and 17th century AD. The presented project investigates their diversity, temporal dynamics and mutual relations. The excavations of Samarra undertaken by E. Herzfeld and F. Sarre one hundred years ago represent a formative step in the archaeology of Islamic cities. Nevertheless, since the end of these excavations, little progress has been made in our knowledge of medieval urbanism in the hinterland of Samarra, i.e. in the Central Tigris region and Northeastern Mesopotamia. Besides general studies on the region (Hoffmann 1880; Le Strange 1905; Fiey 1965; Morony 2005; Wheatley 2001) and historical-topographical analyses (e.g., Heidemann 1996), the published archaeological contributions reflect only a cursory or geographically narrow interest of individual scholarship (Sarre/Herzfeld 1911: 210–212; Sarre/Herzfeld 1920: 322–329; Edmonds 1932; Venco Ricciardi 1971; Córdoba 2005). On one hand, the reasons for this intellectual gap can be seen in the long-lasting cultural and scientific isolation of these regions (now predominantly Kurdish), and, on the other hand, in the shifting priorities within the discipline itself, where a regionally-defined interest in urban networks has found a prominent place only in the most recent years. The Central Tigris area with the adjacent region of the «Land of Hazza» centered around Arbīl which we have chosen in our investigation of the structure and dynamics of medieval Islamic centers, 1 has great potential for several reasons. First, during the Abbasid Caliphate, the region formed part of the communication corridor of key importance between the centre of the empire and its northern periphery, as well as a part of the wider hinterland of the Abbasid capitals of Baghdad and 1

Our three-years project (2013–2015) supported by the Czech Science Foundation (No. 13-19266S), represents a continuation of the previous research dealing with the archaeological topography and long-term urban evolution of the Assyrian Arbail – medieval Irbil (Nováček et al. 2008; Nováček et al. 2013).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

96

Nováček – Melčák

Fig. 1. Map of the region, urban sites under study and reconstructed courses of medieval routes (drawing by L. Starková and K. Nováček).

Samarra. The density and diversity of the central places in this region reflect its importance. According to our preliminary survey, combining analysis of historical topographies with satellite imagery, more than fifteen sites of either confirmed or conjectured urban status appeared in the region delimited by the Tigris, Great Zāb, and Little Zāb rivers, and the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The Medieval Urban Landscape in Northeastern Mesopotamia (MULINEM)

97

Zagros Mountains in the period from the 6th to the 17th century AD (fig. 1). Apart from Arbīl and Altun Köprü, all sites were abandoned during the Ottoman era at the latest. Furthermore, the state of preservation of the sites and their visibility both in the satellite imagery and in the terrain is very good, particularly in the Central Tigris area and Makhmur Plain, where the degree of the modern agricultural impact is still relatively small. Thus the sites are very suitable for investigation using methods of remote sensing and non-invasive survey. In 2013–2014, the western and southern parts of the study area became a war zone and the sites in this region (9 of the total of 16 sites) ceased to be accessible. These circumstances forced us to shift emphasis in the research on these sites to the analysis of written sources as well as satellite and aerial images. In these cases, our findings remain unverified hypotheses. Yet, several towns known from medieval Arabic geographies and biographical dictionaries have been located with high probability. This is the case of Hadīthat al-Mawsil and al-Bawāzij, two towns that played an important role as regional centres on the Central Tigris between the 8th and 12th (13th?) centuries AD. Hadīthat al-Mawsil was re-founded by the last Umayyad caliph Marwān II before the mid-8th century in place of a previous Christian settlement. Some indications in our sources lead us to the tentative conclusion that the city lost its urban character just before the Mongol invasion. The satellite imagery provides a view of a huge, c. 300ha area of low mounds and architectural remains adjacent to the eastern terrace edge of the Tigris river, between the present-day villages of Sultān Abdullāh and Tell al-Ša’īr. The deserted, regularly planned city shows traces of monumental buildings and courses of the town wall. One of the outstanding structures could hypothetically be connected with a deserted congregational mosque, which was described by al-Maqdisī as the only non-mud-brick structure of the city. In 1923, C. J. Edmonds described a group of Middle Islamic architectural monuments in the lower Little Zāb basin, at the place locally called al-Ismā´īlīya, and connected it with the town of al-Bawāzij (Edmonds 1932). In 2013, three monumental mausoleums in two different locations (two of them locally called Imām Ismā´īl and Šaykh Hamad respectively, the third one remains unnamed) were still standing, although in a ruined state. The structures are characterized by lavish interior decoration executed either in stucco or via geometrically-composed fired bricks (hazarbaf ) (fig. 2). The mausoleums were apparently built on the outskirts of medieval settlements but an urban context cannot be assumed here so far. Conspicuous remains of an urban site were identified only c. 7,5 km NE of the mausoleums (fig. 3). In the satellite images, one can recognize a large 138-ha enclosure of trapezoidal plan which was fortified by a rampart, originally a mud-brick wall with bastions. There are also traces of open canal systems that once entered the enclosure. On the west side, a narrow projection extends out of the fortified area, ending with a large rectangular mound, probably a citadel. The position of the abandoned town corresponds well with some descriptions of al-Bawāzij in the works of medieval Arabic geographers, but its dating remains uncertain.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

98

Nováček – Melčák

Fig. 2. Imām Ismā´īl, interior of the ruined mausoleum, state in 2013 (photo by O. Khorshid).

As for our fieldwork, in 2013 and 2014, we concentrated on three central sites in the Adiabene region (Kafr Azza, Old Makhmūr, Altun Köprü), each very different from the other in terms of size, past status and state of preservation. The town of Kafr Azza (probably alternatively referred to, in the medieval period, as Hazza) was founded in the Late Sasanian or Early Islamic period as an alternative centre of the Arbīl province, as is indicated by the name «Land of Hazza», which occurs several times in the sources. The archaeological remains of the city have been identified with a high probability only 10 km SW of Arbīl, between the villages of Azza and Qunyān (Sardar). The settlement area of irregular plan followed both a longitudinal West–East axis and an important canal running NE-SW, whose most distant traces are identifiable in the satellite images in the suburb of Arbīl, in the vicinity of the Kilik Mišik Tell. This canal could be the same as a canal called al-Jathulīk mentioned as still in use in a Chaldean source dated to 1491 AD. According to the image analysis and survey, the total area of the site amounts

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The Medieval Urban Landscape in Northeastern Mesopotamia (MULINEM)

99

Fig. 3. Topographical situation in surroundings of Tell Ali with a deserted city and two groups of mausoleums; traces of deserted canals by dashed lines, deserted villages by hatched areas (image of CORONA mission DS1104-2138DF011, 16 August 1968, drawing by K. Nováček).

to c. 80 ha.2 The settlement area comprises one central tell and several tens of large, mostly low, ploughed-up mounds. We succeeded in corroborating that, in most cases, the mounds correspond to remains of medieval structures or built-up areas masoned of fired bricks. The settlement witnessed a maximum extent in its initial phase, i.e. in the Late Sasanian and Early Islamic Periods (roughly to the 10th century AD) when its area reached c. 75 ha. In the following period (c. the 11th–14th century AD), the site was substantially reduced; according to pottery analysis, the zone of intensive settlement use stayed compact but it shrank to

2

The site reconnaissance was carried out in autumn 2013, courtesy of Directorate of Antiquities in Arbīl. The detailed survey was conducted one year later (October 2014) as a joint activity of MULINEM and Erbil Plain Archaeological Survey (EPAS), dir. by Jason Ur, Harvard University. We thank Jason Ur for his kind cooperation in the preparation of the survey.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

100

Nováček – Melčák

Fig. 4. Settlement area in the Azza–Qunyan zone (EPAS Site No. 215) with survey polygons (black lines), brick fragments scatters (white hatched areas), deserted canal (white dotted line) and zone of most intensive and long-term settlement (crosshatched area). Satellite image Formosat-2, 23 December 2013 (drawing by L. Starková and K. Nováček).

c. 31 ha, abandoning the central tell – a former nucleus of the city (fig. 4). This shrinkage of extent finds parallels in the textual sources: while Ibn Hawqal described Kafr Azza in the late 10th century as a town (madīna), Yāqūt al-Hamawī and Ibn al-Mustawf ī, some 250 years later, mention only a village (qarya). Detailed topography of the city is, nevertheless, a rather complex issue that needs a careful comparison of archaeological records with available textual evidence, which is beyond the scope of this preliminary information. The survey of the second centre – Old Makhmūr (Makhmūr al-Qadīma or Eski Maxmūr) – opens somewhat different questions. The site was identified in the CORONA satellite imagery in 2010 (Mühl 2013: 212) and 2012 (Nováček 2012: 81–83) and has become a subject of detailed surveys and mapping since October 2013. The site located on the SE outskirts of the modern town of Makhmūr is locally known as Old Makhmūr, but no mentions in the sources have been reliably connected with the site so far (Fiey 1965: 127). About 120 ha large, the most likely unfortified site stretches along a barely cultivated landscape, and hence its remains are extremely well preserved in relief. The settlement area comprises

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The Medieval Urban Landscape in Northeastern Mesopotamia (MULINEM)

101

Fig. 5. Eski Makhmūr, situation of the site with the most important features 1. a water reservoir? 2. ‹urban quarters›. 3. a vaulted structure, presumably a church. 4. Early Islamic residences (qusūr) (satellite image WordView 2, 2013, drawing by L. Starková and K. Nováček).

several ruined structures and low mounds (fig. 5). The central part of the site is shaped into a strip with the main axis in a NW–SE direction and a high concentration of architectural features arranged in a seemingly regular pattern resembling an urban neighborhood. All wall courses, visible on the surface, maintain the general orientation of the building area (NW– SE or NE–SW). The western limit of this ‹quarter› is occupied by a remarkable rectangular mound concealing a large structure with maximal dimensions of 57 x 34 m whose longitudinal axis is oriented NE–SW. The northeastern part of the building–originally domed, now collapsed–is divided into four square compartments arranged in a T-shape. This form of subdivision resembles the plan of a church sanctuary and the structure should be tentatively interpreted as an Early Medieval church. The domed rooms could have served as souterrain tomb chambers (a crypt or martyrion) which would mean that the ground floor level of the eastern part of the church used to be above the current level of the surface. The nucleus of the settlement is flanked by large, isolated structures. At least three mounds of a rectangular plan probably contain ruins of fortified building complexes. One of these mounds in the SW part of the site bears visible portions of a rectangular perimeter wall

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

102

Nováček – Melčák

Fig. 6 Eski Makhmūr, an Early Islamic castle (qasr), plan of surface remains of the structure in the modern satellite image (drawing by L. Starková and K. Nováček).

Fig. 7 Eski Makhmūr, an Early Islamic storage jar with honeycomb decoration, coming from illicit excavations (photo and drawing by K. Nováček).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The Medieval Urban Landscape in Northeastern Mesopotamia (MULINEM)

103

(53 x 42 m) on its summit, built with limestone boulders bound by gypsum mortar, as well as sparse traces of inner constructions joining the wall (fig. 6). These fortified residences are typologically very close to qusūr, Umayyad castles in Syropalestine, or Early Abbasid residences at al-Hīra. The isolated features surrounding the compact site nucleus are added by a 1,5-ha square enclosure (c. 120 x 120m) in the north part of the area, created by a 3–5 m high rampart. We consider it to be a large water reservoir. The surface pottery from the area provides uniform evidence of the Late Sasanian to Early Islamic Periods (fig. 7). The ceramics of Uruk, Neo-Assyrian, Hellenistic and Parthian Periods have only been collected at the western and southern edges of the site, where the anthropogenic relief had already been flattened, which suggests a more complex settlement history. Old Makhmūr represents an extraordinarily well preserved Early Islamic site of apparently central function with hints at the presence of an ‹Islamic› elite. Their residences enclosed an ‹urban nucleus› settled most probably by a Christian population. The city was situated in a strategic position at the edge of the fertile Makhmūr Plain, close to Hassan Ghazi Pass, where an important route leading from the Central Tigris area to Arbīl crossed over the mountain ridge. It would be worthwhile to cite al-Hīra, once the capital of the Lakhmids in central Iraq, as a close parallel to the architectural findings at Old Makhmūr (Talbot Rice 1934; Toral-Niehoff 2014: 75–82). Diverse ties also connect the morphology of the site with the Levant of the Umayyad period. The third example introduces a case of a small town in a strategic position which survives to the present. Altun Köprü (Pirdi) was situated on the historical route leading from Baghdad to Mosul via Arbīl. Occupying a 13-ha island in the Little Zāb river and being connected to the banks by two bridges, the town offered the only perennially reliable crossing of the river in the whole region. Our investigation benefited from a continual line of accounts recorded by European travellers visiting the site between the 16th and early 20th century. This material, being confronted with a unique view of the town in an Ottoman miniature from 1534 (Gabriel 1928; Eroğlu et al. 2008: 190) enabled us to define the main trajectories in the urban and architectural development of the site, which were afterwards verified by a survey conducted in the complex environment of the living city. We can preliminarily conclude that the town gained urban status in the early 16th century AD at the latest, but probably much earlier; we are working with the hypothesis of an urban extension from the island out to the western bank of the river in the Middle Islamic period. During the late Ottoman period, however, the town underwent a deep transformation, which resulted in a total change of the urban plan as well as in a structural pauperization accompanied by the loss of all former architectural landmarks (town wall with gates, congregational mosque, governor’s palace). We have been also able to fix the chronology of the famous bridges which gave the town its name.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

104

Nováček – Melčák

To conclude, we have to emphasize a high degree of formal variability of central places in the region under study. The cities differed in terms of size, formal elements of groundplan, presence of perimeter fortifications, inner articulation of space (namely, presence of a citadel), water management, number and size of large structures etc. The diverse forms of these cities suggest a hierarchical pattern within the urban network and possibly also a higher level of cooperation within the settlement structure. Moreover, the constellation of central places in Adiabene was by no means static, although the long-term continuity and high degree of resilience against political changes seems to be the most distinctive feature of the urban landscape in NE Mesopotamia. On the other hand, several towns might have been abandoned or lost urban status within periods that are generally considered as times of stability and prosperity (Mahoze/Tell Mahuz in the 9th century, al-Sinn and Hadīthat al-Mawsil in the 12th century). As the project continues, our aim is to interpret these shifts in meaning of individual nodal points, as well as to search for an explanation of the final collapse of the whole urban network in the Ottoman period.

Bibliography Córdoba, J. M., 2005. Excavando en Tell Maḥūz. Notas sobre la supuesta muralla sasánida de la ciudadela. ISIMU – Revista sobre Oriente Próximo y Egipto en la antigüedad 8, 217–234. Edmonds, C. J., 1932. An Abbasid Site on the Little Zab. The Geographical Journal 80, 332–333. Eroğlu, C./Babuçoğlu, M./Özdil, O., 2008. Osmanlı vilayet salnamelerinde Musul. Ankara. Fiey, J. M., 1965. Assyrie chrétienne. Contribution à l’étude de l’histoire et de la géographie ecclésiastiques et monastiques du nord de l’Iraq. Beyrouth. Gabriel, A., 1928. Les étapes d’une campagne dans les deux ‘Irak d’après un manuscrit turc du XVIe siècle. Syria 9, 328–349. Heidemann, S., 1996. Al-cAqr, Das islamische Assur, Ein Beitrag zur historischen Topographie in Nordmesopotamien. K. Bartl/S. R. Hauser (eds.), Continuity and Change in Northern Mesopotamia from the Hellenistic to the Early Islamic Period. Berlin, 259–285. Hoffmann, G., 1880. Auszüge aus syrischen Akten persischer Märtyrer. Leipzig. Le Strange, G., 1905. The Lands of the Eastern Caliphate. Mesopotamia, Persia and Central Asia from the Moslem conquest to the time of Timur. New York. Morony, M. G., 2005. Iraq After The Muslim Conquest. Piscataway. Mühl, S., 2013. Siedlungsgeschichte im mittleren Osttigrisgebiet: Vom Neolithikum bis in die neuassyrische Zeit. Wiesbaden. Nováček, K., 2012. Ard Hazza, ard Arbíl: Projekt výzkumu středověkých měst v oblasti mezi Velkým a Malým Zábem v severovýchodní Mezopotámii/Ard Hazza, ard Arbíl: A project of the research of medieval urban landscape between Great and Little Zab in northeastern Mesopotamia. L. Pecha

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The Medieval Urban Landscape in Northeastern Mesopotamia (MULINEM)

105

(ed.), Materiální a duchovní kultura Východu v proměnách času. Plzeň, 77–87. Nováček, K./Chabr, T./Filipský, D./Janíček, L./Pavelka, K./Šída, P./Trefný, M./Vařeka, P., 2008. Research of the Arbil Citadel, Iraqi Kurdistan, First Season. Památky archeologické 99, 259–302. Nováček, K./Amin, N. A. M./Melčák, M., 2013. A medieval city within Assyrian walls: the continuity of the town of Arbīl in Northern Mesopotamia. Iraq 75, 1–42. Sarre, F./Herzfeld, E., 1911. Archäologische Reise im Euphrat- und Tigris-Gebiet 1. Berlin. –

1920. Archäologische Reise im Euphrat- und Tigris-Gebiet 2. Berlin.

Talbot Rice, D., 1934. The Oxford Excavations at Hīra, Ars Islamica 1, 51–73. Toral-Niehoff, I., 2014. Al-Ḥīra: eine arabische Kulturmetropole im spätantiken Kontext. Leiden. Venco Ricciardi, R., 1971. Sasanian Pottery from Tell Mahuz (North Mesopotamia). Mesopotamia 5-6/1970, 427–482. Wheatley, P., 2001. The Places Where the Men Pray Together: Cities in Islamic Lands, Seventh through the Tenth Centuries. Chicago.

Karel Nováček, University of West Bohemia, Plzeň, Czech Republic. Miroslav Melčák, Oriental Institute, Czech Academy of Sciences, Prague, Czech Republic.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9th ICAANE, Basel 2014, Vol. 3, 107–117

Maria Grazia Masetti-Rouault – Olivier Rouault

French Excavations in Qasr Shemamok, Iraqi Kurdistan (2013 and 2014 campaigns): the Assyrian Town and Beyond Identified as the Neo-Assyrian town of Kakzu/Kilizu since Layard’s explorations, the site of Qasr Shemamok, in Iraqi Kurdistan, has been excavated since 2011 by a French mission. This paper sums up the preliminary results of the researches carried out on the tell and in the Lower town in 2013 and 2014.

Introduction This paper concerns the last two seasons of excavations at the site of Qasr Shemamok, located in Iraqi Kurdistan around 25 km South-West of Erbil, carried out by the French mission who has been working there with an international team since 20111 (fig. 1). We are conducting our researches with the official agreement of the Iraqi State Board of Antiquities and Heritage in Bagdad, as well as of the Kurdish General Direction of Antiquities and his director, A. O. Zainadin (Mala Awat). With the help and the friendly support of the Direction of Antiquities of the Erbil region, of its director, M. Nader Bubakr Mohamed, and of its representatives, we have been able to carry out the excavations for two seasons of 1

Dr. Narmin Ali Mohamed Amen (Salaheddin University, Erbil & UMR 8167, Orient et Méditerranée, Paris), Dr. J. MacGinnis (Macdonald Institute, Cambridge University), Dr. O. Mahmoud (Soran University, Kurdistan); Dr. M. G. Micale (Sapienza University, Rome & Freie Universität, Berlin), archaeology; Prof. A. Tabbagh and his team (Paris VIII – Pierre and Marie Curie University, UMR 7619), geomagnetic surveys; Prof. M. Chevreuil and his team (Ecole Pratique des Hautes Etudes, Paris, UMR 7619), environmental studies; Prof. P. Azara and his team (Polytechnic Institute, Barcelona), architecture and 3D reconstructions; Prof. J. Ur and his team (Harvard University), archaeological regional survey; Prof. J. Tomczyk and his team (Cardinal Wyszynski University, Warsaw) archaeo-biological studies; Prof. V. Picotti and his team (Bologna University) geological and geomorphological survey.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

108

Masetti-Rouault – Rouault

Fig. 1. Map of Northern Mesopotamia.

eight weeks in spring 2013 and 2014. In one respect the situation at Qasr Shemamok differs from that of many of the other new projects now starting up in Kurdistan. The site has been identified since the 19th century, at the beginning of the archaeological exploration of Mesopotamia. Thanks to an inscription of Sennacherib found on thousands of baked bricks dispersed across the site (Layard 1853: 223–224), Austen Henry Layard himself recognized, in the large, lofty tell of Qasr Shemamok – with its high Citadel and an extensive Lower town – the remains of a neo-Assyrian city founded by this king. This short inscription records the construction of a double system of urban walls around the city as well as its ancient name, first read as Kakzu and now, more correctly, as Kilizu (Postgate 1980: 591). However, after some tunneling around various parts of the tell, Layard declared the site not interesting enough to merit regular excavations, an opinion shared also by the French Victor Place, who visited it at almost the same time (Place 1852: 463–464). When we first started thinking about setting up a project of excavations at Qasr Shemamok we took serious consideration of these first reports and we were also aware of the sad conclusion of the Italian mission directed by G. Furlani, who in 1933 failed to find in this site the Assyrian levels he expected and had to content himself with the publication of materials from a Parthian cemetery found in the western part of the Lower town, never to return to Qasr Shemamok again.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

French Excavations in Qasr Shemamok, Iraqi Kurdistan

109

Fig. 2. Qasr Shemamok: topographical plan upon a Google image (oriented north, each square 10 m).

Notwithstanding these alarming precedents, the urban site of Qasr Shemamok, which covers an area of 50 to 70 hectares on the left bank of the Shiwazor river, midway between Arbela/Erbil and Calah/Nimrud, seemed us to offer a very good opportunity – maybe even an unique one – to study an Assyrian town belonging to the Māt Aššur homeland from Middle Assyrian period, in a site which had never before been the subject of regular investigations. After our researches in two Neo-Assyrian colonies, respectively in the Syrian Lower Middle Euphrates and in the Lower Habur valleys, at Tell Masaikh and Bir el-Haddad (Masetti-Rouault 2013; Rouault/Masetti-Rouault 2014), we wanted to have a fresh, critical look over the Assyrian archaeological record in order to better understand the forms, conditions and also the chronology of the imperial expansion, as well of its crisis. Hoping that Qasr Shemamok’s resistance to previous archaeological studies had been, for the first explorers, mainly a matter of bad luck, maybe a consequence of the still limited knowledge of Iron age ceramics and mud brick structures, we resolved in 2010 to start a project there (fig. 2). To justify our confidence we should add that we knew much better than our predecessors that the toponym Kilizu is often attested in Middle and Neo-Assyrian administrative archives

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

110

Masetti-Rouault – Rouault

and in royal inscriptions too – beyond the text of Sennacherib already mentioned. In the Neo-Assyrian Period, Kilizu was the capital of one of the provinces of the empire.

The 2011 and 2012 Campaigns The presence of Assyrian levels was confirmed by our first surveys on the surface and by archaeological operations carried out in 2011 and 2012 mainly on the tell – the Citadel (Rouault i. p.; Rouault/Masetti-Rouault i. p.; Rouault/Masetti-Rouault i. p.). The study of the Lower town, which is covered by fields, is limited, at least for the moment, to a geomagnetic survey program, which has started to reveal the lay-out of the larger circle of urban walls. Since the very beginning, our excavations have highlighted an important NeoAssyrian occupation in the Citadel on its south-eastern slope – Area A – marked by the presence of a large platform of red mud bricks. This terrace covers a part of the present slope of the tell stretching more than 30 m and it is associated with a monumental ramp. On a baked brick inserted in this staircase, in 2012 we found one more Sennacherib’s inscription mentioning the construction of the double system of city walls, confirming the importance of this king’s investment in a new urban order of the settlement. Middle Assyrian occupations, at that time, were documented mainly through the ceramic and epigraphic materials found in fills under this terrace. Our first excavations on the top of the Citadel, in its North-West corner – Area B (fig. 3) – have revealed that the flatness of the surface is a result of the repeated and almost complete destruction of military and administrative buildings, often of great dimensions. Built on the top of the tell in a dominant, strategic position over the Shemamok plain, these structures were replaced by similar constructions in different historical periods. The foundation of a modern Iraqi army installation with its barracks, clearly visible in the Corona images dating to the end of the 1960s, have surely wiped out the last remains of the Ottoman period fortress seen by Layard – the Qasr giving the name to the site – as well as the traces left on the surface by earlier archaeological excavations. The Iraqi army occupation has in its turn been destroyed by American bombs during the last Gulf war. This event, and the subsequent bulldozing and removal of the ruins, partially pushed down the slopes, have created the situation we were confronted with, with pits, trenches and bomb craters cutting into the stratigraphy. While the remains of an Early Islamic period are extremely limited, we have identified in this part of the Citadel, under a level of mud bricks – possibly still belonging to the Ottoman period fortress – the structure of a Sassanian/Parthian occupation, represented by a large building, with apparently no domestic or residential functions. We could identify at least three phases of construction, the earliest one marked by Parthian ceramic materials. If «Assyrian» ceramic shards are quite common only on the southern slope of the Citadel,

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

French Excavations in Qasr Shemamok, Iraqi Kurdistan

Fig. 3. Area B, Plan 2014.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

111

112

Masetti-Rouault – Rouault

Fig. 4. Area A, Plan 2014.

in these levels of Area B we have often found baked bricks, and fragments of baked bricks, with Sennacherib’s inscription, in fills or reused. Even if these bricks, as well as other more ancient epigraphic material found on the surface or in other levels, speak for the reality of the Assyrian presence, after the first campaigns we could not sometimes help thinking that the Assyrian material and structures which we found were not as abundant and self-explaining as we would have hoped. Ashamed to have the same feelings as Layard or even Furlani had, we maintain however that, if one of the main goals of our project is the identification of the Assyrian levels, our work is centered on the study of all the stratigraphic sequences in which the Assyrian levels are integrated,

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

French Excavations in Qasr Shemamok, Iraqi Kurdistan

113

Fig. 5. Area A, Parthian/Sassanian buildings above Hellenistic and Assyrian levels.

highlighted both in Area A and in Area B. It is in this spirit that we have organized our work at the site.

Campaigns 2013 and 2014, Area A During the last two seasons, in 2013 and 2014, working in Area A (fig. 4) on the south-eastern slope of the tell, we have studied again the organization of the red mud brick platform or terrace, which could be also a part of the wall encircling the Citadel itself. Thanks to a deep Hellenistic pit, we discovered that the structure built in Sennacherib’s time was set upon another similar mud brick terrace. The two platforms are separated by a level of occupation, marked by a path of baked bricks parallel in its function to the ramp associated with the later terrace. We have no way, for the moment, to date the earlier terrace, but it covers a child jar burial of the late Middle Assyrian or maybe Iron I period. This platform is built over fills which also contain early Middle Assyrian epigraphic and ceramic materials, apparently not in situ. In this part of the tell, Neo-Assyrian levels – meaning, actually, the red mud brick

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

114

Masetti-Rouault – Rouault

terrace – are sealed by an important Hellenistic period occupation represented mainly by domestic structures (fig. 5). In the limited area excavated they appear as constructed directly on top of the Neo-Assyrian platform, over an almost completely cleared surface. The ruins of these Hellenistic houses, which in a number of locations are cut by the heavy erosion of the slope, are covered directly by Parthian period structures. The 2014 excavations on the South-Eastern edge of the surface of the Citadel have allowed us to improve our understanding of the stratigraphic relations between the Post-Assyrian levels. We have highlighted the presence of a building with Parthian material which defines, even now, the profile of the tell edge itself. At least three main phases of construction have been identified, the earliest one set upon the Hellenistic ruins. This space was marked by cooking activities and possibly the later structure also corresponds to a house. Another operation was opened last spring in the southwestern part of the tell at a location on the slope which had an extensive surface presence of Middle Assyrian and Mitannian ceramics. We exposed here, again, part of the Neo-Assyrian «red» mud brick terrace which extended up to this point, covering layers with more ancient mixed materials. An even earlier level has been identified through the presence of two walls, defining a built space, with fills and floors delivering clear Middle Assyrian material. An occupation of the same period has also been found in a sounding at the base of the Citadel, at the juncture with the Lower town. These operations will continue next year, focusing on the levels which can give evidence germane to studying the impact of the integration of the local culture of Kilizu into the Middle Assyrian Empire.

Campaigns 2013 and 2014, Area B During the last two seasons excavations in Area B (fig. 3), on the Northern side of the Citadel, have been carried out to improve our knowledge of the Sassanid and Parthian period buildings already identified here, and with the further aim of reaching the Hellenistic and then Assyrian occupations in this part of the site. Confronted with the problem of the huge stratification of the levels of these periods and the presence of the deep bomb crater cutting in its northern slope in the North-Easter corner of the citadel (even though now partially filled by collapse and wash) we realised that this actually presented an opportunity to get a quick insight into the stratigraphy of the Post-Assyrian periods. The 2013 excavations in Area B, the crater area, have been an interesting experience, showing us the effects of the explosion in mud brick structures surrounding the impact point and up to 5 meters below. Actually, the bomb has crushed, burnt and even almost completely erased, an entire part of the Sassanid/Parthian settlement, allowing us to discover only the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

French Excavations in Qasr Shemamok, Iraqi Kurdistan

115

Fig. 6. Area B, sounding toward the Adad-nirari I palace.

ruins of a Post-Assyrian level, where the material seems to point to a Hellenistic date. A clear Assyrian level came up where we were not yet expecting it. Just like an episode out of the Babylonian tradition, a heavy rain storm revealed for us the presence of a line of baked bricks, visible from a distance in the northern slope of the citadel immediately west of area B a little less than two meters lower than the level of the Post-Assyrian, «Hellenistic» floor. After cleaning, these bricks, which belong to a floor, showed to have an inscription of the Assyrian king Adad-nirari I (14th century BC) announcing the presence of his palace in Kilizu (fig. 6). This floor was rebuilt at least three times but only the earliest phase bore the inscription. After this first surprise, we realized that the two later levels had been cut by a very deep pit which stopped just short of the floor with the inscription. We could be wrong, but it is possible that this pit has been excavated by one of our predecessor archeologists in the site in order to study, in his own way, the same line of bricks which we saw. But we were luckier, or more tenacious, and were rewarded with the inscriptions – very useful, because the context has almost no other material to date it. In order to avoid repeating the same methodological mistake, we launched, during the 2014 campaign, a double strategy to approach the Middle Assyrian levels and the structures of the palace. We carried out a

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

116

Masetti-Rouault – Rouault

sounding, in order to analyze quickly and more precisely the stratigraphic context, as well as to understand the origin of the pit and of other perturbations noted in the area. On the other hand, we started a new excavation from the surface, north of the Sassanid building, immediately west of the crater studied last year, and east of the sounding. This work has given us the opportunity to find and to read again the sequence of occupations erased by the bomb, still documented in this sector of the Citadel. We highlighted other structures and spaces of the earlier Parthian building which had been used for storage and conservation, and possibly milling. This occupation was built upon the ruins of a Hellenistic level, with two distinct phases of construction, maybe already transitional with a Persian Achaemenid period settlement.

Conclusions At the end of the 2014 season, we had enough elements to connect our earliest levels with the Post-Assyrian, Hellenistic/Achaemenid structures found under the crater. In the large space created in this way in the North-eastern slope of the Citadel we will be able to start a new phase of excavations during the next seasons in order to study the transition from the NeoAssyrian imperial levels to the Post-Assyrian levels. Toward the west we will surely find some evidence to explain the even more problematic evolution from the Middle Assyrian occupation, and palace, into the Neo-Assyrian one. But we are sometimes still secretly wondering: Where are the Lamassus? The Qasr Shemamok team is convinced future researches in the site will give new evidence to fully understand the functions of this city in the Assyrian imperial system, in both the Middle and the Neo-Assyrian periods, but also on its own evolution and integration in the much longer and complex history of this Northern Mesopotamian region.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

French Excavations in Qasr Shemamok, Iraqi Kurdistan

117

Bibliography Furlani, G., 1934. Gli scavi italiani in Assiria (campagna del 1933). Giornale della Società Asiatica Italiana N.S. 2, 265–276. Layard, A. H., 1853. Discoveries among the Ruins of Nineveh and Babylon, New York. Place, V., 1852. Lettre de M. Place à M. Mohl, sur une expédition faite à Arbèles. Journal Asiatique 20, 441–470. Masetti-Rouault, M. G., 2013. Interpreting the Changes in the Plan of the Assyrian Palace in Tell Masaikh – Kar Assurnasirpal (Lower Middle Euphrates Valley, Syria). D. Kertai/P. A. Miglus (eds.), New Research on Late Assyrian, Conference at Heidelberg, January 22nd 2011, Heidelberger Studien zum Alten Orient 15, 31–40. Postgate, J.N., 1980. Kilizu. Reallexikon der Assyriologie und vorderasiatischen Archäologie 6, 591–593. Rouault, O., i. p.. Qasr Shemamok/Kilizu, a Provincial Capital East of the Tigris: Recent Excavations, and New Perspectives. Proceedings of the Provincial Archaeology of the Assyrian Empire conference (University of Cambridge, December 15, 2012). Rouault, O./Masetti-Rouault, M. G., 2014. From One Valley to Another. Bir el-Haddad, a Neo-Assyrian Trading Post? D. Morandi Bonacossi (ed.), Settlement Dynamics and Human-Landscape Interaction in the Steppes and Deserts of Syria. Studia Chaburensia 4, 243–256. –

under press a. French Excavations in Qasr Shemamok-Kilizu (Iraqi Kurdistan): The First Mission

–

under press b. Recent Researches in the Erbil region: 2011 Excavations in Qasr Shemamok –

(2011). Proceedings of the 57th Rencontre Assyriologique Internationale (Roma, 2011). Kilizu (Iraqi Kurdistan). Proceedings of the 8th International Congress on the Archaeology of the Ancient Near East (Warsaw 2012). Rouault, O./Masetti-Rouault, M. G. (eds.), i. p. Travaux et recherches de la mission archéologique française à Qasr Shemamok – Kilizu (Kurdistan d’Irak) en 2011 et 2012. Etudes Mésopotamiennes – Mesopotamian Studies (EMMS), Brepols. Maria Grazia Masetti-Rouault, Professor, Ecole Pratique des Hautes Etudes, Sorbonne University, Paris, France, Co-Director of the French Archaeological Mission in Qasr Shemamok (Kurdistan, Iraq), UMR 8167 Orient et Méditerranée, CNRS, Paris, France. Olivier Rouault, Professor, Lyon2 University, Lyon, France, Director of the French Archaeological Mission in Qasr Shemamok (Kurdistan, Iraq), UMR 5133 Archéorient, CNRS, Paris, France. All images © French Archaeological Mission in Qasr Shemamok.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9th ICAANE, Basel 2014, Vol. 3, 119–130

Tevfik Emre Şerifoğlu – Jesse Casana – Claudia Glatz – Shwkr Muhammed Haydar

Initial Results of the Sirwan (Upper Diyala) Regional Project The Sirwan Regional Project is an international archaeological survey project that is conducted along the upper Diyala (Kurdish Sirwan) River valley at the south-eastern edge of the Kurdish Region of Iraq. This paper presents an overview of the geography and environment of the study area, our methods for investigating this largely unexplored region, and some initial results of fieldwork conducted between autumn 2012 and spring 2014. Our findings, while still provisional, reveal that the region possesses an extremely rich archaeological record, with major settlements dating from all prehistoric and historic phases, and help to frame some of the key issues the project will address in ongoing investigations.

Introduction The Sirwan Regional Project is an archaeological expedition undertaken jointly by the University of Glasgow (UK), Bitlis Eren University (Turkey), and Dartmouth College (USA), in cooperation with the Garmian Department of Antiquities, headquartered in Kalar, KRG. The project, which was initiated in 2012, undertakes extensive archaeological survey, paleoenvironmental investigations, and targeted excavations throughout our study area in the Sirwan (upper Diyala) River Valley, encompassing all periods of human occupation, from early prehistory through the Ottoman period. The upper Diyala constitutes a key transitional zone between the lowland Mesopotamian plains and Zagros highlands, serving as one of the primary transport and communication routes connecting these historically important regions. As such, archaeological investigations have great promise to shed light on issues ranging from the initial development of sedentary agriculture, the emergence of the world’s first complex societies, and the cultural interactions that took place between imperial centres of power in lowland Mesopotamia and the highland communities of the Zagros.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

120

Şerifoğlu – Casana – Glatz – Haydar

Despite the evident importance of the upper Diyala, the region has scarcely been subjected to any archaeological investigations in the past. The lower Diyala plains were famously surveyed by Robert McAdams (1965), and were home to major excavations by the University of Chicago’s Oriental Institute at Tell Asmar, Khafaje, and elsewhere in the 1920s (Frankfort et al. 1932; 1940; Delougaz 1940; Delougaz et al. 1942; 1967). In the 1970s, in advance of the construction of a major dam just north of the Jebel Hamrin, salvage excavations were undertaken at a number of sites (Roaf 1982; Postgate/Roaf 1982; Killick 1988; Yaseen 2005). However, the region to the north of the Hamrin basin and below the Shahrizor Plain1, has seen almost no archaeological research. Several of the most prominent sites in the upper Diyala corridor were mapped in the 1960s for inclusion in the national Iraqi Archaeological Atlas (Iraqi Directorate General of Antiquities 1970) and small-scale excavations have been undertaken by the Iraqi Department of Antiquities at the multi-period mound of Qalat Shirwana and the Halaf/Ubaid site of Tepe Rahim. But for the most part, the Sirwan Regional Project is based in a region that is, in archaeological terms, a terra incognita. This paper presents a description of the physical landscape of the study area, an overview of our methods for discovery and documentation of archaeological remains, and a brief summary of our initial findings from the 2012–2014 seasons. While much of the region remains yet to be investigated and many of our interpretations are necessarily provisional, our results to date nonetheless clearly demonstrate the enormous potential of the region for continued research and highlight some of the major issues we will be investigating going forward.

Geographic Setting The Sirwan Regional Project’s study area comprises approximately 4000 sq km along the main branch of the Diyala River and into surrounding highlands, extending from the town of Darbandikhan in the north to the southern edge of the Kurdistan Region at the Jebel Nasaz range (fig. 1). The Qara Dagh mountains at the northern side of the area separates it from the Kurdish plateau, home to the Shahrizor Plain and the city of Suleymaniyah, while the Zagros foothills along the modern Iranian border form the eastern edge of the study area. The study area encompasses a wide variety of environmental zones, each with major differences in resource and water availability, as well as in agricultural potential, and thus are key to understanding the settlement history of the region. At the eastern and northern edge of the survey zone are the lofty upland landscapes of the Qara Dagh and Zagros, which are 1

The Shahrizor Plain has been the focus of several recent archaeological projects (e.g., Altaweel et al. 2012; Miglus et al. 2013).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Initial Results of the Sirwan (Upper Diyala) Regional Project

121

Fig. 1. The Sirwan Regional Project research area.

covered in snow for much of the winter and rather arid for the rest of the year. The upper reaches of the main branch of the Diyala are deeply incised into a series of terraces, the youngest of which appears to be Pleistocene in date. Flanking this part of the valley to the east of the river are a series of large drainage channels that flow only during seasonal rains. In most years, the expansive plains on the upper river terrace are too arid to support reliable dry farming agriculture, although today are often farmed using water pumped from the Diyala. On the eastern side of the river there are several plains with higher agricultural potential, because of the orographic rainfall caused by the Zagros, as well as by a series of perennial streams that flow down from the mountains. Near the modern city of Kalar, the high gravel terraces give way to several low-lying basins constrained between a series of small mountain ranges, culminating with the Jebel Hamrin in the south, beyond which the Diyala flows out onto the flat plains of lower Mesopotamia and forms an anastomosing channel system. Between Kalar and the Jebel Hamrin, the Diyala forms a braided channel that is incised into the Pleistocene alluvial deposits to a sufficient depth that flooding and sedimentation on the flanking basins are rare, while traditional irrigation from the Diyala is challenging. However, these basins, particularly on the eastern side of the river, are well-watered by a large number of artesian springs, supplied by groundwater in the Zagros to the north and east. While the southern end of our survey area has far too little rainfall to support dry farming, the rich soils and abundant perennial water

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

122

Şerifoğlu – Casana – Glatz – Haydar

supply make these basins among the most agriculturally productive in the entire region. At the same time, the lack of significant sedimentation from the Diyala during the Holocene results in much better archaeological preservation than seen in the lower reaches of the river basin, where heavy sedimentation along the anastomosing channels have buried much of the landscape below many meters of sediment (Adams 1965).

Methodology The Sirwan Regional Project’s survey area, at around 4000 sq km, is very large by comparison to most survey projects in the Near East, and encompasses a wide variety of challenging archaeological landscapes, including rugged highlands, arid plateaus, and marshy irrigated fields. The near total absence of prior knowledge regarding the archaeology of the region necessitated a survey strategy designed to maximize site discovery. Moreover, because the area is witnessing rapid industrial, urban, and agricultural development, many archaeological sites and features are being destroyed or obscured at an alarming rate, and thus we have a necessity to work quickly, and prioritize our time to documenting as much as possible. In our initial seasons of fieldwork, we relied primarily on the analysis of declassified 1960s-era CORONA satellite imagery, alongside modern high-resolution colour imagery, for the identification of sites. Prior to our first field season, we mapped more than 500 probable archaeological sites, primarily concentrated in the lower plains to the south and east of Kalar. The high concentration of sites in this region may be due in part to the greater visibility of mounded sites in flat plains, but probably also reflects a real concentration of ancient settlement in these agriculturally rich and well-watered areas. Considering both the rapidly evolving security situation and access to the sites, we then select a number of sites to visit and document each day. In the field, we create maps of each site, record detailed physiographic features, and conduct grab collections of surface artefacts, demarcating them into major topographic features of sites (e.g., high mound, lower slope, etc.). Analysis of artefacts is the basis for dating occupations at sites, but as we progressed in our project over the first two seasons, we began to recognize the distinctly local character of much of the material culture assemblage. For some periods, the connections to other well-known assemblages from southern Mesopotamia, the northern Mesopotamian plains, and to a lesser extent the Iranian plateau are clear, but in many cases the ceramics cannot be confidently linked to any known sequence, complicating our ability to date much of the material. Moving forward, one of the key contributions of our work will be in constructing a more comprehensive local material culture sequence, but for the time being many of our chronological assessments must be considered as provisional.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Initial Results of the Sirwan (Upper Diyala) Regional Project

123

In 2014, we also instituted a strategy to both acquire more information from key sites and begin to improve our knowledge of key periods. Alongside more traditional regional survey, we also leave a site-based team for several days at a time at key sites, in order to conduct aerial photogrammetric mapping (using either a kite or small UAV), magnetic gradiometry survey, and targeted excavation of features. This evolving strategy offers to provide more contextual information than surface finds, in addition to datable samples that will help eventually to anchor a new local chronology.

Initial Results To date, our project has visited and recorded 107 sites, recording also a number of ancient irrigation works and other water management features, and assisting with some salvage operations undertaken by the Garmian Department of Antiquities. The majority of our efforts to date have been devoted to recording major sites recognized on satellite imagery in the plains south and east of Kalar, with some exploratory research in the more arid northern parts of the survey area. Among the sites recorded in the southern plains, we have found a rather large number of early prehistoric occupations, a fact which is made more significant given the near total absence of pre-Uruk period settlement in the lower Diyala plains (Adams 1965). The oldest settlement discovered thus far is a Pre-Pottery Neolithic site (SRP 9), with lithic material that has many linkages to finds from Jarmo not far to the west (Braidwood/Howe 1982). Numerous sites show affinities to early pottery Neolithic cultures, with some sites displaying the full range of classic Hassuna-types (e.g., Fallah, SRP 22), and others a chronologically distinct Hassuna-related phase (as at Tepe Sirwan, SRP 36, fig. 2). Several sites have produced high-quality Halaf period wares including Geydan 3 (SRP 35), and at least one, Mirwari (SRP 28), shows strong affinities to southern Mesopotamian Ubaid sites. To date, only one site, Gakol Kale (SRP 79), has produced evidence of Uruk period materials, a fact that raises some questions about the relationship of our region to the Hamrin Basin not far to the south where Uruk materials are abundant (Killick 1988). A number of large Bronze Age sites have also been documented, although dating these occupations confidently remains challenging, as some phases, particularly the 3rd millennium BC, show little resemblance to better known ceramic sequences from southern Mesopotamia or elsewhere. Nonetheless, numerous sites including Tepe Kalan (SRP18), Tepe Bor (SRP 25) and Kani Masi 6 (SRP 46), produced large quantities of materials datable to the Ur III, Old Babylonian, and Kassite periods. Tepe Kalan (SRP18) is among the largest sites in the region, with a prominent mounded feature possessing a rectangular top, suggestive of a temple mound that would be expected at significant Mesopotamian cities (fig. 3).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

124

Şerifoğlu – Casana – Glatz – Haydar

Fig. 2. Tepe Sirwan.

Fig. 3. Tepe Kalan.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Initial Results of the Sirwan (Upper Diyala) Regional Project

Fig. 4. Goblet fragments from Tepe Kalan.

Fig. 5. Tepe Gawur Kebir.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

125

126

Şerifoğlu – Casana – Glatz – Haydar

Fig. 6. Remains of a stone-lined siphon in Kalar.

Collections from the site are primarily 2nd millennium BC, including a large assemblage of so-called Kassite goblets and other similar materials (fig. 4). The site also has medieval architectural remains and a pre-Modern cemetery, and this makes it a good candidate for future excavations which would provide a full chronology of the settlement history in this region. As in other parts of greater Mesopotamia, a very large number of Parthian and Sasanian-period settlements were recorded during our investigations. While most Sasanian sites recorded to date are relatively modest in size, several are quite significant, such as the site of Tepe Gawur Kebir (SRP 11), where a massive architectural structure, likely a fortress city, is still preserved (fig. 5). The proliferation of settlement during the Sasanian period in particular is an important indication of the socio-economic changes that occurred in the region during this time, as well as the construction of grand irrigation projects (Adams 1965: 69–83). The remains of several massive canal and qanat systems were mapped on CORONA imagery, including a 40-km long channel cut to the west of Kalar. This extraordinary canal is difficult to see on the surface today, but at drainage crossings, portions of monumental stone-lined siphon channels are sometimes preserved (fig. 6).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Initial Results of the Sirwan (Upper Diyala) Regional Project

127

Fig. 7. Ottoman-period watchtower overlooking the Sirwan/Diyala.

The semi-arid northern portion of the survey area is more sparsely settled than the wellwatered plains in the south, and has relatively few mounded sites, some covered by modern villages as at the village of Tazadee, where Qalai Gawri (SRP 61) is located. Finds produced by modern residents show that the site was likely inhabited from early prehistoric times until the Sasanian period, but most of the other small sites found to date are later medieval in date, as at the small site of Ain Barlut (SRP 60). A number of Ottoman-period watchtowers (SRP 58, 59, 62 and 63; fig. 7) are also located along this part of the river valley, which functioned as the Ottoman-Safavid border for centuries. In the 2013 season, several of these towers were photographed and mapped, but given their rapidly deteriorating state, we plan to further investigate them and develop a preservation plan in the future. On the eastern side of the upper Diyala terrace, there is somewhat more promise for archaeological discovery, in part owing to the higher agricultural potential of the plains that lie at the base of the Zagros foothills. This region includes the 2nd millennium BC rock relief of Darbend-i-Belula, which is located on a mountain pass linking the area to western Iran (Börker-Klähn 1982: 139–140; Postgate/Roaf 1997), and likely served as an ancient route connecting Mesopotamia to the Iranian plateau. Exploratory survey in 2014 located several

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

128

Şerifoğlu – Casana – Glatz – Haydar

prominent mounded sites in the valley near Darbend-i-Belula, including SRP 106 and 107, both showing long histories of occupation dating back to the Ubaid period. The area thus offers a promising region for further investigation in future seasons.

Future Directions In the relatively brief first field seasons of the Sirwan Regional Project, we have successfully developed a solid foundation to understand the geography of the region, the range and type of archaeological sites located within it, and major differences in land use and settlement that characterize different environmental zones as well as historic time periods. We have shown that the region possesses an exceptionally rich record of early prehistoric sites from the earliest sedentary human settlement, is home to a number of historically significant Bronze Age urban centres, and has an extraordinarily robust record of later period settlement and irrigation, particularly in the Parthian and Sasanian periods. We have also begun to recognize key distinctions in the settlement and material culture across different areas within the larger region, particularly for example in the sites that populate the upland plains that flank the Zagros foothills versus the expansive mounds of the low-lying irrigated basins in the south. These findings are helping to shape our plans for future research in this still poorly-known but promising region. One critical task, as discussed above, will be to work to refine the material culture sequence for the upper Diyala through continued excavation at key sites and absolute dating of finds. It has become clear that a large portion of the artifactual assemblage in most periods, and virtually all of it in some phases, is of a distinctly local character, with few type fossils that can confidently be linked to other regional chronologies. This fact creates a stumbling block in our ability to date occupations on the basis of surface collection, but also offers an opportunity to chart diachronic change in degrees of connectivity between the upper Diyala and surrounding traditional centres of power. Our ongoing work will expand our approach of excavating targeted test soundings at key sites in order to build a better chronological sequence. Our regional survey efforts will also continue, documenting more of the still hundreds of unrecorded mounds in the southern basins, expanding our work considerably in some of the upland plains, as well as exploratory pedestrian survey into the arid drainages to the west of the Diyala. Regional settlement survey will work in tandem with more site-focused studies, especially through geophysical prospection and aerial mapping in order to reveal potential distinctions in the built environments of settlements over time as well as in different areas. Geophysical data will also facilitate more precise excavation of key features. In addition, archaeological survey efforts will be closely coordinated with geomorphological and

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Initial Results of the Sirwan (Upper Diyala) Regional Project

129

paleoenvironmental investigations in order to clarify the ways in which the landscape has been transformed over the Holocene through both climatic and anthropogenic mechanisms.

Acknowledgements We would like to thank Abwbakr Osman Zainadin (Mula Awat) and the General Directorate of Antiquities and Heritage of the Kurdistan Region of Iraq in Erbil as well as the Garmian Department of Antiquities and Heritage and its staff for allowing us to work in this important area and for their on-going support. We are also grateful for the continued assistance and advice of Kamal Rashid, Director of the Suleymaniyah Department of Antiquities. We also thank the University of Glasgow, the University of Arkansas, Dartmouth College and Bitlis Eren University for their institutional support. The work presented in this paper was funded by the John Robertson Bequest of the University of Glasgow, the Carnegie Trust for the Universities of Scottland, the G. A. Wainwright Fund, the British Institute for the Study of Iraq and the King Fahd Center for Middle East Studies at the University of Arkansas.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

130

Şerifoğlu – Casana – Glatz – Haydar

Bibliography Adams, R. M., 1965. Land Behind Baghdad: A History of Settlement on the Diyala Plains. Chicago. Altaweel, M./Marsh, A./Mühl, S./Nieuwenhuyse, O./Radner, K./Rasheed, K./Saber, S. A., 2012. New Investigations in the Environment, History, and Archaeology of the Iraqi Hilly Flanks: Shahrizor Survey Project 2009–2011. Iraq 74, 1–35. Börker-Klähn, J., 1982. Altvorderasiatische Bildstelen und vergleichbare Felsreliefs. Band I–II. Baghdader Forschungen 4. Mainz am Rhein. Braidwood, R. J./Howe, B., 1982. Prehistoric Investigations in Iraqi Kurdistan. SAOC 31, Chicago. Delougaz, P., 1940. The Temple Oval at Khafajah. Chicago. Delougaz, P./Hill, H. D./ Lloyd, S., 1967. Private Houses and Graves in the Diyala Region. Chicago. Delougaz, P./Lloyd, S., 1942. Pre-Sargonid Temples in the Diyala Region. Chicago. Frankfort, H./Jacobsen, T./Preusser, C., 1932. The First Season’s Work in Eshnunna, 1930/31. Chicago. Frankfort, H./Lloyd, S./Jacobsen, T., 1940. The Gimilsin Temple and the Palace of the Rulers at Tell Asmar. Chicago. Iraqi Directorate General of Antiquities, 1970. Archaeological Sites in Iraq. Baghdad. Killick, R. G. (ed.), 1988. Excavations at Tell Rubeidheh: An Uruk village in the Jebel Hamrin. Warminster. Miglus, P. A./Bürger, U./Fetner, R. A./Mühl, S./Sollee, A., 2013. Excavation at Bakr Awa 2010 and 2011. Iraq 75, 43–88. Postgate, J. N./Roaf, M. D., 1982. Excavations in Iraq, 1979–80. Iraq 43.2, 167–198. –

The Shaikhan Relief. Al-Rafidan 18, 143–156.

Roaf, M. D., 1982. The Hamrin Sites. J. Curtis (ed.), Fifty Years of Mesopotamian Discovery: the Work of the British School of Archaeology in Iraq, 1932–1982. London, 40–47. Yaseen, G. T., 2005. Settlements of the Hamrin Basin, Iraq from Early Times to the End of the Old Babylonian Period. Adumatu 11, 27–38.

Tevfik Emre Şerifoğlu, Bitlis Eren University. Jesse Casana, Dartmouth College. Claudia Glatz, University of Glasgow. Shwkr Muhammed Haydar, Directorate of Antiquities and Heritage of Garmian.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 131–145

Tim B. B. Skuldbøl – Carlo Colantoni

First Results of the Rania Plain Survey. Salvage Operations in the Dokan Dam Inundation Zone This paper presents the first results of a salvage survey project being conducted on the Rania Plain in northwestern Iraq. The survey project is a collaboration between the Netherlands Institute for the Near East in Holland and the University of Copenhagen in Denmark.

Introduction This paper presents the first results of the recent archaeological survey and salvage operations on the Rania Plain in the Kurdish region of Iraq. The project is a collaboration between the Netherlands Institute for the Near East in Holland (NINO) and the University of Copenhagen in Denmark.1 The aims of the project are to investigate the development of settlement on the Rania Plain, and to record valuable data to be employed in the cataloguing and protection of the plain’s archaeological heritage. The survey in 2013 was conducted by a team from the University of Copenhagen in collaboration with NINO2 and with the help of colleagues from Bitlis Eren University in Turkey, the University of Pennsylvania, and staff from the Sulaimaniyah Antiquities Directorate.3 The following discussion presents the authors’ preliminary observations of the survey work on the Rania Plain.

1 2

3

The NINO archaeological project is directed by Jesper Eidem and the University of Copenhagen archaeological project is directed by Tim B. B. Skuldbøl. In the autumn of 2014 the NINO project also undertook a trial survey in an eastern quadrant of the plain. The NINO and the University of Copenhagen projects are collaborating to integrate their findings and assemble a comprehensive record of the settlement patterns on the plain. Barzan Baiz (Sulaimaniyah Antiquities Directorate), Carlo Colantoni (Bitlis Eren University), Henrik Brahe (University of Copenhagen), Rasmus Birk Nielsen (University of Copenhagen), Tim B. B. Skuldbøl (University of Copenhagen), Guenever Bjerre Thaarup (University of Copenhagen), and Jill Weber (University of Pennsylvania).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

132

Skuldbøl – Colantoni

Fig. 1. Landsat satellite image from 1999 of the Rania Plain. Lake Dokan, located in the centre and southern parts of the plain, is an artificial lake created in the late 1950s.

The Rania Plain is an area of outstanding beauty, consisting of an enclosed and archaeologically rich valley in the western foothills of the Central Zagros Mountains in northeastern Iraq, about 70 km east of Erbil. The plain is triangular in shape and measures approximately 30 by 20 km. Together with the interconnected Pishdar Plain to the east these plains cover approximately 800 km 2 or about two thirds of the larger Shahrizor Plain to the south (fig. 1). Surrounded by hills and mountains, the Rania Plain is situated in a spatially and ecologically constrained environment. Access to and from the plain is only possible through one of the corridors located in each corner of the plain. The Sungasur Gorge, the most eye-catching

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

First Results of the Rania Plain Survey

133

Fig. 2. Lake Dokan and the Darband pass with Kur in the foreground (Drone photography, Henrik Brahe 2013). The multipurpose Dokan Dam was constructed to regulate the flow of the Lesser Zab River, store water for irrigation and to provide hydroelectric power. The size of Lake Dokan fluctuates considerably over the year but covers, in most years, about a fourth of the plain.

hill pass, gives access to the Rania Plain from Pishdar plain to the east. It is also through the Sungasur Gorge that the Lesser Zab River enters the plain. Together with numerous natural springs, the Lesser Zab River offered very favourable opportunities for human habitation on the plain in antiquity. At present, the Rania Plain is an endangered archaeological landscape. Archaeological sites on the plain are heavily affected by pressure from a range of threats from modern development and from the artificial Lake Dokan created by the Dokan Dam that was constructed in the late 1950s (fig. 2).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

134

Skuldbøl – Colantoni

Previous Research on the Rania Plain Prior to the flooding of Lake Dokan the inundation area was the subject of archaeological research. A survey undertaken by Behnam Abu Al-Soof and a team of fellow Iraqi archaeologists in the mid-1950s revealed that the plain has a temporally profound and complex settlement history, covering the period from the earliest farming cultures until recent times (fig. 3) (Al-Soof 1970). Today almost all of the sites recorded by Al-Soof now lie in the lake’s flood zone. The majority of these are submerged, some are islands and a small number are revealed annually as the lake’s waters recede during autumn and winter. Regrettably, only 10 out of 40 identified sites (including Basmusian, ed-Dem, Kamarian, Qarashina and Shemshara) had been investigated with trial excavations prior to the flooding of the plain. The best-known site on the plain is Tell Shemshara. It was initially excavated by a Danish team in 1957 which found important occupational remains from the early Neolithic and Middle Bronze Age (Mortensen 1970; Eidem 1992; Eidem/Læssøe 2001). The earliest Neolithic levels at Shemshara are currently being investigated by the University of Reading (directed by R. and W. Matthews), while NINO has resumed excavations of the Middle Bronze Age levels (see Eidem 2013).4

Investigating Early Urbanism at Bab-w-Kur In 2012 the University of Copenhagen initiated – simultaneously with the resumed excavations at Shemshara by NINO – an archaeological project on the Rania Plain. This project is investigating early urbanism on the eastern margins of Upper Mesopotamia. The aims are to better understand the nature, chronology and extent of early urban occupation on the Rania Plain and in the Zagros foothills. The salvage excavations conducted at the twin sites of Bab-w-Kur (translating from Kurdish as ‹Father and Son›), which lie deep within the inundation zone of Lake Dokan and therefore are only accessible for a limited period of the year, have uncovered important archaeological architectural and cultural material dating to the Late Chalcolithic period.5 Although Bab-w-Kur is relatively small (Bab is approximately two hectares and Kur is one hectare) it has proved in many ways to be a very interesting archaeological site complex. Erosional action by Lake Dokan’s waters has removed surface soil from large parts of the main mound, Bab, together with yet unknown numbers of archaeological deposits. This has left many archaeological features and remains completely exposed on the surface (fig. 4). 4 5

For more information about Tell Shemshara, see www.facebook.com/NINO.Leiden For more information about Bab-w-Kur, see www.facebook.com/babwkur.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

First Results of the Rania Plain Survey

Fig. 3. Map with distribution of sites surveyed by Al-Soof (redrawn map of Al-Soof 1970: pl.1.)

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

135

136

Skuldbøl – Colantoni

Fig. 4. Drone photograph of the main mound Bab and a photograph of a ‹sherd› beach at the northern end of the site (Photos: Henrik Brahe 2013).

The first season of archaeological work, in the autumn of 2012, at Bab-w-Kur aimed at investigating the nature of the Chalcolithic remains at the site and their state of preservation. Targeted surface scraping and the excavation of two areas revealed well-preserved remains from the Late Chalcolithic (LC2-5, 4200–3000 BC) period including a fortification wall, large tripartite houses, workshops, pottery kilns, seals and sealings, and evidence of extensive garbage depositional events. In 2013 the high water levels of Lake Dokan made it difficult to continue the 2012 investigations at Bab. Consequently, investigations began at the smaller of the two mounds, Kur. Surface scraping and the excavations produced intriguing results. In Level 2 a fairly large monumental niched building from the Late Chalcolithic 3–4 periods was uncovered. The building was solidly built with thick walls, decorated with niches and had white plastered walls. Its tripartite layout and niched architecture finds parallels at Arslantepe in southeastern Turkey and at Tell Brak in northeastern Syria (see Frangipane 2012: fig. 4; Oates et al. 2007: fig. 11; Emberling/McDonald 2001: fig. 4; Oates 2012: fig. 6).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

First Results of the Rania Plain Survey

137

Fig. 5. Landsat satellite image from 1999 showing the 15 km radius survey area around Shemshara.

The Rania Plain Survey The NINO-University of Copenhagen survey project on the Rania Plain was initiated in late 2013 as a follow up to recent salvage excavations conducted at Shemshara and Bab-w-Kur. At present, there are dozens of archaeological projects being undertaken in the Kurdish region of Iraq. The Rania Plain Survey is itself contextualized by these new archaeological surveys and excavation projects (see Altaweel 2013).6 6

At present, there are a number of survey projects being conducted in the Kurdish region of Iraq, these include: the Eastern Habur Archaeological Survey; the Land of Nineveh Regional Project; the Upper Greater Zab Archaeological Reconnaissance Project; the Erbil Plain Archaeological Survey; the Rawanduz Archaeological Program; the Shahrizor Survey Project; the Sirwan Survey Project; a provincial survey of the Governatorate of Sulaimaniyah in collaboration with L’Institut Français du Proche-Orient Iraq that overlaps with the Rania Plain survey project; the MAIKI- Italian Archaeo-

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

138

Skuldbøl – Colantoni

The initial aims of the Rania Plain Survey project are twofold: firstly to identify and record the archaeological sites in danger of destruction by the lake’s waters and modern development, and secondly, to place these and other settlements on the plain into a wider analysis of the development of settlement complexity on the fringes of the Mesopotamian heartlands (fig. 5). Today the Rania Plain is an endangered archaeological landscape (see Skuldbøl/Colantoni 2014). In recent years there has been a huge growth in economic prosperity in Iraqi Kurdistan. This has brought rapid growth to the Rania Plain. The towns and villages are expanding and together with infrastructure development and the intensification of agricultural activities a serious threat to the rich archaeological heritage on the plain and in the wider region is emerging. The plain’s many archaeological sites are heavily affected by pressure from a range of threats from expansive modern development, such as: extensive infrastructure development; large concrete works and factories in the modern towns of Chwarqurna and Rania; rapid expansion, and sprawl of towns and villages (Boskin, Chwarqurna and Rania); by the intensification of agriculture; and by the hydroelectric Dokan Dam lake’s waters. There has been an intensification of agricultural activities on the Rania plain that particularly affects smaller, more ephemeral sites. Clearly visible across the plain is the growth of these destructive large-scale agricultural practices: expanded fields with deep-ploughing and irrigated crops, irrigation channels, and hundreds of large water-capture reservoirs created with earth diggers and bulldozers. Tell Boskin, located in the middle of the modern village of Boskin, is a good example of a site being destroyed by modern urban expansion. Relatively intact in 1956, the site was a 15m high large mound (Al-Soof 1970). Today the site is in a sorry state with less than one hectare of the site not covered by modern construction. The top and sides of the mound have been levelled for a water-pumping station. Numerous graves have also been dug into the flanks of the mound. As town houses encroach onto the mound deep, large pits have been created. These are now being expanded and filled with town garbage. These types of threats to archaeological sites are common across the Middle East. An unsusual, although not unique, additional factor in the accumulation of damage to the archaeological landscape of the plain is the Dokan Dam. Following the dam’s construction many archaeological sites in the southern half of the Rania Plain were flooded by Lake Dokan (the dam’s reservoir) and thus ‹sealed›. However, during the past decades major

logical Mission in Iraqi Kurdistan survey of the Chamchamal region; Upper Tanjero Project; and the Rania Plain Survey. Furthermore, there are a growing number of archaeological excavation projects developing across northeastern Iraq. A gazetteer of archaeological projects in the region is being prepared by John MacGinnis, Konstantinos Kopanias and Jason Ur.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

First Results of the Rania Plain Survey

139

Fig. 6. The destruction of Bab-w-Kur by wave action and the depositing of silt on the Rania Plain (Photos: Henrik Brahe 2013).

fluctuations have been registered in the water level of the lake.7 Many sites previously covered by the lake’s waters are now exposed to erosional processes as the water levels rise and fall within the annual inundation zone. This incessant action is a slow but tangible threat to the archaeological heritage lying within this zone. The Dokan Dam inundation zone is an area along the northern shore of the lake, where the height of the waters fluctuates according to season. Records show that between the spring and summer of 2000–2010 the lake rose, on average, between 7 and 22 metres. During this decade the level of the lake varied as much as 24 metres.8 This annual rise and fall submerges many sites and erodes them due to wave action. It also partially obscures them by depositing silt (fig. 6). Evidence of erosional damage to intact archaeological deposits is visible at many sites located in this inundation zone.

7

8

From the study of CORONA satellite imagery, large fluctuations in the water level of Lake Dokan seem to have occurred in the 1960s through 1990s. Little data is currently available to evaluate the extent of the damage to archaeological sites resulting from these fluctuations. According to records of reservoir level data provided by members of staff at the Dokan Dam.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

140

Skuldbøl – Colantoni

Working in the Inundation Zone of Lake Dokan The Rania Plain in the vicinity of Lake Dokan is an unusual environment in which to survey. As the water recedes the landscape is often muddy and silty making working conditions challenging, but this does have the advantage of good visibility for the identification of surface sherds. As a basis for the survey, a combination of very high-resolution (0.5m panchromatic) WorldView-29 images, open access BING and Google Maps, and older black and white CORONA images that show the landscape of the plain from the late 1960s, well before recent economic prosperity led to major urban and infrastructure development, are used. Unfortunately, the dam was constructed before the CORONA images were taken, so the southern half of the plain in these images lies under the lake. Hence insights into the previous unsubmerged landscape are not possible. The NINO-University of Copenhagen project chose a preliminary 2 by 2 km ‹test area› roughly around the site of Kur and bound, to its east and west, by wadi and spring systems. The intention was to collect data within a discrete environmentally constrained zone. This focus allowed the tracking of the distribution of sites along the north-south wadi systems that originally flowed southwards down the topographic decline of the plain towards where the Lesser Zab flowed in antiquity. The survey project decided on an intensive test area, because by sampling the density and constituent elements of the Bab-w-Kur settlement complex indications of the true complexity of the wider settlement landscape could be discerned. By intensive field walking the survey was able to record landscape features and settlements ambiguous or indistinguishable when examining satellite imagery, such as small, flat mounding; submerged sites; possible settlement sprawl; and potential off-site activity areas and areas of trash dumping. Adding to this approach was the use of drone photography. The project employed a small battery-powered remote-control aerial drone equipped with a high-resolution digital camera capable of flying up to an altitude of around one hundred metres. With this drone the survey was able to record a series of highly detailed real-time aerial images. The project is developing this technique to be a useful tool to monitor potential damage to the archaeological landscape with up-to-date photographic records and also as a means of identifying architecture and archaeological features visible on the surfaces of the mounds that have been scoured or washed clean by the action of the lake’s waters. As a tool for recording and monitoring archaeological heritage drones are a rapid and inexpensive resource.

9

From the DigitalGlobe satellite launched in 2009.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

First Results of the Rania Plain Survey

141

Fig. 7. CORONA satellite image showing the distribution of sites in the 2013 survey test area (CORONA image 1967).

Results of the Rania Plain Survey During the survey in 2013 the project identified 33 sites of which 29 were collected (fig. 7). A number of these were known from Al-Soof’s survey in the late 1950s before the flooding of the southern half of the plain. Al-Soof appears to have focused on the most visible mounds in the landscape. However, during our survey a number of quite prominent mounds and many very low, small and hard to recognize sites were newly identified. It should be noted that the following results are preliminary as the dating of sites is still an ongoing process. Preliminary examination of the material shows that there is a small number of Neolithic and early-middle Chalcolithic sites belonging to the Hassuna, Halaf and Ubaid periods. There then appears to be a rapid growth in the Late Chalcolithic, with more than twenty sites dating to the Late Chalcolithic 2–5 period, many of which with southern Uruk material

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

142

Skuldbøl – Colantoni

Fig. 8. Topographical map of Bab-w-Kur and a CORONA satellite image showing clustering of sites around Bab-w-Kur (CORONA image 1967).

(fourteen sites possess the distinctive Bevelled Rim Bowl). To this number, we can add around twenty Uruk sites from across the Rania Plain accounted for by Al-Soof (1968 and 1970). There is then a decline in settlement numbers for the subsequent periods. However, this may change as the processing and identification of material for the 3rd, 2nd and most of the 1st millennium BC is still an ongoing task. This possible decline appears to have been followed by a second peak in occupation in the Hellenistic through Seleucid, Parthian, Sassanian and Islamic periods. A rise in numbers of Late Chalcolithic and Sassanian through Islamic period sites has also been observed during the recent survey on the Shahrizor Plain (Altaweel et al. 2012: fig. 6). Although this pattern differs from that recorded by the Erbil Plain Archaeological Survey (Ur et al. 2013), undertaken to the northeast of the Rania Plain. As a further comparison, the Sirwan (Upper Diyala) Regional Project’s survey also observed high numbers of Sassanian period settlements in southern Iraqi Kurdistan (Şerifoğlu/Glatz/ Casana 2016). In the test zone around Bab-w-Kur the survey shows that settlement was concentrated on the three wadi systems of which two lay either side of the Bab-w-Kur wadi system, with 10 sites strung along the western wadi system and 11 along the eastern.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

First Results of the Rania Plain Survey

143

Investigations also suggest that in the sample area there is a complex settlement patterning with a clustering of small satellite sites (less than a hectare) around the sites of Bab and Kur (fig. 8). This complex consists of the two main sites and seven, so far identified, satellites (A–G). A number of soundings in the low mounding of these satellites produced large quantities of Late Chalcolithic 3–4 period pottery. The project believes that they could either be areas of limited ‹off-site› occupation, off-site activity areas or even the results of trash dumping.

Insights and Future Work on the Rania Plain The survey work has so far provided a number of insights into the distribution of settlement across the plain: (1) the sites are distributed, unsurprisingly, along perennial water sources; (2) many of the sites are relatively small at 0.5–2 hectares; (3) there are a number of complexes made up of smaller sites; and (4) the plain has surprisingly dense settlement patterns – particularly in the Late Chalcolithic and Hellenistic through Islamic periods. Survey and excavations around the twin mounds of Bab-w-Kur suggest that in the Late Chalcolithic period the site was a small secondary centre with production and administrative functions within a larger early urban settlement system on the Rania Plain (see Skuldbøl et al. 2014). In the Late Chalcolithic period we may be witnessing small satellite sites clustering along water systems and perhaps indicating the formation of an expansive zone of habitation related to a yet undetermined, larger nearby settlement. This is still an ongoing investigation. The project intends to build on these insights in future research seasons. The NINO-University of Copenhagen Survey project plans to continue the survey along the shoreline of the lake. There is a need to assess the damage to the archaeological heritage on the Rania Plain by identifying sites and monitoring threats of damage due to erosion by the lake waters, intensification of agriculture and expanding modern urban development. As the survey area is expanded new data providing a more comprehensive understanding of the development of the settlement patterns over time, especially issues of dispersal and nucleation, settlement density, the intensity of agricultural exploitation and issues of possible settlement coalescence on the plain will become more apparent (some of these issues will be discussed in Skuldbøl/Colantoni, forthcoming).

Acknowledgments We are greatly indebted to the Kurdistan Regional Authorities for their permission and assistance, especially the support and enthusiasm of Kamal Rasheed Raheem, Director of Sulaimaniyah Directorate of Antiquities and Heritage, and the staff at the Sulaimaniyah Directorate and Museum. We would also like to thank the governor of the town of Rania

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

144

Skuldbøl – Colantoni

for his hospitality, interest and assistance in the project, and for the warmth with which the project was received by the inhabitants of Boskin village and the plain itself. The University of Copenhagen Archaeological Project on the Rania Plain is supported by the Danish Agency for Science and Innovation – Danish Council for Independent Research; Department of Cross-Cultural and Regional Studies, University of Copenhagen; The Danish Institute in Damascus; Brødrene Hartmanns Fond; Julie von Müllens Fond – The Royal Danish Academy.

Bibliography Al-Soof, A., 1964. Uruk pottery from the Dokan and Shahrazur districts. Sumer 20, 37–44. –

1968. Distribution of Uruk, Jamdat Nasr and Ninevite V Pottery as Revealed by Field Survey Work

–

1970. Mounds in the Rania Plain and Excavations at Tell Bazmusian (1956). Sumer 26, 65–104.

in Iraq. Iraq 30.1, 74–86. Altaweel, M., 2013. Some Recent and Current Excavations in Iraq. Iraq 75, 279–280. Altaweel, M./Marsh, A./Mühl, S./Nieuwenhuyse, O./Radner, K./Rasheed, K./Saber, S.A., 2012. New Investigations in the Environment, History, and Archaeology of the Iraqi Hilly Flanks: Shahrizor Survey Project 2009–2011. Iraq 74, 1–35. Eidem, J., 1992. The Shemshāra Archives 2. The Administrative Texts. Historisk-Filosofiske Skrifter 15. Copenhagen. –

2013. “Back to Rania”. The Netherlands Institute for the Near East. Annual Report NINO and NIT 2012, 2–11.

Eidem, J./Læssøe, J., 2001. The Shemshara Archives 1. The Letters. Historisk-Filosofiske Skrifter 23. Copenhagen. Emberling, G./McDonald, H., 2001. Excavations at Tell Brak 2000: Preliminary Report. Iraq 63, 21–54. Frangipane, M., 2012. Fourth Millennium Arslantepe: The Development of a Centralised Society without Urbanisation. M. Frangipane, Origini XXXIV. Prehistory and Protohistory of Ancient Civilizations. Rome, 19–40. Mortensen, P., 1970. Tell Shimshara. The Hassuna period. Historisk-Filosofiske Skrifter 5, 2. Copenhagen. Oates, J., 2012. Early Administration at Arslantepe and Tell Brak (Ancient Nagar). M. Frangipane, Origini XXXIV. Prehistory and Protohistory of Ancient Civilizations. Rome, 169–178. Oates, J./McMahon, A./Karsgaard, P./al-Quntar, S./Ur, J., 2007. Early Mesopotamian Urbanism: A New View from the North. Antiquity 81, 585–600. Şerifoğlu, T. E./Glatz, C./Casana, J., 2016. Between the Mesopotamian Plains and the Zagros Mountains: The Sirwan (Upper Diyala) Regional Project 2012–2014. R. A. Stucky et al., Proceedings of the 9th International Congress on the Archaeology of the Ancient Near East (9. ICAANE), Vol 3. Wiesbaden 2016, 119–130.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

First Results of the Rania Plain Survey

145

Skuldbøl, T. B. B./Colantoni, C., 2014. Damage assessment of Iraq’s past: Archaeological Heritage Management on the Rania Plain, Iraqi Kurdistan. Middle East – Topics & Arguments (META) 3, 41–54. Skuldbøl, T. B. B./Brahe, H./Colantoni, C./Hald, M. M./Nielsen, R. B./Thaarup, G. B./Weber, J., 2014. Arkæologi i vandkanten – redningsundersøgelser langs Dokan-søen i det nordøstlige Irak. National Museets arbejdsmark 2014, 144–157. Skuldbøl, T. B. B./Hald, M. M./Colantoni, C., 2013. Danske arkæologer skal redde Iraks Kulturarv. www. videnskab.dk. 10. December 2013. Skuldbøl, T. B. B./Colantoni, C., Forthcoming. Early urbanism on the margins of Upper Mesopotamia – Complex settlement patterns and urban transformations on the Rania Plain in northeastern Iraq. S. al Kuntar/M. Iamoni (eds.), Trajectories of Complexity in Upper Mesopotamia. Studia Chaburensia. Wiesbaden. Ur, J./de Jong, L./Giraud, J./Osborne, J. F./Macginnis, J., 2013. Ancient Cities and Landscapes in the Kurdistan Region of Iraq: The Erbil Plain Archaeological Survey 2012 Season 1. Iraq 75, 89–117.

Tim B. B. Skuldbøl, Department of Cross-Cultural and Regional Studies, University of Copenhagen. Carlo Colantoni, Department of Archaeology, Bitlis Eren University, Turkey.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 147–159

Aline Tenu – Christine Kepinski

Kunara, a Bronze Age City on the Upper Tanjaro (Iraq) The archaeological expedition to Peramagron undertakes excavations at Kunara since 2012. This 7–10 hectare site, situated on the bank of the Tanjaro river, is one of the largest of the area. The results of the two first seasons show that Kunara was a Bronze Age centre characterized by several monumental buildings. In 2011, the Mission archéologique française du Peramagron undertook a survey in the Upper part of the Tanjaro Valley and in the neighboring valley of the Taynal River, between the pass of Bazian and the modern city of Arbat (fig. 1). 31 sites were identified (Kepinski 2014, Tenu i. p.), and three retained our attention, Bingird, Kalespi and Kunara because of their size, of their location, and of the possible occupation levels. Among them Kunara, located 5 km south-west of Suleymaniyah, was the largest, almost 10 ha and showed the longest sequence from the chalcolithic until the 2nd Millennium, that is why we started to excavate it first, in 2012 and 2013. Kunara lies in a meander of one of the main arms of the Tanjaro River. The ancient topography remains all the more elusive since an asphalted road severely damaged the archaeological deposits, but the site clearly comprised a upper town to the west and a lower town to the east (fig. 2).

The Geomagnetic Survey In 2012, before choosing the different excavation areas, a geophysical survey was conducted by Christophe Benech. The most interesting results were obtained on the top of the lower town where a monumental building, more than 60 m long and 30 m wide appears clearly on the magnetic image (fig. 3). It is organized around two courtyards roughly 400 m 2 to the South and 500 m 2 to the North. The building is bordered to the east by several parallel lines, which could indicate rows of rooms around the courtyards.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

148

Tenu – Kepinski

Fig. 1. Map of the surveyed area (H. David-Cuny).

Fig. 2. Kunara seen from the North-West. In the background, the Baranand range (A. Tenu).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Kunara, a Bronze Age City on the Upper Tanjaro (Iraq)

149

Fig. 3. Geomagnetic map of the Lower Town of Kunara (Chr. Benech).

To the south the black dots may reveal buildings erected with baked bricks, or burnt buildings. Two orientations are visible and probably correspond to two different levels. Unfortunately the steep slopes of the upper town preclude conducting such a survey there (see for more details the presentation written by Christophe Benech in Kepinski et al. i. p.).

The Excavations 2012–2013 In 2012, three areas were implemented, one on the upper town, area A, a stratigraphic trench and two on the lower town where remains appeared on the magnetic map. In 2013, a new one, D was opened with two step trenches, one to the west and one to the south in order to investigate if Kunara was surrounded by a defensive system (fig. 4).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

150

Tenu – Kepinski

Fig. 4. Topographical map of Kunara (Boris Hollemaert).

Area A (2012–2013 – Christine Kepinski) The first goal of the area A was to establish a stratigraphic sequence, which could be used as reference for other areas. A trench 5 meters wide was thus dug. It reveals a monumental building, deliberately filled with earth carefully prepared. The dating, and the reasons of this filling (level 1) remain unclear because the upper part of the tell has not yet been explored. So far, only the north–eastern part of the monumental building (level 2) has been cleared. The building comprised a large space, probably a courtyard, which measures at least 100 m2. It was delimited to the north and to the east by two walls, 2.60 m thick, built on massive stone bases (fig. 5). The northern wall is rather well preserved, and shows a very intriguing construction technique: single layers of rectangular mudbricks, protected by diluted bitumen and jointed with a mortar containing crushed bones, alternated with about 0.60 m of pisé. Two different floors, prepared with pebbles, were identified. The lower floor has been abandoned and yielded some crushed pots whereas the upper one was plastered and incompletely cleaned out before a heavy fire. A terracotta water pipe about 10 m long allowed water drainage from the courtyard; it runs north, through a room which marked the northern limit of the monumental building.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Kunara, a Bronze Age City on the Upper Tanjaro (Iraq)

151

Fig. 5. Area A. The north-eastern angle of the monumental building (level 2) (C. Kepinski).

Walls of this building rested on the remains of an older one (level 3), which seems to have roughly the same plan even if walls, reinforced by buttresses, were clearly thinner: 0.70 m instead of 2.60 m. Two different phases of occupation are attested: under the lower floor we excavated a drain covered by flat stones. Surprisingly, floors were laid before the walls were built, without foundations. These two successive buildings were erected on the top of a huge platform made of sand. This artificial terrace is at least 3.5 m high (we do not reach its base), and it covered, sealed and probably partly destroyed the previous levels. A stairway, 7 m long and 0. 90 m wide, gave access to the top of the platform. It was built on stone foundations and comprised at least one landing. On the top of the platform, a free space was left along the building. Originally it reached 4.40 m wide but it was later shortened to 2 m. The platform was surrounded by an enclosure wall, built with an external facing made of large boulders. We do not yet know what may have been the extent of the monumental building and of the platform on which it was erected but they probably occupied a large part of the upper town.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

152

Tenu – Kepinski

Fig. 6. Area B. Schematic map of the architectural remains (B. Hollemaert, R. Douaud).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Kunara, a Bronze Age City on the Upper Tanjaro (Iraq)

153

Area B (2012–2013 – Aline Tenu) Area B is aimed to excavate the monumental building identified by the geomagnetic survey. For this purpose, we opened four different sectors in which three main occupation levels were recognized (fig. 6). The most recent level (level 1) is attested in three sectors. The remains consist in stone foundations, only one course high, and maximum 0.60 m thick. They were very close to the surface and consequently rather badly preserved: the floors as well as the elevation were not found. The dating of this level is thus still uncertain, but one of them was however clearly founded on the decayed pisé of the previous level. Level 2 corresponds to the monumental building we were looking for. Two of the three parallel walls seen to the east of the geomagnetic map were found: they measure at least 35 m in length. Two walls of the northern part of the building were also discovered, and we put in light the wall which marked the southern limit of the northern sector seen on the magnetic image. Generally speaking, walls are built with massive stone footings, more than 1.30 m thick. They usually have two facings made of uncut blocks (sometimes more than 0.80 m long), and a filling of smaller stones. The easternmost wall was completed by a low bench about 0.80 m wide and to the east by a layer of pebbles. The preserved elevation of the walls shows very interesting construction techniques, which surprisingly differ within the building itself. While the superstructure of some presents an alternation of pisé and mudbricks, the elevation of others is made with uneven layers of pisé or of cob. This is the case of the northern walls of the building. Two different techniques seem to have been used for the wall in the northeast trench (still more than 1.80 m high): homogenous earth was laid on the stone footing and then recovered by rammed earth, whose layers are still visible. An oldest level (level 3) has been recognized: indeed we discovered a thick wall deliberately levelled and covered by the pebbles layer associated to level 2. Its different orientation suggests it belongs not to an older phase of the building but to an older edifice, which was also probably a monumental construction.

Area C (2012–2013 – Nordine Ouraghi) Area C was intended to study later occupation of the site. To this end, about 600 m 2 were opened. They conveyed to a depth of less than 0.50 m remains of three different levels. Very near the surface (level 1), three buildings oriented NO-SE and separated by passageways (one measuring more than 5 m wide) have been partly cleaned. The stone footings between 0.45 and 1 m in thickness were laid on the ruins of both level 2 and 3. Their floor and elevation are not preserved but numerous fragmentary mudbricks suggest the superstructures were built that way. A grindstone, several pestles and buffer as well as sherds of storage jars indicate the central building was probably devoted either to food preparation,

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

154

Tenu – Kepinski

Fig. 7. Area C: Sounding in a room of the burnt building (level 3). Above, stone footings of level 1 (N. Ouraghi).

or to craftsmanship. Unfortunately archaeological deposits belonging to that level, significantly eroded, are poorly preserved. The floors of level 2 were found directly on the top or against the walls of level 3. Contrary to level 1, the structures of level 2, as well as those of level 3, are oriented North-South and East-West. Level 2 is characterized by a monumental complex, whose limits are unclear because later works has damaged it. It was reached by an outside staircase located to the east. On the top of it, a paved passage covered with baked bricks marked the entrance to the complex composed by different kind of exterior floors around a public building. Low walls were erected to retain the pebbles layers about 5 cm thick, which constituted, perhaps the first phase, but more probably the substructure of the floors subsequently covered by backed bricks or by clay plaster. Two terracotta pipes, discovered under the retaining walls, served drainage purpose. To the north-east, the corner of a monumental building has been found. Its walls, 1.40 m thick, are carefully built with stone bases. The elevation has almost completely disappeared but was probably made of mudbricks. The doorstep of the entrance, located in the south-east

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Kunara, a Bronze Age City on the Upper Tanjaro (Iraq)

155

Fig. 8. Area D: Test trench D1 from the east (S. Rey).

corner, was marked with fragmentary baked bricks which were partly covered in a second step by three mud-bricks. The previous level (level 3) corresponds to a building of which four rooms, heavily burnt, have been identified as yet. A test trench (fig. 7) dug in one of them in 2012 revealed its floor was covered by storages jars. Even if the excavation of this level has been postponed until next season, interesting features have already been observed because of its exceptionally good state of preservation. Indeed parts of the reed roofing have been found on the top of the filling of the rooms, and the mudbricks vault which covered the passage between two of them is preserved almost to its summit. Generally speaking, architectural remains of this area, whatever the level they belong to, indicate a wider use of mudbricks than elsewhere in the site. The major interest of this area is also its main difficulty: levels are so closely interlocked, that remains of the oldest structures are sometimes higher than those of the recent ones.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

156

Tenu – Kepinski

Area D (2013 – Sébastien Rey) In 2013, two narrow trenches were put on the slope of the lower town in order to check if Kunara was protected by a defensive system, and to study the stratigraphic relationship between the upper and the lower town. The first one, D1, is located on the western side of the lower city. Four stratigraphic levels were identified on nearly 8.30 m of elevation (fig. 8). The remains of the first level are characterized by a floor –perhaps an open space–, a pisé wall, a retaining wall of blocks of limestone and a badly preserved structure whose original nature is no longer visible. The superstructure of the pisé wall, nearly 1.60 m thick, is preserved to about 1.50 m; it shows at least 9 layers of rammed earth including limestone nodules. Its footing was made of two regular faces of uncut limestone blocks, and a filling of smaller stones and of pebbles. The previous level, Level 2, is mainly documented by a tabook floor linked to two big ceramic vessels and a tannour, intentionally levelled. They were buried in the destruction layer that constitutes the level 3. Levels 1 to 3 were separated from the last level by a thick layer of sterile soil. The level 4 was poorly preserved, but provides nevertheless the northern end of a room. Walls are rather thin, with a maximum thickness of 0.85 m, and are erected with uneven limestone blocks. The upper part is not preserved and there is no evidence whether they were made of mudbrick or of pisé. About 1.80 m to the south, separated by a lane, another wall (ca. 0.90 m in width) may have belonged to another building, or perhaps to an enclosure wall. In the small sounding carried out just outside the external wall, the virgin soil was not reached. In the second trench D2 (2.50 m wide and 17 m long), located on the eastern side of the site, two levels have been recognised. The first one, a floor made of gravel, appeared immediately below the surface. It covered a heterogeneous layer, probably a destruction layer, which covered the remains of a domestic occupation. The latter is characterized by the southern part of a room delimited by three walls about 0.50 m thick. Its floor made of djuss rested on a layer of gravel. The walls are heavily eroded and only the stone footings are preserved. At least three broken jars were found in situ. The rather homogeneous layer of soil on which this level was erected was excavated to a depth of 1.20 m without reaching a significant archaeological level.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Kunara, a Bronze Age City on the Upper Tanjaro (Iraq)

157

Material and Dating The architectural structures provided by the four areas are tentatively attributed to three main levels on the basis of the pottery. Whereas the pottery assemblages seem to show a great continuity between level 2 and level 1, ceramics stemming from the earliest level (level 3) is significantly different. Further investigation is needed to confirm, and interpret this rupture in the pottery tradition. Ceramics ascribable to levels 1 and 2 may be related to materiel found in the Diyala area, but displays also good comparisons with corpus from South Mesopotamia. The most noticeable small finds consist in different kind of beads, an obsidian flint (area D2), a bronze pendant (area A), and a lithic arrowhead from a type well attested in the Akkadian period (area D1). We also discovered different kinds of pins and needles, and a cylinder seal finely carved in a greenish stone found in the monumental building of area B. It features a religious scene with two worshippers and a goddess. According to Nordine Ouraghi, it dates back to the Akkadian or to the post-Akkadian period (see Kepinski/Tenu i. p.). It finds close parallels in the Diyala region (Frankfort 1955, n. 438) and in Ur and its vicinity. The material gives a time range from the Early Bronze Age to the Middle Bronze Age, divided provisionally in three main levels, each with several phases: Level 1, dated to the beginning of Middle Bronze Age (2000–1900), is mainly documented in Area C. To Level 2, end of Early Bronze (2200–2000), belong the monumental buildings excavated in Areas A and B. Level 3, which corresponds to the Early Bronze Age between 2350 and 2200, is attested both by Area A and by the lowest level in area D1, but architectural remains ascribable to this level are so far limited. The dating of levels 1 and 2 is also founded on radiocarbon dates. However this general framework needs to be confirmed, and precised by future investigation. If we convert this periodization in historical terms, we might consider level 1 corresponds to Isin-Larsa, level 2 to Ur III and level 3 to Akkad, but the relevance of this historical terminology in our region certainly deserves further discussion.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

158

Tenu – Kepinski

Conclusions Kunara appears thus as a small town with a surprisingly high number of large constructions, which show a great diversity in building techniques. Generally speaking walls rested on stone footings. Their elevation were made either of mudbricks, or built with an alternation of earth and mudbricks or just in earth. In the latter case, techniques may also vary, and future research will be devoted to identify more precisely the construction methods (rammed earth, cob a.s.o). One may also notice the ancient inhabitants of Kunara paid a great attention to the systems of water evacuation: pebbles floors and floor preparation, terra-cotta pipes, stone drain. After just two seasons of excavations it is too early to give a clear picture of the urban layout of the settlement. In particular, the relationship between the different monumental buildings must be cleared up. Some parts of the site might have been surrounded by enclosure walls, but until now no defensive system has been found. In the 3rd millennium BC, the area around the modern city of Souleymaniyah was probably part of the Lullubum. This land, because of its natural resources and of its strategic location on trade roads toward Iran and beyond, arouse keen interest of Mesopotamian rulers as early as the Akkadian period. The stele, now in the Louvre, Naram-Sin erected was to commemorate his victorious campaign in this region. Šulgi also claimed he subjugated the Lullubum, but the repeated campaigns carried out there reveal the conquest has not been that complete. Written sources show the resistance faced by the Mesopotamian kings who entered the Lullubum, but give almost no indication concerning the land itself, its inhabitants or its organization. We hope the work that has begun in Kunara will contribute to their better knowledge.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Kunara, a Bronze Age City on the Upper Tanjaro (Iraq)

159

Bibliography Kepinski, C., 2014. Research in the Suleymaniyah Province (Iraq): The upper Tanjaro Survey. P. Bieliński et al. (eds.), Proceedings of the 8th International Congress on the Archaeology of the Ancient Near East 30 April – 4 May 2012, University of Warsaw, Volume 2, Wiesbaden, 149–164. –

i. p. Rapport préliminaire sur la première campagne de fouilles à Kunara (Mission Archéologique du Peramagron 2012): niveaux fin Bronze Ancien, début Bronze Moyen. Études Mésopotamiennes – Mesopotamian Studies 1.

Kepinski, C./Tenu, A., 2014. Kunara, ville majeure de la haute vallée du Tanjaro (Irak, Kurdistan, province de Souleymanieh), Routes de l’Orient. Revue d’archéologie de l’Orient ancien, hors série 1, 4–19. –

i. p. Two seasons of excavations at Kunara (Upper Tanjaro): an Early and Middle Bronze Age city. K. Kopanias/J. MacGinnis (ed.), Archaeological Research in the Kurdistan of Iraq and the adjacent areas, Oxford.

Kepinski, C./Tenu, A. et al., 2015. Kunara, petite ville des piedmonts du Zagros à l’âge du Bronze: rapport préliminaire sur la première campagne de fouilles, 2012 (Kurdistan Irakien). Akkadica 136, 51–88. Tenu A., i. p. Prospection dans la haute vallée du Tanjaro. Mission archéologique française du Peramagron 2011. Études Mésopotamiennes – Mesopotamian Studies 1. Aline Tenu, CNRS – UMR 7041 ArScAn, Nanterre. Christine Kepinski, CNRS – UMR 7041 ArScAn, Nanterre.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Syria

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 163–175

Cristina Baccarin

Consumption in a Temple? An Interpretation of the Ceramic Repertoire of the Early Bronze Age Temple at Tell Ahmar (North Syria) A temple dating to the Early Bronze Age was recently unearthed on the acropolis of Tell Ahmar. The temple contained a well preserved ceramic collection, with many pots found in situ. This paper focuses on these vessels, describing their shape and possible function based on their form, context of recovery and organic residue analyses performed on several samples. During the 2009 and 2010 archaeological seasons, the discovery of an Early Bronze Age temple on the acropolis of Tell Ahmar revealed a building preserved with a large number of artefacts in situ. The temple includes benches running along the inner face of the eastern and western walls (Bunnens 2016). The northern wall presents a rather complicated arrangement of installations, including three platforms in the western corner, two steps in the eastern one, and four steps in front of a platform and of a possible niche (Bunnens 2016). The temple was erected on an artificial platform on the eastern slope of the tell (trenches S2, S3, S10, S14), facing the Euphrates River in a very visible position. The use of the sacred building came to an end in a massive fire that destroyed the building and caused the roof to collapse, sealing its contents. The largest quantity of material was found on the floor in front of the installations, and it includes wooden items, a decorated spouted jar and a collection of ceramic vessels. The present paper focuses on the description of the ceramic collection from the temple and offers a tentative interpretation regarding the use of the vessels. The first part describes the context in which the vessels were found and presents a general and typological description of the ceramic assemblage. The second part discusses some vessels’ possible function according to their form, context and the data offered by the organic residue analysis performed on a few samples.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

164

Baccarin

Fig. 1. Plan of the temple of Tell Ahmar (courtesy of G. Bunnens).

The Context of Recovery Most of the artefacts found within the temple come from the northern area close to the benches and platforms (marked A and B on fig. 1). This sector is the best preserved, while the southern part was destroyed in the past by later construction and the natural erosion of the slope of the tell. Nevertheless, the concentration of the finds in front of the installations does not seem to be accidental, but may be connected with the ritual activities performed on the benches and platforms. This is also suggested by the discovery in the northern sector of a wooden sculpture of a bull’s head, a charred wood item and a kind of domino (Bunnens 2016; Roobaert, forthcoming).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Consumption in a Temple? … the Early Bronze Age Temple at Tell Ahmar (North Syria)

165

Fig. 2. Complete vessels from the temple (photo by G. Bunnens).

Pottery Analysis The ceramic collection from the temple includes nearly sixty complete or almost complete vessels, a neckless wide-mouth spouted jar with incised decoration, and many sherds (fig. 2). Small goblets and small cups are the most distinctive and abundant shape profiles of the ceramic repertoire. These pots are in Plain Simple Ware, characterized by a fine or medium fine texture, with a few inclusions. The surface is generally even and regular; only a few pots have a corrugated surface. As for morphology, they have been arranged into a preliminary shape typology that includes five types. •

Type A: thirteen plain or corrugated small goblets with straight walls, a thinned or rounded rim and a base that varies from flat, to disk or ring (fig. 3a–c). • Type B: thirty-three small goblets with an ‹S› shaped profile, a rounded or out-turned thinned rim, and a flat or ring base (fig. 3d–i). These small goblets are the most numerous within the temple’s ceramic collection. • Type C: three small goblets with straight walls, a thinned rim and a flat base. The surface is self-slipped and roughly corrugated inside (fig. 3j). • Type D: six small cups with a thinned or rounded rim and a flat or disk base. The profile varies from slightly flared to straight (fig. 3k–l). • Type E: two ovoid goblets with slightly incurving walls, a rounded or vertical rim, a flat or ring base and smoothed surface (fig. 3m–n). One of the small cups bears a potter’s mark. In addition to these small vessels, three small pot-stands have been found among the small goblets and the cups (fig. 3o–p). They are characterized by a slightly concave profile with an extremely smoothed surface. The rim is out-turned and modelled with two or three grooves.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

166

Baccarin

Fig. 3. Ceramic repertoire from the temple (a: F 258 PL 1103; b: F 258 PL 1013-1021-1044-1109-1118; c: F 258 PL 1124; d: F 258 PL 1039; e: F 258 PL 1083; f: F 258 PL 1042; g: F 258 PL 1061-1108-1128; h: F 258 PL 1092; i: F 258 PL 1025; j: F 258 PL 1027; k: F 258 PL 1127; l: F 258 PL 1102; m: F 258 PL 10501073-1074; n: F 258 PL 1013-1018-1043-1059; o: F 258 PL 1056; p: F 258 PL 1054; q: F 258 PL 980; r: F 258 PL 1052; s: F 258 PL 1022-1084-1099-1131; t: F 258 PL 1101). Drawing by C. Baccarin.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Consumption in a Temple? … the Early Bronze Age Temple at Tell Ahmar (North Syria)

167

Fig. 4. Photo of some sherds forming the upper part of the spouted jar decorated with two antithetic young lions.

Fig. 5. Drawing of the figurative zoomorphic scene incised on the upper body of the spouted jar (drawing by C. Baccarin).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

168

Baccarin

The ceramic collection also includes two shallow bowls with an in-turned rim. The first bowl has an angular in-turned pointed rim and a ring base (fig. 3s). The second bowl has a thickened rounded rim while the base is missing (fig. 3r). Their inner surface is burnished with a cross-shaped design. In addition, two jar fragments come from the temple. The first sherd belongs to a shortnecked jar with a thickened rounded rim (fig. 3t). The second sherd belongs to a small bottle-necked jar with an out-turned rim modelled with two grooves and a globular body (fig. 3q). This bottle-necked jar is made in an extremely fine ware with rare inclusions and fired to a high temperature. It is the only piece of the assemblage manufactured in NorthMesopotamian Grey Ware. Finally, a neckless wide-mouth spouted jar with a high carination and a modelled rim was found near the north-west platform (fig. 4). The jar has a partially broken spout attached to the shoulder. The upper body is decorated with a figurative zoomorphic scene consisting of incised figures of animals associated with geometric motifs (figs. 4–5). Below the figurative scene there are dotted triangles with a cord-like band, and circles are impressed below the spout. The scene is dominated by two antithetic young lions placed on either side of the spout. The lions’ bodies and manes are incised, as all the other figures, only their heads have been separately modelled and applied on the vessel surface. The head of the lion on the right is missing. Associated with the lions are two sinuous snakes placed over the scene and decorated with impressions of dots. The snakes’ heads pass between the lions and the spout and attack a bird. Other animals participate in the scene. A group of five quadrupeds (probably gazelles) and three birds (likely ducks) is represented to the right of the lions. To the left of the lions is a group of four quadrupeds, which likely includes two goats, a deer and a gazelle. Together with the quadrupeds there are a flying bird (and possibly a second bird above a tree), a stylized animal (possibly a fish), a scorpion and filling motifs such as a sun and a stylized tree. This type of decoration with zoomorphic motifs, incised or applied, is well attested in the Middle Euphrates, as well as in the Jezirah region, during the second half of the 3rd millennium BC. In the Middle Euphrates region, decorated vessels with zoomorphic motifs come from Tell es-Sweyhat (Holland 2006: Fig. 64.4, Pl. 93), Tell Munbaqa (Werner 2008: pls. 22–23), Tell Kabir (Porter 1995: Fig. 14), and Tell Halawa (Hempelmann 2001: Taf. 14–15). It is interesting to note that only the jar decorated with snakes and lions from Halawa was unearthed in a temple – the religious complex of Halawa Tell A. Another example of a jar with the same motif was used as a burial jar in the north sector of the high terrace Massif Rouge at Mari (Jean-Marie 1999: Pl. 73). This type of decorated vessels comes from different contexts – religious, funerary or domestic – and I evaluate the possibility that its use is connected with both religious and non-religious contexts. Valentini (2013), who has worked on the symbolism of lions associated with snakes, suggested two interpretations for

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Consumption in a Temple? … the Early Bronze Age Temple at Tell Ahmar (North Syria)

169

this association. On the one hand, it may have represented the eternal conflict between the human world, symbolized by lions, and the chaos, symbolized by snakes. On the other hand, it may have represented the conflict between two deities: Ningizzida – god of fertility and health – represented by a snake, and Ninurta – a warrior god – represented by a lion.

Residue Analysis An examination of organic residues was performed on six vessels from the temple, using the Feigl Spot Test.1 The Feigl test is a common method of analysis that registers the color reactions of organic material by adding different chemical reagents (Feigl 1966). Dr Martin Zarnkow2 conducted this test on twelve samples collected from pots from the temple, looking for the presence of wine and beer. The organic acid that is used to identify beer is oxalic, while tartaric is used to identify wine. Residue material for the analysis consisted of the whitish-grey or light brown encrustations adhering to the vessel surface and visible to the naked eye.3 Two vessel samples were collected, one from the inner surface of the vessel and one from the outer surface, the latter to prove that the residual material did not come from the surrounding soil. Two series of tests were performed for each sample: test A and test B (Tab. 1). The Feigl test shows if members of a particular class of compounds are present or absent, within a +/- range that provides one of these values (Perini 2015: 201): ++ when the change in color is very clear; + when a positive result occurred, although some ingredients are disturbing the staining or the quality of the sample is poor; +- when the change in color is not very clear; -- when no change in color occurred, due to an absence of reaction and the lack of a chemical reagent. When discussing the results of this test it is important to remember that organic residue analysis is more appropriate for storage vessels, because these receptacles store liquids or dry substances for a longer period of time, and the storing time may affect the final

1

2 3

I warmly thank Adelheid Otto who provided the opportunity to perform the organic residue analysis of some samples of ceramic material, as part of her project of analysis of beer production at Tell Bazi (Zarnkow et al. 2006a; 2006b). Together with samples from the Early Bronze Age, the organic residue analysis has also been performed on Middle Bronze Age ceramic samples from Tell Ahmar and the results have been discussed by Silvia Perini (Perini 2014, 2015). I thank for the analysis Dr. Martin Zarnkow of the Technische Universität München-Weihenstephan. For a complete description of the residue analysis methodology used at Tell Ahmar, see Perini 2014, Chapter 7; Perini 2015: 199–200.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

170

Baccarin OXALIC ACID (beer)

VESSEL TYPOLOGY

SMALL BOTTLE-NECKED JAR (PL 980)

SMALL GOBLET (PL 1124)

SMALL GOBLET (PL 1020)

SHALLOW BOWL (PL 1052)

SMALL CUP (PL 1073)

JAR (PL 1014)

TARTARIC ACID (wine)

IN

OUT

IN

OUT

Test A

++

--

--

--

Test B

++

--

--

--

Test A

++

++

--

--

Test B

++

++

--

--

Test A

+

+

--

--

Test B

+

+

--

--

Test A

+

--

--

--

Test B

+-

+-

--

--

Test A

--

++

--

--

Test B

--

++

--

--

Test A

--

--

--

--

Test B

--

+-

--

--

Table 1. Results of oxalic acid and tartaric acid in the selected vessels according to the Feigl Spot Test.

concentration of residual material. In this case, the aim of the analysis is to determine if the vessels were used for the consumption of beer or wine rather than to store them. However, it must be noted that this chemical analysis does not necessarily indicate the intended function of a vessel, but can imply the vessel’s actual use. A positive reaction to a chemical in a sample confirms the existence of a compound, not necessarily the specific or exclusive use of the vessel for that beverage (Perini 2015). According to the vessel shape typology, the selected samples have been associated with five vessel shapes: a small bottle-necked jar, two small goblets (belonging to Types A and B), a shallow bowl, a small cup (belonging to Type E) and a jar (Tab. 1). The small bottle-necked jar (PL 980, fig. 3q) clearly reacts positive (++) to oxalic acid. This result is quite surprising because, given the thick consistency of ancient beer and the narrow neck of the small jar, pouring beer from the vessel or into the vessel would not have been easy. Moreover, according to its shape, the small jar may be associated with the Syrian bottle type, which is commonly considered to be a container for aromatics, unguents or oils (Sconzo 2014: 203). The residue of oxalic acid detected in the vessel does not necessarily provide evidence for beer, even if the probability is high. The residue of oxalic acid can come from plants such as rhubarb and spinach or something derived from them, but the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Consumption in a Temple? … the Early Bronze Age Temple at Tell Ahmar (North Syria)

171

connection of these plants with this shape is difficult to establish as well (Michel et al. 1993; Perini 2014). The samples from the two small goblets (PL 1124 and 1020) clearly react positive (++) to oxalic acid for the first pot, and positive (+) for the second. Thus, the two small vessels could be used as beer containers, at least as their final use. Given their small shape it seems more probable that they represented an offering for the god, or that they were used during some rituals as a libation, rather than as pots to drink beer. Considering that samples from both the inner and outer surfaces of the goblets react positively to oxalic acid, three interpretations may be advanced. In the first case, the oxalic acid was both inside the vessels and in the surrounding soil, thus the samples from inner and outer surfaces both reacted positively. In the second case, the oxalic acid was only in the surrounding soil and contaminated the inner surface of the small pots, influencing the result of the test. In the third case, beer residue remained on the outer surface of the vessels as a consequence of use. Considering the small dimensions of the goblets, it is possible that in order to fill them they were dipped in the beer, instead of beer being poured in, explaining thus the remains of beer residue on the outer surface. The shallow bowl (PL 1052, fig. 3r) is positive for oxalic acid in experiment A, but it gives an unclear signal in the second experiment. The positive result is quite surprising because open shapes are generally associated with food rather than liquids. Finally, the small cup (Pl 1073) and the jar (PL 1014) react negatively to oxalic acid in both test A and B. The residue analysis performed on the six vessels from the temple proved inconsistent to determine the content or function of many of the selected pots. From the two vessels that reacted clearly positively to oxalic acid – the small bottle-necked jar (PL 980) and the small goblet (PL 1124) – only the small goblet may represent a beer container, while the small bottle-necked jar for its shape and for the thick consistency of ancient beer may be excluded. It is worth mentioning that none of the vessels was positive for tartaric acid (Tab. 1). This result is quite unexpected because, according to the text evidence from Ebla, wine played an important role in the economic system of Northern Mesopotamia and Syria during the second half of the 3rd millennium, as well as in ritual activities connected with the temple (Milano 1994: 436). The negative result does not imply an absence of vessels related to wine in the ceramic assemblage from the temple, but it may be related to the small number of samples collected.4

4

A predominance of ceramic samples, which resulted positive to oxalic acid, has also been observed among Middle Bronze Age vessels at Tell Ahmar (Perini 2014; Perini 2015).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

172

Baccarin

General Consideration and Chronological Remarks The ceramic assemblage from the temple at Tell Ahmar includes shapes that are well attested in the ceramic horizon of the Middle Euphrates region. The closest parallels are from Gre Virike (Ökse 2006), Jerablus (Peltenburg et al. 1995), Hadidi (Dornemann 1988), Sweyhat (Holland 2006), Tell Halawa A (Meyer 1989; Hempelmann 2005), Selenkahiye (Schwartz 2001), and Wreide (Orthmann/Rova 1991). The small goblets and cups have largely been identified in the Chaliciform Ware horizon, emerged in western inland Syria (Welton/ Cooper 2014) and are well attested in domestic and burial contexts. The association of these shapes in a cultic context of the 3rd millennium finds parallels in the Temple of the Rock at Ebla, where a large number of corrugated goblets were found in two pits from the cella (Sala 2012). Furthermore, it is interesting to note that a similar collection of small goblets was found in the South Unit of the Central Complex of Palace G at Ebla, where many small goblets and goblets in miniature were placed on a tray over a bench (L. 3463) or lined up on a stepped bench (L. 3532) (Mazzoni 1988; 1994: fig. 1; 2013). The collection of small goblets from Palace G does not come from a cultic context, but from storage rooms and it was intended for the official commensality (Mazzoni 2013). As for chronology, this ceramic collection may be dated to the Early Bronze Age IV. The ARCANE periodization offers a narrower range to date the repertoire. According to this new chronology, small goblets and cups are well attested during the Early Middle Euphrates IV/V periods, as well as swallow bowls, small bottle-necked jars and short-necked jars with rounded rim (Sconzo 2015: 103–109). Given that spouted decorated jars increase in number during phase V, and that goblets evolve to smaller varieties during the same phase (Welton/ Cooper 2014: 302), the repertoire from the temple may be dated to the EME V period. The ceramic assemblage from the temple at Tell Ahmar includes mostly shape categories relating to liquid containers, as small goblets, small cups and jars, and the results of the residue analyses, in spite of all their uncertainties, indicate that beer seems to be the favored candidate. Considering the limited capacity of small goblets and cups, I suggest that these were meant to be offerings to the gods, rather than pots used to drink beer.5 A further hypothesis is their use to perform libations of beer during ritual activities in the temple, as suggested by the discovery of one small goblet inside the spouted decorated jar. If this hypothesis is correct the spouted jar may have been used as a beer container both to pour beer and to dip the goblets in.

5

In her study of the goblets from the temple at Tell Bazi, A. Otto suggested the same function for small goblets according to the results of the residue analyses (2012: 187–188).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Consumption in a Temple? … the Early Bronze Age Temple at Tell Ahmar (North Syria)

173

It is difficult to more precisely delineate the rituals performed in the temple at Tell Ahmar on the basis of the archeological evidence, but further indications are available from textual sources. According to administrative texts from Ebla, numerous foodstuffs and various kinds of beer were offered to deities (Milano 1990; Milano/Tonietti 2012). In particular, a ritual text from the Eblaite archive L. 2712 offers further details, describing both the food offerings for the deities – bread, oil, beer, wine and purified water – and the place where these offerings are placed in the temple – in front of the god’s statue (Biga 2003; Fronzaroli 2012). This ritual text also describes a spouted vase included in the vessel repertoire used during the ritual (Biga 2003: 62–63). In accordance with this ritual text and the shape affinities between the small goblets from the temple at Tell Ahmar and those from the Temple of the Rock at Ebla, parallels concerning the vessel assemblage from Tell Ahmar and the ritual activities performed in the Eblaite temple can be drawn.

Conclusion According to the analysis of the ceramic collection from the temple, and the data obtained from the residue analysis, some preliminary considerations can be formulated. The ceramic assemblage from the temple may be dated to the Early Bronze Age IV or, referring to the ARCANE periodization, to the Early Middle Euphrates V period. This repertoire includes mostly liquid containers (small goblets, small cups and jars), and these shape categories were probably connected with rituals performed in the temple as libations and offerings to the deities. In the absence of further evidence, it is difficult to more precisely determine the rituals performed in the temple at Tell Ahmar, but according to Eblaite textual sources and to the positive reaction to the beer acid of a few samples, I may suggest that beer libations were poured to the deities during the rituals, and a jar of beer was placed before their altar as offering or for them to drink.

Bibliography Biga, M. G., 2003. A Ritual from Archive L.2712 of Ebla. P. Marassini et al. (eds.), Semitic and Assyriological Studies Presented to Pelio Fronzaroli by Pupils and Colleagues. Wiesbaden, 54–69. Bunnens, G., 2016. A Third Millennium Temple at Tell Ahmar (Syria). R. A. Stucky et al., Proceedings of the 9th International Congress on the Archaeology of the Ancient Near East (9. ICAANE), Vol. 3. Wiesbaden 2016, 187–198.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

174

Baccarin

Dornemann, R. 1988. Tell Hadidi: One Bronze Age Site Among Many in the Tabqa Dam Salvage Area. Bulletin of the American Schools of Oriental Research 270, 13–42. Feigl, F., 1966. Spot Test in Organic Analysis. London. Fronzaroli, P., 2012. The Eblaic Kings’s Supplication to the Gods of the Night (TM.75.G.756+771+815). RA 106, 165–176. Hempelmann, R., 2001. Menschen- und Tiergestaltige Darstellungen auf frühbronzezeitlichen Gefässen von Halawa A. J.-W. Meyer et al. (eds.), Beiträge zur Vorderasiatischen Archäologie Winfried Orthmann gewidmet, Frankfurt am Main, 150–169. –

2005. Die bronzezeitliche Keramik von Tell Halawa A. Ausgrabungen in Halawa 3. Saarbrücken.

Holland, T. A., 2006. Excavations at Tell es-Sweyhat, Syria, Volume 2: Archaeology of the Bronze Age, Hellenistic, and Romans at an Ancient Town on the Euphrates River. Chicago. Jean-Marie, M., 1999. Tombes et Nécropoles de Mari. Beyrouth. Mazzoni, S., 1988. Economic Features of Pottery Equipment of Palace G. H. Waetzoldt/H. Hauptmann (eds.), Wirtschaft und Gesellschaft von Ebla: Akten der Internationalen Tagung Heidelberg 4.–7. November 1986. Heidelberg, 81–105. –

1994. Drinking Vessels in Syria: Ebla and the Early Bronze Age. L. Milano (ed.), Drinking in Ancient Societies. History and Culture of Drinks in the Ancient Near East, Padova, 245–276.

–

2013. Centralization and Redistribution. The Pottery Assemblage of the Royal Palace G. P. Matthiae/N. Marchetti (eds.), Ebla and its Landscape. Early State Formation in the Ancient Near East, Walnut Creek, 89–110.

Meyer, J.-W., 1989. Tell Halawa A. W. Orthmann (ed.), Halawa 1980 bis 1986, Saarbrücker Beiträge zur Altertumskunde 52, Bonn, 11–63. Michel, R./McGovern, P./Balder, V., 1993. First wine & beer. Chemical detention of ancient fermented beverages. Analytical Chemistry 65, 408a–413a. Milano, L., 1990. Testi amministrativi: assegnazioni di prodotti alimentari (archivio L.2712 – Parte I). ARET IX, Roma. –

1994. Vino e birra in Oriente: confini geografici e confini culturali. L. Milano (ed.), Drinking in Ancient Societies. History and Culture of Drinks in the Ancient Near East (Atti del Convegno Internazionale – Roma, 17–19 maggio 1990), Padova, 421–440.

Milano, L./Tonietti, M. V., 2012. Cerimonialità alimentare ad Ebla. Offerte, pasti e sacrifici. L. Milano (ed.), Mangiare divinamente. Pratiche e simbologie alimentari nell’antico Oriente, Firenze, 33–81. Ökse, A. T., 2006. Early Bronze age graves at Gre Virike (Period II B): an extraordinary cemetery on the Middle Euphrates. JNES 65/1, 1–29. Otto, A., 2012. Defining and Transgressing the Boundaries between Ritual Commensality and Daily Commensal Practices: the Case of Late Bronze Age Tall Bazi. S. Pollock (ed.), Between Feasts and Daily Meals: Toward an Archaeology of Commensal Spaces, Berlin, 179–195. Orthmann, W./Rova, E., 1991. Gräber des 3. Jahrtausends v.Chr. im syrischen Euphrattal 2. Ausgrabungen in Wreide. Saarbrücken.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Consumption in a Temple? … the Early Bronze Age Temple at Tell Ahmar (North Syria)

175

Peltenburg, E. et al., 1995. Jerablus – Tahtani, Syria 1992–94: preliminary report. Levant 27, 1–28. Perini, S., 2014. Ceramic vessel production, use and distribution in Northern Mesopotamia and Syria during the Middle Bronze Age II (c. 1800–1600 BC). A functional analysis of vessels from Tell Ahmar, North Syria. Unpublished PhD thesis, University of Edinburgh. –

2015. Organic remains from Middle Bronze Age Ceramic Vessels at Tell Ahmar (North Syria). G. Affanni et al. (eds.), Broadening Horizons 4. Conferences of Young Researchers working in the Ancient Near East, Egypt and Central Asia, BAR International Series, Oxford, 199–204.

Porter, A., 1995. The third millennium settlement complex at Tell Banat: Tell Kabir. Damaszener Mitteilungen 8, 125–163. Roobaert, A., Forthcoming. Fragments de sculptures en bois retrouvés dans un temple du IIIe millénaire av. J.-C. à Tell Ahmar (Syrie). Sala, M., 2012. An Early Bronze IVB Pottery Repertoire from Favissae P.9717 and P.9719 in the Temple of the Rock at Tell Mardikh/Ancient Ebla. Levant 44, 51–81. Schwartz, G., 2001. Pottery. M. Van Loon (ed.), Selenkahiye. Final report on the University of Chicago and University of Amsterdam Excavations in the Tabqa Reservoir, Northern Syria, 1967–1975, Leiden/Istanbul, 223–340. Sconzo, P., 2014. Syrian Bottles. M. Lebau (ed.), Arcane Interregional vol. I, Ceramics, Turnhout, Brepols, 197–214. –

2015. Ceramic. U. Finkbeiner et al. (eds.), Associated Regional Chronologies for the Ancient Near East and the Eastern Mediterranean: Middle Euphrates, ARCANE IV, Turnhout, Brepols, 85–202.

Valentini, S., 2013. Snakes on pottery. A. Peruzzetto et al. (eds.), Animals, Gods and Men between East and West. Papers on Archaeology and History in Honor of Roberta Venco Ricciardi. BAR International Series 2516. Oxford, 1–11. Welton, L./Cooper, L., 2014. Caliciform Ware. M. Lebau (ed.), Arcane Interregional vol. I, Ceramics. Turnhout, 293–321. Werner, P., 2008. Die Keramik der frühen Bronzezeit. M. Czichon/P. Werner (eds.), Ausgrabungen in Tall Munbaqa-Ekalte IV. Die bronzezeitliche Keramik. Wiesbaden, 5–95. Zarnkow, M./Spieleder, E./Back, W./Sacher, B./Otto, A./Einwag, B., 2006a. Cold mashing process – a technology possibly used in ancient times in the Orient. Brauwelt International 2006/V, 306–310. –

2006b. Interdisziplinäre Untersuchungen zum altorientalischen Bierbrauen in der Siedlung von Tall Bazi/Nordsyrien vor rund 3200 Jahren. Technikgeschichte 73, 3–25.

Cristina Baccarin, Ca’ Foscari University of Venice.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 177–185

Anna Gómez Bach

Characterizing Red Ware : More than a Single Production at Tell Halula (Syria) at the End of the Halaf Period One of the most easily recognized ceramic production in Near East is the Halaf pottery. But, as it is well known, it is not a homogeneous production, and it is possible to identify some specific sets that can be more studied as the well-known red ware. This kind of ware is well known by having characteristic clay with a light gray and dark red slip associated to closed bowls and open plats, some of them painted with dark geometric motifs. Usually this pottery has been defined as a foreign production and it has been related to Late Halaf Levels in a chronologic and cultural period between 5600–5300 cal BC. Also, their presence is well recognized in several archaeological sites as Yarim Tepe II, Tell Halaf or Tell Halula, among others, and their presence has also been attested at surveys developed in Jezirah or upper Tigris. The set recovered from Tell Halula, belonging to a well-known stratigraphic sequence and dated by radiocarbon dates, allow us to characterize the red ware pottery production from an archaeometric point of view. The analysis of 86 potsherds from a comprehensive perspective by using archaeometric techniques, such as chemical analysis, petrographic, and PIXE as well as morphometric and basic techno functional characterization analysis, set new guidelines to identify their spread and production. The Neolithic settlement of Tell Halula is located in the middle Euphrates valley, in a zone of contact between several natural ecosystems. The known site covers nearly 2500 m2 with an uninterrupted chronological sequence ranging from 7800 to 5300 cal BC. Excavations have been ongoing since 1992, under the direction of Miquel Molist (Autonomous University of Barcelona, Spain).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

178

Gómez Bach

Fig. 1. Map of Syria and the location of the Tell Halula site. SAPPO.

Fig. 2. Image of the Halaf levels. SAPPO.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Characterizing Red Ware : … Tell Halula (Syria) at the End of the Halaf Period

179

Fig. 3. Main Late Halaf sets. SAPPO.

1. Halaf Pottery Production in Euphrates Valley The characterization of pottery production in the mid-6th millennium cal BC context at Tell Halula arises from the need to understand how consolidated agricultural and pastoral communities of the Fertile Crescent were structured and organized (fig. 1). The opportunity to study the pottery assemblages from Tell Halula (with 16,668 sherds) allowed us to study, from a comprehensive perspective by using archaeometric techniques (chemical and petrographic) as well as morphometric and basic technofunctional adscription, well stratified sets dated between 5600 to 5300 cal BC (fig. 2). Halula Late Halaf ceramic assemblages include a large variety of wares. There are fine wares, plain or decorated, simple and burnished coarse ware, either with mineral or plant inclusions. Overall it is a finely made pottery with monochrome and little polychrome decoration. This decoration has good adhesion and pigment is highlighted in combinations of matte and glossy, although the latter is less abundant. Most documented colors comprise the variety of

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

180

Gómez Bach

colors of brown, black and red; some productions are orange (fig. 3). Related to this ware a small group has been identified that can be attributed to red ware (0, 04%). Red ware was identified by Oppenheim and Schmidt at Tell Halaf as very late pottery production. These sets were attributed to the sequence called Late Halaf and later to Halaf transitional Obeid (Davidson 1977; Leenders 1989) in identifying these at Tell Aqab. While attribution is doubtful, red slipped productions are also available in sets of Yarim Late Halaf Tepe (Merpert/Munchaev 1993) or Shams ed-Din (Gustavson-Gaube 1981). 26 fragments were recovered in the course of prospecting the Khabur from 4 specific sites. as Aïn el Qerd, Khaneké Tell, Tell Baqar, all in the Wadi Dara (Nieuwenhuyse 2000: 169). Some authors speak of a much more heterogeneous production than might have been expected. Through new archaeometric analysis the red ware characterization will be definitely characterized and bounded in space and time within the horizon Halaf.

Archaeometric Results Ware known as red ware is present at Tell Halula with 7 potsherds. Broadly, these pieces are characterized by the presence of a red slip, often with a polished finish and a black painted decoration in some of their simple geometric patterns (horizontal and vertical lines or triangles). This series has been attributed to Late Halaf and is well documented in Tell Halula deposits – external floors and pits – in sector 30, 31 and 49 (fig. 4). This set has been sampled and numbered as: THL-412, THL-416, THL-417, THL-502, THL-505, THL-506, THL-508, THL-509, THL-512, THL-514, THL-515, THL-517, THL523, THL-530, THL-541, THL-549, THL-601, THL-604, THL-608. This production corresponds to 18’8% of the petrographic group identified as THL3, and chemical group SI-1; it is mainly represented by small and medium sized open bowls (G.III) and closed bowls (G.IV) (fig. 5).

Remarks The archaeometric analysis done at Tell Halula Late Halaf pottery has documented an association between the chemical group SI-1, petrographic group HL3 and red ware. From a petrographic point of view the group HL3 presents abundant mineral inclusions, small to medium with acicular clay (fig. 6). The appearance is heterogeneous using polarized light, while polarized light with analyzer is more anisotropic. The structure of the dough is fluidal. Abundant pores, not too large, elongated and arranged parallel to the vessel wall forms. The degreaser consists of: quartz (wavy extinction); biotite, plagioclase; basalt (few fragments); pyroxene. The presence of carbonates of contamination on the outer surfaces of the sample is observed. From a chemical point of view, the group SI-1 has relatively low

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Characterizing Red Ware : … Tell Halula (Syria) at the End of the Halaf Period

Fig. 4. Red ware shapes from Tell Halula. SAPPO.

Fig. 5. The mineralogical composition of Red ware and finishing techniques. SAPPO.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

181

182 Id.

Gómez Bach Ue.

Image

Petro

Chemical

Id

Group

Id

Ware

Shape

Group

THL 2007-49E10(A1j)-4

E10 (A1j).4

HL512

THL-3

SI-78

SI-1

PLAIN

VIII

THL 2007-49E42(A6b)-1

E42 (A6b).1

HL514

THL-3

SI-80

SI-1

PAINT

IV

THL 2007-49E32(A7c)-8

E32 (A7c).8

HL508

THL-3

SI-76

SI-1

PLAIN

II

THL 2007-49E8(A1i)-4

E8 (A1i).4

HL505

THL-3

SI-74

SI-1

PAINT

III

THL 2007-49E24(A7d)-16

E24 (A7d).16

HL515

THL-3

SI-81

SI-1

COARSE

III

THL2007-49- E24 E24(A7d)-17 (A7d).17

HL530

THL-3

SI-93

SI-1

PAINT

VII

THL199932-A4

A4

HL604 red ware

THL-3

SI-65

SI-1

PAINT

IV

THL 199630-Ag- 3b

A3b

HL608

THL-3

SI-69

SI-1

PLAIN

III

Fig. 6. Table with main sherds that belong to the S1 chemical group. SAPPO.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Characterizing Red Ware : … Tell Halula (Syria) at the End of the Halaf Period

183

SI-1 (n=8) m

ds

Fe2O3 (%)

6.31

0.27

Al2O3 (%)

12.91

0.42

TiO2 (%)

0.76

0.03

MgO (%)

5.00

0.22

CaO (%)

21.75

1.81

SiO2 (%)

53.18

1.59

Rb (ppm)

72

9

Th (ppm)

10

1

Nb (ppm)

19

1

Zr (ppm)

191

11

Y (ppm)

27

2

Ce (ppm)

55

4

Ga (ppm)

15

1

V (ppm)

109

9

Zn (ppm)

108

11

Ni (ppm)

163

9

Cr (ppm)

187

14

Fig. 7. Table with S1 petrographic group composition. SAPPO.

values of CaO and a larger relative presence of SiO2. Also, in relation to the other defined Tell Halula groups, it shows lower values relative to MgO, P2O5, Fe2O3, and MnO. This group was preliminary attributed to Tell Halula local productions and opens a new perspective on red ware provenance (fig. 7). As other archaeological sets, red ware has been associated to open forms, mainly open and closed bowls and, to a lesser extent, to plates and medium jug type; with red slip surfaces applied to a smoothed or burnished surface and may have a geometric decoration in black color. Numerous authors have identified red ware as a foreign and late product. Usually the attestation of these vases at different places has been explained by the existence of social networks, the exchange of ideas, materials or persons between groups. In these contexts variables as human group mobility, the assignment to specific morphometric group or functional one were considered.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

184

Gómez Bach

Acknowledgment This work was performed in the framework of different research teams funded by the Autonomous University of Barcelona and the Agaur/Dursi-Generalitat de Catalunya called: SAPPO (Seminari d’Arqueologia del Pròxim Orient (2005 SGR 00241; 2009 SGR 00607) directed by Miquel Molist (SAPPO, Departament de Prehistòria, Universitat Autònoma de Barcelona). Moreover, this research is part of the main project «Origen de las sociedades agrarias en el Próximo Oriente: Consolidación de las comunidades neolíticas en el norte de Siria y sudeste de Anatolia», Ministerio de Ciencia y Innovación Ref. HAR 2010-18612.

Bibliography Breniquet, C., 1996. La disparition de la culture de Halaf. Les origines de la culture Obeid dans le nord de la Mésopotamie. Ed. Recherche sur les Civilisations. Paris. Campbell, S., 1992. Culture, Chronology and Change in the Later Neolithic of North Mesopotamia. PhD, University of Edinburgh. –

2007. Rethinking Halaf Chronologies. Páleorient 33.1, 103–136.

Campbell, S./Fletcher, A., 2010. Questioning the Halaf–Ubaid Transition. R. A. Carter/G. Philip (eds.), Beyond the Ubaid: Transformation and integration in the late prehistoric societies of the Middle East. Studies in Ancient Oriental Civilization 63. The Oriental Institute of the University of Chicago, 69–83. Davidson, T. E., 1977. Regional Variation within the Halaf Ceramic Tradition. PhD. University of Edinburgh. –

1981. Pottery Manufacture and Trade at the Prehistoric Site of Tell Aqab, Syria. Journal of Field Archaeology 8, 65–77.

Davidson, T. E./McKerrell, H., 1976. Pottery Analysis and Halaf Period Trade in the Khabur headwaters region. Iraq 38, 45–56. –

1980. The Neutron Activation Analysis of Halaf and Ubaid Pottery from Tell Arpachiyah and Tepe Gawra. Iraq 42, 155–167.

Gómez-Bach, A., 2011. Caracterización del producto cerámico en las comunidades neolíticas de mediados del VI milenio cal BC: el valle del Éufrates y el valle del Khabur en el Halaf Final. PhD Universitat Autònoma de Barcelona. –

2013. Caracterización de las produciones cerámicas del Neolítico Final, Horizonte mediados del VI milenio Cal BC de Tell Halula. M. Molist (ed.) Tell Halula: un poblado de los primeros agricultores en el valle del Éufrates, Siria. Tomo I y II. Ministerio de Educación y Ciència, 212–231.

Gómez-Bach, A./Cruells, W./Molist, M., 2014. Late Neolithic pottery productions in Syria. Evidence from Tell Halula (Euphrates valley): a technological approach. M. Martinón-Torres (ed.), Craft and Science: International perspectives on archaeological ceramics. Doha, Qatar: bloomsbury Qatar

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Characterizing Red Ware : … Tell Halula (Syria) at the End of the Halaf Period

185

Foundation http://dx.doi.org/10.5339/uclq.2014.cas.ch14. Gustavson-Gaube, C., 1981. Shams ed-Din Tannira. The Halafian Pottery of Area A. Berytus 29, 9–182. Leenders, R., 1989. The red wash ware ceramic assemblage in Syria: a review. O.M.C. Haex et al. (eds.), To the Euphrates and Beyond. Rotterdam, 89–101. Merpert, N. Y./Munchaev, R. M., 1993. Yarim Tepe II: The Halaf levels. N. Yoffee/J. J. Clark (eds.), Early Stages in the Evolution of Mesopotamian Civilization: Soviet Excavations in Northern Iraq. Tucson, 128–162. Molist, M. (ed.), 1996. Tell Halula (Siria). Un yacimiento neolítico del valle medio del Éufrates. Campañas de 1991 y 1992. Madrid. –

2013. Tell Halula: un poblado de los primeros agricultores en el valle del Éufrates, Siria. Tomo I y II. Madrid.

Nieuwenhuyse, O., 2000. Halaf settlement in the Khabur Headwaters. B. Lyonnet (ed.), Prospection archéologique du Haut-Khabur Occidental (Syrie du Nord-Est). Bibliothèque archéologique et historique CLV. Beirut, 151–260. Oppenheim, M. F., 1943. Tell Halaf I. Die prähistorischen Funde (bearb. von H. Schmidt). Berlin.

Anna Gómez Bach, Universitat Autònoma de Barcelona, Spain.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 187–198

Guy Bunnens

A 3rd Millennium Temple at Tell Ahmar (Syria) Excavations in 2009 and 2010 at Tell Ahmar, on the Upper Syrian Euphrates, revealed an EBA IV temple with a complex system of benches and platforms, fragments of wooden sculpture–especially a bull’s head–and an abundant ceramic corpus. The present paper gives an overview of the results obtained at the end of the 2010 season before further research was made impossible at the site by the outbreak of violence in Syria. More than ten years ago, a long trench dug on the south-eastern slope of the tell exposed the corner of a monumental building which was obviously of some significance, perhaps a temple (Roobaert/Bunnens 1999; Bunnens 2002–2003), but it is only in 2009 and 2010 that the Middle Bronze levels that were covering the structure could be entirely removed and that the identification of the building as a temple could be confirmed. It is located in Area S in the trenches marked 2, 3, 10 and 14 on the south-eastern slope of the tell (fig. 1). When research was abruptly interrupted, the floor had been reached on an east-west strip, which was between 1 m/1.50 m wide, along the installations that were set against the north wall (fig. 2). What has been exposed is nonetheless significant enough to get an idea of the internal organisation of the building. The material collected from the building allows a date in the Early Bronze Age IV in traditional chronology, i.e. in the latter part of the 3rd millennium B.C.E. (Baccarin 2016).

Architecture The structure is a long single-room building extending approximately from north to south (fig. 3). Its southern end is lost, destroyed either by erosion or taken away by the deep sounding in which the French archaeologists found Chalcolithic pottery in the 1930s. With an inner width that is slighltly more than 6 m and a preserved length, from the stone footing of the north wall to the end of the preserved structures, of about 9 m, it is clear that the Tell Ahmar temple is definitely more than a small chapel.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

188

Bunnens

Fig. 1. Map of the tell of Tell Ahmar. The black circle shows the location of the temple.

Fig. 2. The steps, benches and platforms against the north wall of the Tell Ahmar temple as seen from the south-east.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A 3rd Millennium Temple at Tell Ahmar (Syria)

189

Fig. 3. Plan of the Tell Ahmar temple.

The walls are made of mudbricks resting on a stone footing made of irregular blocks. They differ in width and construction. The north wall is irregular. Its north face is concave, which may be due to the circumstances of its destruction, although this cannot be demonstrated. Its width varies from 1.50 m to 2 m. The west wall is about 1.20 m wide and along its exterior side runs a thin line of loose soil, a kind of foundation trench, which separates the temple from a more solid but indistinct aggregation of soil, bricks and blocks of terre pisée extending further west. This solid mass of building material, which is preserved at the same height as the west wall of the temple, does not result from the collapse of another construction. It rather looks like a kind of terrace. The temple would therefore have been erected against a preexisting artificial mound. The east wall is thicker and more complex than its western counterpart. It is about 1.40 m wide. A kind of buttress is flanking its outer face to the east and a kind of bench, about 50 cm wide, extends along the wall on the same side. The outer side of this bench could be followed down to a depth of about 2 m at its northern end. To the south, however, it is no longer

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

190

Bunnens

visible, probably destroyed by the Iron Age fortifications of the acropolis. Rather than a bench it must therefore have been the retaining wall of a platform on top of which stood the temple. Traces of white plaster were found coating the top and side of this retaining wall. We must therefore imagine that the temple was leaning on a kind of terrace to the west whereas, to the east, it was overlooking the surrounding plain from a considerable height. To the north there might have been more constructions extending into the unexcavated part of the site. To the south the temple is destroyed and we can only conjecture what was there. Such a position on the slope of the mound is remarkable. Temples normally stand on firmer ground, although it must be observed that the in antis temple at Qaraquzaq was also built on the slope of the tell (see below). Actually, more evidence for the existence of a terrace has been found in other places of the tell (fig. 4). A similar mixture of soil and bricks was recognized in association with other Early Bronze Age structures, giving the impression that all the tell, at least along its southern flank, was consolidated by terracing work before constructions were erected above it. The temple seems thus to have stood on a high platform and against a terrace. It must have been visible from a great distance in the Euphrates valley for anyone coming from the east and south-east. It was a landmark in the valley and, as we may assume, a monument of some significance to the people of the region. It disappeared in a great conflagration and was abandoned. No attempt seems to have been made to retrieve objects from the debris and no other temple was built on its ruins. A poor occupation, probably dating from the beginning of the MBA, covered the remains of the temple.

Installations A large part of the temple remains were removed down to a few dozens of centimetres above the floor by a cut in the slope of the tell, which was made to accomodate Middle Bronze Age structures. These constructions left untouched the western part of the temple along the west wall, which, as a result of this, is still about 2 m high at its northern end whereas the east wall is only 50 cm. Quite impressive is the number of beams found lying on the floor. They were identified as the famous Populus euphratica by Philippe Gerrienne from the Geology Department of Liège University. Their average diameter is about 20 cm and some are still more than 4 m long. However, the length they had to cover, from the east to the west wall, was, as already said, more than 6 m, which means that the beams must have been originally at least some 7 or 8 m long. This is quite possible as the Euphrates poplar can reach a height of 15 m.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A 3rd Millennium Temple at Tell Ahmar (Syria)

191

Fig. 4. Plan of the Early Bronze Age remains excavated on the tell of Tell Ahmar (a) and section through the same remains showing where a possible terrace has been recognized (b).

However, if the preserved length of the beams is any indication of their original length, it should be assumed that the building was covered by a kind of timber framework. All these beams are hollow as if they were kinds of pipes. As Marie-Agnès Courty explained when she visited the site, it may be due to the fact that the wood was not completely carbonized. Only the exterior was and the inside decayed as it normally happens when timber is buried. The most striking feature is a complex combination of benches, steps and platforms set against the north wall (fig. 2). Two low benches are running along the east and west walls respectively. The most elaborate arrangement is set against the north wall. Three main components can be recognized. First, in the north-west corner, where the installations are entirely preserved, there is a high platform, about 1 m high above the floor, with indentations on its eastern edge. It is associated with two lower platforms projecting in front of it. Secondly, to the east, three steps are preserved. According to the traces that are still visible, a fourth step can be safely restored. These steps stop at slightly less than 2 m from the east wall. Thirdly, between these steps and the east wall, only one step is visible, which does not

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

192

Bunnens

form the exact continuation of the four others, and traces of plaster help to restore a second step behind it. A strange feature is the presence of a thick block of mudbricks between the north wall and the steps and platforms. It may have a structural function because the back wall does not seem to have been very strong. It probably broke up at its eastern extremity when the temple was destroyed, which would be an indication of its fragility. The mudbricks might therefore be a way to strengthen it, but there is another possible explanation. The arrangement of the installations set against this solid mass of bricks may give a clue as to its function. With three platforms to the west and two steps to the east, this arrangement outlines the prominent position of the four steps that are in the middle of the wall. The more so as, behind the four steps, faint traces of what may have been another platform are visible. We can therefore hypothesize that the platform extended into a niche within the mudbricks and conjecture that there was a large niche more or less in the middle of the back wall of the temple, which would have been the focus of the attention of the worshipers entering the cella (fig. 8). Existing remains allow for a niche that would have been about 2.80 m wide and perhaps 1 m deep. The niche would not have belonged to the original plan of the building. It would have been added in the last phase of occupation and the function of the bricks would have been to enlarge the back wall in such a way that a large niche could be accomodated. The installations as well as the walls and the floor are coated with a thick layer of white plaster.

Finds A large quantity of pottery, including many complete vessels, was retrieved from the excavated part of the temple (Baccarin, this volume). We can observe that, although only a strip about 1 m/1.50 m wide was excavated down to the floor level along the benches and platforms, it is probably also the place where most of the pottery concentrated, because previous excavations in trench S10 along the east wall failed to uncover as high a density of pottery. This would suggest that some kind of activity was performed, or that offerings were brought in, in this particular place. The special significance of this place would be confirmed by the discovery of a large jar with incised and applied decoration at the place marked A on the plan (fig. 3) (Baccarin, this volume). Mixed with the sherds of the jar are two remarkable finds. One is an eye made of a combination of wood and stone (fig. 5a). The other find is a piece of charred wood to which a pair of eyes was still adhering (fig. 5b). This must have belonged to a sculpture showing a fullface figure (Roobaert, forthcoming). The wood was unfortunately too damaged to allow a shape to be recognized.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A 3rd Millennium Temple at Tell Ahmar (Syria)

193

Fig. 5. Eye from a lost sculpture (a); pieces of charred wood with two eyes still attached to them (b); domino-like clay object (c).

A reasonable conclusion is to assume that the decorated jar was placed on a piece of furniture which included wooden decoration. A kind of domino has also been found among the sherds of the decorated jar (fig. 5c). It is made of clay, whereas other examples of similar objects, notably from a temple in neighbouring Qaraquzaq, are in stone (Page del Pozo & García Cano in Matilla Séiquer et al. 2012: 78–83, fig. 18–27). Another eye appeared in front of the central component of the platforms, and it quickly became clear that not just the eye but the whole profile of a wooden sculpture of a bull’s head was lying on the floor (Roobaert, forthcoming) (fig. 6). This is probably a fragment of a high relief wooden panel that may have been part of another piece of furniture or of a decorative panel applied to a wall. Although considerably damaged, the sculpture still displays the marks of a delicate craftsmanship. The fragment has been consolidated by the team of the conservation laboratory of the Damascus Museum and it turned out that only the surface of the sculpture was preserved, probably because, as in the case of the beams, it was the only part of the object to have been carbonized. The rest of the sculpture underwent the normal decaying process. The exceptional quality of the objects and pottery retrieved from this building can be taken as a confirmation of its identification as a temple.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

194

Bunnens

Fig. 6. Wooden sculpture of a bull’s head photographed in situ. The elongated white spot below the sculpture is a small ceramic cup.

Tentative Reconstruction of the Temple’s Layout It is not easy to find the exact place of the Tell Ahmar temple in the sacred architecture of 3rd millennium Syria (on which see Werner 1994; Di Michele 2009; Pinnock 2013; Matthiae 2014; Miglus 2014: 555–557; Werner 2014). It does not belong to the multi-chamber buildings as, for instance, temples at Mari (Margueron 2004), Tell Beidar (Lebeau 2006) and most of the temples in southern Mesopotamia (Heinrich 1982; Tunca 1984; Miglus 2014: 530–539). Monocellular temples seem to be more at home in the valley of the Syrian Euphrates and in northern Syria, especially at Ebla. On the other hand, as the cultic installations concentrate on one of the shorter walls, the Tell Ahmar temple can be safely considered as a temple of the long room type, as opposed to the broad room temples in which the focal point is on one of the longer walls.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A 3rd Millennium Temple at Tell Ahmar (Syria)

195

Fig. 7. In antis temple excavated at Qaraquzaq: reconstruction of its back end (a) (after del Olmo Lete et al. 2001: Fig. 10) and plan of the excavated remains (b) (after Valdés Pereiro 1999: Fig. 2a).

A third distinction in temple architecture is usually made according to the way the access to the cella is organized: along a straight or a bent axis. Unfortunately, the destruction of the southern part of the temple, where the entrance must have been, precludes any firm decision on this point. A fourth series of considerations concerns the cultic installations. Complex installations are found in southern Mesopotamia more often than in Syria (Tunca 1984: 148–159), but, in northern Mesopotamia, something similar, to some extent, can be found in the Tell Beidar temples (Lebeau 2006). These temples include a platform in the left corner of the cella and a bench in front, not of a niche, but of a succession of buttresses and recesses. In the right corner, however, there is nothing. The comparison remains superficial and the overall conception of the multi-chamber building is entirely different. A closer, but much less complex, parallel is found in southern Mesopotamia. However it is in a context that is older by hundreds of years. A somewhat similar arrangement has been found in the second of the Sin temples at Khafajah, which is dated to the late Uruk period (Delougaz/Lloyd 1942: 14–18). The installations against one of the shorter walls consist of a platform in the left corner, a higher platform in the middle with steps in front of it and a niche behind. But no special feature is visible in the right corner.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

196

Bunnens

Fig. 8. Reconstruction of the Tell Ahmar temple: axonometric view of the back end (a) and plan of the entire building reconstructed as an in antis temple (b).

A possible, although highly hypothetical, comparison might be made with the level III-2 temple at Qaraquzaq, a few kilometres south-east of Tell Ahmar (Valdés Pereiro 1999: 119; Gallardo Carrillo in Matilla Séquier et al. 2012: 13–23). This temple was of the typically Syrian type of the in antis temples, which is already well attested in the 3rd millennium (Castel 2010; Sala 2010). Its back wall is partially destroyed, but in one corner there are still remains of two steps (fig. 7), exactly as there are two steps in the north-eastern corner of the Tell Ahmar building. The width of the Qaraquzaq temple, c. 6 m, is about the same as that of the Tell Ahmar temple. There is thus enough space to restore a much more complex system of installations than has been proposed by the excavators (Olávarri Goicoechea/Valdés Pereiro in del Olmo Lete et al. 2001: 29 and fig. 10). The main difference would concern the absence of niche at Qaraquzaq, but, as most of the north wall has been destroyed, there may have been a kind of apse to accomodate a niche, as, for instance, there was one in a temple at Halawa (Meyer in Orthmann 1989: 65 and Beilage 10). Another similarity between the Tell Ahmar temple and its Qaraquzaq counterpart is the latter’s location. It was erected on the edge of the tell and on top of a platform overlooking a system of two lower terraces. The two buildings were thus placed in an awkward but highly visible position.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A 3rd Millennium Temple at Tell Ahmar (Syria)

197

Based on this comparison, a possible reconstruction of the Tell Ahmar temple would make it an in antis structure, as many Syrian temples were, but with a sophisticated combination of benches, steps and platforms, which is probably more characteristic of southern Mesopotamia (fig. 8). The entrance to the temple would have been in the south wall. However, the position of the edifice on the edge of the tell would have made the access difficult, especially from the south. A possible solution would have been to set up a flight of stairs rising from the foot of the tell to the temple’s entrance such as, for instance, the monumental stairs leading to the temple of Tell Mozan (Pfälzner 2008; Buccellati/Kelly-Buccellati 2009; Buccellati, F. 2010; Buccellati, G. 2010). These stairs would have been carried away by the erosion of all the southern flank of the mound.

Bibliography Baccarin, C., 2016. Consumption in a Temple? An Interpretation of the Ceramic Repertoire of the Early Bronze Age Temple at Tell Ahmar (North Syria). R. A. Stucky et al., Proceedings of the 9th International Congress on the Archaeology of the Ancient Near East (9. ICAANE), Vol 3. Wiesbaden 2016, 163–175. Becker, J., et al. (eds.), 2009. Kulturlandschaft Syrien: Zentrum und Peripherie – Festschrift für Jan-Waalke Meyer. AOAT 371, Münster. Buccellati, F., 2010. The Monumental Temple Terrace at Urkesh and its Setting. Becker et al. 2009: 71–85. Buccellati, G., 2010. The Urkesh Temple Terrace: Function and Perception. Becker et al. 2009: 87–121. Buccellati, G./Kelly-Buccellati, M., 2009. The Great Temple Terrace at Urkesh and the Lions of Tish-atal. Studies on the Civilization and Culture of Nuzi and the Hurrians 18, Bethesda, 3–69. Bunnens, G., 2002–2003. Til Barsib Before the Assyrians, Annales Archéologiques Arabes Syriennes 45–46, 167–168. Castel C., 2010. The First Temples In Antis: The Sanctuary of Tell al-Rawda in the Context of 3rd Millennium Syria. Becker et al. 2009: 123–164. del Olmo Lete, G. et al. (ed.), 2001. Tell Qaraqūzāq – II: Campañas IV–VI (1992–1994). Aula Orientalis Supplementa 17, Sabadell (Barcelona). del Olmo Lete. G./Montero Fenollos, J.-L. (eds.), 1999. Archaeology of the Upper Syrian Euphrates: The Tishrin Dam Area. Sabadell (Barcelona) Delougaz, P./Lloyd, S., 1942. Pre-Sargonid Temples in the Diyala Region. OIP 58, Chicago. Di Michele, A., 2009. L’architettura sacra nella Siria dell’età del Bronzo antico. Ocnus 17, 119–132. Heinrich, H., 1982. Die Tempel und Heiligtümer im alten Mesopotamien: Typologie, Morphologie und Geschichte. Denkmäler antiker Architektur 14, Berlin.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

198

Bunnens

Lebeau, M., 2006. Les temples de Tell Beydar et leur environnement immédiat à l’époque Early Jezirah IIIb. P. Butterlin et al. (ed.), Les espaces syro-mésopotamiens: Dimensions de l’expérience humaine au Proche-Orient ancien (volume d’hommage offert à Jean-Claude Margueron), Subartu XVII, 101–140. Margueron, J.-C., 2004. Mari, métropole de l’Euphrate au IIIe et au début du IIe millénaire av. J.-C. Paris. Matilla Séiquier, G., et al., 2012. Tell Qara Qūzāq – III: Campañas VII–XI (1995–1999). Aula Orientalis Supplementa 29, Sabadell (Barcelona). Matthiae P., 2014. Temples et palais d’Ébla protosyrienne et le problème de l’unité architecturale de la Syrie au Dynastique archaïque final. Mari, ni Est, ni Ouest, Syria supplément II, Beirut, 483–515. Miglus, P. A., 2014. Tempel. B. I. Archäologisch. In Mesopotamien, Levante, Anatolien und Iran. Reallexikon der Assyriologie 13/7–8, 530–576. Orthmann, W. (ed.), 1989. Halawa 1980-1986: Vorläufiger Bericht über die 4.–9. Grabungskampagne. Saarbrücker Beiträge zur Altertumskunde 52, Bonn. Pfälzner, P., 2008. Das Tempeloval von Urkeš: Betrachtungen zur Typologie und Entwicklungsgeschichte der mesopotamischen Ziqqurrat im 3. Jt. v. Chr. Zeitschrift für Orient-Archäologie 1, 396–433 Pinnock, F., 2013. Syrian and North Mesopotamian Temples in the Early Bronze Age. K. Kaniuth et al. (ed.), Tempel im Alten Orient. Colloquien der Deutschen Orient-Gesellschaft 7. Wiesbaden, 385– 405. Roobaert, A., forthcoming. Fragments de sculptures en bois retrouvés dans un temple du IIIe millénaire av. J.-C. à Tell Ahmar (Syrie). Roobaert, A./Bunnens, G., 1999. Excavations at Tell Ahmar – Til Barsib. Del Olmo Lete/Montero Fenollos 1999: 165–166. Sala, M., 2010. La tipologia del tempio in antis nell’architettura sacra della Siria e del Levante nel III millennio a.C.: da Tell Chuera ad al-Rawda. R. Dolce (ed.), Quale Oriente? Omaggio a un Maestro: Studi di arte e archeologia del Vicino Oriente in memoria di Anton Moortgat a 30 anni dalla sua scomparsa, Palermo, 53–75. Tunca, Ö., 1984. L’architecture religieuse protodynastique en Mésopotamie. Akkadica Supplementum II, Leuven. Valdés Pereiro C., 1999. Tell Qara Quzaq: A Summary of the First Results. del Olmo Lete/Montero Fenollos 1999: 117–127. Werner, P., 1994. Die Entwicklung der Sakralarchitektur in Nordsyrien und Südostkleinasien vom Neolithikum bis in das 1. Jt. v. Chr. Münchener vorderasiatische Studien XV, Munich/Vienna. –

2014. Tempel. B. I. Archäologisch. In Syrien. Reallexikon der Assyriologie 13/7–8, 576–578.

Guy Bunnens, Liège University.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 199–215

Marta D’Andrea

New Data from Old Excavations: Preliminary Study of the EB IVB Pottery from Area H at Tell Mardikh/Ebla, Syria The paper focuses on the Early Bronze IVB (ca. 2300–2000 BC) pottery from Tell Mardikh/ Ebla, retrieved in the 1970–1971 excavations and recently subjected to re-examination and new analyses. Preliminary results of this study are presented and discussed in a regional perspective, comparing Ebla’s EB IVB dataset with that from other sites in Northwestern Inland Syria.

1. EB IVB at Ebla: Summary of the Archaeological Evidence Early Bronze IVB (EB IVB, ca. 2300–2000 BC1) at Tell Mardikh, ancient Ebla (fig. 1), in Northwestern Inland Syria, corresponds to Phase Mardikh IIB2. This is the phase following the destruction of Royal Palace G by the Akkadians and witnessing, at a local level, a re-organisation of the settlement layout of Ebla and, at a regional level, to profound changes in the socio-political balance of Northwestern Inland Syria. In fact, after the collapse of the Early Bronze IVA (EB IVA, ca. 2450–2300 BC) polity, a new settlement pattern emerged in the region, testified by the spread of smaller sites, characterised by a homogeneous pottery horizon (see Ascalone/D’Andrea 2013; with relevant bibliography; see in particular pp. 224–226).

1

All the absolute dates in this article are conventional and follow the Middle Chronology and the absolute periodization for Northwestern Inland Syria proposed within the ARCANE Project for the Northern Levant (see the chronological table available on the ARCANE website: http://www. arcane.uni-tuebingen.de/EA-EM-EL_phasing_v5-4-6.pdf).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

200

D’Andrea

Fig. 1. Topographical map of Ebla with the areas where EB IVB (ca. 2300–2000 BC) materials and structures were uncovered and Area H highlighted (© MAIS).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Data from Old Excavations: … EB IVB Pottery from Area H at Tell Mardikh/Ebla, Syria

201

During this period, Ebla probably witnessed a first phase of crisis, after the substantial destruction of the EB IVA town, followed by progressive recovery and regeneration in the subsequent phases. The presence of temples in several areas (Areas D2, P, N, HH3), of a Royal Palace in the Lower Town north (the Archaic Palace; Matthiae 2010: 197, 396–399, figs. 99, 215; 2013), and the possible re-use of the EB IVA mud-brick fortifications (Dolce 2002: 19–20; Pinnock 2004: 94, fig. 4), testify the urban nature of the settlement during these phases.4 The recovery of inter-regional exchanges, possibly during a late EB IVB phase, is suggested by the mention of messengers and merchants from Ebla, bringing precious wooden furniture and textiles to the Ur III court, in some cuneiform texts, found at Drehem and dating from the reigns of Shulgi (2094–2047 BC), Amar Sin (2046–2038 BC), and Shu Sin (2037–2029 BC).5

2. Ebla’s EB IVB Pottery as a Research Focus Until 2004, EB IVB pottery was retrieved at Ebla mainly in secondary contexts in several areas (Areas AA, DD, FF, H, V, and Z) and in limited soundings (Areas D, E, N, P and P North, T; see Matthiae 1995: 128–129, 132; Dolce 2002: 13–17, 19–23). Stratified materials in contexts dating from a single EB IVB phase were found in the Archaic Palace, in Area P North (Matthiae 2013: 248, pl. 74), and in the dwellings in Area T6) (fig. 1), but no long EB IVB stratigraphic sequences had been excavated until then. In 2004–2007 a long EB IVB sequence was excavated in Area HH, in the Lower Town south-east (Matthiae 2007: 493– 512, see also the chart on p. 488, fig. 6; fig. 2). Thus, P. Matthiae proposed to identify three main EB IVB stratigraphic phases in Area HH, EB IVB1 (ca. 2300–2000 BC), EB IVB2 (ca. 2200–2100 BC), and EB IVB3 (ca. 2100–2025 BC) and a further, very late and very 2 3 4 5

6

See Dolce 2002: 15, 17, fig. 2; Matthiae 2009: 773–777; 2010: 395–396, fig. 214; Pinnock 2004: 92, fig. 2, with bibliography concerning the soundings in the temple’s cella, probed in the late 1960s. See Dolce 2002: 17, in particular footnotes 28, 30; Marchetti/Nigro 1997: 3, fig. 2; Pinnock 2004: 92, fig. 3; Matthiae 2010: 198–199, 396–399, 390–392, figs. 100, 102, 212; 2007: 495–507. On the layout of the EB IVB town of Ebla, see Matthiae 2010: 195–204; Pinnock 2009: 69–72. Considerations on the EB IVB settlement were put forward also by R. Dolce (1999; 2002; 2009). Owen 1992: 117–122. A different synchronism between the Ur III period in Northern and Southern Mesopotamia and Ebla was proposed by C. Schmidt (2012; 2014), who paralleled Phase Mardikh IIIA, i.e. the MBA, to the Ur III period, based on comparisons between «Ur III carinated bowls» and the MBA carinated bowls found at Ebla. This synchronism is debatable and, in general, a correlation between distant areas based on pottery types can be tricky. In a recent study, based on the thorough re-examination of the stratified materials from the latest EB IVB phase at Ebla, we sought to demonstrate, on a sound stratigraphic basis, that, if such parallels exist, we must search for them in some very late EB IVB pottery types at Ebla (D’Andrea i. p.). Matthiae 1993: 615–621, 634–637, figs. 3, 12–13; Dolce 2002: 17, fig. 5; Pinnock 2009: 69, figs. 2–7.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

202

D’Andrea

Fig. 2. Ebla, the EB IVB (ca. 2300–2000 BC) sequence in Area HH, in the Lower Town south-east (re-elaboration according to Matthiae 2007; forthcoming; © MAIS).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Data from Old Excavations: … EB IVB Pottery from Area H at Tell Mardikh/Ebla, Syria

203

short phase, EB IVB4 (ca. 2025–2000 BC). This consists of «squatter’s» occupation and re-use of the structures in Area HH, after the destruction of the EB IVB town, but before the end of the EB IVB cultural horizon at the site (Matthiae 2007: 493–512; 2010: 205–206). Despite a general continuity, each phase is characterised by a discrete pottery repertoire and by the appearance of new types as ceramic markers for relative chronology, differentiating the phases from one another (Matthiae 2007: 509–512, figs. 25–27; 2010: 200–203)7. This provided, for the first time in the research at Ebla, the backbone of the periodization of EB IVB at the site and a reference point to ascribe the EB IVB materials from the other areas to a phase within the site’s periodization of EB IVB (fig. 2). Thus, it is now clear that most of the stratified evidence from the other areas at Ebla that yielded EB IVB materials and structures (see examples in fig. 3) correlate well with the late EB IVB phase identified in Area HH, i.e., EB IVB3 (ca. 2100–2025 BC). In addition, the EB IVB sequence in Area HH provided a means for correlation with the assemblages of other multi-phase EB IVB sites in Northwestern Inland Syria. In fact, at a regional level, this phase is largely comparable with Hama J1 (Fugmann 1958: figs. 103, 106) and Tell Qarqur Stratum 12 (Dornemann 2008: 84, figs. 5:18–32, 6, 7:1–11), i.e., the latest EB IVB phases at these sites. Further close parallels for Ebla’s EB IVB3 occur at Tell Afis Area E3, in Phase IV (Felli/Merluzzi 2008, fig. 6:1–13) and at Tell Tuqan (Baffi/Peyronel 2013: figs. 9.13–9.16). We will elaborate later on the chronological relevance of these data within the present study (see § 4). Given the availability of a large amount of EB IVB pottery from Ebla in the National Museum of Oriental Art in Rome (MNAO)8, we decided to analyse this ceramic corpus in the light of the renewed research on EB IVB at Ebla.9 Thus, the EB IVB pottery retrieved in Area H in the 1970–1971 excavations, directed by P. Matthiae on behalf of the Sapienza University of Rome, were recently subjected to re-examination and new analyses within the Ebla Chora Project.10 The field reports kept in the archives of the Italian Archaeological Expedition to Ebla (MAIS) at the Sapienza University of Rome, were re-examined and a first-hand analysis of the substantial assemblages of EB IVB pottery stored at the MNAO was undertaken.

The author is in charge of the study of the EB IVB pottery from Ebla and a thorough research is in progress; for a presentation of preliminary results see D’Andrea i. p. 8 The author wishes to thank warmly Paolo Matthiae and Frances Pinnock for the possibility to study and publish the materials, and Paola D’Amore, for the kind and prompt support provided to this research during the selection of the samples at the Department of Near and Middle East at the MNAO, Rome. 9 Some EB IVB shards from Area H were published by Mazzoni (1985a: fig. 6:2–4, 6–7, 11, 13–20, 22–23; 1985b: fig. 1:13–16, 27–29) and a preliminary study on Ebla’s EB IVB Painted Simple Ware, including also the Area H materials, was carried out by Cristiana Pezzetta (2003). 10 The integrated, multidisciplinary project was funded by the European Research Council (FP7-IDEAS 249394) and coordinated by P. Matthiae and N. Marchetti; see Matthiae/Marchetti 2013. 7

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

204

D’Andrea

Fig. 3. Ebla, the EB IVB (ca. 2300–2000 BC) Temple D3 on the Acropolis (no. 1) with a selection of EB IVB materials (nos. 2–6) and a selection of EB IVB pottery from Area T (nos 7–12; see Matthiae 1993: fig. 12:11, 13; Pinnock 2009: figs. 4:1–2, 9, 6:4) comparable to late EB IVB materials from Area H (© MAIS).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Data from Old Excavations: … EB IVB Pottery from Area H at Tell Mardikh/Ebla, Syria

a

b Fig. 4. Ebla, Area H: (a) ashy layers in the stratigraphic section north; (b) mud-brick structures uncovered in the sounding; EB IVB (ca. 2300–2000 BC) (© MAIS).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

205

206

D’Andrea

Fig. 5. Ebla, Area H: representative selection of the EB IVB (ca. 2300–2000 BC) pottery repertoire: Painted Simple Ware and Simple Ware (© MAIS).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Data from Old Excavations: … EB IVB Pottery from Area H at Tell Mardikh/Ebla, Syria

Fig. 6. Ebla, Area H: representative selection of the EB IVB (ca. 2300–2000 BC) pottery repertoire: Cooking Ware (nos. 1–5) and Preservation Ware (nos. 6–8) (© MAIS).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

207

208

D’Andrea

Fig. 7. Ebla, Area H: selection of EB IVB (ca. 2300–2000 BC) Painted Simple Ware (© MAIS).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Data from Old Excavations: … EB IVB Pottery from Area H at Tell Mardikh/Ebla, Syria

209

3. Ebla’s Area H: Topography and Stratigraphy Area H is located across the central sector of the Western Rampart (fig. 1), where excavations were carried out in Squares BdVI8i+ii+iv and BeVI8i+iii+iv. In the soundings massive grey ashy fills (Matthiae 1995: 129, 132, fig. 74; 2010: 205; here fig. 4a), containing large amounts of EB IVB shards, were excavated. These were interpreted as the archaeological layers coming from the destruction of the EB IVB town, accumulated on the outer perimeter of the site when the Middle Bronze Age (MBA, ca. 2000–1600 BC) fortification system was built at the very beginning of the 2nd millennium BC. In addition, some structures were uncovered during the excavations. These consist of in situ reddish mud-bricks (fig. 4b) that might belong to the remains of the EB IVA (ca. 2450–2300 BC) city wall, stretches of which were identified in some spots underneath the rampart (e.g., in Areas Z and AA; Matthiae 2010: 67, fig. 26).

4. Ebla’s Area H Ceramic Corpus: Typology, Style, and Chronology The excavation of the ashy fills in Area H yielded more than 800 EB IVB shards, besides some MBA and EB IVA ceramic materials. Although the materials come from secondary contexts, they constitute a considerable corpus of pottery from Ebla’s post-palatial phase, which illustrates a wide typological and functional range of EB IVB vessel shapes and several of the most spread EB IVB wares typical of Northwestern Inland Syria. The Area H assemblages include the complete typological and stylistic repertoire of Painted Simple Ware (fig. 5.1–3, 5–8, 7). Several types of bell-shaped goblets (fig. 7.1, 3, 5–6, 8–14), either with incurved walls (fig. 5.1) or with cylindrical body (fig. 5.2–3), bowls with moulded rims, bowls with thickened rims (figs. 5.5, 7.2), bowls with flaring walls and simple rims (fig. 5.6), and small necked jars (figs. 5.7–8; 7.7) are represented in the assemblages. Some painted shards feature decorative patterns different from the standard, such as rows of triangles filled with a net-pattern and stylised animal figures (figs. 5.8; 7.7, 10–14; see also D’Andrea 2014: 187–191, fig. 2a–c; D’Andrea 2015). In the Area HH sequence, these motifs appear from the central EB IVB phase (EB IVB2) and increase in the late phase (EB IVB3), although still forming a minority of the percentage of the painted pottery. They have several parallels in the ‘Amuq sequence (Second Mixed Range; Braidwood/Braidwood 1960: figs. 343:9, 344:25, 27, 30, 365:10, pls 44:1, 10, 89:4), at Tell Mastuma (Square 15; Tsuneki 2009: fig. 3.21:15), and at Hama (Level J1; Fugmann 1958: fig. 106:4B825).11

11 On these materials, see D’Andrea 2015; i. p.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

210

D’Andrea

Fig. 8. Ebla, Area H: EB IVB (ca. 2300–2000 BC) goblets with band-combing (nos. 1–4), bowls with moulded rims (nos. 5–7), carinated bowls (nos. 8–10), typical of the late EB IVB pottery repertoire of Ebla, and Smeared Wash Ware (no. 11) (© MAIS).

In the Area H assemblages, Smeared Wash Ware is well represented as well, by a fairly large amount of jugs (fig. 8.11). Simple Ware includes unpainted goblets, either plain (fig. 5.9) or with band-combed decoration (figs. 5.10–12; 8.1–4), bowls with folded-over or slightly thickened outward rim (figs. 5.13), bowls with vertical (fig. 5.14; 8.7) or outer-moulded rim (fig. 5.15–16), groovedrim bowls (figs. 5.17; 8.5–6), carinated bowls (figs. 5.18; 8.8–10), and several types of small jars (fig. 5.19–20). Large bowls/basins with scrabbled surface, with or without ledge handles (fig. 6.1–4), are the most representative EB IVB vessel type belonging to Cooking Ware attested in the Area H pottery repertoire, but fragments of globular cooking pots with flaring

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Data from Old Excavations: … EB IVB Pottery from Area H at Tell Mardikh/Ebla, Syria

211

rim (fig. 6.5) are also present. Finally, jars with thickened outside rim (fig. 6.6–8) are the most frequently found vessel shape within the Preservation Ware repertoire from Area H. The ceramic corpus from Area H represents the typological and stylistic EB IVB pottery repertoire of Northwestern Inland Syria and finds parallels at other sites in this region (e.g., Tell Afis, Tell Tuqan, Tell Qarqur, and Tell Mastuma; see Mazzoni 2002: 78–79, pls. XLIV– XLIX; Ascalone/D’Andrea 2013, with relevant bibliography12). Furthermore, although the Area H ceramic materials come from secondary contexts, they can be correlated to the long stratigraphic sequence excavated in Area HH, in the Lower Town south-east, but also with the ceramic materials excavated in Area T, in the Lower Town north-west, which provided late EB IVB materials (Pinnock 2009: figs. 4–7; fig. 3.7–12). Thus, the majority of the Painted Simple Ware, Simple Ware, and Cooking Ware vessels illustrate the «standard» EB IVB regional repertoire, but many vessel types find precise parallels in those from Area HH dating from EB IVB3 and in the contemporaneous ceramic assemblages from Area T (see § 1; Pinnock 2009: figs. 4–7; fig. 3.7–12). In particular, these are unpainted goblets plain and with band-combing (figs. 5.9–12; 8.1–4), grooved-rim bowls (figs. 5.17; 8.5–6), carinated bowls (figs. 5.18; 8.9–10), and storage jars with thickened outside rim (fig. 6.8). These vessel types are late in the EB IVB sequence not only at Ebla, but characterise the latest EB IVB pottery horizon of other sites that yielded long EB IVB sequences as well, such as Tell Qarqur (Stratum 12; Dornemann 2008: figs. 5:25–26, 28–32, 6:2) and Hama (Level J1; Fugmann 1958: figs. 103, 106). These data have a chronological relevance, albeit only in relative terms, in the study of EB IVB at Ebla. In fact, it has been maintained that the ashy layers rich in EB IVB pottery accumulated in the MBA ramparts, such as those excavated in Area H, originate from the debris of the destroyed EB IVB town (Matthiae 1995: 132; 2010: 205). Thus, the presence of late EB IVB pottery types in the Area H assemblages (figs. 5.9–12, 14–19; 8.1–10) supports a relative date of the destruction late into the EB IVB relative sequence of the site.

5. Perspectives of Future Research: Investigating and Reconstructing Pottery Technology and Production at Ebla during EB IVB The Area H pottery was also included into a large archaeometric study of the EB IVB pottery production at Ebla. Comparison with the Area HH sequence allowed us to isolate late EB IVB vessel types from the long-lived ones, and to investigate this topic in a chronological perspective. 12 See also Cooper 2007; Sala 2010; Welton/Cooper 2013; Welton 2014. On connections between the

Northern Levant and the neighbouring regions, see D’Andrea/Vacca 2015.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

212

D’Andrea

A thorough study on pottery technology based on a selection of EB IVB shards from Ebla’s Area H has been undertaken and the results are now being processed and elaborated.13 Chemical and petrographic characterization of the Area H ceramics were defined through thin sections, XRD, and micro-Raman spectroscopy and fabric groups, firing temperatures, and the nature of pigments used in paints and surface treatments were determined. Parallel to this study, also the analysis of manufacturing techniques at Ebla during EB IVB is currently in progress. Both studies will be subject of other publications once the results will be fully elaborated and interpreted. Given the increasingly availability of comparable data from other sites in the region14, such a research will allow a comprehensive reconstruction of pottery production at Ebla during EB IVB and will provide a means of correlation to other sites in Northwestern Inland Syria, aiming at the investigation of regional patterns of pottery production, distribution, and consumption during the late EBA.

13 The study is carried out through the cooperation of the MAIS (Luca Peyronel, Marta D’Andrea and

Agnese Vacca) with the Department of Earth Sciences of Sapienza University of Rome (Caterina De Vito, Laura Medeghini, and Silvano Mignardi). See also the analysis on some EB IVB pottery from Ebla (Areas H and HH) in Ballirano et al. 2014. 14 As an example, see a recent and innovative study, based on the evidence from Tell Mishrifeh/Qatna, in Iamoni 2014.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Data from Old Excavations: … EB IVB Pottery from Area H at Tell Mardikh/Ebla, Syria

213

Bibliography Ascalone, E./D’Andrea, M., 2013. Assembling the Evidence. Excavated Sites Dating from the Early Bronze Age in and around the Chora of Ebla. Matthiae/Marchetti 2013: 215–237. Baffi, F./Peyronel, L., 2013. Trends in Village Life. The Early Bronze Age Phases at Tell Tuqan. Matthiae/ Marchetti 2013: 195–214 Ballirano, P. et al., 2014. A combined use of optical microscopy, X-ray powder diffraction and microRaman spectroscopy for the characterization of ancient ceramic from Ebla. Ceramics International 40, 16409–16419. Braidwood R.J./Braidwood, L., 1960. Excavations on the Plain of Antioch, I: The Early Assemblages, Phases A–J. Oriental Institute Publications 61. Chicago. Cooper, L., 2007. Exploring the Heartland of the Early Bronze Age ‘Caliciform Culture’. Journal of the Canadian Society for Mesopotamian Studies 2, 43–50. D’Andrea, M., 2014. La Black Wheelmade Ware. Originalità e modelli stilistici, tipologici e tecnologici dalla Siria e dal Levante settentrionale in una peculiare produzione dipinta sud-levantina del tardo III millennio a.C. Rendiconti Morali dell’Accademia Nazionale dei Lincei s. IX (2013), 181–220. –

2015. Preliminary Notes on Some EB IVB Painted Simple Ware Sherds from Ebla. Studia Eblaitica 1, 205–209.

–

i. p. The Early Bronze IVB Pottery of Ebla: Stratigraphy, Chronology, Typology, and Style. Remarks from a Work-in-Progress. P. Matthiae/F. Pinnock (eds.), Ebla and Beyond. The Archaeology of the Ancient Near East after Fifty Years of Discoveries at Tell Mardikh. Studies in Ancient Near Eastern Records, i. p.

D’Andrea, M./Vacca, A., 2015. The Northern and Southern Levant during the Late Early Bronze Age: a Reappraisal of the “Syrian Connection”. Studia Eblaitica 1, 43–73. Dolce, R., 1999. The «Second Ebla»: A View on the EB IVB City. ISIMU II, 293–304. –

2002. Ebla after the ‘Fall’ – Some Preliminary Considerations on the EB IVB City. Damaszener Mitteilungen 13 (FS Abu Assaf), 11–28.

–

2009. The Archaeology of a Long Lasting Power: The Ebla Paradigm. A synthesis. Studi Micenei ed Egeo-Anatolici 51, 251–278.

Dornemann, R. H., 2008. Evidence from Two Transitional Periods, Early Bronze IV and Iron I, at Tell Qarqur. H. Kühne et al. (eds), Proceedings of the 4th International Congress of the Archaeology of the Ancient Near East: 29 March – 3 April, Freie Universität Berlin, Vol. 2. Wiesbaden, 81–96. Felli, C./Merluzzi, E., 2008. EB–MB Afis: A Single Cultural Tradition Between Two Phases. H. Kühne et al. (eds.), Proceedings of the 4th International Congress of the Archaeology of the Ancient Near East: 29 March – 3 April, Freie Universität Berlin, Vol. 2. Wiesbaden, 97–110.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

214

D’Andrea

Fugmann, E., 1958. Hama: Fouilles et recherches de la Fondation Carlsberg 1931–38, II: L’architecture des périodes pré-hellénistiques. Nationalmuseets skrifter. Større beretninger 4. Copenhague. Iamoni, M., 2014. Transitions in ceramics, a critical account and suggested approach: case-study through comparison of the EBA–MBA and MBA–LB horizons at Qatna. Levant 46.1, 4–26. Marchetti, N./Nigro, L., 1997. Cultic Activities in the Sacred Area of Ishtar at Ebla During the Old Syrian Period: The Favissae F.5327 and F.5328. Journal of Cuneiform Studies 49, 1–44. Matthiae, P., 1993. L’Aire Sacrée d’Ishtar à Ebla: Résultats des Fouilles de 1990-1992. Comptes Rendus de l’Académie des Inscriptions et Belles Lettres 1993, 613–662. –

1995. Ebla: Un impero ritrovato. Dai primi scavi alle ultime scoperte (3rd Edition). Einaudi Tascabili 273. Torino.

–

2007. Nouvelles fouilles à Ébla en 2006. Le Temple du Rocher et ses successeurs protosyriens et

–

2009. Temples et Reines de l’Ebla Protosyrienne: Résultats de Fouilles à Tell Mardikh en 2007 et

–

2010 Ebla la città del trono. Archeologia e storia. Piccola Biblioteca Einaudi 492. Torino.

–

2013. The Archaic Palace at Ebla: A Royal Building between Early Bronze IVB and Middle

paléosyriens. Comptes Rendus de l’Académie des Inscriptions et Belles Lettres 2007, 481–525. 2008. Comptes Rendus de l’Académie des Inscriptions et Belles Lettres 2009, 747–791.

Bronze Age I. P. Matthiae, Studies in the History and Archaeology of Ebla 1980–2010 (ed. by F. Pinnock). Wiesbaden, 243–257. Matthiae, P. /Marchetti, N. (eds.), 2013. Ebla and its Landscape. Early State Organization in the Near East. Walnut Creek, CA, Mazzoni, S., 1985a. Frontières céramiques à l’Haute Euphrate au Bronze ancien IV, in Actes du Colloque International du CNRS 620 ‘A Propos d’un Cinquantenaire: Mari, Bilan et Perspectives’ (Strasbourg; 29–30 juin–1er Juillet 1983). Mari Annales de Recherches Interdisciplinaires 4, 561–571. –

1985b. Elements of the Ceramic Culture of Early Syrian Ebla in Comparison with Syro-Palestinian

–

2002. The Ancient Bronze Age Pottery Tradition. M. al-Maqdissi et al. (eds.), Northwestern

EB IV. Bulletin of the American Schools of Oriental Research 257, 1–18. Central Syria, in Céramique de l’Âge du Bronze en Syrie, I: La Syrie du Sud et la Vallée de l’Oronte. Bibliothèque Archéologique et Historique 161. Beyrouth, 69–96. Owen, D. I., 1992. Sumerian Sources from the Ur III Period. M.W. Chavalas /J.L. Hayes, New Horizons in the Study of Ancient Syria. Bibliotheca Mesopotamica 25. Malibu, 108–175. Pezzetta, C., 2003. Aspetti morfologici, tipologici e culturali della ceramica del Bronzo Antico IV B di Ebla: studio preliminare. G. Regalzi (ed.), Le discipline orientalistiche come scienze storiche. Atti del 1° Convegno «Orientalisti» (Roma, 6–7 Dicembre 2001), 2003 (2nd electronic edition). Accessed online at: http://purl.org/net/orientalisti/atti2001.htm. Pinnock, F., 2004. Change and Continuity of Art in Syria Viewed from Ebla. J.-W. Mayer/W. Sommerfeld (eds.), 2000 v. Chr. – Politische, wirtschaftliche und kulturelle Entwicklung im Zeichen einer Jahrtausendwende. 3. Internationales Colloquium des Deutchen Orient-Gesellschaft 4–7 April 2000 in Frankfurt/Main und Marburg/Lahn. Saarbrücken, 87–118.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Data from Old Excavations: … EB IVB Pottery from Area H at Tell Mardikh/Ebla, Syria –

215

2009. EB IVB–MB I in Northern Syria: Crisis and Change of a Mature Urban Civilisation. P. J. Parr (ed.), The Levant in Transition. Proceedings of a Conference held at the British Museum on 20–21 April 2004. Palestine Exploration Fund Annual IX. Leeds, 69–79.

Sala, M., 2010. An Early Bronze IVB Pottery Repertoire from Favissae P.9717 and P.9719 in the Temple of the Rock at Tell Mardikh/Ancient Ebla. Levant 44/1, 51–81. Schmidt, C., 2012. The Late Third and Early Second Millennium BC Pottery Tradition in the Syrian Jazirah and Beyond. N. Laneri et al. (eds.), Looking North. Studien zur Urbanisierung Mesopotamiens, Serie D, Supplementa, Band 1. Wiesbaden, 163–174 –

2014. Late 3rd Millennium “Ur III” Carinated Bowls. M. Lebeau (ed.), Ceramics. ARCANE Interregional 1. Turnhout, 365–371.

Tsuneki, A., 2009. Neolithic and Early Bronze Age Layers in Square 15Gc. T. Iwasaki et al. (eds.), Tell Mastuma. An Iron Age Settlement in Northwest Syria. Memoirs of the Ancient Orient Museum III. Tokyo, 69–88. Welton, L., 2014. Revisiting the Amuq sequence: a preliminary investigation of the EB IVB ceramic assemblage from Tell Tayinat. Levant 46/3, 339–370. Welton, L./Cooper, L., 2013. “Caliciform Ware”. M. Lebeau (ed.), Arcane Interregional. Ceramics. ARCANE Interregional I. Turnhout, 293–322.

Marta D’Andrea, Sapienza University of Rome, MAIS – Italian Archaeological Expedition to Ebla, [email protected].

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 217–229

Stephanie Döpper

The LBA pottery of Area BU in the Royal Palace of Qatna, Syria In this paper1, the pottery assemblage of Area BU in the Royal Palace of Qatna is presented. Additionally, correspondence analyses based on published LBA pottery in Syria and the Levant is used in order to find out how the pottery from Area BU fits to existing pottery traditions.

1. The Pottery Assemblage of Area BU Within the scope of my PhD, four different Late Bronze Age pottery assemblages from the Royal Palace of Qatna, which was most likely destroyed by Šuppiluliuma around 1340 BC, are studied. One assemblage comes from Room G in the central part of the palace to the north of the large Hall C, the central court of the palace (fig. 1; Novák/Pfälzner 2000: 267, 270–271, 283–285, 288–290; Novák/Pfälzner 2001: 181–182). It consists primarily of storage jars, which were partly sunken into the floor, and hence represent for the most part an in situ inventory of the room. The second assemblage belongs to Room DK in the north-western part of the palace. The room fill consisted of ashy soil with many animal bones, pottery sherds and clay sealings, including one of the Egyptian Pharaoh Akhenaten (Ahrens et al. 2012). It is most likely a trash deposit. The third pottery assemblage, Tombeau II, comes from a rock-cut tomb beneath the Royal Palace of Qatna. It was excavated previously by du Mesnil in 1924 (du Mesnil du Buisson 1927: 19), who piled up the pottery found within in front of the tomb, which was then rediscovered by the Syrian-German Mission (Dohmann-Pfälzner/Pfälzner 2007: 154–157). The fourth assemblage, Area BU, which will be the main focus of this paper, lies in the north-western part of the Royal Palace, adjacent to the well room U. Area BU most likely represents the remains of a platform at the northern wall of the palace, comparable to the 1

This research is part of my PhD thesis, which was completed in August 2014.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

218

Döpper

Fig. 1. General plan of the Royal Palace of Qatna with the Rooms G and DK, the area BU and the Tombeau II highlighted.

mud-brick platform DP along the same wall further east (Dohmann-Pfälzner/Pfälzner 2008: 73). The area was excavated in 2002 and 2003 and again in 2006 and 2007 by the University of Tübingen under the direction of Peter Pfälzner. Below 60 to 90 cm of modern deposits, the Late Bronze Age layers were reached. They are made up of compact clay soil mixed with mud brick debris with layers of pottery sherds in between (fig. 2). Animal bones were recovered, along with occasional small finds, such as jewellery and objects of bone and stone. The pottery in the area was excavated in six artificially differentiated layers. The whole accumulation rests on top of a pisé layer, which marks the end of the excavations in 2007. Joints between pottery sherds from top to bottom within the accumulation indicate that the pottery assemblage was placed here in a single action. However, the find context of the pottery assemblage from Area BU also points out that the assemblage is, to use Hachmann’s terminology, a tertiary context (Hachmann 1982: 189). He defines material from such contexts as shifted from their original or even secondary position to a tertiary one. Therefore it is not sure whether such finds are still in their original chronological setting or not (Metzger/Barthel 1993: 28), which seems to be the case for the assemblage from Area

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The LBA pottery of Area BU in the Royal Palace of Qatna, Syria

219

Fig. 2. Pottery sherds in Area BU during excavation (Photo: G. Mirsch).

BU. As the Royal Palace of Qatna unfortunately yielded almost no undisturbed or in situ assemblages, except for the royal tombs, this is the material to work with. All in all, 6389 pottery sherds from Area BU were available for analyses. Among them, 1388 sherds had an identifiable shape, while for the remaining 5001 sherds only the ware could be determined. The pottery is predominantly wheel-made and the surface finish is in general unobtrusive; it is rarely slipped, though sometimes a self-slip can be observed. In total, 14 different wares were identified within the pottery of Area BU. Their distribution shows that a coarse mineral tempered ware, Ware 28 in Qatna terminology, clearly dominates, with 80.2 % of the whole pottery assemblage belonging to this ware. Far behind, in second position, is a semi-coarse mineral tempered ware, Ware 6, which accounts for 8.2% of the assemblage, closely followed by a highly fired, mineral tempered, greenish, crumbly ware, Ware 23, at 7.1%. All other wares are represented by less than two percent of the assemblage. The high dominance of mineral tempered pottery also occurs in the other three pottery assemblages studied and is in general very common for the Late Bronze Age pottery of Qatna.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

220

Döpper

Fig. 3. Red slipped pottery (Qatna ware 38) from Area BU (Photo: K. Wita).

One other ware has to be further highlighted, as it only occurs in this assemblage and not within the other three assemblages studied within the scope of my PhD. This is the Ware 38, of which only 27 sherds were found, making up 0.4 % of the assemblage. The ware is characterized by a reddish slip and is often burnished (fig. 3). In the southern Levant, according to Mazar (1998: 373–374) and Holladay (1993: 95; Holladay 1990: 63), a certain type of hand burnished red slipped pottery is described as a hallmark of the Iron Age. In the northern Levant in Beirut (Badre 1997: 68) and Tell Kazel (Badre et al. 1994: 272), red slipped pottery is seen by Badre as a characteristic for the very end of the Late Bronze age and the beginning of the Iron Age. Therefore, it is likely that the assemblage of Area BU is also to be placed in a similar time range. However, red slipped burnished pottery does also occur in small quantities throughout the Late Bronze Age at different sites, for example at Alalakh (pers. comm. Mara Horowitz). Less than five percent of the pottery of Area BU is decorated, mostly with a combed decoration. Some of the more refined decorated pieces include a rim fragment that is painted a reddish-brown colour with horizontal lines and a band filled with crosshatched lines (fig. 4a). Below the crosshatched lines, there is a row of volutes. A second piece (fig. 4b), which is also decorated a reddish brown colour on a light surface, again shows a row of volutes but in a different style than on the first piece. The third sherd features a geometric design structured

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The LBA pottery of Area BU in the Royal Palace of Qatna, Syria

Fig. 4. Pottery from Area BU.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

221

222

Döpper

in metopes and painted in red on a light surface (fig. 4c). The borders of the metopes are made of straight and wavy lines. Within the central metope, there is an object, which is reminiscent of the upper part of an old fashioned key, and, adjacent to the right of it, there is the beginning of a triangle filled with dots, which probably forms the lower part of a ‹sand hour› shaped decoration. This type of decoration has been found on other pottery sherds in Qatna, as well as in many other sites in the Levant during the Late Bronze Age (e.g. Salles 1980: 83 pl. 15.2–3; Saidah 2004: 40 fig. 5.4; Courtois 1969: 125 fig. 4C; Schaeffer 1938: 210 fig. 11ZB) and can therefore be called a typical motif of this period. Concerning the form, open shapes, mainly bowls, dominate the assemblage. In comparison with the assemblages of Room G, Room DK and Tombeau II, their percentage is lower, with more closed forms, like pots and jars, in Area BU than in the other assemblages. Closed forms make up to nearly 30 percent of the pottery of Area BU. The most prominent open shapes within the assemblage are conical and rounded bowls with simple rounded rims (types 01.20.00 and 02.12.00) (fig. 4d, h), conical bowls with slightly flaring rim (type 02.09.00) (fig. 4g), conical bowls with simple rims, that are slightly flatted on top (type 02.15.00) (fig. 4i), and rounded bowls with incurving rim (type 01.65.00) (fig. 4f). The most frequent closed forms are jars with external thickened rims triangular in section in a more and in a less pronounced version (types 32.10.00 and 32.43.00) (fig. 4l, n) or flaring, simple rounded rims (type 32.33.00) (fig. 4m) and pots with different kinds of outwardly-projecting rims (types 53.15.00, and 53.19.00) (fig. 4o–p). Beside these forms mentioned, many other shapes are present as well, indicating the high variability of the assemblage. A specific group of bowls that appears in the pottery assemblage of Area BU should be mentioned here, though they rarely occur in any of the other assemblages studied from the Royal Palace of Qatna. These are deep bowls, characterized by a carination and a high upright or outflared rim (fig. 4j–k). The carination can be more or less pronounced or even slightly rounded and the bases are simple flat bases or ring bases. These deep bowls are responsible for the fact that the bowls of Area BU have in general a higher capacity than for example those from Room DK. In Area BU the bowl capacity is on average around 1.5 litres per bowl whereas in Room DK it is only around 0.6 litres per bowl. Assuming that bowls from both assemblages were used for serving and eating food, the different capacities would make them suitable for different kinds of food. While flat bowls, like the ones from Room DK, are most fitting for dry food, deep bowls, like the ones from Area BU, are more suitable for containing liquid food, like stews and mash. This could be linked to different eating habits that may have changed over time, as Room DK dates to the Late Bronze Age IIA, whereas Area BU dates to the Late Bronze Age IIB (see below). Alternatively, this difference could be related to variation between social groups, as the assemblage from Room DK belongs to a period when the palace was still in use, yet the assemblage from Area BU comes instead from squatter occupations after the destruction of the palace (see below).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The LBA pottery of Area BU in the Royal Palace of Qatna, Syria

223

Comparisons for all the different pottery types mentioned can be found across the Levant and central Syria during almost all phases of the Late Bronze Age. The deep carinated bowls with high upright or out-flared rims have more comparisons in the second half of the Late Bronze Age, particularly from the 13th century BC. In general, the shapes of Area BU have close parallels to the ones found by the Italian mission at Qatna in the operation T1, dating to the 14th and 13th century BC (Morandi Bonacossi et al. 2009: 101–105). Finally, an important sherd that should also be mentioned is a fragment of a milk bowl (MSH06G-i0386) – one of the few White Slip pottery sherds found in Qatna (du Mesnil du Buisson 1935: 34, 35 fig. 3; Dohmann-Pfälzner/Pfälzner 2006: 104; Novák 2004: 309 fig. 10; Luciani 2002: 149 fig. 122, 150) – which is decorated by rather carelessly applied ladder band and vertical line patterns (fig. 4e). According to Celia Bergoffen, it belongs to the White Slip II derivative style, which is to be distinguished from the normal white slip by its coarse ware and decoration, which consists of a ladder lattice with three instead of four bands.2 In summary, the pottery assemblage of Area BU is to be assigned to the second half of the Late Bronze Age, and more specifically to the Late Bronze Age IIB, as the red slipped ware and the pottery types (especially the deep carinated bowls with high upright or outflared rim, and the milk bowl fragment) suggest. This means that the pottery assemblage belongs to a phase after the destruction of the Royal Palace of Qatna around 1340 BC. From the time after the destruction, few archaeological remains are known from the excavations in Qatna. They mostly consist of simple dwelling structures (Morandi Bonacossi 2013: 120– 121; Luciani 2002: 152–158; Al-Maqdissi 2002: 196–198; Morandi Bonacossi et al. 2009: 93–105). The assemblage probably belonged to a squatter occupation of the Royal Palace, where the people who stayed in Qatna after the destruction, tried to rebuild some parts of the palace for new homes. However, three radiocarbon dates from Area BU, which were taken by Felix Höfelmeyer from barley grains (hordeum vulgare) and olive stones (olea europea)3, date to a period between the end of the 16th and the beginning of the 15th century BC, which would be the earliest part of the Late Bronze Age. All dates are very close to each other, which gives some credibility to this date. As described above, the pottery assemblage, however, does not fit within this time period. Therefore, other explanations must be sought. In general, the percentage of pottery types that are already known from older assemblages within the Royal Palace is small, which argues against a mixed assemblage of older and younger types. The latter would be generally possible regarding the tertiary context of the material. My present idea is thus, that the pottery was brought into area BU as part of the building measures that 2 3

Bergoffen 2014: 666. I thank her very much for her sharing her knowledge about white slip with me. The archaeobotanical determination was carried out by Simone Riehl.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

224

Döpper

took place here after some parts of the walls had eroded or were damaged by the destruction of the palace. For this purpose, the pottery was intentionally collected. It was then placed in layers between the pisé that was used for the repair, perhaps with a drainage function, though this remains speculative. Pottery as building material can also be observed at other parts of the Royal Palace of Qatna, notably in the packing material for the basalt steps in the well-room U close by (Dohmann-Pfälzner/Pfälzner 2006: 81 Abb. 18, 83). The clay for the pisé parts of the repair in Area BU would, in my theory, have been taken from a selected area with older occupation that was no longer in use and thus included in the radiocarbon dating samples. For this reason, the radiocarbon dates are then older than the associated pottery.

2. Correspondence Analysis of LBA Assemblages from the Levant and Western Syria Having presented the pottery of Area BU, I would like to come to the second focus of my PhD, which explores the question how pottery assemblages from different sites can be compared with each other and if such comparisons can be the basis to form a supra-regional, relative, comparable chronology. In order to achieve this, more than 15.000 published diagnostic pottery sherds from more than 50 different sites in the Levant and Western Syria, from Tell Beit Mirsim in the South to Alalakh in the north and to the Euphrates in the East, were classified according to their shapes and entered into a database that also lists their technological features, as far as available. Based on this information, correspondence analyses were carried out in order to see similarities between whole assemblages, which is generally assumed to be an indicator for chronological proximity. A first result is that the correspondence analyses revealed few similarities between the assemblages in general, which is due to the high number of different shapes that were recorded within the pottery, with each shape attested only in small quantities in each of the assemblages. Therefore, assemblages with fewer than 10 different types and types that are attested in fewer than five different assemblages from the remaining ones had to be excluded from the correspondence analysis. With this precondition, the first step was to assign assemblages different symbols according to the date they were given in the publications. Assemblages that were not assigned any specific period of the Late Bronze Age within the literature were classified with a black dot. The resulting correspondence analysis shows that, contrary to expectation, assemblages that are assigned to the same period within the Late Bronze Age do not group together but are instead spread all over the scatter plot (fig. 5). This indicates that similarities and differences between the pottery assemblages are not primarily caused by their chronological position.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The LBA pottery of Area BU in the Royal Palace of Qatna, Syria

Fig. 5. Correspondence analysis of Late Bronze Age pottery assemblages with different symbols according to their date.

Fig. 6. Correspondence analysis of Late Bronze Age pottery assemblages with different symbols according to their region.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

225

226

Döpper

Fig. 7. Correspondence analysis of Late Bronze Age pottery assemblages with different symbols according to functional context.

In a second step, I tried to find out what factors may be responsible for the similarities and differences between the assemblages that are obviously stronger than chronological ones (fig. 6). The assemblages were then assigned different symbols according to the regions they belong to. Rectangles were given to assemblages from western Syria, diamonds to ones from the Euphrates region and triangles to the Levant. Although the transition is smooth, the different regional groups can now easily be recognised within the correspondence analysis. The triangles from the Levant are found on the left side of the scatter plot, the diamonds of the Euphrates Region in the middle and the rectangles from western Syria on the right. This signifies that regional proximity between assemblages is much more relevant for understanding their similarities than chronology. Or, in other words, pottery assemblages from geographically close sites from different time periods within the Late Bronze Age are more alike to each other than contemporaneous assemblages from geographically distant sites. In a third step, I attempted to analyse to what extent the functional contexts of the ceramics affected the result (fig. 7). Therefore, I distinguished between assemblages from tomb and settlement contexts, as this is the easiest distinction to be made based on information from publications alone. Tomb assemblages were assigned diamonds and settlement assemblages triangles. Besides some outliers, the correspondence analyses shows two clear groups corresponding to the different functional contexts, with the settlement assemblages represented by triangles on the left and the tomb assemblages indicated by diamonds on the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The LBA pottery of Area BU in the Royal Palace of Qatna, Syria

227

right of the scatter plot. This demonstrates how great an extent the functional contexts of ceramics are when studying the similarities and differences between pottery assemblages. In order to eliminate the regional and functional aspect, only assemblages from the same region and the same functional contexts were taken for further analyses, for example tomb assemblages from the Levant region or only settlement assemblages from the Euphrates region, and assigned symbols according to their chronological assessment within the literature. However, no clear relationships could be observed in the results, indicating that chronological relationships were very weak in this dataset. To sum up, within the relatively short time frame of the Late Bronze Age, similarities and differences between pottery assemblages are primarily caused by regional proximity and functional analogies rather than chronological aspects. However, it has also to be emphasized that these results are based on a frail data set, as most assemblages are not published in their entirety, and it is nearly always impossible to find out how often a specific type is represented in an assemblage from the information given in the publication. Such information would surely add enormously to our possibilities in working with pottery assemblages.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

228

Döpper

Bibliography Ahrens, A./Dohmann-Pfälzner, H./Pfälzner, P., 2012. New Light on the Amarna Period from the Northern Levant. A Clay Sealing with the Throne Name of Amenhotep IV/Akhenaten from the Royal Palace at Tall Mišrife/Qatna. Zeitschrift für Orient-Archäologie 5, 232–248. Al-Maqdissi, M., 2002. Ergebnisse der sechsten Kampagne der syrischen Ausgrabungen in Mišrife-Qaṭna im Jahr 2000. Mitteilungen der Deutschen Orient-Gesellschaft 134, 193–206. Badre, L., 1997. Bey 003 Preliminary Report. Excavations of the American Museum of Beirut Museum, Bulletin d‘archéologie et d’architecture libanaises 2, 6–94. Badre, L./Gubel, E./Capet, E./Panayot, N., 1994. Tell Kazel (Syrie). Rapport préliminaire sur les 4e–8e campagnes de fouilles (1988–1992). Syria 71, 259–359. Bergoffen, C. J., 2014. Imported Cypriot and Mycenaean Wares and Derivative Wares. D. Ben-Shlomo/ G. W. van Beek (eds.), The Smithsonian Institution Excavation at Tell Jemmeh, Israel, 1970–1990. Smithsonian Contributions to Archaeology 50, 658–720. Courtois, J.-C., 1969. Le mobilier funéraire céramique de la tombe 4253 du bronze récent (ville sud d’Ugarit). Ugaritica VI, 121–137. Dohmann-Pfälzner, H./Pfälzner, P., 2006. Ausgrabungen und Forschungen in Tell Mišrife/Qaṭna 2004 und 2005. Vorbericht der deutschen Komponente des internationalen Kooperationsprojektes. Mitteilungen der Deutschen Orient-Gesellschaft 138, 57–107. –

2007. Ausgrabungen und Forschungen 2006 im Königspalast von Qaṭna. Vorberichte des syrischdeutschen Kooperationsprojektes in Tall Mišrife/Qaṭna. Mitteilungen der Deutschen OrientGesellschaft 139, 131–172.

–

2008. Die Ausgrabungen 2007 und 2008 im Königspalast von Qaṭna. Vorbericht des syrischdeutschen Kooperationsprojektes in Tall Mišrife/Qatna. Mitteilungen der Deutschen OrientGesellschaft 140, 17–74.

du Mesnil du Buisson, R., 1927. Les ruines d’El-Mishrifé au nord-est de Homs (Emèse). Syria 8/1, 13–33. –

1935. Le site archéologique de Mishrifé-Qaṭna. Collection de textes et documents d’Orient 1, Paris.

Hachmann, R., 1982. Über die Grenzen der Möglichkeiten einer statistischen Auswertung von Keramik aus Kamid el-Loz. R. Hachmann (Hrsg.), Bericht über die Ergebnisse der Ausgrabungen in Kamid el-Loz in den Jahren 1971 bis 1974. Saarbrücker Beiträge zur Altertumskunde 32, Bonn, 179–208. Holladay, J. S., 1990. Red Slip, burnish, and the Solomonic gateway at Gezer. Bulletin of the American Schools of Oriental Research 277/278, 23–70. –

1993. The use of pottery and other diagnostic criteria, from the Solomonic era to the divided kingdom. Biblical archaeology today, 1990. Proceedings of the second international congress on Biblical archaeology, Jerusalem, June–July 1990. Jerusalem, 86–101.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The LBA pottery of Area BU in the Royal Palace of Qatna, Syria

229

Luciani, M., 2002. Operation K. M. al-Maqdissi et al. (eds.), Excavating Qatna I. Preliminary Report on the 1999 and 2000 Campaigns of the Joint Syrian-Italian-German Archaeological Research Project at Tell Mishrifeh. Documents d’archéologie syrienne IV, Damascus, 145–168. Mazar, A., 1998. On the appearance of red slip in the Iron Age I period in Israel. S. Gitin et al. (eds.), Mediterranean peoples in transition. Thirteenth to early tenth century BCE, Jerusalem, 368–378. Metzger, M./Barthel, U.-R., 1993. Die spätbronzezeitlichen Tempelanlagen: die Kleinfunde, Kāmid el-Lōz; 8, Text, Bonn. Morandi Bonacossi, D., 2013. The crisis of Qatna at the beginning of the Late Bronze Age II and the Iron Age II settlement revival. A regional trajectory towards the collapse of the Late Bronze Age palace system in the Northern Levant. K. A. Yener (ed.), Across the border: Late Bronze-Iron Age relations between Syria and Anatolia. Proceedings of a Symposium held at the Research Centre of Anatolian Studies, Koç University, Istanbul, May 31–June 1, 2010. Ancient Near Eastern Studies. Supplement 42. Leuven, 113–146. Morandi Bonacossi, D./Da Ros, M./Garna, G./Iamoni, M./Merlino, M., 2009. The „Eastern Palace“ and the Residential Architecture of Area T at Mishrifeh/Qatna. Preliminary Report on the 2006–2008 Excavation Campaigns of the Italian Component of the Syro-Italian Archaeological Project. Mesopotamia 44, 61–112. Novák, M., 2004. The Chronology of the Royal Palace of Qatna. Ägypten & Levante 14, 299–317. Novák, M./Pfälzner, P., 2000. Ausgrabungen in Tall Mišrife-Qaṭna 1999. Vorbericht der deutschen Komponente des internationalen Kooperationsprojektes. Mitteilungen der Deutschen OrientGesellschaft 132, 253–295. –

2001. Ausgrabungen in Tall Mišrife-Qaṭna 2000. Vorbericht der deutschen Komponente des internationalen Kooperationsprojektes. Mitteilungen der Deutschen Orient-Gesellschaft 133, 157–198.

Saidah, R., 2004. Sidon et la Phénicie Méridionale au Bronze Récent: à propos des tombes de Dakerman, BAH 170, Beyrouth. Salles, J.-F., 1980. La Nécropole „K“ de Byblos, Recherche sur les grandes civilisations Mémoire no 2, Paris. Schaeffer, C. F. A., 1938. Les fouilles de Ras-Shamra-Ugarit. Neuvième campagne (printemps 1937). Syria 19, 193–255, 313–334.

Stephanie Döpper, University of Tübingen, Institute for Ancient Near Eastern Studies (IANES).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 231–242

Frank Hole

Historical Processes on the Middle Khabur River, Syria, during the Late Ubaid: Intrusion, Attenuation and Divergence The immigration of Ubaid villagers from Southern Mesopotamia into the Khabur Basin of Northern Mesopotamia exemplifies a general historical process: Emigration into virgin territory, attenuation of the homeland culture, adaptation to new circumstances, culture drift, assimilation of new elements, and formation of a distinct cultural entity. During the 5th millennium BC ceramics in a distinctive style known as Ubaid spread across much of the fertile Near East (Carter/Philip 2010). Ubaid ceramics originated in Southern Mesopotamia where there is a millennium-long sequence, but the spread across the Near East occurred during the latest phase of Ubaid and the terminal phase of Halaf. There are many theories about the causes of this diffusion, essentially there are three possibilities: emigration of people bearing Ubaid ceramics and other traits into new territory, the spread of traits through emulation by local people, or through trade of selected elements. This paper discusses how people who immigrated into the Khabur River basin of Northeastern Syria exemplify each mode of transmission. The essentially simultaneous appearance of at least six Late Ubaid settlements along the Middle Khabur River and its upper tributaries is a clear case of immigration (fig. 1). The sites, clustered along the river or its tributaries, that have been excavated or extensively sampled, are: Mashnaqa (Thuesen 2000), Ziyadeh (Arzt 2001; Hole 2000b), Mulla Matar (Sürenhagen 1991), Kashkashok III (Suleiman and Taraqji 1995), Beydar III (Suleiman et al. 2003), and Kuran (Hole 2000a). In each case the new settlers chose to build on fresh ground and there is no evidence that they moved into an occupied territory. These sites appear to be small agricultural settlements, implying that one motivation for immigrating was to obtain farmland. The fact that these sites are strung out at close intervals suggests that they are part of a closely cooperating social unit, as they must have been to sustain life from generation to generation. Assuming that environmental conditions were suitable, and the full range

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

232

Hole

Fig. 1. Six Late Ubaid sites on the middle Khabur River and Beydar III.

of social and technical skills were available (Thuesen 2000), the migrants could establish virtual replicas of their previous communities. The historical processes involve: (1) emigration from a homeland and immigration into a new land, (2) selective transmission of culture traits to a new locale, (3) divergent adaptation, (4) assimilation of new traits, and (5) formation of a new cultural tradition. The processes are largely sequential. (1) We cannot specify the location of the settlements from which the immigrants came, but the ceramics (for example bell-shaped bowls), are similar to those from sites in southern Mesopotamia, such as Eridu (Thuesen 2000). Between the classic Ubaid sites of southern Mesopotamia and those of the north, there is a wide desert, unsuited to rain-fed or irrigation agriculture. Geography thus limits potential sources of homelands and routes of movement, but two river routes that might have led into the western Jazireh are the Euphrates and its tributaries, the Khabur and Balikh. These tributaries provide a route north, as well as access to potentially arable land once they cross the precipitation boundary that allows rain-fed agriculture. In the north, Ubaid settlements followed other smaller perennial streams northward to the base of the Tur Abdin Mountains. Migration up the Tigris and its tributaries would also have been possible, but east-west travel across the semi-arid steppe known as the Jazireh, would have been more

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Historical Processes on the Middle Khabur River, Syria during the Late Ubaid

233

Fig. 2. Late Ubaid vessels, including the distinctive bell-shaped form, from Kashkashok II (Matsutani 1991: Figs. 86-87).

difficult. During most periods the eastern and western parts of the Jazireh may have been culturally separate although some exchange of cultural elements occurred. Barring widespread and universal catastrophe, it is possible that the emigrant families represent a sub-set of the source populations rather than a mass exodus of the homeland culture. In such a case, some cultural elements will almost certainly not make the move; hence the immigrants will carry an attenuated version of the homeland culture. What people took with them would depend on what was important, portable and useful in a new settlement. Things that could be replicated need not have traveled, but almost inevitably some cultural elements would be left behind. We may think of this as the «bottle-neck effect», resulting in some attenuation of the homeland cultural complex. The initial implantation of Ubaid cultural elements in the Khabur represented a subset of things previously available and used, including multi-room (tripartite) houses. The painting styles on bell-shaped bowls, as found in Mashnaqa and Kashkashok II, are particularly noteworthy, with nearly exact parallels in Eridu (fig. 2) (Matsutani 1991; Thuesen 2000). In Mashnaqa also Mesopotamian-style bent ceramic «nails», variously attributed to wall pegs

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

234

Hole

Fig. 3. The middle Khabur region and Kenan Tepe.

and potters’ tools, and a boat model (Akkermans/Schwartz 2003: Fig. 5.7; Thuesen 1996), identical to those found at Ubaid Eridu, Ubaid, Oueili and others. It is likely that there never was a successful duplication of all of the material culture or social practices that might have been available. This is particularly evident with social arrangements. Populations in the north were small and, until now, no religious structures apart from those at Gawra in northeastern Iraq have yet been excavated. That they may exist at sites such as Tell Brak is not questioned but, for the foreseeable future, neither the extent of this potentially large Ubaid site, most of which is buried by later remains, nor its particular features can be ascertained with any reliability. (2) The Middle Khabur River valley is surrounded by semi-arid to dry steppe (fig. 3). Precipitation is too low to support dependable rain-fed agriculture south of the settlements. People traveling from Mesopotamia would have had to traverse many hundreds of kilometers of land unsuitable for agriculture along the river valleys or east-west across the steppe. Movement by boat up the Khabur would have been against the current and not easily accomplished, and pack animals were not available, so all necessary goods must have been carried by people or replicated on site. This implies that people moved from the south to the north, over a linear (not travel) distance of 800 km, and into a distinctly different

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Historical Processes on the Middle Khabur River, Syria during the Late Ubaid

235

environment. That is, they came from a broad riverine plain flanked by desert, to a narrow river valley flanked by semi-arid steppe. In the 5th millennium there may have been considerably more precipitation to allow rain-fed farming, but the south of Mesopotamia would still have depended on irrigation. We cannot determine the precise reasons why colonists moved into this region, but environmental perturbations in the south may have contributed to the move (Hole 1994; Kouchoukos 1998). (3) Depending on the environment of the source populations, the immigrants may have found different circumstances to adapt to: these might include different methods of cultivation, ready accessibility of game animals, wild predators and perhaps hostile human groups. While the new land may not have held permanent settlements, it lay in a zone of transition between steppe and mountain where movement, at least seasonally, is easy and contacts with other groups likely. The absence of intensive surveys of the intermediate landscape between the middle Khabur and the Tur Abdin Mountains precludes definitive knowledge of occupation prior to the incursion of the Ubaid. Nieuwenhuyse’s analysis of sherds recovered by Lyonnet (Lyonnet 2000) indicates that during the late Halaf there were small sites as far south as the limits of risk-free agriculture; however the so-called Halaf–Ubaid transitional period is illdefined and there appears to be a contraction of settlement toward the north in the late 6th–5th millennium. Based on radiocarbon dates, Campbell (2007) argues that we cannot verify an overlap between Halaf and Ubaid. Whatever the case, any preceding settlements were very small and perhaps were seasonal camps rather than permanent settlements. In short there may have been no territorial competition between people of Halaf and Ubaid. (4) To make the most effective use of the land for farming and keeping of livestock the newly settled communities were small and spaced at intervals along the Khabur River (fig. 4). The narrow width of the floodplain and the stony nature of the first terrace would have precluded extensive agricultural fields. Owing to the seasonal pattern of precipitation, the river would have been at its lowest level at the time of planting, thus limiting potential gravity irrigation. The interannual variability in precipitation also made rain-fed agriculture risky. Finally, the spring spate as precipitation and snow-melt swelled the river could lead to flooding of both the settlement and fields. Methods of coping with these exigencies must have occupied much of the effort of the initial colonists. (5) The many similar small settlements comprised a local community for mutual support during better and worse times, and interacted at the level of marriage, necessary communal labor, and protection. The colonists had a truncated version of their homeland culture, but one adapted to the new circumstances with changes in architecture and material culture. Further attenuation of ties with the homeland, too distant for regular interaction, must have occurred with the frictional effect of time and distance.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

236

Hole

Fig. 4. Tell Kuran on the Khabur River.

Once people had become established, their ceramics began to drift away from the original patterns. There is no evidence of influence in the ceramics from any native local population. Trends of simplification in material culture, which occurred elsewhere in the Late Ubaid, also occurred in the Khabur region (Wengrow 2001). On the Khabur the characteristic post Ubaid styles became a recognizably distinct phase that we refer to as the Kuranian. (6) As ties with the homeland loosened, connections with other groups across the Jazireh must have strengthened. The settlements on the Khabur were on an environmental continuum with mountains and greater precipitation in the north, to desert in the south. This transect, and the strongly seasonal weather, encouraged interaction with any transhumant pastoralists and traders in obsidian in the north. Such interaction opened possibilities for exchange, emulation and the use of new techniques. «Contact» is often expressed in the presence of rare or stylistically distinctive artifacts in two places. Among the Ubaid ceramics one element is especially noteworthy. This is a motif on a bowl that features a curving horn enclosing a dot (fig. 5). This design is not known among Ubaid sites in Mesopotamia, nor is it found in the Jazireh outside the Khabur drainage, yet it occurs at the sites on the middle Khabur and at Kenan Tepe (fig. 5) in Anatolia where the closest ceramic affinities are with the Khabur (Parker 2010). Another distinctive trait is rare, but widely distributed, red burnished wares, suggesting exchange among sites.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Historical Processes on the Middle Khabur River, Syria during the Late Ubaid

Fig. 5. Late Ubaid bowls with horn design from Ziyadeh and Kenan Tepe (Parker 2010: Fig. 21.12).

Fig. 6. Shallow bowls with interior incisions and scraped exteriors, from Ziyadeh.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

237

238

Hole

Fig. 7. Stamped and incised designs (Al Kuntar 2010: Fig. 3.12:73-76) and sprig ware (Abu Jayyab 2012: Fig. 3), from Hamoukar.

How might contact/exchange have occurred? The sites on the middle Khabur and Kenan Tepe are separated by the entire upper Khabur and some distance into the mountains. I suggest that the question might be answered through survey of the north-south trending streams, which were probably perennial in the 5th millennium. It is probable that we would find a string of settlements, newly founded in the Late Ubaid, all the way into Anatolia. The situation may be analogous to that of the Balikh where sites occur along the entire length of the river and its tributaries (Wilkinson 1998) (fig. 3). Accurate radiocarbon dating of the foundation of the sites would help answer this question. At the moment evidence suggests that it was not emulation or trade that account for sites with similar ceramics, but immigration and interaction within a common tradition. It is interesting that the large Halaf and Ubaid site of Zeidan is at the confluence of the Balikh and Euphrates, separated from the Ubaid sites upstream (Stein 2010–2011). So far as we know, there is no equivalent site at the confluence of the Khabur and the Euphrates, which might have been a taking-off point for upstream settlement. The only Ubaid site known from the lower Khabur region, Wadi Umm Aqubba, lies east of the river along an east-west route toward (modern) Mosul in Iraq, and close to the Bouara salt source. Late Ubaid ceramics were not found in the region, but the extensive surface collection revealed ceramics like the Kuranian (Bernbeck 1993: 43); hence the site is not part of the initial migration up the Khabur. (7) Concomitant with local adoptions of new styles and products, the entire Jazireh, and much of the Ubaid world experienced a «simplification» in the painting of pottery. In some regions painting ceased entirely while in others it was much reduced in amount as compared

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Historical Processes on the Middle Khabur River, Syria during the Late Ubaid

239

with plain wares, and there was evidence of mass production of standardized vessels. The trends are clear but the specifics vary by region and by site. (8) The two processes of adoption of new traits and simplification of the ceramics led to regionally distinct «post-Ubaid» assemblages. Intrusions, such as with scraped side (Coba) bowls and those with interior incising, seem to have come from the north and west. Somewhat later, intrusions include stamped and incised wares and sprig ware, apparently from the east of the Jazireh (fig. 6–7). The Khabur lies between these two regions and absorbed these new elements at different times from both. The result is that each geographic region was characterized by a shared, but partly unique, set of attributes. The adoption of traits at different sites is uneven in timing and pertains to relative frequency as well as to types themselves. There is no suggestion here that all of the perhaps hundreds of Ubaid sites that formed across the Jazireh from northern Iraq to the Euphrates are of exactly the same age, yet they all have Late Ubaid-style ceramics, so the diffusion was relatively rapid. Presumably in each region there was an initial incursion of immigrants who founded a few settlements and, through further immigration, as well as opportunity and population increase, gradually filled suitable places in the Jazireh and beyond. The process may have taken several hundred years before the local cultures, now referred to as Late Chalcolithic 1 and 2, formed (Marro 2013a, 2013b). With a large resident population in the Jazireh reaching from the valleys of the Tur Abdin to the limits of rain-fed agriculture in the Jazireh, the stage was set for the formation of the distinctive unpainted, chaff tempered LC 2. Maro suggests that there are six post-Ubaid geographic provinces, the Khabur being one (Marro 2013b: 23–24). She further sees the post-Ubaid as a time of changed social and economic order, in which there is the co-existence of several cultural horizons (Marro 2013b:20). A possible disruption before the formation of LC 3, conventionally called Middle Uruk, is attested at many sites. Assuming the scenario of immigration is correct, it leaves unanswered «what happened to the Halaf?» The newly founded Ubaid sites show little or no stylistic affinities with the Halaf, and the few sites alleged to have a «transitional» phase are problematic. Perhaps the idea of a transition conforms more to expectation than to reality (Campbell 2007). Brenequet (1989) argues that there was local evolution of the Iraqi Northern Ubaid from the Halaf, stimulated in large part by what she sees as widespread interregional interaction manifest in the distribution of obsidian. For reasons outlined below, I find it more plausible to think that carriers of Ubaid pottery moved into the Syrian Jazireh whether or not there were local Halaf folk in the region. When the Ubaid appears in the Khabur region it features a distinctly different architecture, ceramic assemblage, and residential stability than did the Halaf with its typical tholoi structures and short-lived settlements (Akahane 1998; Hole/Johnson 1986–87). We should remember, however, that the sites of Arpachiyah, Songor B and Tell Hassan, all on the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

240

Hole

eastern perimeter of Halafian distribution, have rectangular buildings in their latest phases (Breniquet 1989:331). There are not enough data to indicate whether the Ubaid sites are always larger than the Halaf: Akkermans (1989: 341) notes that most Ubaid sites are only a hectare or less in size, although one cannot point to any differences in their geographic distribution from the Halaf (Wilkinson 1995: 40). Moreover, there are «cemeteries» associated with Ubaid sites, including Mashnaqa, Kashkashok II and Arpachiyah, whereas none has yet been found in Halaf sites. These aspects, as well as differences in pottery and architecture, suggest a cultural if not chronological gap. Whatever the case may be, there are now competing hypotheses for researchers in the future to investigate. Targeted excavations, close attention to actual chronology, and technical materials analyses of ceramics and other artifacts may resolve the questions. The post-Ubaid is denoted by the predominance of plain ceramics, albeit with regional variation. The loss of painting on pots, a «simplification», was a widespread phenomenon, not limited to the Ubaid sphere (Wengrow 2001). There is no obvious reason for this, but authors remark on the appearance of mass-produced wares. The appearance of these as well as diagnostic types, such as Coba bowls, and stamped and incised wares, might be explained through interregional contacts. Exchange of material and customs, through transhumance, trade, marriage and the like could lead to adoptions of new elements of culture. Over a geographic region as broad as northern Mesopotamia and beyond, these interactions might result in something like the six provinces suggested by Marro.

Concluding Remarks The six sites on the Khabur River document immigration into a new region by people with probable origin in Southern Mesopotamia. When they arrived they came with traditional knowledge that had to adapt to new circumstances. Over several hundred years they adapted to floods, interannual weather variability, and new opportunities for hunting on the vast steppe. Nevertheless each site remained small, homestead-like and sustained social interaction within a relatively small radius. As time passed the people changed in concert with region-wide trends toward simplification of ceramics and other items of material culture. In relative isolation from their place of origin they developed dialectic differences in their ceramics, a slow but inexorable trend of divergence (Hole 2013). Evidence indicates that Ubaid sites spread across the Jazireh, some no doubt the offspring of the initial Khabur residents; other perhaps from migrants who followed the Tigris and settled at places like Leilan. A critical part of the puzzle is whether Tell Brak was also a pioneer settlement in the Late Ubaid, built on a previously occupied, but vacant mound. The site of Zaidan, which has both Halaf and Ubaid, may also be a critical link in understanding whether Ubaid settlers moved onto a vacant mound or whether there was a period of transition and accommodation between two cultures. Another part of the puzzle of the spread of Ubaid, lies in the foothills

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Historical Processes on the Middle Khabur River, Syria during the Late Ubaid

241

of southern Turkey, where Ubaid ceramics clearly mixed with local non-Halafian traditions to create another dialectic variant of the pan-regional Ubaid phenomenon. There is still a lot to learn.

Bibliography Abu Jayyab, K., 2013 A ceramic chronology from Tell Hamoukar’s southern extension. Marro 2013a: 87–127. Akahane, S., 1998. Environmental and archaeological setting of the middle Khabur region. 1. Landforms and Geology. A. Tsuneki and Y. Miyake (eds.), Excavations at Tell Umm Qseir in Middle Khabur Valley, North Syria. Report of the 1996 Season. University of Tsukuba, Studies for West Asian Archaeology. Tsukuba. Al-Shark 1, 4–10. Akkermans, P. M. M. G., 1989. Tradition and social change in northern Mesopotamia during the later fifth and fourth millennium B.C. Henrickson/Thuesen 1989: 339–367. Akkermans, P. M. M. G./Schwartz, G. M., 2003. The Archaeology of Syria From Complex HunterGatherers to Early Urban Societies (ca. 16, 000–300 BC). Cambridge. Arzt, J., 2001. The “Ubaid Expansion” Reconsidered: Excavations at Tell Ziyadeh, Syria. Ph.D., Yale University. Bernbeck, R., 1993. Steppe als Kulturlandschaft. Berliner Beiträge zum Vorderen Orient Ausgrabungen 1. Berlin. Breniquet, C., 1989. Les origines de la culture d’Obeid en Mésopotamie du nord. Henrickson/Thuesen 1989: 325–338. Campbell, S., 2007. Rethinking Halaf chronologies. Paléorient 33.1, 103–136. Carter, R. A./Philip, G., 2010. Deconstructing the Ubaid. Carter/Philip 2010b: 1–22. –

(eds.), 2010b. Beyond the Ubaid. Transformation and Integration in the late Prehistory of the Middle East. The Oriental Institute of the University of Chicago 63.

Henrickson, E./Thuesen, I. (eds.), 1989. Upon This Foundation – the ‘Ubaid Reconsidered. The Carsten Niebuhr Institute of Ancient Near Eastern Studies 10. Hole, F., 1994. Environmental instabilities and urban origins. G. Stein/M. Rothman (eds.), Chiefdoms and Early States in the Near East: The Organizational Dynamics of Complexity. Monographs in World Archaeology (Madison) 18, 121–151. –

2000a. The prehistory of the Khabur. Rouault/Wäfler 1993: 16–27.

–

2000b. Tell Ziyadeh on the Middle Khabur, Syria. 1st International Congress on the Archaeology

–

2004. Tell Kuran (Syria). M. Lebeau (ed.), Atlas of Pre-Classical Upper Mesopotamia (APUM).

of the Ancient Near East, Rome. Università degli Studi di Roma “La Sapienza”, 609–619. Subartu 13. –

2013. Constrained innovation: Halafian ceramics. O. Nieuwenhuyse et al. (eds.), Interpreting the Neolithic of Upper Mesopotamia. Turnhout, 77–87.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

242

Hole

Hole, F./Johnson, G. A., 1986–87. Umm Qseir on the Khabur: Preliminary report on the 1986 excavation. Annales Archéologiques Arabes Syriennes 36–37, 172–220. Kouchoukos, N., 1998. Landscape and Social Change in Late Prehistoric Mesopotamia. Ph.D., Yale University. Lyonnet, B. (ed.), 2000. Prospection Archéologique Haut-Khabur Occidental (Syrie du N.E.). Beyrouth: Institute Française d’Archéologie du Proche-Orient 1. Marro, C. (ed.), 2013a. After the Ubaid: Interpreting Change from the Caucasus to Mesopotamia at the Dawn of Urban Civilization (4500–3500 BC). Paris. –

2013b. Is there a Post-Ubaid culture? Reflections on the transition from the Ubaid to the Uruk Periods along the Fertile Crescent and beyond. Marro 2013a: 13–38.

Matsutani, T. (ed.), 1991. Tell Kashkashok: The Excavations at Tell No. II. Tokyo: Institute of Oriental Culture, University of Tokyo. Parker, B. J., 2010. Networks of interregional interaction during Mesopotamia’s Ubaid period. Carter/ Philip 2010b: 339–360. Rouault, O./Wäfler, M. (eds.), 1993. La Djezire et l’Euphrate Syriens de la protohistoire à la fin du second millenaire av. J.-C.: tendances dans l’interpretation historique des donées nouvelles. Subartu 7. Stein, G. J., 2010–2011. Tell Zeidan. Oriental Institute Annual Report 2010–2011, 122–139. Suleiman, A. et al., 2003. The Chalcolithic pottery from Tell Beydar III. Subartu 10, 527–553. Suleiman, A./Taraqj, A., 1995. Tell Kashkashok III. Syria 72, 172–183. Sürenhagen, D., 1991. Mulla Matar. American Journal of Archaeology 95, 714–715. Thuesen, I., 1996. Tell Mashnaqa. Syrian-European Archaeology Exhibition. Damascus: Institut Fraçaise d’Études Arabes de Damas, Ministry of Culture/European Union. 47–53. –

2000. Ubaid expansion in the Khabur: new evidence from Tell Mashnaqa. Rouault/Wäfler 1993: 71–79.

Wengrow, D., 2001. The evolution of simplicity: Aesthetic labour and social change in the Neolithic Near East. World Archaeology 33, 168–188. Wilkinson, T., 1998. Water and human settlement in the Balikh Valley, Syria: investigations from 1992– 1995. Journal of Field Archaeology 25.1, 65–67. Wilkinson, T. J./Tucker, D. J., 1995 Settlement Development in the North Jazireh, Iraq: A Study of the Archaeological Landscape. Volume 3. Baghdad: British School of Archaeology in Iraq. Frank Hole, Yale University.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 243–255

Béatrice Muller

Panneaux d’incrustation en coquille de Mari, Ville II: implications des matériaux et des techniques The inlay pieces are considered here under a new light: magnifying methods have been used to observe the surface, the edges and the bitumen-filled carvings; the structural characteristics ensuring a more precise typology of the figures have been collected and the required specific signs linking one piece to another have been emphasized. En vue de l’exposition Voués à Ishtar (IMA, 22 janvier–20 mai 2014, Cluzan/Butterlin 2014) les pièces d’incrustation de l’«Étendard de Mari» (Louvre AO 17572/AO 19820, AO 18965; Parrot 1956: 135–155 et pl. LVI à LX) ont été démontées du support (un tableau en bois) dans lequel elles avaient été encastrées dans les années 1930 et ainsi présentées depuis: occasion unique de nouvelles observations et de photographies grossissantes (macrophotographies et photographies à la loupe binoculaire1). En outre, les réserves du Musée renferment environ 80 pièces d’incrustation qu’André Parrot n’avait pas inventoriées et qui ont été récemment restaurées2: celles-ci aussi méritent une étude approfondie. Enfin, la prise en compte d’un lot de pièces pratiquement inédit recueilli en 1999 dans l’Espace IV du «Palais»3 de Ville II (Margueron 2004: 290–293; Margueron 2014: 273–276, 284–287), permet de préciser le contexte d’un atelier de travail de la nacre, déjà démontré par l’étude de 1639 silex, parmi lesquels 66 microperçoirs et 64 lamelles à troncature concave (Coqueugniot 1993; Margueron 2004: 224–226). Ce travail sur les éléments de mosaïque en coquille, encore en cours, a suscité des données nouvelles

1 2 3

Nikon SMZ 1000 couplée avec appareil de photo Nikon Coolpix P 5100. N° AO, mais pas de n° M. Abréviations dans la suite du texte et légendes des figures: MAM = Mission archéologique de Mari. Dimensions indiquées en mm. Le «Palais présargonique» (Ville II) est considéré maintenant comme un temple-manufacture (Margueron 2014).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

244

Muller

et une méthode dont les résultats présentés ici devront être confirmés par une recherche systématique.4

I. Mises au point sur les pièces d’incrustation comme objets archéologiques 1. Origine de la coquille nacrée: Pinctada radiata Que la coquille provienne d’une variété d’huître perlière était l’hypothèse la plus raisonnable. C’est ce qu’ont confirmé, sur photographies et sur pièces, deux spécialistes du Muséum d’Histoire Naturelle, Bernard Lozouet et Pierre Métivier, qui penchaient d’abord, selon l’opinion générale, plutôt pour la Pinctada margaritifera. Or la dimension des pièces (7 cm maximum) va plutôt dans le sens de la Pinctada radiata, plus petite, vivant principalement dans le golfe de Mannar (au sud-est de l’Inde), le golfe Arabo-Persique et la mer Rouge. L’attache du mollusque, qui se décèle parfois sur la pièce d’incrustation, permet de situer l’endroit où celle-ci a été prélevée (Muller 2014: 292–294 et ill. 6). Excepté la couche extérieure, le périostracum, c’est toute l’épaisseur de la coquille qui est utilisée, laquelle n’est pas constante, ce qui se voit dans les tranches, qui montrent souvent un feuilletage.

2. Origines possibles de la coquille non nacrée Parmi le lot de silex ayant servi à la gravure de la nacre provenant du Temple-manufacture de Ville II figuraient des débris de Pinctada, d’Unio (moules d’eau douce à coquille nacrée pouvant servir à de petits éléments d’incrustation) et un de Conus (non nacré, ne pouvant pas servir à fabriquer des incrustations, mais des sceaux-cylindres). Les dépôts des favissae du temple de Ninhursag, dont certains vestiges au moins datent de la Ville I, contenaient également des objets faits à partir de columelles, de Conus en particulier, mais aussi éventuellement de bivalves (Beyer/Jean-Marie 2007: 101–119). L’origine de la coquille non nacrée est difficile à identifier à partir des pièces d’incrustation.

4

L’inventaire réuni dans la thèse (inédite) de B. Couturaud (2013) m’a été utile. Je remercie vivement Sophie Cluzan, conservateur du Patrimoine au musée du Louvre et ses collaborateurs, de m’avoir donné accès à ces objets, ainsi que Sabine Kessler, restauratrice, pour ses observations; ma gratitude va à Isabelle Sidéra, directrice de l’UMR 7055, Préhistoire et Technologie, de m’avoir permis d’utiliser son matériel pour les prises de vue à la binoculaire ainsi qu’à Alexandra LegrandPineau, USR 3225, responsable du Service d’Imagerie et de Microscopie Optique de la Maison René Ginouvès à Nanterre, pour ses conseils techniques. Je remercie Jean-Claude Margueron, qui m’a permis d’utiliser les documents inédits des nacres exhumées en 1999.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Panneaux d’incrustation en coquille de Mari, Ville II

245

Fig. 1. Avant-train d’équidé (AO 25745-b). a. La pièce dans son intégralité (bord à droite cassé). h. 37; l. 37; ép. 5 mm (© B. Muller). b. Détail de la crinière incisée avec restes de bitume (grossissement initial x 60) (© B. Muller).

Fig. 2. Dignitaire militaire M. 472 («Étendard»). a. Revers de la pièce constituant le buste. h. 30; l. 40; ép. 5 mm. (© B. Muller). b. Dessin (Parrot 1956, fig. 80) auquel a été ajouté le surlignage permettant de distinguer les cinq pièces constitutives.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

246

Muller

Fig. 3. Pièces d’incrustation de chars et comparaison. a. Caisse et roue du char de l’Étendard (M. 158 et 161) avec, à gauche, les deux jambes du soldat debout à l’arrière. h. 26; l. 36; ép. 3 mm. (© B. Muller). b. Tablier de char (Temple-manufacture, niveau P-2, Espace 4), montrant dans son échancrure un tronçon de la paire de guides. h. 67, 5; l. 24, 5 mm. (Archives MAM J.-Cl. Margueron). c. Empreinte de sceau d’Ishqi-Mari (Temple-manufacture, niveau P-1, secteur de l’entrée, sol A du loc. 11 et couche A du loc. 13). h. 30 mm. (Beyer 2007: 254 fig. 17).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Panneaux d’incrustation en coquille de Mari, Ville II

247

Fig. 4. Modes de jonction complexes (encastrement). a. Assemblage techniquement possible (mais iconographiquement non certain) pour un dignitaire militaire (P-2, Espace 4). h. de la figure 120 mm. (Dessins initiaux A. Horrenberger, Archives MAM J.-Cl. Margueron). b. Bras plié à l’équerre avec encoche en arrière de l’aisselle (AO 25742-o). L. 30 mm. (© B. Muller). c. Partie inférieure de jupe à languettes avec tenon d’encastrement sur le bord supérieur (M.2132, temple de Shamash). h. 33; l. 35 mm. (dessin archives MAM A. Parrot).

3. Détail des gravures, usage du bitume et de la couleur Après découpage de la silhouette (cf. fig. 2a) – sciée après avoir été délimitée par une gravure sommaire du contour –, sont ajoutés les détails: yeux, oreilles, mains, corde, dont la gravure, réalisée à l’aide d’outils pointus (lamelles de silex tronquées, microperçoirs, cf. Coqueugniot 1993: 229–234), était soulignée de bitume, lequel se voit encore souvent à l’œil nu, mais que les photos grossissantes permettent de certifier en cas de doute. On ne peut qu’admirer ce remplissage extrêmement fin – généralisé –, de même que la sûreté du trait de la gravure (Parrot 1956: 142–143). Il subsiste également de rares touches de couleur: le rouge marque la caisse du char de l’Étendard (comme comme sur l’Étendard d’Ur) (cf. fig. 3) tout comme le harnais (entre naseau et chanfrein) des équidés de l’attelage.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

248

Muller

Des micro-outils de largeurs différentes étaient utilisés: le contour interne de la crinière d’équidé est bien plus épais que les stries qui en dessinent les poils, lesquels révèlent encore à la binoculaire d’infimes grains de bitume (fig. 1).

4. Observation systématique des bords Jamais réalisée jusque là, l’observation systématique des bords a son importance sur la connaissance de l’état de conservation et de la forme de l’élément iconographique de base que constitue la pièce d’incrustation. Ainsi, le buste du dignitaire militaire (fig. 2), quoiqu’étant une pièce intacte, n’est qu’une partie de la figure – et même du buste lui-même. Or ces questions de découpe ou de cassure n’ont pas du tout préoccupé A. Parrot: les dessins des mosaïques du temple d’Ishtar (Parrot 1956: 137–140, 144) – qui ne sont d’ailleurs pas exhaustifs – présentent les figures sans en détailler les diverses composantes, et le catalogue stipule généralement «fragment»: il y a confusion entre la figure, le motif (personnage, animal, objet ) et la pièce d’incrustation, laquelle est un objet archéologique en soi (Muller 2014: 288–290).

II. Observations techniques et réflexions nouvelles Ces précisions m’ont conduite à considérer ces pièces d’incrustation selon une optique plus analytique et à me poser diverses questions, non pas tant sur le mode d’exécution en luimême que sur la conception de ces images, décomposées en puzzles selon des principes structuraux inhérents aux conventions de représentation, mais aussi aux contraintes imposées par le matériau.

1. Mode de découpage: nombre de pièces par figure Les représentations humaines, animales et d’objets complexes sont, à Mari, faites de plusieurs morceaux jointifs pouvant aller de 2, comme pour les prisonniers, à 5 comme pour les dignitaires militaires, ou même à 7 pour certains soldats (cf. figs. 5b, 7b–c). Le schéma semble différent sur l’Étendard d’Ur (cf. par exemple Aruz [éd.] 2003: n° 52); le char paraît fait de 4 pièes principales (caisse avec son tablier, roues indépendantes et carquois) et d’une ou deux pièces secondaires (marchepied et rambarde), auxquelles il faut ajouter le timon (1 pièce), le passe-guides (1 pièce) et les rênes (3 pièces). L’attelage de l’Étendard de Mari est incomplet, mais on voit tout de suite que, si l’encastrement des roues sous la caisse suit le même principe, celle-ci est indépendante du tablier mais solidaire des deux jambes du soldat reposant sur un marchepied à peine suggéré: le découpage de la figure est donc différent. Quelques comparaisons permettent de constater

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Panneaux d’incrustation en coquille de Mari, Ville II

249

Fig. 5. Militaires: deux schèmes différents. a. M. 466 (temple d’Ishtar, Étendard). h. 64; l. 51; ép. 7, 5 mm. (© B. Muller). b. Tentative d’assemblage (B. Muller) de cinq pièces (sur sept) provenant de l’Espace 4 du «Templemanufacture», niveau P-2.). h. restituée du personnage 120 mm. (Dessins initiaux A. Horrenberger, Archives MAM J.-Cl. Margueron).

que, à Mari, le tablier constitue au moins une pièce à part alors que le personnage à l’arrière peut faire partie de la caisse (fig. 3).5

2. Mode de jonction des pièces: juxtaposition ou encastrement Lorsque la découpe est rectiligne, l’artisan dispose d’une marge d’approximation dans l’assemblage des pièces au moment du montage du panneau: ainsi, certains bustes et jambes de prisonniers sont interchangeables6. De même, éventuellement, pour le bras tendu en avant et plié de certains soldats ou dignitaires militaires (cf. fig. 7) ou des têtes d’équidés.

5

6

Tablier de char sur deux sceaux d’Ishqi-Mari (Beyer 2007: fig. 17–18) semblable à un exemplaire du temple de Ninni-zaza (M. 2587); jupon kaunakès ou peau d’animal intégrés à l’arrière de la caisse cf. M. 2465 et M. 2466 (temple de Ninni-zaza, tous trois cf. Parrot 1967: fig. 262); on trouve un élément semblable sur une plaque perforée sculptée d’Ur (DA II, Aruz [éd.] 2003: n° 31) montrant un char à deux roues. Les restitutions de l’Étendard en témoignent: ainsi, par exemple, les deux pièces du prisonnier M. 159 sont réunies sur celle du Louvre, mais associées à d’autres sur celle de Calmeyer 1967: fig. 6.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

250

Muller

En revanche, lorsque l’angle d’un coude doit s’encastrer dans l’échancrure d’une jupe ou, plus encore, lorsqu’une encoche à l’intérieur de l’épaule, de la largeur maximale de 3 mm, nécessite le tenon complémentaire, ou au contraire qu’un bas de jupe est pourvu d’un saillant (fig. 4) ou d’un rentrant, il s’agit alors véritablement du principe du puzzle.

3. Biseautage des tranches et champlevé: travail sur le volume L’artisan doit jouer sur l’épaisseur inégale de la coquille et obtenir un avers plan, ce qui peut conférer à une même pièce des épaisseurs variables. Un biseautage intentionnel des bords est également souvent pratiqué, même pour des pièces très fines7: il était sans doute plus facile ainsi d’insérer la pièce parmi les plaques de schiste du fond et d’éviter qu’elle se noie dans le bitume servant d’adhésif, un peu selon le principe d’enchâssement des pierres précieuses en joaillerie. Des évidements de matière étaient pratiqués pour les barbes et les chevelures, remplies de bitume, ainsi que pour les moyeux des roues. Ainsi la mise en œuvre de cette iconographie, qui apparaît en deux dimensions, prend en réalité en compte la troisième.

4. Tentatives d’assemblage: structure de l’image Celles-ci doivent tenir compte des détails comme du module (même si les variations en sont infimes), et le décryptage de la structure de la pièce devrait aider à assurer les restitutions. C’est sur ce dernier point que j’insisterai ici. Si l’on s’en tient au thème guerrier proche de celui des Étendards (Mari et Ur), les personnages se répartissent en types plus ou moins stéréotypés, dont il faudra établir un répertoire précis: le prisonnier nu aux bras entravés par des cordes et collés au buste, le soldat casqué au costume protégé de bandes de cuir cloutées qui le pousse, le dignitaire aux bandes cloutées similaires coiffé, pour sa part, d’une sorte de béret à pointes à la façon de la barrette des ecclésiastiques ou du mortier des magistrats Chaque type de figure comporte des caractéristiques iconographiques semblables et immuables qui ont la même structure – dont je pose ici quelques jalons préliminaires, mais qu’il conviendra ultérieurement d’analyser systématiquement en détail. Je me demande s’il n’en va pas de même pour chaque pièce d’incrustation, qui appelle de façon contraignante ses compléments, et si ce ne serait pas une clé pour aider à une meilleure compréhension de ces ensembles disloqués. Pour exemple, le type de militaire (fig. 5b, 7b–c) qui n’a été retrouvé que dans le «Palais» de Ville II8: le signe distinctif est la javeline portée pointe vers le bas, et dont la direction oblique se répercute obligatoirement sur son prolongement au niveau du buste, lequel constitue une autre pièce. En outre, l’arme est associée, 7 8

Par exemple la portion de bras AO 25742-c (L. 35; l. minimale du revers au poignet: 1 mm). À deux niveaux différents: P-1 cf. fig. 6-b et P-2: Espace 4.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Panneaux d’incrustation en coquille de Mari, Ville II

251

d’après l’exemple retrouvé en connexion, à un double pan frangé dont l’interprétation la plus plausible est celle des vêtements de l’ennemi vaincu dénudé (Parrot 1969: 204). Malgré des points communs, la structure (schème, matrice graphique) de la pièce M.466 est totalement différente (fig. 5a). Ces subtilités iconographiques, on l’aura compris, sont susceptibles d’élargir l’éventail typologique des figures et ainsi d’améliorer la compréhension de ces images.

5. La pièce d’incrustation comme signe iconique Si, comme le lot de l’Espace 4 du Temple-manufacture (niveau P-2) le laisse supposer, ces pièces d’incrustation étaient fabriquées en série, stockées en attente par catégories et que la chaîne opératoire nécessitait une certaine spécialisation, l’efficacité même du montage sur les panneaux requérait que l’artisan reconnaisse du premier coup d’œil chaque pièce par rapport au personnage entier à recomposer, ce qui va dans le sens des schèmes évoqués ci-dessus. Ceux-ci prenaient valeur de signe pour les artisans, mais aussi – différemment –, pour les spectateurs, car ils révèlent un sens aigu de l’observation et une aptitude à schématiser, à «croquer» la caractéristique significative d’une attitude symbolisant un statut: les bustes de prisonniers font preuve à cet égard d’une prouesse technique, car la main du second plan, plaquée, comme la première, au buste par les entraves, forme une excroissance à la découpe extrêmement fine (et fragile !) sur la bordure de la pièce, et va jusqu’à détailler les phalangines et les phalangettes (fig. 6)9. D’autres signes sont des indices syntaxiques, car ils créent le lien entre deux figures. Le plus remarqué (Parrot 1969: 204 et n. 2) est le pouce (de longueur exagérée) du militaire enserrant la nuque du prisonnier (cf. fig. 6a). Contrairement à celles du prisonnier, la main du soldat est une sorte de moignon sans doigts10, qui se justifie par le fait que les autres doigts sont cachés par le cou qu’elle saisit: l’absence de détail, ici, est donc aussi un signe. Rappelons, évoquées ci-dessus, les caisses de char incluant une peau ou un tissu (Couturaud 2013: 166), la javeline pointe en bas et ajoutons le tronçon de rênes passant dans l’échancrure du tablier de char (Couturaud 2013: 166: autre exemple; cf. fig. 3b), au contraire du procédé observable sur l’Étendard d’Ur, où les rênes flottent.

6. Éléments d’incrustation retrouvés en connexion: syntaxe assurée Ces indices sont d’autant plus importants que, à Mari, seuls deux ensembles de pièces ont été retrouvés en connexion.

9 Il en existe où cette seconde main n’est pas marquée: cela avait-il une signification ? 10 Vérification faite à la loupe binoculaire sur la pièce AO 25742-a

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

252

Muller

Fig. 6. Bustes de prisonniers: le signe iconique de la main de 2e plan serrée contre le buste. a. M. 438 (Étendard). h. 47; l. 31; ép. 4, 5 mm (© B. Muller). b. AO 25742-s (temple d’Ishtar ?). h. 21; l. 13; ép. 3 mm. (© B. Muller).

L’Étendard du temple d’Ishtar, restitué à partir de 57 pièces environ, a livré un dignitaire à la hache (buste avec tête et bras, jupe disjointe à côté) près du bord vertical droit du tableau: est ainsi assuré un type de militaire ainsi que le fond, fait de plaquettes de schiste noir de forme irrégulière juxtaposées, et le cadre, fait de baguettes dont le rang extérieur en pierre rouge était doublé par un rang intérieur en coquille (fig. 7a). Au niveau P-1 du Temple-manufacture (passage 49/52), le couple constitué du soldat portant le double pan frangé et du prisonnier nu qu’il tient par le cou a été retrouvé en connexion (fig. 7b–c). C’est évidemment maigre et n’assure en rien le nombre de registres ni la disposition des attelages. Et ce ne sont pas les exemples, plus monumentaux, du palais A de Kish ou du palais G d’Ebla, à la syntaxe différente, qui peuvent donner la solution pour les compositions des panneaux d’incrustation mariotes.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Panneaux d’incrustation en coquille de Mari, Ville II

253

Fig. 7. Fragments de panneaux d’incrustation de Mari retrouvés en connexion a. Étendard : dignitaire à la hache M. 474 bordé des baguettes de pierre rouge et de coquille de l’encadrement, in situ au sud du temple d’Ishtar (Parrot, 1956: fig. 32, détail). b. Soldat casqué poussant un prisonnier (M. 4785 et M. 4793) et portant les vêtements de celui-ci au bout de sa javeline pointe vers le bas: ensemble in situ (Temple-manufacture, niveau P-1, passage 52/49). (Parrot 1969: 204 fig. 12). c. Idem, ensemble restitué (Margueron 2004: fig. 280, cl. F. Pons).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

254

Muller

En conclusion, ces pistes de recherche ne doivent pas s’en tenir là, mais être systématisées de façon à ce que • soient mieux cernées les techniques de fabrication, qui témoignent d’un stade extrêmement abouti et attestent certainement des transferts de technologie autant que de matériaux, bien connus par ailleurs pour le IIIe millénaire, en provenance des régions du Golfe qui travaillaient la coquille depuis le Ve millénaire; • soient rediscutées sur des critères tangibles (schèmes) les analyses stylistiques de Rita Dolce (1978), dont les groupements restent globalement une référence; • soit prise en compte la totalité des 165 éléments figuratifs provenant de l’Espace 4 (P-2), même dans l’état actuel de la documentation en attendant que le matériel lui-même redevienne accessible; ce lot constitue en effet la seule référence sûre concernant la stratigraphie et les ateliers; • soient analysées selon la même méthode les pièces relevant des thèmes religieux ou cérémoniels; • soient considérées avec du recul les positions stratigraphiques, mal notées, on le sait, par les fouilles anciennes, mais non significatives stylistiquement dans le Templemanufacture en ce qui concerne les nacres. La facture de type agadéen, perceptible en particulier sur les prisonniers – longues jambes fuselées, rendu de la musculature – (Muller 2014: 287 et ill. 3 p. 289) pose évidemment la question de la date de la présence akkadienne à Mari, qui ne se perçoit par la céramique qu’au niveau incendié P-1: cela irait dans le sens d’une mainmise politique du pouvoir d’Akkad pendant une certaine période (Sargon ?), avant l’incendie qui mit fin à la Ville II (Naram-Sin réprime la révolte de Migir-Dagan ?) – mais ce n’est pas le lieu d’en discuter ici. Cette facture agadéenne caractérise non seulement la production retrouvée dans le Templemanufacture même, mais aussi dans le temple d’Ishtar (Étendard) (Muller 2014: 291), et peut-être celui de Shamash. En revanche, les temples d’Ishtarat et Ninni-zaza forment un groupe distinct, ce qui avait déjà été constaté, en particulier par Rita Dolce. Ceci pose alors la question: proviennentelles d’un autre atelier ou portent-elles la marque d’une autre époque ?

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Panneaux d’incrustation en coquille de Mari, Ville II

255

Bibliographie Aruz, J. (éd.), 2003. Art of the First Cities, The Third Millenium B.C. from the Mediterranean to the Indus. New York, The Metropolitan Museum of Art. Beyer, D., 2007. Les sceaux de Mari au IIIe millénaire, observations sur la documentation ancienne et les données nouvelles des Villes I et II. Akh Purattim 1, 231–260. Beyer, D./Jean-Marie, M., 2007. Le temple du DA III de la déesse Ninhursag à Mari: les dépôts votifs du Lieu Très Saint. Akh Purattim 2, 75–122. Calmeyer, P., 1967. Zur Rekonstruktion der «Standarte» von Mari. J.R. Kupper (éd.): La civilisation de Mari, XVe Rencontre Assyriologique Internationale organisée par le groupe François Thureau-Dangin (Liège, 4–8 juillet 1966). Bibliothèque de la Faculté de Philosophie et Lettres de l’Université de Liège, fascicule CLXXII. Paris. Les Belles-Lettres, 161–169. Cluzan, S./Butterlin, P. (dir.), 2014. Voués à Ishtar. Syrie, janvier 1934, André Parrot découvre Mari. Guides archéologiques de l’Institut français du Proche-Orient n° 1, Beyrouth, Presses de l’IFPO. Coqueugniot, E., 1993. Un atelier spécialisé dans le palais de Mari. M.A.R.I. 7, 205–250. Couturaud, B., 2013. Mise en scène du pouvoir au Proche-Orient au IIIe millénaire: étude iconographique du matériel d’incrustation en coquille de Mari (thèse inédite). Dolce R., 1978. Gli Intarsi Mesopotamici dell’epoca protodinastica. Rome, Istituto di Studi del Vicino Oriente. Margueron, J.-Cl., 2004. Mari, métropole de l’Euphrate, au IIIe et au début du IIe millénaire av. J.-C., 2004, Picard/ERC. –

2014. Mari Ville II: palais ou temple-manufacture? Actes du colloque international Mari, ni Est, ni Ouest? Damas 19–21 octobre 2010. Syria suppl. 2, 265–289.

Muller, B., 2014. Les éléments d’incrustation en coquille: situation, matériau, considérations techniques et esthétiques. Cluzan/Butterlin 2014, 285–294. Parrot, A., 1956. Mission archéologique de Mari. Vol. I: Le Temple d’Ishtar. Institut français d’archéologie de Beyrouth, BAH LXV. Paris. –

1967. Mission archéologique de Mari. Vol. III: Les Temples d’Ishtarat et de Ninni-zaza. Institut

–

1969. Les fouilles de Mari. Dix-septième campagne (automne 1968). Syria 46.3–4, 191–208.

français d’archéologie de Beyrouth, BAH LXV. Paris.

Béatrice Muller, CNRS, Paris.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 257–268

Paola Poli

Seals and Sealings from Tell Masaikh–Kar-Assurnasirpal. Some New Results Aim of the paper is to present some of the most significant pieces, among the glyptic material found at the site, in order to shed light on the role of a colony on the Syrian Lower Middle-Euphrates during the Neo-Assyrian period.

0. Introduction Kar-Assurnasirpal corresponds to the modern site of Tell Masaikh. It is located on the left bank of the lower Middle Euphrates, about 5 km upstream from ancient Terqa. M. G. Masetti-Rouault had directed the excavations there from 1996 till 2010.1 Although the site’s history is long, as attested by different archeological levels datable from the Halaf period until the Early Islamic era, the Neo-Assyrian occupation is certainly the most significant and that which has revealed the most impressive remains. The identification of the settlement as Kar-Assurnasirpal, the «Port of Assurnasirpal» is based on various kinds of data, such as the information found in Neo-Assyrian royal inscriptions and archives and the excavation results, among which an aniconic stele with a cuneiform inscription mentioning the name of the site is of particular interest.2

1

2

For the archaeological investigations see the reports published in Athenaeum – Studi di Letteratura e Storia dell’Antichità, Università di Pavia, every year from 2001 to 2009. See also Masetti-Rouault 2014, i.p. a; Masetti-Rouault/Salmon 2010: 385–395. A large fragment of an aniconic stele, with a cuneiform inscription, was found in area E of Tell Masaikh. The text, dedicated to the god Nabu, quotes the name of Adad-Bel, the donor, who was, at a certain point, the lord of the town. The inscription mentions the toponym «Kar-Assurnasirpal, on the bank of the Euphrates».

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

258

Poli

Founded as a colony during the reign of Assurnasirpal II, in the first half of the 9th century (phase NA0), the town was rebuilt under the government of Nergal-eresh, who at the beginning of the 8th century became the Governor of the Rasappa province (the archaeological level corresponding to phase NA1). Later, probably under the reign of Tiglath-phalazar III, at the end of the 8th century, Kar-Assurnasirpal was attacked by the Assyrian army, and immediately rebuilt following a typical Assyrian organization and taste (phase NA2). Seals and sealings of Tell Masaikh were brought to light in the excavation areas D, E, G, I, F. However, it must be remarked that most of the glyptic material, presented here, was found in fills or in similar archaeological contexts, and therefore not immediately meaningful for the chronological attribution. The collection of seals and sealings from Tell Masaikh is made up of about one hundred and fifty pieces, eleven of which are original seals: two are in stone, six in composite material and three in clay. There are around ninety cylinder seal impressions and about forty stamp seal impressions. In some cases, the same clay support presents a double sealing, with both impressions of cylinder and stamp seals. Some of the glyptic artefacts appeared in preliminary reports or in other papers (see in particular Poli 2010; 2015), but have not yet been published definitively. The objective of this contribution is to present an overview of the glyptic material discovered at the site, and to seek to underline the variety of cultures which characterizes the iconographic patrimony during the Neo-Assyrian period attested in the colony. The material is classified on the basis of the subject of the representation and only a selection of the most representative examples is presented here.

1. Cultic Scenes 1.1. Cultic Scenes with Anthropomorphic Divinities Among the various subjects the cultic scene, with all its iconographic variations, is the most attested among the ensemble of the glyptic material discovered at Tell Masaikh–KarAssurnasirpal. On sealing MK15 105 (fig. 1), an adoration scene with anthropomorphic figures recurs. A deity in long robe, facing right, stands on a couchant quadruped (maybe a bull?). He holds a bow in his right hand, and extends his left. Behind him, a little tasseled globe on a stand is reproduced. In front of him, a beardless worshipper, facing left, in long robe, with vertical straights and patterned borders is depicted. Between the two figures there is another tasseled globe on a stand; crescent on its left and Pleiades on its right. Of particular interest is the square-patterned dress worn by the worshipper, introduced at the beginning of the 8th

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Seals and Sealings from Tell Masaikh-Kar-Assurnasirpal

259

Fig. 1. Sealing MK105 105.

Fig. 2. Drawing of sealing MK13 225.

century (Collon 2001: 238). Beardless figures appear predominantly on the limmu-type seals of the late 9th and early 8th centuries. These are generally interpreted as the representation of eunuch-officials (Watanabe 1993). This particular type of representation, strictly connected with the central administration of the palace, recurs on other examples brought to light at the site (Poli 2010: fig. 1). The presence of the two impressions on the same sealing (cylinder and stamp) proves that the seal was still in use at least at the end of the 8th century.

1.2. Cultic Scenes with Non Anthropomorphic Divinities On another example, MK13 225 (fig. 2), the anthropomorphic rendering of the deity is replaced by a non-anthropomorphic divine image according to a manner of depiction which was particularly common during the last phase of the Assyrian period, in the 7th century (Herbordt 1992: 83; Ornan 2005: 147–149). Two characters stand on either side of a stand on which are set a tasseled spade of Marduk and the wedge or stylus of Nabu. On the left of the impression, a bearded worshipper in a fringed and square robe facing right, carries a tall, flaring cup with one hand and extends the other, palm up. In front, another worshipper in a

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

260

Poli

Fig. 3. Drawing of sealing MK 13 356.

Fig. 4. Drawing of sealing MK 13 434-433b.

long fringed robe with a sword at the waist; he rises one hand and extends the other with the palm upwards. Star in sky, vegetable element made up of a bottle brush with leaves, and wedge. In its general composition, the replacement of anthropomorphic deities by divine symbols, the scene reproduces a common theme of the 7th century. At the same time, a few iconographic details betray contamination with other traditions. The recipient carried by the worshipper finds, in fact, very few attestations in the Neo-Assyrian glyptic documentation3. A similar recipient recurs on Old-Babylonian and Neo-Babylonian cylinder seals and the personage that carries this kind of object is generally identified as a priest (Collon 1999: 24–25). Also of particular interest is the design of the plant. It looks, in fact, like a more elaborate version of the vegetal motifs, generally called «bottle brush tree» that recurs on soft stone seals depicting the king holding a bow with attendants, according to an iconography mainly attested in the 9th century. An attribution to the 7th century can be suggested to the original seal, on the basis of the general plan of the scene, but few details can remind of an earlier tradition or a foreign influence. On sealing MK13 356 (fig. 3), a cylinder seal with metal cap has been impressed twice. The reconstructed design reproduces two tasseled, globe-topped uprights (centre-dot-circle), resting on elaborate paneled stands, alternate with a worshipper and a series of full-motifs: a romp, a pomegranate, a crescent (?) and another symbol not completely preserved, which maybe a fish. The beardless worshipper stands, facing left, with one hand raised up, and the 3

I am very grateful to Professor Rita Dolce for drawing my attention on this iconographic detail.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Seals and Sealings from Tell Masaikh-Kar-Assurnasirpal

261

other at the waist level. He wears a horned head-dress and a long robe with a sword at the waist. The dress is decorated with two vertical bands of dots-in-double squares, which is a typical pattern of another tradition. It compares, in fact, both on Middle Assyrian (Porada 1948: n. 606e) such as Neo-Babylonian (see, for instance, Collon 2001: nn. 326–327; 329) seals and it generally characterizes the tunics worn by divinities or heroes. A similar observation can be proposed even for this artifact. An attribution to the 7th century can be suggested for the original seal, on the basis of the general composition of the scene, but a few details may recall an earlier or a contemporary southern tradition.

1.3.Cultic Scene with Banquet A scene that recurs at least on three examples is the banquet scene cut in drilled or modeled style. It is a very popular Assyrian subject. Watanabe (1993) suggests an Assyrian origin for many of the hard stone examples preserved. On impression MK13 434-433b (fig. 4), a worshipper wearing a long robe, raising one hand, the other is not preserved, stands facing a god. He raises one hand and holds something else, also lost now, with the other one. He sits on a chair, its high back ornamented by six drill-holes. Between the two figures is a cross-legged table on which rest a line (maybe bread ?) and a fish, a wedge. At the top are the crescent and the Pleiades. The seated person is interpreted as a god, whose identity is uncertain. The high-backed throne would favor a date in the second half of the 8th century, even though a later attribution is not to be excluded (Collon 2001: 67).

1.4. Cultic Scene with Tree Motif A very popular Neo-Assyrian scene is the adoration of the sacred tree. Even though the NeoAssyrian origin of the depiction is still debated, it is certain that such a representation has reached its largest diffusion in this period.4 On sealing MK 12 311 (Poli 2010: fig. 11; Poli 2015: figs. 2–3) the typical scene with two genii represented at either side of the tree is strongly influenced by the iconographic tradition of the end of the 2nd millennium by the presence of the bird on a branch of the tree. On another fragment sealing, MK09 027 (fig. 5), the tree is flanked on the right side by a couchant animal, with a figure probably represented posing his foot on its head, even though the image is a little deformed. The scene is framed, in its lower part, by a running spiral border. The tree is highly stylised, consisting of a single trunk and lateral branches ending 4

According to Parpola (1993: 165, note 24) the representation composed by the sacred tree flanked by the king or «mythical sages» is a «genuine Assyrian innovation». Collon (2001: 83) underlines a more ancient origin that goes back to the eighteenth century. Matthews (1990: 107) points out a Middle Assyrian origin of the representation, while Porter (1993: 137–138, n. 22) suggests a Cassite forerunner. For a brief reconsideration of the bibliography of the subject see Poli 2015.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

262

Poli

Fig. 5. Drawing of sealing MK09 027.

in pomegranates. It is a simplified version of the images depicted on two seals conserved in the British Museum (Collon 2001: nos. 151–152). Of a certain interest is the running-spiral border at the bottom of the impression. It is a very uncommon detail, that appears on Middle Assyrian seals (see for instance Porada n. 606e) and on a few Neo-Babylonian seals. It recurs on a seal of the Pierpont Morgan Library Collection (Porada 1948: no. 749) dated by Wittmann (1992: no. 35) to the 10th–9th centuries, and on another conserved in the British Museum and attributed by Collon (2001: no. 276) to the late 9th or early 8th century. Even for our example a chronological attribution to a period between the end of the 9th and 8th century can also be proposed. It is again interesting to note the permanence of iconographic elements of a more ancient repertoire, or the contamination with the NeoBabylonian tradition in a Neo-Assyrian scene. A different stylistic rendering and figurative interpretation of the iconography of the sacred tree recur on another example (MK09 300, Poli 2010: fig. 12), on which a very simplified tree is depicted but not inserted in an adoration scene. The seal is made of clay, namely a very humble material. It was probably an object of modest value, that, by the way, reflects a clear Assyrian tradition, probably adapted both in iconography and rendering to a non official object. It can be attributed to the 9th–8th century.

2. Scenes with Tree Motif On other artefacts, among the glyptic assemblage uncovered at Tell Masaikh, the tree sometime appears only as an element of the design of the scene.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Seals and Sealings from Tell Masaikh-Kar-Assurnasirpal

263

Fig. 6. Seal and modern impression MK01 17020.

On the seal MK01 17020 (fig. 6) the representation consists of two trees posed one next to the other. They recall in some way the depiction of the palmette tree and the rendering of the representation is much more naturalistic than artificial or symbolic. The depiction on the right side of the impression consists of a sort of central trunk from which branches detach irregularly; while the depiction on the left side of the impression is much more schematic. For this kind of iconographic motif I could not find an exact parallel. The closer depiction is a seal of the end of the 2nd millennium (Moortgat 1940: no. 559), although the tree on the seal from Tell Masaikh does not present dates. This kind of iconography is typical of the Babylonian tradition and is frequently attested at the end of the 2nd millennium or later, namely during the Neo-Babylonian period. The large size of the seal from Tell Masaikh, around 4 cm, suggest an attribution to the 9th century, or even earlier.

3. Hunting Scenes This subject is relatively rare on cylinder seals of the first centuries of the 1st millennium and it is generally restricted to seals in the Assyrian and Iranian style (Collon 2001: 59).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

264

Poli

Fig. 7. Drawing of sealing MK08 531.

Fig. 8. Drawing of sealing MK07 172.

As it has been already pointed out elsewhere, the origin and the dating of seal MK06 238 (Poli 2010: fig. 9) are highly uncertain. The lack of proportion of the figures and the carelessness in the carving of the seal shows that it was not produced in Assyria. One may suggest that it is a provincial work, coming perhaps from North-Eastern regions, dating from the 9th century or even earlier.. On sealing MK08 531 (fig. 7) an archer in a chariot, heading left, takes aim with an arrow, behind a goat or a small bull. Behind the archer, a charioteer stands with his arms extended to hold the reins. The two persons wear a long tunic and a feathered headdress. Only one horse is depicted. The chariot has a square box decorated with diagonal lines and a spear rising behind. An elliptical element below the hatch is clearly visible. This particular detail of the representation is common during the 9th century and continued to be attested during the reign of Tiglath-pileser III (744–727). The closer parallels to our impression are two seals of Babylonian origin, attributed to the 10th–8th century. The first one was found at Ur and published by Legrain (1951: no 611), the second belongs to the Bible+Orient Collection of the University of Fribourg and was published by Keel-Leu (2004: no 141z). A Babylonian origin can be suggested, and the same chronological attribution even for the original seal, which had been impressed on our sealing.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Seals and Sealings from Tell Masaikh-Kar-Assurnasirpal

265

4. Scenes with Animals and Monsters A figurative element that appears in various types of scenes, is the depiction of different kinds of animals and monsters. On seal MK07 421 (Poli 2015: fig. 6–7) two horned quadrupeds are represented flanking a tree. While the scene may be common in earlier periods, it is unusual in Neo-Assyrian period, as Parpola (1993: 165, n. 24) has pointed out. Nevertheless the seal should probably be dated to the Neo-Assyrian period, for its particular iconography of the plant between the two animals. Such a motif does not occur among Middle Assyrian materials, while it is very common on Neo-Assyrian seals. The seal can be dated to circa the 9th century, even though the subject of the representation is closer to the Middle Assyrian tradition. The fragmentary impression MK12 355 (Poli 2015: figs. 4–5) shows a very fine design of a standing winged quadruped (a horse or a bull). Matthews notes that the representation of the winged horse is not usual in Middle Assyrian time (Matthew 1992: 203–204), while in Neo-Assyrian this particular iconography is mainly attested in records engraved in the linear style which reproduce hunting scenes; this kind of seal is generally dated to the 9th century. The depiction on the seal impression from Tell Masaikh is not in the linear style, but, on the contrary, is characterized by a modelled carving style, which is close to the Middle Assyrian tradition. The impression is iconographically closer to the Neo-Assyrian tradition, the stylistic rendering seems to belong to the Middle Assyrian tradition. It can be attributed to the 10th–9th century. On the impression MK07 172 (fig. 8) a series of rampant goats looking back over their shoulders towards the right are reproduced. Two impressions of stamp seals without representation are also visible. A very close parallel, in the composition of the scene, was found at Ur (Legrain no 597). Wittmann (1992: 227) suggests a dating between the 11th to 8th century for this seal. Collon (2001: 56) prefers, for the same object, an attribution to the 9th–8th century and she proposes a Babylonian origin of the artifact. A similar chronological attribution can be argued for our example.

5. Stamp Seals The stamp seals and sealings, based on their themes, can be arranged in three main categories. Cultic scenes are classified in a first group: with anthropomorphic divinities; two persons standing on either side of the sacred tree; a divinity in a nimbus. A second group is characterized by the depiction of an animal. Of particular interest is a sealing which reproduces the image of the scorpion inserted in an oval form. It is a very frequent depiction in Neo-Assyrian time and it has been identified as a representation

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

266

Poli

connected with the royal harem (Herbordt 1992: 136–138). Other artifacts reproduce various kinds of quadrupeds, generally walking or running. Abstract representations or symbols are also common; they constitute the third group. All this material reflects a strong Assyrian tradition and it can be dated between the end of the 8th and the 7th century.

6. Conclusion From the selection of seals and sealings, presented here, it is possible to delineate the main aspects characterizing the glyptic material found in the Neo-Assyrian colony of Tell Masaikh–Kar-Assurnasirpal. Concerning the chronological attribution, a general dating between the 10th and the 7th century for the majority of the pieces can be confirmed. The iconographic repertoire is rich from different points of view, and allows further considerations. A conspicuous number of artefacts are clearly the expression of the Neo-Assyrian cultural tradition, as it to be expected in a colony situated in the middle of a province, in the heart of the empire, but actually in close connection with regions and societies which were closely influenced by Aramaic and Babylonian culture. Other examples are much more influenced by the Babylonian figurative patrimony; in some cases, one may argue for a Babylonian origin. Moreover, a group of seals and sealings are clearly imported artefacts; a phenomenon that proves an economic contact of the site with other regions, mainly with the eastern regions. One example can be interpreted as a clear form of emulation of an official artefact. A closer look at a few examples of the first group, which have been defined as Neo-Assyrian artistic products, reveals the presence of archaic features. This situation appears at two levels: most of such iconographic details seem to be borrowed from the Middle Assyrian tradition, while on an impression, attributed to the 7th century, an elaboration of a typical motif of the 9th century is identifiable5. It is also interesting to remark that a few iconographic details can be assigned to the Middle Assyrian tradition as well as to the Neo-Babylonian.6 On the basis of the assumption that artistic objects are the manifestation of a cultural background, it is evident that the adaptation of certain iconographies is the reflection of a wider cultural phenomenon. 5

6

At the moment it is not possible to argue a local production, namely an expression of a local artistic tradition, because at the site, until now, figurative documentation reproduced on other supports has not been found. The considerations proposed by D. Collon (1987: 80) are of a certain interest. She underlines, without giving an explanation, the fact that the Neo-Babylonian seals are much more influenced by the Middle Assyrian tradition than the Neo-Assyrian.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Seals and Sealings from Tell Masaikh-Kar-Assurnasirpal

267

Nevertheless, it is difficult to establish the source of inspiration for material discovered at Tell Masaikh: whether the Middle Assyrian or the Neo-Babylonian figurative patrimony, and that they can be the result of the same cultural situation not be excluded. One may argue that it is an artistic expression of a provincial work, on the basis of the general conviction that in marginal areas new fashions are adopted and assimilated later and older traditions persist much longer than in the centre (the result is the persistence of the Middle Assyrian tradition). In the same way, it can be suggested that in the same provinces artistic foreign influences and artistic contaminations can be adopted much more easily than in the capitals (the result is the presence of the Neo-Babylonian elements). But the glyptic material examined here may also testify to another phenomenon, which, for the moment, is much more difficult to investigate. It seems that in some regions of the Neo-Assyrian empire – which can be Tell Masaikh as well as other centres with a similar political situation – some elites seem to wish to detach themselves from the central imperial power. While being strictly bound to the Neo-Assyrian iconography and style (the subject, the composition of the scene are absolutely Neo-Assyrian), only few features (such as a bird on a three, a decorative motif) belong to another tradition, to show their ideological position. Therefore, this particular artistic expression can be interpreted as a sign of a kind of «colonial» attitude, trying to find a new cultural autonomy from the central seat of power. The study of the glyptic material uncovered at the site of Tell Masaikh is still in progress and, therefore, it is not possible to suggest a conclusive hypothesis. However, it must be noted that if the chronological attribution of the material here presented is correct, the attempt of cultural detachment of some members of the local elites from the central power is attested throughout the period of the Assyrian colonization, from the 9th until the 7th century.

Bibliography Collon, D., 1987. First Impressions: Cylinder Seals in the Ancient Near East, London. –

1999. Depiction of Priests and Priestesses in Ancient Near East. K. Watanabe (ed.) Priests and Officials in the Ancient Near East. Paper of the Second Colloquium on the Ancient Near East – The City and its Life, held at the Middle Eastern Culture Center in Japan (Mitaka, Tokyo), March 22nd–24th, 1996, Heidelberg, 17–46.

–

2001. Catalogue of the Western Asiatic Seals in the British Museum – Cylinder Seals V. NeoAssyrian and Neo-Babylonian Periods. London: British Museum Publications.

Herbordt, S., 1992. Neuassyrische Glyptik des 8.–7. Jh. v. Chr. State Archives of Assyria Studies 1, Helsinki.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

268

Poli

Keel-Leu, H./Teissier, B., 2004. Die vorderasiatischen Rollsiegel der Sammlungen «Bibel+Orient» .

der Universität Freiburg Schweiz. The Ancient Near Eastern Cylinder Seals of the Collections «Bible+Orient» of the University of Fribourg. OBO 200. Fribourg. Legrain, L., 1951. Seal Cylinders (Ur Excavations X), London/Philadelphia, Pa. Masetti-Rouault, M. G., 2014. L’évolution d’une colonie néo-assyrienne dans le bas Moyen-Euphrate syrien (9e–8e siècle av. J.-C).: recherches archéologiques et historiques récentes à Tell Masaikh. L. Marti (ed.), La famille dans le Proche-Orient ancien: réalités, symbolismes, et images. Proceedings of the 55th Rencontre Assyriologique Internationale at Paris 6–9 July 2009. Winona Lake, 689–700. –

i. p. (a). Rapport préliminaire concernant les travaux de la mission française à Tell Masaikh, 13 septembre – 10 novembre 2006. Akh Purattim 3.

Masetti-Rouault, M. G./Salmon, S., 2010. The Neo-Assyrian Colony of Tell Masaikh, in the Syrian Lower Middle Euphrates Valley and its region: a Report about the last excavations. P. Matthiae et al. (eds.), Proceedings of the 6th International Congress on the Archaeology of the Ancient Near East, University La Sapienza, Rome 5th–10th May 2008. Vol II. Wiesbaden, 385–396. Matthews, D. M., 1990. Principles of Composition in Near Eastern Glyptic of the Later 2nd Millennium. Orbis Biblicus et Orientalis 8. Fribourg/Göttingen. Moortgat, A., 1940. Vorderasiatische Rollsiegel, Berlin. Ornan, T., 2005. The Triumph of the Symbol. Pictorial Representation of Deities in Mesopotamia and the Biblical Image Ban. OBO 213, Fribourg/Göttingen. Parpola S., 1993. The Assyrian Tree of Life: Tracing the origins of Jewish monotheism and Greek philosophy. JNES 52, 161–208. Poli, P., 2010. The Neo-Assyrian Glyptic from Tell Masaikh: Preliminary Results. P. Matthiae et al. (eds.), Proceedings of the 6th International Congress on the Archaeology of the Ancient Near East, University La Sapienza, Rome 5th–10th May 2008. Vol I. Wiesbaden, 961–972. –

2015. A Group o Seals and seals Impressions from Tell Masaikh-Kar-Assurnasirpal with more ancient Motifs. A. Archi (ed.), Tradition and Innovation in the Ancient Near East. Proceedings of the 57th Rencontre Assyriologique Internationale at Rome 4–8 July 2011. Winona Lake, 289–300.

Porada, E., 1948. Corpus of Near Eastern Seals in North American Collections I. The Pierpont Morgan Library Collection (The Bollingen Series 14). Washington, D.C. Porter, B., 1993. Sacred Trees, Date Palms, and the Royal Persona of Ashurnasirpal II. JNES 52, 129–139. Watanabe, K., 1993. Ein neuassyrisches Siegel des Mīnu-ahti-ana-ištari. BaM 24, 199–308. Wittmann, B. 1992. Babylonische Rollsiegel des 11.–7. Jahrhunderts v. Ch. BaM 23, 169–289.

Paola Poli

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 269–282

Agnese Vacca

New Data on the EB III–IVA1 of North-Western Syria in the light of Old and Recent Excavations at Tell Mardikh/Ebla and Tell Tuqan In this paper I will focus on the analysis and definition of the EB III–IVA1 stratigraphic and architectural sequences at Tell Mardikh/Ebla and Tell Tuqan, investigated during old and recent excavations. A study of stratified pottery materials is presented, providing a useful relative chronology for the definition of early 3rd millennium BC phases in North-Western Syria.

1. Introduction The Syrian Early Bronze Age is generally divided into five main phases (EB I–III, roughly corresponding to the first half of the 3rd millennium BC, and EB IVA–B covering the latter part of the period), although for some areas this terminology is avoided in favor of a periodisation built upon regional sequences, such as those standing for the Jezirah and Middle Euphrates regions (Pfälzner 1997; Lebeau 2000; Cooper 2006; Porter 2007). Such periodisations, covering the whole 3rd millennium BC, have been developed as a main consequence of long-term and salvage excavations carried out in a large number of sites within the Middle Euphrates, the Balikh and the Khabur basins, which have provided a consistent documentary base to define a relative chronology based on stratigraphy and pottery sequence, anchored to radiocarbon dates. In recent years, under the coordination of the ARCANE Project, such sequences were used to lay the foundations for a cross-regional synchronization of Near East and Eastern Mediterranean areas for the 3rd millennium BC (Lebeau 2011). With regard to North-Western Syria, the chronology is still imperfectly understood: whereas the sequence of the last quarter of the 3rd millennium BC (EB IVA–B) is fairly well outlined, the definition of the early 3rd millennium BC phases is still a problematic task, due to the lack of extensively excavated sites and published sequences (Akkermans/ Schwartz 2003: 224–225; Matthiae 2013). In this respect, the Amuq sequence is still the best

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

270

Vacca

Fig. 1. Tell Mardikh/Ebla: localization of EB III and EB IVA 1-2 structures on the Acropolis.

reference point for the first half of the 3rd millennium BC in North-Western Syria, providing links between coastal, inland Syria and south-eastern Anatolia (Braidwood/Braidwood 1960). Recent excavations at Tell Mishrifeh/Qatna (Morandi Bonacossi 2008), Tell Qarqur (Dornemann 2008, 2012) and Tell Al-Ṣūr (Mouamar 2014) provided new data on the EB I– III periods, although most of the results are still being published. According to this framework, a new analysis of data from old and recent excavations at Tell Mardikh/Ebla and Tell Tuqan will provide the opportunity to better define the chronology of the early 3rd millennium BC in North-Western Syria. In the last decades, the archaeological excavations, carried out at several locations on the Ebla Acropolis, brought

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Data on the EB III-IVA1 of North-Western Syria …

271

Fig. 2. Tell Mardikh/Ebla, Building G2 (EB III). © MAIS.

to light a seamless stratified sequence of several overlapping floors pertaining both to Palace G itself (EB IVA) and to two monumental structures preceding the construction of the palace (Buildings G2 and G5), dating back to EB III (c. 2750/2700–2550 BC) and EB IVA1 (c. 2550–2450 BC) respectively (Matthiae 1987, 2000; Dolce 2009) (fig. 1). In addition to Ebla, a long EB III stratified sequence has been recently investigated at Tell Tuqan, in a deep sounding opened in the lower town (Area P South) (Peyronel 2008, 2011) (figs. 5–6). The two sites provide different kinds of data: excavations at Ebla yielded a nearly continuous sequence of monumental structures identified in different sectors of the Acropolis (Mardikh IIA2–3), while the sounding at Tell Tuqan, even though limited in extension, showed a continuous stratigraphy of 5 main sub-phases all dating to EB III (Tuqan IC1–3), and a further sub-phase dating to EB IVA1/early EB IVA2 (Tuqan IIA1). The analysis of the stratigraphy at both sites, together with the realization of a detailed ceramic sequence, allows to outline a seriation for the periods EB III–IVA1, which finds comparisons with other regional sequences, such as those of Tell Mastuma, Tell Afis, Tell Suffane and the Maath plain survey.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

272

Vacca

2. The EB III–IVA1 at Tell Mardikh/Ebla and Tell Tuqan: Stratigraphy and Architecture 2.1. Tell Mardikh/Ebla Traces of EB III occupation have been identified, at Tell Mardikh/Ebla, in different sectors of the Acropolis and, in particular, on the south-western (Building G2) and south-eastern (Area CC) slopes (fig. 1). In 1983–84 the excavation of the South Unit of the Palace’s G Central Complex, on the southern slopes of the Acropolis, led to the identification of an ancient monumental structure, named Building G2, dating back to EB III (Matthiae 1987: 136–39, 2000: 572–78). A large portion (ca. 175 m 2) of this building was investigated (fig. 2). Building G2 consists of 16 small and medium-sized rooms, usually 4 meters in length and width. Walls are preserved for a maximum height of 80 cm and are made of orange and yellow rectangular mud-bricks of 60x40 cm, and floors are carefully plastered. Building G2 shows two superimposed phases of occupation dating to EB III; the earliest structure is the best preserved of the two, and yielded abundant pottery materials, found in situ in all the 16 rooms, mainly pertaining to storage jars full of charred cereals (fig. 7.12). The storage facilities uncovered in Area CC are contemporary to the first phase of Building G2, and are located 40 meters further east of Building G2 on the south-eastern slopes of the Acropolis (Matthiae 2000: 572) (fig. 1). Here several rebuilding episodes have been detected within the sequence: two phases with large circular storage pits, likely part of a larger open-air stockpiling area (CC 5–6), and a later phase characterized by the construction of a free-standing mud-brick granary associated with other subterranean circular silos (CC-7). As pointed out by P. Matthiae, the remains of mud-brick rooms in areas CC and G2 are exactly at the same level and show a very similar building technique, suggesting they are part of the same building, probably the peripheral region of a public complex, including residential rooms and warehouses (Matthiae 2000: 572). Further, scattered evidence of storage facilities dating to EB III comes from the north-western edge of the Acropolis (beneath the West Unit of the Royal Palace G Central Complex), where a mud-brick silo with an associated floor was uncovered (Matthiae 1993: 619) (fig. 1). In a later phase of use, still dating to EB III, Building G2 is refurbished; floors are renewed, and mud brick walls are added to the structures. While Building G2 maintains its primary storage function as in the previous phase, the storage facilities in Area CC fall into disuse. During the final phase of use, the eastern sector of Building G2 shows traces of partial destruction by fire, while the western sector seems intentionally abandoned. In fact, while in the western sector of the building rooms have been found empty, in the eastern sector a thick layer of collapsed mud-bricks and ashes was found sealing the deposit inside the rooms.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Data on the EB III-IVA1 of North-Western Syria …

273

Fig. 3. Tell Mardikh/Ebla, Area CC (EB III). © MAIS.

The EB III phase (including both earlier and later phases of use of Building G2) is documented in several further find-spots across the Acropolis, which might indicate that the central mound was extensively settled during this period (fig. 1). One relevant piece of evidence comes from the north-western edge of the Acropolis. Here a large Iron Age II pit has cut both the Royal Hypogeum G4 (EB IVA2) and Building G5 (dating to EB IVA1), revealing a stratigraphic sequence underneath the Royal Palace. On the northern section of the pit a thick mud bricks wall is visible, which is cut by the construction of the Royal Hypogeum (fig. 4). A dating of this wall to EB III can be maintained by both its stratigraphic position and associated materials; in fact, the wall is clearly covered by the earliest floor of EB IVA1 Building G5, and all associated sherds date to EB III. Such an evidence suggests the existence of an EB III monumental structure on the north-western edge of the Acropolis and on its south-western slopes (corresponding to Building G2), as well as showing the stratigraphic overlap between the EB III and EB IVA1 phases.1 The same stratigraphic sequence is observable above the remains of Building G2, which are covered, at some points, by fragmentary floors with associated EB IVA1 pottery materials sealed, in turn, by the floors of 1

The adoption of the EB IVA1 terminology stems from the definition provided by S. Mazzoni (1991, 2002).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

274

Vacca

Fig. 4. Tell Mardikh/Ebla, Building G5 (EB IVA1). © MAIS.

Royal Palace G. In Area CC no EB IVA1 structures have been identified: materials dating to this phase come from several later pits, which have deeply cut the underlying EB III–IVA1 stratigraphy. The EB IVA1 phase is well documented in the stratified sequence uncovered below the West Unit of Royal Palace G (Mazzoni 1991; Matthiae 1993; Vacca 2014) (fig. 1). The excavation revealed a monumental structure, named Building G5, with thick walls and hard lime floors carefully laid (fig. 4). Two phases of occupation can be recognized, corresponding to two overlapping floors with abundant in situ ceramic materials all dating to EB IVA1 (fig. 7.16–27). Immediately to the west of Building G5, on the bottom of a large Iron Age pit, a floor and a silo dating to the same phase (S.4843) have also been identified. The silo (3 m in diameter) contained a large quantity of charred seed, mainly pertaining to barley (Wachter-Sarkady 2013). Further west, another room (L.4470), likely belonging to the same complex, has been identified underneath the massive retaining wall of Royal Palace G (Mazzoni 1991). The latest phase of occupation of Building G5, corresponding to a third floor refurbishment, can be dated, on the base of associated pottery materials, to an early moment of the EB IVA2 (i.e., an earlier phase of the Royal Palace G pottery horizon). Rooms of Building G5 were ultimately sealed with thick mud-bricks and earth fillings, over which the first floor of Royal Palace G was built. Pottery materials from the obliteration levels of Building G5 date to Palace G horizon (Vacca 2015: 9–10).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Data on the EB III-IVA1 of North-Western Syria …

275

Fig. 5. Tell Tuqan, Area P South Phases 8A and 7. © MAIS.

2.2. Tell Tuqan The excavation in Area P South (Lower City North) at Tell Tuqan provided a long sequence, covering the whole time-span between the Iron and the Early Bronze Age. The EB III stratigraphic sequence (corresponding to Tuqan IC in the site periodisation) consists of several superimposed structures and associated deposits, for a total thickness of about 3 meters (Peyronel 2008, 2011). The internal chronological sequence of phase Tuqan IC is articulated into three main distinct ceramic phases, each embracing one or more architectural phases (Tuqan IC1–3; Vacca 2014). Phase Tuqan IC1 includes a ceramic workshop (Phase 10) and a level with cobbled floor and beaten earth surfaces, cut by large refuse pits filled with ashy layers (Phase 9). These phases correspond to the earliest levels exposed in the sequence. A neat chronological assessment is problematic due to the scarce quantity of materials, and the absence of earlier excavated evidence makes it difficult to define the upper limits of Tuqan

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

276

Vacca

Fig. 6. Tell Tuqan, Area P South Phase 6B. © MAIS.

IC1. However, an EB III, or even earlier (EB II/III), date can be suggested on the base of some ceramic types (Vacca 2014). Phases Tuqan IC2-3 encompass the architectural Phases 6–8; the ceramic assemblage is perfectly comparable to that documented at Tell Mardikh/Ebla, and precisely with that pertaining to the first phase of use of Building G2 and from Area CC. During Phase 8, the area is equipped with mud-brick structures, installations and working surfaces, probably connected to handicraft activities or to the processing of agricultural products, as suggested by the recovery of a large number of charred seeds (fig. 5). Phases 6–7 are characterized by a change in the function of the area, as indicated by the presence of silos and plastered surfaces related to storage and food processing activities (figs. 5–6). Charred plant remains (such as barley, wheat and olives) have been collected from the burnt deposits inside the silos and in the area between them. The latest Phase 6A, which corresponds to the abandonment of the area (with thick layers of collapsed debris and mud-bricks), is sealed by a compact layer of clay in which two infant burials were cut (Phase 5, EBIVA1/early EB IVA2; Peyronel 2011).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Data on the EB III-IVA1 of North-Western Syria …

277

Fig. 7. EB III-EB IVA1 pottery from Tell Mardikh/Ebla (TM) and Tell Tuqan (TT). Nos. 1-15: EB III; 16-27: EB IVA1. © MAIS.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

278

Vacca

3. The EB III–IVA1 at Tell Mardikh/Ebla and Tell Tuqan: the Ceramic Evidence 3.1. EB III The EB III phase is characterized, both at Ebla (Building G2, Area CC) and Tell Tuqan (Tuqan IC), by a coherent pottery assemblage, marked by the recurrence of peculiar shapes, which appear in statistically significant quantities through the whole period. The most represented ware type is Simple Ware (70%), followed by Kitchen (19%) and Painted Simple Ware (8%), while Late Reserved Slip and Incised Wares are attested only in low percentages (3%). Simple Ware fabrics are generally mineral tempered, highly fired, with colour ranging from pinkish to pale yellow-brown, suggesting controlled firing temperatures. Among open shapes the main types attested are: large shallow or deep hemispherical bowls (fig. 7.1–3), large squat platters (fig. 7.4–5) and small bowls with curved profiles and everted rim, or with flaring wall and thin rounded lip (fig. 7.6–10). Among closed shapes, jars with outflared rim and rounded or modelled lips are quite characteristic, often bearing potmarks incised on the shoulder (fig. 7.11–12). Painted Simple Ware is attested in low percentage both at Ebla and Tell Tuqan, with vessel profiles falling in the range of Simple Ware. The decoration consists of a thin fading monochrome paint, with colour ranging from red-orange to brown and black, while decorative motifs mainly consist of cross-hatching patterns framed by horizontal bands (fig. 7. 8, 13). Similar decorative motifs occurring on Painted Simple Ware also recur on Incised Ware, such as wavy lines or cross-hatching patterns. Kitchen Ware includes different pot types: with everted rounded rim or with vertical stepped rim, the latter exclusively attested at Tell Tuqan (fig. 7.14–15). Pastes are red-brown in colour and show a high percentage of calcite aggregates, beside quartz, gehlenite and hematite. Among imported pottery few fragments of Red-Black Burnished Ware are documented both at Ebla and Tell Tuqan (Vacca 2014). A certain number of comparisons can be established with the nearby site of Tell Mastuma, located 15 kilometres to the west of Ebla (Tsuneki 2009). The few stratified materials are strictly comparable with those attested at Tuqan and Ebla, and contribute to the definition of the regional ceramic horizon (Vacca 2014: fig. 5). The EB III pottery horizon shows affinities, in particular, with the western area; several comparisons can be singled out with Phase H of the Amuq (excluding level JK3:11 with cyma recta bowls; Braidwood/Braidwood 1960: figs. 269.9–10, 270.8, 271.2), while to the South, comparisons can be established with sites of the Orontes Valley and the Syrian steppe, such as Hama K1-4 (Fugmann 1958: figs. 49, 54), Tell Mishrifeh/Qatna Period II (Besana et al. 2008) and Tell Al-Ṣūr VIII (Mouamar 2014). To the east some parallels can be made with sites located along the Tabqa Dam area,

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Data on the EB III-IVA1 of North-Western Syria …

279

such as Tell es-Sweyhat Period G (Danti/Zettler 2007: fig. 11.5) and Tell Bi’a graves (such as Grabbau 3 and 6; Strommenger/Kohlmeyer 1998).

3.2. EB IVA1 The EB IVA1 ceramic horizon (well documented in the stratified sequence uncovered in Building G5 at Ebla) has been defined by S. Mazzoni (1991, 2002) as an emergent phase, characterized by the introduction of new pottery shapes, such as the corrugated goblets and the hole-mouth cooking pots, which will become peculiar of the whole EB IVA period (fig. 7.16–17, 19, 24–27). Diagnostic types are the corrugated goblet with inner-stepped rim and with short everted rim (fig. 7.16–19). Among bowls, rounded or carinated profiles prevail (fig. 7.20–22). Along with new shapes introduced at this phase, such as fine corrugated wares (which represents ca. 1/3 of Building G5 pottery assemblage), some other vessel types continue from the previous EB III period, such as large platter-bowls, which totally disappear in later EB IVA2 (fig. 7.23). This EB IVA1 horizon can be recognized in several excavated sites within the Northern Levant, such as Tell Qarqur, Area A, stratum 14 (Dornemann 2008: fig. 3.1, 3–4, 15–16, 25), Hama J7–8 (Fugmann 1958: figs. 58, 62), Tell Al-Ṣūr VIIA (Mouamar 2014) and Mishrifeh IIIA period (J38–36, Besana et al. 2008), whereas in the Amuq plain this phase corresponds to an early Phase I (Braidwood/Braidwood 1960). Several surveyed sites in the Math and Jazr plains, such as Tell Atchaneh and Tell Suffane (Mazzoni 2006: fig. 4.k) yielded some EB IVA1 materials, allowing to outline a regional North Inner Syria ceramic province.

4. Conclusions The long and articulated sequence identified at Ebla testifies to a continuous occupation of the Royal Palace G area since, at least, the EB III period. The reconstruction of the stratigraphic sequence allows to identify two main architectural phases underlying Royal Palace G, each related to major architectural episodes together with changes in the pottery repertoire: an EB III phase (corresponding to Building G2 and the remains in Area CC) and an EB IVA1 phase (represented by Building G5). The evidence, on the central mound, of large storage facilities during both EB III and EB IVA1 suggest a centralized management of primary products, at least from the topographical point of view, probably performed by emerging urban elites, who based their power upon the ability to control agricultural products. The location and the monumental character of Buildings G2 and G5 suggest a public function for both structures. Evidence from Tell Tuqan, albeit from a limited excavation, are particularly helpful in defining ceramic phases and outlining a comparative sequence. Noteworthy is the change in the use of the area between Phases 8 and 7. While during Phase

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

280

Vacca

8 the excavated evidence (mud-brick structures, a small kiln and plastered vats containing charred seeds) may suggest the processing of agricultural products at the household level, it is during the last two phases (6–7) that the area seems devoted to the intensive warehousing of primary resources, as indicated by the presence of large storage pits, silos, and plastered surfaces for crop activities, which occupied the whole excavated area. A similar evidence of intensive storage of agricultural products beyond the domestic level is also documented at Tell Mishrifeh/Qatna, where numerous large granaries, silos and storage pits, with associated facilities for crop processing, have been discovered on the summit of the Mishrifeh Acropolis (Operation J, Morandi Bonacossi 2008). Data from Ebla and Tell Tuqan contribute to both the chronological and interpretative framework outlined for the early 3rd millennium BC. The existence of centralized collection and control of agricultural surplus by emerging elites, if confirmed by future researches, will shed new light on the still little-known framework of the formative urban stages in North Western Syria.

Bibliography Akkermans, P. M. M. G/Schwartz, G. M., 2003. The Archaeology of Syria: From Complex HunterGatherers to Early Urban Societies (ca. 16,000–300 BC), Cambridge. Besana, M./Da Ros, M./Iamoni, M., 2008. Excavations on the Acropolis of Mishrifeh, Operation J. A New Early Bronze Age III – Iron Age III Sequence for Central Inner Syria. Part II: the Pottery. Akkadica 129, 79–130. Braidwood, R.J./Braidwood, L., 1960. Excavations on the Plain of Antioch I. The Early Assemblages. Phases A–J. Oriental Institute Publications 41, Chicago. Cooper, L., 2006. Early Urbanism on the Syrian Euphrates. New York/London. Dolce, R., 2009. The Archaeology of a Long Lasting Power: The Ebla Paradigm. A Synthesis. Studi Micenei ed Egeo-Anatolici 51, 251–278. Danti, M. D./Zettler, R. L., 2007. The Early Bronze Age in the Syrian north-west Jezireh: the Tell es-Sweyhat region. E. Peltenburg (ed.), Euphrates River Valley Settlement. The Carchemish Sector in the Third Millennium B.C. Levant Supplementary Series 5. Oxford, 164–184. Dornemann, R. H., 2008. Evidence from Two Transitional Periods, Early Bronze IV and Iron I at Tell Qarqur. H. Kühne et al. (eds.), Proceedings of the 4th International Congress of the Archaeology of the Ancient Near East: 29 March – 3 April, Freie Universität Berlin, vol. 2. Wiesbaden, 81–96. –

2012. The Qarqur Challenge: The Bronze Age and Earlier. Near Eastern Archaeology 75/4, 220– 230.

Fugmann, E., 1958. Hama. Fouilles et recherches de la Fondation Carlsberg, 1931–1938. II.1. L’architecture des périodes pré-hellénistiques, Nationalmuseets Skrifter, Større Beretninger 4, Copenhague.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

New Data on the EB III-IVA1 of North-Western Syria …

281

Lebeau, M., 2000. Stratified Archaeological Evidence and Compared Periodizations in the Syrian Jezirah during the Third Millennium BC. C. Marro/H. Hauptmann (eds.), Chronologie des Pays du Caucase et de l’Euphrate aux IVe–IIIe millénaires, Actes du Colloque d’Istanbul, 16–19 décembre 1998. Institut Français d’Etudes Anatoliennes. Varia Anatolica 11. Istanbul/Paris, 167–192. Lebeau, M. (ed.), 2011. Arcane. Jezirah (= ARCANE 1), Turnhout. Matthiae, P., 1987. Les dernières découvertes d’Ebla en 1983–1986. Comptes rendus des séances de l’Académie des Inscriptions et Belles Lettres 1987/1, 135–161. –

1993. L’aire sacrée d’Ishtar à Ebla: résultats de fouilles de 1990–1992. Comptes Rendus de

–

2000. Nouvelles fouilles à Ébla (1998–1999): Forts et palais de l’enceinte urbaine. Comptes rendus

l’Académie des Inscriptions et Belles-Lettres, 613–62. des séances de l’Académie des Inscriptions et Belles-Lettres 2000/2, 567–610. –

2013. The Third Millennium in North-Western Syria: Stratigraphy and Architecture. W. Orthmann et al. (eds.), Archéologie et Histoire de la Syrie. La Syrie de l’époque néolithique à l’âge du fer. Wiesbaden, 181–198.

Mazzoni, S., 1991. Ebla e la formazione della cultura urbana in Siria. La Parola del Passato 46, 163–94. –

2002. The Ancient Bronze Age Pottery Tradition in Northwestern Central Syria. M. Al-Maqdissi et al. (eds.), Céramique de l’Âge du Bronze en Syrie, I: La Syrie du Sud et la Vallée de l’Oronte. Bibliothèque Archéologique et Historique 161. Beyrouth, 69–96.

–

2006. Tell Suffane, an Early and Middle Bronze Age Site in the Idlib Plain. Baghdader Mitteilungen 37, 381–392.

Morandi Bonacossi, D., 2008. Excavations on the Acropolis of Mishrifeh, Operation J. A New Early Bronze Age III – Iron Age III Sequence for Central Inner Syria. Part 1: Stratigraphy, Chronology and Architecture. Akkadica 129, 55–127. Mouamar, G., 2014. Tell Al-Ṣūr/Al-Sankarī: une nouvelle agglomération circulaire du Bronze ancien IV à la lisière de la steppe syrienne. F. Baffi et al. (eds.), Tell Tuqan Excavations and Regional Perspectives. Cultural Developments in Inner Syria from the Early Bronze Age to the Persian/ Hellenistic Period, Proceedings of the International Conference, May 15th–17th 2013 Lecce. Galatina (Lecce), 95–114. Peyronel, L., 2008. The Lower Town. Area P. F. Baffi (ed.), Tell Tuqan. Excavations 2006–2007. Galatina (Lecce), 21–70. –

2011. The Lower Town. Area P. F. Baffi (ed.), Tell Tuqan. Excavations 2008–2010. Galatina (Lecce), 61–110.

Pfälzner, P., 1997. Wandel und Kontinuität im Urbanisierungsprozeß des 3. Jtsds. v. Chr. in Nordmesopotamien. G. Wilhelm (ed.), Die orientalische Stadt: Kontinuität, Wandel, Bruch. 1. Internationales Colloquium der Deutschen Orient-Gesellschaft, 9.–10. Mai 1996 in Halle/Saale. Saarbrücken, 239–265. Porter, A., 2007. The ceramic assemblage of the Third Millennium in the Euphrates region. M. al-Maqdissi et al. (eds.), La céramique de l’Age du Bronze en Syrie II. Beyrouth, 3–22.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

282

Vacca

Strommenger, E./Kohlmeyer, K., 1998. Tall Bi’a/Tuttul I. Die altorientalischen Bestattungen, Saarbrücken. Tsuneki, S., 2009. Neolithic and Early Bronze Age Layers in Square 15Gc. T. Iwasaki et al., Tell Mastuma: An Iron Age Settlement in Northwest Syria. Memoirs of the Ancient Orient Museum 3. Tokyo, 69–88. Vacca, A., 2014. The Tuqan IC Pottery Sequence from Area P South. F. Baffi et al. (eds.), Tell Tuqan Excavations and Regional Perspectives Cultural Developments in Inner Syria from the Early Bronze Age to the Persian/Hellenistic Period, Proceedings of the International Conference May 15th–17th 2013 – Lecce. Galatina (Lecce), 45–84. –

2015. Before the Royal Palace G. The Stratigraphic and Pottery Sequence of the West Unit of the Central Complex: the Building G5. Studia Eblaitica 1, 1–32.

Wachter-Sarkady, C., 2013. Consuming Plants. Archaeobotanical Samples from Royal Palace G and Building P4. P. Matthiae/N. Marchetti (eds.), Ebla and its Landscape: Early State Formation in the Ancient Near East. Walnut Creek 2013, 376–402.

Agnese Vacca, Sapienza University of Rome.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Turkey

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 285–292

Gulan Ayaz

The Early Iron Age Jewellery from the Karagündüz Necropolis in Eastern Anatolia and Its Relationship to Southern Caucasia Karagündüz is an Early Iron Age (EIA) necropolis in Eastern Anatolia. The present study revisits the jewellery finds including bracelets, rings, earrings, and pins comparing them with similar finds from other sites in Southern Caucasia of the Early Iron Age. Lake Van basin in Eastern Anatolia houses fertile plains, home of ancient settlements. Karagündüz Tombs are located at the fertile Erçek Düzü plain watered by the Memedik River (fig. 1). About 1.5 km west of the tombs is the Karagündüz Höyüğü first identified by Charles Burney in 1956 (Burney 1958: 192). Tombs were identified in 1991 and excavated by Van Museum Directorate and Istanbul University Van Region History and Research Centre under the scientific supervision of Prof. Dr. Veli Sevin from 1992 to 1996. After the tombs, the same team continued with the excavations at the höyük. A calibrated C14 dating from the building layer 5 of the höyük gave 1092–956 BC. Tombs dated to the Early Iron Age were thought to be linked with this building layer of the höyük (Sevin 2005: 362). Excavations brought to light eight chamber tombs and one pit grave (K9). Chamber tombs are categorised into two groups: the first group including K2, K4 and K6/7 have simple doors whereas the second group including K1, K3, K5, K8 and K10 have a corridor-like entryway (dromos). Graves were dated by Veli Sevin considering their finds, the burial traditions and architecture; thus, chamber tombs with simple doors were set to the early phases of Early Iron Age (EIA I 1250–1000 BC) and chamber tombs with dromos to later phases (EIA II 1000–850 BC) (Sevin 2005: tab. 2). This dating conforms to the recently revised dating of Hasanlu (EIA I 1250–1050 BC; EIA II 1050–850 BC) (Danti 2013: 327–376). Karagündüz tombs had multiple burials and contained pottery, weapons, and personal adornment accessories. The jewellery studied here includes more than 100 items, intact or broken, displaying a variety of bracelets, earrings, rings, and pins. Most of the jewellery

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

286

Ayaz

Fig. 1. Map of Southern Caucasia Early Iron Age sites.

Fig. 2. a. Showing the range of metal jewellery at chamber tombs of K2, K4, K6/7; b. Showing the range of metal jewellery at chamber tombs of K1, K3, K5, K8, K10.

finds are made of iron, and the second biggest group is of bronze. There exist only one example of lead, and one item of gold (fig. 2). The beads included in the jewellery and constitute a big group; they have been assessed earlier (Arslan Sevin 2005; Arslan Sevin/Sevin 2008). The majority of jewellery are iron bracelets, which are subgrouped into two groups: those with circular section and those with flat band. Bracelets with circular section are either simple rings or open-ended, which sometimes overlap at the tips. Bracelets with circular

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The Early Iron Age Jewellery from the Karagündüz Necropolis …

287

Fig. 3. Table of the finds from chamber tombs with simple door (K2, K4, K6/7).

section are more in number than flat band ones, and tombs K2 and K6/7 contained more than ten, K8 a few, and K1 and K5 contained one each. Those bracelets with flat band and continuously open ends are again contained in K2. and amount to more than ten, whereas K6/7 had three and K8 had two (figs. 3–4).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

288

Ayaz

Fig. 4. Table of the finds from chamber tombs with dromos (K1, K3, K5, K8, K10).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The Early Iron Age Jewellery from the Karagündüz Necropolis …

289

Like the bracelets, the rings too can be divided into two subgroups as those with circular cross section and those with flat band. Those with circular section display a variety as simple loop, overlapping ends, or spiral; all but one were uncovered in K2 (ten circular rings) and the last one in K6/7. K10 contained iron loops, which may be rings. The bronze examples of simple loop and circular section were found in K5 and K8. Most of the rings with flat band were uncovered in K2; then K6/7 contained some, and K8 and K10 had one each. All of these flat rings are of iron. In addition to these two main ring groups there is one bronze ring with knobs, uncovered in K4 (figs. 3–4). Earrings of simple loop shape constitute a small group. Bronze examples were uncovered in K3, K10 and K6, one in each. The one from K6 is the only bronze find from this tomb. The simple loop shaped earring from K2 is the only gold find from Karagündüz. This gold earring is embellished with a notched agate ornament. The other two earrings have pendants. The earring from K8 has a bronze loop embellished with a frit bead attached with bronze pin, which is embellished with a fine wire wrapped around it. The only lead find from these tombs is an earring with an attachment found in K10. The simple loop is extended with a mushroom-like projection (figs. 3–4). The last group of jewellery finds are pins of iron and bronze. The iron pins uncovered in tombs K2, K4, K5, K6/7, K8 and K10 have a circular cross section and taper towards the tip; some have eyes (figs. 3–4). Some of the pins with eyes opened at the top or middle were forged flat. The iron pin with a carnelian top uncovered in K2 is worth noting. The bronze pins from Karagündüz, considered to be early examples of Urartian pins, have figural heads (Sevin 2005: 371). Bronze pins were produced by casting and comprise body and head. The slim body flares toward the head and has an eye. The head is adorned with animal protomes placed on top of ornamental elements, or with opium poppy. Six examples of such bronze pins were uncovered in tombs K3, K5, K8 and K10 with a dromos (fig. 4). Two pins with non-figural adornment distinguish themselves: A simple pin from K5 has a round section and a flattened eye whereas a bronze pin from K9 has a top adorned with five small spheres (Sevin/Kavaklı 1996: 37).

Conclusion Most of the jewellery finds from Karagündüz tombs were made by forging iron with low carbon content (Sevin 2005: 368). Tombs with simple doors dating to the earlier period contained one bronze ring with knobs (from K4) and a simple loop type bronze earring (from K6/7). All the remaining jewellery from tombs K2, K4 and K6/7 including bracelets, rings, earrings, and pins are of iron (fig.2a). Bronze earrings, rings, and cast pins constitute a

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

290

Ayaz

smaller group compared to the iron finds and were found in the tombs with dromos K1, K3, K5, K8, K10 (fig. 2b) and pit grave K9. Sites of Early Iron Age containing similar metal jewellery include Hasanlu IV (Muscarella 1988: figs. 14–27, 29, 35–39) and Dinkha III, II (Muscarella 1974: figs. 6–7, 12, 15, 36, 39, 43, 45, 48) in northwest Iran; Artik, Katnaghbyur, Mastara (Badalyan/Avestiyan 2007: 75, 79, 155, 219) in Armenia; Tlia IB, IIA, IIB (Apakidze 2009: figs. 63, 66, 70) in Georgia; and Kızılvank, Karabağlar, Culfa (Bahşaliyev 1997: pls. XXXI, XXXII, XXXIV), Kolanı (Bahşaliyev 1998: fig.10; Bakhshaliyev, Marro 2009: 113), Sarıdere, Demyeler, Muncuklutepe (Bakhshaliyev, Marro 2009: 104, 115–116, 118) in Nakhichevan, and many other sites in southern Caucasia. Here we attempt to compare the jewellery finds from Karagündüz with those from other sites as a whole (fig. 1). A majority Karagündüz finds are made of iron. As mentioned above, all but two of the finds from the EIA I tombs, namely K2, K4 and K6/7 are of iron, whereas the finds from EIA II tombs (K1, K3, K5, K8 and K10) indicate that iron continued to be used but the number of bronze finds increases (fig. 2a–b). The finds from the abovementioned sites in southern Caucasia indicate that bronze was more widely used compared to iron since the early phases of EIA. Dinkha contained relatively more iron finds, and the bracelets and rings from there exhibit a similarity to those of simple workmanship in Karagündüz. Bronze bracelets from all the sites can be compared with the iron examples of Karagündüz in typology only with regard to their circular section and flat band, because bronze bracelets have frequently decoration revealing finer workmanship. The motifs attested on them include triangles, hatching, zigzags, waves, crescents, and fine dashed lines, which are all considered distinctive features for the bracelets of Southern Caucasia during EIA. As inferred from flat band examples uncovered at Hasanlu, Dinkha and Kızılvank this model is intensively decorated, sometimes all over whereas bronze bracelets with circular section are plainer with the décor clustering around the tips. This group sometimes has grooved decoration. Bracelets with round section and snake-head tips, of high plastic quality, appear in the Urartian period in Eastern Anatolia; however, their simpler and stylised versions are found at some sites like Hasanlu and Tlia. It should not be too daring to propose that decoration of bracelet tips advanced with the Urartians. Close parallels of the knobbed ring from K4 and its bracelet versions are also seen. This bronze find from K4 tomb with simple door is somewhat early for Eastern Anatolia; perhaps it was not produced locally. Simple loop-shaped gold earring from tomb K2 has a parallel, but without the bead, found at Dinkha. Embellishment of earrings with beads is frequently encountered in the tombs in Nakhichevan (Kolanı, Demyeler, Muncuklutepe); however, no parallelism is attested between the examples from there and the golden earring from K2 and the bronze earring with bead from K8.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The Early Iron Age Jewellery from the Karagündüz Necropolis …

291

Iron pins are also found at Dinkha but their types are not parallel to those at Karagündüz. Other sites have bronze pins and it is not actually plausible to compare the coarse iron pins of Karagündüz with the others. Only the iron pin with carnelian head from K2 has a bronze parallel at Hasanlu. The bronze pin with a head embellished with small spheres is partially similar to those from Artik and Tlia, close to northern examples. Bronze pins with figural heads, which appeared in the EIA II at Karagündüz and continued in widespread use with the Urartu, can easily be claimed to be Eastern Anatolian model. Other EIA finds from other sites in East Anatolia like Hakkari (Sevin et al. 2001: dwg.1), Evditepe (Sevin 1987: figs. 5–6) and Yoncatepe (Belli, Konyar 2003: fig. 54) are mostly of iron and have a low quality workmanship as in Karagündüz. In comparison to other regions of Southern Caucasia bronze objects appear later here. The use of bronze beside iron and workmanship on the objects are attested starting with the EIA II. It looks plausible to think that in the EIA, Eastern Anatolia was a sub-region of culture, poorer than that in Southern Caucasia until the era of the Urartians.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

292

Ayaz

Bibliography Apakidze, J., 2009. Die Spätbronze-und Früheisenzeit in West- und Zentral-Kaukasien – Chronologische Studien zur Kolchis-Kultur 1600–700 v. Chr, Rahden/Westf. Arslan Sevin, N., 2005. Karagündüz ve Altıntepe Kazı Malzemeleri Işığında Van Gölü Havzası Boncukları, Arkeoloji ve Sanat 120, 45–50. Arslan Sevin, N./Sevin, V., 2008. Boncuklardan Okunan Tarih. Urartu Boncukları, Toplumsal Tarih 171, 32–38. Badalyan, R. S./Avestiyan, P. S., 2007. Bronze and Early Iron Age Archaeological Sites in Armenia I, Mt. Aragats and its Surrounding Region, Bar International Series 1697, England. Bahşaliyev, V., 1997. Nahçıvan Arkeolojisi/The Archaeology of Nakhichevan. İstanbul. –

1998. Nahçivan’da Bir Erken Demir Çağı Nekropolü: Kolanı, Belleten LXII/233, 1–13.

Bakhshaliyev, V./Marro, C., 2009. The Archaeology of Nakhichevan. Ten Years of New Discoveries, İstanbul. Belli, O./Konyar, E., 2003. Doğu Anadolu Bölgesi’nde Erken Demir Çağı Kale ve Nekropolleri. Early Iron Age Fortresses and Necropolises in East Anatolian, İstanbul. Burney, C. A., 1958. Eastern Anatolia in the Chalcolithic and Early Bronze Age. Anatolian Studies 8, 157–209. Danti, M. D., 2013. The Late Bronze and Early Iron Age in Northwestern Iran. D. T. Potts, (ed.), The Oxford Handbook of Ancient Iran, Oxford, 327–376. Muscarella, O. W., 1974. The Iron Age at Dinkha Tepe, Iran. Metropolitan Museum Journal 9, 35–90. –

1988. Bronze and Iron. Ancient Near East Artifacts in The Metropolitan Museum of Art, New York.

Sevin, V., 1987. Urartu Oda-Mezar Mimarisinin Kökeni Üzerine Bazı Gözlemler. A. Çilingiroğlu (ed.) Anadolu Demir Çağları I, İzmir, 35–55. –

2005. Son Tunç/Erken Demir Çağı Van Bölgesi Kronolojisi. Kökeni Aranan Bir Devlet: Urartu, Belleten LXVIII/252, Ankara, 355–386.

Sevin, V./Kavaklı E., 1996. Bir Erken Demir Çağ Nekropolü Van/Karagündüz. An Early Iron Age Cemetery, İstanbul. Sevin, V./Özfırat, A./Kavaklı, E., 2001. 1997–1999 Hakkari Kazıları, 22. Kazı Sonuçları Toplantısı, 1, Ankara, 355–368.

Gulan Ayaz, Turkey, Yüzüncü Yıl University, Van, [email protected].

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 293–300

Alice Boccia Paterakis – Sachihiro Omura

Gold Cloisonné from the Assyrian Colony Period in Central Anatolia A unique example of gold cloisonné dating to the 19th c. BCE, the period of the Assyrian Trade Colonies in Central Anatolia, was discovered in the Kaman-Kalehöyük excavations in 2010 by the Japanese Institute of Anatolian Archaeology. XRF analysis was carried out to study the composition and techniques of manufacture.

Description Kaman-Kalehöyük is a rural settlement along the ancient Silk Road trade route dating from the Bronze Age (2300 BC) through the Ottoman Empire (Omura 2008). The site is located 100 km southeast of Ankara, 3 km east of Kaman in Kırşehir province, and consists of a tell, or mound, 280 meters in diameter and 16 meters high. Excavation began in 1986 and has proceeded annually by the Japanese Institute of Anatolian Archaeology under the auspices of the Middle East Culture Center in Japan. Cultural levels have been traced from the Assyrian Colony Period, the Old Hittite Kingdom, and the Hittite Empire Period. The Japanese Institute of Anatolian Archaeology has conducted yearly archaeological surveys in Central Anatolia since 1986 resulting in the addition of two sites to their excavation roster: Yassıhöyük (Kırşehir) and Büklükale (Kırıkkale). In 2010 a gold object was unearthed in Room 409 in Kaman-Kalehöyük that may have been a workshop dating to the Period of the Assyrian Trade Colonies, approximately 1800 BCE (fig. 1). Preliminary reconstruction of the disfigured object suggests a winged lion rearing on its hind legs (fig. 2). Similar Assyrian motifs have been recovered on seals at Kaman-Kalehöyük. The gold weighs 104 grams and is hammered into a sheet and cut to form the figure.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

294

Boccia Paterakis – Omura

Fig. 1. Photo of gold lion showing lower legs © Middle Eastern Culture Center in Japan.

Fig. 2. Reconstruction drawing of gold lion © Middle Eastern Culture Center in Japan.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Gold Cloisonné from the Assyrian Colony Period in Central Anatolia

295

Fig. 3. Empty cells on gold sheet, hollow gold tubes and gold foil are visible © Middle Eastern Culture Center in Japan.

The Kaman object may be divided into three major components: a) a gold sheet that has been hammered to a thickness of approximately 0.5 mm and cut to form the figure and to create cloisons on one side, b) gold that has been hammered to a much thinner foil, and c) hollow gold tubes probably formed by rolling hammered sheet gold (fig. 3). Two rows of hollow tubes resemble threading tubes found on Egyptian gold pieces for stringing onto the body (Reeves 2012). Smaller pieces of gold are cut and adhered to one side of the gold plate forming cloisons or cells (fig. 3). Cloisonné is an ancient technique for decorating metal objects; in older periods with inlays of cut gemstones, glass, and glass paste and in more recent centuries with vitreous enamel. Examples of cloisonné have been found that pre-date the Kaman object. A gold ring and gold bead dating to 2600 BCE from the Royal Cemetery at Ur were inlaid with lapis lazuli (BM 121378) (Woolley 1934; Maxwell-Hyslop 1974; Tait 1976). Contemporaneous or slightly later examples of cloisonné, for example a falcon pendant (BM 1876, 1120.2) and gold pectoral (BM 1892, 0520.7), have been discovered in Greece (Higgins 1979). A gold pectoral (MMA16.1.3a, b) from the Middle Kingdom in Egypt, that

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

296

Boccia Paterakis – Omura

preserves inlays of carnelian, feldspar, garnet, and turquoise, is contemporaneous with the Kaman gold object (Brunton 1920; Dodson and Hilton 2004). The reverse is decorated by chasing. Since the pectoral was worn around the neck by stringing through two hollow attachments, either side could be displayed. Another gold pectoral also dating from the Middle Kingdom, 12th Dynasty, 1897–1841 BC (probabIy from Dahshur) displays many empty cloisons resembling the Kaman object (Eton College ECM 1585). Again the reverse was decorated by chasing so it could be hung around the neck displaying either side. While the deformed state of the Kaman object prevents a complete assessment of the object, examination to the extent permitted suggests that the cloisons in the Kaman gold object are devoid of inlays. The reverse has not been chased or decorated in any way. The presence of intentional holes that pierce the thickness of the object suggests it may have been mounted on something, perhaps a piece of furniture. The function of this object is subject to conjecture. The object has undergone preliminary examination and analysis to determine the composition of the gold and to ascertain the manufacturing techniques. The means of adhering the gold cloisons to the sheet gold, for example gold/silver hard soldering, copper colloidal soldering, and fusion bonding through heating, are of primary interest.

XRF Analysis The gold object was analyzed for 22 elements in 11 spot analyses with a Niton XL3t Thermo Scientific Handheld X-ray Fluorescence spectrometer. The analysis was carried out on August 20, 2013 by Daichi Sawamura and Izumi Nakai from the Tokyo University of Science (Sawamura and Nakai 2013). Areas on the solid sheet gold, the gold foil, and areas of attachment for the hollow tubes and cloisons were selected for analysis.

Results of XRF Analysis The XRF results indicate a gold content ranging from 87 to 96%, silver from 2 to 11% and copper 2 to 3%. The copper content is insufficient to suggest copper colloidal soldering. The gold lion may be divided into 3 alloy groups based on the compositional analysis of silver and gold: 1) the 0.5 mm thick gold sheet making up most of the figure and the gold foil on the first row of threading tubes (1.98–3.93% Ag, 93.25– 95.79% Au); 2) the gold foil on the second row of threading tubes (4.55–5.35% Ag, 92.46–92.49% Au); and 3) the threading tubes themselves (7.83–10.50% Ag, 87.26–88.38 % Au). 10 of the point analyses indicate Type I alloy of AuAgCu homogenous solid solution (< 10% silver) (MacDonald and Sistare

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Gold Cloisonné from the Assyrian Colony Period in Central Anatolia

297

1978). One point analysis with a silver content exceeding 10% indicates a Type II alloy. A higher silver content would suggest a gold-silver solder or the intentional hardening of the alloy to accomodate the manufacture or functionality of the object.

Discussion of the Results The alloy composition approximates that of a Sumerian necklace from Ur (Early Dynastic Period) predating the Kaman object (Maxwell-Hyslop 1977). The purity of gold in the Kaman object exceeds that found in most Egyptian gold of the 2nd millennium BC (Ogden 2000; Troalen et al. 2009). These preliminary results cannot identify the materials and techniques used to adhere the cloisons to the gold figure. While a gold/silver solder is suggested, fusion bonding by heating the object and slightly melting the gold (i.e. autogenous fusing) cannot be ruled out (Schorsch 1995). Evidence of partial melting is present on both sides of the joined cell walls in some areas supporting this theory (fig. 4); they have become detached from the gold sheet in many places (fig. 3). This overheating may have occurred during the attachment of the cloisons by soldering (Guerra 2014) or fusion bonding. Heating to fuse the cell walls to the gold sheet could have softened this Type I AuAgCu alloy object (MacDonald and Sistare 1978). The suggestion of fusion bonding is supported by Sumerian gold objects that pre-date the Kaman object (Early Dynastic Period of Sumer), specifically from the Tomb of Queen Puabi in the Royal Cemetery of Ur, that were constructed by fusion bonding without the use of solder (Maxwell-Hyslop 1977). Another example of damage to gold from overheating during manufacture can be seen on a 17th c. BCE gold necklace from Dilbat in Mesopotamia in the Metropolitan Museum of Art, New York (MMA 47.1a-h) (Lilyquist 1994). The mode of manufacture and the function of the hollow gold tubes remain to be ascertained. While examples of Egyptian gold jewelry present hollow gold tubes for stringing a bracelet or armband to the body, it is unlikely that the Kaman object was worn on the body. One piece of evidence to discourage the theory of body jewelry is a pivot joint that connects one row of hollow tubes to the object (fig. 5). If the hollow tubes formed the lion’s wing perhaps this joint provided mobility to the wing? The presence of intentional holes that pierce the sheet gold of the lion figure and the absence of decoration on the reverse also discourage the theory of body jewelry and suggest a decorative plaque perhaps applied on furniture. No evidence of inlays can be seen in the cloisonné cells of the Kaman object and there is little in the way of gouges or tool marks to indicate the removal of any inlays. Since the Kaman object presents little evidence of wear and preserves rough, cut edges of the gold sheet, perhaps this object was never finished. Damage to the object by overheating during

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

298

Boccia Paterakis – Omura

Fig. 4. Overheating of the gold is visible © Middle Eastern Culture Center in Japan.

Fig. 5. Pivot joint © Middle Eastern Culture Center in Japan.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Gold Cloisonné from the Assyrian Colony Period in Central Anatolia

299

formation of the cloisons perhaps was reason enough for its manufacture to be discontinued. Since is was found in a crumpled state it may have been set aside to be melted down for reuse. Location of manufacture is a matter of speculation. The Assyrian colony at KültepeKanesh was known to trade in metals with Ashur, the capital of the Old Assyrian Kingdom, in the period 1950 to 1750 BC (Maxell-Hyslop 1974). Kültepe-Kanesh is located approximately 175 kilometers from Kaman-Kalehöyük. This object represents a unique example of cloisonné in Central Anatolia.

Conclusions Plans are underway to have the object restored to its original configuration by a goldsmith who is also trained as a metals conservator. This combination of expertise will expedite the maximum recovery possible of all original information. Examination with the Scanning Electron Microscope (SEM) and analytical techniques providing greater accuracy of results could address many unanswered questions regarding technique and manufacture. Security measures in the Kaman-Kalehöyük Archaeological Museum limit access to the object and the possibility for its transport to Ankara for further examination and analysis. It is believed that the unique qualities and value of this artifact will stimulate and encourage further study, examination, and analysis in the future.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

300

Boccia Paterakis – Omura

Bibliography Brunton, G., 1920. Lahun I: The Treasure (BSAE 27 en ERA 20 (1914). London. Dodson, A./Hilton, D., 2004. The Complete Royal Families of Ancient Egypt. London. Guerra, M. F., 2014. Directrice de recherché au CNRS, February 9 personal communication. Higgins, R., 1979. The Aegina Treasure – An Archaeological Mystery. London. Lilyquist, C., 1994. The Dilbat Hoard, Metropolitan Museum Journal, v. 29. MacDonald, A./Sistare, G. H., 1978. The Metallurgy of Some Carat Gold Jewellery Alloys, Part 1: Coloured Gold Alloys. Gold Bulletin 11.3, 66–73. Maxwell-Hyslop, K. R., 1974. Western Asiatic Jewellery, c. 3000 – 612 B.C. London. –

1977. Sources of Sumerian Gold: The Ur Goldwork from the Brotherton Library, University of Leeds, A Preliminary Report. Iraq 39.1, 83–86.

Ogden, J., 2000. Metals. P. T. Nicholson/I. Shaw (eds.), Ancient Egyptian Materials and Technology. Cambridge, Cambridge University Press, 148–175. Omura, S., 2008. Preliminary Report on the 22nd Excavation Season at Kaman-Kalehöyük in 2007. Anatolian Archaeological Studies, 17, 1–43. Reeves, N., 2012. Metropolitan Museum of Art, personal communication. Sawarmua, D./Nakai, I., 2013. XRF Analysis of the Gold Lion Plaque excavated in 2010 (KL 10000009, C10-102), Tokyo University of Science, unpublished report. Schorsch, D., 1995. The Gold and Silver Necklaces of Wah: A Technical Study of an Unusual Metallurgical Joining Method. C. E. Brown et al. (eds.), Conservation in ancient Egyptian collections. London: Archetype, 127–135. Tait, H., 1976. Jewellery Through 7000 Years. London. Troalen, L.G./Guerra, M.F./Tate, J./Manley, B., 2009. Technological study of gold jewellery pieces dating form the Middle Kingdom to the New Kingdom in Egypt, ArchéoSciences. Revue d’Archéométrie 33, 111–119. Woolley, C. L., 1934. The Royal Cemetery, II. London.

Dr. Alice Boccia Paterakis. Dr. Sachihiro Omura, Japanese Institute of Anatolian Archaeology, Kaman, Kirsehir, Turkey.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 301–314

Müge Bulu

An Intact Palace Kitchen Context from Middle Bronze Age Alalakh: Organization and Function This paper presents a Middle Bronze Age palace kitchen recently excavated at Tell Atchana, ancient Alalakh. Due to its destruction by fire, the well-preserved context with in situ pottery provides the opportunity to understand the structural and functional organization of a palace kitchen. Tell Atchana is located on the Orontes River within the Hatay district of modern Turkey (fig. 1, left). It was the capital city of the small regional kingdom of Mukish in the 2nd millennium BC. Its location at the intersection of significant routes from central Anatolia leading to the Mediterranean Sea coast and the Syrian inlands confirms the strategic purposes of the site for its inhabitants throughout its occupational history (Yener 2007). The first excavations at Tell Atchana were conducted by Sir Leonard Woolley in the 1930’s and 1940’s where eighteen occupation levels defined the site’s stratigraphy (Woolley 1955). The problematic nature of the site’s chronology as defined by Woolley has been noted following the initial publication (e.g. Gates 1981; Heinz 1992; Bergoffen 2005). With the goal of re-defining the stratigraphy and chronology, the new round of excavations has been conducted at the site from 2003 onwards under the direction of K. Aslıhan Yener, on behalf of Koç University (Yener 2010). The largest dataset concerning the Middle Bronze II (MB II, c.1800–1600BC1) of Alalakh comes from Woolley’s Level VII palace complex, temple, city gate, ramparts, silos and fortification wall (Woolley 1955; Yener 2005: 105–106). According to cuneiform tablets retrieved from the Level VII palace, the Kingdom of Mukish was ruled from Alalakh by the Yarim-Lim dynasty and was a vassal of the Yamhad Kingdom centred in Aleppo (Klengel 1992: 60–63). In the Woolley excavations, exploration of the earlier phases of the Middle 1

This paper uses the Middle Chronology throughout and is not meant to make any statement on the absolute chronology of Tell Atchana.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

302

Bulu

Bronze Age was only performed in two deep soundings, the Great Stratification Pit and the Temple Sounding (Woolley 1955). As part of the efforts in re-defining the stratigraphy of the royal precinct during the Yener expedition, Square 33.322 was opened in the courtyard (room 9) of the Level VII palace (fig. 1, right). After three seasons of excavation (2007–2009), three occupation phases were revealed. The subject of this paper, referred to as Local Phase 3c3, is the earliest phase reached in the square; it is identified as a burnt kitchen based on architecture and in situ pottery.

The Palace Kitchen Context The context consists of two rooms, Room A and Room B (figs. 2–5), with evidence for two sub-phases of use in the floors. The scale of the walls4 indicates a very large building likely with upper floors (fig. 2a–c). The two rooms are separated by a wall with a doorway allowing passage between the two (fig. 2d). The other side of the western wall faces an outdoor space, indicating that it is an exterior wall. The southern space, Room A, measures 6 x 2 m and was fully exposed (fig. 3). Attached to the northern wall of the room is a horseshoe-shaped hearth made of clay (fig. 2e) with completely vitrified inner face and bottom. Immediately adjacent to the hearth is a platform of rubble and clay plaster with three sunken pithoid jars (fig. 2f, nos. 25, 27, and 285). The platform was added in the second phase of use as evidenced by the discarded pottery sherds on the floor beneath it. Opposite the hearth and platform, a bench runs along the southern wall (fig. 2g). The area around the hearth and the rest of the floor was covered in burnt rubble. Large pieces of carbonized timber were found intact in the middle of the room and may have belonged to the roof, collapsed during the fire. Through a doorway and a step down (fig. 2h), Room A is connected to Room B. Only a small part of Room B (ca. 1.5 x 6.5 m) was exposed and the rest remains in the northern baulk. Similar to Room A, the fill in Room B was burnt but in an even more intensely

2 3

4 5

The square measures 9 x 12 m. The Local Phase system at Atchana is used to track occupational phases in each square and begins with ‹Local Phase 0› for topsoil or ‹Local Phase 1› if the square begins within an area previously excavated by Woolley. The western and eastern walls are at least 2 m thick whereas the southern wall is at least 3 m thick. All are approximately 1 m high as preserved. Vessel 28 is the largest pithoid jar in the platform. The illustration of this vessel will be available after conservation.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

An Intact Palace Kitchen Context from Middle Bronze Age Alalakh

303

Fig. 1. Location of Tell Atchana, ancient Alalakh (left). Square 33.32 within Level VII Palace, superimposed on the modern contour of the site, contour size 1 m (right) (M. Akar).

vitrified state (fig. 5). A large piece of burnt timber found on the floor of Room B is similar to those of Room A and could also belong to the roof. The distinct burnt area on the west side of Room B in front of the doorway revealed many in situ vessels containing burnt cereals and possible residues from liquid contents. Right under this small area is a chunk of burnt wood and a rectangular clay bench along the western wall (fig. 2i). Another feature attached to the southern wall of Room B is a mud brick semicircular bench, the surface of which was plastered, and its eastern part collapsed (fig. 2j). While Rooms A and B are similar in construction methods and connected by a doorway, there are some distinctions between the two. The northern face of the wall separating the rooms (the southern wall of Room B) is well-plastered, a feature not found on any wall in Room A. The semicircular mud brick bench also has a well-plastered surface. These features can suggest differences in the functions of the rooms and possible activities carried out there.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

304

Bulu

17 4 18

i

2

3

23

1

Room B

15

14

j

h

d

d

N

e a

Room A

7

f

19 26

Outdoor Space

28

25

22 21

24

20

8 6 9

b

Tell Atchana 33.32 Local Phase 3c

27

5

g

c

Burnt material 0

1

2

3

4

5m

Distinct burnt area

1-28 In situ vessels

a-j

Features in the context

Fig. 2. The plan of Square 33.32 Rooms A and B (M. Akar and M. Bulu).

Although the excavated area is limited, the hearth, platform, benches, the pottery repertoire and botanical finds indicate that Rooms A and B might be part of kitchen rooms reminiscent of the subsequent Level VII palace of Alalakh. The row of rooms against the east wall of Palace VII, which was also the city wall, was identified by Woolley as ‹a long partitioned corridor› (1955: 93) and two of these rooms (29 and 33) are possibly the kitchen of the palace since these are the only rooms recorded as containing cooking pots and storage jars (1955: 105). The kitchen rooms from Square 33.32 then would be an earlier example of the same pattern and similarly, the eastern wall of Rooms A and B, which disappears into the eastern baulk, is thick enough to be both the wall of the palace and the city.

The Pottery Assemblage The intact contexts of Rooms A and B are important for their considerable amount of pottery ranging from discarded sherds accumulated over time to the 23 in situ vessels from the day of the destruction. The prevalence of in situ pottery within a limited space reveals

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

An Intact Palace Kitchen Context from Middle Bronze Age Alalakh

305

Fig. 3. Room A, view from southeast (Tell Atchana Excavations Archive).

the functional profile of the pottery assemblage as well as the logic of the kitchen organization. A total of 3840 sherds that can be broken down into a minimum number of 335 vessels (MNI)6 were recovered from the floors of the rooms. The whole pottery corpus7 is local and wheel-made with the exception of six small sherds of an unknown dark handmade ware. The majority of the pottery is Simple Ware, a medium-hard fired calcareous fabric containing low amounts of mixed minerals coming from local river sand deposits (Horowitz 2015). Another common ware is Mineral Cookware, a hard gritty brown fabric with crushed calcite temper that is typical of Middle Bronze Tell Atchana (Horowitz 2012). The third

6

7

Minimum number of individual (MNI) vessels in a context is calculated based on the diagnostic parts recorded in individual batches. Thus, the MNI is likely to be higher than reality since it is hard to detect if a rim and base belonged together when dealing with plain ware, therefore they may be batched separately during the recording process. The overall sherd count includes both undiagnostic and diagnostic fragments as well as those belonging to the complete vessels. The ceramic ware and shape typology at Tell Atchana was created by Dr. Mara Horowitz, the senior ceramic specialist of the Alalakh Excavations, and will be fully published in an upcoming volume of the Alalakh Excavations series; see also Horowitz 2015.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

306

Bulu

Fig. 4. The platform with its in situ vessels (Tell Atchana Excavations Archive).

common ware type is Syro-Cilician (Amuq-Cilician) ware, which is a painted subtype of Simple Ware (Horowitz 2015) and is a common Middle Bronze painted pottery tradition seen in Cilicia, the Amuq and northern Syria (Seton-Williams 1953; Tubb 1983; Bagh 2003). The most abundant vessel shape coming from the context is the bowl. The bowls are without surface finishing and only a few of them have painted decoration. The majority of the bowls are ‹s-curve bowl› type (fig. 6.1–10), with an out-turned rim and a rounded or carinated high shoulder (Horowitz 2015). All s-curve bowls are of Simple Ware, with the exception of a single Syro-Cilician rim sherd (fig. 6.10). Parallels are found at Levels X–VIII of Tell Atchana as well as at Ebla, Hama and Mari (Heinz 1992: pls. 20.28, 38.46, 57.22; Matthiae 1980: fig. 33; Nigro 2002: pl. XLVI.7, Pons 2007: pl. II). There are also Simple Ware shallow and hemispherical bowls (fig. 6.11–12) in the assemblages but these are less frequent and preserved only fragmentarily. The second most abundant vessel type is the jar. All are of Simple Ware, the surfaces are wet-smoothed and horizontal straight or wavy comb incisions were preferred for decoration. The most abundant type is the globular jar (fig. 7.21–22) which lacks a distinct shoulder-neck join and varies in the exact ratio of height to mouth opening (Horowitz 2015). Parallels are

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

An Intact Palace Kitchen Context from Middle Bronze Age Alalakh

307

Fig. 5. Distinct burnt area in Room B, view from north (Tell Atchana Excavations Archive).

found at Tell Afis, Ebla and Tell Hadidi (Mazzoni et al. 2002: fig. 6.10; Matthiae 1980: fig. 38; Dornemann 1992: fig. 5.4). Another common vessel type is the cooking pot. The only complete example (fig. 7.19), found crushed in front of the hearth in Room A, is biconical with a slightly carinated shoulder, a round base consistent with the use of a horseshoe shaped hearth, and two knobs projecting from the everted rim. Similar cooking pots are found at Tell Afis, Ebla, Tell Hadidi and Tarsus (Mazzoni et al. 2002: fig. 18.5; Nigro 2002: pl. LV.86; Dornemann 2007: pl. III.28; Slane 1987: pl. 8.24). The pithoid jar repertoire of the kitchen consisted of two types. The first has a hemispherical shape with a ‹rail rim› (a thick flanged rim with incised grooves at the top) and ring base with impressed finger marks (fig. 7.26–27). Parallels are found at Tell Atchana Levels X and VIII, Ebla, Tell al Rimah, and Tell Hadidi (Heinz 1992: pls. 61.55, 25.53; Nigro 2002: pl. LIII.71; Postgate et al. 1997: pl. 65.624, 628; Dornemann 1992: fig. 6.3). The second variety has a globular body with flared rim and convex base (fig. 7.23–25). Two globular pithoid jars with perforated bases (rim diameters of 18 and 24 cm, respectively) were found in situ in Room A (fig. 7.24–25). This vessel type has been associated with either beer or

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

308

Bulu

Fig. 6. Rooms A and B in situ vessels (1-9, 14-15) and fragments (10-13 and 16) (M. Bulu).

cheese production (Ellison 1984), which gives us a clue for the activities carried out in this space. Globular pithoid jars with similar profiles are also found at Ebla, Tell al Rimah and Tarsus (Nigro 2002: pl. XLVII.18; Postgate et al. 1997: pl. 65.629; Slane 1987: pl. 34.138). The pithoid jars of the context are of Simple Ware except for a single Heavy Coarse Ware hemispherical pithoid jar (fig. 7.26), which has a medium-low fired coarse fabric with abundant organic temper and gray-black cores (Horowitz 2008). No finishing was applied to the surfaces of either type and they usually had applied rope decoration on the shoulder. The krater is a less frequent type in these assemblages. It is defined as a high-shouldered medium-sized vessel with a wide mouth and flanged rim. There is only one partially complete krater with two rounded strap handles (fig. 6.17), and the rest are of the hole-mouth type without handles (fig. 6.15). All of them are of Simple Ware except for a single SyroCilician example (fig. 6.16). Parallels are found at Tell Atchana Level X and at Ebla (Heinz 1992: pl. 59.43; Nigro 2002: pl. XLVII.21). The kitchen revealed only one complete pitcher, which is a typical Syro-Cilician ‹eye› pitcher with trefoil rim, globular body and convex base (fig. 6.18). Its shoulder is decorated with a black painted band of repetitive bird figures (resembling swans) alternating with

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

An Intact Palace Kitchen Context from Middle Bronze Age Alalakh

309

Fig. 7. Rooms A and B in situ vessels (M. Bulu).

vertical line metopes, and there are two applied knobs on the shoulder opposite the handle. Exact parallels are found at Tell Atchana Level XII–XI and, although they lack the bird motif, ‹eye› pitchers are seen at various sites such as Kinet Höyük, Mersin, Qatna, and Ebla (Heinz 1992: pl. 69.8; Woolley 1955: pl. XCI; Gates 2000: fig. 6; Garstang 1953: fig. 143; Bagh 2003: fig. 3e; Matthiae 1989: fig. 4). Two small Syro-Cilician juglets were exposed from the context (fig. 6.13–14). They both lack the neck and rim and are decorated on the shoulder with diagonal and vertical lines between two horizontal bands. Rooms A and B also revealed some unarticulated handle fragments, which would indicate the presence of jugs or more pitchers during the life span of the spaces.

Discussion The rooms analyzed in this study had a long life span, indicated by two floor sub-phases, and their function could have changed over time. However, identifying certain activity areas within Rooms A and B is attempted here by examining the overall distribution of the minimum number of individual (MNI) vessel shape types including those in situ at the time

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

310

Bulu

of destruction (fig. 8). Moreover, the distribution of the in situ vessels reveals the activities carried out on the day of the destruction. The numbers on the plan in fig. 2 represent the locations of in situ vessels illustrated in figs. 6–7. The overall pottery assemblage coming from floor and fill of Room A indicates that various activities took place in this space. As predictable from the hearth, cooking pots constitute 16% (MNI: 10) of the whole assemblage found in this room. With the platform next to it, cooking processes took place in the northeastern corner. The broken in situ cooking pot (fig. 7.19) found in front of the hearth suggests that it was placed on the hearth right before the destruction. The high number of bowl fragments (36%, MNI: 23) could indicate preparation for serving or that the bowls were stored here. However, only one in situ s-curve bowl was found in Room A (fig. 6.5). The percentages of the jars (27%, MNI: 17) and pithoid jars (16%, MNI: 10) suggest that a considerable amount of commodity storage took place in this room. Two jars (fig. 7.21–22) and two pithoid jars (fig. 7.24, 26) found in situ confirm storage in the room. Located next to the hearth in Room A, the three built-in pithoid jars suggest that the platform was used for storage. The largest pithoid jar (fig. 2.28) was placed in a way that its previously broken rim was facing the hearth and it was full of emmer wheat (pers. comm., Emel Oybak Dönmez).8 In addition, an s-curve bowl found in this jar (fig. 4.7) might have been used to transfer the grains to the cooking pot that was placed on the hearth. The wide break at the rim of the hemispherical pithoid jar (fig. 4.27) could be intended to increase ease of access to the contents. The in situ cooking pot (fig. 4.20) and s-curve bowls (fig. 4.6, 8–9) found on the area around the necks of the larger jars likely imply that the remaining part of the platform was used as a countertop-workspace and an open storage to keep the vessels for food-related activities. Another possible activity in Room A is beer production. The textual evidence from various sites in the ancient Near East informs us about the production, consumption and distribution of beer with details about each process, including the types and amounts of ingredients and preparation methods (Hartman and Oppenheim 1950; Damerow 1996; 2011). Jars with perforated bases were used for filtering the beer and were placed on top of the collector vessel (Damerow 2011: 15, Bottéro 2004: 90). Therefore, the perforated pithoid jar found on the floor of Room A (fig. 7.24) could have been used to brew fermented beer and placed on top of the hemispherical pithoid jar found on the floor nearby (fig. 7.26). In this case, the emmer wheat in this room could also have been used for beer brewing. Similar contexts associated with beer production are found at Tell Hadidi, Tell Bazi, and Kuşaklı in central Anatolia, based on ceramic, textual and floral evidence (Gates 1988; Zarnkow et al. 2006; MüllerKarpe 2006). Large jars were used at Tell Bazi (with a capacity of around 100 liters) and Tell 8

The radiocarbon dates acquired from these samples are still in process.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

An Intact Palace Kitchen Context from Middle Bronze Age Alalakh

311

Fig. 8. Distribution of vessel shape types by MNI within Rooms A and B (M. Bulu).

Hadidi (90–175 liters) while others from Tell Hadidi (23 liters) and Kuşaklı (48–50 liters) are smaller (Zarnkow et al. 2006: fig. 2a–b; Gates 1988: 68; Müller-Karpe 2006: fig. 9). Although only a small part of Room B is available for study, the amount of bowls in the total sherd collection (64%, MNI: 56) is higher than that of Room A (including the ones from the platform) that was fully exposed. Moreover, kraters (5%, MNI: 4), the pitcher, and two juglets indicate serving rather than cooking, or a staging area for serving where these vessels were stored. Four s-curve bowls (fig. 6.1–4), two kraters (fig. 6.15, 17), the pitcher (fig. 6.18) and the juglet (fig. 6.14) found in situ on and in front of the rectangular bench support the serving activities in this room. However, the presence of cooking pots (7%, MNI: 6), jars (14%, MNI: 12), and pithoid jars (8%, MNI: 7) might point to other activities as well. One of the in situ vessels on the rectangular bench is a pithoid jar (fig. 6.23) and is related to storage rather than serving. Since only a small part is exposed, making accurate assessments about the functions of this room is difficult.

Conclusions The results of this contextual study conducted on the recently excavated Middle Bronze Age palace kitchen at Tell Atchana suggest that these rooms, at least the fully excavated Room A, functioned as food preparation spaces. The prevalence of s-curve bowls and some vessel types that can be associated with drink consumption such as krater, pitcher and juglet could

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

312

Bulu

indicate that Room B was related to serving activities as opposed to food and drink preparation in Room A. In addition, regarding the presence of jars with perforated bases and the abundance of emmer wheat recovered in this context, this kitchen might have been part of beer production. The interpretation for the function of these kitchen rooms remains incomplete due to the limited excavated area as well as the lack of comparable contexts from the renewed excavations. As more contexts dating to pre-Level VII periods at the site are excavated, the results of this research will contribute to the interpretation of cooking and serving practices on a site-wide scale as well as to differences in the pottery traditions in elite and non-elite parts of the site, if any. Furthermore, residue analysis on some of the vessels is in progress and will add more information regarding the topic of food and drink preparation and serving in the Middle Bronze Age palace of Alalakh.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

An Intact Palace Kitchen Context from Middle Bronze Age Alalakh

313

Bibliography Bagh, T., 2003. The Relationship between Levantine Painted Ware, Syro/Cilician Ware and Khabur Ware and the Chronological Implications. M. Bietak (ed.), The Synchronisation of Civilisations in the Eastern Mediterranean in the Second Millennium BC. II. Proceedings of the SCIEM 2000 EuroConference in Haindorf, 2nd–7th of May 2001. Haindorf, 219–238. Bergoffen, C., 2005. The Cypriot Bronze Age Pottery from Sir Leonard Woolley’s Excavations at Alalakh (Tell Atchana). Contributions to the Chronology of the Eastern Mediterranean 5. Vienna. Bottéro, J., 2004. The Oldest Cuisine in the World: Cooking in Mesopotamia. Chicago and London. Damerow, P., 1996. Food Production and Social Status as Documented in Proto-cuneiform Texts. P. Wiessner/W. Schiefenhövel (eds.), Food and the Status Quest: An Interdisciplinary Perspective, edited by Providence, 149–169. –

2011. Sumerian Beer: The Origins of Brewing Technology in Ancient Mesopotamia. Cuneiform Digital Library Journal 3, 1–20.

Dornemann, R. H., 1992. Early Second Millennium Ceramic Parallels between Tell Hadidi-Azu and Mari. G. D. Young (ed.), Mari in Retrospect. Fifty Years of Mari and Mari Studies. Indiana, 77–112. –

2007. The Pottery of the Middle Bronze Age in the Euphrates River Valley, in the Area Affected by the Basins of the Tabqa and Tishrin Dams. M. al-Maqdissi et al. (eds.), Céramique de l’âge du Bronze en Syrie, II. L’Euphrate et la région de Jézireh. Beyrouth, 43–52.

Ellison, R., 1984. The Uses of Pottery. Iraq 46.1, 63–68. Garstang, J., 1953. Prehistoric Mersin: Yümük Tepe in Southern Turkey. Oxford. Gates, M.-H., 1981. Alalakh Levels VI and V: A Chronological Reassessment. Syro-Mesopotamian Studies 4.2, 11–50. –

1988. Dialogues between Ancient Near Eastern Texts and the Archaeological Record: Test Cases from Bronze Age Syria. BASOR 270, 63–91.

–

2000. Kinet Höyük (Hatay, Turkey) and MB Levantine Chronology. Akkadica 119–120, 77–101.

Hartman, L. F./Oppenheim, A. L., 1950. On Beer and Brewing Techniques in Ancient Mesopotamia According to the XXIIIrd Tablet of the Series HAR.ra = hubullu. Journal of the American Oriental Society Supp. 10. Baltimore. Heinz, M., 1992. Tell Atchana/Alalakh: Die Schichten VII–XVII. Alter Orient und Altes Testament 41. Kevelaer/Neukirchen-Vluyn. Horowitz, M. T., 2008. Building an Integrated Typology for the Alalakh Ceramic Assemblage. Paper presented at the 2008 ASOR Annual Meeting, Boston. –

2012. History in a Cookpot: New Approaches to Local Pottery at Middle/Late Bronze Age Tell Atchana/Alalakh. Paper presented at the 113th AIA Annual Meeting, Philadelphia.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

314 –

Bulu 2015. The Evolution of Plain Ware Ceramics at the Regional Capital of Alalakh in the 2nd Millennium BC. C. Glatz (ed.), Plain Pottery Traditions of the Eastern Mediterranean and Near East: Production, Use, and Social Significance. Walnut Creek, 153–182.

Klengel, H., 1992. Syria, 3000 to 300 BC: A Handbook of Political History. Berlin. Matthiae, P., 1980. Ebla. An Empire Rediscovered. London. –

1989. Jugs of the North-Syrian/Cilician and Levantine Painted Wares from the Middle Bronze II Royal Tombs at Ebla. K. Emre et al. (eds.), Anatolia and the Ancient Near East. Studies in Honor of Tahsin Özgüç. Ankara, 303–313.

Mazzoni, S. et al., 2002. Tell Afis. Siria – 2000–2001. Pisa. Müller-Karpe, V., 2006. Anadolu’da Bronz Çağında Bira. Ü. Yalçın et al. (eds.), Uluburun Gemisi. 3000 Yıl Önce Dünya Ticareti. Bochum, 173–186. Nigro, L., 2002. The Middle Bronze Age Pottery Horizon of Northern Inner Syria on the Basis of the Stratified Assemblages of Tell Mardikh and Hama. M. al-Maqdissi et al. (eds.), Céramique de L’Âge du Bronze en Syrie I: La Syrie du Sud et la Vallée de l’Orontepp. Beyrouth, 97–128. Pons, N., 2007. Essai de caractérisation de la poterie du Bronze moyen de Mari. M. al-Maqdissi et al. (eds.), Céramique de l’âge du Bronze en Syrie, II. L’Euphrate et la région de Jézireh. Beyrouth, 85–117. Postgate, C./Oates, D. /Oates, J., 1997. The Excavations at Tell al Rimah. The Pottery. Iraq Archaeological Reports 4. London. Seton Williams, M. V., 1953. A Painted Pottery of the Second Millennium from Southern Turkey and Northern Syria. Iraq 15.1, 56–68. Slane, D. A., 1987. Middle and Late Bronze Age Architecture and Pottery in Gözlükule, Tarsus: A New Analysis. Unpublished PhD dissertation Bryn Mawr College, Pennsylvania. Tubb, J. N., 1983. The MB IIA Period in Palestine: Its Relationship with Syria and Its Origin. Levant 15, 49–62. Woolley, C. L., 1955. Alalakh: An Account of the Excavations at Tell Atchana in the Hatay, 1937–1949. Reports of the Research Committee of the Society of Antiquaries of London 18. London. Yener, K. A., 2005. The Amuq Valley Regional Projects. Vol. 1, Surveys in the Plain of Antioch and Orontes Delta, Turkey, 1995–2002. Chicago. –

2007. The Anatolian Bronze Age Kingdoms and Alalakh: Mukish, Kanesh and Trade. Anatolian Studies 57, 151–160.

–

2010. Tell Atchana, Ancient Alalakh. Volume I: The 2003–2004 Excavation Seasons. Istanbul.

Zarnkow, M. et al., 2006. Interdisziplinäre Untersuchungen zum altorientalischen Bierbrauen in der Siedlung von Tell Bazi/Nordsyrien vor rund 3200 Jahren. Technikgeschichte 73.1, 3–25.

Müge Bulu, PhD Candidate, Koç University.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 315–331

Nilgün Coşkun

Middle and Neo-Assyrian Periods of the Harran Plain in Light of a Survey An archaeological survey has been carried out in the plain of Harran by Veli Sevin between 1987–1989 and numerous settlements have identified. The Middle and Neo Assyrian pottery collected at this work and it shows a strong continuity in form and technic. Surveys conducted in the Harran Plain under the direction of Prof. Dr. Veli Sevin between 1987–1989 have identified numerous settlements. Some of the Middle and Neo-Assyrian potsherds collected from these settlements will be presented here. The Harran Plain is located in southeast Turkey (fig. 1) at an altitude of 550–400 m. The plain is an extension of the desert to its south and it is watered by Çullap (Jullab) River, a tributary of Belih (Balikh) River (Özfırat 2005: 20). Sites explored include: Anaz: Located at the village of Anaz about 30 km north of Harran. Destroyed by the modern settlement. Potsherds of Middle and Neo-Assyrian periods were collected. Aşağı Yarımca: Located to the north of Aşağı Yarımca village about 5 km northwest of Harran. The modern settlement occupies the southern foot of the höyük. S. Lloyd and N. Gökçe conducted a small sondage in 1951. Potsherds of Middle Assyrian, Early Iron and Neo-Assyrian periods were collected. Göktepe: Located to the north of Göktepe village about 35 km north of Harran. Potsherds of Middle Assyrian, Early Iron and Neo-Assyrian periods were collected. Kepirli (Kantara): Located at the village of Kepirli about 35 km north of Harran. Potsherds of Middle and Neo-Assyrian periods were collected. Mamuca (Mezarlıktepe): Located to the north of Mamuca village about 23 km north of Harran. The north part of the settlement is destroyed. Potsherds of Middle and Neo-Assyrian periods were collected.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

316

Coşkun

Fig. 1. Sites explored.

Sultantepe: Located at the Sultantepe village about 24 km north of Harran. Modern settlement has destroyed particularly the eastern slopes. A. Dönmez and W. C. Brice conducted surveys in 1947 and S. Lloyd conducted excavations in 1951–52; G. Kozbe carried out extensive surveys in 2011. Potsherds of Middle and Neo-Assyrian periods were collected. Yukarı Yarımca: Located at the Yukarı Yarımca village about 8 km northwest of Harran. It is occupied by the modern settlement. Potsherds of Middle and Neo-Assyrian periods were collected. Çamlıdere: Located at the Çamlıdere village about 33 km north of Harran. It is occupied by the modern settlement. Potsherds of Middle and Neo-Assyrian periods were collected. Akçalı (Mineyçir): Located to the north of Akçalı village about 35 km north of Harran. Potsherds of Neo-Assyrian period were collected. Erbaş (Harabeysito): Located 1.5 km southwest of Yarımsu village about 30 km north of Harran. Potsherds of Neo-Assyrian period were collected. Para Para: Located at the Para Para village about 12 km northwest of Harran. It is occupied by the modern settlement. Potsherds of Neo-Assyrian period were collected.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Middle and Neo-Assyrian Periods of the Harran Plain in Light of a Survey

317

Middle Assyrian potsherds collected can be categorized into three groups: Red-brown wares: Their clay has a colour ranging from red to brown and they are slipped in the same colour. Cream-green wares: Clay colour varies from cream to buff and light green. Slipped in the same colour as the clay. Cream-green slipped wares: Clay colour varies from cream to brown, to red; slip colour is cream or light green. Most of the Middle Assyrian pottery has sand and intensive plant temper. Examples are usually moderately or well fired; some have their surfaces slightly treated or left plain. All are wheelmade. Frequent examples of this period are shallow pots/plates1 (fig. 2.1–5), globular bodied2 (fig. 2.8), deep3 (fig. 2.10, 13) and carinated4 (fig. 2.6–7, 14), «S» profiled jar5 (fig. 2.10), ring base6 (fig. 2.11–12) and pedestal (fig. 2.15) bottoms. Forms of Middle Assyrian pottery do not vary according to ware groups. Close parallels of both types and wares are encountered at the sites in Upper and Central Tigris, Middle Euphrates and Habur Valley. Collected potsherds include a few grooved examples dated to the Early Iron Age and these can be categorised into two groups based on their ware groups: Red-brown wares: Their clay has a colour ranging from red to brown and are slipped in the same colour. 1

2

3

4

5

6

Parallels: Tarsus LBA II (Goldmann 1987: Fig. 4: 182; 25: 200), Tell Barri/Kahat 3 (Anastasio 1998: Fig. 21: 3), Tell Şeh Hamad A a–e (Pfälzner 1995: Fig. 77: a; 98: h), Tell Brak HH 1 (Oates et al. 1994: Fig. 181: 24). Parallels: Tell Şeh Hamad E IIb (Pfälzner 1995: Fig. 98: h; 100: c), Tell Taban 9b (Ohnuma/Numoto 2001: Fig. 7: 17), Tell Bderi 2 (Pfälzner 1995: Fig. 136: q), Sabi Abyad (Rossmeisl 1989: Fig. XII.2: 15). Parallels: Üçtepe 9 (Köroğlu 1998: Fig. 8: 3), Tell Barri/Kahat 1 (Anastasio 1998: Fig. 4: 1), Sabi Abyad LBA (Rossmeisl 1989: Fig. XII.2: 14), Tell Şeh Hamad A a–e (Pfälzner 1995: Fig. 67: e), Tell Taban 9b (Ohnuma/Numoto 2001: Fig. 7: 19), Khirbet esh Shenef LBA (Bartl 1990: Fig. 5: 2), Tell Brak HH1 (Oates et al. 1994: Fig. 182: 12). Parallels: Üçtepe 9 (Köroğlu 1998: Fig. 5: 15; 9), Tell Şeh Hamad A a–e (Pfälzner 1995: Fig. 67: h), Khirbet esh Shenef LBA (Bartl 1990: Fig. 4: 5), Tell Barri/Kahat 3 (Anastasio 1998: Fig. 6: 2), Mohammed Arab (Roaf 2001: Fig. 6: 1), Tell Taban 9a (Ohnuma/Numoto 2001: Fig. 48: d), Tell Rimah A1 (Postgate et al. 1997: Fig. 30: 1), Gricano (Schachner 2002: Fig. 5:5). Parallels: Üçtepe 10 (Özfırat 2007a: Fig. CIII: 11–13), Tarsus LBA II (Goldmann 1987: Fig. 22: 133), Tell Hadidi LBA I (Dornemann 1981: Fig. 15: 17), Tell Nebi Bend E–F (Bourke 1993: Fig. 11: 1), Tell Şeh Hamad A a–e (Pfälzner 1995: Fig. 75: c). Parallels: Tell Şeh Hamad A a–e (Pfälzner 1995: Fig. 97: a), Tell Taban 8b (Ohnuma/Numoto: 2001: Fig. 9: 46).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

318

Coşkun

Fig. 2.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Middle and Neo-Assyrian Periods of the Harran Plain in Light of a Survey

Fig. 3.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

319

320

Coşkun

Pink-buff wares: Clay colour varies from pink to cream and to buff. Slipped in the same colour as the clay. Grooved pottery examples have sand temper of various sizes and mostly were moderately fired. Beside burnished examples there are also examples with no surface treatment. Most were shaped with hand. Bowls with grooved rims (fig. 3.1–3, 5–7)7 and jars with closed mouths (fig 3.4)8 are the forms of this period. Potsherds with grooves collected in the course of surveys in the Harran Plain, albeit few in number, are parallel to the pottery of Early Iron Age from Eastern Anatolia. Their parallels are known from Eastern Anatolia, South Caucasia, north of Middle Euphrates basin and Upper Tigris basin. Potsherds of Neo-Assyrian periods studied are categorised into three groups just like the pottery of the Middle Assyrian period: Red-brown wares: Their clay has a colour ranging from red to brown and are slipped in the same colour. Cream-green wares: Clay colour varies from cream to buff and light green. Slipped in the same colour as the clay. Cream-green slipped wares: Clay colour varies from cream to brown, to red; slip colour is cream or light green. Neo-Assyrian pottery has intensive plant temper. They are usually moderately or well fired. Some have surfaces slightly treated whereas some others are left coarse. All are wheelmade.

7

8

Parallels: Kavuşan F14 (Kozbe/Köroğlu 2004: Fig. 6: 4), Gricano (Schachner/Schachner 2003: Fig. 2), Salattepe (Ökse et al. 2004: 332), Dilkaya (Çilingiroğlu 1993: Fig. 14:1; 16), Ernis (Sevin 2004: Fig. 4: 1–4), Karagündüz Cemetery (Sevin/Kavaklı 1996: 25, 46–47), Norşuntepe (Müller 2005: Fig. 2: 7), İmikuşağı (Sevin 1995: Fig. 13: 1–4), Siyahtaş (Özfırat 2007b: Fig. 6: 1–7), Syunik (Xnkikyan 2002: Fig. XXI: 5–10), Metsamor (Khanzadyan et al. 1973: Fig. XV: 1). Parallels: Tille IV–V (Blaylock 1999: Fig 3: 10), Lidar (Müller 1999: Fig. 5: BB03), Dilkaya (Çilingiroğlu 1993: Fig. 17), Norşuntepe (Müller 2005: Fig. 2: 3–4), İmikuşağı (Sevin 1995: Fig. 15: 7–9), Syunik (Xnkikyan 2002: Fig. XXI: 18).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Middle and Neo-Assyrian Periods of the Harran Plain in Light of a Survey

321

Neo-Assyrian pottery finds include fragments of globular bodied (fig. 3.8–14; fig. 4.1, 2, 6–8; fig. 5.1, 5–10, 12, 13; fig. 7.3, 7–9)9, carinated shouldered (fig. 3.15–17; fig. 5.11; fig. 7.1)10 and carinated bellied (fig. 4.3–5; fig. 5.2–4; fig. 7.2, 4–6)11 bowls, jars with closed mouths

Parallels: Gre Dimse (Karg 1999: Fig. 11: 10), Hakemi Use (Tekin 2009: Fig. 13: 3), Hirbemerdon IVb (Guarducci/Laneri 2010: Fig: 53), Lidar (Müller 1999: Fig. 7: 2), Üçtepe (Köroğlu 1998: Fig. 9: 17), Ziyarettepe (Matney et al. 2007: Fig. 18: d), Assur Cemetery (Haller 1954: Fig. 6: v, w), Khirbet Qasrij (Curtis 1989: Fig. 1, 4, 5), Nimrud (Lines 1954: Fig. XXXVII: 6; Oates 1959: Fig. XXXV: 7, 18, 21), Ninive (Lumsden 1999: Fig. 4: 1), Tell al Rimah (Postgate et al 1997: Fig. 56: 497), Tell Afis (Oggiano 1999: Fig. 5: 10, 12), Tell Jurn Kabir (Eidem/Ackermann 1999: Fig. 6: 7; 8: 6), Tell Rad Shaqrah (Reiche 1999: Fig. 6: b), Tell Sheikh Hassan (Schneider 1999: Fig. 4: 1), Tell Taya II (Anastasio 2010: 58: 1), Til Barsip (Jamieson 1999: Fig. 4: 2). 10 Parallels: Hakemi Use (Tekin 2004: Fig. 9: 6, 7), Hirbemerdon IVb (Guarducci/Laneri 2010: Fig: 72, 77), Lidar (Müller 1999: Fig. 17: 12), Tille (Blaylock 1999: Fig. 5: 18), Assur Cemetery (Haller 1954: Fig. 6: b, g, h), Kar Tukulti Ninurta 2–4 (Schmidt 1999: Fig. 5a: 12; 6a: 22, 24), Khirbet Khatuniyeh 6–5 (Curtis/Green 1997: Fig. 85114), Khirbet Qasrij (Curtis 1989: Fig. 57, 79, 81), Nimrud (Lines 1954: Fig. XXXVI: 1; Oates 1959: Fig. XXXV: 2), Qasrij Cliff (Curtis 1989: Fig. 3, 4, 6, 13, 25), Tell al Rimah (Postgate et al. 1997: Fig. 495), Tell Abu Dhahir 11 (Green 1999: Fig. 5: 5), Tell Taya II (Anastasio 2010: 58: 2). 11 Parallels: Hakemi Use (Tekin 2004: Fig. 9: 8), Hirbemerdon IVb (Guarducci/Laneri 2010: Fig. 53), Tille (Blaylock 1999: 5: 22), Üçtepe 7–8 (Sevin 1990: Fig. 8: 1), Assur Cemetery (Haller 1954: Fig. 6: ab, c, l), Kar Tukulti Ninurta 2–4 (Schmidt 1999: Fig. 6a: 1), Khirbet Khatuniyeh 6–5 (Curtis/ Green 1997: Fig. 90, 115), Nimrud (Lines 1954: Fig. XXXVII: 2, 7), Ninive (Lumsden 1999: Fig. 5: 15, 19, 17, 22, 23), Qasrij Cliff (Curtis 1989: Fig. 11, 16, 18), Tell Jurn Kabir (Eidem/Akermann 1999: Fig. 4: 12; 5: 14; 8: 7), Tell Afis (Oggiano 1999: Fig. 3: 4, 5).

9

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

322

Coşkun

Fig. 4.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Middle and Neo-Assyrian Periods of the Harran Plain in Light of a Survey

323

(fig. 6.1)12 or with short or tall necks (fig. 4.9–12; fig. 6.2–14; fig. 7.10–14; fig. 8.1–14)13, ring bases (fig. 4.13–15)14, nipple bases (fig. 6.15)15 and pedestal bases (fig. 8.15)16. Neo-Assyrian pottery forms do not vary according to ware groups. Except for the nipple base fragment studied (fig. 6.15) there is no fragment reminiscent of «palace ware»; all the rest are of «common ware» for daily use. Their parallels with respect to forms and wares are found at Upper and Middle Tigris, Middle Euphrates basins and Habur Valley. Neo-Assyrian pottery studied here is entirely parallel with the homeland of the Assyrian Empire. This cultural unity had already existed during the Middle Assyrian period, although in a more limited area. This unity seems to have been interrupted during the 12th– 10th centuries BC when the Middle Assyrian Kingdom shrank to its core region. It may be proposed that the political stagnation of the Kingdom observed in the north of the Middle Euphrates basin, the Upper Euphrates basin and partially in the Harran Plain facilitated the spread of «grooved wares» influenced from Eastern Anatolia. With the Neo-Assyrian Kingdom gaining in power as of the 9th century BC and extending its territory up to the Anti-Taurus Mountains and East Mediterranean the interrupted Assyrian wares or Assyrian influenced wares started to prevail in Southeast Anatolia again. Ware, technical properties 12 Parallels: Lidar (Müller 1999: Fig. 13: 15; 20: 1), Kar Tukulti Ninurta 2–4 (Schmidt 1999: Fig. 5b:

13

14

15

16

1; 7a: 9), Khirbet Khatuniyeh 6–5 (Curtis/Green 1997: Fig. 250), Khirbet Qasrij (Curtis 1989: Fig. 181, 182), Ninive (Lumsden 1999: Fig. 7: 39, 41), Qasrij Cliff (Curtis 1989: Fig. 68, 24), Tell al Rimah (Postgate et al. 1997: Fig. 56: 504), Tell Afis (Oggiano 1999: Fig. 8: 1, 2). Parallels: Aşağı Salat (Şenyurt 2006: Fig. 3, 4), Gre Dimse (Karg 1999: Fig. 10: 10–12; 11: 1), Hakemi Use (Tekin 2004: Fig. 8: 2, 3), Hirbemerdon IVb (Guarducci/Laneri 2010: Fig. 138, 140– 143, 146, 147), Kavuşan (Kozbe 2007: Fig. 4), Lidar (Müller 1999: Fig. 20: 20; 21: 13), Salat Tepe (Ökse/Görmüş 2014: Fig. 9: 2, 3), Tille (Blaylock 1999: Fig. 8: 1, 3, 4), Ziyarettepe (Matney et al 2007: Fig. 19: a, c, e, f), Assur Cemetery (Haller 1954: Fig. 3: al, ac, ar, ad, g, ag), Kar Tukulti Ninurta 2–4 (Schmidt 1999: Fig. 4: 9, 22; 5b: 9, 12, 23, 51; 7a: 12), Khirbet Khatuniyeh 6–5 (Curtis/Green 1997: Fig. 242, 248, 236, 264), Khirbet Qasrij (Curtis 1989: Fig. 150–154, 164, 200, 202, 209), Nimrud (Lines 1954: Fig. XXXIX: 3; Oates 1959: Fig. XXXVIII: 83, 84, 93, 94, 99), Ninive (Lumsden 1999: Fig. 6: 30–35; 7: 38, 39; 8: 45–47), Qasrij Cliff (Curtis 1989: Fig. 49, 50, 54–56, 59, 60–64), Tell Afis (Oggiano 1999: Fig. 9: 5, 6), Tell Jurn Kabir (Eidem/Akermann 1999: Fig. 7: 12–14, 17; 8: 14–16, 20), Til Barsip (Jamieson 1999: Fig. 2: 2–4). Parallels: Tille (Blaylock 1999: Fig. 5: 16, 19; 7: 5), Assur Cemetery (Haller 1954: Fig. 6: o, t, r, n), Khirbet Khatuniyeh 6–5 (Curtis/Green 1997: Fig. 86), Khirbet Qasrij (Curtis 1989: Fig. 30, 184), Nimrud (Lines 1954: Fig. XXXVII: 2, 3, 5; Oates 1959: Fig. XXXVI: 31), Qasrij Cliff (Curtis 1989: Fig. 17–19), Tell al Rimah (Postgate et al 1997: Fig. 56: 492–495, 497), Tell Afis (Oggiano 1999: Fig. 6: 7), Til Barsip (Jamieson 1999: Fig. 1: 1, 2, 8–10, 12, 13). Parallels: Girnevaz (Şenyurt 1988: Fig. IV: 2, 6), Hakemi Use (Tekin 2004: Fig. 8: 2, 6; Tekin 2009: Fig. 13: 2), Tille (Blaylock 1999: Fig. 9: 5; 10: 12), Üçtepe 7–8 (Sevin 1990: Fig. 9: 1, 8), Assur Cemetery (Haller 1954: Fig. 5: d, e), Nimrud (Lines 1954: Fig. XXXIX: 2), Til Barsip (Jamieson 1999: Fig. 2: 5). Parallels: Girnevaz (Şenyurt 1988: Fig. IV: 4), Tille (Blaylock 1999: Fig. 10: 18), Üçtepe 7–8 (Köroğlu 1998: Fig. 10: 10), Assur Cemetery (Haller 1954: Fig. 5: u), Nimrud (Lines 1954: Fig. XXXIX: 3), Til Barsip (Jamieson 1999: Fig. 2: 4).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

324

Coşkun

Fig. 5.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Middle and Neo-Assyrian Periods of the Harran Plain in Light of a Survey

Fig. 6.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

325

326

Coşkun

Fig. 7.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Middle and Neo-Assyrian Periods of the Harran Plain in Light of a Survey

Fig. 8.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

327

328

Coşkun

and forms of Middle Assyrian pottery changed very little and continued their existence through this period as well. N. Yardımcı also conducted surveys in Harran Plain for many years. Potsherds collected from more than 200 settlements include examples from the Middle and Neo-Assyrian periods (Yardımcı 2004). It is likely that Southeast Anatolia housing trade routes transforming to a systematic transportation network within the imperial borders during the Neo-Assyrian Kingdom period, housing routes important for Anatolia and East Mediterranean, new cities and related settlements with a variety of functions that formed when the entire region was incorporated into the provincial system and the bureaucratic system in the region brought with it cultural influences. As a matter of fact, that the pottery culture emerging in Southeast Anatolia as of the 1st millennium BC exhibits unity with the Neo-Assyrian pottery tradition may be related with the intensive connections and political developments between regions in close proximity with each other. The Harran Plain was always an important region for the Assyrian Kingdom. In one of his inscriptions King Adad-Nirari I of Middle Assyrian Kingdom (13th c. BC) states that Harran fortress was conquered; his son and successor Sulmanu-Asherid I (13th c. BC) mentioned, like his father, that he conquered Harran on his campaign to Hanigalbat (Luckenbill 1968: 27, 40). The name of Harran is not encountered in the reign of Tukulti-Apil-Assara (12th–11th c. BC) because almost all the campaigns in this period targeted the Aramaeans who pressurised increasingly. Texts from the reign of Adad-Nirari II (end of 10th – beginning of 9th c. BC) reveal that Harran paid taxes and tributes (Luckenbill 1968 I: 115). The inscriptions of Assur-nasir-apli II, one of the first influential kings of the Neo-Assyrian Empire (9th c. BC), do not mention Harran but state that the king founded a provincial centre called Tusha/ Tushan on his campaigns to the Upper Tigris basin and took under his control the area from the Middle Euphrates basin to the East Mediterranean (Luckenbill 1968 I: 138–199). The texts of Salmanu-asherid (9th c. BC) do not mention Harran but it is known that he ruled over the entire Southeast Anatolia. The name of Harran is not attested in the texts for a long time until the reign of Sarrukin II (8th c. BC), one of the strong kings of the Neo-Assyrian Empire, who exempted Harran from taxes (Luckenbill 1968 II: 25). Texts of Assur-ban-apli (7th c. BC) state that the king rebuilt the temple of Sin at Harran (Luckenbill 1968 II: 377). Texts of the 8th c. BC from Nimrud and of the 7th c. BC from Nineveh contain the names of governors appointed to Harran. Although it is not clear, it is thought that Harran was made the provincial centre in the early phases of the 8th c. BC. The Harran Plain retained its importance throughout the entire Assyrian political history. In the region, Huzurina (Sultantepe) was also made a provincial centre. Finally when Nineveh fell in 612 BC Assyrians took refuge in Harran and proclaimed it the capital but soon thereafter, in 610, they lost Harran as well, thus being wiped out from the history.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Middle and Neo-Assyrian Periods of the Harran Plain in Light of a Survey

329

The Harran Plain located at the crossroads of the main routes to the East Mediterranean and Central Anatolia and secondary routes to the south of the Anti-Taurus Mountains was of great importance especially during the Neo-Assyrian period. Pottery collected in the course of surveys is in full coherence with the excavation finds from North Mesopotamia and North Syria. However, these strata at Harran, the most important Assyrian centre in Southeast Anatolia, have not been excavated yet. Harran distinguishes itself from all other settlements through its location at crossroads of all the roads coming from Mesopotamia via Euphrates and Tigris and going towards the East Mediterranean and through its political status first as a provincial centre and then as the capital.

Bibliography Anastasio, S., 1998. La ceramica medioassira di Tell Bari/Kahat. P. E. Pecorella (ed.), Tell Bari/Kahat 2. Relazione sulle campagne 1980–1993 a Tell Bari/Kahat nel bacino del Habur (Siria). Documenta Asiana V. Roma, 135–186. –

2010. Atlas of the Assyrian Pottery of the Iron Age. Subartu XXIV. Brepols.

Bartl, K., 1990. Khirbet Esh-Shenef. A Late Bronze Age Settlement in the Balikh Valley, Northern Syria. Akkadica 67, 10–32. Blaylock, S. R., 1999. Iron Age Pottery from Tille Höyük, South-Eastern Turkey, Preliminary Information on the Pottery of the Iron Age. Hausleiter/Reiche 1999: 263–286. Bourke, J. S., 1993. The Transition from the Middle to the Late Bronze Age in Syria: The Evidence from Tell Nebi Mend. Levant XX, 155–195. Curtis, J., 1989. Excavations at Qasrij Cliff and Khirbet Qasrij. British Museum Western Asiatic Excavations 1, London. Curtis, J./Green, A., 1997. Excavations at Khirbet Khatuniyeh, Saddam Dam Report II, London. Çilingiroğlu, A., 1993. Van-Dilkaya Höyüğü Kazıları Kapanış. Kazı Sonuçları Toplantısı XIV.I, 469–492. Dornemann, H. R., 1981. The Late Bronze Age Pottery Tradition at Tell Hadidi, Syria. BASOR 241, 29–47. Eidem, J./Ackermann, R., 1999. The Iron Age Ceramics from Tell Jurn Kabir. Hausleiter/Reiche 1999: 309–324. Green, A., 1999. The Ninevite Countryside. Post and Places of the Eski-Mosul Region in the Neo-Assyrian and Post-Assyrian Periods. Hausleiter/Reiche 1999: 91–126. Goldman, G. H., 1987. Materialheft spätbronzezeitliche Keramik, Tarsus. Hannover. Guarducci, G./Laneri, N., 2010. Hirmebedon Tepe During the Iron Age Period: A Case Study in the Upper Tigris River Region. Anatolica XXXVI, 17–65. Haller, A., 1954. Die Gräber und Grüfte von Assur. Berlin. Hausleiter, A./Reiche, A. (eds.), 1999. Altertumskunde des Vorderen Orients. Archäologische Studien zur Kultur und Geschichte des Alten Orients 10. Münster. Jamieson, A. S., 1999. Neo-Assyrian Pottery from Tell Ahmar. Hausleiter/Reiche 1999: 287–308

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

330

Coşkun

Karg, N., 1999. Gre Dimse 1998: Başlangıç Raporu. N. Tuna/J. Öztürk (eds.), Ilısu ve Karkamış Baraj Gölleri Altında Kalacak Arkeolojik ve Kültür Varlıklarını Kurtarma Projesi 1998 Yılı Çalışmaları, Ankara, 237–270. Khanzadyan, E. V./Mkırtçıvan. K. A./Parsamyan, E. S., 1973. Metsamor: Issledovanie po dannim raskopok 1965–1966. Erevan. Kozbe, G., 2007. Kavuşan Höyük 2005 Yılı Kazısı. Kazı Sonuçları Toplantısı XXVIII.1, 573–588. Kozbe, G./Köroğlu, K. 2004. Kavuşan Höyük 2002 Yılı Kazısı. Kazı Sonuçları Toplantısı 25.2, 279–289. Köroğlu, K., 1998. Üçtepe I. Ankara. Lines, J., 1954. Late Assyrian Pottery from Nimrud. Iraq 16.2, 164–167. Lumsden, S., 1999. Neo-Assyrian Pottery from Nineveh. Hausleiter/Reiche 1999: 3–15. Luckenbill, D. D., 1926–1927. Ancient Records of Assyria and Babylonia. Vols I+II. Chicago. Matney, T./Rainville, L./Köroğlu, K./Wickle, D./MacGinnes, J., 2007. Report on Excavations at Ziyaret Tepe, 2006 Seasons. Anatolica XXXIII, 23–74. Müller, U., 1999. Die eisenzeitliche Keramik des Lidar Höyük. Hausleiter/Reiche 1999: 405–433. –

2005. Norşun Tepe and Lidar Höyük. Two Examples for Cultural Change During the Early Iron Age. A. Çilingiroğlu/G. Darbyshire (eds.), British Institute at Ankara Monograph 31. Anatolian Iron Ages 5, Ankara, 107–115.

Oates, J., 1959. Late Assyrian Pottery from Fort Shalmaneser. Iraq XXI-2, 130–146. Oates, D./Oates, J./ McDonald, H., 1994. Excavations at Tell Brak, Vol. 1: The Mitanni and Old Babylonian period. Oxford. Ohnuma, K./Numato, H., 2001. Excavations at Tell Taban, Hassake, Syria (3): Report of the 1999 Season of Work. Al Rafidan XXII, 1–65. Oggiano, I., 1999. Cataloguistics: The Experience at Tell Afis, Syria. Hausleiter/Reiche 1999: 377–402. Ökse, A. T./Alp, O./İnal, N., 2004. Ilısu Barajı-Salat Tepe 2000–2002 Yılı Kazıları. Kazı Sonuçları Toplantısı XXV-1, 331–339. Ökse, T./Görmüş, A., 2014. Demir Çağında Salat Tepe. A. Özfırat (ed.), Arkeolojiyle Geçen Bir Yaşam İçin Yazılar Veli Sevin’e Armağan. Scripta. Essays in Honour of Veli Sevin. A Life Immersed in Archaeology. İstanbul, 233–257. Özfırat, A., 2005. Eskiçağ’da Harran. İstanbul. –

2007a. Üçtepe II, Tunç Çağları (Kazı ve Yüzey Araştırması Işığında). İstanbul.

–

2007b. Van, Ağrı ve Iğdır İlleri Yüzey Araştırması, 2005. Araştırma Sonuçları Toplantısı 24.1, 143–163.

Pfälzner, P., 1995. Mittanische und mittelassyrische Keramik Eine chronologische, funktionale und produktionsökonomische Analyse. Berlin. Postgate, C./Oates, D. /Oates, J., 1997. The Excavations at Tell Al Rimah: The Pottery. Warminster. Reiche, A., 1999. Iron Age Pottery from Tell Rad Shaqrah (North-East Syria). Hausleiter/Reiche 1999: 231–259.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Middle and Neo-Assyrian Periods of the Harran Plain in Light of a Survey

331

Roaf, M., 2001. Continuity and Change from the Middle to the Late Assyrian Period. R. Eichmann/ H. Parzinger (eds.), Migrations und Kuturtransfer. der Wandel vorder- und zentralasiatischer Kulturen im Umbruch vom 2. zum 1. vorchristlichen Jahrtausend: Akten des internationalen Kolloquiums, Berlin, 23. bis 26. November 1999. Kolloquien zur Vor- und Frühgeschichte 6, 357–369. Rossmeisl, I., 1989. Late Bronze Age Pottery of Tell Sabi Abyad. P. M. M. G. Akkermans (ed.), Excavations at Tell Sabi Abyad. Oxford, 337–356. Schachner, A., 2002. Ausgrabungen in Giricano (2000–2001). Neue Forschungen an der Nordgrenze des mesopotamischen Kulturraums. Istanbuler Mitteilungen 52, 9–57. Schachner, A./Schachner, Ş., 2003. 2000–2001 Yılı Gricano Kazıları. Kazı Sonuçları Toplantısı XXIV.2, 447–461. Schmidt, C., 1999. Die Keramik der Areale A–F in Kar-Tukulti-Ninurta. Hausleiter/Reiche 1999: 61–90 Schneider, E., 1999. Die Eisen Keramik von Tell Sheikh Hassan (Syrien). Hausleiter/Reiche 1999: 325–346 Sevin, V., 1990. 1988 Yılı Diyarbakır/Üçtepe Kazısı. Kazı Sonuçları Toplantısı XI, 103–125. –

1995. İmikuşağı I. (6.–1. Yapı Katları). Ankara.

–

2004. Son Tunç/Erken Demir Çağı Van Bölgesi Kronolojisi Kökeni Aranan Bir Devlet: Urartu, Belleten LXVIII, 355–386.

Sevin, V./Kavaklı, E., 1996. Bir Erken Demir Çağ Nekropolü Van/Karagündüz. İstanbul. Şenyurt, Y., 1988. Girnevaz Kazıları Işığında Geç Assur Devri Düğme Dipli Vazoları, Ankara Üniversitesi Dil ve Tarih-Coğrafya Fakültesi Dergisi XXXII.1/2, 281–291. –

2006. Aşağı Salat Tepe’den Bir Yeni Assur Mezarı (M-29). B. Avunç (ed.), Hayat Erkanal’a Armağan Kültürlerin Yansıması, Studies in Honor of Hayat Erkanal Cultural Reflections, İstanbul, 697–705.

Tekin, H., 2004. Hakemi Use 2001 Yılı Kazısı İlk Sonuçları 2001. N. Tuna et al. (eds.), Ilısu ve Karkamış Baraj Gölleri Altında Kalacak Arkeolojik ve Kültür Varlıklarını Kurtarma Projesi 2001 Yılı Çalışmaları. Ankara. Ortadoğu Teknik Üniversitesi, TAÇDAM, 425–449. –

2009. Hakemi Use 2008 Yılı Kazılarında Keşfedilen Geç Assur Dönemi Mezarları. Hacettepe Üniversitesi, Edebiyat Fakültesi Dergisi 26-2, 193–211.

Xnkikyan, O. S., 2002 Syunik During the Bronze and Iron Ages. Barrington. Yardımcı, N., 2004. Harran Ovası Yüzey Araştırması I–II, Archaeological Survey in The Harran Plain I–II. İstanbul.

Dr. Nilgün Coşkun, Mustafa Kemal University, Faculty of Sciences and Letters, Department of Archaeology, Tayfur Sökmen Campus, Serinyol, Antakya, Hatay, Turkey. E-mail: [email protected].

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9th ICAANE, Basel 2014, Vol. 3, 333–346

Anacleto D’Agostino – Valentina Orsi

Researches at Uşaklı Höyük (Central Turkey): Survey, Surface Scraping and First Digging Operations The Italian Archaeological Expedition in Central Anatolia has carried out a multidisciplinary research on the site of Uşaklı Höyük and its territory in the years 2008–2013. This paper will present an overview of the results obtained by the survey and 2013 season of excavations.1

1. From Survey to Excavations: Context, Methods and Results of the Research The research project of the University of Florence at the site of Uşaklı Höyük has been active for the last seven years under the direction of Stefania Mazzoni. Between 2008 and 2013, the Italian Archaeological Expedition in Central Anatolia has carried out an intensive survey of the site and its territory, producing a rich corpus of materials and information that allows the investigation of the landscape and settlement pattern of this distinct and partially under-explored area of the plateau. This paper will present a summary of the archaeological results obtained from the survey and a preliminary overview of the 2013 excavation season.2 Uşaklı Höyük lies on the southern bank of the Egri Öz Dere, at about 1120 m above sea level, in a land characterised by plains and low hills delimited to the south by the high profile of the Kerkenes Dağ and to the northwest by the mountains of Yozgat (fig. 1). The site is clearly visible from the route connecting Yozgat and Sivas, and its position along one of the main E–W roads crossing the plateau favoured visits by travellers and scholars since the first half of 20th century. In the reports of the first western visitors about the area of

1 2

A. D’Agostino is the author of section 1; V. Orsi of section 2; the conclusion has been co-written. Further details and information on the results of the project are available in the preliminary reports published regularly during the last years. In particular see Mazzoni 2012; Mazzoni et al. 2011; 2014; Pecchioli et al. 2014.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

334

D’Agostino – Orsi

Sorgun, the site is referred to also as Uçaklı, Kuşaklı and Kusachakly (Mazzoni et al. 2011: 91; Summers et al. 1995: 53–59). The importance of this area within the regional road network is undeniable, since ancient periods when the visibility from the bottom of the valley of the Eğri Öz Dere of important landmarks, such as the Kerkenes Dağ (Summers 2014: 41–44), as well as the mountains surrounding Yozgat the north Pontic chains represented valid visual supports along the itinerary from north-western and northern areas of the plateau towards Kayseri. The position of the site along the route probably linking the Hittite capital Boğazköy/Khattusa to Alişar Höyük/Ankuwa, its distance from these two sites and its visual relation with the imposing high granitic peak of the Kerkenes Dağ were some of the elements that supported the identification of Uşaklı Höyük with the holy city of Zippalanda, mentioned in the textual sources as an important cult centre of the Storm God Tahawašezzu, son of the Storm God of Hatti.3 Uşaklı consists of a somewhat oval flat mound named the terrace, with an extent of around 10 ha, at the south-eastern edge of which rises an approximately circular high mound covering an area of about 2 ha (fig. 2a). The mound has steep slopes covered by vegetation and used for grazing cows. The terrace and its slightly sloping sides have been cultivated, until recent times. The resulting periodic ploughing activity, that inevitably disturbed the buried strata, brought to light a variety of archaeological materials that attracted our interest. The Uşaklı Höyük Archaeological Project included different field activities, to which the different seasons of work were dedicated. The survey focused on the mound of Uşaklı and a small area immediately around the site, with a range of ca. 5/6 km. This surrounding area represents the hinterland of the site, a variegated environment encompassing hilly lands and rocky outcrops, where streams and springs, field and pasturelands are the main components of the physical landscapes. Firstly a team of archaeologists and geologists examined the territory of Uşaklı, their work aimed at understanding the anthropic and physical landscape. Consequently an integrated and multidisciplinary prospection has been planned and carried out on the field, including geomorphological and geophysics researches addressing different questions. The archaeological survey of the surrounding area was primarily intended to illustrate its long-term settlement pattern and the development of its occupation. Through using satellite imagery and topographic purpose-built maps, in addition to walking among fields and hills, 18 sites of archaeological interest have been found. Only two mounded sites gave back a discrete collection of sherds: site 3, close to the village of Büyük Taşlık and site 9, close to Aşağı Karakaya Köy, dating respectively to the Late Chalcolithic and Late Roman/ Byzantine periods. Most of the sites were identified due to the occurrence of few pottery sherds scattered over a field or flanks of hills (sites 2, 5–7, 11–12, 14); in some cases due to the concentration of some stone blocks and sherds (site 2) or the presence of small silos and 3

See Mazzoni/Pecchioli 2011 for arguments and bibliography concerning the proposed identification.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Researches at Uşaklı Höyük (Central Turkey)

Fig. 1. a–b. The location of the site and a map of Central Anatolian Plateau showing some sites dating to the Hittite period. c–d. Views of Uşaklı Höyük from N and from NNW (in the background the Kerkenes Dağ).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

335

336

D’Agostino – Orsi

some pits/cisterns (sites 15, 16, 19); and in one case (site 4) we found probable traces of a large buried wall in connection with a concave area. Given the small amount of potsherds present on the surface, it is hard to assign a specific time period to these sites. According to the evaluation of the few diagnostic types collected, only an approximate date between the late 1st millennium BC and, mainly, the Common Era can be suggested. In addition, other places have been registered for their potential relevance: a quarry (site 13), the location of a worked stone of Hittite manufacture (site 8) and a mound but apparently without sherds (site 10). The emerging pattern from the survey is that of a territory occupied by sparse small villages or more probably farmsteads of Later Periods, with no sites dating to the Bronze and Iron Ages, with the exception of Taşlık Höyük and Uşaklı (site 1). It could also be true that the potential of identifying the earlier sites, often not on mounds, could have been hindered by the intense agricultural use of the land as well as by the heavy erosion and sediment accumulation that characterise this region. The geological analysis of the territory was aimed at identifying geologic and geomorphologic structures useful in reconstructing the ancient natural background of human settlements and the identification of exploitable sources of raw materials. One of the targets of this research, is to determine the area of provenience of the large granitic blocks used for buildings and often found protruding from the surface of Uşaklı. With the permit of the Turkish authorities, a large number of rock samples collected during the prospection will be analysed in the laboratory of Siena University. The fieldwork at Uşaklı, the main site of the area, consisted of geophysical and archaeological prospections to prepare for the subsequent more invasive scraping and digging operations. The geophysical prospection and collecting activities have been carried out together, proceeding gradually in both and using the same georeferenced grid. The results of the geomagnetic and the geoelectric prospections, from both the terrace and the high mound, turned out to be fruitful (fig. 2b–d). Several anomalies have been registered under the surface soil and their processing produced clear results that can be interpreted as large articulated buildings, city walls, dwellings, alignments of walls, possible streets as well as other isolated structures of different sizes and shapes. An urban layout seems to emerge with distinct features of an ancient 2nd millennium BC city with public buildings, such as Building II reminiscent of a typical Hittite temple. It is noteworthy to highlight here, that the structures individuated from the geophysical analyses are not all necessarily contemporaneous and could belong to different chronological phases. During the 2008–2010 seasons, the systematic collecting of the materials scattered on the site demanded a great deal of efforts and organization. The planned intensive sampling strategy of all the artefacts relating to the main topographical or surveyed units aimed at achieving precise spreadsheets of the different categories of findings for scatter analysis. The collection was carried out in several areas of the mound surface, namely the top of the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Researches at Uşaklı Höyük (Central Turkey)

337

Fig. 2. a. Topographic map of Uşaklı Höyük. b–c. Geomagnetic anomalies. d–e. Geoelectric anomalies and location of Area A. f. Scraping on the southern slope of the high mound.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

338

D’Agostino – Orsi

terrace and its low slope that gradually descends to the valley floor, and the south-eastern slope of the high mound. In fact, in the course of a preliminary extensive inspection these areas of the site showed a major concentration of sherds, whereas the other sectors were characterised by a very low occurrence of materials. The surface of the site was divided into precise survey units, of square or irregular shapes, depending on the morphology or the topographic grid. All the scattered artefacts and diagnostic sherds were collected from and assigned to these units during the systematic fieldwalking. Rakes, shovels and rods were used to clean the selected sectors from the residuals of the crop season and natural vegetation, which would have compromised the visibility of the soil surface. Furthermore, the whole site was surveyed in order to evaluate the validity of the representative sample and test the quality of findings from different sectors of the settlement in terms of functionality or at least of archaeological deposits. The same portions that disclosed clearer evidence during the geophysics prospection gave back more substantial collection of materials, in terms of state of preservation, variety and relevance of documented categories. This is due to the fact that in some areas of the site, namely in the eastern sector of the flat terrace, significant archaeological layers are nearer to the surface and are not covered by a thick layer of eroded deposits coming from the high mound or from decayed mud-brick structures. Herbaceous plants have always covered the surface of the high mound during our fieldwork period, preventing a good visibility of the soil. Because of its steep sides the high mound was not used for agriculture and it has not been ploughed. The cutting and removal of the vegetation with sickles and rakes, which proved to be an efficacious approach in the survey of the terrace, turned out to be ineffective here given that the surface soil was very compact and had very few exposed materials. Therefore, the scraping operations on the slopes of the high mound of Uşaklı that took place in 2012 was necessary to gather evidence on the occupational sequence of this area of the site, where the few sherds collected in the previous years did not produce enough satisfactory information and left us a little bit puzzled concerning the rarity of findings dated to the Later Periods, within our first poor collection. Focused deep scraping operations of the superficial layers of soil have been effectuated on the southern and south-eastern slopes, which correspond to a long anomaly in the upper portion of the slope revealed by the geomagnetic prospection and interpreted as a well-structured wing of a large building with large rooms and courtyards of regular plan. Removing by scraping the topmost layer of soil by means of trowels and shovels was then the best method to obtain results from the slope surface (fig. 2f). This method allowed us to find a good amount of sherds and expose also the upper surface of sizeable archaeological features, such as traces of collapsed mudbrick walls that were in some cases burned. Hence, the location and limits of the reddish, brownish and buff spots were measured and their layout was overlapped with the geomagnetic anomalies of the buried building, resulting in some cases with a good correspondence (Mazzoni et al. 2014: fig. 12). Because of the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Researches at Uşaklı Höyük (Central Turkey)

339

sloping of the surface soil, a fine-grained characterization in chronological terms of the different collections is obviously hard to be determined, nonetheless interesting data have been collected and are reported in the following section. Noteworthy was the finding of three fragmentary tablets with cuneiform inscriptions, possibly in relation with the geomagnetic anomalies mentioned above. Two small test soundings investigated the depth of some significant anomalies and features of a probable surfaced segment of a wall. The first was located at the base of the southern slope, where the geomagnetic prospections signaled a large 3/4m thick anomaly. The excavation of a homogeneous deposit of soil reaching a depth of 1.20m allows us to exclude the proximity of a buried structure to the surface; therefore, the structure potentially related with the anomaly is likely to be located at a deeper level. The second sounding was done on the top of the mound, where some large stones protruded from the surface, possibly related with the strong pseudo-circular anomaly clearly visible in the geomagnetic map. Here, the emerging stones were cleaned and part of a wall has been exposed in a superficial dig; it consisted of two rows of ashlars separated by a row of middle-sized stones, with no trace of mortar. One of the purposes of our project was to approximately locate the levels dating to the 2nd millennium BC. The extensive techniques adopted in the site survey allowed us to correlate the distribution of artefacts (see below in section .2) with the large structures detected under the surface by the geophysical prospection, determining the two areas to be examined in details and allocating the excavation trenches according to the periods to be investigated. Consequently, in 2013 the first large excavation area was opened in the eastern portion of the terrace, where scattered massive granite blocks were preserved on the surface and partially in situ till 2011 (during the winter of 2012 some stones have been removed and piled up by the landowner in order to facilitate his ploughing activity). This area was also selected for two other reasons: the geoelectric survey was especially productive and registered the presence of a structure likely interpreted as a large building of about 875m² (fig. 2e); the scattering of sherds indicated that this area was potentially recommended given the concentration of Late Bronze Age materials and low occurrence of material dating to the later periods. The first digging operations exposed directly under the topsoil the groundwork of a large building preserved for a unique course (fig. 3). A series of well structured walls (width of 2.10 m and 2.70 m) made of large rough granite blocks (fig. 3b) laid on a sort of deep under-foundations of middle-sized stones were traced (fig. 3c); they delimit elongated spaces (measuring between 4,25 x 1,60 m and 4,60 x 7,50 m). The orientation of the longer rooms, probably a sort of cellars, is NW–SE. The building shows no traces of the original collapse and/or destruction; and only very small segments of badly preserved beaten earth floors have been found in two different areas. Some of the stone blocks, originally part of the foundation, were relocated, occupying the inner space between two walls. There is evidence of

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

340

D’Agostino – Orsi

Fig. 3. a. Plan of the Building II in Area A. b–c. Particular of the walls and platform in course of digging. d. The portion of Building II exposed in the 2013 season.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Researches at Uşaklı Höyük (Central Turkey)

341

Chart 1.

a partial reuse of some rooms, such as the presence of different built walls bonding with the main walls or overlaying them and marginal addition of light structures. The building, after its first phase of use, was left exposed for a long time and used for different purposes, and lastly progressively dismantled and its materials reused. This fact, in addition to the intense farming activity that levelled eventual relics of the mud bricks upper part of the walls, could explain the absence of any debris of the original collapse of the elevation and the preservation of only large heavy boulders of the foundation. The regular plan and the general concept of the building, as visible through the processing of geoelectric results and results of the first excavations, as well as its monumental character reminds of the layout of the temples of Boğazköy. In particular the thickness of the walls, sustaining probably an upper storey, finds comparisons with the Temple I of the Lower City. A small sounding (1.80 m deep) on the north-eastern side of the building has revealed a sequence of eight cobbled floors separated by thin accumulations of soft clayish earth, containing pottery and animal bones. It is probably a part of some preparatory work intended to regularize and strengthen the area for the building of the structure; however, at the moment a cultual purpose (a ritual to clear or to purify the area?) has not been ruled out. The impressive architecture we started to expose in Area A in the course of the 2013 campaign, the fragments of tablets and a clay bulla found in the survey suggest the importance of the settlement at the end of 2nd millennium BC, at the time of the Hittite rule over the region and – as the following section will illustrate – probably already existing during the older phases.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

342

D’Agostino – Orsi

2. The Findings, Their Dispersion and Preliminary Dates of the Occupation The corpus of archaeological artefacts recovered after the surface survey and scraping consists for the greatest part of potsherds, but also a notable amount of roof tiles and slags were collected (chart 1). Lithics on the other hand were only sporadically found. The distribution pattern of roof tiles points out to some form of occupation of the site during the Late Roman/ Byzantine periods, apparently clustered on the top of the northern sectors of the lower terrace. Roof tiles are essentially absent from the slopes of the höyük (fig. 4a–b). Samples of amorphous slags, composed of a variety of melted combustion residues, were found scattered all over the surveyed areas with a remarkable density noted in some of the eastern sectors of the terrace (fig. 4a–b). They are likely to be connected with ancient fire events. Potsherds are ubiquitous (fig. 4a–b). Simple body-sherds without additional features (like paint or plastic elements) have been sorted out as ‹generic›, whereas all of the other sherds generally connectable with vascular shapes or decorative styles have been sorted out as ‹diagnostic›. The great majority has been classified according to broad chronological periods, finer dating was arduous to achieve in many cases, given that our effectiveness in dating non-stratified assemblages is deeply affected by the relative continuity in the central Anatolian ceramic productions over long chronological periods. As far as the results of the surface survey on the terrace are concerned, a few sherds of glazed pottery dated to the medieval period document the most recent occupation of the site. The ceramics dated between the late Iron Age and the Byzantine periods were mainly scattered over the western portion of the terrace. Iron Age period sherds, possibly as a result of disturbances from modern ploughing activities, were homogeneously distributed over the surveyed area. Most of the representative inventories dated to the Late Bronze Age period are clustered on the northern and eastern slopes of the terrace, possibly in connection with the buried structures identified by geophysics survey. A considerable amount of sherds, more likely connected with a Middle Bronze Age ceramic horizon was concentrated on the south-eastern slopes of the terrace. A group of hand-made sherds with red slipped surface, probably to be dated to late Early Bronze Age, account for the most ancient occupation of the site (Mazzoni et al. 2010: 112–118; 2011: 94–96; 2012: 322–324; D’Agostino/Orsi 2011). In all likelihood, the substantial post-depositional displacement of surface archaeological finds connected with the steepness of the southern slopes of the high mound of the höyük deeply affects a fine chronological classification of the units selected for scraping. The percentage of ceramic sherds per m2 recovered on the slopes after the scraping activity is decidedly higher if compared to that acquired after the surface survey.4 The ceramic 4

On the slopes of the höyük the ratio between scraping and surface survey is 8,67 to 1 ceramic items per m2. The method of scraping resulted in 5,2 ceramic items per m2, whereas the sole surface

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Researches at Uşaklı Höyük (Central Turkey)

Fig. 4. a. scattering of materials (survey). b. scattering of materials (scraping). c–d. sherds dating to LBA and MBA periods (plates and red-slip specimens from deep sounding).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

343

344

D’Agostino – Orsi

inventory included a significant amount of Painted Wares of the Middle and Late Iron Age periods, largely comparable to the so-called Phrygian and Alishar IV–V repertoire. The distribution pattern of Middle Iron Age sherds over the sloping area is quite homogenous, hinting to a corresponding uniform distribution of contemporaneous buried archaeological evidence, or possibly resulting from the erosion processes affecting the upper part of the mound. Despite some difficulties in the chronological attribution of some of the ceramic types (see Mazzoni et al. 2014: 256–257 for more details), Middle and Late Bronze Age periods are well attested by the presence of Red Slip and Drab Ware diagnostic shapes. Their distribution over the sloping area is also unexpectedly consistent, thus suggesting that the contemporaneous deposits have not been substantially concealed by later accumulation. Similarly, unforeseen was the substantially low percentage of later ceramics of Hellenistic or Byzantine periods, that leads us to hypothesize that the occupation of the main mound during these periods was not particularly conspicuous. Minor evidences of late 3rd and early 2nd millennium BC periods are provided by Hand-made red-slip wares and painted sherds, which distribution pattern is mainly scattered and sporadic (Mazzoni et al. 2014: 256–257). Among the surface and scraping finds, remarkable is the recovering, concentrated between the southern and south eastern low slopes of the höyük, of 4 fragmentary tablets with cuneiform inscriptions, dated to the Middle to Late Hittite period.5 Preliminary readings interpret them as letters and a mythological text. Moreover, a clay bulla with two sealed impressions made by a circular stamp, dating to the Hittite period, was recovered from the northern slope of the terrace (Mazzoni et al. 2011: 92). Unexpectedly, the superficial fillings of Building II, the excavations of which started in 2013, yielded only very sporadic fragments of roof tiles as well as slags, suggesting that the samples recovered from the surface survey in the same area must have originated from archaeological remains located elsewhere. No primary context with associated materials has been reached so far, but the secondary context accumulations covering the structure provided a rich ceramics inventory. The chronological span is quite large, reflecting the same sequence attested on the surface. The prevailing ceramic class is represented by Common ware; nonetheless, Kitchen along with Storage wares are well attested. Contrarily, Fine ware was only sporadically found. In the accumulation layer immediately above the poorly preserved floor of one the rooms of the building, a set of 3 miniature vessels has been found (an ointment vase, a conical cup and a plate). They belong to cultic pottery equipments of the Late Bronze period,

5

survey on the slopes of the höyük returned approximately 0,6 ceramic items per m2. The first and most sizeable tablet has been recovered during the surface survey of the terrace southern slopes (Mazzoni et al. 2011: 92; Pecchioli et al. 2014); the other three minor fragments result from the scaping of the lower band of the höyük southern and south-eastern slopes (Mazzoni et al. 2014: 257 and fig. 11).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Researches at Uşaklı Höyük (Central Turkey)

345

thus providing additional support for the general functional interpretation of the original structure. A large assemblage of Common and Kitchen ware can be attributed to the Late Bronze period (15th–13th cent. B.C.). Notable is the presence of the coarse platters with string impressions on the rim, which constitute an important bulk of the whole assemblage. Drab ware, in the prevailing shape of simple rim bowls, is also abundant. The typology of Common ware production seems to be quite standardized, with marked predominance of open shapes, namely small bowls with rounded profile and simple rim, and medium size bowls with inner-thickened or in-turned rim. A medium percentage of wheel made red slip ware is also present. Most recurring shapes are upper-convex side bowls with vertical, triangular lugs; small, carinated bowls with vertical upper section; necked jars and spouted vessels. A small corpus of painted sherds, usually hand-made with burnished surface, belongs to the Early Cappadocian/Intermediate and Middle Bronze Age horizons, whereas a group of hand-made red-slip ceramics, mainly chaff tempered with unevenly burnished surface, should be assigned to the Early Bronze Age. The Middle and Late Iron Age horizon is also well documented and ubiquitous, especially in the upper portion of the filling layer, and consists of painted geometric wares. The pottery inventory from the deep sounding under Building II is not homogeneous, but typical late Bronze Age ceramic typologies, including platters with string impressions on the rim (fig. 4c) constitute the main and the more recent attested ceramic horizon. Wheel made Red slip ware sherds are also quite frequent (fig. 4d), as well as hand-made sherds both with red slip or painted surface dating to the late Early Bronze Age and the Middle Bronze Age.

3. Conclusions The analysis of the ceramics assemblages leads us to suggest that the settlement of the region around Uşaklı Höyük reached its peak most probably during the later periods (from late 1st millennium BC to Late Roman/Byzantine periods). During the Iron Age, the area was apparently scantily settled, as no mounds or scatters of materials belonging clearly to this phase were identified. The site of Uşaklı itself was occupied from the end of the Early Bronze Age to the Byzantine period. The fact that most of the Late Bronze Age and even earlier materials dating to the Middle Bronze Age period were found on the outskirts of the terrace can substantiate the hypothesis of a significant settlement consisting of a lower town and an acropolis already existing during this older phase. Its floruit, corresponding to an intensive occupation, was in the course of the entire 2nd millennium BC. During this phase, the site already covered 10 ha, as the distribution of diagnostic sherds suggests. Noticeably,

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

346

D’Agostino – Orsi

no materials of the Old Hittite or Imperial periods have yet been collected in the territory. However, even if still inadequate to offer a definitive confirmation to the identification of Uşaklı Höyük with ancient Zippalanda, the results of the first archaeological researches on the site provide a strong support to the identification. They testify to a markedly important city of the Hittite period, well situated within the geographical setting associated with the ancient holy centre.

Bibliography D’Agostino, A./Orsi, V., 2011. I materiali della ricognizione archeologica nell’area di Uşaklı Höyük. S. Mazzoni et al. (eds.), Ricerche italiane in Anatolia. Risultati delle attività sul campo per le età del Bronzo e del Ferro. Studia Asiana 6. Roma, 129–156. Mazzoni, S., 2012. Uşakli Höyük A Hittite Site in Central Anatolia. Origini XXIV, 401–415. Mazzoni, S./D’Agostino, A./Orsi, V., 2010. Kuşaklı 2008 Survey Season (Yozgat-Sorgun): 1st Preliminary Report. 27.Araştırma Sonuçları Toplantısı 3.Cilt, 333–342. –

2011. Uşaklı/Kuşaklı Höyük (Yozgat-Sorgun): the 2009 Survey. 28. Araştırma Sonuçları Toplantısı

–

2012. Uşaklı 2010 Survey Season (Yozgat). 29. Araştırma Sonuçları Toplantısı 2. Cilt, 317–333.

– 3. Cilt, 91–104. Mazzoni, S./D’Agostino, A./Orsi, V., 2014. Uşaklı Höyük 2014: Surface Scraping and Collecting on the High Mound. 35. Kazı Sonuçları Toplantısı – 1. Cilt, 253–270. Mazzoni, S./Pecchioli, F., 2011. La ricognizione archeologica nel territorio di Uşaklı Höyük (Yozgat). S. Mazzoni et al. (eds.), Ricerche italiane in Anatolia: risultati delle attività sul campo per le età del Bronzo e del Ferro. Studia Asiana 6. Roma, 105–128. Mazzoni, S./Pecchioli, F./Torri, G./D’Agostino, A. (eds.), 2011. Ricerche italiane in Anatolia: risultati delle attività sul campo per le età del Bronzo e del Ferro. Studia Asiana 6. Roma Pecchioli, F./Torri, G./Corti., C., 2014. The Survey in the Area of Usakli Hoyuk (Yozgat): Archaeologic and Epigraphic Findings. P. Taracha (ed.), Proceedings of the Eighth International Congress of Hittitology, Warsaw, 5–9 September 2011. Warsaw, 671–681. Summers, M. E. F./Summers, G. D./Ahmet, K., 1995. The Regional Survey at Kerkenes Dağ: an Interim Report on the Seasons of 1993 and 1994. Anatolian Studies 45, 43–68. Summers G. D., 2014. East of the Halys: thoughts on settlement patterns and historical geography in the late 2nd millennium and first half of the first millennium B. C. H. Bru/G. Labarre (éd.), L’Anatolie des peuples, des cités et des cultures (IIe millénaire av. J.-C.-Ve siècle ap. J.-C.), vol. I. Besançon, 41–51.

Anacleto D’Agostino & Valentina Orsi, University of Florence.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 347–359

Sevinç Günel

A New Centre of Intercultural Relations in Western Anatolia during the Late Bronze Age: Çine-Tepecik The geographical location of Western Anatolia plays an important role in the cultural development of the region. Tepecik lies on the Çine Plain to the south of the Meander. The material culture of Tepecik demonstrates that the settlement was part of both the Mycenaean and the Hittite spheres of influence. In the region of the Meander the valleys opening over the Büyük Menderes River (the Meander in antiquity) to the coast of the Aegean Sea on one hand and the natural ways stretching out to the north, south and east on the other hand are an important factor in the cultural history of the region. In the region of the lower Maeander, the Çine Plain is one of these natural ways opening towards the south. The Çine region connects the Aegean Coast and Central Anatolia by means of natural passes between the mountain ranges. Tepecik lies on the Çine Plain, to the east of the Çine Çayı (fig. 1). Tepecik has a long settlement history spanning from the Neolithic until the end of the Late Bronze Age (Günel 2008: 129–139; Günel 2011: 217–232; Günel 2012a: 55–80; Günel 2014a: 243–256; Günel 2014b: 83–103; Günel 2015: 627–643). In addition to early settlement remains, Carian-Geometric remains were found on the surface of the mound (Günel 2008: 133, fig. 7). In the stratified development of the mound, Levels II 1 and II 2 of the settlement date to the 2nd millennium BC. Culture Level II 1, belonging to the Late Bronze Age, reveals a fortified settlement plan (fig. 2). The wall was strengthened by square towers. In terms of both construction technique and the thickness of the wall, the fortification system reflects a strong defensive purpose. The settlement spreads to the south and east of the fortification wall. The wall and building remains reflect a settlement model extending over a large area. Architectural remains from this settlement contain finds dated to the latest period of the Late Bronze Age. In this Level II 1, the cultural remains draw attention, the data helping to explain the economic forces of the settlement and the economic elements of its administrative system. The buildings have

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

348

Günel

Fig. 1. Location of Çine-Tepecik in the region of the Meander.

rectangular plans and are single-roomed (fig. 2; ‹Building 1-2›). The walls of the buildings are constructed of mud bricks over stones. The architecture, with foundations built of small stones, reflects the building tradition known from Western Anatolia. The finds in the settlement context show signs of an intense fire. A large number of in situ vessels, pithoi, metal finds and seal impressions have been discovered on the floor belonging to the destruction horizon in the architectural remains. Large storage jars sitting on the floors of the buildings indicate a storage system. The various shapes among the vessels and pithoi support the identification of these buildings as ‹storage›. The pottery in this level consists predominantly of local fine, medium-fine and coarse wares. The fine, medium-fine and coarse ware is a well-known type of the Bronze Age fabric of the Meander region. Among the vessels, plates, bowls with red painted decoration, deep bowls and jugs reflect the local Western Anatolian pottery tradition of fine and medium-fine wares and shapes. The buff-ware plates and carinated bowls are characteristic of local wares, a well-known type of pottery from 2nd millennium BC Western Anatolia (Blegen et al. 1953; Blegen et al. 1958; Günel 1999a; Günel 1999b: 51–52, fig. 12). On the floor of the one of the buildings, there was a large quantity of vessels, and pithos fragments

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A New Centre of Intercultural Relations … Çine-Tepecik

Fig. 2. Settlement plan and stratigraphical development of the mound.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

349

350

Günel

Fig. 3. Stirrup jars in the Building 1.

lying beside and on top of each other (fig. 2; ‹Building 2›). Pithoi supported by small stones laid around them were found in the building. These had been broken by a destruction event which led to the collapse of the buildings. The pithoi and the vessels were scattered especially in a southwesterly direction from this building. Within the pottery finds there are Mycenaean fine and medium-fine wares in this level (fig. 3). In ‹Building 1›, the stirrup jars show very typical Mycenaean motifs that are often found on various vessel types in the Aegean (Günel 2010: 32–35, fig. 8–10). The stirrup jars are short conical (FS 181) or conical body forms (FS 182). The main motifs on the shoulder of the stirrup jars are dot-fringed semicircles (FM 43i), ‹flowers› (FM 18) and circle motifs (FM 27: 35) (fig. 3). The top of the false neck is decorated with concentric circles. The body of the stirrup jars is decorated with horizontal bands. The stirrup jars can be related to the forms and decorative motifs found in the Aegean region. The motif of dot-fringed semicircles on the shoulder is seen on stirrup jars dated to the Late Helladic phase in Argolid (Mountjoy 1999: 179, fig. 54: 410), Berbati (Holmberg 1983: 30, fig. 16: 23), Patras from Achaea (Mountjoy 1999: 435, fig. 155: 123), Lefkandi on Euboea (Popham et al. 2006: 206, fig. 2.32: 5), from Greece and Kephallonia – Lakkithra in the Ionian islands (Souyoudzoglou-Haywood 1999: 73–74, fig. 11:7, pl. 13: A 1339), Ialysos on Rhodes (Mountjoy 1999: 1053, fig. 431: 204), and Langada on Kos (Mountjoy 1999: 1080, fig. 441: j). Parallels to the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A New Centre of Intercultural Relations … Çine-Tepecik

351

Fig. 4. Mycenaean vessel fragments (krater or deep bowls) with figurative designs.

semi-circles motif used on stirrup jars of the LH III B2 phase are also found on the stirrup jars of the LH C early phase in the Aegean. The motif of semi-circles (FM 43) can be seen on different vessel forms in a similar chronological development in the Aegean (Mountjoy 1999: 390, fig. 135: 69; 1122, fig. 461: 178; Thomatos 2006: 94, fig. 1. 293). Besides on the stirrup jar, the motif of semi-circles is present also on the sherds in Level II 1 from Tepecik (Günel 2010: 38, fig. 22, tab. 1:4). Furthermore, the ‹flower› motif on the shoulder of the stirrup jar is known from Attica (Mountjoy 1999: 548, fig. 198: 227). In the light of all the comparable material, the Mycenaean examples in ‹Building 1› display characteristics pointing to a date in the LH III B2 and LH III C early phases (Günel 2010; Günel 2015: 632–635, fig. 7). In the Level II 1 Mycenaean-decorated vessels, the subjects portrayed on the deep bowls and kraters reflect a tradition parallel to the ‹pictorial style› of vessels known from the Aegean and the Eastern Mediterranean world. Mycenaean pottery with figurative designs is present in ‹Building 2› and in the debris in Level II 1 along with local pottery finds. Among the Mycenaean vessels with figurative designs, the krater fragments are significant with regard to their ‹pictorial style› (Günel in: Günel-Herbordt 2014, 4–8, fig. 3–7). On one of the kraters, a warrior with a ‹hedgehog-type› helmet is depicted throwing a spear (Günel in: Günel/Herbordt 2014: 4–5, fig. 3–5). Examples of the warrior with a ‹hedgehog-type›

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

352

Günel

Fig. 5. A Mycenaean vessel fragment (krater) with figurative design.

helmet on kraters are known from pottery dated to the LH III C Early and Middle phase in Tiryns (Güntner 2006: 177, fig. 1), and Kynos (Wachsmann 1998: 131, fig. 7, 8; Wachsmann 2000: 115, fig. 6, 13–15; Dakoronia 2006: 24, fig. 1) in Greece. On the other krater is a scene in which animals attack a deer. This scene depicts a bird between the legs of a deer (Günel in: Günel/Herbordt 2014: 5–7, fig. 6–7). Parallels to this hunting scene on the krater are also known from Pisaskion in Messenia in the LH III C phase (Blegen et al. 1973: 226, fig. 289; Mountjoy 1999: 355, fig. 122:128) and Aradippo on Cyprus (Vermeule-Karageorghis 1982: 50, 203–204, V. 60). Such depictions on kraters or bowls of birds between the legs of horses and goats are well known from the Mycenaean pictorial pottery tradition. Examples of birds together with other figures and motifs on vessels are known from Mycenae (Crouwel 1991: fig. 10; Crouwel 2006: 20, fig. 16) and Tiryns (Güntner 2000: 58–59, pl. 24.1b) in Greece and Seraglio on Kos (Crouwel 2006: 20, fig. 17) in the Aegean in the LH IIIB2–LH III C phase. In addition to these examples, on a fragment, a goat is depicted besides a plant (fig. 4). A similar pictorial design is known from Mycenaea (Crouwel 2006: 16, fig. 4) and Perati (Crouwel 2007: 78, fig. 15) in Greece. Another fragment from Tepecik shows a warrior or a hunter with his bag (fig. 5). A similar figure on a krater is known from Kalapodi (JacobFelsch 1987: 39, fig. 5; Mountjoy 1999: 815, fig. 325: 28). In addition, scenes depicting birds with regional characteristics are well known from the pictorial pottery tradition. Various

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A New Centre of Intercultural Relations … Çine-Tepecik

353

Fig. 6. Stone-paved platform with storage pithoi.

types of bird representations can also be seen among the pottery repertoire from Tepecik (Günel 2015: 637, fig. 15–16). The Mycenaean vessels prove that the settlement in Level II 1 was heavily influenced by the Mycenaean culture. The examples with figurative designs, typical for the early and middle phases of the LH III C in the Aegean, are of interest. The Mycenaean pottery finds from the buildings as well as from the debris in this settlement are representative of a period from the LH III B to the middle phase of the LH III C. Considering the distribution of the Mycenaean pictorial pottery tradition in the Aegean and wider Mediterranean region, such magnificent depictions on the kraters and deep bowls from Tepecik are very meaningful in regard to the intercultural relations in the Aegean world during the Late Bronze. The pictorial-style Mycenaean kraters among the LH III C period Mycenaean pottery show the influence of the Aegean world. Besides pottery, two seal impressions dating to the period of the Hittite Empire have also been found in ‹Building 2›. Seal impressions with hieroglyphic-Luwian inscriptions are important for the historical geography of Western Anatolia and also for documenting the presence and impact of the Hittites in this region (Herbordt in: Günel-Herbordt 2010: 5–6, figs. 4–6; Herbordt in: Günel-Herbordt 2014: 8–11, fig. 8–10). The seal impressions from Tepecik present new information contributing to discussions on the historical geography of Western Anatolia as understood from Hittite texts. Furthermore, the seal impressions are

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

354

Günel

Fig. 7. Spool-shaped loomweights. Fig. 8. Pyramid-shaped loomweight.

unique to date in Western Anatolia, not least because they originate from an archaeologically secure context in a building, socialised with pottery and metal finds. Otherwise, seal impressions with hieroglyphic-Luwian inscriptions have so far only occurred in Western Anatolia as isolated finds or in secondary contexts. The seal impressions from Çine-Tepecik attest that the local administration of the settlement stood in direct connection with the Hittite Kingdom (Herbordt in Günel-Herbordt 2014: 10–11). In this building context, the pithoi, Mycenaean-decorated vessels with figural scenes and seal impressions with hieroglyphic inscriptions dating to the Hittite imperial period show that ‹Building 2› was involved in interregional trade and exchange, and functioned within the context of the administrative buildings of the settlement. In the settlement plan of Level II 1, the stone-paved area and the large storage pithoi point to this definition in relation to economic organisation. In the storage system of the settlement, a platform of small stones is located beside these buildings (fig. 2; ‹Platform›; fig. 6). In this area, apart from pithos fragments, two pithoi were discovered in situ. Each pithos was closed with an even stone. It becomes clear that this is a warehousing system in which big pithoi were positioned for storage and encircled by a stone platform. In the settlement, another important group of artefacts within the finds consists of a great number of various loomweights and spindle whorls, unearthed together. Conical spindle whorls, spools and pyramid-shaped loomweights found together at Tepecik, as well as the positions in which they were found, are clues to the nature of textile production (Günel 2012a: 59, fig. 12; Günel 2012b: 24, fig. 7). A large number of these weights belong to a workshop or workshops related to the textile industry in this level (fig. 7–8). The spool-shaped examples in this group are of unbaked clay. It was not fired, and the clay has large inclusions. The distribution of spool loomweights is observed frequently in the Aegean and also the Eastern Mediterranean, particularly in the LH III C period (Rahmstorf 2003: 404; Rahmstorf 2005: 147–149; Rahmstorf 2008; Rahmstorf 2011: 320–322). These

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A New Centre of Intercultural Relations … Çine-Tepecik

355

clay loomweights are seen at Lefkandi (Rahmstorf 2011: 321, fig. 1.j–k), at Karphi on Crete (Kanta and Kontopodi 2011: 130, fig. 16), at Maa-Palaeokastro on Cyprus (Karageorghis 2011: 23, fig. 3), and also at Tell Afis in north Syria (Cecchini 2011: 196, fig. 1–2, 4), and Qubur el-Walaydah in southern Palestine (Lehmann 2011: 289, 294, fig. 15) in the Late Bronze Age and the beginning of the Iron Age. The evidence of spool loomweights in this large geographical distribution shows the migration of Aegean groups to Cyprus and the Levant in the Eastern Mediterranean (Rahmstorf 2011: 321, fig. 7). In the light of all of the data in Level II 1, the Late Bronze Age settlement may be evaluated in terms of a chronology extending from about 1350/1300 to 1240/1100 BC. With reference to the distribution area of comparable pottery, the cultural relations of Tepecik may be followed into the Late Bronze Age. Among the comparable pottery, the shapes, as well as the decorative motifs, correspond to the Mycenaean painted pottery tradition, which had a wide distribution not only in mainland Greece and the Aegean Islands, but also in the Western Anatolian coastal region. The later phases of Mycenaean culture are noteworthy for their distribution across a wide area, including the eastern Aegean islands and the coastal region of Western Anatolia. The Mycenaean pottery finds from Tepecik necessitate a reevaluation of the borders of the Mycenaean cultural sphere in this period so as to include Çine within its area. Among the settlements of the 2nd millennium BC in Western Anatolia, Tepecik takes its place as a town within the dispersion area of main centres with defensive systems. In accordance with the model of settlements with fortification systems in the Late Bronze Age in Western Anatolia, the settlement structure with fortification wall of Tepecik stretches out according to the topography of the mound and has a regional plan. The architectural remains of Tepecik which shed light on the period between the beginning of the Late Bronze Age and the end of the Middle Bronze Age reveal another dimension in the southerly direction from the mound. The settlement shows the impacts of a thick ash layer. In particular, the southern part of the settlement is covered by a thick layer of volcanic ash over a large area. The analysis of the volcanic ashes was executed by Dr Johannes Sterba at the Atominstitut of Vienna. He determined that the volcanic ash is connected with the eruption of Thera. It appears that volcanic ash from the Thera eruption was carried eastwards by winds and reached the Level II 2 settlement in Tepecik. These traces show that the volcanic eruption in the vicinity heavily affected the site of Tepecik. In this Level, II 2, single-handled conical cups among the pottery, reminiscent of Minoan vessel forms, give a parallel chronology. The Thera volcanic eruption is associated with the destruction of Miletos IVa in the Meander valley, and the sites of Iasos on the southwestern coast and probably Çeşme-Bağlararası in the Izmir region were also affected by this natural event (von Graeve, Niemeier 2002: 77–78; Momigliano 2009: 129; Momigliano 2012: 32; Şahoğlu 2007: 318–319).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

356

Günel

Archaeological excavation from Çine-Tepecik carried out in the Meander region in recent years sheds new light on cultural influences and external contacts impacting the region. The cultural history of the 2nd millennium BC is illuminated by cultural remains at Çine-Tepecik and is given a certain chronological definition. At the same time, this data from the Tepecik settlement indicates the local cultural structure, as well as the intensity of cultural unity in its long chronological development. In conclusion, Çine-Tepecik in the southern part of Western Anatolia constitutes a new centre positioned between the Aegean and central Anatolian regions.

Acknowledgments The Çine-Tepecik excavations are supported by the Ministry of Tourism and Culture, Hacettepe University Scientific Research Fund Project and the Turkish Historical Society, to whom I would like to express my gratitude. I would also like to give my thanks to Prof. Dr S. Herbordt for her contribution regarding the analysis of these seal impressions. I would also like to give thanks to Dr J. Sterba for the analysis of the volcanic ashes and the results. I would like to give my warmest thanks the committee of the 9th International Congress on the Archaeology of the Ancient Near East (9ICAANE) symposium and the Basel University for the kind invitation.

Abbreviations FM FS LH

Furumark Motif Furumark Shape Late Helladic

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A New Centre of Intercultural Relations … Çine-Tepecik

357

Bibliography Blegen, C. W./Caskey, J. L. / Rawson, M., 1953. Troy III. The Sixth Settlement. Princeton. Blegen, C. W./Boulter, C./Caskey, J. L./Rawson, M., 1958. Troy IV. Settlements VIIa, VIIb and VIII. Princeton. Blegen, C. W., /Rawson, M. /Taylour, L. W./Donovan, W. P., 1973. The Palace of Nestor at Pylos in Western Messenia III. Princeton. Cecchini, S. M., 2011. Loomweights and the textile industry in north Syria in the early Iron Age. Karageorghis/Kouka 2011: 195–202. Crouwel, J. H., 1991. Mycenaean Pictorial Pottery: Well Built Mycenae 21, Oxford. –

2006. Late Mycenaean Pictorial Pottery: a Brief Review. E. Rystedt/B. Wells (eds.), Pictorial Pursuits. Figurative painting on Mycenaean and Geometric Pottery, Stockholm, 15–22.

–

2007. Pictorial Pottery of LH III C Middle and its Antecedents. S. Deger et al. (eds.), LH III C Chronology and Synchronisms II LH III C Middle, Vienna, 73–88.

Dakoronia, F., 2006. Mycenaean pictorial style at Kynos, East Lokris. E. Rystedt / B. Wells (eds.), Pictorial Pursuits. Figurative Painting on Mycenaean and Geometric Pottery, Stockholm, 23–29. Günel, S., 1999a. Panaztepe II. M.Ö. 2. Bine Tarihlendirilen Panaztepe Seramiğinin Batı Anadolu ve Ege Arkeolojisindeki Yeri ve Önemi/Die Keramik von Panaztepe und ihre Bedeutung für Westkleinasien und die Ägäis im 2. Jahrtausend, Ankara. –

1999b. Vorbericht über die mittel- und spätbronzezeitliche Keramik vom Liman Tepe. Istanbuler Mitteilungen 49, 41–82.

–

2008. Çine-Tepecik Kazıları ve Bölge Arkeolojisine Katkıları. A. Erkanal-Öktü et al. (eds.), Batı Anadolu ve Doğu Akdeniz Geç Tunç Çağı Kültürleri Üzerine Yeni Araştırmalar, Ankara, 129–139.

–

2010. Mycenaean cultural impact on the Çine (Marsyas) plain southwest Anatolia: the evidence from Çine-Tepecik, Anatolian Studies 60, 25–49.

–

2011. Çine-Tepecik Kazıları Işığında Bölgenin Tarihöncesi Kültürleri Üzerine Bir İnceleme. A. Öztan/Ş. Dönmez (eds.), Karadeniz’den Fırat’a Bilgi Üretimi. Önder Bilgi’ye Armağan Yazılar, İstanbul, 217–232.

–

2012a. Menderes Bölgesinde M.Ö. 2. Binyıl Kenti: Çine-Tepecik, Colloquium Anatolicum XI, 55–80.

–

2012b. Çine-Tepecik 2010 Yılı Kazıları, 33. Kazı Sonuçları Toplantısı, 2, 23-28 Mayıs 2011

–

2014a. M.Ö. Binde Çine-Tepecik Yerleşmesinin Batı Anadolu Arkeolojisindeki Yeri ve Katkıları.

Malatya, Ankara, 19–37. N. Çınardalı-Karaaslan et al. (eds.), Armağan Erkanal’a Armağan, Anadolu Kültürlerine Bir Bakış, 243–256.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

358 –

Günel 2014b. New Contributions regarding Prehistoric Cultures in Meander Region: Çine-Tepecik. B. Horejs/M. Mehofer (eds.), Western Anatolia before Troy. Proto-Urbanisation in the 4th Millennium BC? Vienna.

–

2015. Çine-Tepecik: New Contributions on Late Bronze Age Cultures in Western Anatolia. N. Chr. Stampolidis et al. (eds.), NOSTOI. Indigenous Culture, Migration, Integration in the Aegean Islands, Western Anatolia during the Late Bronze Age-Early Iron Age. Koç University, İstanbul, 627–646.

Günel, S./Herbordt, S., 2010. Ein hethitischer Siegelabdruck aus Çine-Tepecik, Archäologischer Anzeiger 1, 1–11. –

2014. Mykenische Kraterfragmente mit figürlichen Darstellungen und ein Siegelabdruck eines hethitischen Prinzen aus der spätbronzezeitlichen Siedlung von Çine-Tepecik. Archäologischer Anzeiger, 2014/1, 1–14.

Güntner, W., 2000. Figürlich bemalte mykenische Keramik aus Tiryns. Mainz. –

2006. Same motif, another context. E. Rystedt / B. Wells (eds.), Pictorial Pursuits. Figurative Painting on Mycenaean and Geometric Pottery. Stockholm, 177–182.

Holmberg, E. J., 1983. A Mycenaean Chamber Tomb Near Berbati in Argolis, Humaniora 21, Göteborg. Jacob-Felsch, M., 1987. Die Entwicklung der Keramik der Phase SH III C fortgeschritten und spät anhand der Schichtenfolge von Kalapodi und ihre Relation zu vergleichbaren Fundkomplexen. E. Thomas (Hg.), Forschungen zur Aegaeischen Vorgeschichte: Das Ende der mykenischen Welt, Akten des internationalen Kolloquiums, 7–8. Juli 1984, Köln, 37–52. Kanta, A./Kontopodi, D. Z., 2011. Kastrokephala (Crete): strangers or locals in a fortified acropolis of the 12th century BC. Karageorghis/Kouka 2011: 129–148. Karageorghis, V., 2011. What happened in Cyprus c. 1200 BC: hybridization, creolization or immigration? An introduction. Karageorghis/Kouka 2011: 19–28. Karageorghis, V./Kouka, O. (eds.), 2011. On Cooking Pots, Drinking Cups, Loomweights and Ethnicity in Bronze Age Cyprus and Neighbouring Regions, An International Archaeological Symposium in Nicosia, 6–7 November 2010. Nicosia Lehmann, G., 2011. Cooking pots and loomweights in a ‘Philistine’ village: preliminary report on the excavations at Qubur el- Walaydah, Israel. Karageorghis/Kouka 2011: 287–314. Momigliano N., 2009. Minoans at Iasos. C. F. Macdonald et al. (eds.), The Minoans in the Central, Eastern and Northern Aegean – New Evidence. Athens, 121–140. –

2012. Bronze Age Carian Iasos. Structures and Finds from the Area of the Roman Agora (c. 3000– 1500 BC). Roma.

Mountjoy, P. A., 1999. Regional Mycenean Decorated Pottery. Rahden 1999. Popham, M./Schofield, E./Sherratt, S., 2006. The Pottery. D. Evely (ed.), Lefkandi IV. The Bronze Age. The Late Helladic III C Settlement at Xeropolis. The British School at Athens, 137–231.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

A New Centre of Intercultural Relations … Çine-Tepecik

359

Rahmstorf, L., 2003. Clay Spools from Tiryns and other contemporary sites. An indication of foreign influence in LH IIIC? N. Kyparissi-Apostolika/M. Papakonstantinou (eds.), 2nd International Interdisciplinary Colloquium. The Periphery of the Mycenaean World, 26–30 September 1999, Lamia, 397–415. –

2005. Ethnicity and changes in weaving technology in Cyprus and the eastern Mediterranean in the 12th century BC. V. Karageorghis et al. (eds.), Cyprus: Religion and Society from the Late Bronze Age to the End of the Archaic Period. Proceedings of an International Symposium on Cypriote Archaeology, Erlangen, 23–24 July 2004, Möhnesee, 143–169.

–

2008. Tiryns XVI: Kleinfunde aus Tiryns. Terrakotta, Stein, Bein und Glas/Fayence, vornehmlich der Spätbronzezeit. Wiesbaden.

–

2011. Handmade pots and crumbling loomweights: ‘Barbarian’ elements in the eastern Mediterranean in the last quarter of the 2nd millennium BC. Karageorghis/Kouka 2011: 315–330.

Souyoudzoglou-Haywood, C., 1999. The Ionian Islands in the Bronze Age and Early Iron Age 3000–800 BC. Liverpool. Şahoğlu, V., 2007. Çeşme-Bağlararası: A New Excavation in Western Anatolia. F. Felten et al. (eds.), Middle Helladic Pottery and Synchronisms. Proceedings of the International Workshop held at Salzburg, Wien, 309–322. Thomatos, M., 2006. The Final Revival of the Aegean Bronze Age. A Case Study of the Argolid, Corinthia, Attica, Euboea, the Cyclades and the Dodecanese During LH IIIC Middle. BAR InterSeries. Oxford. Vermeule, E./Karageorghis, V., 1982. Mycenaean Pictorial Vase Painting. London. von Graeve, V./Niemeier, W. D., 2002. 1998–2000 Yılı Milet Çalışmaları, 23. Kazı Sonuçları Toplantısı, 2, 28 Mayıs-1 Haziran 2001. Ankara, 75–88. Wachsmann, S., 1998. Seagoing Ships and Seamanship in the Bronze Age Levant. College Station. –

2000. To the Sea of the Philistines. E. D. Ơren (ed.), The Sea Peoples and Their World: A Reassessment. Philadelphia, 103–143.

Prof. Dr. Sevinç Günel, Hacettepe University, Faculty of Letters, Department of Archaeology, Beytepe/ Ankara/TURKEY, [email protected]

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 361–376

Sara Pizzimenti – Federico Zaina

The Iron Age at Karkemish between Tradition and Innovation. The Case Study of the Pottery Assemblage from Area C This paper analyzes the pottery assemblage from the palatial contexts excavated in area C at Karkemish between 2011 and 2014. The aim is to understand the degree to which either slow or abrupt socio-political changes that occurred between Iron Age I and Iron Age III at the site could be reflected in the material culture. A substantial continuity in the ceramic shapes from the Neo Hittite to the Neo Assyrian occupations has been observed, while changes can be mostly detected through the quantitative occurrence of certain types.

1. Introduction1 The aim of the Turco-Italian Archaeological Expedition at Karkemish, directed by Nicolò Marchetti (Alma Mater Studiorum – University of Bologna) since 2011, is to extensively expose the Iron Age town in order to understand the urbanism of the site, to better define its history and the development in material culture (Marchetti 2014).2 In this respect, the renewed investigations are of importance to further investigate the role of the city during the Iron Age I–II (hereafter IA) as the capital of a Neo Hittite kingdom, and in the IA III as part of the Neo Assyrian Empire.

1 2

S. Pizzimenti wrote sections 2.1 and 3.1 and F. Zaina wrote sections 2.2, 2.3, 3.2, 3.3. Sections 1 and 4 have been written jointly. The Turco-Italian Expedition at Karkemish of the Universities of Bologna, Gaziantep, Istanbul is directed by Nicolò Marchetti (Alma Mater Studiorum – University of Bologna) to whom we would express our gratitude for his guidance and support. Thanks are also due to all our Turkish and Italian colleagues and friends for their help and support.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

362

Pizzimenti – Zaina

Fig. 1. Topographic map of Karkemish.

The excavations in the Lower Palace Area and in the Outer Town at Karkemish, together with the numerous soundings opened in the Yunus cemetery, provided a significant ceramic corpus mostly dating to the IA II and III (Bonomo and Zaina 2014). Different types of contexts have been explored so far, such as domestic, religious, palatial, funerary and military. In particular, the archaeological sequence of area C has provided some clues as to the development of the pottery horizon in a palatial context from the IA I to the IA III.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The Iron Age at Karkemish between Tradition and Innovation Date

363

Phase

Type of context

British Museum excavation (20 cent. AD)

0

Excavation dump

Middle Islamic to Ottoman (11th-19th cent. AD)

NOT ATTESTED

th

Early Islamic (8th-10th cent. AD)

Byzantine (4th-7th cent. AD) Roman (1st- 3rd cent. AD) Hellenistic (4th-1st cent. BC) Persian (6th-5th cent. BC) Iron Age III - (7th cent. BC) Iron Age II - (9th-8th cent. BC) Late Iron Age I - (10 th cent. BC)

1a-b

Domestic buildings

2a-b

Domestic buildings

3a-b

Open area

4a-b

Open area

NOT ATTESTED 5

Monumental Forum

6

Domestic building

7

Domestic building

8a-c

Production area

9a-c

Palatial compound (Sargon II)

10a-b

Palatial compound (Katuwa and Yariris)

11

Proto-Palatial compound

Table 1 Stratigraphical sequence of area C.

The aim of this paper is to understand the degree to which slow or abrupt socio-political changes occurred during the IA I–III, at the site could be reflected by the material culture. To this end, the rich pottery assemblage uncovered during the 2011–2014 excavations in area C will be considered here.3

2. The Stratigraphical Sequence of Area C Four excavation campaigns in area C (2011–2014) revealed a complex stratigraphical sequence (Table 1) consisting of 11 structural phases ranging from the IA I (10th century BC) to the early Islamic period (9th–10th century AD).4 Despite much disturbance, mostly 3

4

This paper focuses on the assemblages retrieved from area C East and area C South. Further ongoing studies on the assemblages from area S and area G are part of a wider research on the Iron Age pottery horizon from Karkemish. Most of the area was also used as a dump during the British excavation campaigns (phase 0). Several retaining walls made with large stone blocks by the British archaeologists have been newly excavated and documented.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

364

Pizzimenti – Zaina

caused by the numerous Islamic-period pits and channels, three IA occupations dating from the late IA I (10th century BC) to the IA III (7th century BC) have been distinguished. The architectural evidence revealed a sequence of large-scale buildings, two of which could be interpreted as palaces based on the architectural evidence as well as on the analysis of the finds, including textual data.5 Due to the obliteration, re-use or even destruction occurring in subsequent phases, each plan can only be partially reconstructed.

2.1 Phase 11 (late Iron Age I) The earliest occupation discovered to date in area C has been mainly exposed immediately to the east of the so-called King’s Gate (fig. 2a). Here, the destruction of the later Iron Age structures caused by the Islamic occupation and the 20th century AD activities allowed to study the archaeological sequence from the early 1st millennium BC. Moreover, a few, rather isolated, find-spots revealed a well preserved stratigraphy highlighting the clear architectural difference between the early Neo Hittite and later Iron Age structures. The reconstructed plan consists of a building with at least seven rooms of various dimensions. The building is badly preserved due to the destruction and leveling works of phase 10 and other later phases. Furthermore, the paucity of finds from primary contexts do not permit a full understanding of the activities taking place there – with the exception of a toilet, whose bench and drains in baked bricks are still preserved.

2.2 Phases 10a–b (Iron Age II) The first massive architectural and urban re-organization of the area probably took place at the end of the 10th century BC, under the reign of the ruler Katuwa6 (fig. 2a). Both the external perimeter of the building, with its monumental orthostat reliefs and the textual evidence confirming the name of the builder, had been already retrieved during the British excavations, directed by C. L. Woolley (Woolley/Barnett 1952). Nevertheless, it was only during the 2013 campaign that the first internal sectors were uncovered. Of particular relevance is the south-western sector, with its elaborate architecture consisting of some rooms with alternated limestone and basalt sculpted orthostats, as well as limestone slab floors. A fragment of an inscription with the name of Katuwa also comes from this context.

5

6

Two rectangular stamped baked bricks bearing the name of Sargon II were found during the excavations in area C. One was recovered in the central courtyard sealed by the bricks’ collapse (phase 9c), while the other comes from a later context belonging to the Persian period (phase 8). Two inscriptions of Katuwa were found on the floors of phase 10b. Gianni Marchesi and Hasan Peker are respectively publishing these inscriptions. A radiocarbon sample from a wooden pole in the foundation masonry of wall W.203, built in phase 10a, provided a date between 940–810 calBC (92.1 %).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The Iron Age at Karkemish between Tradition and Innovation

365

Fig. 2. Plan of the Protopalatial compound (Phase 11), the Palace of Katuwa (Phases 10) and the Palace of Sargon II (Phases 9).

Later structural modifications associated with the regent Yariris (late 9th century BC), such as the addition of the Royal Buttress (Gilibert 2011: 47–49; Orthmann 1971: 35), did not present any further architectural evidence and are preliminary included in sub-phase 10b of the Palace of Katuwa (10a).

2.3 Phases 9a–c (Iron Age III) At the end of the 8th century BC,7 the inner plan of the Neo Hittite palace (phases 10a–b) was deeply transformed. The presence in situ of a complete baked brick with the inscription of Sargon II suggests that the reconstruction took place between the conquest of Karkemish by the Assyrians (717 BC) and the death of Sargon II (705 BC). The new palace probably consisted of at least four blocks enclosed within the same perimeter of the previous one (phases 10a–b) and arranged around a large pebbled courtyard. The previous structures were only partially reused, as indicated by the different architectural layout and the general floors rebuilding above the previous ones. Among the most relevant renovations it is worth noting that the southern court was paved with a black and 7

The break between phases 10 (Neo Hittite) and 9 (Neo Assyrian) may correspond to the conquest of the city by the Assyrian army in 717 BC. The lack of archaeological evidence testifying this violent event may be explained with the massive removal of the previous structure to set up the new building.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

366

Pizzimenti – Zaina

white mosaic forming a pebbles checkerboard pattern. This architectural feature has contemporary parallels from neighboring sites such as Arslan Tash (Thureau-Dangin et al. 1931), Tell Ahmar (Bunnens 1998) and Ziyaret Tepe (Matney and Rainville 2005) as well as more distant areas like the recently discovered Phoenician Santuario de Càstulo in Spain (Blásquez Martínez 2006).

3. The Pottery Assemblages The Iron Age sequence retrieved in area C is of particular importance as it circumscribes some ceramic horizons within an absolute chronological frame. Indeed, the assemblages can be correlated with the textual evidence and a body of datable categories of finds (e.g. Pizzimenti 2014). As a consequence, the last occupation of phase 11 can be dated no later than the end of the 10th century BC, phases 10a–b range from the late 10th century BC to the end of the 8th century BC, while phase 9a–c, corresponding to the Neo Assyrian domination, covers the end of the 8th through all the 7th century BC. In terms of relative dating, the periodization proposed by Wilkinson (1995: 142) and updated by Mazzoni (2000: 56, table 1), suggesting three main phases, has been applied also to the Middle Euphrates (Jamieson 2012: 16). This has been also preliminary applied to the Iron Age sequence from area C, although a more in-depth study, considering the entire corpus, is necessary to confirm such synchronization. The Iron Age sample from area C consists of approximately 6,500 collected sherds, among which are c. 1,500 selected diagnostic fragments (rims, walls or bases) and complete shapes from contexts, including baked bricks collapse layers, clay or ashes covering the floors of the buildings. The methodology is primarily based on a functional division of three functional classes: Simple Ware, Kitchen Ware and Preservation Ware.8 Simple Ware open shapes are the most common types so far attested (c. 90% of SW) while both Kitchen and Preservation Wares are poorly represented. Preliminary analysis revealed that either morphological or technological variations in the assemblages are somewhat nuanced within the three phases. Indeed, those are mainly represented by the quantitative occurrence of certain shapes or surface treatments rather than their abrupt emergence or disappearance.

8

For a definition of the functional type and the morphological determination, see Zaina 2013: 66, 72, figs. 3–4.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The Iron Age at Karkemish between Tradition and Innovation

Fig. 3. The pottery assemblage from phase 11, late Iron Age I.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

367

368

Pizzimenti – Zaina

3.1 Phase 11 The pottery assemblage from Phase 11 is still quite limited, due to the demolition and leveling of the building during the construction of the Palace of Katuwa and the King’s Gate (phase 10). The pottery assemblage clearly reflects the latest use of the palace during the IA I, as confirmed by parallels with both Euphrates valley sites such as Tille Höyük (Blaylock 1999), and the Northern Levant such as Tell Taynat (Harrison 2010), Tell Afis (Venturi 2007) and Tell Mastuma (Iwasaki et al. 2009). Among the open vessels, bowls with an outer inflated rim (fig. 3.4–5) and bowls with a pointed, slightly in-turned rim (fig. 3.6) are attested. Closed shapes present a strong continuity during the entire IA period. Jugs and juglets with in-turned thick rim (fig. 3.7) first appear in this phase, and increase in frequency in the following IA II–III periods. Other diagnostic shapes include jug and juglets with thick outturned rims (fig. 3.8, 11–12), out-turned squared rims (fig. 3.9) and out-turned and exteriorly folded rims (fig. 3.10, 13). Preservation ware includes craters with expanded rectangular rims (fig. 3.18), and jars with out-turned thick rims (fig. 3.15–16), the former type probably belonging to a LB II tradition. Two types of Kitchen Ware are associated with phase 11: hole-mouth pots with pointed outer inflated rims (fig. 3.20), and molded rims (fig. 3.17, 19). Both shapes are chronologically restricted to this period, which is also confirmed by the Iron Age sequence of area G. Surface treatments include mostly White Slip, while burnishing is limited. The assemblage from phase 11 mostly consists of Simple ware shapes, 90% of which is made of fine homogeneous fabrics with a low frequency (> Fig. 7-8. 1. Haft Tape in Susa 2. Jame Mosque of Isfahan 3. Jame Mosque of Isfahan 4. Jamee Mosque of Isfahan 5. Ivan-e-Karkheh Palace in Susa, Art Bazar in Isfahan 6. Karim Khan Castle in Shiraz 7. Jonban Minaret in Isfahan 8. Shourjestan Caravan-serai in Shourjestan 9. Jame Mosque of Isfahan 10. Bam Castle 11. Gangali Khan Bazar in Kerman, Bazar of Tabriz 12. Borujerdiha Home in Kashan.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

An Introduction to the Traditional Arch and Vault …

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

437

438

Hassas

Bibliography Golabchi, M., 2013. Iranian Architecture technology. Tehran University. Pirnia, M. K., 1995. Archs and Vaults. Asar No.24. Tehran. Memarian, GH. H., 2011. Iranian Architect. Naghme no andish. Fakhar Tehrani, F., 2008. Traditional Buildings. Tehran University.

Najmeh Hassas, MA. Conservation Architect, Islamic Azad University, Yazd, Iran. [email protected].

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 439–448

Vito Messina – Jafar Mehr Kian

The Religious Complex at Shami. Preliminary Report on the Research of the Iranian-Italian Joint Expedition in Khuzestan at Kal-e Chendar The Iranian-Italian Joint Expedition in Khuzestan started fieldwork at Kal-e Chendar, valley of Shami (Iranian Khuzestan). The site, investigated in 1936 by Aurel Stein and never visited again, was one of the most important sanctuaries of Hellenistic and Parthian Elymais. This is the report on the first season of excavation.

1. Introduction In 2013, the Iranian-Italian Joint Expedition in Khuzistan started its 1st season of excavation (6th campaign) at Kal-e Chendar (fig. 1), in the Valley of Shami. This project falls in the frame of a Memorandum of Understanding signed between the Research Centre of the Iranian Cultural Heritage and Tourism Organization (RICHT-ICAR) and the Centro Ricerche Archeologiche e Scavi di Torino per il Medio Oriente e l’Asia (CST).1 The Valley of Shami extends about 30 km north of the modern city of Izeh (Iranian Khuzestan). The site of Kal-e Chendar, located in its northern part, was investigated in 1936 by Marc Aurel Stein, the famous Hungarian-British explorer, who during one of his 1

Other institutions involved in the project, as partners of the CST, are the Dipartimento di Architettura e Design, Polytechnic of Torino (DAD), and the Dipartimento di Studi Storici, University of Torino (DSS). In Izeh, the Expedition has been supported by the Ayapir Cultural Heritage NGO. Members of the Expedition were Mr. Jafar Mehrkian, Mr. Alireza Bagherian and Mrs. Mana Rohani, for the Iranian side, Mr. Vito Messina, Mrs. Alessandra Cellerino, Mr. Enrico Foietta and Mrs. Ilaria Bucci, for the Italian side. The Iranian-Italian Joint Expedition in Khuzestan does thank Mr. Jalil Golshan, Siamak Sarlak, and Mrs. Susan Cheragchi for their constant and precious support; in Izeh, Mr. Darvishpour, Qasemi, Mohammad Hassan Jafari, Gholam Reza and Hamid Bouzarjoumeri, Shahpur Kiani, Gheisar Saidipur, Mrs. Dehbanipour, Mr. Gaestouni and all the kind people of Izeh; in Kal-e Chendar, Mr. Ibrahim Baqheri.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

440

Messina – Mehr Kian

Fig. 1. Map of Iran and location of Kal-e Chendar.

last journeys through Inner Asia decided to remain a few days in the village after he was informed of the accidental recovering of ancient marble statues, broken into fragments, and a well preserved and known bronze statue, portraying a nobleman in Parthian dress, now in the Iran Bastan Museum, Tehran (Cumont 1939; Kawami 1987: 59–64, 169–174; Mathiesen 1992: 165–168; Curtis 1993). Stein opened a small trench and brought to light the remains of a rectangular enclosure of unclear purpose, built on stone foundations, a rectangular baked brick altar, and several stone bases, which appeared to be related to ancient statues now lost (Stein 1940: 141–159). Both the structures and findings seem what remain of an ancient sanctuary, and the high quality of the sculpture’s fragments recovered indicate that it must be one of the most reputed religious places of ancient Elymais, at least in the Hellenistic and Parthian periods (Sherwin-White 1984). Despite the importance of the site and findings, no further research was systematically carried out after Stein’s fieldwork terminated. Indeed a village was built in the area and both surveys and excavations appeared hindered by the presence of modern houses.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The Religious Complex at Shami

441

Fig. 2. The site of Kal-e Chendar after the survey of the Iranian-Italian Joint Expedition in Khuzestan. Satellite image: WW2 scene, 2012 (© Digital Globe).

2. Survey at Kal-e Chendar Given the high interest of the site, the Iranian-Italian Joint Expedition in Khuzestan carried out preliminary surveys at Kal-e Chendar in 2011 and 2012 (Messina/Mehr Kian 2014), after having examined the documents of Aurel Stein preserved in the Bodleian Library, Oxford, and National Library, London, and identified the area investigated in 1936, which was only roughly sketched and published (Stein 1940). The survey, conducted by GPS locator over an area of about 0.34 km 2, allowed to recognize that the archaeological site extends at least for 50 ha on the western slope of a narrow valley and has the shape of a triangle with its narrower vertex to the west, being apparently delimited on the north and south by the beds of two streams, now dried up, flowing westeastwards into a small river flowing north-southward, and delimiting the site on its eastern side: the so-called Rud-e Shami. The valley slope approximately ranges from 920 to 1040 m ASL. To the south, a hill rising up to ca. 1070 m ASL supports the remains of a very small fort (qala) of about 210 m 2.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

442

Messina – Mehr Kian

The surveyed slope revealed traces of ancient structures made in undressed stones that can be identified as wide terraces progressively allowing the ascent to the highest terrace: the latter extends for about 4,600 m 2, having an irregular quadrangular shape and overlooking the southern stream: its central part must correspond to the area investigated by Stein. The south-east corner of this terrace is now occupied by one of the eight modern houses built along the southern stream and partially overlapping the archaeological site. Cultivated fields extend over the rest of the slope and are delimited by low enclosures made in undressed irregular stones, which are sometimes superimposed on the remains of more ancient walls. At least three subsequent terraces have been identified. The most impressive of the ancient structures is the south wall of the upper terrace, which appears to be more than 90 m long and, at some points, up to 3 m high. Its façade has collapsed, probably in ancient times, but it is possible to see that its lower part is made by a row of huge irregular stones, onto which cut stones of smaller size are arranged in courses. In the area of its southeast corner, which has been used as foundation for a modern house, squared and rounded ancient column bases have been recognized, while ancient stone blocks of masonry have been re-used in the walls of modern houses. The latter findings are particularly important for they testify to the fact that a monumental building was risen on the top of the upper terrace, at about 1015 m ASL, and confirm the preliminary results of Stein’s research. However, the possibility that many other ancient elements, re-cut or broken on purpose, were used during the building of modern houses and field enclosures cannot be ruled out. These are indeed surely recognizable when they are in a quite good state of preservation, because of their size, which as a rule is larger than the other stones used in the courses of the walls, and polished surface, but become almost undistinguishable where they are broken into small pieces or re-cut. It is no surprise that the site, during the ages, was used as an emerging quarry of easy-access stone materials. Further to the N of the upper terrace, which we called the ‹Stein Terrace› (figs. 3–4), a smaller squared terrace of about 56 x 60 m, Terrace 2, can be clearly seen both from the satellite image of the site and on the ground. About 110 m north-east of Terrace 2, the corner of a third terrace, Terrace 3 (fig. 2), still emerges from the present ground level, as well as its well preserved eastern façade, which seems to be built following the same technique of the Stein Terrace; it is slightly differently oriented, for it follows the natural conformation of the ground. The preserved surface of Terrace 3 is about 9 m lower than that of the other terraces; it revealed the presence of a wide number of baked bricks fragments. The terrace’s size cannot be determined, for its southern and western fronts are unknown. Remains of other structures of unclear purpose have been even identified in the northernmost part of the valley slope and east of the modern road, as well as a number of tombs of different types. The latter, located on the whole area surrounding the terraces, are underground saddle-roofed chambers built in undressed stones.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The Religious Complex at Shami

443

Fig. 3. The Stein Terrace from the south (© CST and RICHT-ICAR).

In the upper part of the Kal-e Chendar slope, a very small valley was supplied by a spring now dried: this is known as ‹Chesmeye Sar-e Murd› (the spring flowing in the Murd plain); it can be reached by a well preserved stone-paved road running from the Stein Terrace to the spring itself via two large terraces. The road continues up to the mountain cliffs that rise west of Kal-e Chendar, known as Bileva peaks. A survey has been conducted also on this area: here, at an elevation of about 1709 m ASL, archaeological remains have been marked at 32°04’03.1’’ N, 049°40’50.2’’ E. In this area, huge walls made in undressed stones still emerge from the present ground level. These walls appear to be related to one great fort (qala) or, maybe, a complex of small fortresses that overlooks the Shami valley and offers a view to the Susan plain: it is remarkable that the area is known as ‹Char Qala› (the four fortresses). On the road that reaches the Char Qala area, a rock grave (Bard-e Gawri) has been recognized: this is a small chamber carved into a boulder having a polished façade and a small square opening, once blocked by a flat stone cut on purpose, now lost.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

444

Messina – Mehr Kian

Fig. 4. Trench 1 from the north (© CST and RICHT-ICAR).

3. Excavations at Kal-e Chendar Six trenches have been opened during the 1st season of excavation (6th campaign): four on the top of the Stein Terrace, one at the feet of Terrace 2, one west of the south stream bed. These have been named ‹Trench 1›, ‹Trench 2›, ‹Trench 3›, ‹Trench 4›, ‹Trench 5› and ‹Trench 6›. Trench 1 (fig. 4) is located in the centre of the Stein Terrace, in an area that could be identified, approximately, as the area already investigated by Stein. Trench 2 is located at the south-west corner of the terrace, in a point that seems to correspond to the ancient enclosure of the terrace itself. Trench 3 (fig. 5) is north-east of Terrace 2, not far from its NE corner. Trench 4 crosses the presumed NE wall of the Stein Terrace. Trench 5 is at the terrace’s presumed NW corner. Trench 6 is on the other bank of the south stream. Trench 1 measures 6 x 8 m east-westwards, having the shape of a T. Its south-west corner is located at 32°03’34.9’’ N, 049°41’54.9’’ E. On the surface, traces of mechanic cultivation were clearly visible: excavation revealed that cultivated soil reaches a depth of about 30 cm. Below it, a layer of deposits, containing very few potsherds and baked brick fragments (SU 3) covered the remains of a structure made of undressed stones (SU 4). The latter appears to be a wall or the foundation (?) of a wall, measuring at least 8 m in length and 1 m in width, and is built by stones roughly cut or re-cut on purpose, as is shown at least by two fragments

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The Religious Complex at Shami

445

Fig. 5. Trench 3 from the east (© CST and RICHT-ICAR).

having a regular rectangular shape. As far as it is possible to see at present, SU 4 is a retaining wall almost perfectly oriented to the north, which contains a filling of stones and rubble extending beyond the western limits of the excavated area. It appears to have been laid on a layer of rubble (SU 8), which has been reached at about 1.2 m below the present surface and includes huge stones and boulders. The latter hindered the continuation of the excavation: for this reason, SU 8 has been excavated only on a limited area of 1.5 x 2 m for about 80 cm in depth. No potsherds have been found in this layer. Excavation was suspended at about 2 m below the surface Trench 2 measures 6 x 5 m. Its south-west corner is located at 32°03’33.8’’ N, 049°41’52.6’’ E. In this point, the surface is about 2 m deeper than the surface of Trench 1. Just below the surface, a huge structure has been unearthed. It is made of undressed roughly cut stones that appear to have been disposed, at least, on 7 rows and seem to retain a filling of rubble. These rows could be interpreted as a huge retaining wall, approximately oriented to the NE, measuring at least 3.8 m in width and extending for more than 5 m in length. Given that this wall is very close to the supposedly identified corner of the Stein Terrace, it could be stressed that these rows are what remain of the terrace west wall and that two huge boulders still in their position at the west limit of the trench could be part of the terrace west façade.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

446

Messina – Mehr Kian

Fig. 6. Digital Elevation Model of the Kal-e Chendar slope elaborated by Aster-GDEM data (© CST and RICT-ICAR).

In any case, the upper rows of this structure appear to be partly loose and decayed. Excavation in depth, hindered by the presence of rubble and stones over all the exposed area, has been stopped at about 50 cm below the surface. It must be remarked that no potsherds have been found in Trench 2. Trench 3 is roughly rectangular, measuring 2.5 x 7 m and oriented to the east. It is located at 32°03’39.2’’ N, 049°41’56.4’’ E. Here a complex structure made of baked bricks has been unearthed immediately below the surface and partially below a modern retaining wall of undressed stones. This structure, which is oriented to the E and is preceded on its east side by a paved baked brick floor (measuring 2.5 x >1.8 m), is a rectangle measuring 2.5 x 1.8 m. It has been partially destroyed, probably in modern times, by the opening of an unauthorized trench of irregular shape. Indeed, the courses of bricks alternating in its masonry have been broken for reaching the layers below them. The latter have been cleaned and revealed the presence of cut stones incoherently disposed below part of the structure: it is still unclear whether the structure was built over this layer of loose stones or the latter have been throw in the unauthorized trench and pushed below the bricks in order to fill the trench itself. In the filling of the trench and the disturbed layer on the paved floor, several baked bricks fragments have been found and, among them, 5 complete baked bricks having the shape of a

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The Religious Complex at Shami

447

column segment: these are of two sizes, the bigger measuring 34 x 35 x 30 x 8 cm (35 being the radius), the smaller measuring 30 x 32 x 29 x 6 cm (32 being the radius). The function of the whole structure is still unclear and needs to be more accurately investigated, but it could be supposed that it must have been a kind of platform that supported another structure (an altar?). Bricks of two sizes have been recognized in their original position: the standard is a square brick measuring 36 x 36 x 8 cm, but even some rectangular samples measuring 36 x 16 x 8 cm have been found. Very few potsherds have been found in the surface layer covering the paved floor: these are in common ware and seem to be generally dated to the Parthian period. Trench 4 is a narrow rectangle, measuring 1 x 10 m. It has been opened across the presumed NE wall of the Stein Terrace and its NW corner is located at 32°03’35.5’’ N, 49°41’56.4’’ E. Only the surface layer has been removed in order to clarify the position of some huge stones that appeared to be part of ancient wall masonry. These have been cleaned and seem to be in their original position, being aligned at least on 3 with the purpose of retaining the terrace filling: the latter is composed of rubble and soil. No potsherds have been found. Trench 5 is a rectangle measuring 3 x 10 m, roughly oriented to the east. Its NE corner is located at 32°03’36.0’’ N, 49°41’54.6’’ E. It follows the same orientation of a modern retaining wall, which could have been built over the remains of a more ancient structure. The surface layer revealed the presence of few potsherds and, among others, many fragments of a big storage jar decorated with a ribbing. Given that no old structures have been discovered during the present fieldwork, excavation stopped at 50 cm below the surface. Trench 6 has been opened on the place of a tomb already identified during the survey conducted in 2012 (5th campaign) and named T9. It is located at 32°3’40.10”N, 49°41’57.90”E and is a chamber of 4.5 x 2.4 m, made in undressed stones and roughly oriented to the east, which has been partially excavated by an unauthorized trench opened in modern times. Part of the collapsed saddle (?) roof has been removed and half of the chamber has been plundered. The remaining part has been excavated by the present expedition, starting from the surface, by removing the remaining part of the roof, which was instable and endangered the whole structure. Below this debris filling, a floor made in cut and flat stones has been reached. On the chamber’s N wall, a rectangular low bench, made following the same technique of the floor, has been discovered. It has been broken by the trench that also destroyed the roof; its top is at about 20 cm over the floor. The chamber must have been at least 2 m high and, given the present sloping surface, it could be assumed that part of the chamber itself, likely its façade, now lost, emerged from the ancient ground level, and could be seen when arriving in the area: this area seems to have been part of one of the cemeteries surrounding the terraces complex. In any case, this type of chamber can be interpreted as a kind of family tomb. All the trenches have been covered after the end of the excavation.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

448

Messina – Mehr Kian

4. Conclusions Even if cleared to a limited extent, the structures unearthed by our expedition attest that stone and baked brick buildings stood on the top of the recognized terraces, and seem to show that the present surface, used for cultivation purposes, approximates the ancient ground level, being coincident in some points. Archaeological stratification was indeed hindered by the continuous re-use of the stone blocks employed in ancient masonries, and, as a consequence, even layers containing ancient objects and materials did not accumulate during the centuries. Despite the fact that our excavations provided indirect indication on the presence of buildings on all the terraces, it may be supposed that not all them were built for cult purposes: given that loose-on-surface ancient blocks of masonry and architectural decorations were found only nearby the Stein Terrace, it would not be surprising to discover, in the future, that only the latter supported a monumental building, like a temple, while less monumental structures made in baked bricks, like small altars, could have been built random, in many other points of the site. In any case, the building size and monumental setting that can be inferred from the data acquired on the Stein Terrace’s top and other points lead to assume that the sanctuary at Kal-e Chendar could have rivalled those at Majid-e Sulayman and Bard-e Nechandeh.

Bibliography Cumont, F., 1939. Les bronzes gréco-parthes de Shami. Syria 20.2, 167–168. Curtis, V., 1993. A Parthian Statuette from Susa and the Bronze Statue from Shami. Iran 31, 63–69. Kawami,T.S.,1987.MonumentalArtoftheParthianPeriodinIran.ActaIranica26.TextesetMémoiresXIII.Leiden. Mathiesen, H. E., 1992. Sculpture in the Parthian Empire. A Study in Chronology. Aarhus. Messina, V./Mehr Kian, J., 2014. Return to Shami. Preliminary Survey or the Iranian-Italian Joint Expedition in Khuzistan at Kal-e Chendar. Iran 52, 65–77. Sherwin-White, S. M., 1984. Shami, the Seleucids and Dynastic Cult: a Note. Iran 22, 160–161. Stein, M. A., 1940. Old Routes of Western Iran. Narrative of an Archaeological Journey. London. Vito Messina, Università di Torino, Dipartimento di Studi Storici. Via Sant’Ottavio, 20, 10124 Torino, Italy. Jafar Mehr Kian, RICHT-ICAR. Tehran, Iran.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Israel – Lebanon – Levant

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 451–462

Eva Katarina Glazer

Sedentary and Nomadic Population during Bronze Age Southern Levant: An Example of Cultural Contacts Bronze Age southern Levant is most commonly recognized as the first great urban period in the area. Processes of urbanization were accompanied by a variety of contacts between urban and rural population and the always present nomadic people of the region. This paper will present several examples and aspects of such cultural contacts.

Introduction The Bronze Age was a dynamic period in southern Levant history, most commonly recognized as the first great urban period in the area. It was marked by the emergence of urban centers followed by their collapse and subsequent reemergence all of which is well documented in historiography based on many decades of intensive archaeological research combined with intensive study of ancient archives. This picture of Bronze Age Levant history points out to a very versatile society. To simplify the different components of society we can observe the rural-urban-nomad networks. It is exactly those components of Bronze Age society that must be observed when trying to decipher the historical processes of the period. Interrelations between nomads, semi-nomads, and urban city dwellers along with farmers and sedentary village community residents have been considered in many terms and under different perspectives, resulting in more and more complicated terminology issues and analysis with no coherent synthesis. This paper will not try to provide such a synthesis but it will try to highlight the cultural contacts between nomadic and sedentary populations in the Bronze Age southern Levant by presenting several interesting aspects of the interactions between urban, rural and nomadic population.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

452

Glazer

Examples of Rural - Urban - Nomad Networks At the beginning of the 3rd millennium BC Levant was undergoing a process of urbanization. This is not to say that the entire population moved into towns, it may be assumed that the majority continued to live in villages scattered over the countryside. However, the crystallization of the city or town as a form of settlement is the single most important social phenomenon during Bronze Age in the Near East as a whole and appears to be one of the essential conditions for the development of civilization (Ben Tor 1992: 85). Various explanations for the origin of cities have been advanced and debated at length in the literature, ranging from the monumental work of L. Mumford to the most recent anthropological works on this inexhaustible subject.1 The definition of the city or urban center is a subject often studied and much debated. Its intricacies cannot be fully treated here, but it seems that some conditions, such as size, fortification, population density, public building, social stratification, a part of the population engaged in non-rural pursuits are essential for the definition of a settlement as urban, though not all of them are required in every case. The established Levant Bronze Age urban centers were large, surrounded by fortifications, divided into public areas such as sacred spaces and markets and communal areas of dwellings and private houses (Mazar 1992: 108–109). Most of those urban centers were founded at the beginning of the EB II period, increasing in size and strength of fortifications in EB III. The cities evolved out of an agricultural economy, that is, they evolved from rural settlements of the EB I period in a way that left some rural settlements abandoned while others grew in size which resulted in the formation of a two or three level hierarchical system in which the largest center became the urban center which took control over smaller rural settlements.2

The Case of Arad – Questioning the Norm «Rural to Urban» One of the leading examples of a large Bronze Age urban center in southern Levant is Arad. This city has yielded plentiful remains which testify to its rich history and continuity through many periods. Its location on the fringe of the Negev desert in the south and the fruitful valley in the north has provided a better insight in understanding the interaction

1

2

Lewis Mumford’s book The city in history was first published in 1961 in New York and very soon it was translated and published all over the world, cf. Mumford, L. Grad u historiji, Zagreb, 1968. One of the recent works worth mentioning is a book that remarkably combines anthropological archaeology with contemporary urban anthropology: The city in time and space by Aidan Southall published in 1998 by Cambridge University Press. In the three level system those rural settlements dominated over smaller villages and hamlets, cf. Frick 1997: 15.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Sedentary and Nomadic Population during Bronze Age Southern Levant

453

between nomadic populations of the desert and semi arid areas and the rural and urban population of the agricultural areas. From the beginning of the research Arad offered a new and better insight into the urban culture of the Bronze Age period, but also raised many questions and after half a century of investigations it seems that many problems have yet to be answered.3 The earliest settlement of Arad was founded at the beginning of the EB I period in the vicinity of the available water sources and construction materials in an excellent transit location. It has provided less architectural remains (no public buildings or any architectural remains have been found so far) and we can only assume that it was a typical rural settlement of the EB I period. This oldest settlement is important because it provided important finds which helped to understand and construct early Palestine chronology. Ceramics from the level IV period of occupation has been recognized as Egyptian which was a major breakthrough in establishing a chronology for the southern Levant area in general.4 As already mentioned, the urbanization process in Arad can’t be compared with simultaneous processes in the region because of the lack of other rural EB I settlements around it, so in Arad’s case there was definitively no obvious relocation of the rural population and formation of a multi-hierarchical system in which Arad would take the role of the main urban centre. That is why we must consider the idea of migrations of nomadic populations and the possibilities that the newcomers who built the urban city at the beginning of EB II might have come from the south and southeast. The nomadic population was surely in contact with the rural population of Arad during the EB I period and it is possible that those connections, rooted in the common interest of trade, represent the building blocks of Arad’s further development.

3

4

During the ’80 of the 20th century Ruth Amiran has concluded, based on numerous studies, which have yielded rich finds and witnessed extremely high degree of urbanization, that Arad developed into an urban center like similar examples from Palestine, from the rural settlement which was created in EB I into an urban center in EB II era. Despite the deviations such as the close proximity of the desert and the fact that there were no other small rural settlements from the period at hand, it was believed that Arad was a typical urban center just as many others from the EB II period. But Israel Finkelstein has presented an interesting thesis during ’90 of the 20th century, that Arad was actually a trading centre of the southern desert nomads and semi-nomads. The idea that a nomadic population, who by its very nature do not build permanent housing units, let alone cities, raised such a significant urban center itself has imposed more questions than answers, and made Finkelstein the target of numerous attacks by researchers who have claimed that Arad was more associated with rural agricultural population oriented to the northern areas; this debate has its opponents still arguing on. See further in: Amiran 1970: 96; Finkelstein 1995: 81. The beginning of the EB II period can be correlated with the time of Djer, the 3rd king of the 1st Dynasty of Egypt (ca. 3050 BC), cf. Ben Tor 1992: 122.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

454

Glazer

Considering Arad’s location on rather dry territory, its economy was quite diverse. Carbonized cereals, mostly barley and wheat, and edged sickles testify to an advanced agriculture. There is also evidence of many legumes: peas, chickpeas and lentils. Olives are evidence of an evolved horticulture. At the same time, many remains of sheep and goat bones indicate that there was production of dairy products, meat, skin and wool. Cattle bones indicate the likely use of a plow and donkey bones indicate their use as a main form of transportation (Mazar 1992: 118). Food production was the main occupation and grinding stones and mortars have been found in every household while spindles, looms and needles point to other domestic activities, such as spinning and sewing. Considering the similarities between the material culture of Arad and the contemporaneous settlements in Sinai Finkelstein has assumed that Arad was a merchant centre of the inhabitants of the desert (Finkelstein 1995: 69–86).5 Accordingly, Arad was the main supplier of copper for the northern regions during the EB I and EB II period. During the EB II period Arad was a large urban center6, ca. 10 hectares in size, surrounded with fortification walls over 1176 m long with many city gates and towers (Mazar 1992: 124). From this period there are many remains that testify to an urban center: significant fortifications, a palace, a temple, a water reservoir and a complex of public buildings. This urban center was destroyed around 2800 BC, well attested by large burnt deposits found all over the tell. Nevertheless, immediately after the destruction of the city from level III the rebuilt city of level II emerged. The continuity is obvious from the architectural style, construction methods, material culture and social structures. So urban Arad revived once again until 2650 BC when it was (unexplainably) abandoned. As this author’s recent publication has implied, some of the inhabitants of Arad left the city and moved to Bab ed Dhra where they played a role in the transformation of the dominant burial practices in their new city.7

The «Otherness» of Nomads – Questioning the Tradition The case of Arad is a vivid example of how nomads (non sedentary population) contributed to the development of a major urban center of the southern Levant, while confronting the established view of nomads as «the others». The traditional scholarly view of nomadicsedentary relations in the Near East was influenced by ancient historical sources, all written 5

6

7

There is evidence of many copper mines in Sinai and some authors confirm that people from the desert traded copper for the products of the north, such as olive oil and cereals, see Ilan/Sebbane 1989: 139–162. If we compare sizes of excavated sites the most prominent cities were Beth-Yerah with 22 hectares, Yarmuth with 16 hectares, Tell el-Hesi, ‘Ai and Arad with 10 hectares each, cf. Mazar 1992: 111–112. This new ritual might have been a result of the remainder of the former urban lifestyle in Arad, and it might have better suited to their religious needs, their cult, their life style and also as a distinction from the nomadic lifestyle of their ancestors, cf. Glazer 2014: 161–162.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Sedentary and Nomadic Population during Bronze Age Southern Levant

455

by urban-based scribes, who emphasized the «otherness» of pastoral nomadic groups (Schwartz 1995: 250). For example a late 3rd millennium text from Mesopotamia The Curse of Agade describes the pastoralist group Gutians as a group «not classed among people, not reckoned as part of the land/… people who know no inhibitions/with human instinct but canine intelligence and monkeys’ features» (Schwartz 1995: 255) in short, people completely beyond the pale of urban civilization and, perhaps more to the point, beyond the control of urban political authority. The nature of such contacts is similar in other regions, for example Andrew B. Smith mentions «Saharan nomads are notorious for their personal aversion to agriculture. They are, however, quite happy for other people to grow grains for them and to extract their tribute at harvest time». (Smith 2005: 263) This rather negative aspect of nomadic society is also present in recent works, such as the following description of the ruthless nature of nomads «Nomadic people have always been warriors, ‹lean and mean›; they not only fight continuously with other nomads over wells and pastures, but constantly hover, like birds of prey, on the edge of the fertile fields, ready to pounce on the stores of the farmers at the slightest sign of weakness». (Issar/Zohar 2007: 102) If we put the negative aspect aside we should focus on the considerable evidence that testifies to the symbiotic, mutually dependent relationship between the nomadic and sedentary society. If nomads were able to play a role in establishing the urban center such as the previously mentioned Arad, then what stopped them from turning to a rural sedentary life style in general?

Rural vs Nomadic Lifestyle Our knowledge of the economy and particularly of animal husbandry and agriculture is insufficient and random, both because organic remains (with the exception of bones) are poorly preserved in the Levant climate and because sufficient attention has not so far been given to the retrieval of relevant data. We have already mentioned that urban centers evolved out of an agricultural economy. The urban unit dictated the political organization of the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

456

Glazer

Levant for the following two thousand years.8 Nevertheless, agricultural productivity, a higher standard of living, and the resultant increase in population also led to the foundation of new smaller settlements, farmsteads and villages (Ben Tor 1992: 96). This period also witnessed the great shift in agriculture, the introduction of new crop plants that were first widely cultivated, the olive and the vine.9 «Permanent fields, residential stability, and general tranquility are the prerequisites for the commercial production of horticultural crops» (Stager 1985: 177). The most characteristic feature of the Early Bronze Age economy seems to have been the investment in large-scale horticultural production, implying a settled and stable mode of life as fruit trees and vines can only be reproduced by cloning and do not produce fruit until many years after planting (Grigson 1998: 259). Agriculture on any scale is bound to be a sedentary occupation (Grigson 1998: 247). Floral, faunal, artifactual and textual evidence all point to a settled agricultural life style in the Bronze Age Levant.10 But villages were not only agricultural settlements, some villages were seasonal settlements of pastoral nomads. This phenomenon of coexisting urban and nonurban population is well known in the Levant Bronze Age.11 Towns provided villagers and semi-nomads with various spiritual and material services, with a market for their produce, protection in times of danger, and perhaps a political framework and a sense of community. The villagers and semi-nomads, in return, brought their surplus produce to the urban settlements, and played an important part in inter-site communication and in the trade network.12

The city-state was a unit in the center of which stood the city (urban centre) surrounded by villages and farmsteads, which formed its agricultural hinterland and for which the town provided protection and services, cf. Ben Tor 1992: 86. 9 The immense contribution of these plants to the history of Mediterranean civilization has often been remarked: the olive and vine adapt well to poor soil, require relatively little attention, produce highly nutritious fruit and by-products (mainly oil and wine) that keep well for long periods. It is not surprising that oil and wine soon occupy a prominent position among the exports of Palestine, cf. Ben Tor 1992: 84. 10 Thus it is the agriculture of fellahin, not the semi-nomadic practices of the Bedouin, which is the relevant ethnographic parallel to studies of early agricultural systems in the southern Levant, cf. Grigson 1998: 259. 11 Strong bonds undoubtedly existed between these two components of the population as can be observed in the archaeological record or inferred from documents of the 2nd millennium BC, cf. Ben Tor 1992: 96. 12 Sometimes the very existence of the villages depended on the urban markets, as was the case with the miners’ settlements of southern Sinai, see Ben Tor 1992: 96.

8

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Sedentary and Nomadic Population during Bronze Age Southern Levant

457

To put aside this rather logical explanation of nomadic sedentary interaction we can observe similarities between the two groups: settled way of life for some nomads and for all rural people, as well as practicing animal husbandry for all nomads and for most of the rural population. There is a variety of ways in which herd animals can be managed, varying from the keeping of small numbers within sedentary village communities, through small scale movements around the landscape, to large scale, nomadic movements although there is no evidence that full-scale nomadic pastoralism was practised in the Levant during Bronze Age (Grigson 1998: 259). Each of these stages implies an increasing degree of mobility on the part of animal keepers and an increasing dislocation between animal keeping and agriculture, which is primarily a sedentary occupation.13 An interesting illustration of the socio-economic system of the Levant is found in the story of Sinuhe14: «I went with him to his tribe. What they did for me was good. He set me at the head of his children. He married me to his eldest daughter. He let me choose for myself of his country, of the choicest of that which was with him on his frontier with another country. It was a good land named Yaa. Figs were in it, and grapes. It had more wine than water. Plentiful was its honey, abundant its olives. Every kind of fruit was on its trees. Barley was there, and emmer. There was no limit to any kind of cattle. Moreover, great was that which accured to me as a result of the love of me. He made me ruler of a tribe of the choicest of his country. Bread was made for me as daily fare, wine as daily provision, cooked meat and roast fowl, beside the wild beasts of the desert, for they hunted for me and laid before me, beside the catch of my own hounds. Many … were made for me, and milk in every kind of cooking» (Wilson 1973: 7). Sinuhe lived with the tribes in the land of the Retenu, i.e. in Canaan. These people fit well in our understanding of the rural communities because they enjoyed the products of typical Mediterranean types of fruit orchards, like figs, grapes and olives. On the other hand, we can easily identify them as nomads with their tribal organization and their cattle herding. Or were they simple marauders? Which brings us back to the question of the intensity and nature of cultural contacts between nomads and sedentary populations. 13 The four main stages of pastoral systems are mixed farming, semi-sedentary pastoralism, semi-

nomadic pastoralism and nomadic pastoralism, See: Khazanov 1984, or compare with Grigson 1998: 264, fig. 1.1. 14 An Egyptian courtier who fled to Canaan when the Pharaoh Amen-em-het I was murdered in a rebellion at around 1960 BC.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

458

Glazer

Nomadic pastoralism, agricultural sedentism and a combination of the two must be seen as part of a spectrum of potential economic strategies. When faced with severe environmental or climatic difficulties farmers had the option of adopting a pastoral nomadic life style. And so it happens that most of the explanations of abandonment of the EB III urban system simply consider the population shift to pastoralism. «Archaeological evidence throughout the region shows that most of the urban centers, which were established during the exceptionally cold and humid climate of the Early Bronze Age, were deserted towards the end of the third millennium B.C.E. and the surviving inhabitants turned to animal husbandry and became increasingly mobile and nomadic. They supplemented their food supply with seasonal planting of cereals and vegetables near steady water sources. Along with this, agricultural technology and international trade – the main source of urban elite wealth – dwindled and pastoralism became the dominant branch of economy spreading along the belt bordering the desert.» (Issar/Zohar 2007: 139) But a shift from an economy based on sedentary agriculture to a system primarily based on animal husbandry with less emphasis on agriculture didn’t affect the whole region. Although most of the papers focus on the shift to pastoralism or nomadism in general there are interesting examples that don’t support this simplistic explanation.

The Case of Tell Abu en-Ni’aj – Questioning Generalizations The excavations carried out by the East Jordan Valley Survey in the Jordan Valley at the edge of ghor (agricultural terrace of the Jordan Rift) overlooking the zor (the active floodplain of the Jordan River; Falconer/Fall 2009: 99.; Falconer et al. 2007: 262) discovered the site Tell Abu en-Ni’aj and recognised it as a rare example of an EB IV farming village (Falconer et al. 2007: 263). Occupied ca. 2300–2000 BC, the site provides our most substantial body of evidence for a year-round farming community during the pervasive abandonment of Levantine towns and cities in the EB IV period; this interval is most commonly interpreted in terms of nonsedentary pastoralized society. Rectilinear mudbrick architecture, sherd-paved streets, large storage bins and a possible olive or grape press, coupled with floral and faunal assemblages dominated by domesticated farm crops and herded animals, clearly portray this site as a long-lived sedentary community (Falconer/Fall 2009: 99). As with most agrarian communities in the Near East, the majority of bones represent sheep (Ovis aries) and goat (Capra hircus). There is a relative abundance of cattle (Bos taurus) bones and most intriguingly, the frequency of pig (Sus scrofa) is high at the presumed

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Sedentary and Nomadic Population during Bronze Age Southern Levant

459

«pastoral» settlement of Ni’aj.15 Unexpectedly high frequencies of pig bones are significantly higher than those for other Near Eastern Bronze Age communities, which commonly produce frequencies around 10 percent (Horwitz 1989; Wapnish/Hesse 1988). Low pig frequencies at other sites reflect their relatively high water needs and poor suitability for mobile pastoralism (Horwitz 1989). High pig-bone frequencies provide strong circumstantial evidence that Tell Abu en-Ni’aj was a sedentary agrarian settlement. The general agricultural orientation of the village is underscored by the near absence of wild taxa, despite the availability of migratory birds along the Jordan Valley and fish in the nearby Jordan River (Fall et al. 1998: 116). Agricultural crops are grouped into four main categories: cultivated cereals, orchard crops, cultivated legumes, and wild species. Annual and perennial crops often require very different cultivation strategies. Cereals and legumes require the planting of mature seeds but may not entail intensive maintenance before producing a harvest of new seeds within several months. Thus the cultivation of annual crops permits discontinuous seasonal cultivation by mobile populations (La Bianca 1990; Zohary/ Hopf 1988). In contrast, the cultivation of fruit trees requires more continuous horticultural management and a shift from sexual (i.e., involving seeds) to vegetative reproduction (Liphschitz et al. 1991). Fruit trees represent a distinct long-term investment, since they do not bear fruit for three to eight years after planting and may not reach full productivity for ten to twenty years (Stager 1985, Hopkins 1985). This detailed profile of village agriculture in the Jordan Valley provides fundamental, and unexpected insights in Bronze Age rural economy and ecology in the midst of urban collapse. Animal management and orchard cultivation reflect the element of rural sedentarism that is seemingly distinct from the «pastoralized» EB IV period (Fall et al. 1998: 120). The community of Tell Abu en-Ni’aj provides a rare, intimate portrait of agrarian ecology in an increasingly anthropogenic landscape during the collapse of early towns and cities. It exemplifies the importance of sedentary communities and the persistence of orchard cultivation during the urban collapse and regional «pastoralization» of the EB IV period. Tell Abu en-Ni’aj exemplifies the spectrum of agrarian strategies that provided the basis for village economy and ecology in emerging urbanized societies such as those of the southern Levant.

15 From a total of 1056 bone samples 28 percent belonged to Sus scrofa specie (pig), cf. Fall et al.

1998: 116.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

460

Glazer

In Conclusion So if we consider our rural - urban - nomad network there are several important conclusions that can be drawn. Firstly, most of the historical sources were biased, written by urban based scribes generally considering nomads as «others» and giving primacy to urban populations, i.e. in favor of the states, on the other hand almost every anthropological research can seem biased as well giving nomadic life style a sort of mystical and romantic aura. Secondly, not every shift in economical practice is a result of a nomad penetration into an urban state, or a result of warfare, and as we have seen in the case of Tell Abu en-Ni’aj generalizations concerning urban abandonment and transitions to a nomad pastoral economy should be avoided; it should be kept in mind that rural villages could have existed if sufficient water resources were available. Thirdly, nomadic populations could have played an important role in the process of urbanization, such as in the case of Arad. More examples can be possibly drawn from recent excavations and papers dealing with Kranzhügel sites in Syria and their connection with pastoral nomads (Lyonnet 2009). All papers that deal with this interrelations and dynamics of society have in common that they acknowledge the ever present symbiosis and the continuous existence of nomads as well as sedentary populations in Levant. As Ibn Khaldûn put it in 14th century «While the Bedouins need the cities for their necessities of life, the urban population needs the Bedouins for conveniences and luxuries» (Ibn Khaldûn 2005: 122) If we consider the long history of urban and rural settlements in the Levant and the presence of nomads even today, while modernization has taken over all and globalization is forcing a uniform life style all over the world, we must conclude that cultural contacts between nomadic and sedentary population still exist, as well as the fact that our research hasn’t even scratched the surface of this interesting topic. Future research should try to determine the status of the nomadic population during specific phases of the Bronze Age, in particular in connection to the processes of urbanization during EB II and III phase and during the re-urbanization in the Middle Bronze Age. It should also examine the contact between the nomadic, semi-nomadic and rural population in Jordan Valley sites, because of the specific character of those sites during the Bronze Age. Then maybe we will understand better the interrelations and cultural, socio-political and economical relations in the urban - rural - nomad network.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Sedentary and Nomadic Population during Bronze Age Southern Levant

461

Bibliography Amiran, R., 1970. The Beginnings of Urbanization in Canaan. J. A. Saunders (ed.), Near Eastern Archaeology in the 20th Century. New York, 83–101. Ben Tor, A., 1992. The Archaeology of Ancient Israel, New Haven. Falconer, S. E./Fall, P. L., 2009. Settling the Valley: Agrarian Settlement and Interaction along the Jordan Rift during the Bronze Age. E. Kaptijn/L. Petit (eds.), A Timeless Vale: Archaeological and Related Essays on the Jordan Valley. Leiden, 97–107. Falconer, S. E./Fall, P. L./Jones, J. E., 2007. Life at the foundation of Bronze Age Civilization: agrarian villages in the Jordan Valley. T. E. Levy et al. (eds.), Crossing Jordan: North American Contributions to the Archaeology of Jordan. London, 225–232. Fall, P. L./Lines, L./Falconer, S. E., 1998. Seeds of Civilization: Bronze Age Rural Economy and Ecology in the Southern Levant. Annals of the Association of American Geographers 88, 107–125. Finkelstein. I., 1995. Living on the fringe. The Archaeology and History of the Negev, Sinai and Neighbouring Regions in the Bronze and Iron Ages, Sheffield. Frick, F. S., 1997. Cities: An Overview. E. M. Meyers (ed.), The Oxford Encyclopedia of Archaeology in the Near East. Oxford, 14–19. Glazer, E. K., 2014. Nomadic urban centres Arad and Bab edh Dhra: A framework of displacement. M. Tomorad (ed.), Research of the history and culture of the Ancient Near East in the Southeast Europe: 1. Croatia – Selected problems. Saarbrücken, 150–165. Grigson, C., 1998. Plough and Pasture in the Early Economy of the Southern Levant. T. E. Levy (ed.), The Archaeology of Society in the Holy Land. London, 245–268. Hopkins, D. C., 1985. The Highlands of Canaan. Agricultural Life in the Early Iron Age, Sheffield, UK. Horowitz, A., 1979. The Quaternary of Israel. New York. Horwitz, L., 1989. Diachronic Changes in rural Husbandry Practices in Bronze Age Settlements from the Refaim Valley, Israel. Palestine Exploration Quarterly 121, 44–54. Ibn Khaldûn, 2005. The Muqaddimah. An Introduction to History, Princeton. Ilan, O./Sebbane, M., 1989. Metallurgy, Trade, and the Urbanization of Southern Canaan in the Chalcolithic and Early Bronze Age. P. de Miroschedji (ed.), L’urbanisation de la Palestine à l’age du Bronze ancient: Bilan et perspectives des recherches actuelles; Actes du Colloque d’Emmaus, 20–24 octobre 1986. Oxford, 139–162. Issar, A. S./Zohar, M., 2007. Climate Change. Environment and History of the Near East. Berlin. Khazanov, A. M., 1984. Nomads and the outside world. Cambridge. LaBianca, O. S., 1990. Hesban 1: Sedentarization and Nomadization, Berrien Springs.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

462

Glazer

Liphschitz, N./Gophna, R./Hartman, M./Biger, G., 1991. The Beginning of Olive (Olea europaea) Cultivation in the Old World: A Reassessment. Journal of Archaeological Science 18, 441–453. Lyonnet, B., 2009. Who Lived in the Third-millennium “Round Cities” of Northern Syria? J. Szuchman (ed.), Nomads, tribes and the State in the Ancient Near East. Crossdisciplinary perspectives. Chicago, 179–200. Mazar, A., 1992. Archaeology of the Land of the Bible 10000–586 B.C.E., New York. Mumford, L., 1968. Grad u historiji, Zagreb. Schwartz, G. M., 1995. Pastoral Nomadism in Ancient Western Asia. J. M. Sasson (ed.), Civilizations of the Ancient Near East. New York, 249–258. Smith, A. B., 2005. Desert Solitude. The Evolution of Ideologies Among Pastoralists and Hunter-Gatherers in Arid North Africa. P. Veth et al. (eds.), Desert peoples: archaeological perspectives. Oxford, 261–275. Southall, A., 1998. The City in time and space, Cambridge. Stager, L. E., 1985. The Firstfruits of Civilization. J. N. Tubb (ed.), Palestine in the Bronze and Iron Ages. Papers in Honor of Olga Tufnell. Institute of Archaeology Occasional Papers 11. London, 172–188. Wapnish, P./Hesse, B., 1988. Urbanization and the Organization of Animal Production at Tell Jemmeh in the Middle Bronze Age Levant. Journal of Near Eastern Studies 47.2, 81–94. Wilson, J. A., 1973. The Story of Sinuhe. J. B. Pritchard (ed.), The Ancient Near East. Princeton, 7. Zohary, D./Hopf, M., 1988. Domestication of Plants in the Old World. Oxford.

Eva Katarina Glazer, Department of History, Centre for Croatian Studies – University of Zagreb, Borongajska cesta 83d, 10 000 Zagreb, Croatia, [email protected].

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9th ICAANE, Basel 2014, Vol. 3, 463–473

Aaron Greener

Analyzing the Late Bronze Age Imported Pottery Distribution in the Southern Levant: Overcoming Methodological Challenges It is the main goal of this paper to demonstrate the biases and difficulties of analyzing wide scale pottery distribution patterns based on data from excavations and publications of differing qualities and to offer methods to overcome them. I will focus on the problematics of assessing shard material remains.

Introduction In my recent PhD research, I attempted to outline the major developments in the circulation of the Cypriot and Aegean pottery which was imported to the Southern Levant during the Late Bronze Age (ca. 1570–1180 BCE), and formulate questions arising from their comparative examination. These ceramics remain as intriguing, tangible clues to what was once lively commerce and interaction connecting the Southern Levant to other parts of the Eastern Mediterranean during the 2nd Millennium BCE (e.g., Leonard 1994; Steel 2013). These interactions were not static but fluctuated with time and with the rise and fall of empires and cultures. Alongside the trade in raw materials, pottery vessels (and their contents) appear to have been part of a growing body of manufactured objects that circulated widely and symbolized an international Late Bronze Age culture (e.g. Yasur-Landau 2010). My interest in the imported vessels focused on their typological, chronological and geographical distribution, their places of origin, and also on how these vessels were integrated within the Near Eastern cultural assemblages (see Van Wijngaarden 2002). Among scholars who have studied the diffusion and function of Cypriot or Aegean pottery overseas diachronically, few have tried to treat them in parallel. Although we had a general knowledge of the facts that Cypriot exports outnumbered Aegean ones at most sites, and that Aegean pottery became common in the Eastern Mediterranean only in the 14th and

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

464

Greener

13th centuries BCE, the paucity of updated quantitative and qualitative data did not allow for more refined observations. As a result, we know little about their relative frequency per region and period, and the extent to which their repertoires and contexts of use coincided or differed. These are all crucial for evaluating the dynamic Eastern Mediterranean trade networks and geo-political situation, and the vessels’ economic, social, and cultural significance for the people of the Southern Levant. My study and analysis are based on a large and extensive databank which I assembled, and includes all the published imported pottery from sites across the Southern Levant. These include sites of all sizes as well as tombs, and the information derives from formal excavations as well as many surveys. This database allows for an in-depth study of the chronological and spatial distribution of the imported wares at the various sites, and a comprehensive analysis of the Southern Levantine distribution of the main Cypriot and Aegean pottery types.1 The results and ratios between the various pottery types, interesting chronological and regional differences, as well as the cultural significance of these wares, are currently still being evaluated and will hopefully be published in the near future. However, it is the main goal of this paper to demonstrate the biases and difficulties of assessing large bodies of data from excavations and publications of differing qualities, and to offer possible ways to overcome them. Beyond the obvious contribution to anyone studying Late Bronze Age imports, the methodologies outlined below should be of interest to anyone studying large bodies of shard material. After briefly addressing several problems which archaeologists must overcome when attempting to analyze wide scale pottery distributions, I will focus on the problematics of assessing distribution patterns while relying on shard material remains.

Possible Reservations and Biases of Pottery Distribution Analysis (and Ways to Overcome Them) 1. Studying the distribution patterns of specific imported pottery types over space and time presents a challenge, as the overall quantity of excavated imports at each site is influenced by the varying excavation scopes and intensity in each stratum. Moreover, the excavation and publication methods and quality greatly vary from site to site; even within each site it is often challenging to integrate the data from separate excavations and reports which were published decades apart (see Whitelaw 1999). To overcome these challenges, one must first compare the occurrences of the various imported pottery types within the same site or 1

The dissertation (Greener 2015) includes also a catalogue of all the Canaanite Late Bronze sites from which Cypriot or Aegean pottery have been published, and which are mentioned in this study.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Analyzing the Late Bronze Age Imported Pottery Distribution in the Southern Levant

465

stratum. These ratios can then be compared to other sites and contexts. Only when specific patterns are repeated consistently throughout many sites and regions, can their reliability be trusted. 2. Naturally, the quantities of imported vessels and shards vary from site to site. To prevent a situation where the results from a small number of sites (with a great quantity of wares) overshadow all other sites, the percentages of the shards/vessels from every site must be «normalized». In other words, the quantities of shards should not all be grouped together, but rather the results or ratios from each site, large or small, should be treated equally. This method was routinely utilized in my study, both when discussing the relative abundance of the Cypriot or Aegean individual vessel types and when comparing between the Cypriot and Aegean wares. Thus, the results represent the average abundance of the investigated wares at the sites, and do not necessarily indicate the numerical ratio representation of those vessels in the Southern Levant as a whole. 3. In accordance with the previous comment, one must avoid reading too deeply into the data originating from sites with meagre finds. Utilizing the statistics from such sites within a larger framework would result in meaningless results. Thus, only sites with at least a minimum number of imported wares (relevant for the study at hand) should be included in the general statistical analysis. 4. The imported pottery was often only partially published, or there were differences in the publication level of the Cypriot and Aegean imports.2 Additionally, older publications tended to over represent the Aegean wares, often publishing them all, while saving or publishing only samples of the Cypriot ones. Such unequal publications make straightforward comparisons between the Cypriot and Aegean imports tricky. In order to overcome this problem, I rated the archaeological publication according to the confidence level of the comparisons between both types of imports. Only publications with a confidence level 1 were utilized in my study when comparing between the Cypriot and Aegean wares. 1 – Full publication. Comparison is valid. 2 – Full publication, but with differing chronological definitions. Comparison is valid, though not within the LBII (between the 14th and 13th centuries BCE). 3 – Partial/selective publication of both types of imports (though without an apparent preference for either of them). Comparison is not valid, though the general

2

For example, the great quantities of the Aegean pottery from the ‘Amman airport were published in detail (Hankey 1974), while the Cypriot ones were never adequately published; unpublished Cypriot pottery from the Southern Coastal Plain and Western Negev was catalogued by Bergoffen (1989), while substantial Aegean pottery from these sites was never published.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

466

Greener

trend/preference can be assumed. 4 – Varying publication levels of the imported wares. Comparison is not valid. 5. Many of the published settlement strata and graves have very limited chronological use since they were continuously used for long periods of time, or the finds in them were insufficient to determine an exact dating. Many of the finds are designated as deriving from «Late Bronze Age II» or even a more general «Late Bronze» stratum. This of course limits the utilization of the finds from these contexts for most chronology-related analysis. Thus, only selected sites and strata were incorporated when attempting to compare the finds and distribution patterns between the 14th and 13th centuries BCE.

Assessing Past Distribution Patterns Based on Shard Material Remains The large majority of the finds from the settlement contexts are broken non-restorable pottery shards, whose find spots cannot be trusted to be the vessel’s actual last place of use in terms of strata and exact location (beyond the general settlement or funerary context designation). In other words, since small shards may have been moved as part of landfill or used within mud bricks in antiquity, their exact find spots within the settlement (both horizontal and vertical) do not necessarily indicate their deposition spot. Moreover, since different types of vessels may tend to break differently, the drawbacks of quantitative analysis relying on such shards must be evaluated. In the following section, I will offer a method to overcome these issues in order to achieve a comparative distribution pattern which is closest to the Late Bronze Age reality. Only the presence of a considerable number of shards from a specific vessel within one locus can be considered as an indicator that these shards were not moved or incorporated within mud bricks (see Stockhammer 2011: 287; 2012: 91). Throughout my study, only vessels from which at least 50% was found were considered «complete», and their find spot could be assumed to represent their deposition spot (usually indicating destruction levels). The assumption is that the complete vessel distribution ratios more accurately reflect the true ratios during the Late Bronze Age. They reflect the reality during times of destruction, and were less influenced by depositional and post depositional processes. Naturally, a greater percentage of complete vessels were found within the funerary contexts than within the settlement ones. Unfortunately, a distribution study based only on the complete vessels is not possible due to the meager quantities of complete vessels found within the Late Bronze Age settlement contexts.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Analyzing the Late Bronze Age Imported Pottery Distribution in the Southern Levant

467

Thus, the ratios and results of the analysis of the complete vessels were factored into the results arising from the analysis of all the sherds. The general ratio between the percentages of the sherds and the percentages of the complete vessels was determined at the beginning of each chapter (concerning the Cypriot and Aegean wares, and the comparison between them). The percentages of various specific distribution results of the sherds throughout the study were then multiplied according to this ratio. The results in percentages were mathematically adjusted in order to reach a sum of 100%. The differences between the two types of analysis can teach also of depositional, as well as post-depositional, processes which took place after the vessel went out of use, or during their collection and publication.3 It must be stressed that the tables below do not represent in any way the true ratios between the various types of vessels.4

Southern Levantine Distribution of the Cypriot Vessel Types When comparing the complete vessel analysis to the shard analysis regarding the ratios of open vs. closed vessels, we notice that the shard analysis resulted in a greater presence of open wares (see table 1). These results indicate that the open vessels tended to break into more pieces than the closed ones, and were thus discovered in greater quantities than what would reflect the true ratio during the Late Bronze Age (when examining all strata together – 73% of all shards were from open Cypriot vessels, while only 55% of the complete Cypriot vessels were open).5 Another possibility is that publications in which only the rims were studied would naturally identify more rims from open vessels as their wider mouths produced more shards than the narrow mouths of the closed jugs and juglets; indeed, an examination of the rims shows a great overrepresentation for the open vessels. Finally, it has been noted that vessels that were repeatedly used and broken within the domestic context, will consequently be represented in larger quantities in the archaeological record (Shott 1996; Varien/Potter 1997).

3

4

5

Another method which may be used for approximating the number of vessels deposited in a defined area is to determine the minimum number of items. Some publications include this information, while others do not. It was thus impossible for me to use this quantification parameter in my study. The methods mentioned in the notes above were utilized in my general study of the shard material in order to reach meaningful results. However, the data used for the comparison between the shards and the complete vessels could not be «normalized» between the sites (see notes 2 and 3 above) – since there were not many sites with a significant amount of complete vessels (from any given stratum). See also Steel (2006:162) who wrote that open vessels tend to break into more shards, as opposed to closed vessels which have a better state of preservation.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

468

Greener

Fig. 1. Example of a Cypriot Base Ring II Jug (© The Trustees of the British Museum). Fig. 2. Example of a Cypriot White Slip II Bowl (© The Trustees of the British Museum).

When examining the different types of imported Cypriot vessels (see table 2), we notice that the base ring and white slip wares are overrepresented among the shard material (figs. 1–2). Thus, the monochrome, and especially the white shaved vessels probably had a much greater presence than is reflected in the shard material. The reason for this phenomenon may be double fold. Firstly, the white shaved shards may not have always been recognized as imports by the excavators (as opposed to the more recognizable base ring and white slip shards). Secondly, perhaps these sturdy closed vessels tended to break into fewer shards than other types of vessels. The base ring material contained also many open bowls (see below), which contributed to its apparent overrepresentation when examining all the shards. Finally, we notice that the earlier base ring I wares were underrepresented among the shard material; perhaps many of the undistinguishable base ring shards (usually small non-indicative shards) should actually be identified with the base ring I types.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Analyzing the Late Bronze Age Imported Pottery Distribution in the Southern Levant Closed Cypriot

Open

469 Total N

N

%

N

%

All shards

1662

27.02

4489

72.98

6151

Complete vessels

128

44.76

158

55.24

286

Table 1. Cypriot pottery distribution: Open/closed vessels – settlement context (excluding unknown shapes). Base Ring I

Base Ring II

Base Ring

Bucchero

Monochrome

Red lustrous

White shaved

White Slip I

White Slip II

White Slip

Other vessel types

All shards - 7421 shards

8.77%

15.98%

43.18%

0.08%

7.09%

0.39%

3.92%

5.04%

37.30%

45.33%

796

Complete vessels 293 vessels

15.47%

18.35%

37.05%

0.36%

10.79%

0.36%

16.19%

1.44%

30.22%

35.25%

15

All LB stratasettlement context

Table 2: Cypriot pottery distribution: major vessels types – settlement context. Closed Aegean

Open

Total N

N

%

N

%

All shards

869

61.50%

544

38.50%

1413

Complete vessels

44

91.67%

4

8.33%

48

Table 3. Aegean pottery distribution: Open/closed vessels – settlement context (excluding unknown shapes). All stratasettlement context

Jarpiriform FF7

Jarstirrup FF46

Alabastra FF19

Flasks FF47-49

Amphoroid krater FF8

Cup FF58

Bowldeep FF80

Bowlshallow FF85

Kylix FF79

Unknown and mic.

All shards -1817 shards

11.90%

35.69%

5.65%

6.85%

9.98%

11.59%

5.85%

3.73%

8.77%

825

Complete vessels -51 vessels

9.30%

58.14%

4.65%

20.93%

4.65%

2.33%

0.00%

0.00%

0.00%

8

Table 4. Aegean pottery distribution: All vessel types – settlement strata.

Southern Levantine Distribution of the Aegean Vessel Types When comparing the complete vessel analysis to the shard analysis regarding the ratios of open vs. closed vessels (table 3), we notice that, as with the Cypriot wares, the shard analysis resulted in a greater presence of open wares (when examining all strata together – 38% of all shards were from open Aegean vessels, while only 8% of the complete Aegean vessels were

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

470

Greener

Aegean/Cypriot wares

Cypriot

All shards Complete vessels

Aegean

Total

N

%

N

%

4716

86.83%

715

13.17%

5431

82

73.21%

30

26.79%

112

Table 5. Aegean/Cypriot ware distribution: settlement context. Aegean/Cypriot wares – open vessels

All shards Complete vessels

Cypriot

Aegean

Total

N

%

N

%

2879

96.26%

112

3.74%

2991

50

98.04%

1

1.96%

51

Table 6: Aegean/Cypriot pottery distribution: Open vessels – settlement context. Aegean/Cypriot wares – closed vessels

All shards Complete vessels

Cypriot

Aegean

Total

N

%

N

%

1213

77.86%

345

22.14%

1558

30

52.63%

27

47.37%

57

Table 7: Aegean/Cypriot pottery distribution: closed vessels – settlement context.

open). These results may indicate that the open vessels tended to break into more pieces than the closed ones, and were thus discovered in greater quantities than what would reflect the true ratio during the Late Bronze Age. Another possibility is that publications in which only the rims were studied would naturally identify more rims from open vessels as their wider mouths produced more shards than the narrow mouths of the closed jugs and juglets; indeed, an examination of the rims shows a great overrepresentation for the open vessels. Finally, it has been noted that vessels that were repeatedly used and broken within the domestic context, will consequently be represented in larger quantities in the archaeological record.6 When analyzing the different types of vessels, we notice that all the open wares are overrepresented among the shard material, as opposed to the closed vessels which were underrepresented (see table 4), especially the closed stirrup jars and the flasks (figs. 3–4). 6

Another possible reason for the great discrepancies between the analyses of the shard material and of the complete vessels is the fact that the destruction levels of the end of the Late Bronze age occurred after the end of the Aegean importation, thus a larger percentage of the material is residual, leading to a heightened presence of the open Aegean wares. All in all, the complete vessels constituted a smaller percentage of the total number of shards when compared to the Cypriot imports (ca. 5% of the Cypriot wares compared to ca. 3% of the Aegean ones).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Analyzing the Late Bronze Age Imported Pottery Distribution in the Southern Levant

471

Fig. 3. Example of an Aegean Stirrup Jar (photo taken by the author at the Corinth Museum Warehouse).

Fig. 4. Example of an Aegean Flask (photo taken by the author at the Tiryns Excavation Warehouse).

It is possible that many of the flask shards were not always recognized as imports by the excavators.

Comparing the Cypriot and Aegean Pottery Distribution Patterns When comparing the complete vessel analysis to the shard analysis (from all LBII strata), we notice that the Aegean wares may have actually suffered from an underrepresentation in the shard material, and their true presence in the settlement context during the Late Bronze Age was probably greater than what would appear from this analysis (see tables 5–7). The

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

472

Greener

reason/s for the underrepresentation of the Aegean wares is unclear. One option may be related to the fact that these mostly small closed vessels tended not to break as easily into many shards. In general, there were hardly any complete open Aegean vessels, and their relative occurrence within both types of analyses remains very small.

Conclusions This paper is based on a database which includes the published imported shards from many excavations and surveys conducted throughout the Southern Levant. Unfortunately, the amount of high quality Late Bronze Age data available for my research is limited (highlighting the complexities and pitfalls of assessing large bodies of data from a range of geographic areas and sites, each with different excavation and publication standards), and so the analysis of the complete imported vessels had to be factored into the general results. This demonstrated that the open vessels tended to break into more sherds than the closed vessels, and were thus over-represented within the sherd material. The Aegean sherds seem to have also been over-represented among the imported sherds. This information allowed us to “manipulate” the results and reach more trustworthy conclusions regarding the true ratios between the various vessel types during the Late Bronze Age, as well as to better understand the deposition and post deposition processes which these vessels and shards went through. Only if we take these issues into consideration can we reliably examine the overall incorporation and major developments in the circulation of the imported vessels in Canaan, and formulate questions arising from their comparative examination. I have focused on the differences between the analysis of all shards and the analysis of only the complete vessels. Since almost all the pottery within settlement contexts is found broken into shards, this factoring method can and should be used by any archaeologists evaluating the distribution of pottery wares.

Acknowledgments This study is based on my PhD Thesis on the Cypriot and Aegean Imported Wares during the Late Bronze Age, written in the Martin (Szusz) Department of Archaeology and Land of Israel Studies at Bar-Ilan University, under the guidance of Prof. Avraham Faust (Bar-Ilan University) and Prof. Assaf Yasur-Landau (Haifa University). I would like to thank them for their valuable help and comments. Part of this study was conducted while I was an Associate Member of the American School of Classical Studies at Athens.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Analyzing the Late Bronze Age Imported Pottery Distribution in the Southern Levant

473

Bibliography Bergoffen, C., 1989. A Comparative Study of the Regional Distribution of Cypriot Pottery in Canaan and Egypt in the Late Bronze Age, New York University: PhD Thesis. Greener, A., 2015. Late Bronze Age Imported Pottery in the Land of Israel: Between Economy, Society and Symbolism, Bar-Ilan University: PhD Thesis. Hankey, V., 1974. A Late Bronze Age Temple at Amman: I. The Aegean Pottery: II. Vases and Objects Made of Stone. Levant 6, 131–178. Leonard, A., Jr., 1994. An Index to the Late Bronze Age Aegean Pottery from Syria-Palestine, Jonsered. Shott, A., 1996. Mortal Pots: on Use Life and Vessel Size in the Formation of the Ceramic Assemblage. American Antiquity 61, 463–482. Steel, L., 2006. Cypriot and Mycenaean Pottery. A. Mazar (ed.), Timnah (Tel Batash) III. Jerusalem, 151–172. –

2013. Materiality and Consumption in the Bronze Age Mediterranean. New York.

Stockhammer, P. W., 2011. An Aegean Glance at Megiddo. W. Gauss et al. (eds.), Our Cups are Full: Pottery and Society in the Aegean Bronze Age. Oxford, 282–296. –

2012. Entangled Pottery: Phenomena of Appropriation in the Late Bronze Age Eastern Mediterranean. J. Maran/P. W. Stockhammer (eds.), Materiality and Social Practice: Transformative Capacities of Intercultural Encounters. Oxford, 89–103.

Van Wijngaarden, G. J., 2002. Use and Appreciation of Mycenaean Pottery in the Levant, Cyprus and Italy (1600–1200 B.C.), Amsterdam. Varien, M. D./Potter, M., 1997. Unpacking the Discard Equation: Simulating the Accumulation of Artifacts in the Archaeological Record. American Antiquity 62, 193–213. Whitelaw, T., 1999. Value, Meaning and Context in the Interpretation of Mycenaean Ceramics. Archaeological Dialogues 6, 31–35. Yasur-Landau, A., 2010. The Philistines and Aegean Migration at the End of the Late Bronze Age, Cambridge.

Aaron Greener, Bar-Ilan University, W.F. Albright Institute of Archaeological Research.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 475–489

Haskel J. Greenfield – Itzhaq Shai – Aren M. Maeir

Understanding Early Bronze Age Urban Patterns from the Perspective of Non-Elite Neighbourhood: The Excavations at Tell es-Safi/Gath, Israel This paper describes some of the results of the excavations of the Early Bronze Age III nonelite neighbourhood in Area E of Tell es-Safi/Gath, Israel. The neighbourhood contained a layout that is found at other EB urban centres, including sturdy small multi-room houses, often with a courtyard and in few cases a small storage room. Occupants had access to local and some interregional trade goods, used various recording methods, sacrificed expensive animals, and built and maintained the neighbourhood over a long period of time. These results are discussed in conjunction with other EB finds from the site, both from the surface survey and from the excavations in other areas. Based on these data and analysis, we discuss the role of Tell es-Safi/Gath during the Early Bronze Age in a regional perspective. The results of the excavation suggest that Tell es-Safi/Gath was an important EB III political and economic centre in this region.

Introduction Until recently, most research on the EB of the Near East has focused on monumental and/ or public structures and areas of sites. Hence, relatively little is known about lower stratum neighbourhoods. This was particularly true of the southern Levant region. Beginning in 2004, we began the excavation of a non-elite neighbourhood at the site of Tell es-Safi/Gath in central Israel. In this paper, we present some preliminary results of our excavations of the Early Bronze (EB) Age layers at the site of Tell es-Safi/Gath in central Israel and discuss their implications for increasing our understanding of EB urban societies, particularly from the perspective of a non-elite neighbourhood. While we have been working on the Early Bronze Age of the site since 2004, we will be summarising the information up until 2010.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

476

Greenfield – Shai – Maeir

Fig. 1. Location of Tell es-Safi/Gath and some important EB sites in the region.

Tell es-Safi/Gath is a multi-period site, occupied from the proto-historic through modern periods and is most famous in the literature for its substantial Philistine Iron Age occupation (Hitchcock/Maeir 2014; Maeir 2012). However, in this paper, we will introduce another substantial occupation at the site, the EB. This is when the site first achieves its maximum size and becomes one of the largest EB urban centers in the southern half of the southern Levant.

Location Geographically, Tell es-Safi/Gath is located halfway between the coastline and the pinnacle of the mountain spine where Jerusalem sits (fig. 1). It is on the edge of the Judean foothills (Shephelah) and the coastal plain. On a clear day, one can see the Mediterranean from the top of the site. Tell es-Safi/Gath is in a region which includes a number of large and medium sized tell sites with extensive EB occupations. These include Yarmouth to the east, and Erani, Lachish, Hesi, Halif, ‘Ira and Sakan to the south and southwest. Gezer is the closest urban EB site to the north.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Understanding Early Bronze Age Urban Patterns …

477

Fig. 2. Map showing the size of Tell es-Safi/Gath in different time periods based on typochronological analysis of surface remains.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

478

Greenfield – Shai – Maeir

Fig. 3. Map of site showing location of various excavation areas.

Size of Site An extensive surface survey of the site was conducted and enables reconstruction of the extent of different periods of occupations (Uziel/Maeir 2005, 2012a). While some Chalcolithic sherds were recovered, it was not possible to plot the spatial extent of the site for this period. During the EB, the site emerges as a major urban centre for the region (fig. 2). In the EB I, the site was relatively small and restricted to part of the Tell or upper mound. But, this situation changes dramatically during the EB II/III period when the site spreads across the entire upper mound and covers an area of c. 24 ha. During the ensuing MB, the scale of occupation shrinks dramatically. It was only during the later periods (LB and IA) that the occupation of the site extends beyond the tell to create a lower city and the site reaches its largest extent. Hence, the EB II/III is the largest occupation for the pre-Iron Age periods on the site.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Understanding Early Bronze Age Urban Patterns …

479

The results of the surface survey have been verified in a number of excavations across the site (fig. 3). At present, wherever large or small excavations have been opened, substantial EB III deposits have been discovered from the west to the east across the entire extent of the site (Areas F, P, A, and E) (Shai et al. 2014).

Location of High Status and Public Buildings At this point, no clear-cut evidence of high status and/or public buildings has been uncovered at the site. However, the topography of the site suggests that such structures would have been found at the western end of the site, with its high acropolis-like feature, that dominates the landscape. It has a view of all the lowlands to the west and hill country to the north, south and east, and has constant fresh air blowing in from the west. Given the topographical prominence of this end of the site, it is thought that this is the most likely location for palaces, temples and other important public buildings. However, it is impossible to extensively investigate the EB in this area since the layers are too deeply buried beneath later deposits, ranging from the 4th millennium BCE to the 20th century CE. A deep step trench at this end of the site, Area F excavated under the directorship of J. Chadwick, has confirmed the presence of the EB in this area.

Fortifications Evidence for a substantial fortification system with a stone foundation (that was topped by a mud-brick superstructure) and composed of insets and outsets has been found in at least two parts of the site (Area F to the west and Area P [directed by J. Katz] in the centre). It is consistently 2.7 m wide and preserved to a height of 2–3 m (fig. 4). In both locations, long sections of this large and thick stone fortification wall have been excavated. Furthermore, it seems to us that the fortification reported by Bliss and Macalister is also part of the EB wall. As well, a third potential outset has been recently identified at the far eastern end of the site below Area E (Shai et al. 2015).

EB Excavation at East End of Tell The excavations in Area E at the eastern end of the site tell a very different kind of story. The EB III deposits are only partially covered by later remains. This has allowed us to open up a large area of the EB occupation without having to deal with the thick later deposits that

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

480

Greenfield – Shai – Maeir

Fig. 4. Map showing extent of fortification system.

cover it elsewhere on the site. In some locations, the EB deposits are only a few cm below the modern surface. Chronologically diagnostic ceramics from all three phases of the EB (I–III) have been recovered in this part of the site. Stratigraphically, three stratigraphic horizons (Strata) have already been identified: E5 (EB IIIB), E6 (EB IIIA), and E7 (EB II). The E6 and 7 Strata have not been extensively excavated yet. The ensuing section of the paper will focus on the E5 Stratum.

Radiocarbon Dates for EB III Occupation Horizon Radiocarbon dates for the EB III occupation have been obtained through high precision dating and Bayesian modeling. It relies upon the use of short-lived organics, such as olive pits that derive only from very secure depositional contexts that reflect their use and/or discard. It avoids the problems of later contamination. In general, the 1 sigma calibrated range for the EB III occupation at the site is c. 2830–2500 BCE (Regev 2013), which is very similar to the terminal EB III occupation at other major sites in the region, such as nearby Tel Yarmouth (Regev et al. 2012). These dates provide support for the new «high» radiocarbon chronology for the southern Levant.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Understanding Early Bronze Age Urban Patterns …

Fig. 5. Plan of houses excavated in Stratum E5 from Area E.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

481

482

Greenfield – Shai – Maeir

Stratum E5 Architecture The excavations in Area E of the site represent the remains of a domestic urban neighborhood from the EB III. From fig. 5, the remains of several building complexes are visible. Architecturally and stratigraphically, we have been able to define three separate architectural and artefactual phases dating to the EB III alone (Stratum E5). The plan shows the various internal phasing of some of the buildings during the later EB III (known as Strata E5a–c). Most of the structures were used in all three phases, but some changes in their use were observed between phases as walls go in and out of use or are remodeled. In Stratum E5, at least two building complexes have been identified, with one on either side of the alley. Each complex includes courtyards, rooms (for cooking, sleeping, and storage), and various installations. It is likely that the building complex to the east of the courtyard includes several (at least 3) different house complexes based on wall thickness, doorways and hearth distribution. In total, four house complexes have been partially or totally uncovered to date. Not only is there evidence for continuity in the ceramic inventory (Uziel and Maeir 2012b), there is also architectural continuity between Strata (Shai et al. 2012). As the neighbourhood is being renewed, the mud-brick superstructure of the underlying Stratum E6 is mostly leveled off and the new stone foundations for the houses in the later Stratum E5 are laid directly on top of the stumps of the old mud-brick walls. Similarly, the Strata E5 and E6 buildings follow same orientation. The stumps of the Stratum E6 mud-brick wall superstructure are visible below the Stratum E5 stone foundation. The nature of the architecture is relatively simple. All walls have a stone foundation of approximately three courses. This is topped with a mud-brick superstructure of indeterminate height. The foundation of the exterior walls for each building is built of two rows of large field stones, while the internal walls use two rows of smaller field stones.

Stratum E5 Neighbourhood Orientation and Slope In contrast to later periods at the site where whole areas are leveled off and the orientation of buildings follow a more regular plan, the EB neighbourhood appears to be built more haphazardly as it follows the natural slope of the terrain. The entire area slopes downhill to the east. The structures descend in elevation from NW to SE.

Alleyway In between the two building complexes, there is a narrow alley that divides them. It was not roofed over and the sediments were exposed to torrential rain and runoff. The runoff made a V- or U-shaped depression through the centre of the alley. To counter erosion from rain and to make it a more comfortable walking surface, the ruts in the alley were periodically

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Understanding Early Bronze Age Urban Patterns …

483

and purposefully filled in with small and medium-sized stream pebbles, broken ceramics, and other artefacts and ecofacts. While there is substantial evidence for disposal of organic remains in the alley (such as carbonised grains), there is very little evidence of animal bones.

Installations Associated with each of the building complexes are various installations. One type is the large circular stone installations that are about a metre high and found in the middle of rooms that are probably open-courtyards. Some are built of large field stones and are solid, while others are built of smaller field stones and are hollow. The hollow ones were probably granaries since high quantities of carbonised plant remains have been found within. In contrast, the function of the solid installation is unknown and may have included use as altars or working platforms since it is topped with a large capping stone. It was used throughout the E5 Stratum. Interestingly, the solid installation is found in the same room as the donkey sacrifice (to be discussed later).

Cooking/Heating Installations Another type of installation is more clearly recognisable for cooking and/or heating. These have the more obvious qualities of hearths. They are generally circular pebble-based affairs which are found in each group of rooms. They are made of relatively small pebble stones mixed with crushed chalk and are formed into a circle. Generally there is one in each house, but there is at least one example where two are found next to each other. In some cases, they have an underlying mud-brick platform that sits on an EB III earthen floor. This would have raised them up slightly. In general, they were used for a low fire since none of the pebbles were heated (thermally altered) beyond 400°F (or 200°C). They were probably used for low temperature cooking and/or warming of food or perhaps to heat the room itself.

Artefactual Remains E5 Pottery Assemblage The Stratum E5 (later EB III) ceramic assemblage is the most comprehensive of all the strata since this layer has been the focus of the excavations. The wares are largely reflective of household functions. Few classes of fine wares were recovered. The Stratum E5 pottery assemblage from Area E is dominated by a few ceramic types. The assemblage is composed of types that are commonly found elsewhere in later EB III sites across the region. The most common is that of storage vessels (some of which are shown here). There are jars with a flaring rim and hole-mouth jars. Other types of vessels are for serving and cooking, such

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

484

Greenfield – Shai – Maeir

as bowls and platters, some of which are decorated with web-pattern burnishing (Shai et al. 2014; Uziel/Maeir 2012b). Petrographic analysis of the EB pottery has amply demonstrated that production was largely a local affair (Ben-Shlomo et al. 2009). The pottery was made from the local clays that are found within the Elah River valley, which runs along the base of the site. Two small bowl fragments of Khirbet Kerak Ware have been found in EB layers at the site. While they are commonly thought to be non-local products, they were probably locally produced as well. Petrographic analysis of the clays shows that they include local rendzina or rendzinaderived soils that are found around the site. Also, the non-plastic (clay) components found in the thin-sections are well-represented in the local lithology of the site.

Pot Marks A large number of EB vessels have pot marks. So far, over three hundred specimens have been recovered making this one of the largest samples from a domestic neighbourhood complex. They are incised while the clay is still wet and probably don’t reflect what is in the vessels. Kisos has suggested that these marks perhaps indicate symbolic messages derived from the ruling strata of this society. Other possible explanations of these marks, such as potters’ marks, definition of content and volume, were demonstrated as irrelevant (Kisos 2014a).

Cylinder Seal It is very rare to find cylinder seals in EB contexts from the southern Levant. Yet, in Area E, in one of the residential complexes, we recovered an exquisitely carved hippopotamus ivory cylinder seal (Maeir et al. 2011). The skill level to make it suggests that the seal was made by a highly proficient artist. The quality of the raw material and the drawing also suggest that the owner was of relatively high status. The presence of such items hints that the socio-economic organisation of the local inhabitants was relatively complex and would have included administrative or exchange specialists.

Game Pieces Recreation was also part of life during the EB of Tell es-Safi/Gath. At least three game board fragments have been found thus far. They are made of chalk, probably locally produced. They have vertical and horizontal lines incised on them. In addition, there are a few small chalk pieces that may be the gaming pieces. The shape of the board is reminiscent of the well-known Egyptian game called senet. It is the most common type of gaming board known from the EB II–III in the southern Levant. Many astragali with use-wear polish, suggesting they were used as gaming pieces, have also been recovered (Kisos 2014b).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Understanding Early Bronze Age Urban Patterns …

485

Exotic Trade Goods Even though hippopotamus ivory is usually assumed to be a highly desired trade good, it may have been of local origin since Hippopotamus’ were available in local rivers near the coast until the Early Iron Age (Horwitz/Tchernov 1990). There is the already mentioned Hippopotamus ivory cylinder seal, as well as other pieces of hippopotamus ivory, which are shaped and drilled. In addition, there are maceheads made of non-local stone. Clearly, expensive trade goods are being brought into the site from outside the immediate vicinity of the site and of the region. Their discovery in a commoner’s house is indicative that such goods were not simply signatures of elite trade networks.

EBA Lithics Even though the period is known as the Early Bronze Age, it is well known that bronze tools are not represented in the southern Levant during this period. The production of chipped stone tools remains a major technology in the Early Bronze Age. There is a strong tradition of local production since all phases of production are present at the site, including cores, debitage and finished tools. They are littered across the excavation area. Most are probably locally produced since the site sits atop a limestone outcrop filled with large flint nodules. Grinding stones are found dispersed across the neighbourhood, both within and discarded outside of each house complex. They appear to be associated with food preparation areas, when found in situ. Petrographic analysis suggests that they were brought from the Galilee or Golan region, in the north of present-day Israel (Beller 2014). In spite of the fact that few metal objects have been found in the EB levels at the site, the presence of metal knife blades can be tracked by examining slice- or butchering marks on bones. Microscopic analysis demonstrates that all of the EB butchering marks on animal bones are from chipped stone tools. There is no evidence of metal use in the butchering process during the period (Greenfield et al. 2015). Most of the marks were probably made by simple unmodified flakes since there is little evidence for retouch. There is very little evidence for metal being used for such purposes until the latter half of Middle Bronze Age.

Ritual A domestic ass (donkey) was buried in a deposit that was sealed beneath EB IIIB house floor. It was buried up against the EB IIIA wall as the house was being rebuilt. It is very likely that it was a sacrificial deposit since the head was cut off and laid backwards on the back of the animal. The head is not in a natural position and there is a space between cervical and thoracic vertebrae. It appears to be a foundation deposit that protects the inhabitants

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

486

Greenfield – Shai – Maeir

of the house complex or neighbourhood and sanctifies it by its sacrifice (Greenfield et al. 2012). A few fragments of zoomorphic (donkey) figurines have been recovered. From Old and Middle Kingdom Egyptian texts, donkeys are well-known as pack animals to carry long-distance trade goods. But, donkeys were also used to carry elite individuals and were sacrificed under various conditions in ancient Near Eastern cultures (Way 2011). Could the presence of the figurines and the donkey sacrifice be indicative of the use of the donkey as a totem for a community of merchants? Could this neighbourhood be for merchant families? It does not appear to be a high class neighbourhood, yet they have access to exotic and probably costly goods.

Conclusions The presence of an extensive fortification system that encircled the entire tell suggests that Tell es-Safi/Gath was a major regional polity. It probably competed with the other nearby centres, such as Yarmouth (de Miroschedji 2006, 2009). There is little reason that a neighbouring kingdom would have wanted to have second tier nearby centres with a substantial fortification system. The surface survey indicates that the EB II/III occupation at Tell es-Safi/Gath covered an area of 24 ha. This dimension may be used to suggest that the site was hierarchically subservient to neighbouring Yarmouth, which has a slightly larger area of occupation (32 ha) and evidence for a large palace and similar fortification system. The lack of evidence for a palace and elite complex at Tell es-Safi/Gath is not indicative of their absence since it is thought that they are probably deeply buried at the western end of the site. Given the evidence described below, it is likely that Tell es-Safi/Gath was an independent polity that competed with Yarmouth for influence in the region. Our research suggests that Tell es-Safi/ Gath was an important EB III regional political and economic centre. It is either the centre of an independent polity or positioned just below the larger centre at nearby Tell Yarmouth. The excavations in Area E demonstrated the presence of an EB III domestic quarter with several houses and a small street or alleyway. There is no evidence of industrial activities. The houses were built and maintained over a long period of time. They are typical small multi-room structures characteristic of the period. Given the size of the rooms, they are 1–2 rooms associated with a courtyard, and in few cases also a small storage room. The courtyard was used for domestic activities, which included grain storage and other working platforms (stone built circular installations). Comparison with other sites suggests that this is a common layout in EB urban centers.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Understanding Early Bronze Age Urban Patterns …

487

In spite of the absence for any elite or administrative structures in this area, the occupants had access to exotic trade goods, used various recording methods, and sacrificed expensive animals. When we began our excavations of this area, we originally thought that this area would be a low class neighbourhood since it far from the center of power at the other end of city. In contrast, we discovered that it was a residential area for probably middle-class families who were possibly merchants involved in transport and trade.

Acknowledgments We would like to acknowledge the following organizations for generously funding our research: The Ackerman Family Bar-Ilan University Expedition to Gath and its staff and team members, Bar-Ilan University and the Department of Land of Israel Studies and Archaeology at Bar-Ilan University, the Social Science and Humanities Research Council of Canada (PG# 895-2011-1005), The University of Manitoba, St. Paul’s College and Near Eastern and Biblical Archaeology Laboratory of the University of Manitoba, and a host of colleagues and students without whose unstinting support this research would not have been successfully carried out.

Endnotes The overall excavations of the site are directed by AMM, while the excavations of the Early Bronze Age occupation at Tell es-Safi/Gath are conducted as a joint operation by the University of Manitoba and Bar-Ilan University, co-directed by HJG and AMM, with IS as Area Supervisor.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

488

Greenfield – Shai – Maeir

Bibliography Beller, J., 2014. The Life History of the Basalt Ground Stone Artefacts from an Early Urban Domestic Neighbourhood in the Southern Levant: The Remains from Early Bronze III Tell es-Safi/Gath, Israel. Unpublished MA Thesis. Winnipeg, University of Manitoba. Ben-Shlomo, D./Uziel, J./Maeir, A. M., 2009. Pottery production at Tell es-Safi/Gath: a longue durée perspective. Journal of Archaeological Science 36, 2258–2273. de Miroschedji, P., 2006. At the dawn of history: sociopolitical developments in southwestern Canaan in Early Bronze Age III. A. M. Maeir/P. de Miroschedji (eds.), ‘I Will Speak the Riddles of Ancient Times’: Archaeological and Historical Studies in Honor of Amihai Mazar on the Occasion of His Sixtieth Birthday. Winona Lake, IN, 55–78. –

2009. Rise and collapse in the southern Levant in the Early Bronze Age. Scienze dell’antichità. Storia Archeologia Antropologia 15, 101–129.

Greenfield, H. J./Shai, I./Maeir, A. M., 2012. Being an “ass”: An Early Bronze Age burial of a donkey from Tell es-Safi/Gath, Israel. Bioarchaeology of the Near East 6, 21–52. Greenfield, H. J./Brown, A./Shai, I./Maeir, A. M., 2015. Preliminary analysis of the fauna from the Early Bronze Age III neighbourhood at Tell es-Safi/Gath, Israel. N. Marom et al. (eds.), Bones and Identity: Zooarchaeological Approaches to Reconstructing Social and Cultural Landscapes in Southwest Asia (Proceedings of the ICAZ-SW Asia Conference, Haifa, June 23–28, 2013), i. p. Oxford. Hitchcock, L. A./Maeir, A. M., 2014. Yo-Ho, Yo-Ho, a Seren’s Life for Me! World Archaeology 46.3, 1–17. Horwitz, L. K./Tchernov, E., 1990. Cultural and environmental implications of Hippopotamus bone remains in archaeological contexts in the Levant. Bulletin of the American Schools of Oriental Research 280, 67–76. Kisos, S., 2014a. The Definition and Understanding of Marks on Pottery from the Early Bronze III in the Southern Coastal Plain of Canaan: A Case Study from Tell es-Safi/Gath. Unpublished Master’s Thesis. Ramat-Gan, Bar-Ilan University. –

2014b. Leisure activity in the EB III: game boards as a reflection of social interaction at Tell es-Safi/Gath. Paper presented at the 2014 ASOR conference (San Diego).

Maeir, A. M. (ed.), 2012. Tell es-Safi/Gath I: Report on the 1996–2005 Seasons. Ägypten und Altes Testament 69. Wiesbaden. Maeir, A. M./Shai, I./Horwitz, L. K., 2011. ‘Like a Lion in Cover’: a cylinder seal from Early Bronze Age III Tell es-Safi/Gath, Israel. Israel Exploration Journal 61 (1), 12–31.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Understanding Early Bronze Age Urban Patterns …

489

Regev, J., 2013. Chronology of the Early Bronze Age in the Southern Levant Based on 14C Dates in Relation to Context, Stratigraphy and Cultural Remains, Modeled with Bayesian Analysis. Ramat Gan, Israel, Ph.D. thesis, Bar-Ilan University. Regev, J./de Miroschedji, P./Greenberg, R./Braun, E./Greenhut, Z.,/Boaretto, E., 2012. Chronology of the Early Bronze Age in the southern Levant: new analysis for a high chronology. Radiocarbon 54, 525‒566. Shai, I./Uziel, J./Maeir, A. M., 2012. The architecture and stratigraphy of Area E: Strata E1–E5. Maeir 2012: 221–234. Shai, I./Greenfield, H. J./Eliyahu-Behar, A./Regev, J./Boaretto, E./Maeir, A. M., 2014. The Early Bronze Age remains at Tell es-Safi/Gath, Israel: an interim report. Tel Aviv 41.1, 20–49. Shai, I./Chadwick, J. R./Welch, E./Katz, J./Greenfield, H. J./Maeir, A. M., 2015. The Early Bronze Age fortifications at Tell es-Safi/Gath, Israel. Palestine Exploration Quarterly i. p. Uziel, J./Maeir, A. M., 2005. Scratching the surface at Gath: implications of the Tell es-Safi/Gath surface survey. Tel Aviv. Journal of the Institute of Archaeology of Tel Aviv University 32.1, 50–75. –

2012a. The location, size and periods settlement at Tell es-Safi/Gath: the surface survey results.

–

2012b. The Early Bronze Age III pottery from Area E. Maeir 2012: 235–239.

Maeir 2012: 173–181. Way, K. C., 2011. Donkeys in the Biblical World: Ceremony and Symbol. Winona Lake, IN.

Haskel J. Greenfield, University of Manitoba, Department of Anthropology and St. Paul’s College, FA 432, Winnipeg, MB, R3T5V5, Canada. [email protected]. Itzhaq Shai, Israel Heritage Department, Ariel University, P.O.B. 3, Ariel 40700 Israel; [email protected]. Aren M. Maeir, Bar-Ilan University, Institute of Archaeology, the Martin (Szusz) Department of Land of Israel Studies and Archaeology, Ramat Gan 52900, Israel; [email protected].

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9th ICAANE, Basel 2014, Vol. 3, 491–502

Ann E. Killebrew – Jamie Quartermaine

Total Archaeology@Tel Akko (The 2013 and 2014 Seasons): Excavation, Survey, Community Outreach and New Approaches to Landscape Archaeology in 3D The 2013–2014 excavation seasons at Tel Akko entailed excavation, pedestrian survey of the mound, a multi-faceted landscape survey of the Akko Plain, and a community outreach program. Highlights include the discovery of iron working smithies dating to the Persian period and new insights regarding Tel Akko and its environs.

Introduction The 22-hectare maritime harbor settlement of Tel Akko has dominated the Plain of Akko’s ancient landscape for millennia (fig. 1). First inhabited in the Early Bronze Age, Tel Akko served as a major city for much of the 2nd and 1st millennia BCE. During the Hellenistic period, probably as a result of the changing coastline, the population of Akko migrated westward off the tell, to an area now under the modern city and its historic walled Old City. Today the mound is situated ca. 600 m inland and ca. 1.1 km east of the modern town of Akko and the UNESCO World Heritage site of Ottoman- and Crusader-period Acre (fig. 1). (for a summary of Akko’s history and primary sources mentioning the site and an overview of the first expedition [1973–1989] to Tel Akko, see e.g. Dothan 1976; 1993; and Artzy/Beeri 2010). Led by the late Moshe Dothan and a team from the University of Haifa in co-operation with Diethelm Conrad from Marburg University, the 1973–1989 expedition opened five major excavation areas (A/AB/B, C, F, H and K) and several smaller trenches (G, P, PH, and S; fig. 2). Renewed excavations and survey under the direction of Ann E. Killebrew and Michal Artzy began in 2010. The Tel Akko Total Archaeology Project, as the name implies, incorporates a holistic approach to archaeology. It integrates an innovative field school focusing on skills necessary for 21st century archaeology and a multi-dimensional research agenda, under the direction of Killebrew. The project encompasses archaeological

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

492

Killebrew – Quartermaine

Fig. 1. Tel Akko and its Environs. Graphics by R. Stidsing. Tel Akko Total Archaeology Project.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Total Archaeology@Tel Akko (The 2013 and 2014 Seasons)

493

Fig. 2. Areas of Excavation and Survey at Tel Akko. Graphics by J. Skinner and R. Stidsing. Tel Akko Total Archaeology Project.

survey, systematic excavation, a robust conservation plan, archaeological sciences, a public outreach program, and the incorporation of the largely unpublished results of Dothan’s excavations focusing on the extensively excavated Area A/AB/B (for a recent overview see Killebrew/Olson 2014). Under the direction of Artzy, Tel Akko’s topography, site formation, and location of its ancient harbors are being investigated by documentary studies, GPR (Ground Penetrating Radar) and ERT (Electrical Resistivity Tomography) geophysical surveys. Combined with coring both on and off the tell, these techniques make it possible to examine the sub-surface make-up of the mound and its environs to reconstruct past landscapes (see e.g. Artzy 2012; Kaniewski et al. 2013; and Artzy/Quartermaine 2014). Key to our understanding and interpretation of Tel Akko and its surroundings is a multi-faceted 3D approach to the documentation and analysis of Tel Akko within its broader landscape (for a summary of our earlier use of 3D documentation and analyses see e.g. Olson et al. 2013; Quartermaine et al. 2014).*

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

494

Killebrew – Quartermaine

Excavation During the 2013–2014 excavation seasons, excavations continued in Dothan’s unpublished Area A/AB (fig. 2). Massive amounts of iron slag, found by the kilo, were recovered in Persian and Late Iron Age contexts. The appearance of hearths, tuyeres and significant quantities of iron hammer scales, indicating the presence of iron smithies, attest to the importance of large-scale metal working at Phoenician Akko. The realization of the massive scale of the metal working industry at Akko, not noted in the first series of excavations, contextualizes one of Dothan’s most remarkable discoveries from Area A, uncovered in 1980, which was a large Persian-period administration building that contained a 7-line Phoenician ostracon (Dothan 1985). A monumental wall, since removed by Dothan, may have been part of an associated Persian period cultic area. This early 5th century BCE Phoenician inscription was written in cursive script on both sides of a discarded late Iron Age II sherd. The inscription, as published by Dothan (1985), was apparently an order by a local authority or governor to Akko’s metal-workers’ guild regarding the donation of metal vessels to a temple official. Our documentation of extensive metal working in Area A and Dothan’s discovery (unpublished) of a large public structure, identified as a possible temple, adds to the archaeological evidence indicating the close connection between Akko’s role as a Phoenician maritime industrial and commercial center, and as a Phoenician cultic site linked to the metal working craft.

Survey In an attempt to better understand the settlement patterns and site formation of Tel Akko, a test pitting survey was initiated in 2011 (fig. 3). Our approach entails the excavation of numerous 0.4 m x 0.4 m x 0.4 m test pits laid out on a grid in order to examine the occupation history beneath the topsoil across the site. Thus far, a total of 997 test pits have been dug. The 3D locations of each test pit, including the all-important elevation, were incorporated into a GIS system; the pottery was analyzed and the broad chronology of each test pit was incorporated into the GIS, allowing the analysis of the spatial, stratigraphic and chronological patterns of different surface deposits. Our initial study focused on the southern section of the site and its unusual crescent shape. In the past it had been considered that this area may have been formed as a result of extensive extraction and quarrying by the British during the Mandate period to backfill the swamps associated with the Naaman River. However, an analysis of historical photos and maps of the tel shows that the irregular crescent shape predates the 18th century. The current shape is

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Total Archaeology@Tel Akko (The 2013 and 2014 Seasons)

Fig. 3. Detailed Plan of the Intensive Pedestrian Test Pit Survey: Seasons 2011–2014. Graphics by J. Skinner and R. Stidsing. Tel Akko Total Archaeology Project.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

495

496

Killebrew – Quartermaine

not due to modern damage to the site during the past few centuries but is likely to reflect its form in antiquity (Artzy/Quartermaine 2014). The 2011, 2012 and 2013 test pit surveys in the western area of the south-central section of the tell (figs. 2–3) revealed that the deposits got progressively earlier in date as the pits descended down the steep slope. As one would expect, Hellenistic and Persian pottery was recovered in the uppermost test pits. As the test pits descended down the slope, Iron, Late Bronze, Middle Bronze and Early Bronze artifacts were recovered, indicating that the section edge had been truncated exposing the horizontal stratigraphy. During the 2014 season, however, test pitting in the area further to the east (fig. 3), uncovered clean Persian deposits extending from top to bottom, descending down the steep slope and across the base of the lower southernmost section of the tel. It also identified a section of a rampart against the edge of the higher tell. The implication is enormous as it would indicate that, what we had previously thought of as a disturbed landscape, was in fact the intact, cohesive Persian-period landscape of the lower southeastern section of the tell. The paucity of Hellenistic-period pottery suggests this section of the site was not inhabited during the last century of its occupation. Analysis of the extant ground surface in the area could therefore reveal the function and form of this lower eastern part of the tel. It may help us better understand what we now perceive as a Persian-period landscape and part of what may have formed a lower settlement area of the tel.

Community Outreach As an ancient mound associated with the Canaanites and Phoenicians, and today a municipal park with a modern statue of Napoleon, Tel Akko lacks any clear connection to contemporary local populations. Our community outreach program builds on this site’s non-contested «neutrality». Beginning in 2013, in partnership with the Society of Humanitarian Archaeological Research and Exploration (SHARE), directed by Dana DePietro, the Israel Antiquities Authority’s International Conservation Center, Città di Roma in Akko’s Old City, under the leadership of Shelley-Anne Peleg, and various local community groups, we instituted a heritage project whose goal is the promotion of dialogue and understanding between local Arab and Jewish youth ages 14 to 16 by means of a collective exploration of Akko’s shared cultural heritage (fig. 4). Because archaeological work is highly physical, cooperative, and goal-oriented, it encourages and facilitates the creation of meaningful and lasting dialogue and relationships between individuals that seldom interact. In addition to the experience of working with our team and students on the tell, the young people also participate in an archaeological conservation program based in the UNESCO World Heritage Site of Akko’s Old City, conducted in coordination with the International Conservation

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Total Archaeology@Tel Akko (The 2013 and 2014 Seasons)

497

Fig. 4. Youth from Akko and Society of Humanitarian Archaeological Research and Exploration team members at Tel Akko, 2014.

Centre. As these heterogeneous groups come together through archaeology, they receive not only a hands-on introduction to field methodology, but also a sense of the diversity of historical perspectives and periods. This counteracts the common perception on both sides that archaeology only serves to legitimize each side’s conflicting claims to the land. Our most recent project, exploring the impact of the designation World Heritage Site on the local communities, business and tourism, commenced in 2014. The project includes participant observations of visitors at a variety of tourist sites including Crusader city, Treasures in the Wall museum, and the British Mandate period Museum of Underground Prisoners. Faculty and graduate students conducted extensive interviews with tourists, local business owners, community leaders and residents to get their perspectives on topics ranging from their interest in the history of the city to local economics and tourism. With over a million visitors per year, and as the Old City grows in prominence as a major tourist destination, this research has the potential to contribute to sustainable models of heritage management and development, while preserving the character and population make-up of the Old City.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

498

Killebrew – Quartermaine

Landscape Archaeology in 3D New innovative breakthroughs in 3D technology are revolutionizing archaeological recording, analysis, presentation, and digital preservation. In 2011, we began to develop and implement an inexpensive, easy-to-use, and very effective documentation system to produce a highly accurate means of daily 3D recording on the artifact, square and field excavation, and the general site scale (Olson et al. 2013; Quartermaine et al. 2014). During the 2013 and 2014 seasons, we expanded our 3D documentation and analysis to include a 100 sq. km region surrounding Tel Akko (fig. 1). Utilizing LiDAR, 3D photogrammetry previously developed for the tell, and GIS, we are investigating the tell in relationship to its immediate surroundings, hinterland and satellite settlements, and agricultural catchment. This approach entails the «marriage» of LiDAR and 3D photogrammetry to identify and record the key features of the landscape, utilizing a GIS system to collate all the pertinent data sets that relate to the topography, communication routes, agricultural fields, drainage patterns and settlement distribution in Plain of Akko over the broad lifetime of the tell. The base tapestry for the survey is the incorporation of 1m resolution LiDAR tiles across a study area of 100 sq. km of the Akko coastal plain centered on the city itself, as well as very detailed geo-referenced aerial orthophotos. Although LiDAR, which is a Digital Terrain Model of the ground surface generated by airborne laser scanning, is not a particularly new technique, until now, LiDAR has seldom been applied to archaeological fieldwork in Israel. Our project is distinguished by its combination of several effective survey techniques including ground survey/ground truthing, orthophotos and 3D photogrammetry, which is then overlaid with the results of previous studies into the archaeology of the plain, notably the most recent surveys by Gunnar Lehmann in the Akko Plain during the early 1990s (see e.g. Lehmann/Peilstöcker 2012 and bibliography there). The final stage of the process incorporates a detailed 3D photographic record using a UAV (Unmanned Aerial Vehicle), which generates a 3D model of each identified feature or site, a process that takes minutes to implement, and provides a particularly effective and precise record of the archaeological landscape. The resulting photogrammetric models of the sites are incorporated and geo-referenced into the GIS alongside database descriptions and outline chronology from surface ceramics. An unexpected benefit of our use of LiDAR has been to refine and correct coordinate readings and locations of previously surveyed sites that used maps to estimate coordinates, which lack the precision of the LiDAR and rectified orthophotos. A preliminary study to check the viability of the use of LiDAR and orthophoto techniques in Akko’s hinterland was undertaken during the 2014 season. It entailed a desk-top analysis of LiDAR data and orthophotos to identify potential sites, followed by very limited ground truthing of several of the computer-identified sites. This approach proved to be especially successful in locating archaeological features and sites across the upland areas or other

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Total Archaeology@Tel Akko (The 2013 and 2014 Seasons)

499

Fig. 5a. LiDAR DEM model of Horbat Turit. Tel Akko Total Archaeology Project.

Fig. 5b. Othophotograph of Horbat Turit. Tel Akko Total Archaeology Project.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

500

Killebrew – Quartermaine

regions that had not been subjected to intensive cultivation within the study area. At Horbat Turit, surface features such as walled structures, a rock-cut tomb and a building complex, subsequently confirmed by ground truthing, were clearly visible beneath dense woodland canopy via LiDAR and the orthophotgraph (fig. 5a–b; see Abu Raya 2014 regarding recent salvage excavation in the area). In the intensively cultivated flat lands, the surface features highlighted by LiDAR tended to be modern or relatively recent, thus of lesser significance. However, based on our preliminary study in 2014, our examination of the orthophotos noted areas of soil marks that apparently reflect sub-surface sites or features that have been disturbed by the plough, resulting in soil differences on the surface. This may hold potential for the identification of new sites, which will be confirmed and documented via pedestrian survey. In future seasons, the Tel Akko Total Archaeology Project will continue to pursue its goals of integrating Akko’s past, present and future. Excavation, which is by definition destructive, is only one of many components of our work. No less important is the documentation of the visible surface remains on the tell and its surrounding environment through intensive test pits, pedestrian survey, geophysical and topographic surveys, environmental investigations, and 3D documentation and analysis. While the Plain of Akko survey is still in its infancy, it has demonstrated the potential of a methodological approach that integrates ground truthing, LiDAR for high resolution terrain/topography, GIS for mapping, and photogrammetry for 3D imaging. We hope to be able to produce a precise and hyper-detailed record of the environment of the Plain of Akko that will provide invaluable information for landscape interpretation of Akko and its hinterland through the ages, and will help to recreate the landscape prior to the substantial human interventions of the last century. Akko’s present and future is addressed by our community outreach program, conservation training course, and our ongoing dialogue with the relevant shareholders in Akko’s future on the prospective public interpretation and presentation of Tel Akko.

* The 2013 and 2014 excavations were conducted on behalf of the University of Haifa, in collaboration with the Pennsylvania State University, Claremont Graduate University, Claremont McKenna College, Trinity College (Hartford, CT), the University of Massachusetts at Amherst and the Israel Antiquities Authority. The excavation seasons were carried out under the auspices of the Israel Exploration Society. The excavations were directed by Professors Ann E. Killebrew (Pennsylvania State University) and Michal Artzy (University of Haifa), with the assistance of Professors Tammi Schneider (head of the Claremont consortia and education program director), Martha Risser (project ceramicist and head of the Trinity College, Hartford, Connecticut expedition), Gary Gilbert (square supervisor; Claremont McKenna College), Michael Sugerman (University of Massachusetts at Amherst), and Jolanta Mlynarczyk (project ceramicist; University of Warsaw). Nicholas Pumphrey (Claremont Graduate University), Mikael Bruun-Schmidt (2013; University of Copenhagen), Amani Abu Hmed (Israel Antiquities Authority and the University of Haifa) and Brady Liss (2014; University of California at San Diego) served as Area A field supervisors. Assistant square and square supervisors included Brady Liss (2013), Zachary Devlin, Amanda Pumphrey (Claremont Graduate University), Matthew Hunt (Penn State University), Eliza Wallace (2013; Boston University), Monica Genuardi (2014; Penn State), Hande Köpürlüoğlu (2014; Bilkent

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Total Archaeology@Tel Akko (The 2013 and 2014 Seasons)

501

University) and Jane Skinner (2013; Yale University). Justin Lev-Tov (Statistical Research Inc.) and Melissa Rosenzweig (University of Chicago) were the project’s zooarchaeologist and archaeobotanist respectively. Evan Taylor (2013; University of Massachusetts at Amherst), Emma Kellman (2013) and Mattew Hunt (2014) served as the on-site registrars for the survey and dig respectively. The site surveyor was Jane Skinner (Yale University). Jamie Quartermaine (Oxford Archaeology North, Great Britain) and Matthew Howland (2013; University of California at San Diego) supervised the 3-D digital recording system. Dana DePietro (Society of Humanitarian Archaeological Research and Exploration) directed the Akko youth community outreach program. The geomorphological and geophysical survey team included Christophe Morhange Matthew Giaime, Matteo Vacchi, Alexander Baralis (Aix en Provence University, Geoarchaeological Studies), David Kaniewski Toulouse University, Palynology), Dov Zvieli (Recanati Institute for Maritime Studies at the University of Haifa, Geologist), Idan Shaked, Yossi Salmon (Hatter Laboratory, landscape mapping project), Gloria Lopez (Hatter Laboratory, Recanati Institute for Maritime Studies at the University of Haifa, Optically Stimulated Luminescence). Staff from the Israel Antiquities Authority conservation unit, including Ra‘anan Kislev (Director of the IAA Conservation Department) and Shelley-Anne Peleg (Director of the International Conservation Center) advised and participated in the educational component of the conservation program. Additional staff included Rachel Ben-Dov (pottery and artefact registrar; independent scholar), Rachel Merhav (volunteer coordinator/camp administrator), Yossi Salmon (computer consultant and database designer; Hatter Laboratory, University of Haifa), Ragna Stidsing (database designer, database maintenance and data entry; Hatter Laboratory, University of Haifa), Roee Shafir (conservator/restorer; University of Haifa), Yachin Yankelevich (2013; majordomo/ administrator) and Oren Gutfeld (2014; majordomo/administrator), assisted by Ahiad Ovadia (2014; Hebrew University of Jerusalem). Grey Gundaker (2013; College of William and Mary), Rachel Pollak (2013; University of Haifa) and Ragna Stigsing (2014: University of Haifa) served as the artifact draftspersons. Special thanks are due to Shelley-Anne Peleg (Israel Antiquities Authority), director of the Akko International Conservation Center, who provided invaluable assistance during the planning stages and implementation of the education program. We express our appreciation to Ra‘anan Kislev, Dror Barshad, Eliezer Stern, Hanaa Abu Uqsa, Kamil Sari, and Danny Syon, all of the Israel Antiquities Authority who provided much appreciated advice and assistance for the Total Archaeology@Tel Akko Project and its implementation. Participating in the excavation were students from the Pennsylvania State University, Claremont Graduate University, Claremont McKenna College, Trinity College and the University of Massachusetts at Amherst. Invaluable assistance, infrastructure and professional services were provided by the Hatter Laboratory, Recanati Institute for Maritime Studies at the University of Haifa, directed by Professor Michal Artzy. Special thanks are due to Nira Karmon of the Recanati Institute for Maritime Studies at the University of Haifa for her administrative assistance. We also express our gratitude and appreciation to Ragna Stidsing, Hatter Laboratory Recanati Institute for Maritime Studies at the University of Haifa) for her significant contribution to the graphic component of this article. Lastly, sincere thanks are also extended to the Akko Nautical Academy, in particular to Yael Ban Cohen (Director, Akko Nautical Academy), Reuven Wagner (former Chief Engineer, Akko Nautical Academy: retired) and to their excellent staff, for the exceptional accommodations, network and computer support, and hospitality extended to the Total Archaeology@Tel Akko expedition.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

502

Killebrew – Quartermaine

Bibliography Abu Raya, R., 2014. Horbat Turit (South). Hadashot Arkheologiyot Excavations and Surveys in Israel 126. (http://www.hadashot-esi.org.il/Report_Detail_Eng.aspx?id=11634. Accessed February 28, 2015). Artzy, M., 2012. Return to Tel Akko, its Anchorages, Harbor, and Surroundings. The Leon Recanati Institute for Maritime Studies News, 5–14. Artzy, M./Beeri, R., 2010. Tel Akko. A. E. Killebrew/V. Raz-Romeo (eds.), One Thousand Nights and Days: Akko through the Ages. Haifa. Artzy, M./Quartermaine, J., 2014. How and When Did Tel Akko Get its Unusual Banana Shape? Y. Galanakis et al. (eds.), Critical Essays on the Archaeology of the Eastern Mediterranean in Honour of E. Susan Sherratt. Oxford, 11–22. Dothan, M., 1976. Akko: Interim Excavation Report First Season, 1973/4. Bulletin of the American Schools of Oriental Research 224, 1–48. –

1985. A Phoenician Inscription from ‘Akko. Israel Exploration Journal 35, 81–94.

–

1993. Tel Akko. E. Stern (ed.), The New Encyclopedia of Archaeological Excavations in the Holy Land 1, 17–23.

Kaniewski, D./Van Campol, E./Morhange, C./Guiot, J./Zvieli, D./Shaked, I./Otto, T./Artzy, M., 2013. Early Urban Impact on Mediterranean Coastal Environments. Nature Scientific Reports 3: 2540 (DOI:10.1038/srep03540), 1–5. Killebrew, A. E./Olson, B. R., 2014. The Tel Akko Total Archaeology Project: New Frontiers in the Excavation and 3D Documentation of the Past. P. Bieliński et al. (eds.), Proceedings of the 8th International Congress on the Archaeology of the Ancient Near East, 30 April–4 May 2012, University of Warsaw. Vol. 2: Excavation and Progress Reports – Posters. Wiesbaden, 559–574. Lehmann, G./Peilstöcker, M., 2012. Map of Ahihud (20). Archaeological Survey of Israel. Jerusalem. Online at (Hebrew): http://www.antiquities.org.il/survey/newmap.asp#zoom=8.0000;xy:3874867.2872378, 3671517.394016;mapname=20; (English) http://www.antiquities.org.il/survey/newmap_en.asp#zoo m=8.0000;xy:34.80852508545,31.298049926757;mapname=20. Accessed February 28, 2015. Olson, B. R./Placchetti, R. A./Quartermaine, J./Killebrew, A. E., 2013. The Tel Akko Total Archaeology Project (Akko, Israel): Assessing the Suitability of Multi-Scale 3D Field Recording in Archaeology. Journal of Field Archaeology 38.3, 244–262. Quartermaine, J./Olson, B. R./Killebrew, A. E., 2014. Image-Based Modeling Approaches to 2D and 3D Drafting in Archaeology at Tel Akko and Qasrin. Journal of Eastern Mediterranean Archaeology and Heritage Studies 2.2, 110–127. Ann E. Killebrew, The Pennsylvania State University, Classics and Ancient Mediterranean Studies, Jewish Studies and Anthropology, 108 Weaver Building, University Park, PA 16802, USA, [email protected]. Jamie Quartermaine, Oxford Archaeology North Mill 3, Moor Lane Mills, Moor Lane, Lancaster LA1 1QD, Great Britain, [email protected].

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 503–513

Florine Marchand

L’industrie lithique de Tell ‘Arqa (Plaine du Akkar, Liban Nord) Tell Arqa, a prominent archaeological site in North Lebanon, has a precisely documented stratigraphical record spanning the 3rd and 2nd millennia. A rich lithic industry, from EBA contexts, includes thousands of artifacts. We present our research focused on three main topics: technological study, traceology and experimentation of the débitage technique.

1. Le site de Tell ‘Arqa Tell ‘Arqa est un site archéologique majeur du Liban-Nord, occupé du Néolithique à l’époque mamelouke. Situé en bordure méridionale de la plaine du Akkar, le tell de 7 hectares de superficie et 40 mètres de haut contrôle la partie sud d’une des rares plaines côtières à fort potentiel agricole du littoral syro-libanais. La plaine du Akkar forme le débouché maritime de la «trouée de Homs»; cette région est l’une des principales voies de communication stratégique et commerciale à travers la chaîne montagneuse qui isole la côte méditerranéenne de la Syrie intérieure proche et du monde mésopotamien. L’exploration archéologique du site fut entreprise dès 1972 et confiée par le directeur des Antiquités d’alors, l’Émir M. Chéhab, à l’Institut français de Beyrouth sous la direction du prof. E. Will. Après la guerre civile libanaise, la fouille fut reprise et étendue par le Dr. J.-P. Thalmann, directeur de la mission à partir de 1995.1 L’objectif des nouvelles recherches était d’assurer la stratigraphie et la chronologie de la fin du Bronze Ancien et du Bronze Moyen, qui sont parmi les périodes les plus importantes et prospères dans le développement du site

1

Depuis l’origine, la mission d’Arqa est financée par le Ministère français des Affaires étrangères, et entretient une collaboration étroite et amicale avec la Direction libanaise des Antiquités; ces deux institutions sont ici chaleureusement remerciées pour leur soutien constant à une entreprise qui, à plusieurs reprises et encore récemment, a connu des moments difficiles en raison de la situation locale.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

504

Marchand

Fig. 1. Tableau résumant les correspondances entre les niveaux, les dates absolues, le phasage culturel local, la périodisation ARCANE et la périodisation traditionnelle.

et de sa région. De 2005 à 2012, dans le cadre du projet européen ARCANE2, une opération spécifique fut consacrée aux niveaux de la première moitié du 3ème millénaire. Enfin, en 2010, fut entrepris un projet pluridisciplinaire visant à restituer, sur la base de l’ensemble de la documentation disponible (architecture, céramique, botanique, zoologie, lithique, etc.), l’évolution technique, économique et sociale de la plaine du Akkar aux 3ème et 2ème millénaires. Notre recherche sur les industries lithiques d’Arqa s’inscrit dans le cadre de ce projet. La stratigraphie d’Arqa est organisée par niveaux numérotés de haut en bas en chiffres arabes; chaque niveau est subdivisé en couches, désignées par une lettre majuscule additionnelle (Thalmann 2006a: 7–16). Le Bronze Ancien correspond aux niveaux 20 à 15, le Bronze Moyen et Récent aux niveaux 14 à 11.3 Les vestiges dégagés sont, à chaque niveau, des structures d’habitat. La chronologie absolue de la séquence du Bronze Ancien repose sur une quarantaine de déterminations radiocarbone (Thalmann et al. à paraître; Thalmann 2013: fig.1). 4 Elle sert de référence pour la nouvelle périodisation ARCANE du 3ème millénaire sur la côte syro-libanaise (Central Levant): le tableau (fig. 1) résume les correspondances entre les niveaux et dates absolues, le phasage culturel local (Thalmann 2006a: 13–16, 92–95), la nouvelle périodisation ARCANE et la périodisation traditionnelle.

2 3 4

Ce projet, financé par l’European Science Foundation, a fait l’objet de plusieurs présentations: dans le premier volume de la série ARCANE (Lebeau 2011) et sur le site www.arcane.uni-tuebingen.de. Les niveaux de la seconde moitié du 3ème et du 2ème millénaire ont été publiés (Thalmann 2006a). Les niveaux anciens n’ont fait l’objet que de rapports préliminaires (Thalmann 2009, 2010, 2013, à paraître). Les dates données sont issues d’une sériation préliminaire à l’aide du logiciel OxCal v. 4.1.7.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

L’industrie lithique de Tell ‘Arqa

505

2. Objectifs de notre recherche Notre recherche s’articule autour de cinq axes principaux: (1) l’étude technologique nous permet de reconstruire les chaînes opératoires propres au débitage des lames et des éclats. Nous étudions le matériel lithique associé aux lames dites «cananéennes», afin de mieux connaître l’économie de débitage et la replacer dans le contexte économique et technique d’un site de l’âge de Bronze. Il sera important, dans ce cas, de définir si plusieurs schémas opératoires coexistent; (2) Nous utilisons les méthodes de l’archéologie expérimentale5 afin de reproduire le débitage par pression au levier et constituer une banque de données de référence que nous comparerons aux matériels archéologiques; (3) Nous constituons une typologie, propre au site, et nous souhaitons définir la provenance de la matière première puisque, dans les autres sites connus, pour leur lames «cananéennes», elle s’avère être exogène (Chabot 1998: 167); (4) Notre étude tracéologique prolonge celle menée par P. Anderson, afin de déterminer si les lames «cananéennes» de Tell Arqa ont la fonction d’armature de tribulum. Pour le reste des outillages lithiques, nous réalisons une analyse tracéologique afin de préciser la place qu’avaient les outils en pierre dans l’économie du site; (5) Nous comparons les données obtenues à partir de nos axes de recherche avec les données préexistantes de la côte levantine, principalement, les sites de Byblos (Cauvin 1969), de Tell Fadaous-Kfarabida (K. Badreshany et al. 2005)6 et de Sidon (Yazbeck 2006), …) mais aussi du Levant Nord comme Tell Atij (Chabot 1998), Tell Gudeda (Chabot 1998), Kutan (Anderson et Inizian 1994), Amuq (Braidwood et Braidwood 1960) ou encore Megiddo (Blockman/ Groman-Yeroslavski 2006; Gersht 2006) au Levant Sud.

3. Question de terminologie Depuis de nombreuses années, les grandes lames en silex sont désignées par le terme ambigu de «lames cananéennes» suite à leur découverte et description dans les années 30 par R. Neuville en Palestine (Neuville 1930). Les lames «cananéennes» sont toujours perçues comme de grandes lames à sections trapézoïdales dont les nervures sont parallèles aux bord avec, lorsqu’il est présent, un talon punctiforme. Cette première définition, comprenant peu de critères distinctifs, a engendré la problématique suivante: les lames de grandes dimensions découvertes sur les sites de l’âge du Bronze sont le plus souvent qualifiées de

5

6

Nous faisons partie du groupe d’archéologie expérimentale de taille de silex de Jales (Université Lyon 2-CNRS) dirigée par le Dr. F. Abbès. Un programme d’expérimentation de production de grandes lames est actuellement en cours. Communication personnelle du Dr. H. Genz, AUB, Beyrouth.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

506

Marchand

«cananéennes», quelle que soit leur technique de débitage. Plus tard, suite aux études de J. Pelegrin (Anderson et Inizian 1994; Chabot 1998), dans les années 90, l’appellation de «débitage cananéen» a été créé pour désigner plus spécifiquement la technique de débitage par pression au levier. Depuis, aucun autre système de débitage n’est envisagé et cette technique est, elle aussi, systématiquement associés aux grandes lames sans exceptions. Suite à nos observations, issues de l’étude de la collection d’artefacts lithiques de Tell ‘Arqa, nous avons pris le parti d’utiliser une terminologie que nous voulons mieux adaptée et fondée sur des critères distinctifs. Nous désignerons les lames de grandes dimensions sous le terme générique de «macro-lames», en y ajoutant un second critère: «macro-lame (de facture) locale» ou «macro-lame d’importation». Cette distinction est fondée sur la matière première et sur la présence ou non des produits de débitage sur le site (cf. infra).

4. La technologie lithique La collection d’outillage lithique comporte environ 3000 pièces en silex et une centaine en obsidienne. A côté de plusieurs types de silex de couleurs et de textures variées, la matière première majoritaire et locale est un silex de couleur brun se poursuivant par du violacé ou du rouge brique vers le cortex, à grains moyens ou fins, avec des inclusions en forme de taches et points gris caractéristiques. L’abondance de ce type de silex, ainsi que la présence des produits secondaires et des déchets propre au débitage permet de distinguer une production laminaire locale d’une production laminaire plus vraisemblablement importée. L’origine de cette matière première «locale» au sens large, n’est pas actuellement déterminable à cause des conditions sécuritaires qui interdisent toute prospection régionale. Plusieurs hypothèses sont envisagées tels des gisements de silex remarqués par J.J. Ibañez7 dans la vallée de la Bekaa.

4.1. Les différents types de produits sur silex La production locale d’Arqa comporte des nucléus, des lames à crêtes, des éclats, des lamelles et des lames. Les nucléus sont dans une phase d’exhaustion complète. Ils sont de formes sphéroïdales et revêtent des négatifs de petits éclats et de lamelles ainsi que, dans certains cas, des traces d’impacts de percussion directe. Les éclats corticaux ou non-corticaux sont issus du dégrossissage ou du façonnage du nucléus et montrent des stigmates classiques d’un débitage en percussion directe. Les lamelles en silex sont extrêmement rares.

7

Communication personnelle du Dr. J.-J. Ibañez.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

L’industrie lithique de Tell ‘Arqa

Fig. 2. Lame de facture locale, pièce n°08/272.001.

507

Fig. 3. Lame importée, pièce n°96/274.001.

Fig. 4. Lame denticulée, pièce n°12/911.001.

4.2. La production laminaire de Tell ‘Arqa

Les produits laminaires locaux A Tell Arqa, les produits laminaires locaux sont présents dès le Bronze ancien I (Niveau 20/19) jusqu’au Bronze récent. Une production de macro-lames unipolaires de facture locale (fig. 2) a été repérée, à partir du niveau 17 (2700–2500 Cal B.C.), suite aux considérations suivantes: leur matière première est similaire aux déchets de débitage, impliquant un débitage direct sur le site; leurs stigmates techniques, leur longueur qui dépasse difficilement une quinzaine de centimètres pour une largeur maximum de 3 centimètres, les talons lisses avec des bulbes plus proéminents, les sections majoritairement trapézoïdales, les nervures ondulant fortement, tout comme les négatifs d’enlèvements, indiquent un débitage par percussion directe; cette technique ne correspond pas au débitage des macro-lames d’importation (cf. infra). Parallèlement aux macro-lames de facture locale, nous signalons la présence de deux autres types laminaires: le premier type comprend des lames unipolaires étroites à sections peu épaisses trapézoïdales, toujours utilisées pour un type d’outil particulier, les lames

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

508

Marchand

Fig. 5. Pointe de projectile, pièce n°09/544.001.

denticulées (fig. 4). Dans ce cas-ci, les extrémités sont toujours tronquées ou cassées, ne donnant que des informations partielles sur la technique de débitage; le second type regroupe des lames épaisses (0,9–1 cm) unipolaires de sections strictement triangulaires dédiées à un type d’outillage particulier, les projectiles (fig. 5). Dans ce cas, les retouches envahissantes et fortement couvrantes ne donnent que peu, voire pas, d’informations technologiques.

Les macro-lames d’importation Les macro-lames d’importation sont associées à la fin du 4ème millénaire et au début du Bronze ancien. Dans le cas de Tell ‘Arqa, ce type de lames est retrouvé dans des contextes archéologiques bien définis et strictement contrôlés: les niveaux 16 et 15 exclusivement (Bronze ancien IV, entre 2500 et 2000 Cal. B.C.), datation plus récente que sur les autres sites ayant livré ce type de produits. Ces lames sont issues d’un débitage unipolaire. Leurs matières premières ont été rapidement considérées comme exogènes pour diverses raisons: (i) la qualité et la diversité des matériaux utilisés ne correspond pas au silex considéré comme «local»; (ii) aucun déchet de débitage n’est présent, nous possédons seulement des lames entières ou des tronçons de lames, des produits finis. La facture de ces lames est bien caractérisée (fig. 3): une longueur impressionnante (pouvant aller jusqu’à 25cm), une largeur comprise entre 3 et 4 cm, un talon punctiforme ou dièdre avec une quasi absence de bulbe, la rectitude extrême des arrêtes totalement parallèles aux bords, une section généralement trapézoïdale (des sections triangulaires étant aussi connues). Tous ces stigmates techniques ne se retrouvent pas dans le débitage local. La technique la plus vraisemblablement utilisée,

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

L’industrie lithique de Tell ‘Arqa

509

dans ce type de production, est la pression au levier, technique qui n’a pas été utilisée dans le cadre des autres productions du site.

4.3. Les différents types de produits sur obsidienne Une centaine de pièces en obsidienne proviennent des niveaux du Bronze ancien sur le tell: des lamelles, des petits/moyens éclats et quelques lames. Plusieurs types de matière première ont été distingués: de l’obsidienne gris clair, anthracite, bleue, verte (fumée) et grise à veines blanches. La plupart de ces types proviennent d’Anatolie centrale (Gölü Dag-Est et Nenezi Dag) (Coqueugniot 2006; Thalmann 2006b) tandis que d’autres plus rares sont issus de l’Est de l’Anatolie (Bingöl A-Nemrut Dag et Bingöl B) (Thalmann 2006b). Le débitage, dans le cas de l’obsidienne, est bipolaire. Il s’agit d’une méthode néolithique, hors contexte dans des niveaux du 3ème millénaire. De fait, une «aire de débitage» de silex et d’obsidienne a été découverte au pied du tell (1 km à l’ouest), où une occupation néolithique a été reconnue suite au ramassage d’artefacts comprenant environ 400 fragments d’obsidienne ainsi que des blocs de grandes dimensions. Le Bloc exposé au Musée de Beyrouth (22 kg) montre des zones corticales et de grands enlèvements d’éclats. Nous n’avons pas de datation précise pour ce bloc mais il indique qu’à une certaine période la matière première arrivait sous forme brute. Nous considérons, dans le cas de Tell ‘Arqa, que les produits en obsidienne retrouvés sur le tell ne résultent pas d’un débitage sur obsidienne ayant eu cours au 3ème millénaire mais plutôt d’une récupération sur l’aire de débitage de produits manufacturés durant le Néolithique. Un cas particulier renforce cette hypothèse: un fragment mésial de lame retouchée et lustrée a été perforée à ses deux extrémités pour en faire un élément de parure. Ces constatations expliquent la présence du débitage bipolaire mais aussi la faible taille des produits durant le 3ème millénaire à Tell ‘Arqa: nous envisageons que les grands produits débités plus anciennement ont été retouchés/retravaillés/réutilisés engendrant la présence de petits produits comme les lamelles, les petits éclats et les esquilles.

5. Typo-chronologie des outils en silex Nous signalons la présence d’un panel d’outillage «classique» tels les grattoirs, les racloirs, les perçoirs/forets ou encore des haches aux travers des différents niveaux. Les niveaux 20/19 (3000–2800 Cal. B.C.) se caractérisent pas la présence de fragments de racloirs en éventails (Fan Scrapers) qui sont typiquement associés à la fin du Chalcolithique et au début du Bronze ancien mais qui, contrairement aux autres cas connus, ne sont pas associés aux macro-lames.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

510

Marchand

Les niveaux 18/17 (2800–2500 Cal. B.C.) sont définis pas la présence de lames denticulées et de pointes de projectiles. Les lames denticulées (fig. 3), à grosses retouches directes (dents), sont, généralement, associées à des périodes plus anciennes comme c’est le cas au Levant Sud (Rosen 1997); à Megiddo par exemple, dans un assemblage du Bronze ancien I–II, elles sont éliminées sur des critères typologiques, sans discussion sur l’intégrité des contextes dont elles proviennent (Blockman/Groman-Yeroslavski, 2005). Toutefois, nous insistons sur le fait que ce type d’outil est représenté, durant le Bronze ancien, sur d’autres sites libanais tels que Byblos (Enéolithique ou Bronze ancien ?, Liban, Cauvin 1968), Mengez (fin Chalcolithique-Bronze ancien, plaine du Akkar, Liban, Ibañez et al. 2010); dans le domaine mésopotamien, il s’agit des outils caractéristiques du Dynastique Archaïque I au début du 3ème millénaire sur des sites comme Larsa (Irak, Coqueugniot 2003) ou Abu Salabikh (Irak, Crowfoot Payne 1980), sites sans occupation antérieure et sans possibilité de pollution par du matériel ancien. Nous n’avons aucune raison de penser que les lames denticulées, à Tell ‘Arqa, ne soient pas en place dans les niveaux dont elles proviennent. Dans le cas des projectiles (fig. 4), leur aspect rappelle fortement les pointes de flèche de l’Amuq pour certaines. J. Cauvin signale déjà ce type de projectile à Byblos (Cauvin 1968). Leurs retouches sont directes et/ou inverses envahissantes et fortement couvrantes, quelques unes d’entre elles sont pédonculées. L’étude de la faune a démontré une quantité importante d’animaux issus de la chasse d’environ 20% pour les niveaux 18/17 (Chahoud 2013). Tell ‘Arqa n’est pas le seul site où des projectiles ont été découverts durant l’Âge du Bronze, nous pensons notamment à Byblos (Liban, Cauvin 1968), à Tell Fadaous-Kfarabida (Liban, Genz 2013)8, à Tell Bazi (Syrie)9 ou encore à Norsun Tepe (Turquie, Schmidt 1996) et à Arslan Tepe (Turquie, Frangipane et al. 1993; Frangipane 2010). Les niveaux 16/15 (2500–2000 Cal. B.C.) sont les niveaux à macro-lames par excellence. C’est durant cette période que la majorité des macro-lames importées sont retrouvées, même si les macro-lames locales sont toujours présentes en moindre quantité comparativement aux phases précédentes et suivantes. Les deux types montrent d’importants lustrages et portent des retouches directes ou indirectes, des troncatures simples, des bi-troncations ou encore des fractures à leurs extrémités. Souvent fragmentaires, nous possédons, tout de même, quelques cas entiers. Pour les périodes plus tardives, seul le niveau 12 présente un type d’outil caractéristique: les segments de lames lustrées trapézoïdaux, déjà remarqués par E. Coqueugniot en 2006 (Coqueugniot 2006).

8 9

Communication personnelle du Dr. H. Genz, AUB, Beyrouth. Communication personnelle de B. Einwag, C. Fink et A. Otto.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

L’industrie lithique de Tell ‘Arqa

511

6. Les analyses tracéologiques Les analyses tracéologiques, pour le 3ème millénaire, sont dédiées à l’identification des tâches agricoles et à l’utilisation, ou non, des macro-lames lustrées en tant qu’élément de tribulum. Un premier jeu d’analyses réalisé par E. Coqueugniot sur un petit échantillonnage a démontré la présence d’éléments de faucille uniquement. Un deuxième jeu d’analyses a été réalisé par nos soins, sous la supervision de J.J. Ibañez, chercheur au CSIC (Consejo Superior de Investigationes Cientificas), Institut Mila y Fontanals de Barcelone, en 2013. Notre objectif était alors de vérifier les observations d’E. Coqueugniot et de nous assurer de la présence ou de l’absence d’éléments de tribulum. Tout comme E. Coqueugniot, les analyses de grandes lames ont toutes montré les caractéristiques des faucilles qu’elles soient de facture locale ou importées. Aucune macro-lame importée ou locale n’a été utilisée comme élément de tribulum. Généralement, au Proche-Orient, l’emmanchement est réalisé grâce au bitume. Au Liban, le bitume n’est pas présent naturellement, un autre type de ‹colle› a donc été utilisé: la présence de résidus organiques a été observée sur de nombreuses lames.

7. Conclusion L’industrie lithique de Tell ‘Arqa est l’une des rares collections d’artefacts lithiques stratifiées pour l’Âge du Bronze sur la côte levantine. Son caractère paraît à première vue atypique, de par la présence de denticulées, de pointes de flèches, d’une industrie de grandes lames locales, des macro-lames importées tardives, de l’absence de tribulum. Elle révèle surtout les lacunes de notre connaissance au sujet d’un matériel trop souvent négligé et les particularismes régionaux dans ses diverses utilisations, nous offrant une vision jusqu’alors méconnue des industries lithiques de l’Âge du Bronze au Liban et, de manière plus globale, au Levant Central. Quelques points de notre recherche doivent être impérativement approfondis comme les différentes chaînes opératoires, l’expérimentation du débitage des macro-lames (en cours) ou encore les analyses tracéologiques des outils non-lustrés. Nous souhaitons que notre étude aboutisse à une meilleure connaissance du rôle de la taille du silex, de ses statuts et fonctions sur les sites de l’âge du Bronze au Levant.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

512

Marchand

Bibliographie Anderson, P./Inizian, M.-L., 1994. Utilisation du tribulum au début du IIIe millénaire: des lames «cananéennes» lustrées à Kutan (Ninive 5) dans la région de Mossoul (Iraq). Paléorient 20/2, 83–103. Badreshany, K./Genz, H./Sader, H., 2005. An Early Bronze Age Site on the Lebanese Coast, Tell FadaousKfarabida 2004 and 2005: Final Report. Bulletin d’archéologie et d’architecture Libanaises (BAAL), Vol.9. Beyrouth Blockman, N./Groman-Yeroslavski, I., 2006. The Early Bronze Age Flint Assemblage. I. Finkelstein et al., Megiddo IV: The 1998–2002 Seasons, Monograph Series No. 24, The Emery and Claire Yass Publications in Archaeology of Tel Aviv University. Tel Aviv, 315–342. Braidwood, R. J./Braidwood, L. S., 1960. Excavations in the Plain of Antioch, I. The Earlier Assemblages, Phases A-J, Oriental Institute Publications 61. Chicago. Cauvin, J., 1968. Les outillages néolithiques de Byblos et du littoral libanais, Fouilles de Byblos, tome IV. Paris. Chabot, J., 1998. Etude des artefacts en pierre taillé découverts à Tell ‘Atij et Tell Gudeda en Mésopotamie Septentrionale (Syrie du Nord-Est, Age du Bronze Ancien: 3000–2500 av. J.-C.). Thèse de la Faculté des Lettres, Université de Laval, Québec, Canada. Chahoud, J., 2013. Diversité faunique, économie alimentaire et pratiques socio-culturelles au Levant à l’Âge du Bronze: Une approche archéozoologique, Thèse, Université Lumière Lyon 2. Lyon. Coqueugniot, E., 2003. Les outils de pierre taillée de Larsa 1989 (IIIe et IIe millénaire av. J.-C.). J.-L. Huot (éd.), Larsa travaux de 1987 et 1989, BAH 165. Beyrouth, 385–412. –

2006. Les outillages en pierre taillée et la question des lames «cananéennes»: étude préliminaire. J.-P. Thalmann, Tell Arqa-1, Les niveaux de l’âge du Bronze, BAH 177. Beyrouth, 195–202.

Crowfoot Payne, J., 1980. An Early Dynastic III Flint Industry from Abu Salabikh. Iraq 42.2, 105–119. Frangipane, M./Hauptmann, H./Liverani, M./Matthiae, P./Mellink, M. (eds.), 1993. Between the River and Over the Mountains: Archaeologica Anatolica et Mesopotamica Alba Palmieri Dedicata, Dipartimento de Scienza Storiche, Archeologiche e Anthropologiche dell’Antichita. Universita di Roma “La Sapienza”. Roma. Frangipane, M. (ed.), 2010. Economic Centralization in Formative State. The Archaeological Reconstruction of the Economic System in 4th Millennium Arslantepe, Dipartimento de Scienza Storiche, Archeologiche e Anthropologiche dell’Antichità. Studi di Preistoria Orientale, Vol. 3, Università di Roma “La Sapienza”. Roma. Genz, H., 2013. The Introduction of the Light, Horse-Drawn Chariot and the Role of Archery in the Near East at the Transition from the Middle to the Late Bronze Age: Is There a Connection? A. J. Veldmeijer/S. Ikram (éd.), Chasing Chariots, Proceedings of the First International Chariot

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

L’industrie lithique de Tell ‘Arqa

513

Conference (Cairo 2012). Leiden, 95–106. Gersht, D., 2006. The Flint Assemblage of Area K. I. Finkelstein et al., Megiddo IV: The 1998–2002 Seasons, Monograph Series No. 24, The Emery and Claire Yass Publications in Archaeology of Tel Aviv University. Tel Aviv, 343–352. Ibáñez, J.-J./Haïdar-Boustani, M./Al-Maqdissi, M./Armendáriz, A./Urquijo, J. G./Teira, L., 2010. Découverte de nécropoles mégalithiques à l’ouest de Homs. M. Al-Maqdissi et al., Hauran V, La Syrie du Sud du Néolithique à l’Antiquité tardive, Recherches récentes, Actes du Colloque de Damas 2007, BAH 191. Beyrouth, 359–366. Neuville, R., 1930. Notes de préhistoire palestinienne III, Les industries lithiques de l’âge du Bronze. Journal of the Prehistoric Society X, 199–216. Rosen, S., 1997. Lithics After the Stone Age, A Handbook of Stone Tools From the Levant. Walnut Creek/ London. Schmidt, K., 1996. Norsuntepe. Kleinfunde I. Die lithische Industrie, Archaeologica Euphratica 1. Mainz. Thalmann, J.-P., 2006a. Tell Arqa-1, les niveaux de l’âge du Bronze, BAH 177. Beyrouth. –

2006b. Obsidian at Tell Arqa, North Lebanon: a stop-over point on a trade route? Baghdader

–

2009. The Early Bronze Age: Foreign Relations in the Light of Recent Excavations at Tell Arqa.

Mitteilungen 37, 575–592. A.M. Afeiche (ed.), Interconnections in the Eastern Mediterranean – Lebanon in the Bronze and Iron Ages. Proceedings of the International Symposium, Beirut 2008. Bulletin d’Archéologie et d’Architecture Libanaises supplément VI, 15–38. –

2010. Tell Arqa: a Prosperous City during the Bronze Age. Near Eastern Archaeology 73/2, 2–17.

–

2013. Le lion, la chèvre et le poisson. A propos d’une jarre à empreintes de sceaux-cylindres de Tell

–

à paraître. Rapport sur les campagnes de 2008 à 2012 à Tell Arqa. Bulletin d’Archéologie et

Arqa (Liban). Syria 90, 255–312. d’Architecture Libanaises. Thalmann, J.P./Weninger, F./Wild, E., à paraître. A Third and Early Second Millennium Radiocarbon Sequence from Tell Arqa, Lebanon. Yazbeck, C., 2006. The Early Bronze Age Industry of Sidon (“College Site Excavations”). C. DoumetSerhal, The Early Bronze Age in Sidon, “College Site” excavations 1998–2001. BAH 178, Beyrouth, 291–300.

Florine Marchand, PhD student, Université Libre de Bruxelles, Belgium.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Jordan

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9th ICAANE, Basel 2014, Vol. 3, 517–531

Don Boyer

Aqueducts and Birkets: New Evidence of the Water Management System Servicing Gerasa (Jarash), Jordan A preliminary plan has been drawn up of the aqueduct system and sources that supplied the Gerasa area. Evidence is presented of aqueducts approaching the city from the west and north, and of a source recently identified in the eastern part of the city.

Introduction This paper presents some preliminary results of a study into the water management system at Gerasa and in its hinterland in the Classical period (3rd century BC to 7th century AD). Gerasa was one of the cities of the Decapolis, a group of Hellenistic colonies established in the 3rd and 2nd centuries BC in an area centred in modern-day north-western Jordan and adjacent parts of Israel and Syria (fig. 1). Archaeological evidence (Barghouti 1982: 218–228; Braemer 1989: 318; Kehrberg/Manley 2002: 197) supports the founding of a Hellenistic colony (known as Antioch-on-the-Chrysoroas) on the site of another settlement (probably called Gerasa or the Aramaic form of that name) by the middle of the 2nd century BC, though little is known of this colony. Gerasa, along with the other Decapolis cities, flourished in the early centuries following the establishment of Roman rule. The city was directly connected by road to other Decapolis cities and these trade links supported the city’s emergence as a flourishing provincial centre in the early Roman period. An impressive building program implemented in the 1st and 2nd centuries AD saw the establishment of a planned city on both banks of the Wadi Jarash and the construction of a number of major public buildings and monuments.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

518

Boyer

Fig. 1. Plan of the Decapolis. Source: Author.

Study Objectives The study area covers approximately 100 km 2 and includes two valleys that lay in the centre of Gerasa’s territory (fig. 2). The study seeks answers to a number of fundamental questions including the sources of water supply and the nature and dating of the distribution network to and within the city. Gerasa is one of the most intensely studied cities of the Decapolis1 and there have been individual studies on Gerasa’s baths (Lepaon 2008; 2012) and fountains 1

Raja (2012: 137–189) provides a synthesis of previous work.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Aqueducts and Birkets … Water Management System Servicing Gerasa (Jarash)

519

Fig. 2. Plan of the study area showing the location of springs mentioned in the text. Source: Author.

(Seigne 2008). However, there has been no published study of the whole water management system within and – even more important, beyond the city boundary, comparable to the studies on the systems that supplied the Decapolis cities of Gadara – Abila and Adraa.2 A brief assessment by Seigne includes preliminary observations on Gerasa’s water supply in the context of its hinterland (Seigne 2004) and Sapin’s pioneering geoarchaeological survey of an area east of the city provides a brief outline of the region’s ancient rural water management system (Sapin 1998: 112). The present study was inspired by the results of the Jarash Hinterland Survey conducted in 2005, 2008 and 2010 that included records of many

2

For hydraulic studies on Gadara see Kerner et al. 1997; Al-Daire 2004; Döring 2005; for regional hydraulic studies in northern Jordan see Döring 2004; 2008; Kerner 2004.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

520

Boyer

Fig. 3. Plan of Gerasa and main water installations (after Lepaon 2011: fig. 2, with modifications).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Aqueducts and Birkets … Water Management System Servicing Gerasa (Jarash)

521

water installations in the city’s immediate hinterland (Kennedy/Baker 2009; see also Baker/ Kennedy 2010; 2011).

The Water Consuming Installations in the City The main water consuming installations within the city comprise bathing complexes, fountains and a monumental nymphaeum, with most of the installations being located on the west side of the city. As only 15% of the city has been excavated it is likely that many more installations lie hidden beneath the ruins. A total of six bathing complexes have been identified within the ancient city (fig. 3), of which two – the Large East Baths covering 1.5 hectares (Lepaon 2008: 62) and the West Baths covering 0.5 hectares – can be described as thermae bathing and recreational complexes. Many of the water installations in the city date to the 2nd century AD including the two thermae (Lepaon 2008: 65; 2012: 117), a number of street fountains (Seigne 2008: 39–49) and the nymphaeum. After a hiatus the Central Baths and Small East Baths were constructed in the 4th century AD and the Placcus Baths were constructed in the 5th century AD. Undated bathing complexes are also known outside the city at Birketein (Lepaon 2008: 67–68), 1.6 km north of the city, and at the el-Hammam mineral spring 7.0 km to the south (Glueck 1939: 220–221).

The Water Sources Gerasa is centrally located in a valley richly endowed with strong springs of good quality and average annual rainfall of 356 mm (Al-Qaisi 2010, table 11.1.). There is palaeoclimatic proxy evidence of moist climatic conditions from the Hellenistic period to the 6th century AD (Rambeau/Black 2011: fig. 7.2), and it seems likely that climate conditions in this period may have been similar to or wetter than those currently existing. Until recently the combined drainage of Wadi Suf/Wadi ed Deir/Wadi Jarash (hereafter Jarash Valley) carried a perennial stream sustained by seasonal rainfall and surplus flows from at least four highyielding springs within the catchment upstream of Jarash and from Qairawan spring within the city and it is probable that, historically, Tannur spring and other sources sustained a perennial stream in Wadi Tannur. The stronger springs draw on aquifers in the Upper Cretaceous Na’ur Limestone, however there are a number of weaker springs in the lower Jarash Valley that draw from aquifers in the underlying Kurnub Sandstone. Water from the main sources has been harnessed for human use over a long period, and evidence of Roman-Byzantine structures or human settlement has been recorded at or near all the main spring sites. Qairawan spring was monumentalised at some point – possibly in

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

522

Boyer

Fig. 4. Location of spring source in eastern part of city. Source: RAF Aerial photograph AP-1072, courtesy of APAAME.

the 2nd century AD.3 It was clearly an important source but its low elevation (566 m) limited its use to supplying the lower east side of the city and providing irrigation water to fields on both banks of the wadi south of the city. Most of the ancient city lies above the elevation of Qairawan, including most of the western side where many of the water consuming 3

Seigne suggests the second half of the 2nd century AD (2004: 175, note 13).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Aqueducts and Birkets … Water Management System Servicing Gerasa (Jarash)

523

Fig. 5. Preliminary aqueduct plan. Source: Author.

monuments are located, which means that the west side and the upper east side of the city must have been supplied from elsewhere. The general view among scholars has been that Birketein reservoir was the principal water source for the west side of the city (see, for example, McCown 1938: 162; Harding 1967: 87; Freeman 2008: 434) – some specifically identifying it as the source for the nymphaeum and other monuments (see, for example, Rostovtzeff 1932: 83; McCown 1938: 167; Browning 1982: 213). Seigne, however, concluded that the open reservoir at Birketein was an unlikely source of the city’s potable water supplies and suggested instead that the water supply came from another source or sources upstream from Birketein (Seigne 2002: 16; 2004: 176–177).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

524

Boyer

Fig. 6. North-west Aqueduct-stepped chute cascade locality: (a) View from south-west showing specus 40 cm wide cut into limestone outcrop; (b) View down-gradient showing a fracture with vertical and horizontal movement dislocating the specus. Source: Author.

No in-situ elements of a water reticulation system have yet come to light in the upper east side of the city, but the system’s existence is suggested by ceramic water pipe fragments visible in the fill adjoining the Procopius Church (elevation 594 m) in the city’s south-east quarter. Burckhardt referred to a source on the eastern hillside during his visit in 1812 (Burckhardt 1822: 263) but its precise location is unclear. This conundrum has been resolved by the recent recognition of photographic evidence of a spring approximately 90 m west of the east wall of the city. Historic photographs show a water channel draining a 16 m x 5 m reservoir or cistern cut into limestone bedrock. The reservoir is fed from a source that appears to be located immediately above it at an elevation of c. 605 m, however further details must await a field survey (fig. 4). The spring was flowing as recently as the late 1920s.

The Aqueducts It has been possible to define a network of ancient aqueducts approaching Gerasa from the north along both banks of the Jarash Valley using a combination of hard evidence from field survey and photographic interpretation. It will never be possible to fully recover all the details of this network as a result of the loss of evidence through agricultural and building activities, but there is sufficient evidence to define the routes of a number of aqueducts with reasonable confidence (fig. 5). Most of the aqueducts identified are of simple construction with a ‹U› shaped specus cut directly into the bedrock, but there are two examples of important aqueducts constructed of masonry; one (DW01) crosses the floor of the valley north of

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Aqueducts and Birkets … Water Management System Servicing Gerasa (Jarash)

525

b

a

Fig. 7. North-west aqueduct. a. Cross-sectional profile of canal exposed in road cutting. b. Detailed view showing plaster and calcareous sinter accumulation. Source: Author.

the city, and the other is a section of the aqueduct from Qairawan to the Small East Baths on the east bank that was carried on a four metre high frame wall. Rock-cut aqueduct SW02 from esh Shawahid spring north-west of Birketein can be traced along the western edge of the valley floor to the northern city wall. When used in the late Ottoman period this aqueduct crossed the city wall at 603 m elevation, but no ancient crossing point has been identified. Aqueduct DW01 can be traced from a point close to the Birketein reservoir along the west bank of the wadi to the northern city wall; in the late Ottoman period this aqueduct crossed the city wall at 599 m elevation, but no ancient crossing point has been identified. It may have been sourced from Birketein reservoir but this is not certain. Where it crosses the valley floor it was constructed of large masonry blocks, and a number of these blocks were located during the Jarash Hinterland Survey (Kennedy/ Baker 2009: 23). Although displaced they lie close to the position of a 360 m long aqueduct shown in situ on an unpublished 1:2000 scale plan of the valley probably drawn in 1931 and preserved in the Yale University archives (Yale University Art Gallery Negative number: 1938.5999.5004.50). Aqueduct DW05 is based on one short rock-cut section and its route is uncertain, but its elevation points to Birketein being the likely source. Aqueduct SE01 from Bisas er Rum spring, located on the east bank 1.0 km north of Birketein, can be traced for 1.3 km along the eastern slope of the valley. Its destination is uncertain but its elevation and

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

526

Boyer

the absence of cultivable land on this side of the valley suggests that its likely destination was the east side of the city 2.0 km to the south. On the west side of Gerasa a 1.0 km section of rock-cut aqueduct (North-west Aqueduct) has been traced approaching the city’s north-west gate from the north-west, however its exact entry point into the city is not visible. The source has not yet been identified but probably lies between Deir el Lyiat and Suf. The observed section has an average gradient of c. 10% and would have entered the city at c. 630 m elevation. Aqueducts with such steep gradients provided a number of challenges to Roman hydraulic engineers. One device for reducing flow velocity was the introduction of stepped chutes (Chanson 2000: 47–48), and a likely example has been identified on a section of the North-west Aqueduct (fig. 6). Earthquake damage at the same location offers a possible reason for the aqueduct going out of service. Nearby a cross-sectional profile of the aqueduct is exposed in a road cutting, and a long history of use and maintenance is attested in the plaster and calcareous sediment (calcareous sinter; Sürmelihindi et al. 2013: 962) lining it (fig. 7). Away from the Jarash Valley evidence of ancient aqueducts in a rural context is sparse; however, a section of masonry aqueduct lined with opus signinum plaster and completely choked with calcareous sinter is preserved at Tannur spring. Extensive re-use of the ancient spring-fed aqueduct irrigation system in the late Ottoman period over the entire length of the Jarash Valley and Wadi Tannur is evident from early aerial photographs and field survey and gives a sense of the likely extent of the system in antiquity. In the late Ottoman period water was also taken from the wadi stream to augment irrigation water supplied from springs to feed the many watermills that lined the Jarash Valley, and it is likely that similar use was made of this supply in antiquity.

The Reservoirs Large storage reservoirs (or birkets) are not common within the project area (fig. 8). The largest is the masonry reservoir built at Birketein north of the city prior to AD 209–2114 to hold water from the adjacent spring. Based on surface dimensions of 88.5 x 44.5 m and a depth of 4.8 m estimated from historic photographs, it had a theoretical maximum capacity of c. 18000 m3 (after allowing for the volume of the internal dividing wall) but its effective capacity was potentially reduced to 11000 m3 by the lower wall that divides it into two compartments. These estimates are substantially higher than the previously published figure of ‹more than 7000 m3› (Seigne 2004: 175) due to the greater estimated depth. The compartments were probably part of the system controlling the supply of irrigation water 4

The colonnade around the pool is dated by inscription to AD 209–211 (McCown 1938: 167).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Aqueducts and Birkets … Water Management System Servicing Gerasa (Jarash)

527

Fig. 8. Location of main reservoirs in the Gerasa area. Source: Hunting Aerial Survey Photograph 6.113, courtesy of APAAME.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

528

Boyer

to consumers in the valley downstream. While there are larger Roman reservoirs in Jordan – examples include the 19765 m3 Umm el-Quttein 1 reservoir in the Hauran, 70 km ENE of Gerasa, and the huge 68180 m3 Jiza reservoir south of Amman (Kennedy 1995: Table 1) – Birketein is one of the largest masonry reservoirs in the Jordanian Decapolis. The next largest reservoir outside the city was located at Bab Amman 750 m south of the city’s south gate. It now lies beneath modern housing and is only known from early aerial photographs. It was c. 25 x 15 m in size, and had a capacity of c. 1100 m3 based on an estimated depth of 3.0 m and was cut into a west-facing slope of exposed bedrock. While it probably stored rainwater harvested from the adjoining slopes, there is also photographic evidence of a possible aqueduct from the north. No entry points of major aqueducts to the city dating to antiquity have been identified with certainty and there is no published evidence of castella divisiori. The high ground on the west side of the walled city would be the logical location of reservoirs storing water for gravity distribution to consumers in the city below, and it is here that we find evidence of two reservoirs. The largest currently known is located at 612 m elevation on the hillside c. 130 m west of the Temple of Artemis. Based on an estimated depth of 4.0 m this c. 41 x 17 m rock-cut reservoir had a capacity of c. 2800 m3. Its high elevation and location suggests that it was probably supplied by the North-west Aqueduct. A rectangular c. 35 x 15 m rock-cut structure located at 610 m elevation on a hill in the south-west quarter of the city, 50 m south of the Church of Saint Peter and Saint Paul, is interpreted to be another reservoir.

Preliminary Conclusions The city’s early aqueduct delivery system has yet to be dated absolutely but archaeological evidence from the major water-consuming installations within the city points to a need for a reticulated water supply from the beginning of 2nd century AD. This does not preclude the existence of a reticulated water supply prior to this – perhaps even of Hellenistic date – but no evidence of such a system has yet come to light. While the bathing complexes on the east side of the city were supplied from Qairawan spring inside the city, on present evidence it seems likely that the water installations on the west side would have had to rely on aqueducts bringing water from sources outside the city. The west side of the city was supplied from an important aqueduct – the North-west Aqueduct – that entered the city close to the city’s north-west gate, although the end-users of this supply are currently unknown. The preliminary aqueduct plan shows that both sides of the city may have also been supplied via one or more aqueducts bringing water from known sources to the north of the city. The entry points of these northern aqueducts in antiquity have not been located, although irrigation water was channelled to the city from these sources until recently. The spring and/or reservoir at

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Aqueducts and Birkets … Water Management System Servicing Gerasa (Jarash)

529

Birketein may have been one of the sources of water for the city, but this is not proven and as a consequence the notion of a direct link between Birketein and specific water installations in the city is not substantiated. It is, however, very likely that water from Birketein reservoir would have irrigated the valley to the north of the city just as it does today. Gerasa’s aqueduct system differs in several respects from the systems that supplied the Decapolis cities of Gadara, Abila and Adraa in northern Jordan, which have been studied in more detail. On present evidence the sources of Gerasa’s aqueducts probably lie within 5 km of the city; much closer than most of the sources identified for Gadara, Abila and Adraa which were – in part – supplied from a source in southern Syria. While Gerasa was supplied via simple rock-cut and masonry aqueducts lying at or close to ground level, the water supply to Gadara, Abila and Adraa from the source in Syria was transported by a tunnel aqueduct system 100 km in length. The only tunnel aqueduct attested near Gerasa so far is a short section of c. 100 m south of the city that was part of a system that supplied irrigation water to the southern valley. The water storage component of Gerasa’s water management system is poorly understood, as is the reticulation system within the city. Two unexcavated water storage reservoirs have been identified at high elevations on the west side of the city, but they cannot yet be directly linked to any aqueduct. Two large reservoirs have been identified outside the city, the largest being Birketein – which was also one of the largest in the Decapolis – but both are likely to have been part of the irrigation system in the hinterland rather than being part of the city’s delivery system. Perennial stream flow in the Jarash Valley would have provided irrigation opportunities during the classical period, just as it does today, but there is also evidence of the extensive use of springs to supply irrigation networks that likely encompassed the entire Jarash Valley and the lower section of the Wadi Tannur valley.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

530

Boyer

Bibliography Al-Daire, M., 2004. Water Management in Trans-Jordan Byzantine Architecture with Respect to Excavated Monuments in the City of Gadara/Umm Qais. Bienert/Häser 2004: 219–230. Al-Qaisi, B.M., 2010. Climate Change Effects on Water Resources in Amman Zarqa Basin-Jordan. Individual Project Report for Climate Change Mitigation and Adaptation, Ministry of Water and irrigation. Amman. APAAME, Aerial Photographic Archive for Archaeology in the Middle East. www.apaame.org. Baker, F. M. C./Kennedy, D. L., 2010. Report on the Jarash Hinterland Survey 2005 and 2008. AJA 114, 519–521. –

2011. Jarash Hinterland Survey 2010 – An overview of results. ADAJ 55, 451–466.

Barghouti, A. N., 1982. Urbanization of Palestine and Jordan in Hellenistic and Roman times. SHAJ 1, 209–229. Bienert, H-D./Häser, J. (eds), 2004. Men of Dikes and Canals. The Archaeology of Water in the Middle East. International Symposium held in Petra, Wadi Musa (H. K. of Jordan) 15–20 June, 1999. Rahden, Braemer, F., 1989. History of Exploration at Jerash. D. Homès-Fredericq/J. B. Hennessy (eds.), Archaeology of Jordan II, vol.1, Field Reports, Surveys and Sites (A–K). Leuven, 316–319. Browning, I., 1982. Jerash and the Decapolis. London. Burckhardt, J. L., 1822. Travels in Syria and the Holy Land. London. Chanson, H., 2000. Hydraulics of Roman Aqueducts: Steep Chutes, Cascades, and Dropshafts. AJA 104, 47–72. Döring, M., 2004. Wasser für die Dekapolis. Römisches Wasserversorgungssystem im Norden Jordaniens. Schriften der Dt. Wasserhist. Ges. 5, 183–212. –

2005. Wasser für Gadara: Antikes Wasserversorgungssystem in Jordanien entdeckt. ImpulsE 10.1,

–

2008. Qanat Firaum – über 100 km langer unterirdischer Aquädukt im nordjordanischen Bergland.

41–50. C. Ohlig (ed.), Cura Aquarum in Jordanien. Proceedings of the 13th International Conference on the History of Water Management and Hydraulic Engineering in the Mediterranean Region Petra/ Amman 31 March – 09 April 2007. Siegburg, 189–204. Freeman, P., 2008. The Roman Period. R. B. Adams (ed.), Jordan: An Archaeological Reader. London, 413–441. Glueck, N., 1939. Explorations in Eastern Palestine III. AASOR XVIII–XIX (1937–1939). New Haven. Harding, G. L., 1967. The Antiquities of Jordan. London. Kehrberg, I./Manley, J., 2002. The 2001 Season of the Jarash City Walls Project: preliminary report. ADAJ 46, 197–203.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Aqueducts and Birkets … Water Management System Servicing Gerasa (Jarash)

531

Kennedy, D. L., 1995. Water supply and use in the Southern Hauran, Jordan. Journal of Field Archaeology 22.3, 75–90. Kennedy, D. L./Baker, F. M. C., 2009. Jarash Hinterland Survey 2005 and 2008. Council for British Research in the Levant Bulletin 4, 19–26. Kerner, S., 2004. Water-systems in Gadara and other Decapolis cities in Northern Jordan. Bienert/Häser 2004, 187–202. Kerner, S. et al., 1997. Water management in northern Jordan: the example of Gadara Umm Qays. SHAJ VI, 265–270. Lepaon, T., 2008. Les édifices balnéaires de Gerasa de la Décapole: Premières observations. Syria 85, 51–70. –

2011. Un Nouveau Plan Pour Jarash/Gerasa (Jordanie). ADAJ 55, 409–420.

–

2012. The Western baths of Gerasa of the Decapolis: original or standard building in the Near Eastern bathing context? R. Kreiner/W. Letzner (eds.), Proceedings of the International FrontinusSymposium on the Technical and Cultural History of Ancient Baths. Aachen, March 18–22, 2009. Leuven, 117–124.

McCown, C. C., 1938. The Festival Theater at the Birketein. K. Kraeling (ed.), Gerasa: City of the Decapolis. New Haven, 159–167. Raja, R., 2012. Urban Development and Regional Identity in the Eastern Roman Provinces, 50 BC–AD 250. Copenhagen. Rambeau, C. M. C./Black, S., 2011. Palaeoenvironments of the Southern Levant 5000 BP–present: linking the geological and archaeological records. S. Mithen/E. Black (eds.), Water, life and civilisation: climate, environment and society in the Jordan Valley. Cambridge, 94–104. Rostovtzeff, M. I., 1932. Caravan Cities. Oxford. Sapin, J., 1998. A l’est de Gerasa. Aménagement rural et reseau de communications. Syria 75, 107–136. Seigne, J., 2002. Gerasa-Jerasch – Stadt der 1000 Säulen. A. Hoffmann/S. Kerner (eds.), Gadara – Gerasa und die Dekapolis. Mainz, 6–22. –

2004. Remarques préliminaires à une étude sur l’eau dans la Gerasa antique. Bienert/Häser 2004:

–

2008. Fontaines et adduction d’eau à Gerasa (Jerash, Jordanie). Syria 85, 33–50.

173–185. Sürmelihindi, G. et al., 2013. Laminated carbonate deposits in Roman aqueducts: Origin, processes and implications. Sedimentology 60, 961–982.

Don Boyer PhD Candidate, Classics and Ancient History, The University of Western Australia, Australia.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9th ICAANE, Basel 2014, Vol. 3, 533–548

Marta D’Andrea

Pottery Production at Khirbat Iskandar, Jordan. Preliminary Results of the Technological Study of EB IV Pottery from the Site The article presents the preliminary results of the technological study of EB IV pottery from Khirbat Iskandar, Jordan, carried out to reconstruct manufacturing techniques at the site during this period. In addition, the paper will compare Khirbat Iskandar’s ceramic production with other southern Levantine sites to address regional patterns of technological, cultural, and socio-economic interactions during EB IV.

1. Khirbat Iskandar: Introduction and General Setting Khirbat Iskandar (fig. 1) lies in the Central Plateau of Transjordan, on the west bank of the Wadi Wala and along the ancient King’s Highway, one of the main routes linking the southern Levant to Egypt and to north-eastern Syria. The site has been investigated since 1981 by an archaeological expedition directed by Suzanne Richard first and subsequently by Richard and Jesse C. Long, still ongoing, which uncovered archaeological evidence possibly spanning the entire EBA. The earliest phases of the settlement are not yet well known thus far, but it is certain that Khirbat Iskandar was one of the main fortified settlements of the southern Levant during EB III, when walled towns spread all over the region (see Richard 2016). The EB III settlement was brought to an end by a destruction, uncovered in Area B. However, three EB IV phases in Area C and two EB IV phases in Area B testify that the site was settled continuously during EB IV1, a period characterised by shifts in the settlement patterns, village life, and higher groups mobility in the southern Levant (see Palumbo 2008, with references to 1

For Area C see Richard et al. 2010, with complete and updated bibliography; for Area B, see Richard 2000; Richard/Long 2009 and Richard 2016.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

534

D’Andrea

Fig. 1. Khirbat Iskandar, topographical map of the site (© Archaeological Expedition to Khirbat Iskandar).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Pottery Production at Khirbat Iskandar, Jordan.

535

previous works; D’Andrea 2014: 223–235, 270–278). Subsequently, the site was abandoned, possibly because of drier climatic conditions (Cordova/Long 2010: 34). The present paper focuses on ceramic materials from Areas C (§§ 2.1–2.2), where one of the longest EB IV chronological and typological sequences published so far were uncovered, and B (§ 2.3), which provided substantial EB III and IV stratified sequences (fig. 2). We studied ceramic materials from both areas from a technological perspective in order to reconstruct manufacturing techniques during EB IV and to investigate if changes occurred between the different phases, and if these could be linked to different socio-economic patterns within the period. The preliminary results of this research2 can be used to compare Khirbat Iskandar’s EB IV ceramic production with other southern Levantine sites, and address regional patterns of technological, cultural, and socio-economic interactions during the late EBA (§§ 3–4).

2. Pottery Production at Khirbat Iskandar during EB IV 2.1 Methods and Stratigraphic Basis of the Research Valentine Roux and Marie-Agnés Courty (1998: 748–750) identified four manufacturing methods accomplished with the tournette on ancient ceramics: 1) Method 1 (wheel-finished vessels): coils are built, joined, and thinned without the slow wheel, and then the vessel is finished on the tournette; 2) Method 2 (wheel-finished vessels): as no. 1, but coils are thinned and shaped on the slow wheel; 3) Method 3 (wheel-finished vessels): coils are built without the wheel, and then joined, thinned, and shaped on the wheel; 4) Method 4: the entire process is carried out on the tournette, i.e., the vessels are really wheel-fashioned. Roux and Courty (1998: 750–753; see also Courty/Roux 1995) also identified the macro-surfaces associated to each method and recognisable on the vessels surfaces at a macroscopic observation, providing a dataset to use as a reference point for the investigation of ancient pottery technology, followed within the present study.

2

I wish to thank deeply Suzanne Richard for allowing me to carry out and present preliminarily in these proceedings the technological study of the EB IV pottery from Khirbat Iskandar.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

536

D’Andrea

Fig. 2. Khirbat Iskandar, the EB IV architectural sequences in Areas C and B (© Archaeological Expedition to Khirbat Iskandar).

In our study, Khirbat Iskandar Area C was chosen as a starting point, since it provided a long EB IV architectural and stratigraphic sequence. This includes two phases of domestic occupation and a third phase in which the area was transformed into a multifunctional complex of rooms interpreted as a «gateway», with further architectural sub-phases (Richard/Long 2009: 94–95; Long 2010; here fig. 2). Three ceramic repertoires illustrative of each stratigraphic phase were isolated by Richard and Holdorf on a statistic basis (Holdorf 2010: 113–122; Richard 2010; 2013; here fig. 3). Despite the overall typological continuity of the pottery repertoires during the different EB IV phases at the site and the presence of long-lasting EB IV vessel types during the whole period, although decreasing in percentage in the later assemblages, there are vessel types representative only of the ceramic repertoire of each phase (fig. 3). The EB IV assemblages from Area C (§ 2.2) were first analysed in a comparative perspective, and confronted with the EB III pottery from Khirbat Iskandar to look at continuity or break between the two periods at the site. Since EB III has not yet been reached in Area C thus far, we used the EB III ceramic corpus from Area B (fig. 4) for a preliminary comparative analysis between the two periods at the site. The aim of this comparison was to identify possible changes in the manufacturing techniques between this period and the first EB IV phase.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Pottery Production at Khirbat Iskandar, Jordan.

537

Fig. 3. Khirbat Iskandar, EB IV pottery from Area C, Phases 1-3 (© Archaeological Expedition to Khirbat Iskandar).

Secondly, the analysis concentrated on the Area C pottery with the goal to investigate whether technological changes occurred; and, if so, if these changes could be linked to the site’s trajectory of crisis, recovery, and growth over three stratified EB IV phases identified in this sector of the site (Richard/Long 2010: 278; § 3). In addition, ceramics from the two EB IV phases in Area B (Richard/Long 2009: 95–99) were analysed for technological changes in order to compare the results with those from Area C (§ 2.3).

2.2 Summary of the Results of Technological Study of EB IV Pottery from Khirbat Iskandar’s Area C As we said above (§§ 1, 2.1), three EB IV phases, Phases 1–3, were identified in Area C, each corresponding to a pottery repertoire. Area C Phase 1 pottery features red-slipped and burnished platter-bowls (Richard 2010: fig. 4.5:3–5; here see figs. 3, 5.8–18) pointing to EB III/IV technological and typological continuity (Long 2010: 63). On the other hand, a group of coarsely handmade vessels (fig. 5.1–7), different from the EB III wheel-fashioned vessels (fig. 4), argue also for changes in the manufacturing processes in the early EB IV phase at the site. From a technological point of view, the Phase 1 red-slipped and burnished ceramics are different from the other contemporary vessels dating from Phase 1, being definitely finer. This raises the question

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

538

D’Andrea

Fig. 4. Khirbat Iskandar, EB III pottery from Area B, Phase C showing traces of wheel-fashioning (photos by the author, © Archaeological Expedition to Khirbat Iskandar).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Pottery Production at Khirbat Iskandar, Jordan.

539

Fig. 5. Khirbat Iskandar, early EB IV pottery from Area C, Phase 1. Nos. 1-7: handmade pottery; nos. 8-18: red-slipped and burnished pottery (photos by the author, © Archaeological Expedition to Khirbat Iskandar).

of whether to consider Phase 1 as a transitional phase or not, since, as recently pointed out by Marco Iamoni (2014: 20), «the occurrence of transition does not permit the existence of two parallel traditions filling similar roles», and radical changes characterise the appearance of new archaeological horizons. However, Khirbat Iskandar Phase 1 is still poorly known

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

540

D’Andrea

Fig. 6. Khirbat Iskandar, Area C, manufacturing techniques of Phases 2 (nos 1-3) and 3 (nos 4-8) pottery (photos by the author, © Archaeological Expedition to Khirbat Iskandar).

compared to Phases 2–3, and more materials from stratified contexts and data on the local EB III ceramic assemblage are needed for an in-depth analysis of this issue. On the Phase 2 ceramics, we have identified traces of the simplest methods according to Roux and Courty’s classification of the wheel-fashioning techniques, i.e. using the slow wheel only to finish the vessels. Some vessels are hand-coiled and wheel-finished (Roux and Court’s Methods 1; fig. 6.1). According to Roux and Courty (1998: 752, tab. 3, fig. 2:a), on vessels made with this method, the different coils are no longer visible, due to prolonged

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Pottery Production at Khirbat Iskandar, Jordan.

Fig. 7. Khirbat Iskandar, Area B, manufacturing techniques of Phases B–A pottery (photos by the author, © Archaeological Expedition to Khirbat Iskandar).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

541

542

D’Andrea

irregular pressure applied on the vessels’ surfaces while shaping the roughouts without rotatory kinetic energy (henceforth RKE), and the vessels’ walls are not stretched. However, concentric parallel striations on both surfaces testify to the use of a rotational device to finish the vessels at the end of the process. Other vessels are hand-coiled and wheel-thinned and shaped (Roux and Courty’s Method 2; fig. 6.2–3). Hand-coiling is noticeable thanks to irregular coil-bindings, and concentric parallel striations due to wheel-finishing are visible on the inner and outer surfaces of the vessels (Roux/Courty 1998: 752, tab. 3: fig. 2:b). Phase 3 pottery was made with the same manufacturing techniques as the Phase 2 ceramics. Thus, we identified vessels hand-coiled and shaped and wheel-finished (Roux and Courty’s Methods 1–2; fig. 6.4–5), and vessels hand-coiled and wheel-thinned and shaped (Roux and Courty’s Method 2; fig. 6.6). Furthermore, we noticed a third method, consisting in a mixed technique, according to which vessels are partly handmade and wheel-finished and partly wheel-fashioned. In such vessels, the body is hand-coiled, as usually visible from the transversal thumb-impressions resulting from binding the coils by hand, then the outer surface is wheel-finished, and the neck is wheel-coiled and inserted into the body (fig. 6.7–8). This results in a sharp inner ridge at the junction between the two parts.

2.3 Summary of the Results of the Technological Study of EB IV Pottery from Khirbat Iskandar’s Area B Based on pottery types and styles, the two EB IV phases excavated in Area B seem to compare well with Area C Phases 2–3 (compare, e.g., fig. 6.5 and fig. 8), while no earlier materials comparable to Area C Phase 1 have been found thus far. From the technological point of view, the slow wheel is largely employed in pottery manufacturing, utilising a variety of techniques, including the same as those identified in Area C. We identified vessels hand-coiled and wheel-shaped (Roux and Courty’s Method 1; fig. 7.1), vessels hand-coiled and wheel-thinned and shaped (Roux and Courty’s Method 2; fig. 7.2), and vessels made with a mixed technique (fig. 7.3). Beside these cases, there are also wheelfashioned vessels (Roux and Courty’s Method 4; fig. 7.4–5), i.e., vessels built, thinned, and shaped on the slow-wheel. These are recognisable thanks to the presence of regular coilbindings, resulting from the fact that each new coil was applied on the even surface of the underlying one, due to the fact that the slow wheel was employed continuously within the process (Roux/Courty 1998: 753, tab. 3, fig. 2:d). A general typological continuity with Phase B seems to characterise the Phase A pottery repertoire, as suggested by the presence of the same late EB IV vessel types, such as cymaprofiled bowls (fig. 8). The main change between the ceramic assemblages of the two EB IV phases in Area B is the presence of pottery imported from the neighbouring areas in Phase B (Goren 2010), and a locally produced pottery repertoire in Phase A. On the other hand, manufacturing techniques show a pattern of strong technological continuity between Phase

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Pottery Production at Khirbat Iskandar, Jordan.

543

Fig. 8. Khirbat Iskandar, cyma-profiled bowls from Area B, Phases B-A (photos by the author, © Archaeological Expedition to Khirbat Iskandar).

B and A. Vessels are either hand-coiled and wheel-shaped (Roux and Courty’s Method 1; fig. 7.6), hand-coiled and wheel-thinned and shaped (Roux and Courty’s Method 2; fig. 7.7), or made with a mixed technique (fig. 7.8). Compared to Phase B, wheel-coiling is possibly more diffused, and a fourth manufacturing method could be identified, although very rarely, according to which the coils are built separately and then joined, thinned, and shaped on the tournette (Roux and Courty’s Method 3; fig. 7.9). Such a method is easily recognisable through the associated macro-surfaces, due to the presence of a band crossed by

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

544

D’Andrea

a non-rectilinear groove in the middle, resulting from «intermittent differential pressure» applied to the vessels’ surfaces (Roux/Courty 1998: 752–753, tab. 3, fig. 2:c).

3. Pottery Technology at Khirbat Iskandar during EB IV in the Regional Context Results of the technological study of the EB IV pottery from Khirbat Iskandar are preliminary and some issues still need to be solved, such as those related to Area C Phase 1 and its possibly «transitional» character. Yet, these can be used to put forward some observations related to technology, chronology, and diachronic socio-economic patterns, which could apply to a regional scale. From the technological point of view, a datum arguable from the study of the EB IV pottery from Khirbat Iskandar is the use of the slow wheel during the central and late EB IV phases. This remark is not irrelevant, since Roux has repeatedly ruled out that the tournette was used in pottery manufacturing during EB IV in the southern Levant (Roux 2013, with references to previous bibliography). Yet, our examination of the Khirbat Iskandar materials offered several parallels for the macro-surfaces associated to the use of the slow wheel as resulting from the dataset provided by Roux and Courty themselves (§§ 2.2–2.3). Microscopic analysis of the micro-surfaces associated to the possible use of wheel-coiling on the Area B materials from Khirbat Iskandar are currently in progress in Rome, led by Laura Medeghini and Caterina De Vito (Sapienza University of Rome, Department of Earth Science). The study, which will be subject of future publications, confirmed the use of the slow wheel in the pottery making at the site during EB IV. In some previous studies, we have already pointed out that the stratified ceramic assemblages of the EB IV key-sites show a clear-cut technological difference between EB III and early EB IV and the «early» and «late» EB IV phases (D’Andrea 2012: 21–26; 2014: 59–94; 2015: 33, figs. 1–4). The main distinction was the use of the slow wheel in EB III, its demise in the «early» EB IV phase, and its reintroduction in the «late» EB IV phase.3 We noticed that morphological and stylistic features correlated to different manufacturing techniques can be recognised on pottery from different parts of the southern Levant, which can thus be linked to an «early» and a «late» EB IV relative chronological horizons. This allowed us to use pottery technology as a marker for relative chronology cutting across the sub-regional differences in the vessels’ types and styles, which, otherwise, make inter-regional correlations difficult, due to the marked ceramic regionalism. 3

On the use of the slow wheel in the southern Levant during EB IV, see also Kamaiski in Milevski et al. 2012: 102–106; London 2014.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Pottery Production at Khirbat Iskandar, Jordan.

545

Svend Helms (1986: 30–31, 42–49) and Lorenzo Nigro (2003: 133–134) had already identified diachronic changes in technology in the EB IV stratified pottery sequences of Tell Umm Hammad and Jericho respectively. At both sites, it was clearly noticeable that the early EB IV pottery (Tell Umm Hammad Stage 5; Jericho Phase Sultan IIId1) was entirely handmade. At Tell Umm Hammad the tournette was gradually reintroduced to finish the vessels’ rims and necks in an «intermediate» EB IV phase (Stage 6), and, since then, systematically used in the following EB IV phases (Stages 7/8). This parallels well the late EB IV phase at Jericho (Phase Sultan IIId2), characterised by wheel-finished and wheel-fashioned pottery. Khirbat Iskandar, with its long stratigraphic and ceramic sequences and its phased assemblages, offered a great opportunity to test hypotheses on the development of pottery manufacturing techniques in the southern Levant during EB IV. The differences in pottery technology between the early (Area C Phase 1) and late (Area C Phases 2–3, Area B Phases B–A) EB IV phases that we have identified in the Area C sequence at Khirbat Iskandar could be compared with those noticeable at other southern Levantine sites with long EB IV stratigraphic sequences with the aim of relative chronology. However, the coexistence of coarse handmade vessels and finer red-slipped and burnished vessels in Area C Phase 1 (§ 2.2; fig. 5) represents a picture slightly different from that of the Jordan Valley sites. Pottery production in the early EB IV phase could have been more articulated in a sub-regional area, such as central Transjordan, featuring a strong EB III/IV continuity. In fact, the redslipped and burnished vessels found in early EB IV assemblages at Khirbat Iskandar are not isolated, but have close parallels in the ceramic materials from Khirbet al-Minsahlat (Chesson et al. 2005: fig. 26:c, e). The site, which is also located in the eastern Dead Sea basin, seems to belong as well to an early EB IV horizon or even to a possibly EB III/IV transitional phase. The diachronic developments of manufacturing techniques of the ceramic materials from Area C at Khirbat Iskandar, observable from one phase to the next, are consistent with the tripartite EB IV trajectory of the site inferred from stratigraphy and architecture of Phases 1–3 (Richard/Long 2010: 278). This trajectory includes: 1) a first phase of crisis, corresponding to the EB IV village occupation at the site after the destruction of the EB III fortified settlement, and possibly characterised by a «transitional» EB III/IV ceramic horizon, featuring a possible technological impoverishment; 2) a central phase of recovery, in which the village prospered, with a ceramic horizon characterised by technological improvements and the introduction of new vessel types; 3) a final phase of growth, when the Area C structures were transformed into a multifunctional «public» complex, interpreted as a «gateway», distinguished by a skilled pottery technology and by the possible presence of «transitional EB IV/MB I» elements within the ceramic assemblages.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

546

D’Andrea

4. Implications and Hints for Future Research The possible diversified socio-economic patterns outlined in the Khirbat Iskandar sequence could offer a means to reconstruct the regional EB IV relative sequence for the central plateau of Transjordan. This could be relevant when the new higher absolute dates for the southern Levantine EB IV are taken into account. In fact, absolute dates of the beginning of EB IV at ca. 2500 BC and of the end of the period at ca. 1920 BC were proposed. These dates are based respectively on Bayesian modelling of 14C dates for the end of EB III (Regev et al. 2012: 559–561; Höflmayer 2014: 125–132) and on new 14C dates for the beginning of MB I (Marcus 2003; Cohen 2012). The resulting longer duration for EB IV compels us to reappraise the question of the internal periodization of the last phase of the EBA and to try to identify progressive socio-economic developments within such a long time-span, if these chronological suggestions, which currently meet growing favour, would prove correct. Thus, the Khirbat Iskandar sequence, finding parallels also at other southern Levantine EB IV key-sites, could allow us to highlight phenomena of socio-economic change during EB IV, usually concealed by the consolidated practice of considering this period as a static and homogeneous interlude of regional collapse. This would offer the basis for a long-time awaited re-evaluation of the southern Levantine EB IV.

Bibliography Chesson, M. S. et al., 2005. Results of the 2001 Kerak Plateau Early Bronze Age Survey. Annual of the American Schools of Oriental Research 59. Part I. Boston MA, 1–62. Cohen, S., 2012. Synchronisms and significance: Reevaluating interconnections between Middle Kingdom Egypt and the southern Levant. Journal of Ancient Egyptian Interconnections 4.3, 1–8, accessed online at http://jaei.library.arizona.edu. Cordova, C. E./Long, J. C. Jr., 2010. Khirbat Iskandar and its Modern and Ancient Environment. Richard et al. 2010: 19–35. Courty, M.-A./Roux, V., 1995. Identification of Wheel-Throwing on the Basis of Ceramic Surface Features and Microfabrics. Journal of Archaeological Science 22, 17–50. D’Andrea, M., 2012. The Early Bronze IV Period in South-Central Transjordan: Reconsidering Chronology through Pottery Technology. Levant 44, 17–50. –

2014. The Southern Levant in Early Bronze IV. Issues and Perspectives in the Pottery Evidence. Contributi e Materiali di Archeologia Orientale XVII, Rome.

–

2015. The socio-economic landscape of the Early Bronze IV period in the Southern Levant: a ceramic perspective. G. Affanni et al. (eds.), Broadening Horizons 4. Proceedings of the 4th International Conference of Young Researchers Working in the Ancient Near East, Egypt, and

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Pottery Production at Khirbat Iskandar, Jordan.

547

Central Asia, University of Torino, October 2011. British Archaeological Report. International Series 2698. Oxford, 31–38. Goren, Y., 2010. Ceramic Technology and Provenance at Khirbat Iskandar. Richard et al. 2010: 133–140. Helms, S., 1986. Excavations at Tell Umm Hammad, 1984. Levant 18, 25–49. Höflmayer, F., 2014. Dating Catastrophes and Collapses in the Ancient Near East: The End of the First Urbanization in the Southern Levant and the 4.2 ka BP Event. L. Nigro (ed.), Reading Catastrophes. Proceedings of the International Conference “Reading Catastrophes: Methodological Approaches and Historical Interpretation. Earthquakes, Floods, Famines, Epidemics between Egypt and Palestine – 3rd–1st millennium BC” held in Rome, 3rd–4th December 2012. Rome “La Sapienza”. Studies on the Archaeology of Palestine and Transjordan 11, Rome, 117–139. Holdorf, P. S., 2010. Quantitative Analysis of the Early Bronze IV Tell and Tomb Ceramic Assemblage. Richard et al. (eds.) 2010, 113–132. Iamoni, M., 2014. Transitions in Ceramics, A Critical Account and Suggested Approach: Case-Study through Comparison of the EBA–MBA and MBA–LB Horizons at Qatna. Levant 46.1, 4–26. London, G. A., 2014. Manufacturing Technique of Late Third Millennium BCE Flat-Bottomed Pots. W. G. Dever, Excavations at the Early Bronze IV Sites of Jebel Qa‘aqir and Be‘er Resisim. Harvard Semitic Museum Publications 6. Winona Lake, In., 253–278. Long, J. C., 2010. The Stratigraphy of Area C. Richard et al. 2010: 37–68. Marcus, E., 2003. Dating the Early Middle Bronze Age in the Southern Levant: A Preliminary Comparison of Radiocarbon and Archaeo-Historical Synchronizations. M. Bietak (ed.), The Synchronisation of Civilisations in the Eastern Mediterranean in the Second Millennium B.C. II. Proceedings of the SCIEM 2000-EuroConference Haindorf, 2nd–7th of May 2001 (= Contribution to the Chronology of the Eastern Mediterranean 4 / Denkschriften der Gesamtakademie Band 19). Vienna, 95–110. Milevski, I., et al., 2012. Er-Rujum (Sha’alabim East): An Intermediate Bronze Age (EB IV) Site in the Ayyalon Valley. ‘Atiqot 69, 76–140. Nigro, L., 2003. Tell es-Sultan in the Early Bronze Age IV (2300–2000 BC.). Settlement vs Necropolis – A Stratigraphic Periodization. Contributi e Materiali di Archeologia Orientale IX, 121–158. Palumbo, G., 2008. The Early Bronze IV. R. B. Adams (ed.), Jordan: An Archaeological Reader, London, 227–262. Regev, J. et al., 2012. Chronology of the Early Bronze Age in the Southern Levant: New Analysis for a High Chronology. E. Boaretto/N. R. Rebollo Franco (eds.), Proceedings of the 6th International Radiocarbon and Archaeology Symposium. Radiocarbon 54.3–4, 525–566. Richard, S., 2000. Chronology vs Regionalism in the Early Bronze IV. L. E. Stager et al. (eds.), The Archaeology of Jordan and Beyond: Essays in Honor of James H. Sauer. Studies in the Archaeology and History of the Levant 1, Winona Lake, IN, 399–415. –

2010. The Area C Early Bronze IV Ceramic Assemblage. Richard et al. 2010: 69–112.

–

2013. Khirbat Iskandar Tall, Tombs and Bab adh-Dhra‘: A Multi-Dimensional Quantitative Approach. al-Hmoud (ed.). Studies in the History and Archaeology of Jordan XI, Amman, 783–791.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

548 –

D’Andrea 2016. Recent Excavations at Khirbat Iskandar, Jordan. The EB III/IV Fortifications. R. A. Stucky et al., Proceedings of the 9th International Congress on the Archaeology of the Ancient Near East (9. ICAANE), Vol 3. Wiesbaden 2016, 585–597.

Richard, S. et al. (eds.), 2010. Khirbat Iskandar Final Report on the Early Bronze IV Area C ‘Gateway’ and Cemeteries. ASOR Archaeological Reports 14/Archaeological Expedition to Khirbat Iskandar and its Environs, Jordan Vol. 1, Boston, MA. Richard, S./Long, J. C. Jr., 2009. Khirbet Iskander, Jordan and Early Bronze IV Studies: A View from a Tell. P. J. Parr, The Levant in Transition. Proceedings of a Conference held at the British Museum on 20–21 April 2004. Palestine Exploration Fund Annual 9, 90–100. –

2010. Summary and Conclusions. Richard et al. 2010: 271–279.

Roux, V., 2013. Spreading of Innovative Technical Traits and Cumulative Technical Evolution: Continuity or Discontinuity? Journal of Archaeological Method and Theory 20.2, 312–330. Roux, V./Courty, M.-A., 1998. Identification of Wheel Fashioning Methods: Technological Analysis of 4th–3rd Millennium BC Oriental Ceramics. Journal of Archaeological Science 25, 747–763.

Marta D’Andrea, Sapienza University of Rome, [email protected].

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 549–559

Sumio Fujii

Slab-lined Feline Representations: New Finding at ‘Awja 1, a Late Neolithic Open-air Sanctuary in Southernmost Jordan Our recent excavations at ‘Awja 1, a small-scale open-air sanctuary in southernmost Jordan, revealed several slab-lined features representing feline animals. Similar representations are known in the Negev and Sinai, suggesting close cultural contact between the two areas. This paper discusses their date, function, and archaeological implications in a broader context.

1. Introduction The site of ‘Awja 1 is located near the Jordan/Saudi border, in the middle of a sandstone desert that extends from the escarpment fringing at the southern edge of the Transjordanian Plateau toward the at-Tubayq Highlands in the northwestern part of Saudi Arabia (fig. 1). It is among five archaeological sites (i.e. ‘Awja 1–5) that we have investigated in the area. The excavations, which have taken place in 2011 and 2012, revealed that the site contained an elongated open sanctuary probably dated to the Late Neolithic, and that the sanctuary was associated with several slab-lined features representing feline animals (Fujii/Adachi et al. n. d.; Fujii/Yamafuji et al. n. d.). As mentioned below, similar combinations are known in the contemporary Negev Highlands and Sinai Peninsula, suggesting that there was close cultural contact between the two areas under the initial process of pastoral nomadization. This paper briefly reviews the investigation results at the desert site and discusses the date, function, and archaeological implications of the unique features.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

550

Fujii

100 km

‘Amman

Jerusalem

Beersheba

Karak

Ramat Saharonim

Harrat al-Juhayra Qa’ Abu Tulayha West

Wadi Musa Ma’an

Jafr Basin

Biq’at ‘Uvda 6

Khashm at-Taref sites

‘Aqaba

‘Awja 1-5

Late Neolithic open sanctuary: not associated with feline representations associated with feline representations

Fig. 1. Late Neolithic open sanctuaries in the southern Levant.

2. The Site ‘Awja 1 is situated in a flat fluvial plain c. 5 km northeast of Tell ‘Awja, a landmark of the area. The annual average rainfall in this area is less than 50 mm, and no perennial natural water source is available around the site. Understandably, no settlements exist; even pastoral nomads are rarely encountered. Being located in such a harsh natural environment, the site is both small in scale (c. 0.1–0.2 ha) and thin in artificial deposit (less than a few cm). Thus it is defined as a typical desert site. Aside from several small stone concentrations dotted in its northern half, the site consists of the following two structural entities: Complex I in the northeast and Complex II in the southwest. The construction method of the two complexes is very simple. Elongated sandstone slabs c. 10–30 cm long are merely put down on the then ground surface in an upright position with their longer sides facing up and down. The walls are two-rows or c. 15–20 cm wide and a single course or c. 5–10 cm high. Thus it follows that the complexes consist only of the foundation course. No clear evidence for clay mortar or foundation banks is recognized, although it is possible that it weathered away over a long period of time. Alluvial and/

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Slab-lined Feline Representations: New Finding at ‘Awja 1 … Jordan

551

Fig. 2. ‘Awja 1: general view of Complex I (looking N).

or aeolian sandy deposits of Layer 3 are used as a natural floor without any special treatment. This unique construction method (i.e. the two-rowed upright slab wall technique) is ubiquitous in the Neolithic Badia, and many examples have been found in the Negev and Sinai (e.g. Bar-Yosef 1982: 11) as well as southern Jordan (e.g. Fujii 2013: figs. 17, 19, 20). Of interest is the existence of corner stones, which marked the four corners (and, in some cases, the middle point of two long sides) of a rectangular unit constituting the complexes. This fact demonstrates that the complexes were constructed under a systematic architectural design. Complex I, our main concern, contains a total of four rectangular units c. 4–4.5 m wide and c. 2–2.5 m deep, which are divided into the following two parts (figs. 2–3). One is a laterally connected body of the northerly three units (i.e. Units A–C). This continuum stretches in the NEN-SWS direction, measuring c. 13 m in total length. The other is the freestanding Unit D, which is c. 1 m separated from the neighboring Unit C and, at the same time, slightly differs in general orientation from the continuum. This minor structural discontinuity possibly indicates a small temporal gap between the two. Every unit is rectangular in general plan and equally incorporates several rectangular or round features into its left sidewall. In addition, a round feature c. 1 m in diameter is added to the frontal space of every unit. It appears that both components were paired to form a composite unit or a pseudo-house burial cairn

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

552

Fujii

Fig. 3. ‘Awja 1: Plan of Complex I.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Slab-lined Feline Representations: New Finding at ‘Awja 1 … Jordan

553

Fig. 4. ‘Awja 1: close-up view of feline representations (looking W).

characteristic of the Late Neolithic open sanctuaries in the neighboring Jafr Basin (Fujii 2002; 2013: 70–78, 96–98). The wall-sharing relationship between any two adjacent units suggests that the elongated complex gradually developed southwestward, namely, from Unit A toward Unit D. Given this, it would follow that the indoor layout of each unit was gradually simplified in the course of the lateral development. As mentioned below, this perspective provides a key to approaching the chronological seriation of the five ‘Awja sites. Neither human skeletal remains nor faunal/floral remains, to say nothing of artifacts, were found at the complex. Even hearths and ashy deposits were not confirmed. The series of unique traits – the location in the middle of the hyper-arid land, the adoption of the two-rowed upright slab wall technique, the combination of an empty rectangular unit and an again empty circle, the lateral development of such homogeneous components and consequent formation of a continuum facing southwest, and the absence of traces of real life – are common to Late Neolithic open sanctuaries in the neighboring Jafr Basin (Fujii 2014b: 107–114), suggesting that Complex I shares a similar character and date to them.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

554

Fujii

Fig. 5. ‘Awja 1: close-up view of the tail tip of a feline representation (looking W).

3. Slab-lined Feline Representations Complex I was associated with a total of eight slab-lined features representing quadrupeds (fig. 4). They were produced using the same building material and technique as the adjacent four rectangular units. Every feature was filled with sandy soil including abraded flint pebbles c. 5–10 cm long, which might possibly represent black spots on the body surface of the quadrupeds. A similar representation is known in the Negev Highlands (e.g. Avner 2002: 114, figs. 5.153, 5.155). The eight features occupied the frontal space of Units C and D, being laid out in two rows and four tiers. The upper (i.e. westerly) two tiers comprised four large quadrupeds, which measured c. 1.5–2 m in trunk length, c. 1 m in trunk height, and c. 1 m in tail length, respectively. They probably represent adult individuals. The lower two tiers, on the other hand, contained four smaller quadrupeds measuring c. 1 m in trunk length, c. 0.5–0.6 m in trunk height, and c. 0.5 m in tail length. Unlike the upper four features, they used the single-row upright slab wall technique, which is probably attributable to their small size. It appears that they represent cubs of the adults arranged in the upper two tiers. (Such a regular layout differentiates ‘Awja 1 from the Negev and Sinai sites referred to below.) In terms of

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Slab-lined Feline Representations: New Finding at ‘Awja 1 … Jordan

555

iconography, all of the eight quadrupeds were represented in a side view with their small head being equally oriented southward and their long tail stretching northward. A few of them curled their tail tip upward, suggesting that they belong to feline animals. Furthermore, one of them took the trouble to use small pebbles to depict a rounded tail tip, another remarkable trait of the feline (fig. 5). It is evident that the unique features were combined with the adjacent units to form a unified complex, because both of them share the same site stratigraphy, construction material, and construction method. The question is which of the four units the features belonged to. The nearest two units, Units C and D, would be likely candidates, but further narrowingdown is difficult. A hint, if any, would be that extremely elongated sandstone slabs over 60 cm long are shared between one of the features and Unit D only. This fact seems to suggest that the features were attached to the final component of Complex I. As discussed below, this perspective becomes important in dating the features.

4. Discussion The following discussion deals with, first, the family identification of the quadrupeds and, then, their date, function, and archaeological implications in a broader context. The first issue requires no further argument. The series of remarkable traits – the slender trunk, relatively small and rounded head without horns, and the long tail curling only its rounded tip upward – allows us to regard them as representations of feline animals (i.e. Felidae). Especially important is the last item, which clearly distinguishes feline animals from the other quadrupeds. Although subfamily- or genus-level identification is not easy, cheetahs (Acinonyx jubatus) or leopards (Panthera pardus) seem likely in view of the slender trunk, long tail, and probable representation of black spots. The second issue (i.e. the date of the features) is difficult to approach due to the absence of direct evidence such as C-14 data and diagnostic artifacts. However, the following two clues are available. One is the local chronologies of open sanctuaries in the neighboring areas, which would provide a general framework for seriating the five ‘Awja sites including ‘Awja 1. The other is the typological sequence of the ‘Awja open sanctuaries themselves (Fujii et al. op. cit.). It is our present perspective that the ‘Awja open sanctuary started with the freestanding unit at ‘Awja 2, developed to its connected body at ‘Awja 5 and, through the typological simplification at ‘Awja 1, ended with bipartite complexes at ‘Awja 4 (fig. 6). (According to the typological sequence in the Jafr Basin, the unique complex at ‘Awja 3 is thought to have followed the series of rectangular units.) Both clues enable us to tentatively date the feline representations to the Late Neolithic, especially its latter half.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

556

Fujii

Fig. 6. Tentative chronology of open sanctuaries in the Levantine arid peripheries.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Slab-lined Feline Representations: New Finding at ‘Awja 1 … Jordan

557

What is important here is that all of the major open sanctuary sites associated with feline representations – ‘Awja 1 in southernmost Jordan, the Khashm at-Taref sites in the northeastern corner of the Sinai Peninsula (Avner op. cit.: 113-115, figs. 5.134–5.141; Eddy/ Wendorf 1999: 72–73, 74–75, figs. 3-37, 3-44, 3-45), and Biqat Uvda 6 in the southeastern Negev (Avner 1984; 2002: 105–106, 114, figs. 5.149–5.155; Goring-Morris 1993: 82–83, fig. 12; Yogev 1983: 121) – not only concentrate on a relatively limited range on both sides of the Lower Jordan Valley, but they also consist equally of simple complexes without any substantial indoor space division. (A small circle often incorporated into the center of a rectangular unit might possibly be understood as an indoor version of an outdoor circle recognized at earlier open sanctuaries.) It would follow that the feline representation was a cultural component proper to the southern edge of the Levantine Badia in the latter half of the Late Neolithic. Available evidence seems to suggest that it originated in the context of open sanctuaries in southern Jordan and, then, spread westward together with the bipartite complex. However, the final conclusion must await further investigation. What, then, were the unique features or Complex I associated with them used for? As for the complex, it is most unlikely that it represents a settlement, workshop, storehouse, or any other structure for practical use. The absence of traces of real life speaks for itself. Thus, it is more reasonable to assume that the complex was used as an open sanctuary of initial pastoral nomads who migrated in the arid peripheries of the southern Levant. The function of the feline representations can also be understood in this context, but their specific use is still obscure. We should note, however, that they are represented in a small group, and that they include cubs in the group. Both traits are shared among most of the sites referred to above, suggesting that the unique features were related to some fertility/prosperity rite (Avner 2002: 114–115). Avner (ibid.) identifies them as leopards and, on this base, claims that they actually represent deities of fertility, a likely interpretation in view of the series of collateral evidence he listed. However, his identification as leopards is still debatable, because the possibility that they represent cheetahs or hunting-cheetahs (i.e. tamed cheetahs for hunting) cannot be ruled out (Fujii 2008, 2010). In this case, the argument might be forced to make a significant turn. Anyhow, further scrutiny is needed to develop the details of the issue. The new finding at ‘Awja 1 has a twofold archaeological implication. To begin with, it has provided a glimpse at the ideological aspect of initial pastoral nomads already scarce in information. Although its iconographical interpretation is not yet fully determined, it has the potential to enrich our understanding of the LN culture. Second, it has enlarged the distribution range of the unique features eastward to a large extent and, by doing so, bridged a spatial gap between the Late Neolithic Negev/Sinai and the contemporary southern Jordan. As a result, it has enabled us to trace the process of pastoral nomadization in the two adjacent areas within a unified chronological framework. Though not associated with feline representations, a similar open sanctuary is recognized at the northern edge of the Baidat

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

558

Fujii

ash-Sham as well (Fujii et al. 2013; Fujii 2014a: 72–75). This means that the unique open sanctuary culture holds a key to approaching the issue of pastoral nomadization in the whole range of the Levantine arid peripheries. The feline representations must also be understood as an aspect of this far-reaching post-PPNB socio-cultural reorganization.

5. Concluding Remarks The new discovery at ‘Awja 1 has proved that the slab-lined feline representation is not a local phenomenon in the Negev and Sinai but can be understood as a common cultural index in the arid margins of the Late Neolithic southern Levant. In this sense, it provides valuable insights into the socio-cultural dynamics of the areas under the initial process of pastoral nomadization. However, much still remains obscure including its ritual meaning. Further investigation is required for a better understanding of the unique features.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Slab-lined Feline Representations: New Finding at ‘Awja 1 … Jordan

559

Bibliography Avner, U., 1984. Ancient cult sites in the Negev and Sinai deserts. Tel Aviv 11, 115–161. –

2002. Studies in the Material and Spiritual Culture of the Negev and Sinai Populations, during the 6th–3rd Millennia B.C. Ph. D. Thesis, Hebrew University. Jerusalem.

Bar-Yoser, O., 1982. Pre-Pottery Neolithic sites in southern Sinai. Biblical Archaeologists 45/1, 9–12. Eddy, F. W./Wendorf, F., 1999. An Archaeological Investigation of the Central Sinai, Egypt. Boulder. Fujii, S., 2002. Pseudo-settlement hypothesis: Evidence from Qa’ Abu Tulayha West, southern Jordan. Archaeozoology of the Near East 5, 181–194. –

2008. The origin of cheetah-hunting: A new perspective from Wadi Abu Tulayha in southern Jordan.

–

2010. The origin of cheetah-hunting: A new perspective from Wadi Abu Tulayha in southern Jordan.

Bulletin of the Ancient Orient Museum 28, 23–38 (in Japanese with a brief English summary). Oral presentation in ICHAJ-11 (the 11th International Conference of History and Archaeology of Jordan), Paris, June 7–12. –

2013. Chronology of the Jafr Prehistory and Protohistory: A key to the process of pastoral nomadization in the southern Levant. Syria 90, 49–125.

–

2014a. Chronology of the Bishri pastoral prehistory and protohistory: A cross-check against the Jafr chronology in southern Jordan. Studia Chaburensia 4, 63–92.

–

2014b. Make-believe playhouses at Wadi Burma East: A cognitive approach to the Neolithic unilinear settlement in the Jafr Basin, southern Jordan. B. Finlayson/C. Makarewitz (eds.), Settlement, Survey and Stone. Essays on Near Eastern Prehistory in Honour of Gary Rollefson. Berlin, 101–116.

Fujii, S./Adachi, T./Endo, H./Yamafuji, M., n.d. ‘Awja sites: Supplementary investigations of Neolithic open sanctuaries in southernmost Jordan. ADAJ 57 (in print). Fujii, S./Adachi, T./Yamafuji, M., 2013. Fakat Bidewy 1 and 2: Archaeological investigations around Bir Rahum, 2011 (Spring). Al-Rafidan 34, 3–12. Fujii, S./Yamafuji, M./Nagaya, K., n.d. ‘Awja 1–3: Neolithic and Chalcolithic open sanctuaries in southernmost Jordan. ADAJ 56 (forthcoming). Goring-Morris, A. N., 1993. From foraging to herding in the Negev and Sinai: The early to late Neolithic transition. Paléorient 19/1, 65–89. Yogev, O., 1983. A fifth millennium B.C.E. sanctuary in the ‘Uvda Valley. Qadomoniot 16/4, 118–132 (in Hebrew).

Sumio Fujii, Kanazawa University.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 561–570

Lucas Petit

Tell Damiyah. A Small Settlement Mound with a Remarkable International Role Tell Damiyah, a small settlement mound located on the eastern bank of the river Jordan, contains occupation remains from Bronze Age until the Ottoman Period. In order to clarify the remarkable international role of the site during the Iron Age, the Dutch National Museum of Antiquities started a new excavation in 2012.

1. Introduction The ancient site of Tell Damiyah is situated in the Central Jordan Valley, directly south of the confluence of the Zerqa and the Jordan River (fig. 1). The settlement mound consists of an upper tell and a lower terrace that occupies the western and southern sides. The upper tell covers an area of approximately three hectares at the bottom, has relatively steep slopes and rises almost 20 meters above the present valley floor (fig. 2). The exact extent of the lower terrace has to be investigated in the nearby future. Tell Damiyah has a strategic position on important trade routes running north-south and east-west, and commanding one of the few fords over the Jordan River.

2. History of Study Victor Guérin was the first who recognised the importance of Tell Damiyah (Guérin 1869: 238–240), although others like Irby and Mangles in 1818, William Lynch in 1848 and Charles van de Velde in 1851 must have passed the site during their travels (Irby/Mangles 1823: 325–326; Lynch 1849: 249–250; Van de Velde 1854: 321). John William McGarvey, visiting the site in 1879, mentioned the ruins of a building on the top (1881: 350). He was also one of the first scholars who equated Tell Damiyah with Adam(ah), a city mentioned in the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

562

Petit

Fig. 1. Iron Age sites in the Central Jordan Valley.

Old Testament (e.g. Joshua 3:16, I Kings 7:46, II Chr. 4:17; see for a discussion MacDonald 2000: 45–61) and on the victory stele of Shoshenq I in Karnak (Kitchen 1973: 298). From 1880 onwards the site was visited and surveyed frequently (e.g. Albright 1926: 47; Glueck 1951: 329–331; Mellaart 1962: 148; Yassine et al. 1988: 191). Those scholars found pottery from LB II, Iron I, Iron II, Persian, Early Roman, Byzantine and the Islamic Period.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Tell Damiyah. A Small Settlement Mound with a Remarkable International Role

563

Fig. 2. Tell Damiyah.

First archaeological excavations were carried out in 2004 and 2005 by the Leiden University and the Yarmouk University (Kaptijn et al. 2005; Petit et al. 2006; Petit 2009a). During these two seasons the main objective was to rescue and to document the uncovered archaeological remains in the bulldozer cut (Squares I–III). This recent cut had not only destroyed much of the southern summit, but had also initiated extensive erosion. More substantial archaeological research was started in 2012 with the goal to investigate the remarkable continuing occupation history at Tell Damiyah during the Iron Ages (Petit et al. 2006; Petit 2008; 2009a: 103–149; 2013; cf. Yassine 1988; Van der Kooij 2001). Under the auspice of the Dutch National Museum of Antiquities and with the help of the Department of Antiquities of Jordan, an international team of archaeologists and specialists directed by the present author, opened four squares on the summit (Petit i. p.). In 2013 the excavations were continued in the same squares. What follows is a brief preliminary overview of the occupation history of Tell Damiyah based on the results of four seasons of excavations.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

564

Petit

Fig. 3. Female figurine.

3. Iron Age II (10th–8th century BC) The oldest occupation remains were encountered during the 2005-season inside a step trench along the eastern slope: a large mudbrick wall, parallel to the site’s slope, and some circular cooking installations. Pottery and one 14C sample date these occupation remains somewhere at the end of the 10th or early 9th century BC (Petit 2009a: 103–109). The parallel wall remained in use for at least one hundred years with only minor architectural alterations. In the 8th century BC, the outlook and orientation of the buildings on Tell Damiyah changed considerably. It is too early to explain the difference in terms of newcomers since only limited remains were excavated. A fire destroyed these mudbrick structures not very long after their arrival.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Tell Damiyah. A Small Settlement Mound with a Remarkable International Role

565

4. Iron Age II (8th–7th century BC) After a period of abatement, somewhere during the late 8th century BC, Tell Damiyah experienced new construction activities. At least two multi-roomed complexes were built, divided by a small alley. The walls were of mudbricks and the flat roof consisted of large wooden beams and clay. A courtyard with flagstones was encountered towards the east of these buildings. Much of this occupation phase was damaged by the bulldozer, but also by post-Iron Age occupation (see below). A devastated conflagration destroyed the buildings, preserving some floor deposits and intriguing finds. Besides restorable pottery, one of the most amazing finds was a clay bulla with a cuneiform inscription (Petit 2009a: fig. 8.38). The two lines contain numbers and an unknown Akkadian word, possible a personal name. The discovery of this broken bulla points certainly to contact between the inhabitants of Tell Damiyah and the Neo-Assyrian empire (Petit 2009a: 226). This was also stated by the discovery of the so-called Assyrian Palace Ware (cf. Anastasio 2010). The seasons of 2012 and 2013 revealed fascinating information about this phase, such as the discovery of a number of female and animal figurines on the courtyard to the east of the two buildings (fig. 3). Those figurines were surprisingly intact and found together with two anthropomorphic statues that have close parallels with objects from Central Jordan (Petit 2009b: 145–153; cf. Daviau 2001: 322 and fig. 5). Domestic household equipment, a clay bulla, imported pottery, statues and clay figurines suggest a varied range of activities at Tell Damiyah during this phase. The exact nature of the buildings will be further investigated in 2014. After the fire, the inhabitants abandoned the site. A Neo-Babylonian seal suggests the presence of people during the late 7th and 6th century BC.

5. Persian-Hellenistic Period (5th–3rd century BC) After the prosperous Iron Age II-levels, the Central Jordan Valley experienced a decrease in settlements (e.g. Petit 2009a). This decline was most likely caused by a sharp drop in precipitation (e.g. Enzel et al. 2003; Migowski et al. 2006). But although all sites were abandoned people did return regularly to the valley. Excavations on sites in this area have revealed courtyard layers, oven remains, small ash lenses and dozens of pits. Especially these last features can be called the hallmark of this interlude. Tell es-Sa’idiyeh phase IV and IIIE-F (Pritchard 1985; Tubb 1988; 1990), Tell ‘Ammata phase 9–8 (Petit 2009a), Tell Deir ‘Alla phase IV/V (Van der Kooij and Ibrahim 1989; Ibrahim and Van der Kooij 1997; Van der Kooij 2006; Groot 2009; 2011), but also other sites in the Levant like Tell el-Hesi (Bennett/ Blakely 1989: 63–67) contain mudbrick-lined as well as unlined pits. Most of the fill consists of older settlement debris but the bottom was often covered with organic residues and

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

566

Petit

Fig. 4. Mudbrick lined pit.

exceptional high numbers of loomweights and spindle whorls. The organic content is generally interpreted as animal fodder. The excavations at Tell Damiyah revealed a similar phase with at least 21 pits of which six were lined with mudbricks (fig. 4; Petit 2014). The sizes of the pits differ considerably but most of them were several meters in diameter and had a depth of 1.5 meters. No complete pottery was found inside, making it hard to date each of them precisely. The many small bi-conical and spherical loom weights and a few Attic pottery sherds suggest a Persian or Hellenistic date (cf. Boertien 2013).

6. Byzantine Period (c. 5th century AD) After the Persian-Hellenistic Period the site was deserted for a while, although erosion process could have removed some of the occupation remains on top. Some Roman pottery on the surface at least suggests the presence of people during this period. During the Byzantine Period, Tell Damiyah was used as a cemetery. Several burials were excavated containing the skeletal remains of both adults and children (fig. 5). The orientation differs considerably and it was extremely difficult to date each of those burials since finds in situ were scarce. A glass bottle and some beads suggest a date in the middle of the 1st millennium AD.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Tell Damiyah. A Small Settlement Mound with a Remarkable International Role

567

Fig. 5. One of the burials.

7. Ottoman Period (15th–18th century) Most problematic in dating the Byzantine graves, was the fact that the site was probably reused as a cemetery in the Ottoman Period. At this stage it remains obscure which of the burials were Ottoman and which had a Byzantine date. The Ottoman graves include wooden structures over the skeletal remains and most of those later burials have cut the earlier remains. The excavation revealed that the Ottoman people reburied the earlier human remains on top of their deceased, some parts anatomically correct. Further research is necessary to solve this puzzle

8. Conclusion The excavations at Tell Damiyah will be continued in 2014 with the primary task to solve the questions about the two cemeteries and the intriguing international role of the site during the Late Iron Age. What was the function of this small site that the larger empires like Egypt and Assyria were interested in? Also the extend of the site – the upper and lower tell – will

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

568

Petit

be studied. Tell Damiyah might have been much larger than the settlement mound now visible above the valley floor. Geomorphological study in 2005 have proven that there exist at least several meters of clay deposits on top of the valley floor of the Iron Age.

Bibliography Albright, W. F., 1926. The Jordan Valley in the Bronze Age. Annual of the American Schools of Oriental Research 6, 13–74. Anastasio, S., 2010. Atlas of the Assyrian Pottery of the Iron Age. Subartu XXIV. Turnhout. Bennett, W. J./Blakely, J. A. (eds.), 1989. Tell el-Hesi. The Persian Period (Stratum V). Winona Lake. Boertien, J. H., 2013. Unravelling the Fabric. Textile Production in Iron Age Transjordan. Ph.D. diss., University Groningen. Daviau, P. M. M., 2001. New lights on Iron Age Religious Iconography. The evidence of Moab. Studies in the History and Archaeology of Jordan VII, 317–326. Enzel, Y./Ken-Tor, R./Sharon, D./Gvirtzman, H./Dayan, U./Ziv, B./Stein, M., 2003. Late Holocene Climates of the Near East Deduced from Dead Sea Level Variations and Modern Regional Winter Rainfall. Quaternary Research 60.3, 263–273. Glueck, N., 1951. Explorations in Eastern Palestine, IV. Part I: text. Annual of the American Schools of Oriental Research XXV–XXVIII. New Haven. Groot, N. C. F., 2009. The Early Persian Period at Tell Deir ‘Alla: a Ceramic Perspective. Kaptijn/Petit 2009: 167–180. –

2011. All the Work of Artisans: Reconstructing Society at Tell Deir ‘Alla Through the Study of Ceramic Traditions: Studies of Late Bronze Age Faience Vessels and Iron IIc–III Ceramics from Tell Deir ‘Alla, Jordan. Ph.D. diss., Delft University.

Guérin, V., 1869. Description géographique, historique et archéologique de la Palestine accompagnée de cartes détaillées. Ibrahim, M. M./van der Kooij, G., 1997. Excavations at Tall Dayr ‘Alla, Season 1987 and 1994. Annual of the Department of Antiquities of Jordan 41, 95–114. Irby, C. L./Mangles, J., 1823. Travels in Egypt and Nubia, Syria, and Asia Minor during the years 1817 and 1818. Kaptijn, E./Petit, L.P. (eds), 2009. A Timeless Vale. Archaeological and related essays on the Jordan Valley in honour of Gerrit van der Kooij on the occasion of his sixty-fifth birthday. ASLU 19. Leiden. Kaptijn, E./Petit, L./Grootveld, E./Houran, F./Kooij, G. van der/al-Ghul, O., 2006. Dayr ‘Alla Regional Project: Settling the Steppe (First Campaign 2004). Annual of the Department of Antiquities of Jordan 49, 89–99. Kitchen, K. A., 1973. The Third Intermediate Period in Egypt (1100–650 B.C.). Warminster.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Tell Damiyah. A Small Settlement Mound with a Remarkable International Role

569

Kooij, G. van der, 2001. The Vicissitudes of Life at Dayr ‘Alla during the First Millennium BC, Seen in a Wider Context. Studies in the History and Archaeology of Jordan VII, 295–303. –

2006. Tell Deir ‘Alla: The Middle and Late Bronze Age Chronology. P. M. Fischer (ed.), The Chronology of the Jordan Valley during the Middle and Late Bronze Ages: Pella, Tell Abu al-Kharaz, and Tell Deir ‘Alla. Wien, 199–226.

Kooij, G. van der/Ibrahim, M. M. (eds.), 1989. Picking up the Threads. A continuing Review of Excavations at Deir ‘Alla, Jordan. Leiden. Lynch, W. F., 1849. Narratives of the United States Expedition to the River Jordan and the Dead Sea. Philadelphia. MacDonald, B., 2000. East of the Jordan. Territories and Sites of the Hebrew Scriptures. Boston. McGarvey, J. W., 1881. Lands of the Bible: A Geographical and Topographical Description of Palestine, with Letters of Travel in Egypt, Syria, Asia Minor, and Greece. Philadelphia. Mellaart, J., 1962. Preliminary report on the archaeological survey in the Yarmouk and Jordan Valley for the Point Four Irrigation Scheme. Annual of the Department of Antiquities of Jordan 6–7, 126–157. Migowksi, C./Stein, M./Prasad, S./Negendank, J. F. W./Agnon, A., 2006. Holocene Climate Variability and Cultural Evolution in the Near East from the Dead Sea Sedimentary Record. Quaternary Research 66, 421–431. Petit, L. P., 2008. Late Iron Age Levels at Tell Damiyah: New Excavation Results from the Jordan Valley. J. M. Córdoba et al. (eds.), Proceedings of the 5th International Congress on the Archaeology of the Ancient Near East, Madrid, April 3–8 2006. Madrid, 177–187. –

2009a. Settlement Dynamics in the Middle Jordan Valley during Iron Age II. BAR International Series 2033. Oxford: Archaeopress.

–

2009b. A wheel-made anthropomorphic statue from Iron Age Tell Damieh, Jordan Valley. Kaptijn/ Petit 2009: 151–160.

–

2013. The Jordan Valley. D. Master (ed.), The Oxford Encyclopedia of the Bible and Archaeology, Volume 2. New York, 47–53.

–

2014. Understanding the ‘Pit people’. An imaginary conservation in the Central Jordan Valley during the late 7th or 6th century B. C. E. van der Steen et al. (eds.), Exploring the Narrative. Jerusalem and Jordan in the Bronze and Iron Ages. London 171–179.

–

i. p. Recycling the Valley. Preliminary results of the excavations at Tell Damiyah, 2013. Annual of the Department of Antiquities of Jordan 57.

Petit, L. P./Kaptijn, E. /Hourani, F. /al-Ghul, O. /Grootveld, E./Kooij, G. van der, 2006. Dayr ‘Alla Regional Project: Settling the Steppe. Second Campaign 2005. Annual of the Department of Antiquities of Jordan 50, 179–188. Pritchard, J. B., 1985. Tell es-Sa’idiyeh: Excavations on the Tell 1964–1966. Pennsylvania University Monograph 60. Philadelphia. Tubb, J. N., 1988. Tell es-Sa’idiyeh: preliminary report on the first three seasons of renewed excavations. Levant 20, 23–88.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

570 –

Petit 1990. Preliminary report on the fourth season of excavations at Tell es-Sa’idiyeh in the Jordan Valley. Levant 22, 21–42.

Velde, C. W. M. van der, 1854. Narratives of a Journey through Syria and Palestine in 1851 and 1852. Edinburgh-London. Yassine, K. (ed.), 1988. Archaeology of Jordan. Essays and Reports. Amman. Yassine, K./Sauer, J./Ibrahim, M. M., 1988. The East Jordan Valley Survey II. K. Yassine (ed.), The Archaeology of Jordan. Essays and Reports. Amman 189–207.

Lucas Petit, The Netherlands, National Museum of Antiquities, Leiden.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9th ICAANE, Basel 2014, Vol. 3, 571–583

Andrea Polcaro – Juan Muñiz – Valentín Alvarez

The New Spanish-Italian Expedition to the EB I site of Jebel al-Mutawwaq, Middle Wadi az-Zarqa, Jordan: Preliminary Results of the 2012–2013 Campaigns The paper summarizes the main results of the 2012–2013 Spanish-Italian archaeological excavations of the site of Jebel al-Mutawwaq, located on the top of a mountain along the Middle Wadi az-Zarqa, Jordan. The site is characterized by an Early Bronze Age I settlement of about 13 hectares and by a large dolmen field.

1. Introduction to the Site1 Jebel al-Mutawwaq is a site located on the Middle Wadi Zarqa, 11 km south from Jerash, at the confluence between Wadi az-Zarqa and Wadi Hmeid, near the spring of Qreisan, actually in the modern village of Quneye. It was first surveyed by Hanbury-Tenison (1986; 1989a), which first identified, along the southern cliff of the mountain, a village of the beginning of the Early Bronze Age of about 13 hectares encircled by a stone settlement wall and characterized by houses of apsidal shape.2 He noted also the presence of hundreds of dolmens around the village and scattered all over the mountain. From 1989, for more than twenty years the site was excavated by a Spanish mission led by the late professor Juan Antonio Fernandez-Tresguerres Velasco. Two main important discoveries were made by Tresguerres excavating the village of Mutawwaq.

1 2

By Andrea Polcaro, Juan Muñiz, Valentín Alvarez. The problem of the terminology of the EB I house architecture was deeply investigated by E. Braun (1989).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

572

Polcaro – Muñiz – Alvarez

First, he excavated about 5% of the 300 houses,3 identifying the principal architectonical features of domestic buildings: the east west orientation following the slope of the mountain, the entrance made of megalithic stones, the hearth of the house located on one limit.4 The second main discovery was a cultic complex denominated «Temple of the Serpents». The main building of the sanctuary is House 76. Inside this complex cultic vessels decorated with snakes, similar to that discovered in later Early Bronze Age II–III sites like Khirbet ez-Zeraqon or Khirbet al-Batraway were also discovered in the main building.5 Flint blades and tubular scrapers were discovered in the minor House 75 and in the small rooms near altar probably used for animal sacrifices and for processing the food, perhaps during important festivities of the community (see Fernandez-Tresguerres 2008 and Polcaro et al. 2014). All the pottery recovered by Fernandez-Tresguerres was attributed to the first part of the Early Bronze Age, and in different articles he spoke about their similarities with some Ghassulian vessels, enforcing the hypothesis of dating the village to the EB IA (see Fernandez-Tresguerres 2005). The only C14 data available from the House 76 in the cultic complex point to a period between 3300 and 3200 BC.6 The Spanish excavation stopped in 2011 after the death of Fernandez-Tresguerres.

2. The New Campaigns of Excavations7 In 2012 the Perugia University, together with the Pontificia Facultad San Esteban of Salamanca, restarted a Spanish-Italian archeological excavation to this site, now co-directed by Juan Muñiz and Andrea Polcaro. In the 2012–2013 campaigns six dolmens located in the area of the megalithic necropolis were excavated (fig. 1), just outside the eastern village border (Area B). Another area along the eastern side of the settlement wall was opened, near the southeastern door of the village, leading to the necropolis (Area A).

3

4 5 6

7

Only for 182 houses it was possible to recognize the plan; the Spanish expedition recognized three main sector in which the settlement was divided along the East-West axis: The Western Sector with 84 houses, the Central Sector with 71 houses, and the Eastern Sector with 22 houses (see Muñiz et al. 2013). The houses have the base of the walls made of large stone blocks, located directly on the bedrock (see Muñiz et al. 2013). See for the sherds discovered in Zeraqon: Ajlouny et al. 2011: fig. 24:2; for Khirbet al-Batrawy: Nigro/Sala 2012: fig. 5. C14 analysis was performed on animal bones discovered inside a pit in the northern part of the sacred building. The results are 5290–5040 BP = 3340–3090 b.c.e. (Beta Analytic 194526) and 5270–5170 BP = 3320–3220 b.c.e. (Beta Analytic 194527). By Andrea Polcaro, Juan Muñiz.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The New Spanish-Italian Expedition to the EB I site of Jebel al-Mutawwaq … Jordan

573

Fig. 1. 2012–2013 excavation areas of Jebel al-Mutawwaq, from North.

The first objective of the excavations was to clarify the chronological relationship between the sectors of the EB I village, the stone settlement wall, the Temple of the Serpents and the dolmen field. In the 2012 excavation campaign, a first result concerning the original aspect and the construction technique of the megalithic structures was achieved. Looking at the hundred of dolmens it seems that they did not have originally any covering tumulus, but only circles of stones, sometimes called «fence».8 However, the archaeological sections of the first dolmens excavated, nos. 228, 232 and 318, proved that there was in origin a stone layer filling the space between the external stone circle and the lateral slabs of the dolmen (fig. 2). The circular stone fence was thus the retaining wall of a tumulus, originally covering the whole structure to the covering slab, the last stone posed (Alvarez et al. i. p.). The 2013 excavation to two well preserved dolmens, dolmen no. 317 and dolmen no. 321, definitively proved this interpretation (fig. 3). These two dolmens were recovered perfectly intact with the tumulus covering all the structure to the capstone. Looking at dolmen no. 317 it is clear that there was a typical shape of the retaining wall; the left wall was straight and the right wall curved, assuming an apsidal shape perhaps functional to the erection of the covering tumulus (Polcaro et al. 2014). 8

This kind of feature was observed in many other dolmen fields in Jordan (see Polcaro 2013).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

574

Polcaro – Muñiz – Alvarez

Fig. 2. Sections of dolmens excavated in 2012 in Jebel al-Mutawwaq: a. no. 228; b. no. 232; c. no. 318.

The excavations of the six dolmens in these two first campaigns identified also a difference in the entrance architecture of the megalithic tombs: some of them had in fact a dromos entrance, obtained with three steps made of large flat stones (Polcaro et al. 2014: fig. 11). These features seem to be not related to the dimension of the dolmens, but to their topographical position, in particular when the structure, oriented toward north, was built between two different levels, following the natural slope of the bedrock. A second result achieved in the 2013 excavation campaign concerns the connection of the dolmen field with the settlement. In Area A (fig. 4a), between the eastern wall of the village and the nearby dolmen area, a street was discovered clearly in connection with a south-eastern door of the EB I settlement (fig. 4b). This street, called L. 5, was made of two lines of large stones and a filling inner layer of pebbles and compact earth. It connected the eastern sector of the village with the necropolis, as it is shown by the other part of the same street, called here L. 21, discovered in Area B in stratigraphical connection with some of the excavated dolmens (Polcaro et al. 2014: figs. 8–9). In fact, the excavation of Area B had shown that first, connected with the southern dolmen no. 316, there was the street L. 21, then dolmen no. 317 was built in a crossroad between L. 21 and another street running from south

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The New Spanish-Italian Expedition to the EB I site of Jebel al-Mutawwaq … Jordan

575

Fig. 3. Dolmen no. 321 with the intact tumulus in place, before the excavation of the burial chamber.

to north, called L. 22, and finally dolmen no. 321 was erected cutting the street L. 21, clearly no more in use. This indicates perhaps the continuity of the use of the necropolis also after the abandonment of the village. In particular, the construction of the last dolmen suggests that the stratigraphy of the relationship between the necropolis and the village could be more complicated and that in the long period of the Early Bronze Age I there were different phases of use of the dolmens by the permanent settlers of the village and the seasonal shepherds, occasionally using the same area. The investigation of Area A was also useful to achieve another objective, the better understanding of the nature of the settlement wall. The preparation of the wall construction was obtained by leveling the natural slope of the bedrock with layers of small stone and pebbles. Then large flat stones were posed vertically one against the other above the foundation, apparently without any kind of superimposing structure (fig. 5). The height of the wall probably did not reach more than 1.50 meter and in the south-eastern corner of the fence there are no signs of towers or defensive features. It is thus possible that the settlement wall had no military function9, but could have had an ideological meaning of separating the land of 9

Hypothesis yet proposed by Hanbury-Tenison (1989b). About the interpretation of the Jebel alMutawwaq settlement wall see also Philip 2008: 170.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

576

Polcaro – Muñiz – Alvarez

Fig. 4. a. Plan of Area A, excavated in 2013. b. The Southeastern settlement door, from West.

the living from the land of the dead. Comparing the part of the settlement wall excavated in Area A with the other portions of the same wall excavated by the past Spanish expeditions in the southern and western borders of the settlement, it is possible to affirm that the fence was built with different methods of construction. In particular in the southern sector, running along the southern cliff of the mountain, the settlement wall was built in its upper part with more squared regular stone blocks, similar to the foundation of the house walls (Muñiz/Polcaro/Alvarez 2013). The south-eastern door of the settlement wall discovered in 2013 excavations is similar to the other doors identified along the other sectors of the fence,

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The New Spanish-Italian Expedition to the EB I site of Jebel al-Mutawwaq … Jordan

577

Fig. 5. East-West section of Area A: in the center is visible the North-South (W. 2) and the East-West (W. 7) parts of the settlement wall.

built with two large stone slabs, about one meter wide. Furthermore, during the first survey of the EB I village, Hanbury-Tenison (1989a: fig. 4) discovered a larger door, called the Southern Gate, not related to a private building of the village but to another wall separating the eastern sector of the village from the central one. In fact, this door was located inside the eastern part of the settlement, about 130 m from the southern door identified in 2013, and it is possible to identify an inner path between the two gates of the village. The presence of inner divisions and the difference in the construction technique of the eastern and southern part of the settlement wall suggest that this fence was built in several chronological phases, perhaps following the expansion of the Early Bronze Age I village. The third important result achieved concerns the function of the dolmens that, in Jebel al-Mutawwaq, were surely used as tombs: compared with the houses identified in the settlement, the large number of megalithic monuments suggests that dolmens were used for the same purpose not only by the population of the nearby village, but also by the entire population of the countryside. The inner stratigraphy of the megalithic chamber of all the dolmens excavated till now (nos. 228, 232, 318, 316, 317, 321) presents layers of stone and earth filling completely the burial space. This layers represent the last sealing phase of the structure. In this upper layers just a few EB I pottery sherds and some flints were recovered, but in the deeper layer, directly above the floor slab, small fragments of human bones like teeth, finger bones and ribs, were recovered in all the dolmen excavated. This indicates that originally the dead were deposed, decomposed and rearranged inside the burial chamber. Then, at the end of the use of the dolmen, all the most important human remains, like the long bones and the skulls, were removed together with the funerary gifts. Finally, the burial chamber was sealed forever with layers of stones and earth.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

578

Polcaro – Muñiz – Alvarez

Furthermore, the final prove came in the 2013 campaign with the discovery of an entire burial (denominated B. 25) left inside dolmen no. 317 (Polcaro et al. 2014: fig. 17). The burial was hidden in the lower layer excavated inside the burial chamber of the dolmen through a flat stone separating the lower part of the burial chamber. Inside this lower layer, Stratum 51, an entire human skeleton was found (Polcaro et al. 2014: fig. 12). The bones of the burial, after the decomposition of the body, were re-arranged inside the chamber. Another concentration of fragmentary human bones (denominated B. 26), belonging to at least other five people, was discovered pushed on the back of the chamber (Polcaro et al. 2014: fig. 13). These burials were deposed in dolmen no. 317 and then their skulls and long bones were moved to another location: the only dead left in the burial chamber was burial B. 25. No vessels have been found with this burial, but just two objects have been discovered in 2013: two flat flint scrapers, with double blade and the cortex visible on one side, recalling the so-called fan scrapers typical of the Late Chalcolithic and Early Bronze Age periods.10 In the Temple of the Serpents similar flint tools were recovered in House 75 near an altar (Polcaro et al. 2014: fig. 6). This indicates that these objects were perhaps used in ritual context, probably for butchering or working animal skins. The anthropological analysis made on the bones of burial B. 25, showed that he was in very good health condition, no disease, a balanced alimentation and no muscular stress, for an age of 30–40 years (Polcaro et al. 2014). Moreover, a particular feature was observed on the skull: on the left parietal, in the ectocranial side, a hole is visible with a triangular section, ending in a small hole in the endocranial side of the skull (Polcaro et al. 2014: fig. 18). Due to the absence of the cranial bone growth around it, the wound could be interpreted as perimortem or postmortem. We are carrying out further analysis on the skull of B. 25, in order to find a flint object or weapon consistent with the hole in the cranium.

3. The Pottery of Jebel al-Mutawwaq11 In April 2014, we started to study and draw all the pottery coming from the past excavations of the Oviedo expedition to the Temple of Serpents and still unpublished. We started with ten pottery boxes coming from House 76. In the future campaigns we will continue to study the other pottery materials coming from the other structures of the sacred area. It is now possible to compare provisionally the pottery coming from the necropolis with the ones coming from the village and from the sanctuary: the pottery of Jebel al-Mutawwaq is 10 Objects very similar to these two finds were recovered as funerary gifts in cave tombs of the western

southern Levant, and in Jordan in the EB II–III temples (see Polcaro et al. 2014: figs. 14–15).

11 By Andrea Polcaro, Juan Muñiz.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The New Spanish-Italian Expedition to the EB I site of Jebel al-Mutawwaq … Jordan

Fig. 6. Examples of painted, incised and applied decoration on pottery sherds from House 76 of the Temple of the Serpents.

Fig. 7. Pottery shapes from the Temple of the Serpents, Area A and Area B of Jebel al-Mutawwaq. a. Holemouth jars. b. Bowls.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

579

580

Polcaro – Muñiz – Alvarez

generally rough, low fired, hand made, with red-orange paste. In the necropolis we found just red slip as surface treatment but in Area A inside the village and in the temple area also a heavy whitish slip was recovered on some sherds. Only in the temple we registered painted red decoration on the surface of some vessels, mostly wave lines or parallel vertical lines (fig. 6.1). Incised or impressed decoration and applied roped decoration are common in all the contexts (fig. 6.2–3). Just in the temple area a lot of sherds presenting applied knobs of different shapes have been recovered (fig. 6.4). Also typical of the temple area are decorations of snakes applied to the neck and the body of the vessel, and pointed band impressed decoration in parallel vertical lines, recovered on spouted bowls and hole-mouth jars (fig. 6.5–6). The most common pottery shape in the village and in the temple is hole-mouth jars (fig. 7a), only sporadically discovered in the necropolis. This shape, common in all the Early Bronze Age periods, is not chronologically relevant. As to the bowls (fig. 7b), the most typical shape is the hemispherical bowl with rounded or pointed rim, recovered in all the areas of investigation. In the temple area bowls, slightly carinated, with everted pointed rim and sometime ring loop handle, are also present. This typology is not present in Area A, but it is present in Area B connected to the dolmens, so it is perhaps linked to a more cultic function. Big storage jars, with ledge handle are common in all the area investigated. In the temple area, as well as in the eastern sector of the village and in the necropolis, the ledge handles are of the simple plain typology, typical of the EB I, without any wavy shape typical of the following EB II–III periods or pushed up ledge handle of the late EB III (fig. 8a). These storage jars usually have everted rims and present often a roped applied decoration on the neck (fig. 8b). Another jar typology is the one with ribbed everted plain rim: also this typology is present just in the temple area and in the necropolis (fig. 8c), but further investigations in Area A in the following campaigns could change this presence. Also this shape could be inserted in the EB I period. Another typology of vessels typical of the EB I are jars with simple flat vertical rim and wide neck, till now recovered just in the temple area. The last typology of vessels recovered in the temple are decorated jugs with loop handles, often painted with red band decorations. In conclusion, all the pottery recovered till now from the village and the necropolis in the shape and in the surface treatment have good parallel with the pottery of the EB I period from Jebel Abu Thawwaq, Jawa and Tell Umm Hammad.12 The absence of Grey Washed

12 See in particular the pottery from Jebel Abu Thawwab (Douglas/Kafafi 2000: 105–108, figs. 6.2, 6.4,

6.5). Some shapes and decorations, like the hole-mouth jars with pushed-up lug handles and pointed impressed decoration are typical from Jawa in the eastern Black Desert to Tell Umm Hammad; they were also recovered along the Wadi az-Zarqa valley in EB I sites (see Philip 2008: fig. 6.8:3).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The New Spanish-Italian Expedition to the EB I site of Jebel al-Mutawwaq … Jordan

581

Fig. 8. Pottery shapes from the Temple of the Serpents, Area A and Area B of Jebel al-Mutawwaq: a. Plain ledge-handle. b. Jars with everted rounded or flat rims. c. Jars with ribbed everted plain rim.

Ware points to ascribe Jebel al-Mutawwaq in the EB I culture of central Transjordan.13 However, it is not clear yet if the site continued to be used in EB II, even if reduced in dimension. Further excavation areas inside the village and the necropolis, and further surveys around the mountain will be done to better identify the chronological extension of the site occupation, related to the reuse of the ruins of the village or of isolated dolmens.

13 The presence of the Grey Burnished Ware is considered a distinctive trait of the EB IA northern

culture of Galilee and Western Palestine and, even if present in some sites of Northern Jordan, it is uncommon in the eastern Jordanian highlands and in the southern desert during the same chronological phase (see Philip 2008: 167).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

582

Polcaro – Muñiz – Alvarez

4. Further Considerations14 The Spanish-Italian excavation at Jebel al-Mutawwaq in 2013 definitively proved that Hanbury-Tenison was wrong affirming the dolmen field and the village are unlike to be contemporary (1989a). The presence of pattern of streets connecting the village and the megalithic monuments proved that, at least for some time, the two areas were connected and the large necropolis was used also by the EB I inhabitants of the nearby settlement. Moreover, it is possible to say that the dolmens scattered on the mountain and in particular the ones on the eastern border of the village, were funerary monuments used as family tombs by the Jebel al-Mutawwaq population. The study of the dolmens architecture and of the typology of these megalithic constructions it is also interesting. The discovery of the presence of a tumulus covering the dolmens connects this kind of tombs to the Southern Levant megalithic funerary tradition: in particular, dolmens appear to be the natural evolution of the Late Chalcolithic cairns and cists attested near Tuleilat al-Ghassul and Bab edh-Dhra. These kinds of tombs are considered a funerary costume typical of the eastern side of the Jordan Valley, as distinctive from the ossuary and cave tombs tradition of the western region of Palestine. In the changed landscape and society of Early Bronze Age I the Chalcolithic underground cist was taken above the soil, and in the dolmens the megalithic burial chamber is enlarged and built above the ground. Also the tumulus covering the monument thus assumes a different shape, with the introduction of a dromos leading inside the burial chamber. The architectonical changes testify to a different ideology pointing to physically show the presence of the tomb on the landscape. Moreover, the enlargement of the megalithic chamber for burials also testifies to the transformation of the funerary monument in a family tomb, linking a clan to a landscape in a striking and evident way. The concentrations of dolmen fields in the Southern Levant along the main rivers and lakes could suggest also different scenarios of relationship between the new Early Bronze society and economy with the surrounding water landscape; further excavations and surveys must delineate too this aspect of the EB I megalithic tradition.

Bibliography Ajlouny, F./Douglas, K./Khrisat, B., 2011. Spatial Distribution of the Early Bronze Clay Figurative Pieces from Khirbet ez-Zeraqon and its Religious Aspects. Ancient Near Eastern Studies 48, 88–125.

14 By Andrea Polcaro, Juan Muñiz.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

The New Spanish-Italian Expedition to the EB I site of Jebel al-Mutawwaq … Jordan

583

Alvarez, V. A./Muñiz, J. R./Polcaro, A., i. p. Preliminary Results of the First Spanish-Italian Excavation Campaign to the Jebel al-Mutawwaq Dolmen Field, August–September 2012. Annual of the Department of Antiquities of Jordan 57. Braun, E., 1989. The Problem of the Apsidal House: New Aspects of Early Bronze I Domestic Architecture in Israel, Jordan and Lebanon. Palestine Exploration Quarterly 121, 1–43. Douglas, K./Kafafi, Z., 2000. The Main Aspects of the EB I Pottery from Jebel Abu Thawwab, North Jordan. G. Philip/D. Baird (eds.), Ceramics and Change in the Early Bronze Age of the Southern Levant. Sheffield, 101–111. Fernandez-Tresguerres, J. A., 2005. Jabal Muṭawwaq. Annual of the Department of Antiquities of Jordan 49, 365–372. –

2008. The “Temple of the Serpents”, a Sanctuary of the Early Bronze Age I in the Village of Jabal al-Muṭawwaq (Jordan). Annual of the Department of Antiquities of Jordan 52, 23–34.

Hanbury-Tenison, J. W., 1986. The Late Chalcolithic to Early Bronze I Transition in Palestine and Transjordan. BAR International Series 311. Oxford. –

1989a. Jabal Mutawwaq 1986. Annual of the Department of Antiquities of Jordan 33, 137–144.

–

1989b. Desert Urbanism in the Fourth Millennium. Palestine exploration Quarterly 121, 55–63.

Muñiz, J. R./Polcaro, A./Alvarez, V. A., 2013. Jebel al-Mutawwaq: La evolución del estudio de un yacimiento de la Edad del Bronce Antiguo I en la estepa jordana. ISIMU: Revista sobre Oriente Próximo y Egipto en la antigüedad 16, 79–95. Nigro, L. /Sala, M., 2012. Preliminary Report on the Seventh (2011) Season of Excavation by “La Sapienza” University of Rome at Khirbet al-Batrawy, Upper Wadi az-Zarqa. Annual of the Department of Antiquities of Jordan 56, 45–53. Philip, G., 2008. The Early Bronze Age I–III. R. B. Adams (ed.), Jordan. An Archaeological Reader. London, 161–226. Polcaro, A., 2013. The Stone and the Landscape: The Phenomenon of Megalithic Constructions in Jordan in the Main Historical Context of Southern Levant at the Beginning of the 3rd Millennium bc. L. Bombardieri et al. (eds.), SOMA 2012: Identity and Connectivity; Proceedings of the 16th Symposium on Mediterranean Archaeology, Florence, Italy, 1–3 March 2012, Vol. 1. BAR International Series 2581. Oxford, 127–135. Polcaro, A./Muñiz, J. R./Alvarez, V. A./Mogliazza, S., 2014. Dolmen 317 and Its Hidden Burial: An Early Bronze Age I Megalithic Tomb from Jebel al-Mutawwaq (Jordan). Bulletin of the American Schools of Oriental Research 372, 1–14. Andrea Polcaro (Università degli Studi di Perugia). Juan Muñiz (Pontificia Facultad San Esteban de Salamanca). Valentín Alvarez (Association Professional of Independent Archaeologist of Asturias).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 585–597

Suzanne Richard

Recent Excavations at Khirbat Iskandar, Jordan. The EB III/IV Fortifications. Although known primarily for its EB IV (2500–2000 BCE) multi-phase occupational stratum, Khirbat Iskandar also includes a multi-phase EB III stratum. This paper focuses on the construction history of the EB III fortifications at the site, including evidences for reuse and rebuilding of the fortifications in the EB IV period.

Introduction This article concentrates on the Early Bronze III settlement at Khirbat Iskandar, especially the construction history of the fortifications, as revealed in Area B at the northwest corner of the mound. The discovery in 2013 of a new fortification line is, we believe, the missing link to unraveling the phasing of the fortifications at the site. As the chronology of the fortifications has come into sharper focus, so has the dating of the various defensive lines, thanks to another important discovery in 2013 – a substantial wall of early EB III date – found to run under the fortifications. This new piece of evidence appears to resolve the question of the date of the inner wall, previously thought to be EB II. As the discussion will show, the length of the EB III occupation at the site of Khirbat Iskandar is more extensive than indicated in previous seasons. Coupled with investigations of the northwestern perimeter defenses, the expedition has sought to expose more of the EB III occupational levels in Area B in order to grasp the nature of the urban settlement preceding the well-known EB IV settlements. Work over the past three seasons allows us to contextualize and tentatively interpret the activities in the EB III settlement. Finally, a series of stratigraphic links to the fortifications now provides the dataset required to revisit the controversial topic of the existence of fortifications in the EB IV period. Although the expedition has uncovered EB IV reuse of the fortifications – even Nelson Glueck observed this in his survey (1939) – the question of whether or not the fortifications

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

586

Richard

were rebuilt in EB IV has long been outstanding. Our excavations of the wonderfully preserved and extensive settlements including, e.g., a gateway, a public building, storeroom, etc. have demonstrated that there is a high level of complexity in the rural EB IV period. For more information on the EB IV period at Khirbat Iskandar, see recent field reports and bibliography cited there (Richard et al. i. p.; Richard, et al. 2010; Richard/Long 2005). A word on the currently understood phasing at the site is in order prior to discussion of the fortifications. In Area B – the primary concern of this article – there are thus far five well-defined phases: A – B are EB IV; C – E are EB III. From the top, there are two major EB IV settlements (with sub-phasing): Phase A was built immediately above the destroyed Phase B settlement. There is a destruction level separating Phase B from the EB III Phase C1 settlement. Below is the founding Phase C2 settlement. [Phase C includes the outer EB III fortifications.] Phase D, known primarily as the earlier stage of fortifications [the inner wall] now appears to be EB III (not EB II). This suggested revision derives from a newly discovered wall (mentioned above) running under the fortifications, whose diagnostic ceramics were early EB III. The latter discovery calls into question the provisional dating (EB I/II) originally given to two superimposed fragmentary walls found to run under the EB III northwest tower, tentatively identified as Phases E-F (Richard 1990). With future excavation, more exposure should clarify the site’s occupational history preceding the EB III Phases C–D fortifications. Moreover, we are mindful of Peter Parr’s work on the eastern side of the mound, where he appears to have uncovered evidence for an early Early Bronze Age defensive wall in his Period 1 (1960: 129).

The Site The 2.7 ha site of Khirbat Iskandar sits on the north bank of the Wadi al-Wala straddling the «King’s Highway», and lies some 19 km south of Madaba and ca. 10 km north of the Wadi Mujib (fig. 1). As Glueck (1939) noted, Khirbat Iskandar comprises both a «tell» and to the east a «lower» mound approximately the same size on the other side of a narrow wadi. Glueck also described occupational remains on the plain north of the site and across the wadi to the south. If we include the lower mound, the «site» would be at least 5.4 ha or more if we add the area to the north (and maybe to the south), perhaps expanding the original «site» to at least 6 ha (or around 15 acres). Factors such as the perennial wadi stream, the proximity of the «King’s Highway», and a fertile floodplain all combined to support a prosperous, fortified Early Bronze Age agricultural site until about mid-3rd millennium BCE (see Cordova/Long 2010: 21–35; Cordova 2007: 141, Fig. 5.7). As the latter geomorphological and environmental studies have shown, the onset of dryer conditions, as evidenced by erosion of the wadi and subsequent

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Recent Excavations at Khirbat Iskandar, Jordan.

587

Fig. 1. Map showing the location of Khirbat Iskandar, Jordan, on the Wadi al-Wala. Looking North.

diminishment of the floodplain, was a major factor in the de-urbanization process at this site. Such conditions, not to say probable human overuse of the land, were the likely mechanisms operating to reduce the carrying capacity of the site in the EB IV, and to render difficult – if not impossible – its continued occupation into the Middle Bronze Age, despite numerous indications (gate, pottery, tin-bronze, rural complexity and village economy, etc.) linking the site with the subsequent age (Richard 2006; Richard et al. 2010 eds; D’Andrea 2014). Khirbat Iskandar was abandoned ca. 2000 BCE. This data set of climatic and environmental variables informs our research design as we study the rise and collapse of urbanism at Khirbat Iskandar. Significantly, the data set (and 14C dates) likewise resonates with the new, higher chronology, in which the EB III/EB IV nexus is ca. 2500, rather than 2300 BCE (Regev et al. 2012; Höflmayer 2014).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

588

Richard

Fig. 2. Plan showing excavation areas at Khirbat Iskandar. Looking north.

The Fortifications In Area B (fig. 2), the most significant discovery of the 2013 season was a new fortification wall on the west (B2A120/B4A006). It came to light during excavation of a section against W. B2A005, the remains of the western perimeter wall (hereafter, the «rubble» wall, and see below), for the purpose of determining whether underlying Phase D curved structure (B2A077) was a tower. As fig. 3 shows, the latter curved wall does continue for a short distance, but was apparently cut by the newly discovered (Phase C) defensive line. The season ended before we could determine the exact nature and extent of the Phase D curved wall. The new wall line (B2A120/B4A006) is earlier and far more substantially built than the «rubble» wall abutting the tower at its southwest corner. The new wall abuts the tower at its northwest corner (fig. 4). As alluded to earlier, the new fortification line provides the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Recent Excavations at Khirbat Iskandar, Jordan.

589

Fig. 3. Khirbat Iskandar northwest fortifications: Phase C «bastion/tower» (top), «rubble wall» overlying curvilinear Phase D wall, and new fortification line at left. Looking northeast.

missing link needed to fully comprehend the construction history of Khirbat Iskandar’s system of defenses. The new wall also provides clarification to the somewhat mystifying series of parallel wall lines observed along both the eastern and western perimeters and slopes of the site.

Previously Exposed Fortifications Prior to the 2013 season, the assessment of the fortification system at the site consistently held that there were two phases to the defenses, as observed on the north: an early inner mudbrick/stone wall (Phase D) that curved at the northwest corner; a parallel later outer stone wall (Phase C) that terminated in what appeared to be a square tower at the northwest corner (fig. 4). There was a rubbly fill between the two wall traces. The western trace of the Phase C fortifications appeared to be the «rubble» wall that abutted the tower. Exposure of additional architectural components confirmed the expedition’s assessment of two phases of wall traces at the site of Khirbat Iskandar. The phasing of the two defensive lines is based on a number of stratigraphic pieces of evidence. A section (fig. 5) through the outer and inner traces on the north revealed that the outer (Phase C) line appeared to buttress the inner (Phase D), which was slumped thirty degrees inward suggestive of a collapse (see Richard/Boraas 1984: fig. 14; Richard 1990;

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

590

Richard

Fig. 4. Sketch (not to scale) of fortifications: dark grey (Phase D), light grey (Phase C), light grey outline («rubble» wall). Looking north.

fig. 5). There is evidence that Phase D was destroyed. The two distinctly constructed wall lines combined to provide a 3.0m wide defensive system in Phase C, strengthened at the northwest corner to 6.5m with the construction of a bastion/platform (see below). The «rubble» wall was a 3.0m wide defensive line on the western perimeter of the site. The wall was traced southward to Area A. When sectioned, this sloping wall proved to consist of three elements: a one-row 1.65m high inner wall (W. B2024B), an outer wall of 2–3 courses and 2 rows (B2A005), and, primarily, a rubble interior (B2A005A) built up around various discernible wall segments (fig. 6). A rough abutment to the tower, including a massive monolithic stone, gave the appearance of a block (Richard/Long 2005: fig. 4). Despite the remarkable difference in construction between the northern and western traces, the «rubble» wall was the only candidate for a Phase C EB III western perimeter wall. This assessment is coming into serious question in light of the 2013 season. Finally, the Phase D system, besides the inner line on the north, included two juxtaposed curved or curvilinear walls (B2108/B2A077) separated by a 1.0.m wide stone threshold-like feature (figs. 3–4). We believe they are towers. Additionally, a stretch of a well-preserved

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Recent Excavations at Khirbat Iskandar, Jordan.

591

Fig. 5. Khirbat Iskandar fortifications: collapsed inner Phase D wall on right, outer Phase C wall on left, rubble in-between. Looking east.

mudbrick/stone wall (B5A043) came to light just to the south (see Richard et al. i. p., fig. 9). It is the only candidate for a Phase D western perimeter wall trace. The stratigraphic evidence for the mudbrick/stone wall being earlier is significant. Besides its destruction and collapse mentioned earlier, there are several examples of a Phase C stone buttressing other than on the north. On the west, there is a stone «pier» buttressing curvilinear W. B2081 (Richard/Long 2005: fig. 9). There is a stone buttress against the possible western trace wall, B5A043 as well. Moreover, as fig. 3 shows, the newly discovered Phase C wall (B2A120/B4A006) cuts Phase D curvilinear W. B2A077, which, moreover, is superimposed by the «rubble» wall (see further below).

A New Look at the Northwest Corner Fortifications and Settlement New wall B2A120/B4A006 is clearly the original Phase C western perimeter wall constructed contemporaneously with the northern outer wall trace (see Long et al. 2014). The substantial (2.0m wide) wall, of a construction technique similar to the northern outer (Phase C) wall, was founded some 1.50m lower than the «rubble» wall. Indeed, as shown in Fig. 3, the bottom course of the «rubble» wall and upper course of W. B2A120/B4A006 are roughly coeval. One can now explain the disparity between the «rubble» wall and the northern Phase C line: the «rubble» wall is a later rebuild of the western defenses, apparently after W. B2A120/B4A006 had gone out of use. This clarification of the phasing of the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

592

Richard

Fig. 6. Khirbat Iskandar «rubble wall», looking east. Fig. 7. Khirbat Iskandar Phase C central room, platform to right, looking west.

various wall traces likewise clarifies the nature of the «tower» at the northwest corner, and, as discussed below, the question of EB IV fortifications. Prior to the discovery of the new Phase C fortification, our view of the northwest defenses was of a tower projecting beyond the curtain wall, given the abutment of the «rubble wall» against its southwest corner. As is now evident, originally the Phase C population erected an interior bastion/platform at the northwest corner, encompassing the earlier destroyed Phase D fortifications. A four-step stairway leads up to the «platform» and a guardroom, indicated by what appears to be an interior space (see Richard/Boraas 1984: figs. 12–13). The stairway widens the platform, which extends eastward for 10.0m up to a transverse wall (fig. 4). As mentioned previously, the addition of a bastion/platform extended the width of the fortifications to 6.5m at this point. As at other Early Bronze Age sites, for example, Bab adh-Dhraʻ (Rast/Schaub 2003: 280–83) and Numayra (Coogan 1984), the builders incorporated transverse walls perpendicular to the fortification line, whether functioning as an earthquake device or simply a segmental construction method is unknown. Moreover, Khirbat Iskandar now joins other sites, where the erection of bastions and platforms attest to the strengthening of defenses within the EB III period, as at Tell Yarmut (de Miroschedji 1990). The strengthened northwest corner, the highest point on the mound, along with its occupational evidence suggests the location of a possible public area to the Phase C settlement.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Recent Excavations at Khirbat Iskandar, Jordan.

593

It is the settlement built against the fortifications/platform that provides the ceramic evidence for an EB III date. Summarily, excavation below a massive destruction layer, has thus far exposed an upper Phase C (C1) settlement of a central room with pillar base, an outer room with well-preserved doorway and threshold with steps, a courtyard, and a work area, along with ancillary rooms (to the east). A great amount of whole and restorable pottery (primarily storejars with carbonized grain) was found in the destruction debris lying on the surface of the central room, suggestive of a storeroom (fig. 7). The work area with enigmatic mudbrick feature/bench contained a wealth of materials, including a macehead, a tournette, grinding stones, and numerous bowls with ground depressions, as well as a stamp seal and wavy-handled teapot. The mudbrick feature sat contiguous to a nicely made stone bin with associated pavement, a nearby mortar surrounded by smaller stones, and two small post holes, all part of what appears to be a large open courtyard (two large hearths were also found nearby). We have tentatively interpreted the workshop and courtyard (along with a wealth of material remains) in relation to the central building as evidence suggestive of a public area within the considerably strengthened northwest corner defenses. Less is known, at this point, about the founding Phase C2 settlement, although the Phase C fortifications were constructed at that time. The EB III date for construction derives from a study of the pottery, including a Syrian goblet discovered near the bottom of the foundation trench on the north side of the fortifications (Richard/D’Andrea i. p.). For additional details about the Phase C settlement, see Richard/Long 2005; Richard et al. i. p.; Long et al. 2014. To sum up the emended phasing of the Khirbat Iskandar fortifications: Phase D fortifications (inner mudbrick/stone base wall and curvilinear structures, as well as the mud brick/ stone base western curtain wall) are the earliest, now more confidently dated to the early EB III period. Subsequently constructed, the Phase C fortifications (bastion/platform associated with outer walls on north and west, along with other Phase C components (piers/buttresses/ transverse walls) also date to the EB III, based on interior occupational surfaces associated with these fortifications. When the Phase C western curtain wall (B2A120/B4A006) went out of use, there subsequently followed the construction of the «rubble wall». This less substantial construction roughly abutted the southwest corner of the bastion. We are presently identifying it as an EB III/EB IV (Phases C-B) wall, based on the sum of the evidence.

EB IV Fortifications The 2013 discovery of the original Phase C western perimeter wall sheds light on the «rubble» wall, now more comprehensible as a later rebuild. But, was it erected by the Phase C population following a destruction at which time W. B2A120/B4A006 went out of use? Or, was it constructed in Phase B, the EB IV period? A review of some of the salient pieces of evidence would be in order. It has been known for some time that the Phase B population reused the fortifications, as mentioned previously. Specifically, our excavations delineated

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

594

Richard

Fig. 8. Khirbat Iskandar EB IV wall under meter stick, «rubble»wall at right.

a public building/storeroom built against the outer (Phase C) wall on the north. Along the west, Phase B walls and features have been found in association with the «rubble» wall. As a matter of fact, there are Phase A walls in association with the western perimeter wall as well (Richard et al. i. p.). The 2013 season brought to light the most compelling evidence thus far for an EB IV, Phase B rebuild of the fortifications, in particular, the «rubble» wall. In order to ascertain whether or not Phase B wall B19A019 continued as W.B19A043, which intersected with the «rubble» wall, the team removed part of overlying Phase A wall, B19A003. In the 2010 season, excavation of a surface in association with B19A019 uncovered a cache of whole and restorable EB IV vessels (Richard et al. i. p.). As fig. 8 clearly shows, walls B19A019 and B19A043 are continuous and intersect with the «rubble» wall, evidencing that the two were at some point contemporaneous. The probe in Square B19A, where four deep walls converge and intersect and where Phases A, B, and C walls intertwine is, at this stage, stratigraphically challenging without more horizontal exposure. Despite the limited exposure of the four walls, however, the data are compelling enough to posit not just a reuse, but a rebuild of the defensive line termed the «rubble» wall. Combined W.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Recent Excavations at Khirbat Iskandar, Jordan.

595

B19A019/B19A043 is founded at a higher level than the three other intersecting walls with which it is associated, ending at Surface 19A041 on a layer of destruction. The other three walls continue lower through a layer of silt and also appear to end at a layer of destruction (B19A044). We are positing as a hypothesis to test next season that the «rubble wall» was constructed in the EB IV period. At the present time, it would be more prudent to consider the «rubble» wall an EB III/IV fortification. The cumulative data that the expedition has uncovered regarding the EB IV fortifications at Khirbat Iskandar correlate nicely with the description of the site provided by Nelson Glueck (1939: 127–28). As he noted, there were prominent fortifications visible around the site, along with towers, and an east-west wall bisecting the site in the middle. We have identified the latter wall as connected with the Area C gateway (Richard et al. 2014); we believe that we have sufficient evidence now to associate the «rubble» wall and the reused northern outer wall with the EB IV period as well.

Summary The 2013 excavations at Khirbat Iskandar have dramatically transformed our interpretation of the site’s fortifications. New defensive line B2A120/B4A006 is the missing western trace wall to the outer (Phase C) EB III fortifications, which were constructed following destruction of the Phase D inner wall (now dated to EB III). The new Phase C western trace wall has likewise engendered a revised interpretation of the western perimeter («rubble») wall previously discovered, now more cogently interpreted as a later phase. We are calling the «rubble» wall an EB III/IV fortification, possibly constructed in the EB IV period. In 2013, an EB IV, Phase B wall was found to intersect with three other deeply founded walls, one of which was the «rubble» wall. The recent excavation of EB IV sites with defensive structures, such as Khirbet el-Meiyiteh (Bar et al. 2013) and Dhahrat Umm al-Marrār (Jones et al. 2013), along with sites known from survey, such as Jabal Rahil, Khirbat Umm ar-Rujum and ar-Rasayfah (Palumbo 2001: 242–52), resolves the controversial question of whether or not fortifications existed in the period, and adds substantial weight to the suggestion that considerable evidence for rural complexity existed in the EB IV.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

596

Richard

Bibliography Bar, S./Cohen, O./Zertal, A., 2013. New Aspects of the Intermediate Bronze Age (IB/MBI/EBIV): Khirbet el-Meiyiteh – A Fortified Site on the Eastern Fringe of Samaria. Revue Biblique 120/2, 161–181. Coogan, M. D., 1984. Numeira 1981. Bulletin of the American Schools of Oriental Research 255, 75–82. Cordova, C. E., 2007. Millennial Landscape Change in Jordan: Geoarchaeology and Cultural Ecology. Tucson. Cordova, C. E./Long, J. C., Jr., 2010. Modern and Ancient Environment. Richard et al. (eds.) 2010: 21–35. D’Andrea, M., 2014. Townships or Villages? Remarks on Middle Bronze IA in the Southern Levant. P. Bieliński et al. (eds.), Proceedings of the 8th International Congress on the Archaeology of the Ancient Near East (8ICAANE) 30 April – 4 May 2012, University of Warsaw. Wiesbaden, 151–172. de Miroschedji, P., 1990. The Early Bronze Age Fortifications at Tell Yarmut – An Interim Statement. Eretz Israel (Amiran Volume) 21. Jerusalem, *48–68*. Glueck, N., 1939. Explorations in Eastern Palestine III. Annual of the American Schools of Oriental Research 18–19. New Haven. Höflmayer, F., 2014. Dating Catastrophes and Collapses in the Ancient Near East: the End of the First Urbanization in the Southern Levant and the 4.2 ka BP Event. L. Nigro et al. (eds.), Reading catastrophes. Proceedings of the International Conference “Reading Catastrophes: Methodological Approaches and Historical Interpretation. Earthquakes, Floods, Famines, Epidemics between Egypt and Palestine – 3rd – 1st millennium BC” held in Rome, 3rd – 4th December 2012. ROSAPAT 11. Rome, 117–139. Jones, J. E./Falconer, S. E./Fall, P. L./Metzger, M., 2012. Excavations at Early Bronze IV Dhahrat Umm al-Marrār in the Jordan Valley. Annual of the Department of Antiquities of Jordan 56, 381–403. Long, J. C., Jr./Richard, S./Wulff-Krabbenhøft, R./Ellis, S., 2014. Khirbat Iskandar. G. J. Corbett et al. (eds.), Archaeology in Jordan, 2012 and 2013 Seasons. American Journal of Archaeology 118/4, 648–650. Palumbo, G., 2001. The Early Bronze Age IV. B. MacDonald et al. (eds.), The Archaeology of Jordan. Sheffield, 233–269. Parr, P. J., 1960. Excavations at Khirbet Iskander. Annual of the Department of Antiquities of Jordan 4–6, 128–133. Rast, W. E./Schaub, R. T., 2003. Bab edh-Dhraʻ: Excavations at the Town Site (1975–1981). 2 vols. Winona Lake. Regev, J./de Miroschedji, P./Greenberg, R./Braun, E./Greenhut, Z./Boaretto, E., 2012. Chronology of the Early Bronze Age in the Southern Levant: New Analysis for a High Chronology. E. Boaretto/N. R. Rebollo Franco (eds.), Proceedings of the 6th International Radiocarbon and Archaeology Symposium. Radiocarbon 54.3–4. Tucson, 525–566.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Recent Excavations at Khirbat Iskandar, Jordan.

597

Richard, S., 2006 Early Bronze IV Transitions: An Archaeometallurgical Study. S. Gitin et al. (eds.), Archaeological and Historical Essays on Ancient Israel in Honor of William G. Dever. Winona Lake, 119–132. –

1990. The 1987 Expedition to Khirbet Iskander and Its Vicinity: Fourth Preliminary Report. W. E. Rast (ed.), Bulletin of the American Schools of Oriental Research Supplement 26. Baltimore, 35–56.

Richard S./Boraas, R. S., 1984. Preliminary Report of the 1981–82 Seasons of the Expedition to Khirbet Iskander and its Vicinity, Bulletin of the American Schools of Oriental Research 254, 63–87. Richard, S./D’Andrea, M., i. p. A Syrian Goblet at Khirbat Iskandar, Jordan: A Study of Interconnectivity in the EB III/IV Period. Studies in the History and Archaeology of Jordan 12. Amman. Richard, S./Long, J. C., Jr., 2005. Three Seasons of Excavations at Khirbat Iskandar, 1997, 2000, 2004. Annual of the Department of Antiquities of Jordan 4, 261–275. Richard, S./Long, J. C., Jr./Wulff-Krabbenhøft, R./Ellis, S., i. p. Three Seasons of Excavation at Khirbat Iskandar: 2007, 2010, 2013. Annual of the Department of Antiquities of Jordan. Amman. Richard, S./Long, J.C., Jr./Holdorf, P./Peterman, G. (eds.), 2010. Archaeological Expedition to Khirbat Iskandar and Its Environs. Vol. 1: Khirbat Iskandar Final Report on the Early Bronze IV Area C “Gateway” and Cemeteries. ASOR Archaeological Reports 14. Boston.

Suzanne Richard, Gannon University, [email protected].

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Uzbekistan – Turkmenistan – Afghanistan

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 601–614

Joaquín María Córdoba – Muhammed Mamedov

L’âge du fer au Dehistan. Nouvelles recherches archéologiques turkmènes et espagnoles sur les sites de Geoktchik Depe et Izat Kuli (Province de Balkan, Turkménistan) Recently a Turkmen and Spanish archaeological project has been initiated in the ancient region of Dehistan (Balkan Province, Republic of Turkmenistan). In the fifties of the last century, V. M. Masson discovered in the desert plain of Misrian, a new culture that he dated to the Iron Age, and called Archaic Dehistán Culture. A large network of canals, settlements of different types related to it and also some major archaeological sites whose big surface suggested an urban or proto-urban status. Since then, brief Soviet, Turkmen and French interventions have been providing more information. But much remains to understand the characteristics, relationships and real history of this culture. This joint project has started operating on three routes: the territory and the control of the environment, the hierarchy of settlements (Izat Kuli), and a possible ceremonial centre (Geoktchik Depe). We present the first results.

1. Introduction En général, nous disons que dans la grande zone géographique couvrant l’histoire et la culture de l’Orient ancien, l’âge du Fer est une période très mal connue. Pour la seconde moitié du 1er millénaire av. J.-C., nous avons tendance à utiliser les auteurs grecs comme source d’information, en dépit de ses limites évidentes. Mais dans la première moitié du millénaire, les textes orientaux posent aussi des problèmes. Parce que la diffusion progressive de la langue araméenne et l’abandon des écritures cunéiformes a eu des conséquences: l’histoire d’aujourd’hui met l’accent sur les événements politiques décrits dans les annales assyriennes et les Chroniques babyloniennes.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

602

Córdoba – Mamedov

D’un point de vue archéologique, même à Babylone et en Assyrie, nous avons des connaissances sur les palais et les temples, mais presque rien sur les villes, villages, communications et l’exploitation économique du territoire. Pour cette raison l’âge du Fer est devenu l’objectif de beaucoup des nouvelles recherches archéologiques. L’Asie Centrale est une des principales régions culturelles de l’Orient ancien. Après un âge du Bronze brillant avec la civilisation étonnante de l’Oxus, mise en valeur par la culture et l’histoire de Margus-Margiana, au coeur du Turkménistan (Sarianidi 2006, 2007; Kozhin et al. 2010), nous sommes confrontés à un âge du Fer qui soulève de nombreuses questions. La recherche soviétique a défini le paysage dans deux domaines principaux: la Culture du Dehistan archaïque, au sud-est de la mer Caspienne, et le Complexe Yaz, dans les piémonts de la Kopet Dag, la région de Merv et le delta de la rivière Murghab (Masson/Sarianidi 1972: 155–166). De récentes études sur la poterie peinte de Yaz I (Lhuillier 2013) et les résultats des études à quelques sites dans le distant Dehistán, comme Geoktchik Depe (Lecomte 1999), nous ont fait corriger les limites chronologiques de l’âge du Fer. Notre projet porte sur cette question.

2. Dehistan : de la géographie et de l’histoire Entre les contreforts du Grand Balkan et le lit fossile de la riviere de l’Ouzboï au nord, les montagnes du Kopet Dag à l’est, l’Elburz au sud et la côte de la mer Caspienne à l’ouest, s’étend une grande plaine qui a été appelée à l’époque achéménide, Vehrkānā, d’où dérive le toponyme grec Hyrcanie/Hyrcania (Lecomte 199: 135). Les cartes historiques de l’Empire achéménide varient dans la localisation de Vehrkānā-Hyrcania, mais on s’accorde à l’identifier telle qu’elle est indiquée sur la carte (fig. 1). Du point de vue géographique, il faut distinguer deux régions: l’Hircanie du Sud, limitée à la plaine de Gurgan, abondamment arrosée par des rivières et de fortes précipitations, occupée depuis l’âge du Bronze ancien, et l’Hyrcanie du Nord, une plaine alluviale plate, formée par les alluvions de la rivière de l’Ouzboï et l’érosion des montagnes, aujourd’hui transformée en désert (Lecomte 1999: 135–136). Le grand Takyr est limité par de grandes dunes à l’est et au nord. Extrêmement aride, les archéologues soviétiques y ont découvert un réseau dense de canaux alimentés par la rivière de l’Atrek, qui auraient dû faciliter l’occupation humaine et l’agriculture. Dépourvue de rivières et de sources, jouissant de très peu de précipitations, la vie dans le Nord de l’Hircanie n’était possible que grâce à l’irrigation. Les sources historiques nous renseignent. Hérodote (III, 117) parle d’une région où l’irrigation est arrêtée selon le bon vouloir du Grand Roi, référence considérée comme une «fable historique» (Briant 1996: 428), mais que l’archéologie semble confirmer; le débit de

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

L’âge du fer à Dehistan

603

Fig. 1. Carte du Turkménistan indiquant l’Hyrcanie et Geoktchik Depe (selon H.-P. Francfort/O. Lecomte 2002: carte 10).

l’eau prélevée depuis la rive droite de la rivière Atrek pouvait être contrôlé, bloqué à l’entrée du canal.

3. Redécouvrir une culture Aujourd’hui nous savons que depuis le 13e siècle et la destruction du Grand Khorezm par les Mongols, la plaine du Dehistan est restée déserte. Vers le milieu du 19e siècle, A. Vámbéry a traversé la région avec un groupe de derviches, en notant le désert aride du Takyr, la présence d’importantes ruines et l’absence totale de d’eau, de villages et de populations (Vámbéry 1864: 90–92). Voyageurs et géographes russes ont décrit cela aussi (Gorshenina/ Rapin 2001: 29–33). En 1875, le général Lamakhin a étudié la zone et noté l’existence d’un réseau d’anciens canaux associés à des ruines et des monticules. Il a proposé de dégager le réseau et de remettre la plaine en culture. La réouverture d’un réseau complexe de plus de 150 kilomètres du nord au sud, et de plus de 80 km d’est en ouest, a semblé une tâche excessive et coûteuse à l’administration russe (Lecomte, communication personnelle).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

604

Córdoba – Mamedov

Des fouilles ont été menées durant la période tsariste près d’Ashgabat, dans Merv, Samarkand et à Tachkent, menées en particulier par le général A. V. Komarov (Gorshenina 2004: 39–40). Mais ce n’est qu’à l’époque soviétique que des fouilles ont été engagées dans le lointain Dehistan (Atagarryev/Berdyev 1970: 288; Muradova 1991).1 En 1931, M. E. Masson a exploré la vallée du Sumbar (Masson 1931). Un an plus tard, A. A. Maruschenko a exploré certaines parties de la plaine de Misrian et le piémont du Kopet Dag (Maruschenko 1935). Après la Seconde Guerre mondiale, en 1946, une grande mission interdisciplinaire a été envoyée au sud du Turkmenistan (Yutake) ayant pour mission de restituer le passé de la région et de cataloguer les sites archéologiques et les monuments (Atagarryev/Berdyev 1970: 291). Plus tard ont été menées des prospections méthodiques dans la plaine de Misrian qui ont mis en lumière une culture particulière, un peuplement médiéval important et un ancien réseau de canaux d’irrigation. En 1947, M. M. Masson a le premier exploré des sites de la culture qui nous intéresse ici: Tangsikyl’dia et Tchialyk Depe (Masson 1951). Un an plus tard, il a travaillé dans le Madau Depe (Masson 1955). En 1950, les fouilles de Izat Kuli ont été engagées, et V. M. Masson s’est intéressé au réseau des canaux (V. M. Masson 1953, 1954). Trois périodes chronologiques ont été distinguées : la période du Dehistan archaïque, la période partho-sassanide et la période du Khorezm. Entre 1951 et 1953, V. M. Masson a fouillé quatre sites (Tanjsikildya, Chialik Depe, Izat Kuli et Madau Depe) à partir desquels la culture du Deshistan archaïque (1500–500 av. J.-C.) a été décrite (Masson 1956). A la fin des années soixante, les archéologues russes ont fini le catalogue complet des sites du Dehistan archaïque (Atagarriev/Lisitsyna 1970), et ont fouillé le site de Tchyglyk Depe, découvrant une grande citadelle fortifiée, et les sites de Geoktchik et Tangsikyldia. En 1980, E. A. Charieva reprend les fouilles d’Izat Kuli, en particulier un ensemble de fours à céramique dans le cadre d’une recherche sur la poterie (Charieva 1981). E. A. Muradova a fouillé aussi dans ces années sur plusieurs sites, à Benguvan, et aussi à Izat Kuli, où elle a établi une séquence stratigraphique et une chronologie de la céramique (Muradova 1991). Un peu plus tard, entre 1994 et 1997, une mission conjointe du CNRS français et l’Académie des Sciences du Turkménistan, sous la direction de O. Lecomte et E. Atagarriev, a travaillé à Geokchik Depe et dans les ruines de la cité de Misrian (Lecomte 1999: 135–170; 2005: 461–478). A la fin de ce projet et jusqu’en 2009 plus aucune activité archéologique n’a eu lieu dans la région. Dès la période soviétique, les recherches ont mis en évidence une culture de l’âge du Fer que V. M. Masson a appelé le Dehistan archaïque et daté de 1500/1300–500 av. J.-C. L’importance du peuplement à la période médiévale a aussi été soulignée, ainsi que le rôle central joué par la grande ville de Misrian, point important sur la route de la soie. Il faut 1

Nous nous référons aux recherches les plus importantes. E. A. Muradova a écrit un compte rendu détaillé de la recherche soviétique au Misrian-Dehistan (Mouradova 1991: 3–7)

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

L’âge du fer à Dehistan

605

reconnaître un certain courage aux scientifiques soviétiques, puis russes et turkmènes, d’avoir mené ces travaux dans des conditions difficiles et un climat très rigoureux.

4. Le projet Dehistán Nous avons engagé nos travaux dans cette voie ouverte par nos collègues soviétiques, russes et turkmènes, puis par la mission française. Le Dehistan est situé dans une partie très reculée du Turkmenistan. Les conditions climatiques y sont pénibles, l’environnement difficile, l’eau manque et les moyens de communication difficiles alors que les distances sont immenses. En revanche, la carte archéologique est très complète et l’état des sites archéologiques plutôt favorable à l’exploration. Or les périodes anciennes du Dehistan sont d’un grand intérêt, et malgré les travaux déjà menés beaucoup de domaines restent à explorer. Les objectifs de nos travaux sont : de vérifier la nature du site de Geoktchik Depe et sa relation avec les autres sites de même culture ; d’étudier la géologie de la plaine, le réseau de canaux de la période du Dehistan archaïque ; de fouiller un grand site tel que Izat Kuli – ou du moins le comprendre et obtenir une stratigraphie de référence pour préciser la chronologie des trois phases de la période du Dehistan archaïque ; enfin, définir globalement la culture du Dehistan archaïque et préciser son identification avec l’état pré-achéménide de Verkhana (Dandamaev 1994: 40–43). Durant la dernière saison, la mission conjointe turkmène et espagnole a privilégié les travaux topographiques, tout en poursuivant la fouille de Goektchik Depe et en commençant l’exploration du site d’Izat Kuli.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

606

Córdoba – Mamedov

Fig. 2. Vue de Geoktchik Depe depuis le sud (photo: mission archéologique conjointe).

5. Geoktchik Depe: résultats de la saison mai–juin 2014 Geoktchik Depe (fig. 2) est situé à l’extrémité nord de la plaine structurée par le réseau de canaux, environ 25 kilomètres à l’ouest des ruines de Misrian. Il se situe à une altitude de seulement 16 mètres au-dessus de la mer. La carte archéologique soviétique signale une occupation des périodes du Dehistan archaïque et sassanide (Atagarryev/Lisitsyna 1970: 180). Entre 1994 et 1997, la mission franco-turkmène a constaté la présence d’un village sassanide au sud du tell formé par un immense bâtiment (Lecomte 1999: 135–170; 1999b: 54–66; 2001: 289–302; 2005: 461–478; 2009: 69–77).2 Dans la masse du tell, au nord a été dégagée une immense salle de 16 x 15 mètres, et 13 mètres de haut (Room 459) comblée par du sable dans l’antiquité. Le sol n’a été atteint que dans l’angle nord-est, la fouille ayant été interrompue. Les murs étaient couverts d’un revêtement de briques qui s’est érodé depuis l’arrêt des fouilles.

2

Trés généreusement, le Dr. Lecomte nous a fourni toutes les archives de la mission franco-turkmène.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

L’âge du fer à Dehistan

607

Fig. 3. Carte topographique de Geoktchik Depe, indiquant les secteurs explorés durant la dernière saison archéologique (T. Fernández Pareja).

La partie non fouillée de la salle, et cinq autres plus petites pièces avaient été laissées, protégées par plus de 5 mètres d’épaisseur de sable. Les observations faites dans les sondages est et ouest, et dans les sondages ouverts en haut et en bas du monticule à l’ouest semblent indiquer que le bâtiment est formé par deux grandes plates-formes en terrasses. Toutes deux sont datées du Dehistan archaïque, entre 1100 et l’époque achéménide. Au pied du monticule ont été partiellement dégagées des pièces destinées aux activités artisanales (Lecomte 1999: 135–170). Nos premiers objectifs ont été de confirmer la présence des plates-formes, de poursuivre le dégagement du bâtiment et de préciser la fonction des espaces sassanides rencontrés. Une longue tranchée a été ouverte, du haut vers le bas, que nous appelons le Sondage Sud. Il a permis de confirmer la continuité d’un grand bâtiment d’environ 60 mètres de projection nord-sud, conservé sur près de 14 mètres de haut dans la Room 459 fouillée par la mission franco-turkmène. À l’extrémité sud de la tranchée, un sondage stratigraphique profond a été fait. Des pluies torrentielles ont forcé à interrompre le travail.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

608

Córdoba – Mamedov

Fig. 4. Vue des couches de briques crues de la plate-forme de la période du Dehistan archaïque, dans le Secteur I (photo: mission archéologique conjointe).

Côté sud, différents niveaux aménagés en briques crues (caractéristiques du Dehistan archaïque) ont été dégagés (fig. 4), témoignant de l’ampleur et de la solidité de la construction, ainsi que les vestiges des aménagements en relation avec l’artisanat d’époque sassanide au pied du grand bâtiment partiellement ruiné (Córdoba 2010: 59–62). Un long sondage a aussi été ouvert sur le flanc ouest de l’édifice, pour reconnaître l’extension de la plate-forme, et l’éventuelle présence d’une occupation sassanide. La plateforme est ici formée de briques semblant s’ouvrir en éventail, comme si l’étage du bâtiment avait un plan polygonal au lieu d’un simple carré (fig. 5). Ce point demande une exploration complémentaire. Du matériel sassanide a été recueilli, des restes de faune et de la céramique aux formes très répétitives (fig. 6). Le sondage Nord va progressivement nous livrer une stratigraphie des dépôts situés à l’extérieur de la grande salle fouillée par la mission franco-turkmène.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

L’âge du fer à Dehistan

609

Fig. 5. Plan des secteurs ouverts sur les pentes méridionale et occidentale de Geoktchik Depe au cours de la dernière campagne (secteurs K, I, J, L) (dessin: M. Núñez).

Le bâtiment est maintenant dégagé sur 533 m2, ce qui en fait la plus grande construction du Dehistan archaïque. Il mesure près de 80 x 90 m, et est conservé sur 14 m de hauteur. Le nouveau relevé topographique (fig. 3) intègre le site dans le réseau des canaux qui structure le territoire du site. Nous y intègrerons les vestiges et constructions découvertes. Geoktchik nous apparait comme un monument unique de la culture du Dehistan, et avait probablement une fonction religieuse ou funéraire (Lecomte 2005: 466). Son exploration doit se poursuivre.

6. Izat Kuli. Résultats de la saison mai–juin 2014 Un autre volet du travail mené durant la dernière saison a été de commencer une nouvelle exploration d’Izat Kuli, un site de plus de 100 hectares, de 1200 x 1500 m (Atagarryev/ Lisitsyna 1970: 181), à environ 29 kilomètres au sud de Geoktchik Depe (fig. 7). Il se situe dans le centre de la région, apparemment loin de la plus grande concentration d’établissements du Dehistan archaïque à 32 mètres au-dessus du niveau de la mer. Les données des anciennes fouilles, la topographie des vestiges et le matériel de surface (Masson 1956: 390–402; Charyeva 1981; Muradova 1991: 51–71; 2010: 214–226; 2013: 113–120) semblent indiquer que Izat Kuli était une sorte de proto-ville.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

610

Córdoba – Mamedov

Fig. 6. Cette même forme de céramique sassanide se trouve répétée dans plusieurs des espaces fouillés au pied du monticule de Geoktchik Depe (photo: mission archéologique conjointe).

Fig. 7. Vue aérienne du site de Izat-Kuli (Google Earth).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

L’âge du fer à Dehistan

611

Fig. 8. Vue d’une partie de la présumée plate-forme centrale d’Izat-Kuli (photo: mission archéologique conjointe).

Nous avons visité et prospecté plusieurs fois le site, toujours surpris par ses caractéristiques et le matériel de surface. Un sondage stratigraphique profond sur ce site nous fournira une séquence chronologique plus précise pour tout le Dehistan archaïque: selon Masson et Muradova, il a été longuement à cette période. L’objectif pour cette campagne était d’en faire un plan topographique pour mieux en comprendre l’organisation. Mais les pluies et les tempêtes de sable l’ont empêché. Les grands centres de la culture de l’âge du Fer dans cette région se définissent communément par la présence d’une plate-forme de brique crue sur laquelle les résidences de la classe dirigeante étaient construites (Masson/Sarianidi 1972: 155–158). Masson avait interprété les vestiges au centre du site comme étant ceux d’une citadelle. La topographie nous suggère à la fois la présence d’une citadelle fortifiée avec des tours et d’une plate-forme. Une fouille archéologique limitée a été menée dans la partie haute du monticule central, un sondage de 17 x 7 m. Une construction en brique crue a été mise au jour, sorte de plateforme, munie peut-être d’un accès au nord, dont les matériaux, la forme et les mesures sont comparables à celle de Geoktchik Depe (fig. 8). Cette construction est bordée de pièces, la

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

612

Córdoba – Mamedov

mieux dégagée mesurant 3,25 x 2,50 m, remplies avec de la terre et des fragments de poterie. Dans l’état actuel des travaux, notons seulement que pour la première fois, nous avons une partie de la surface supérieure d’une des plates-formes constituant l’élément central de la planification urbaine des grands centres du Dehistan archaïque. L’ensemble céramique recueilli est très caractéristique du Dehistan (pourcentages entre rouge, gris, clair, et céramique de cuisine) et plus précisément de la troisième phase du Dehistan archaïque (Muradova 1991: 54), correspondant aux 7e–6e s. av. J.-C. Puisque ce matériel provient du sommet du monticule central il peut être un indicateur de la phase finale de l’occupation de la plate-forme.

7. Conclusions Provisoires Le site de Geoktchik Depe est formé principalement par un bâtiment de très grande taille en briques crues, d’environ 80 x 90 m à la base, conservé sur plus de 14 m de hauteur. A la période sassanide, à la base de ce bâtiment ruiné, presque au niveau de la plaine où se dresse la construction sassanide, ont été creusées des fosses et aménagés des espaces rectangulaires, utilisés comme dépotoirs, et d’autres activités difficiles à déterminer (préparation d’enduits?). Le bâtiment de la période du Dehistan archaïque a dû être abandonné pendant la période achéménide. Une phase de violence a pu forcer à l’abandon des principaux sites de la région. Izzat-kuli, qui s’apparente à Geoktchik, par les matérieux et modes de construction, semble avoir été occupé jusqu’à la fin de la période du Dehistan archaïque. Le fonctionnement du réseau de canaux suppose la présence continue d’une population rurale. Pour obtenir progressivement une vision plus précise de la chronologie du peuplement de la région il est nécessaire d’affiner la séquence de la culture matérielle, en particulier pour déterminer la limite supérieure de l’occupation à la période du Dehistan archaïque (Lecomte 1999: 159–160).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

L’âge du fer à Dehistan

613

Bibliographie Atagarryev, E./Berdyev, O., 1970. The Archaeological Exploration of Turkmenistan in the Years of Soviet Power. East and West 20 (3), 285–306. Atagarryev, Е./Lisitsyna, G. N., 1970. Raboty nad sostavlenie arkheologuicheskoy karty MeschedMisrianskoy raviniy – Chaskogo masiva. Karakumskie Drevnosti III, 166–183. Briant, P., 1996. Historire de l’Empire Perse. De Cyrus à Alexandre. Paris. Córdoba, J. Mª., 2011. Turkmen-Spanish Archaeological Mission in Geoktchik Depe”/“Gökçikdepe Türkmen-ispan Arheologik Toporynyn geçíren ilkinji ylmy-barlaglary”/“Первая туркменоиспанская археологическая экспедиция на Геокчик-Депе в 2010 г”. Ashgabat, 59–62. Charyeva, E. A., 1981. Raskopki goncharnykh pechey na Izat-kuli. Arkheologuicheskie otkrytnya 1980. Moscou. Dandamayev, M. A., 1994. Media and Achaemenid Iran. J. Harmatta (ed.), History of civilizations of Central Asia. Volume II. Paris, 35–65. Gorshenina, S., 2004. The Private Collections of Russian Turkestan. Berlin. Gorshenina, S./Rapin, C., 2001. De Kaboul à Samarcande. Les archéologues en Asie centrale. Paris. P. M. Kozhsin/ M. F. Kosarev/N. A. Dubova (eds.), 2010. On the Track of Uncovering a Civilization/ Naputi otkrytuya sivilizasii. Sankt Petersburg. Lecomte, O., 1999. Vehrkānā and Dehistan: late farming communities of South-west Turkmenistan from the Iron Age to the Islamic Periods. Parthica 1, 135–170. –

1999b. Le complexe cultuel de Geoktchik Depe. Archeologia 352, 54–66.

–

2005. The Iron Age of Northern Hyrcania. Iranica Antiqua XL, 461–478.

–

2009. Origine des cultures agricoles du Dehistan (Sud-Ouest Turkménistan. M. Al-Dbiyat/ M. Mouton (eds.), Stratégies d’acquisition de l’eau et société. Presses de l’IFAPO. Beyrouth, 69–77.

Lhuillier, J., 2013. Les cultures à céramique modelée peinte en Asie centrale méridionale. Dynamiques socio-culturelles à l’âge de Fer ancien (1500–1000 av. n.è.). Paris. Maruschenko, A. A., 1931. Syschectvennye popravki. Tyrkmenovedenie, 11. –

1935. Arkheologuicheskie otkrytniya poslednikh let v Turkmenictane. Ashgabat.

Masson, M. E., 1931. Siluety Sumbarkogo raniona. Istoriko-arkheologuicheskii ocherk. Turkmenovedenie nº 3–4. –

1951. Yuzhno-Turkmenistanskaya arkheologuicheskaya kompleksnaya ekspedisiya (YUTAKE) 1947. YUTAKE 2, Ashgabat.

Masson, V. M., 1953. Izuchenie kultyry drebnego Dakhistana. AN TSSR nº 1. –

1954. Misrianskaya ravnina v epokhy pozdnei bponzy i rannego zheleya. AN TSSR nº 2.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

614 –

Córdoba – Mamedov 1956. Pamyatniki kultury arkhicheskogo Dakhistana v Yugo-Zapadnoy Turkmenii. YUTAKE 7, Аshgabat 1956, 390–402.

Masson, V. M./Sarianidi, V. I., 1972. Central Asia. Turkmenia before the Achaemenids. London. Muradova, E. A., 1991. Poseleniya arkhaicheskogo Dakhistana. Ashgabat 1991. –

2010. Raskopki v zapadnoi chacti poceleniya izat-kuli. P. M. Kozhsin et al. (eds.), On the Track of Uncovering a Civilization/Na puti otkrytuya sivilizasii. Sankt Petersburg, 214–226.

–

2013. Archaeological explorations at Izat-kuli, ISIMU 16, 113–120.

Sarianidi, V. I., 2006. Gonurdepe. Ashgabat. –

2007. Necropolis of Gonur. Athens.

Vámbéry, Á., 1864. Travels in Central Asia. London 1864.

J. Mª. Córdoba, Universidad Autónoma, Madrid. M. Mamedov, National Department for the Protection of Historical Monuments, Ashgabat.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9th ICAANE, Basel 2014, Vol. 3, 615–629

Fabiana Raiano

Pottery from the joint Uzbek-Italian Archaeological Mission at Kojtepa (Samarkand Area – Uzbekistan) The paper presents the results achieved regarding the ceramic materials from the excavation activities of the joint Uzbek-Italian Archaeological Mission, directed by Bruno Genito and Kazim Abdullaev at Kojtepa from 2008 to 2013.

1. Kojtepa Site The site of Kojtepa (UTM 42 S 300104.00 m E, 4386581.00 m N) is close to Chandyr village, in the Pastdargom district, about 30 km SW of the modern city of Samarkand, in the Middle Zeravšān Valley (fig. 1) and covers an area of 2,62 ha (175 × 150 m). It is located in an area between the southern piedmont region of Karatyube chain, populated by nomadic tribes, and the northernmost more extensive area is cultivated with wheat and cotton. It consists of a central citadel (tepa), 9 meters high, encircled by fortified walls and a moat1. The main canal irrigating the area are Dargom, Eski Anghor and Čilibursay (Abdullaev/Genito 2011: 8–9). Between 2008 and 2013, four different areas were excavated: – AREA 1, close to the eastern walls (Trenches nos. 1, 2, 3, 4, 7); – AREA 2, the top of the tepa (Trenches nos. 5, 11); – AREA 3, close to the southern walls (Trenches nos. 6, 8); – AREA 4, in the damaged zone between eastern and northern walls (Trench no. 9).

1

The joint Uzbek-Italian Archaeological Mission is directed by B. Genito (Università degli Studi di Napoli «L’Orientale», hereinafter UNIOR) and co-directed by A. Gricina (2008–2009), K. Abdullaev (2011–2012) and M. Pardaev (2013) of the Institute of Archaeology of Academy of Sciences of Uzbekistan (hereinafter IAASU). Many thanks are due to the members of the Mission: G. Ciucci, E. Cocca, F. Franzese, M.M. Lamberti, Li Yusheng, D. Lunelli, S. Pardaev, L.M. Rendina, and F. Spinelli.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

616

Raiano

Fig. 1. The middle Zeravšān Valley (after Mantellini 2013).

2. Pottery Two different approaches were used in the study of pottery: the archaeometric method, in order to observe the petrographic composition of the body clay, and the typological method, based on the analysis of all the morphological features of the potsherds. In addition, the methodology included quantitative study, fabric analysis and comparisons between ceramic materials from other archaeological contexts in Sogdiana, Bactria and surrounding areas.

2.1. Petrographic Analysis After the autoptic observation of all the clay pastes, a fabric has been assigned to each potsherd on the basis of some specific criteria (Genito/Raiano 2011: 104) according to the parameters of the Norma UNI:10739482, accompanied by fabric samples. The petrographic 2

The norm was intended to establish the terms of ceramic materials providing definitions applicable to production technologies, with particular attention to the raw materials, techniques, processes, tools, and products http://www.uni.com.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Pottery from the Joint Uzbek-Italian Archaeological Mission at Kojtepa (Uzbekistan)

617

Fig. 2. Excavated Trenches 2008–2013 (after Lunelli 2013).

analysis was conducted on six samples chosen among the most representative fabrics at the laboratory of the Dipartimento della Terra, dell’Ambiente e del Territorio, of the Università degli Studi di Napoli «Federico II» with the contribution of Dr. A. De Bonis3: – KTO: Trench no. 4, US 17, fabric over-fired; – KT2b: Section no. 1, US 5, fabric 2; – KT4b: Trench no. 2, US 26, fabric 4; – KT4c: Trench no. 1, Room no. 1, US 21, fabric 4; – KT6: Trench no. 4, US 19, fabric 6; – KT7: Section no. 2, fabric 7.2.

3

I would like to thank Prof. V. Morra and Dr. A. De Bonis of the Dipartimento di Scienze della Terra, dell’Ambiente e delle Risorse of the Università degli Studi di Napoli «Federico II» for the great availability and professionalism with which they conducted the analysis.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

618

Raiano

Fig. 3. Thin sections under a polarizing microscope (after Raiano 2014b).

The observation in thin sections was performed using an optical microscope in polarized light (Leitz Laborlux 12 POL) interfaced to the computer with a camera (Leica DFC280) in order to acquire the images (fig. 3). Preliminarily, we can say that there is a correspondence between the majority of the petrographic inclusions observed in thin section and the sediments of the areas surrounding the site of Kojtepa. These, in fact, are made up of alluvial sediments, the composition of which

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Pottery from the Joint Uzbek-Italian Archaeological Mission at Kojtepa (Uzbekistan)

619

Fig. 4. Quantitative Study of Kojtepa Pottery (after Raiano 2013).

was attributable to the lithology forming the Karatyube Mt., and from which most of the sediments carried by the watercourses from the mountains would come. The petrographic composition of the inclusions, observed in all the fabric samples, is compatible with the geo-lithological composition of the area, confirming a local production of pottery.

2.2. Quantitative Study The total amount of ceramic materials collected during the excavation activities is 12484 (fig. 4). Within them, the diagnostic potsherds are 1395; the un-diagnostic are 10750; lastly, significant potsherds are 339 and constitute the Inventory.

2.3. Classification and Typology (fig. 5) The only class of pottery found in the site is the «common ware», inside which three different functional categories have been identified: cooking ware, storage ware and tableware. Each of these categories, in the frame of the further distinction between the closed and the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

620

Raiano

Fig. 5. Classification and typology of Kojtepa pottery (after Raiano 2013).

open shapes, has been distinct in morphological differentiations according to the shape of the vessel. The main collected shapes are cooking pots, jars, pots, miniature pots, jugs, flasks, lids, goblets, bowls and dishes. Finally, for each shape recurring types have been identified. Cooking Ware is the 4% of the total, exclusively pertinent to fabric 2 and represented by cooking pots. The technological features of this pottery types are closely related to their function. During the cooking process, the pot needs to tolerate the heat and the changes of temperature. The quality of clay, the shapes and the firing temperature affect the resistance to thermic shock and the time of the heat exposure. Table Ware is used to prepare or consume food; it is represented by stemmed goblets, cups, bowls, jugs and flasks and represents the 41% of the total. The majority of the potsherds have thin walls and fine paste: cups belong to fabrics 6 and 6.1; goblets to fabric 7.2; bowls to fabric 4. Storage Ware is the 55% of the total amount and is represented mostly by jars and pots of fabrics 1, 3, 4, and 14.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Pottery from the Joint Uzbek-Italian Archaeological Mission at Kojtepa (Uzbekistan)

621

Fig. 6. Decorative Techniques: a. rounded beads (Inv. 0266). b. herringbone (Inv. 0207). c. glossy blackgray slip (inv. 0028). d. splashes (Inv. 0223). e. spots of color (Inv. 0290). f. Inv. 0152. g. Inv. 0180. h. Inv. 0188 (after Raiano 2014b).

2.4. Decorative Techniques (fig. 6) Pottery from Kojtepa shows a rather homogeneous decoration: vessels are always slipped, light brown or red colored, or, sometimes, the surface is burnished, especially dark brown colored. The main decorative techniques are: 1. Incision: with single or multiple lines, waving or horizontal, sometimes appearing with squared incised motifs or rounded beads (Inv. 0266, fig, 6a), lozenges and herringbone (Inv. 0207, fig. 6b). From a «morphological» point of view, this kind of incised motives appear on jars (of small/medium size) and on pots. 2. Slip: mainly light brown or red light, which can cover: 2a. the entire surface, internally and/or externally. The presence of a glossy black-gray slip (inv. 0028, fig. 6c), which is somewhat reminiscent of metallic luster, has so far been rare in the corpus of Kojtepa pottery: this kind of surface treatment is attested in Afrāsīāb pottery between the 1st century BC and the 2nd AD, most likely due to a possible influence of the Roman pottery favoured by commercial development along the Silk Routes; 2b. only a portion (Inv. 0223, fig. 6d), especially in correspondence with the rim, usually in dark brown or red (splashes). Sometimes

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

622

Raiano

the slip could cover the entire surface of the vessel, just a specific portion of it (rim, decoration, careen, etc.) or appears in irregular spots of colour (Inv. 0290, fig. 6e). 3. The burnishing, giving a metallic appearance to the surface, seems to be pertinent only of fine and depurated pottery, as goblets and cups. The splashes are very widespread, especially on jugs (Inv. 0152, fig. 6f), pots (Inv. 0180, fig. 6g), and lids (Inv. 0188, fig. 6h).

3. Comparisons The comparative method used between materials coming from archaeological contexts, geographically far and contemporary, is useful for obtaining a more complete picture of both the productions and the commercial and cultural interactions. This method, however, has been difficult, especially in relation to the relative chronology of the materials, since there are no complete ceramic repertoires that enclose all the pottery productions coming from the main archaeological contexts in the central-Asiatic area. Comparisons were conducted on the basis of the publications of individual archaeological investigations where possible even if, in most cases, they are quite old and do not provide a detailed periodization of the material, chronologically grouped for large periods.

3.1. Comparative Study4 The sites under consideration are mainly those in the Valley of the Surkhan Dar’ja (Dal’verzin Tepe, Termez), eastern Bactria (Aï Khanum) and Sogdiana (Afrasiab, Kumiškent Tepe, Tali Barzu) in addition to the nomadic cemeteries in the Kafirnigan Valley, Zeravšān Valley and Bukhara oasis (fig. 7). To integrate the ceramic material discovered so far during the systematic excavation of the site of Kojtepa in the broader context of Central Asia, we attempted to define a terminus post quem corresponding to the final chronological stage of Kok Tepe5, dated from the end of the 2nd millennium BC until the 4th century BC (excluding the nomadic 4

5

The comparisons were made only from the morphological point of view, and only in the case of Afrāsīāb they were conducted, at least for a part, directly. In all other cases, comparisons were conducted on the basis of publications based on the ceramic data published by the Délégation Archéologique Française en Afghanistan (Lyonnet 1997), those published in the general work conducted by G. A. Košelenko (1985) and those published by G. A. Pugačenkova (1989). In addition, we also used the publications in four volumes of the series Afrasiab (1969–1975), some papers and reports published in the series Istorija material’noj kul’tury Uzbekistana (IMKU), the volumes by A. M. Mandel’štam Kočevniki na Puti v Indiju (1966) and Pamjatniki kočevnikov Kušanskogovremeni v Severnoj Baktrii (1975), and the results presented in an unpublished doctoral dissertation (Dimartino 2011). The site was investigated by MAFOuz under the direction of C. Rapin between 1994 and 2008 (Lyonnet 2001, 2010, 2013; Lhuillier 2010; Lhuillier, Isamiddinov, Rapin 2012; Rapin, Isamiddinov 2013).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Pottery from the Joint Uzbek-Italian Archaeological Mission at Kojtepa (Uzbekistan)

623

Fig. 7. Main sites interested by comparative study (adapted from Lyonnet 1997).

tomb dated to the 2nd–3rd century AD), and a terminus ante quem, corresponding to the initial phase of the site of Kafir Kala6, dated from the 6th–7th century AD (Dimartino 2011).

Hellenistic Period (3rd Century BC) In the site of Kojtepa, up to now, after four seasons of excavations, we did not reach stratigraphic levels prior to the Hellenistic period. It is not possible to define whether or not this stage is actually «Hellenistic» because of the absence of obvious confirmations: some shapes, compared with the «Hellenistic» materials found at Afrasiab and Aï Khanum would suggest an influence of Greek type, but, more critically, we may lean towards a possible legacy of the later Greek elements. Among the so-called «Hellenistic» shapes, there are two potsherds of dishes similar to those attested at Aï Khanum (Lyonnet 1997: fig. 43, 1). Similar shapes are also attested in the Surkhan Dar’ja Valley at Termez (3rd–2nd century BC; Pidaev 1987: fig. 1, 14), at Mirzankul’ Tepe (1st century BC–1st century AD; Pidaev 1978: tav. IV, 6), at Dal’verzin Tepe (2nd

6

Kafir Kala has been investigated by the Italian Archaeological Mission of the Università degli Studi di Bologna «Alma Mater Studiorum» under the direction of M. Tosi since 1999 (Mantellini 2003b; Mantellini, Berdymuradov 2005; Dimartino 2011).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

624

Raiano

century BC–1st century AD; Pugačenkova/Rtveladze 1978: fig. 101, 10), and in the Zeravšān Valley, at Afrāsīāb (4th–2nd century BC; Šiškina 1974: fig. 6, 19). Goblets with disc-shaped base are attested in eastern Bactria, at Aï Khanum (4th–3rd century BC; Lyonnet 1997: fig. 45, 1) and in the plain of Bactra at Dil’berjin (4th–3rd century BC; Kruglikova 1974: fig. 12). During the so-called «Greco-Bactrian period» (250–130 BC circa) this shape appeared in the Surkhan Dar’ja Valley, at Dal’verzin Tepe (Pugačenkova/ Rtveladze 1978: fig. 100, 1), Kampyr Tepe (Rtveladze 1987a: tav. XX, 5) and Termez (Pidaev 1987: fig. 1, 1–2). Along the middle course of the central Amu Dar’ja river, the shape is attested in the Babašov necropolis (2nd century BC–3rd century AD; Mandel′štam 1975: tav. XXII, 2–8) and in the Zeravšān Valley, at Afrāsīāb (3rd–2nd century BC; Kabanov 1969: fig. 3, 9; Šiškina 1974: fig. 5, 17, 20, 21, 27; 1975: fig. 2).

Post-Hellenistic Period (2nd–1st Century BC) On the basis of our comparisons and of the stratigraphic provenance of potsherds, goblets on stem are the main shapes datable to the post-Hellenistic period in Kojtepa. During the post-Hellenistic period, they are attested at Aï Khanum (Lyonnet 1997: fig. 48, 7–11; 2001: fig. 4), in the tomb no. 3 of Tillja Tepe (Sarianidi 1989:69, 82), at Dil’berjin (Kruglikova/ Pugačenkova 1977: fig. 16, 1–12), in the Surkhan Dar’ja Valley, at Kalčajan (Pugačenkova 1966: fig. 15), at Dal’verzin Tepe (Pugačenkova/Rtveladze 1978: 101, 3–4); at Kampyr Tepe (Rtveladze 1987: tav. XX, fig. 11, 29; fig. 13, 1). In the Kafirnigan and Biškent Valleys, they are attested in the Tulkhar (Mandel’štam 1966: tav. XX, 6, 10–11; tav. XXI, fig. 11–15) and Kokkum necropolis (Mandel’štam 1975: tav. X, fig. 1). In the Zeravšān Valley, they are attested in a group of sites west of Varakšā (Šiškin 1963: fig. 68, 4–6), in Bukhara (Mukhamedjanov et al. 1982: fig. 1) and in the Afrāsīāb southern (Nemceva 1969: fig. 5, 2–5) and northern area (Kabanov 1973: fig. 12, 3–6; fig. 13, 3–5, 8, 9). In Kojtepa, we found just one potsherd of tripod base vessels. Similar shapes were found in Aï Khanum (Lyonnet 1997: fig. 48, 4–6), in the plain of Bactra, in the site of Bactra (1st century BC–3rd century AD; Gardin 1957: fig. 4, a), and at Dil’berjin (1st century BC; Kruglikova/Pugačenkova 1977: fig. 21, IV). In the Surkhan Dar’ja Valley, in the Aïrtam necropolis (2nd–1st century BC; Turgunov 1973: fig. 14), at Kalčajan (2nd–1st century BC; Pugačenkova 1966: fig. 16), in Dal’verzin Tepe (2nd century BC–1st century AD; Pugačenkova, Rtveladze 1978: fig. 100, 31), and in Termez (1st century BC–1st century AD, Pidaev 1987: 90). In the Kafirnigan Valley, in Tulkhar necropolis (Mandel’štam 1966: tav. XVI, 9, 11). In the Zeravšān Valley, in Kumiškent Tepe (Pugačenkova/Bel’jaeva 1987: fig. 7, 10), and in Orlat area (Pugačenkova 1989: fig. 17) between 1st century BC and 1st century AD.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Pottery from the Joint Uzbek-Italian Archaeological Mission at Kojtepa (Uzbekistan)

625

Mustahara flasks7, with asymmetrical body laying on one side, are dated in Aï Khanum to the Hellenistic period. Shapes of this type were found in Dil’berjin at the beginning of the Kuśan period (Kruglikova 1986: fig. 26, 12); in the Surkhan Dar’ja Valley in the Aïrtam necropolis (2nd century BC–1st century AD; Turgunov 1973:66); in the Zeravšān Valley, in the Bukhara oasis and in Ustrušana (1st century BC–1st century AD; Mukhamedjanov et al. 1982: fig. 2; Košelenko 1985: fig. 135, 144). This shape is also attested in Afrāsīāb (3rd–1st century BC; Kabanov 1973:63), in the tomb of the nomadic princess at Kok Tepe (Rapin et al. 2001: fig. 8) and in the Orlat necropolis (2nd–3rd century AD; Pugačenkova 1989: fig. 54, 64).

Kuśan and Kuśano-Sasanid Period (1st–4th Century AD) Chronological difficulties recur in the study of material culture attributable to the Kuśan period: in contrast to the Bactria, Kuśans had never a real control of Sogdiana, as evidenced by the absence of numismatic elements and by a ceramic production clearly attributable to their presence in the region. The same problem also occurs in the Kuśano-Sasanid period, for the same reasons. In the present work, therefore, «Kuśan» and «Kuśano-Sasanid» are nothing more than a purely chronological reference (1st–4th century AD) and have not to be seen as indicative of the actual presence of the Kuśan and Kuśano-Sasanian governors in Sogdian territory. In the complex framework of the Kuśan and Kuśano-Sasanid period, we relied on two criteria in order to date the materials: comparisons with some sites of the Zeravšān Valley and the stratigraphic sequence of the levels in Kojtepa to establish a relative chronology of the potsherds. The Kojtepa pottery referring to this period do not deviate from the previous period pottery production, mainly from a morphological point of view, but a difference can, however, be identified on the basis of engraved wave decorative patterns. The main comparisons came from the site of Kumiškent Tepe (1st–4th century AD; Pugačenkova 1969: 35– 65, fig. 17–19). The wave engraved motive attested in the Surkhān Dar’ja Valley at Dal’verzin Tepe (Pugačenkova/Rtveladze 1978: 156) and in Termez (Pidaev 1987: 92) is datable to the same period. In the Valley of Kaška Dar’ja, this decorative motive is attested in the site of Erkurgan (1st–5th century AD; Košelenko 1985: fig. 133).

The so-called «Second Nomadic Phase» (4th–6th Century AD) Many difficulties have been encountered in the conduction of typological comparisons of this period too. Our starting point was the stratigraphic provenance of the potsherds. The quantitative diffusion in the more superficial levels and areas close to the eastern and southern walls suggested a later period from the 4th century AD up to the early medieval period (6th–7th century AD). 7

Typical flasks with hemispherical shape, imitating leather wineskins.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

626

Raiano

Up to now, on the basis of comparisons between the corpus of Kojtepa and some pottery from Afrāsīāb, jars seems to be datable to the 5th–7th century AD, because the fragments come from the later levels of the area near the eastern and southern walls (Lyonnet, personal communication 2011). The same shape has similarities with attested types in Erkurgan, in the Kaška Dar’ja region (5th–6th century AD; Turebekov 1981: fig. 3, 10, 11). This pottery should be the later shapes found in Kojtepa, both on the basis of the comparisons with the pottery of Kafir Kala (Dimartino 2011:113–243) dated for the most part to the 8th–10th century AD, and correspond to the period of «decline» or abandonment of the Kojtepa site.

4. Conclusions On the basis of comparisons between the different pottery productions in the area, up to now, there is no material attributable to the Achaemenid period (6th–4th century BC), just as there are no materials of the Islamic period (from 8th century AD): starting from these two chronological terms, in fact, one may put Kojtepa in a time span ranging from the Hellenistic period (4th–3rd century BC) and the early Middle Ages (5th–7th century AD). One may also identify four phases: • the so-called «Hellenistic» phase, corresponding to the stratigraphic levels achieved so far in Trench no. 5; • the so-called «post-Hellenistic» phase, starting from the 2nd century BC up, probably, to the 1st century AD, and corresponding to the lower layers of Trench no. 6; • the Kuśan chronological stage, corresponding to the layers of the central citadel (Trench no. 5) and to the area of reutilization of the walls (Trench nos. 1, 2, 3, 4, 6); • the last phase, from 5th to 7th century AD, corresponding to the superficial layers of the site.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Pottery from the Joint Uzbek-Italian Archaeological Mission at Kojtepa (Uzbekistan)

627

Bibliography Abdullaev, K./Genito, B., 2011. Trial Trenches at Koj Tepa, Samarkand Area (Sogdiana), Third Interim Report. Newsletter Archeologia (CISA) 2. Napoli, 7–72. –

2012. Trial Trenches at Koj Tepa, Samarkand Area (Sogdiana), Fourth Interim Report 2012. Newsletter Archeologia (CISA) 3. Napoli, 9–85.

Dimartino, R., 2011. Studio analitico della cultura materiale fra VII e IX secolo d.C. nella Regione di Samarcanda (Uzbekistan): analisi morfo-tipologica, produzione e commercio della ceramica di Kafir Kala,. Tesi di Dottorato (inedita) in Bisanzio ed Eurasia, XXIII Ciclo, Università di Bologna. Genito, B./Gricina, A., 2009. The Achaemenid Period in the Samarkand Area (Sogdiana). Newsletter Archeologia (CISA) 0. Napoli, 122–141. –

2010. The Achaemenid Period in the Samarkand Area (Sogdiana): Trial Trenches at Koj Tepa 2009 Campaign. Newsletter di Archeologia (CISA) 1. Napoli, 113–161.

Genito, B./Raiano, F., 2011. Ceramics from Koj Tepa (Samarkand Area-Uzbekistan). A Preliminary Study Report (2009–2010). Newsletter di Archeologia (CISA) 2. Napoli, 103–177. Grigor’ev, G. V., 1946a. Tali Barzu kak pamjatnik domusul’manskogo Sogda. KSIIMK 12. Moskva, 150–153. –

1946b. K voprosu o hudojestvennom remesle domusul’manskogo Sogda. KSIIMK 12. Moskva, 94–103.

Guljamov, Ja. G. (ed.), 1969. Afrasiab. Afrasiabskaja kompleksnaja arheologičeskaja ekspedicija, Vol. I. Taškent. –

1973. Afrasiab. Material’ij Afrasiabskoj arheologičeskoij kompleksnoij ekspedicii. Sbornik Statej, Vol. II. Taškent.

–

1974. Afrasiab. Sbornik Statej, Vol. III. Taškent.

–

1975. Afrasiab. Sbornik Statej, Vol. IV. Taškent.

Isamiddinov, M.K., 2002. Istoki gorodskoj kul’tury samarkandskogo Sogda (Problemy vzaimodejstvija kul’turnyh tradicii v epohu ranneželeznogo veka i v period antičnosti). Taškent. Kabanov, S.K., 1969. Izučenie stratigrafii gorodišča Afrasiab. SA 1. Moskva, 183–198. –

1973. Stratigrafičeskij raskop v severnoi časti gorodišča Afrasiab. Ja. G. Guljamov (ed.), Afrasiab, Vol. II. Taškent, 16–84.

Košelenko, G. A. (ed.), 1985. Drevneišje gosudarstva Kavkaza i Srednej Azii. Arheologija SSSR. Moskva. Kruglikova, I. T., 1974. Dil’berdžin (raskopki 1970–1972 gg.). Moskva. –

1986. Dil’berdžin – Hram Dioskurov, Moskva.

Kruglikova, I. T., Pugačenkova, G.A., 1977. Dil’berdžin (raskopki 1970–1973 gg.). Moskva. Lhuillier, J./Isamiddinov, M./Rapin, C., 2012. Rannedželeznij vek severnogo Sogda: harakteristika i predvaritel’naja tipologičeskaja hronologija. In IMKU, Vol. 37. Samarkand, 57–66.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

628

Raiano

Lhuillier, J., 2013. Les cultures à céramique modelée peinte en Asie Centrale méridionale. Dynamiques socio-culturelles à l’âge du Fer Ancien (1500–1000 av. n.è.). Mémoires de la Mission Archéologique Française en Asie Centrale, Tome XIII. Paris. Lyonnet, B., 1997. Prospections archéologiques en Bactriane orientale, Vol. 2 (1974–1978). Céramique et peuplement du chalcolithique à la conquête arabe. Mémoires de la Mission Archéologique Française en Asie Centrale, Tome 8. Paris. –

2001b. Les Grecs, les Nomades et l’indépendance de la Sogdiane, d’après l’occupation comparée d’Ai Khanoum et de Marakanda au cours des derniers siècles avant notre ère. BAI 12. Bloomfield, 141– 159.

–

2010a. D’Aï Khanoum à Koktepe. Questions sur la datation absolue de la céramique hellénistique d’Asie Centrale. In K. Abdullaev (ed.) Tradicii Vostoka i zapada v antičnoij kul’ture Sredneij Azii. Samarkand, 141–153.

–

2013a. La céramique hellénistique en Asie Centrale. N. Fenn/C. Römer-Strehl (eds.), Networks in the Hellenistic World according to the pottery in the Eastern Mediterranean and beyond. Oxford, 351–368.

Mandel’štam, A. M., 1966. Kočevniki na puti v Indiju. MIA N. 136. Moskva-Leningrad. –

1975. Pamjatniki kočevnikov kushanskogo vremeni v severnoj Baktrij. Leningrad.

Mantellini, S., 2003a. The Dargom Canal and the Early Settlement of the Middle Zeravshan Valley. S. Pagani (ed.), Italo-Uzbek Scientific Cooperation in Archaeology and Islamic Studies: An Overview (Rome, January 30, 2001). Roma, 41–48. Mantellini, S./Berdymuradov, A. E., 2005. Archaeological Explorations in the Sogdian Fortress of Kafir Kala. Ancient Civilizations from Scythia to Siberia 11.1–2. Leiden, 107–131. Mukhamedjanov, A. R. et al., 1982. Keramika nižniklh sloev Bukhary (opyt predvaritel’noj periodizacii). IMKU 7. Taškent, 81–97. Nemceva, N.B., 1969. Stratigrafija južnoj okrainy gorodišča Afrasiab. J. G. Guljamov (ed.) Afrasiab, Vol. I. Taškent, 153–205. Pidaev, Š., 1978. Poselenja kušanskogo vremeni severnoj Baktrii. Taškent. –

1987. Stratigrafija gorodišča Starogo Termeza v svete novijh raskopok. Gorodskaja kul’tura Sogda i Baktri-Toharistana v antičnuju i rannesrednevekovuju epohu. Taškent.

Pugačenkova, G. A., 1966. Halčajan. K probleme hudožestvennoj kul’tury Severnoj Baktrii (Halčajan). Taškent. –

1989. Drevnosti Miankalja. Taškent.

Pugačenkova, G. A./Bel’jaeva, T.V., 1987. K stratigrafii Kumijškent-Tepe. SA, Vol. 4. Moskva, 74–86. Pugačenkova, G. A. et al., 1978. Dal’verzintepe. Kušanskii gorod na juge Uzbekistana. Taškent. Raiano, F., 2012. Ceramics from Kojtepa (Samarkand Area – Uzbekistan): Second Interim Report – 2011. Newsletter di Archeologia (CISA) 3. Napoli, 337–373. –

2013. Ceramics from Kojtepa (Samarkand area – Uzbekistan): Third Interim Report – 2012. Newsletter di Archeologia (CISA) 4. Napoli, 303–340.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Pottery from the Joint Uzbek-Italian Archaeological Mission at Kojtepa (Uzbekistan) –

629

2014a. Hellenistic and post-Hellenistic Pottery from Kojtepa. Newsletter di Archeologia (CISA) 5. Napoli, 455–477.

–

2014b. La Sogdiana tra il periodo ellenistico (III sec. a.C.) e le “invasioni nomadiche” (II sec. a.C. – VI sec. d.C.). L’evidenza archeologica di Kojtepa (area di Samarcanda – Uzbekistan). Tesi di Dottorato (inedita), sotto la direzione di Bruno Genito e Bertille Lyonnet. Università degli Studi di Napoli “L’Orientale” – École Pratique des Hautes Études, Paris.

Rapin, C./Isamiddinov, M., 2013. Entre sédentaires et nomades : les recherches de la Mission archéologique franco-ouzbèke (MAFOuz) de Sogdiane sur le site de Koktepe. J. Bendezu-Sarmiento (éd.), L’Archéologie française en Asie centrale. Cahiers d’Asie centrale N. 21–22. Paris, 113–133. Rtveladze, E., 1987. Les édifices funéraires de la Bactriane septentrionale et leur rapport au zoroastrisme. F. Grenet (éd.) Cultes et monuments religieux dans l’Asie centrale préislamique. Paris, 29–40. Šiškin, V. A., 1963. Varakša. Moskva. Šiškina, G. V., 1974. Keramika hontsa IV–II vv. do n. e. Ja. G. Guljamov (ed.) Afrasiab. Material’ij Afrasiabskoj arheologičeskoij kompleksnoij ekspedicii. Sbornik Statej, Vol. II. Taškent, 28–51.

Fabiana Raiano, Università degli Studi di Napoli «L’Orientale» – École Pratique des Hautes Études, Paris.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 631–642

Victor Sarianidi – Nadezhda Dubova

Types of Graves at Gonur Depe Bronze Age Site in Turkmenistan Types of more than 5000 funeral construction (4408 tombs with remains) and their distribution around the territory of Gonur Depe Bronze Age site of Turkmenistan are described. Shafts (53,4%), ordinary (30,5%) and huge (0,15 %) pits, cists (6,1%), chambers (1,7%), sepultures (0,2%), burials in jars (2,1%) and in ovens (0,2%), one dakhma and five fractional burials were found. All information in this article was prepared and discussed together by prof. V. Sarianidi and N. Dubova. The small abstract of it, which is published in the volume of 9. ICAANE, was also written together. Unfortunately, after the passing of Prof. Victor Sarianidi on December 22th, 2013, the preparation of the text and the illustrations were made by only one author, so all mistakes and misunderstandings are her’s only. Gonur Depe (2300–1600 Cal BC), situated in the South-Eastern Turkmenistan in the KaraKum desert, is a typical and the brightest monument of the Bactria-Margiana archaeological complex (BMAC) and has some cultural parallels with Near Eastern cultures and with Indus civilization as well (Sarianidi 1990, 2002, 2005, 2007 etc.). It has been excavated since 1974. Since the 2000s, the work at the site is conducted within the framework of a Cooperation Agreement between the Ministry of Culture of Turkmenistan and the Institute of Ethnology and Anthropology named by. N. N. Miklukho-Maklay RAS. These excavations have shown the high level of the culture of the population of Mouru-Margush-Margiana-Merv oasis in the Bronze Age, and made possible to assert that this region formed one of the Centers of ancient Oriental civilization (Sarianidi 2006; Turkmenistan 2006). In addition to the monumental architecture, the cults of fire, water and, animals, fine examples of jewelry, a significant number of gold, silver artifacts, many testimonies of crafts, mosaics unique in their technique and image quality, and other finds, the different burial structures occupy a prominent place showing a significant social stratification of Margiana society. More than 5000 tombs in various cemeteries have been excavated:

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

632

Sarianidi – Dubova

Fig. 1. General scheme of Gonur Depe and the satellite sites.

2831 tombs at the Main Gonur necropolis (2250–1600 Cal BC) (Sarianidi 2001, 2007), 1626 in different parts of the Palace temple complex (they were made when Gonur lost its significance as a «holy city»; 1800–1600 Cal BC), 109 in the Royal Gonur necropolis (61 of them belong to 2250–2000 Cal BC) and 106 in some satellite settlements close to Gonur (2000–1600 Cal BC) (Sarianidi/Dubova 2012: 39–52) (fig. 1). All graveyards show diversity of burial constructions. Shafts (pits with shaft in one wall), cist (pits with mud-brick walls and roof), ordinary pits, chambers (underground models of one-two rooms houses) graves, sepultures (large houses under the ground), burials in jars and in ovens, one dakhma in Gonur Palace and five fractional burials (3 at Main Gonur necropolis and 2 in the ruins of North Gonur) were found. Only 12,8% of the funeral constructions cannot be identified. Shaft graves (53,4%) make up the majority of the total amount of graves found at Gonur oasis, ordinary pits (30,6%) take the second place, cists (6,1%) the third one, the burials in jars (2,4%) the fourth one, and burials in chambers (1,7%) the fifth. Very rare burials were

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Types of Graves at Gonur Depe Bronze Age Site in Turkmenistan

Fig. 2. Types of tombs at Gonur Depe.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

633

634

Sarianidi – Dubova

Type of tomb Shaft Shaft without step Pit

Main Necropolis

Royal Necropolis N

%

R1

N

%

N

2096

79, 2

7

7, 9

37

0

0, 0

0

0, 0

R2 %

R3

All territories

N

%

N

%

N

%

35, 9

119

9, 6

77

23, 6

2336

53, 0

2

1, 9

13

1, 0

1

0, 3

16

0, 4

225

8, 5

63

70, 8

34

33, 0

868

69, 9

155

47, 5

1345

30, 5

Large pit

0

0, 0

1

1, 1

1

1, 0

3

0, 2

1

0, 3

6

0, 1

Burnt pit

194

7, 3

0

0, 0

1

1, 0

4

0, 3

5

1, 5

204

4, 6

Fire place

31

1, 2

0

0, 0

0

0, 0

0

0, 0

0

0, 0

31

0, 7

Chamber

52

2, 0

2

2, 2

0

0, 0

10

0, 8

10

3, 1

74

1, 7

Sepulture

0

0, 0

5

5, 6

1

1, 0

4

0, 3

0

0, 0

10

0, 2

49

1, 9

8

9, 1

23

22, 3

124

10, 0

62

19, 0

266

6, 0

Large cist

0

0, 0

0

0, 0

0

0, 0

0

0, 0

4

1, 2

4

0, 1

In Jars

0

0, 0

3

3, 4

4

3, 9

78

6, 3

8

2, 5

93

2, 1

Jar in the pit

0

0, 0

0

0, 0

0

0, 0

10

0, 8

2

0, 6

12

0, 3

In old oven

1

0, 0

0

0, 0

0

0, 0

9

0, 7

1

0, 3

11

0, 2

Cist

All types

2648

89

103

1242

326

4408

R 1 – Tombs outside North Gonur - at South Gonur, Gonur 20 and Gonur 21. R 2 – Tombs inside the encircling wall of North Gonur R 3 – Tombs outside the encircling wall of North Gonur, but close to it

Table 1. Distribution of different types of funeral constructions by the territories of Gonur oasis.

made in huge (diameter about 5–7 m) pits (0,1%), in sepultures (0,2%) and in old ovens (0,2%) (fig. 2). Huge cists (length about 6 m) were found only four and only in one area (Area 19 on the North-West of North Gonur). Specific type of pits, where only sick people were buried, are named ‹burnt pits›, pits with burnt walls (but not floors!). Most of these pits (77% of 204) were empty; they were found, as the 31 fire places (small pit, diam. not more than 50 cm, with traces of hard fire) at the Main necropolis (in table 11 only constructions with identifiable type are mentioned).

1

In the tables and figures only tombs with human or animal remains are analyzed.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Types of Graves at Gonur Depe Bronze Age Site in Turkmenistan

635

Fig. 3. The comparison of the distribution of different tombs’ types at North Gonur and other territories (South Gonur, Gonur 20 and Gonur 21).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

636

Sarianidi – Dubova

The comparison of the distribution of different types of tombs by territories provides interesting information. As can be seen on fig. 3 in Gonur and at other settlements about 30% of the burials are in pits; shaft tombs are rarer at others settlements, but not in North Gonur. The difference is 17,5%. More significant is the difference in the share of cist: more often (4 times!) they are found outside North Gonur. But at the same time only one sepulture from the total 10 ones was found outside the main site. It was found at Gonur 21, a satellite settlement 4 km to the south of North Gonur that looks like a small «feudal castle». It seems that such a «feudal lord» or «large landowner» (we have no idea about the form of the property at that time) was to be buried in a sepulture but not in the cist or shaft. Additionally it can be said that not one chamber tomb was made outside the main city. The large share of cists outside the main city corresponds to the twice larger percentage of them outside the encircling wall2 of North Gonur (fig. 4) than inside it. It may be a sign that even representatives of the top of Margush community (the ruler or governor-High Priest was the only exception) could not bury their dead in the sacred part of the city. But they tried «to be so close to it» as possible and made tombs=cist just near the encircling wall. Noteworthy are the percents of the burnt pits: they are significant (4,7%) at Gonur Depe and only occasional at other settlements (1,0%, only one case). The fireplace like ones were found at the Main Gonur Necropolis, but they are absent in other territories. That situation together with an amount of empty burnt pits there highlights the specific rituals associated with fire which took place at this graveyard itself. There were real fire-altars near the sepultures in the Royal Gonur necropolis (see for example a reconstruction of thouse near the tomb 3235 – Sarianidi/Dubova 2013: 206–207) and many other traces of fire cult at this site (see Sarianidi 1998; 2001; 2010: 27–39, 130–134). The signs are not as clear as, a little later (2nd mill. B.C.), at the graveyard of the Sherabad oasis in Southern Uzbekistan (Avanesova 2013). But many parallels can be seen between these two famous monuments. Some of them are very important, and we need to take them into account when discussing two problems: 1) the interactions between nomadic (and semi-nomadic) tribes with agriculturalists communities; 2) only correspondences between the fire worshipers, early Indo-Iranians and nomads/ cattle breeders of the Eurasian Steppe. It has been mentioned earlier that the main part of the tombs at Main Gonur necropolis were made as shafts (Sarianidi 2001; 2007). At the famous Chalcolithic-Bronze Age site Altyn Tepe situated near Kopet-Dagh foothills there are no such types of tombs (Kircho/Alekshin 2005: 473–474). This type of tombs is rare (9,6%) among the burials inside the encircling wall (fig. 4); they are three times more frequent outside it (23,6%). At the small graveyard close to the main site, the so called «Cemetery of the Temenos» (Dubova/Muradova 2008) 2

The encircling wall – is the third, not a defensive wall surrounding the city (fig. 1). It was, as it seems to us, the border between sacred, sacral (inside) and common (outside) space at Gonur.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Types of Graves at Gonur Depe Bronze Age Site in Turkmenistan

Fig. 4. The comparison of the distribution of different tombs’ types inside and outside the encircling wall of North Gonur.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

637

638

Sarianidi – Dubova

9 out of 11 (81,8%) tombs were shaft ones. So were 36,6% of the 71 tombs at Gonur 20 and 19% of the 19 ones at Gonur 21; only 2 tombs out of 236 (0,8%) in the territory of Gonur Palace and Kremlin and 7 out of 108 (6,5%) tombs of the Royal necropolis were shafts. If we analyze the distribution of shaft tombs in the territory of the North Gonur, it can be seen that all excavation areas3 can be divided into two groups: 1) eight defined areas with only occasional shafts (1,6–10,4% of the total amount of the graves); 2) five areas, where shafts amount to one-third (or slightly more) from the total number of graves (30,3–42,9%). Area 22 which has the maximum of such constructions among 28 tombs is situated to the south of the Royal necropolis (outside the encircling wall). Close to it is area 6 (west face of the site between second and encircling wall) which has 41,5% shafts among 65 graves. This type of graves accounts for about one-third (30–34%) of the total amount of graves at areas 12, 16 and 17. Two of the last ones are marked at the south-western side of Gonur, outside the encircling wall, the 17th – to the north of Gonur, again outside the third wall. Area 16 (242 tombs) is situated on the right bank of ancient Murghab river while on the left bank just in front of it the Main Gonur necropolis ends (Sarianidi/Dubova 2008: 28–49). When the Murghab river became smaller and smaller and the water disappeared from that place people of Margush began to use this area for burial. So, area 16 became an extension of the Main burial ground at the end of the second chronological period of the site, although in the earliest times there were some inhabited areas. The placement of area 12 is also special: it is marked to the north-east of area 16 by the remains of a wall that may have encircled the whole city, though it is very far to the west of the main wall. The excavations in 2013–2014 showed that there was a graveyard of one family or of a group of relative families (it became clear thanks to the anthropo-genetic traits of the people buried and the construction of the tombs). We may suppose that this burial ground was not the only one. The similar shares of shaft tombs at different territories tell us about the possibility of presence of some local graveyards at Gonur Depe. At the same time this data may correct ideas on the social structure of Margush country. As it has been proved, the ratio of different types of graves at the Gonur necropolis is a reflection of the social structure of the country: 80% of the middle class (shaft tombs), 10% of poor (pit tombs) and 10% of rich people (chambers and cists) (Sarianidi 2001, 2007). Here it is necessary to clarify that Gonur was not a city where people lived. It was a palace-temple center, where the inhabitants of different settlements came to carry out ceremonies and rituals, to make sacrifices and pray. Only the ruler and may be his family lived there. As it can be supposed the cemetery on the other side of the ritual center bank of the Murghab river was 3

The location of twenty archaeological areas recorded at Gonur Depe was described many times in publications. On the cover of Volumes 4 and 5 of Transactions of Margiana Archaeological expedition (Moscow 2012, 2014) their positions are marked on the schema.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Types of Graves at Gonur Depe Bronze Age Site in Turkmenistan

639

Fig. 5. The comparison of the distribution of different tombs’ types at Main Gonur Necropolis and at the territory of the central part of Palace-Temple complex of North Gonur («Ruins of the Complex»).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

640

Sarianidi – Dubova

Fig. 6. Schema of the cist 2900 at area 5 of North Gonur.

organized by people of the surrounding settlements because the area near the «holy city» had become sacred. And nowadays modern Turkmen try to bury their dead close to the old cemeteries, close to the tomb of famous or holy men. But to carry the body and perform the necessary ceremonies far from home requires considerable effort which only the wealthy are able to make. Only few common people will do that. The proportion of different types of tombs at small graveyards, especially those which belong to one settlement (for example our Gonur 20) is more demonstrative. As we have tried to show, about 30–40% was above the middle class (sign of which may be the shaft tombs). The share of pits as well as cists is about 30% each. May be this proportion is closer to the real one? Burials in large ceramic vessels, jars are absent in the main Gonur necropolis. They were found only on the territory of the main part of the palace-temple complex. These burials were organized in the last period of occupation of the city (later they are named as «Tombs in the ruins of the Complex»). Only babies and children were buried in them. The large difference can be seen when comparing the distribution of types of tombs in the Main Gonur necropolis and in the «Ruins»: the majority of tombs in the Necropolis, as it was said above, are shafts (79,2%), but in the «Ruins» pits (65,2%) (fig. 5). Following the logic just described, we can assume that the last inhabitants of Gonur Depe were poor people who could not leave this place together with the top of the society which had moved to the another oasis when there was no longer enough water. One more sign of the same phenomenon is the absence of the sepultures and chamber tombs

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Types of Graves at Gonur Depe Bronze Age Site in Turkmenistan

641

which are dated later than 19th–18th cc. B.C.4 The latest data from the main part of Gonur belongs to unplundered cist 2900 (Sarianidi 2007: 146–155). It got two names – «A Burial of Noble Warrior» and «A Burial of the Last Ruler of Gonur» due to the richest funerary offerings among contemporaneous graves. But it was not a chamber or sepulture, it was a typical cist (fig. 6). So in the last period of life at Gonur those who buried their dead in cists were the richest people. Cultural differences between people buried in different types of graves are not traced. This short analysis of the distribution of the types of tombs in the territory of Gonur Depe and some satellite settlements shows some regularity. It helps us to understand not only the dynamics of formation of some graveyards there but also to come closer to the knowledge of the social structure of the society. One can hope that future excavations will give us new information, including those about the funeral rituals. For example huge cists mentioned above were found in Gonur only in 2010. And their placement was unexpected: outside the encircling wall, close to it, but not in a free space where there were most of the previously found burials of this kind but close to some buildings and other constructions on the northwestern corner of the site – Area 19 (Sarianidi/Dubova 2014: 98–99). Also new information was obtained during the excavation on Area 12 in the south-western part of the site, where among other things some shafts without step and some variants thereof were excavated.

Bibliography Avanesova, N. A., 2013. Buston VI – Necropol ognepoklonnikov dourbanisticheskoi Bactrii [Бустон VI – некрополь огнепоклонников доурбанистической Бактрии] [Buston VI – Necropolis of the Fire-Worshipers of the Preurban Bactria]. Samarkand. Dubova, N. A./Muradova, E. A., 2008. “Mogilnik Temenosa” Gonura [“Cemetery of the temenos” of Gonur], Transactions of Margiana Archaeological Expedition 2. Moscow, 105–111 (This book and other publications on Gonur Depe are available in Internet: http://margiana.su/ index.php/publications.html). Kircho, L. B./Alekshin, V. A., 2005. Khronologiya epoh pozdnego eneolota – sredney bronzy Srendey Asii (pogrebeniya Altyn Depe) [Хронология позднего энеолита – средней бронзы Средней Азии (погребения Алтын Депе)] [Chronology of the Late Chalcolithic – Middle Bronze Age of Central Asia (burials of Altyn Depe)]. Sankt-Petersburg. Sarianidi, V. I., 1990. Drevnosti strany Margush [Древности страны Маргуш] [The Antiquity of Margush country]. Ashgabad. 4

There are more than 80 C14 data for Gonur Depe, the main part of which is published: Zaytseva et al 2008.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

642

Sarianidi – Dubova

–

1998. Margiana and Protozoroastrism. Athens.

–

2001. Necropolis of Gonur and Iranian paganism. Мoscow.

–

2002. Margush: Ancient Oriental Kingdom in the Ancient Delta of Murghab River. Ashgabat.

–

2005. Gonur-Depe. Turkmenistan. City of Kings and Gods. Ashgabat: Turkmendowlethabarlary.

–

2006. Tsarskiy necropol na Severnom Gonure [Царский некрополь на Северном Гонуре] [The royal necropolis in North Gonur], Vestnik drevney istorii [Вестник древней истории] 2(257), 155–192.

–

2007. Gonur necropolis. Athens.

–

2007. Dvortsovo-khramovyi kompleks Severnogo Gonura [Дворцово-храмовый комплекс Северного Гонура] [The palace-temple complex of North Gonur], Rossiyskaya arkheologiya [Российская археология] 1, 49–63.

–

2008. Margush. Mystery and truth of the great culture. Ashgabat.

–

2010. Long before Zaratushtra. Moscow.

Sarianidi, V. I./Dubova, N. A., 2008. Archeologicheskiye raboty na yugo-zapadnom kholme Gonur Depe (Raskop 16) [Археологические работы на юго-западном холме Гонур Депе (Раскоп 16)] [Archaeological work on the south-western hill of Gonur Depe (Area 16)], Transactions of Margiana Archaeological Expedition 2. Moscow, 28–83 . –

2012. Archeologicheskiye raboty Margianskoy archeologicheskoi expeditsii v 2008–2011 godah [Археологические работы Маргианской археологической экспедиции в 2008–2011 гг.] [Archaeological Excavations of the Margiana Archaeological Expedition during 2008–2011 years]. Transactions of Margiana Archaeological Expedition 4. Moscow, 29–55.

–

2013. Treasures of Ancient Margiana. Photoalbum/Wiktor M. Hramov (Ed.), Ashgabad.

–

2014. Raboty Margianskoy archeologicheskoi expeditsii v 2011–2013 gg. [Работы Маргианской археологической экспедиции в 2011–2013 гг] [The work of the Margiana archaeological expedition in 2011–2013]. Transactions of Margiana Archaeological Expedition 5. Moscow, 92–111.

Turkmenistan, 2006. Ancient Margiana is the New Centre of the world civilization. Ashgabad. Zaytseva, H. I./Dubova, N. A./Sementsov, A. A./Reimer, P./Mallory, J./Jungner, H., 2008. Radiouglerodnaya hronologiya pamyatmika Gonur Depe [Радиоуглеродная хронология памятника Гонур Депе] [Radiocarbon chronology of the Gonur Depe site], Transactions of Margiana Archaeological Expedition 2. Moscow, 166–179.

Victor Sarianidi & Nadezhda Dubova, Institute of Ethnology and Anthropology, Russian Academy of sciences, Moscow, Russia. [email protected].

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 643–645

Liliya Sataeva – Robert Sataev

Wood Using at the Bronze Age Site Gonur-Depe (Ancient Margiana, South Turkmenistan) Wood remains are presented by coals from furnaces, building constructions, fragments of products. There were saxaul, tamarisk, maple, arboresced thistle, camel-thorn as fuel. Poplar and willow were used in buildings structures. Cartwheels from the rich burial 3900 were made from elm. Stalks of shovels were made from Cotinus coggygria (Smoke tree). Studying plant remains from archaeological excavations brings very important information about culture and everyday life of ancient peoples. Gonur-Depe (Grey Hill) is one of the largest and most well-known sites of the Middle Bronze Age of Central Asia, the capital city of Ancient Margiana (Margush) country. It is located in the sands of the southeastern part of Karakum Desert, South Turkmenistan. The site has been investigated since the early 1970th by the Margiana archaeological expedition under the leadership of Victor Sarianidi. It was inhabited from 2300 to 1500 BC (Sarianidi 2004; Zaytseva et al. 2008). During archaeological excavations at Gonur-Depe different forms of conservation of plant residues were identified: charcoal, carbonized seeds and fruits, wood (mostly conserved copper salts), imprints of plants (on ceramics, bricks, daubing, various structures in the floor) (Sataeva 2014, Sataev/Sataeva 2014). The study of the coals showed that saxaul (Haloxylon sp.), shrubby forms thistle (Salsola sp), willow (Salix sp.) and tamarisk (Tamarix sp.) were used as fuel. In ritual and pottery furnaces only wood fuel was used. The coals from the territory of the palace-temple complex belong to tamarisk, haloxylon, buckwheat, poplar (Populus sp.), Spireantus schrenkianus. The last species is endemic to the other major desert – Kyzyl Kum (where this shrub is used as fuel); nowadays it does not grow in the territory of the Karakum desert, but in the past its area was probably much wider. Next to this, in residential furnaces, lignified stems of camel thorn (Alhagi pseudalhagi) were burned after fruiting and drying on the vine, perhaps gathered from dry plants occurring in late autumn or winter.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

644

Sataeva – Sataev

In the studied material, small twigs of haloxylon prevail; this can not be explained by taphonomic reasons, because large coals remain not worse. Probably this is due to the difficulty to provide whole trunks – hardwood of haloxylon is difficult even for modern cutting tools, the more to the ancient ones; such a strategy of fuel provision also prolongs timber reserves. Haloxylon is very resistant to cuttings, the tree does not die even after considerable denudation, and regains its crown. Considering the amount of wood fuel used, including haloxylon, we may assume significant primary reserves of woody vegetation in the surrounding city area. We also analyzed the accumulations of charred wood remains under the internal pilasters that strengthened the outer wall of the Kremlin. They are the remains of the roof which extends between the inner and outer walls of the Kremlin, which had fallen at a heavy fire. In the construction the bearing beams were poplar trunks (diameter of beams about 15 cm, length about 10 m). Across them were installed thick branches from poplar wood with a diameter of 5 cm. Over the poplar beams were placed willow branches (perhaps the fence), higher on there were mats of reed stems; the top side was covered with daubing reinforced with willow twigs. A special place is taken by finds of the decayed wood representing fragments of products and designs. Their analysis also shows wide taxonomical composition. M. Tengberg (Tengberg 2008) identified wood from the «Royal tombs» as willow (Salix sp.), elm (Ulmus sp.) and ash (Fraxinus sp.); elm and ash were only single samples. M. Tengberg suggests that part of the wood (especially elm) used for making items, could come from the foothill areas of the Kopet Dagh. On the settlement a few «royal tombs» had carts (# 3200, 3225, 3240, 3900, 3915). The study of the rich burials # 3900 and 3200 showed that the wheels of the carts were made from elm (Ulmus sp.). The wheels had bronze cover plates («tires»); where the wood had contact with bronze, it was preserved, and its identification possible. Growth rings are visible to the naked eye. Microscopic study showed the ringed wood with large spring vessels with a diameter of 200–250 microns. Small vessels of the autumn wood collected in strips have the oblique tangential direction. Such microscopic structure is typical for elm (Ulmus), and some features are transitional between several kinds of elm. At the moment we have not observed elm near Gonur Depe, but all of these species are found in the foothills of Kopet Dagh, where the wheels could have been made. The same identification as elm was made by M. Tengberg on burial 3200. In the rich burial 3900 a bronze shovel was found with handle that was made of wood of the smoke tree (Cotinus coggygria). The wood of the smoke tree contains a milky sap that is resistant to decay, and can be used for tanning leather and hides, fabric dyeing, as rot-props in the vineyards, as well as for inlaying art products. It is not currently found in Turkmenistan. Its natural habitat is the Mediterranean, the Caucasus, the Crimea, Asia Minor, China

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Wood Using at the Bronze Age Site Gonur-Depe

645

and the Himalayas. Perhaps in the 3rd millennium BC the area of distribution of the smoke tree included oases of Central Asia, but it is possible that for this shank the tree was imported from the west (Caucasus or Asia Minor) or east (from the Himalayas through Indus valley). It is archaeological evidence of contacts between the Margiana and other civilization. Analysis of prints of branches detected on the inner layers of gypsum plasters covering massive vessels – hulls and walls of stationary tanks («bath» and «tanks») made of adobe clay – shows that they belong to the willow. Likely the willow twigs were used for reinforcing the binder. Thus, the inhabitants of Gonur-Depe used a wide range of woody plants, some growing in the area and some imported from afar. Researches were supported by RFBR grant № 13-06-00233.

Bibliography Sarianidi, V., 2004. Palace-temple ensemble of North Gonur. Near the Sources of Civilizations. Moscow, 229–253. Sataeva, L., 2014. Plants in Life of the Population of Gonur-Depe (Ancuent Margiana). Bulletin of Moscow University. Series XXIII. Anthropology 3, 109. Sataev, R./Sataeva, L., 2014. Results of Archaeozoological and Archaeobotanical Research at the Bronze Age Gonur Depe Site (Turkmenistan). Proccedings of the 8th International Congress on the Archaeology of the Ancient Near East (8. ICAANE), vol. 3. Wiesbaden, 369–372. Tengberg, M., 2008. Analysis of wood samples from Gonur (Margush), Turkmenistan. Transactions of Margiana archaeological expedition 2. Moscow, 163–164. Zaytseva, H. I./Dubova, N. A./Sementsov, A. A./Reimer, P./Mallory, J./Jungner, H., 2008. Radiocarbon chronology of the Gonur Depe site. Transactions of Margiana archaeological expedition 2. Moscow, 166–179.

Sataeva L., Bashkir State Agrarian University, Ufa, Russia. Sataev R., Institute of Ethnology and Anthropology RAS, Moscow, Russia.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 647–661

Judith Thomalsky

Afghanistan: Ancient Mining and Metallurgy: Initial Project Stage The rich mineral deposits in Afghanistan are widely known and were used since prehistoric times. Lapis lazuli was highly favoured for precious items and thus widely spread as far as Egypt from the 4th millennium BCE onwards. A similar distribution pattern is documented for gold and tin. For the latter, an important linkage with the early appearance of tinbronzes in the 3rd millennium BC Mesopotamia suggests that Afghanistan’s deposits may play a major role in the invention of Bronze metallurgy.

Introduction Ancient mining traces became recently known from the huge copper deposit of Mes Aynak, located 30 km southwest of Kabul (Coats 2005; Litecka et al. 2013). The site became particularly famous for the ancient Buddhist city on the route of the Silk Road, dating from the early centuries AD, now under the surveillance of UNESCO and other institutions, but still endangered by commercial exploration in near future. Rescue excavations supervised by the Afghan government cover the archaeological investigation of the direct vicinity of Mes Aynak, which unearthed ancient copper working, smelting workshops and miners habitations.

Project Aims and Schedule The Afghanistan Project of the Eurasia Department of the German Archaeological Institute aims to develop a cooperation of scientific, economic and cultural policy related institutions, and to provide a long-lasting partnership for future interdisciplinary researches. Recent engagement of the German Archaeological Institute in Afghanistan is primarily associated with Ute Franke in Kabul, Bamyan and Herat since 2005. She further conducted surveys in

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

648

Thomalsky

the vicinities of Bamiyan and Herat, hand in hand with special educational programmes in archaeology and restoration and in cooperation with the Afghan’s Ministry of Information and Culture and the Agha Khan Trust Organisation (Franke 2008; Franke-Vogt et al. 2005). Contemporaneous investigations of the Eurasia Department on ancient tin mining in Central Asia give a broad and interdisciplinary approach in early mining archaeology and its dimension for the social and technical evolution of pre-industrial societies (Alimov et al. 1998; Boroffka et al. 2002; Parzinger/Boroffka 2003). A final impulse to develop this new project was the colloquium «Cultural Heritage and Archaeology in Afghanistan», 21.–22.11.2012 in Berlin. Applied methods comprise the collection of archaeological and mineral data into a GIS supported database and the analysis of metallurgical samples. Assuming the initial stage of the Afghanistan project, one estimated and concentrated on the following major issues: • The systematic sampling of ores and ancient metal objects from Afghanistan for geochemical composition and provenance studies, including • Geochemical analysis • Neutron activation and isotope analysis • RFA (X-ray fluorescence analysis) • The creation of a comprehensive GIS database of the mineral occurrences and ancient sites as a documentation framework, in a web-based open access solution.

Collecting Data The Afghanistan Project acquires data from archaeological and geological literature, and historical sources. The database of the Afghanistan Project consists of three major data sheets: 1) Archaeological sites (of all periods) 2) Mineral occurrences and geological data 3) Sample analyses – so far published/made Up to now, the Afghanistan Project registered roughly 3000 sites – that can be sorted and mapped through general determinations such as chronology, location and type of investigation, and more distinctive attributes such as site characteristics and material (fig. 1). All site data – whether mineral occurrences or archaeological sites – are provided with their relevant coordinates in order to easily connect them with GIS structures. As concerns the geological and mineral data, huge collections of different satellite imageries1 and maps are available

1

Common providers are the US American satellite programs CORONA or Landsat – see http:// landsat.usgs.gov/; http://www.nro.gov/history/csnr/corona/.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Afghanistan: Ancient Mining and Metallurgy: Initial Project Stage

649

Fig. 1. Map of registered sites in the database (all periods; Warwick Balls gazetteer– completed by the Russian investigations) © DAI, J. Thomalsky.

for free, and are consolidated for example in the listed compound of Bliss and Orris (Orris/ Bliss 2002). Of major significance are the investigations of the Afghan Geological Survey (AGS), in collaboration with different other non-governmental organisations (NGO) that collected geological data on mineral resources in Afghanistan, actually by Helicopter- Sampling – geologists and engineers were brought by military helicopters in order to gather hand-pieces of samples from selected mineralisation areas – and the systematic scanning and evaluation of the landscape by satellite and aerial hyper spectral measurements. On the other hand, any geochemical analysis that was sampled for sure during these modern commercial investigations, are not available for free. Moreover, these data cannot be regarded as statistically valued, since the open-web accessible data are not representing their true potential, as an occurrence of geological-mineral interest or as a geological deposit of commercial interest. Such an approach is currently in process, again under supervision of the already mentioned NGOs that primarily focus on commercial benefit. The possible deposits of commercial deposits are termed as Mineral Areas of Interests. Nevertheless, such investments in mapping of resource potentials recover immense information for mining archaeology. As geologists are observing and collecting ancient traces of exploration and mining artefacts such as slags and mining tools, important indicators are given for the archaeologist to define Archaeological Areas of Interests where ancient

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

650

Thomalsky

Fig. 2. Archaeological Areas of Interests (major mineral occurrences, hatched) and prehistoric sites; labeled sites with indicators of metallurgy © DAI, J. Thomalsky.

occupation and metallurgical activities are to expect and should become archaeologically surveyed (fig. 2): • Mes Aynak (30 km from Kabul in Loghar Province), polymetallic copper/iron deposit • Panjshir valley, in Eastern Afghanistan (valley of the four lions) • Balkhab copper mine Central Afghanistan • Mineral deposits in Badakhshan, primar gold mines; lapis lazuli and emeralds • Copper and tin occurrences southwest of Herat • Zarkashan-Anguri Copper and Gold Deposits, in the Prov. Ghazni and Kandahar Systematic geochemical analyses of the collected ore and slag samples – in addition to ancient artefacts – establish a geo-archaeological register of the mineral resources of Afghanistan and therefrom produced objects that were circulated throughout Western and Central Asia. Similar to the mineral data, the actual value of the site’s attributes, in respect of time depths or simply the site’s lay-out is in a very preliminary state of registration. This is especially the case for the prehistoric sites, biased by the fact that less than 1% of prehistoric sites in Afghanistan were examined through systematic excavation so far. The major amount of

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Afghanistan: Ancient Mining and Metallurgy: Initial Project Stage

651

Fig. 3. Distribution of lapis lazuli from Afghanistan, during the 4th-3rd millennium BCE (preliminary) © DAI, J. Thomalsky.

the points on the archaeological map became known during survey investigations. Thus, major aspects that concern the major period under study – the Bronze Age in Afghanistan – will be discussed in the following passages.

Archaeological framework Particularly the rise of tin bronze was the starting question of the «Afghanistan Project», considering that Afghanistan’s rich mineral deposits yield both major ingredients – the copper and tin (Muhly 1985; Muhly 1995). In other words, one can rather easily obtain a natural bronze alloy by accidental melting of these ores. For the given distances of raw material circulation throughout the Old World that we have to consider, these distinctive multi-mineral deposits are to be regarded as an almost unique phenomenon worldwide. This special precondition highly suggests Afghanistan as being the motor of one of the most important inventions of the metal periods of Western and Middle Asia. Moreover, not only copper and tin were in the focus of the ancient explorers, but any other mineral resource – respectively the lapis lazuli (Brown 1991; Casanova 2013; Herrmann 1968). Lapis lazuli objects in Predynastic Egypt indicate an already established network that linked Central Asia – the region

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

652

Thomalsky

of the later Indus civilization, Iran, Mesopotamia and the Southern Levante and Egypt, as well as the Northern regions such as Maikop in the Northern Caucasus (fig. 3). From the mid 3rd millennium BCE onwards, tin bronzes increasingly occur in the Old World, particularly far from any possible local source as it is the case in the famous necropolis of Ur in Mesopotamia. Archaeometallurgical investigations on several objects from Bronze Age Mesopotamia could establish that the use of tin bronze had begun in the first half of the 3rd millennium BCE. Later, the percentage of the total metal inventory became the dominant alloy type in the 2nd millennium (Alimov et al. 1998; Bátora/Pernicka 1999). Not only the very long distances, but the technological and social preconditions for its production are to be considered when dealing with early mining and metal production in prehistory. The metal alloy – the intentional combination of different metallic opponents in order to change the physical characteristics – is a major innovation in metal technology. The mixture of copper with other metals, initially with arsenic and antimony, causes an welcome improvement of the material properties: the arsenic copper alloy is, at the same time, more flexible and harder than native copper; and it can be cast in better and thinner conditions. Moreover, alloying metal directly affected the color of the finished product: compared to the reddish brightened copper the arsenical copper is of a silver brilliance. Tin again served for both, hardening and enhancement of casting properties but further produces a golden sheen of the metal. The increasing distribution of tin bronzes illustrates an impressive dynamic in social-economical skills during the Bronze Age and a contemporaneous pattern of globalism: as copper was indeed easily available in several local deposits, tin was quite less visible and required the logistic organization of reliable access and safe transport over enormous distances. For the archaeological context of 3rd millennium BCE Afghanistan, several aspects are to be considered: First of all, our regional and chronological focus remains difficult to determine, as we have only a very flimsy knowledge about the prehistoric cultural developments in Afghanistan; they can be summarized as follows: From mid-3rd millennium onwards, the so-called Bactro-Margiana Archaeological Complex, or BMAC – or «Oxus civilization» – stretches from the desert oasis of Margiana along Uzbekistan and Tajikistan, including Northern Afghanistan and NE Iran (Biscione/Vahdati 2011; Sarianidi et al. 2012; Tosi et al. 1992). Part of this complex is the prominent treasure (of silver and gold vessels) from Tepe Fullol (dated to the transition of the 3rd to 2nd mill. BC); an outstanding example for the metallurgical and technological know how; commonly considered to be an expert work of Mesopotamian artisans – as design and composition suggest (Hiebert 2009; Jarrige 2009; Tosi/Wardak 1972). Wide parts of South-Eastern Iran were traditionally assigned to the Helmand civilization. This Iranian-Afghanistan-Pakistan borderland established a broad cultural network already during the late Neolithic that reached a peak with the establishment of production centres

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Afghanistan: Ancient Mining and Metallurgy: Initial Project Stage

653

such as Shahr-i Sukhte. Analyses of ores, slags and bronze objects of Shar-i Sokhta indicate that arsenical copper metallurgy characterizes the assemblage – with the intentional addition of lead and arsenic to the copper (Artioli et al. 2005; Hauptmann et al. 2003). The deposits are located not far away in the mountains of Southern Dasht e-Lut. Central Iran with metal production sites such as Arisman (Helwing 2014; Vatandoust et al. 2011), located along the Western Fringe of the High Plateau evidences an contemporary local arsenical copper production (Thornton 2009; Thornton 2010; Thornton/Rehren 2009). The intentional alloy of arsenical copper was also established in the Southern Levant during the 4th millennium BCE while tin bronzes appeared clearly later in relevant percentages in the 2nd millennium BCE (Levy/Shalev 1989). On the Afghanistan’s side, there are weak indicators that similar production centres have existed. A highlight is the hoard of «Tepe Fullol» that is dated to the end of the 3rd millennium BCE. References of sites that evidenced metal production are few and not clearly determined yet (fig. 2). To be added are the sites that were mentioned in Ball’s gazetteer, e.g. site 9 in the Dashli oasis with kilns and other remains of metal production and Shortughai, where gold and copper were processed as well as lapis lazuli in large quantities (Ball 1982; Kruglikova 2005: 328 fig.13). However, with the exception of Shortughai, the information about these earliest findings of metal production remains very sparse (Francfort 1983; Francfort 1989; Francfort/Pottier 1978). Archaeological investigations in the site cluster of Gardan Reg in Sistan Province (Weisgerber 2004:196) provided 3rd millennium pottery and strong evidence of metal production, including slag and other metallurgical residue. The material culture shows strong relations to the western «neighbours» across the Afghan-Iranian border, to contemporary sites in the vicinity of Shar-e Sokhta. However, a comparative analysis of these assemblages has not been conducted, yet, a task for future research. The all-over lack of data and the very limited prospects of fieldwork caused by actual security conditions, though, the cultural dimension of mining archaeology (not only in Afghanistan) can be illustrated by the outstanding and prominent investigations in the Mes Aynak district that is shortly summed up in the following passages.

Case Study 1: Mes Aynak The copper deposit of Mes Aynak, 30 km south of Kabul, found considerable international attention since UNESCO stopped the preparations of commercial exploration in order to conduct rescue excavations in the mine district under the direction of the Afghan Ministry of Information and Culture and its affiliated Department of Archaeology (Litecka et al. 2013). All work was and is actually still under the supervision of the Délégation d’ Archéologique

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

654

Thomalsky

Fig. 4. Mes Aynak: view from SW to Southern Slope of Mes Aynak mound © DAI, J. Thomalsky.

Française en Afghanistan (DAFA).2 The work since 2006 has unearthed broad parts of a historic settlement of the Kushan and Shahi period, of the 2nd to 7th century A.D. Evidence of an older occupation – in form of Bactrian coins dating to the 3rd century BC – may suggest even older metal exploration in the vast mining district. According to the excavators it is still unclear if one can expect older occupation (and mining traces) in Mes Aynak. The work is still going on, focused on the remains of the Buddhist settlement together with stupas and up to three meters sized terracotta statues; the fortune of this impressive ancient site is clearly determined since the copper deposit is claimed as being the largest occurrence in Afghanistan and even in Central Asia.

2

I would like to express my deepest gratitude to Philippe Marquis and his DAFA team, for their outstanding engagement and enthusiasms without whom the archaeological work would not have taken place.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Afghanistan: Ancient Mining and Metallurgy: Initial Project Stage

655

Fig. 5. Mes Aynak: Traces of Extraction © DAI, J. Thomalsky.

Buddhist temples and stupas are distributed over the whole Mes Aynak district, including monasteries built in extern locations to surround and fortify the region. The internal area is densely occupied with buildings. The dwellings were built directly onto the greenish stone; green stones and metal production waste – large slag cakes – were used for floor and wall construction. As concerns ancient mining, there is only rare information – contrary to the huge slag heaps and the vast remains of copper production-related activities, scattered all over the site (fig. 4). To this day, the full assemblage of mining and metallurgical artefacts in Mes Aynak is not examined yet – if actually identified. Its registration and documentation remain for future archaeological work. The southern slope of Aynak Mountain is covered with slags that came down from the upper areas of the mountain. Here, the French colleagues excavated a one-roomed fortification wall along the top, while directly below that structure the ores have been extracted and melted (Litecka et al. 2013). Several metallurgical installations such as furnaces have been recorded, but remain unpublished. In direct vicinity of a slag heap the inhabitants or craftsmen constructed a small niche into the rock, possibly a small altar. Furnace walls suggest remains of melting ovens still in context. Other traces of extraction can be observable such as small pits that derive from controlled rock-breaking and ore hackling, as well as scratches

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

656

Thomalsky

of picks (metal tools) in the hosted mineral rocks (fig. 5). In lower elevations, several corridors were worked into the mountain, still not systematically investigated, but very likely of historic age, as a few ceramic potsherds found here date to the same period as the main settlement phase.

Case Study 2: The Panjshir Silver Mining District The above mentioned mineral district of Panjshir came recently came to the fore as the Afghan Geological Survey detected a high potential of galena and silver ores for commercial exploration (fig. 6). Nowadays, these occurrences are not in the focus of the economical investors, but exhibit a high potential for archaeologists, as the expeditions describe several areas with slag and melting residues in the mountains of Panjshir. Indeed, the mineral deposits were not unknown in the past, but in contrast intensively exploited since Saffarid times. According to Arab geographers as al Idrisi (ca. 1100–1165 AD) or Jakut (1179–1229 AD; Weisgerber 2004), the huge amount of silver from Central Asia during that era came principally from the Panjshir district and was distributed as far as to the Viking lands in Scandinavia (Steuer 2004). A number of analysed Ghaznavid and Samarid silver coins shows that the Hindukush silver is characterized by a very low gold content and often high bismuth (as high as 3 percent) that makes it «easy» to differ between distinctive silver productions. On the other hand, there was no direct sampling conducted in Panjshir concerning the ores (galena). According to geological reports the hosted rock bodies exhibit 60–65 percent iron and minor manganese, copper, lead, zinc and silver – the latter form economic concentrations. The Afghanistan project aims to fill this lack of data, and started geochemical analyses on the recently collected slags and ores. Exciting results were gained through few carbon remains inhered in the slag pieces. Radiocarbon analyses date the pieces to the 3rd and 2nd millennium BCE (fig. 7). In regard to the prehistoric occupation pattern given by the site register and data collection, no relevant prehistoric culture has been determined in the Panjshir region so far.

Perspectives Assuming, where metal working is evidenced and in particular the intentional composition of metal alloys, the use of arsenical copper is very common throughout the Ancient Old world from 4th millennium until the late 3rd millennium – thus into the Bronze Age.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Afghanistan: Ancient Mining and Metallurgy: Initial Project Stage

657

Fig. 6. Panjshir valley: view to mountain slopes covered with slag; slag and galena sample (photo: B. Bräutigam, Freiberg) © DAI, J. Thomalsky.

However, the given archaeological evidence is still not so strong in the earlier periods though relevant enough, in comparison to the collected Afghan data so far. A glimpse at the map of prehistoric sites (fig. 2) clearly demonstrates the enormous underrepresentation of adequate archaeological data. Nevertheless, one can expect ancient mining and related settlements (possibly in form of craftsmen camp-sites) in Panjshir and other mining districts.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

658

Thomalsky

Fig. 7. Panjshir valley: silver slags and radiocarbon results (CEZA Mannheim) © DAI, J. Thomalsky.

It is clear, that Bronze Age mining activities must have existed during these days, which is once more underlined by the radiocarbon data that we obtained from the slags from the Panjshir district. Finally, the commercial investments will consequently lead to the destruction of archaeological monuments in these areas. It is thus urgent and important to create a sensibility for the prominent history of mining tradition in Afghanistan and to develop a comprehensive programme for the investigation, documentation and conservation of the Cultural Heritage of Afghanistan including capacity-building; in form of educational programmes for Afghan Colleagues on modern measurement and laboratory techniques; or simply the regular access to relevant literature in form of a digital library, a work currently in progress by the Afghanistan Project of the Eurasia Department (DAI). Such a program can only be achieved with the support, generosity and strong enthusiasm of our cooperation partners and colleagues; without them this project would not have been developed so far.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Afghanistan: Ancient Mining and Metallurgy: Initial Project Stage

659

Acknowledgements

Funding The Federal Foreign Office of the Republic of Germany

Project Directors Dr. N. Boroffka, Dr. J. Thomalsky

Cooperation Partners • • • • • •

Afghan Ministry of Information and Culture & Archaeological Department Afghan National Museum, Kabul P. Marquis, J. Bendezu-Sarmiento, Délégation Archéologique Française en Afghanistan (DAFA) Dr. Bernd Bräutigam, University Freiberg Prof. Dr. Ernst Pernicka, Curt Engelhorn Center of Archaeometallurgy Prof. Dr. A. Hauptmann, German Mining Museum Bochum

Bibliography Alimov, K. et al. 1998. Prähistorischer Zinnbergbau in Mittelasien. Vorbericht der Kampagne 1997. Eurasia Antiqua 4, 137–200. Artioli, D., et al. 2005. On the Exploitation of Copper Ores at Shahr-i Sokhta (Sistan-Iran) in the 3rd Millennium BC. U. Franke-Vogt/H.-J. Weisshaar (eds.), Proceedings of the seventeenth international conference of the European Association of South Asian Archaeologists: (7–11 July 2003, Bonn), vol. 1. Forschungen zur Archäologie Aussereuropäischer Kulturen 1. Aachen, 179–184. Ball, W., 1982. Archaeological gazetteer of Aghanistan. Paris. Bátora, J./Pernicka, E., 1999. Chemische Zusammensetzung der Kupferartefakte aus dem Frühbronzezeitlichen Gräberfeld von Jelsovce, Südwestslowakei. A. Hauptmann et al. (eds.), The beginnings of metallurgy: proceedings of the International Conference “The Beginnings of Metallurgy”, Bochum, 1995. Der Anschnitt 9. Bochum, 247–254. Biscione, R./Vahdati, A., 2011. Excavations at Tepe Chalow, Northern Khorasan, Iran. Studi Micenei ed Egeo-Anatolici 53, 236–241. Boroffka, N. et al., 2002. Bronze Age Tin from Central Asia: Preliminary Notes. K. Boyle et al. (eds.), Ancient interactions: east and west in Eurasis. Cambridge, 135–159. Brown, S. C., 1991. Lapis Lazuli and its Sources in Ancient West Asia. Bulletin of the Canadian Society for Mesopotamian Studies 22, 5–13.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

660

Thomalsky

Casanova, M., 2013. Le Lapis-Lazuli dans l’Orient Ancien. Production et Circulation du Néolithique au IIe Millénaire av. J. -C. Documents Préhistoriques 27. Paris. Coats, S., 2005. Minerals in Afghanistan. The Aynak Copper Deposit. Afghanistan Project Manager, British Geological Survey. Kabul. Francfort, H.-P., 1983. Excavations at Shortughaï in Northeast Afghanistan. American Journal of Archaeology 87.4, 518–519. –

1989. Fouilles de Shortughaï: recherches sur l’Asie centrale protohistorique. Mémoires de la Mission archéologique française en Asie centrale t. 2. Paris.

Francfort, H.-P./Pottier, M.-H. 1978. Sondage préliminaire sur l’établissement protohistorique harappéen et post-harappéen de Shortugaï (Afghanistan du N.-E.). Arts asiatiques 34, 29–86. Franke, U. (ed.), 2008. National Museum Herat. Areia antiqua through time – Muzīm Millī Hirāt. Berlin. Franke-Vogt, U., 2006. Alte Kulturen neu entdeckt. Neue Forschungen in Sohr Damb (Pakistan) schließen eine Lücke in der Kulturgeschichte Baluchistans. Antike Welt 37.2, 85–93. Franke-Vogt, U. et al., 2005. Bagh-e Babaur, Kabul: Excavating a Mughal Garden. U. Franke-Vogt/ H. J. Weisshaar (eds.), South Asian Archaeology 2003. Aachen, 533–540. Hauptmann, A./Rehren, T./Schmitt-Strecker, S., 2003. Early Bronze Age copper metallurgy at Shahr-i Sokhta (Iran), reconsidered. T. Stöllner et al. (eds.), Man and Mining – Mensch und Bergbau. Der Anschnitt Beiheft 16. Bochum, 197–213. Helwing, B., 2014. Trade in Metals in Iran and the Neighboring Areas, a Reflection based on Results of the Excavations at Arisman (Iran). M. Casanova/M. Feldman (eds.), Les produits de luxe au Proche-Orient ancien, aux âges du Bronze et du Fer. Travaux de la Maison de l’Archéologie et de l’Ethnologie, René-Ginouvès 19. Paris, 73–84. Herrmann, G., 1968. Lapis Lazuli: The Early Phases of Its Trade. Iraq 30.1, 21–57. Hiebert, F. T., 2002. The Kopet Dag Sequence of Early Villages in Central Asia. Paléorient 28.2, 25–41. –

2009. Tepe Fullol Catalogue. F. Hiebert/P. Cambon (eds.), Afghanistan: Hidden Treasures from the National Museum, Kabul. Washington D.C., 78–80.

Jarrige, J.-F., 2009. The Treasure of Tepe Fullol. F. Hiebert/P. Cambon (eds.), Afghanistan: Hidden Treasures from the National Museum, Kabul. Washington D.C., 67–77. Kruglikova, I. T., 2005. Razvedka archeologičeskich pamjatnikov, provodivšajasja sotrudnikami SovetskoAfganskoj archeologičeskoj ėkspedicii (SAĖ) na severe i severo-zapade Afganistana v 1969–1976 gg. Problemy istorii, filologii, kul’tury IX, 309–437. Levy, T. E./Shalev, S., 1989. Prehistoric Metalworking in the Southern Levant: Archaeometallurgical and Social Perspectives. World Archaeology 20.3, 352–372. Litecka, S. et al. (eds.), 2013. Recent Archaeological Works in Afghanistan. Preliminary Studies on Mes Aynak Excavations and other Field Works (Kabul). Muhly, J. D., 1985. Sources of Tin and the Beginnings of Bronze Metallurgy. American Journal of Archaeology 89, 275–291.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Afghanistan: Ancient Mining and Metallurgy: Initial Project Stage –

661

1995. Mining and Metalwork in Ancient Western Asia. J. M. Sasson (ed.), Civilizations of the Ancient Near East, vol. 1. New York, 1501–1519.

Orris, G. J./Bliss, J. D., 2002. Mines and Mineral Occurrences of Afghanistan. Open-File Report. Parzinger, H./Boroffka, N., 2003. Das Zinn der Bronzezeit in Mittelasien I. Die siedlungsarchäologischen Forschungen im Umfeld der Zinnlagerstätten. Archäologie in Iran und Turan 5. Mainz. Sarianidi, V. I., 1976. Bactria in the Bronze Age. Anthropology & Archeology of Eurasia 15.1. Sarianidi, V. I./Boroffka, N./Dubova, N. A., 2012. Cultural Contacts of Margiana, Turkmenistan, in the 3rd Millennium BC. New Evidence from Gonur Depe, Burial 4150. Gandhāran Studies 6, 1–18. Steuer, H., 2004. Die Ostsee als Kernraum des 10. Jahrhunderts und ihre Peripherien. Siedlungsforschung. Archäologie – Geschichte – Geographie 22, 59–88. Tosi, M. et al., 1992. The Bronze Age in Iran and Afghanistan. History of civilizations of Central Asia. Vol. 1: The dawn of civilization: earliest times to 700 B.C. Paris, 184–216. Tosi, M./Wardak, R., 1972. The Fullol Hoard: A New Find from Bronze-Age Afghanistan. East and West New Series, 22.1–2, 9–18. Vatandoust, A./Parzinger, H./Helwing, B. (eds.), 2011. Early Mining and Metallurgy on the Western Central Iranian Plateau. The first five years of work. Archäologie in Iran und Turan. Mainz. Weisgerber, G., 2004. Prähistorischer und historischer Bergbau in Afghanistan. Teil 2. Der Anschnitt 56.4, 190–211.

Dr. Judith Thomalsky; [email protected], German Archaeological Institute, Eurasia Department, Berlin, Im Dol 2-6, Haus II, D-14195 Berlin.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Oman

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 665–678

Michele Degli Esposti

Excavations at the Early Bronze Age Site «ST1» near Bisya (Sultanate of Oman): Notes on the Architecture and Material Culture The paper offers an overview of the excavations at the site of a 3rd millennium tower near the town of Bisya in central Oman, conducted by the University of Pisa. Its monumental architecture, the associated imposing water management system will be briefly discussed, along with local and imported materials.

Introduction The area around Bisya, in central Oman (fig. 1a), first caught the attention of archaeologists after the inaugural volume of the Journal of Oman Studies included a report on the results of the Harvard Archaeological Survey in Oman (1973), which could best be attributed to «Oman in the 3rd millennium BCE» (Hastings et al. 1975: 9–55). Amongst the sites mentioned for the first time were several Bronze Age sites located in the vicinity of Bisya, south of Bahla, where two major wadis, Wadi Sayfam and Wadi Bahla converge (fig. 1b). The Bronze Age sites included several large circular towers built of massive masonry blocks which are located on low terraces adjacent to the main wadis and sometimes built around a low natural rock outcrop. These sites are complemented by hundreds, perhaps thousands, of stone built burial cairns that cover many of the surrounding hills. In a separate article, Humphries (1974: 49–77) also described two sites tentatively dated to the «Iron Age»; the more impressive of the two (BB-15) was located near Bisyah, at a place named Salut, on a prominent hilltop in the middle of a wide wadi bed. Thirty years after the Harvard team had visited the site, in 2004, the Italian Mission to Oman (IMTO)1 1

Prof. Avanzini, IMTO director, is gratefully acknowledged for appointing the author as Field Director at ST1 and allowing the publication of this paper. A special thank to Carl Phillips, who corrected

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

666

Degli Esposti

Fig. 1. a. The location of Bisya and other main Early Bronze sites mentioned in the text. b. The area of Salut and nearby Early Bronze Age towers (circles).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Excavations at the Early Bronze Age Site «ST1» near Bisyah (Sultanate of Oman)

667

Fig. 2. General view of the site from the East. In the foreground: the outer channel of the main ditch.

started to investigate BB-15, from then on simply referred to as Salut, at the behest of the Office of the Adviser to His Majesty the Sultan for Cultural Affairs. The dense Early Bronze Age settlement of the area around Salut is indicated by the presence of at least three large towers distributed at irregular intervals along the course of the wadi. One of the towers was first identified by the Harvard survey (‹BB-16›: Humphries 1974: 50) and subsequently Beatrice de Cardi located a second tower, the furthest from Salut (‹site 39›: de Cardi et al. 1976: 164). The third, and nearest to Salut, appears first to have been identified by the Birmingham Archaeological Expedition (‹Building 5›: Orchard/Orchard 2006: 158, pl. 6d) who have also speculated on the presence of a fourth tower which would be located somewhere to the northwest of ‹Building 5› (‹Building 6›: ibid.). In 2010, at the request of the Adviser to His Majesty, the IMTO extended the focus of its work to include ‹Building 5›, which was given the site name Salut Tower 1 (ST1). More than 90 towers similar to ST1 are known throughout South East Arabia (A. Mortimer, pers. comm.), some are believed to date from the early 3rd millennium BC, whilst others appear to have been established later, nearer the middle of the millennium. This appears also to be the case with ST1.

an earlier version of the text and to Anne Mortimer, who revised it. The responsibility for any mistakes rests entirely on the author.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

668

Degli Esposti

Fig. 3. General plan of the site (Bronze Age phase).

Architecture The tower comprises a circular stone wall (W1), approximately 22 m in diameter, built from roughly hewn boulders that form the external face (fig. 2). The wall was constructed mainly above ground, with only a relatively shallow foundation trench which appears more substantial on the inside of the tower, due to the fact that the latter was built around a low natural mound that forms the building’s core. After the tower foundations had been established, the trench on the inner side of the wall was backfilled with medium and small size rough stones. The external wall now stands no higher than 2.30 m from the surrounding original ground surface, which corresponds to the top of a thick caliche2 formation underlying the whole plain. It is no longer possible to accurately predict the original height of the tower and, more importantly, impossible to say what was on top of the tower, since no traces of the upper or any intermediary floor levels survived. It seems that the area contained by the outer ring wall was backfilled with predominantly sterile sand and gravel deposits to form a base for whatever existed above. There are not even any internal walls/foundations that might help 2

The term caliche indicates a hardened deposit of calcium carbonate which cements together other materials, including gravel, sand, clay, and silt. It generally occurs in arid and semiarid regions.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Excavations at the Early Bronze Age Site «ST1» near Bisyah (Sultanate of Oman)

669

indicate the layout of any overlying structures. The only remaining feature inside the tower, which again is characteristic of most of the excavated tower sites in Oman, e.g., Maysar (Weisgerber 1981: 199, Abb. 26), Bat (Frifelt 1976: 64, fig. 3)) and the UAE ((Hili 1 (Frifelt 1975: 370, fig. 3), Hili 8 (Cleuziou 1989: 64, pl. 11–12), Tell Abraq (Potts 1997: 66), Kalba (Eddisford/Phillips 2009: 117, fig.7), is a central well. At ST1 an almost circular well was located in the centre of the tower (fig. 3). The shaft has an average diameter of 1.75 m that widens near the surface of the caliche, the level from which the well was presumably excavated prior to the construction of the surrounding tower. The narrower, lined well head would no doubt have been built upwards in tandem with the infilling of the tower. It has not been possible, however, to ascertain the dimensions of the original Bronze Age well due to reuse in the Iron Age, when it was probably re-dug and enlarged. This much later re-use of the well is clearly indicated by dumped Iron Age pottery found deep inside the well, and by the fact that the partially surviving wellhead, with a diameter of approximately 2.3 m, can likewise be dated to the Iron Age (Degli Esposti 2011: 195–196). It can also be noted that where original Bronze Age wellheads have survived they clearly have a much more constricted opening – as seen at several of the sites referred to previously. Another recurrent feature of the towers (e.g. Bidya 2 (al-Tikriti 1989: 108, pl. 78), Hili 8 (Cleuziou 1989: 65, pls. 11–15) and Kalba (Eddisford/Phillips 2009: 116, fig. 6b; 117, fig. 7)) is a large, surrounding ditch. ST1 is no exception as the excavations revealed a 11 to 13 m wide, up to 3.50 m deep ditch. A unique feature, a narrow strip of caliche was left standing in the centre of the ditch, thus dividing it into two, concentric channels. The elevation of this central septum is non uniform, and the two channels were touching at places. A part of the inner side of the ditch, nearest the tower, was lined with a wall made from large boulders (W28), packed by stone rubble laid inside its foundation cut, which corresponds with the inner line of the ditch. At a later stage, a massive terrace (Structure 3) contained by three large boulders walls, blocked the outer channel of the ditch with its northernmost wall crossing it and abutting against W28. This, however, did not block completely the inner channel, and a meter wide passage was left open. Two further structures, similarly employing large boulder masonry, were excavated at ST1. In both cases, only the lower courses of the perimeter walls remain, in both cases they form a roughly oval shape. Structure 1 rests on a wider portion of the central caliche strip of the ditch. It comprises two «rooms» with the smaller one built against the larger. A third small rectangular feature, W47, lies adjacent to the smaller room and above the sloping cut of the ditch. Structure 2 is located on the outer edge of the ditch, slightly south of Structure 1. It comprises a single oval space with clear evidence that broken pieces of caliche were used to create a level floor surface.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

670

Degli Esposti

Structure 2 is also surrounded by a shallow channel, which appears to have been used to divert water in to a deep and large, square pit excavated on its northern side. The pit, which for the moment has only been partially investigated and its bottom not yet reached, is separated from the main ditch by a relatively narrow ridge of in situ caliche. Others wall features were discovered inside the ditch. It is interesting to note that one of these was a sturdy, rectangular buttress, W50, located immediately below Structure 2 and facing W47, mentioned above. It is tempting to interpret these two features as «piers» for some sort of pathway linking Structure 1 and Structure 2, prior to the establishment of Structure 3, which blocks the outer channel and thus provides a connection between Structure 1 and the outermost side of the ditch.

Water Management and Climatic Change Apart from the well and the ditch described above, and the drain and cistern near Structure 2, excavations at ST1 revealed a much more complex network of excavated features. All are part of an impressive system of water management (fig. 3), which shows modifications over time, some of which clearly reflecting the community response to climatic change. Three rather substantial channels (1.50 m wide) have been located so far, which are connected to the main ditch. Their plan is currently only partially known, however it seems of some significance that they all come from the direction where the actual wadi is located. One of these channels has been completely excavated at the junction with the main ditch, where it is 1.50 m deep and an oval sedimentation pit was cut through its bottom. It cannot be stated whether these channels derived water from the wadi bed, or tapped into the wadi fans, or drained some closer underground water; be that as it may, they provide strong evidence for the water storage/management function of the main ditch. Two other deep channels were located approximately 25 m east of ST1. At almost 2 m wide on the surface, they exhibit a roughly V shaped profile with a flattened bottom. The depth varies between around 2 and 2.30 m. The uncovering of these channels is still at a preliminary stage and their overall plan is far from being understood. However, they show different orientations and irregular plans, with one showing a large squared widening south of the other’s end. These channels likely represent only part of a much more extended network, if a comparison is made with Hili 8, though not all the channels there belong to the same phase (Cleuziou 1989: pl. 11–15), and al-Ghubra near Bahla, where magnetic prospection revealed numerous ditches around a partially destroyed, but presumably contemporary site, dated to the 3rd millennium (Orchard/Stanger 1994: 83, fig. 12).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Excavations at the Early Bronze Age Site «ST1» near Bisyah (Sultanate of Oman)

Fig. 4. Wells dug through the ditch’s marl pavement. W28 stands on the right, Structure 3 on the left and Structure 1 in the background.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

671

672

Degli Esposti

The stratigraphy, morphology and composition of the deposits inside the main ditch show that they can be attributed to three Early Bronze Age phases. During the first phase, the ditch was active and filled with backwater. The second phase corresponds to the natural silting up of the ditch, for more than half its depth. This could have happened over a very short period of time during a particularly wet season combined with a lack of maintenance. After the ditch had thus fallen out of use, it became the place for random dumping, with localised contexts containing abundant pottery and occasionally burnt material also discarded. The most important discovery however was that of three wells dug through the ditch’s cemented marl pavement, all ending at the same depth which corresponds to an horizon of not cemented marls – the level of a lower water table than the one exploited during the early phase of occupation at ST1 (fig. 4). The excavation of these wells represents the final attempt of ST1 dwellers to cope with the changing climate, which is to a drier period right towards the close of the 3rd millennium (e.g. Mayewski et al. 2004). Until now there is no evidence for any specific activities having taken place at ST1. That all the large ditches can be associated with water management might be indicative of irrigation and agriculture. Several saddle querns, as well as some grinding/crushing stones, might indicate the processing of plant foods at the site. It is known from contemporary sites elsewhere in South East Arabia that cereals and dates were cultivated (Potts 1994). Unfortunately no direct evidence for a similar range of crops has been found at ST1, despite the flotation of a limited number of burnt deposits retrieved from the site. The only identification so far has been of the fruits and thorns of Zizyphus spina christi. It is possible that the fruits were collected for food. The presence of thorns and other fragments, however, might be more indicative of the tree simply being used as a source of firewood. In addition to this, animal bones have also been retrieved from many contexts and await further study.

Material Culture: Local Productions At ST1, earlier Islamic and Iron Age horizons were often completely mixed together below the recent agricultural soil. Moreover, the evidence indicates an absence of occupational deposits between the Early Bronze Age and the Early Iron Age, when the immediate surroundings of ST1 were clearly reoccupied. Iron Age activities in some cases took place almost directly on top of Early Bronze Age deposits. Fortunately, however, the deep Bronze Age ditches and channels had already been abandoned and filled completely with various deposits before the end of the 3rd millennium BC. Sealed contexts have been excavated inside the main ditch, as well as in the other cut features. The following comments on the Bronze Age pottery and other finds from ST1 will, therefore, be restricted to material coming from these safe contexts.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Excavations at the Early Bronze Age Site «ST1» near Bisyah (Sultanate of Oman)

Fig. 5. Local and imported materials from ST1 (n° 9 drawing S. Martelli).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

673

674

Degli Esposti

Fig. 6. a. Calibrated date for SU55, a context belonging to the last Early Bronze Age phase of the site. b. The stone «Indus» seal discovered at ST1 (photo M. Kenoyer, sketch drawing the author).

The presumably local Omani wares are characterised throughout the whole stratigraphic sequence by a high degree of uniformity, with some slightly differing fabrics evident, notably in the earliest phase. The pottery fabrics also appear to be the only element that can distinguish the assemblage collected at ST1 from that published from other comparable sites. Despite the lack of mineralogical analyses, eight pottery fabrics (A–G) have been distinguished, on the basis of fabric colour and type of inclusions, and in one case also of slip, which was considered distinctive and consistent (see fig. 7). Fabrics A, A2, and F are the most abundant at ST1, whilst the remaining fabrics (B, C, D, E) are much less abundant and in cases rare (G). The abundance of A2 sherds appears at the moment as the element that may distinguish ST1 assemblage from other contemporary sites. Given its abundance fabric A is considered to most likely indicate a local production. In terms of shapes and decoration, the assemblage from ST1 is in total keeping with the general repertoire of the so-called Umm an-Nar period that characterises South East Arabia in the latter half of the 3rd millennium BC. No exclusive associations appear to exist between shapes and fabric. Suspension vessels, with the typical net pattern over the body, below the pierced cordon (fig. 5.5), come for example from almost every context of the sequence and were made using all the main fabric types. A shape that is known elsewhere (e.g. Méry 2000: fig. 92.2) but appears to be more widely represented at ST1 than is usually the case, is that of small beakers. These have a distinct short foot, are mainly undecorated and are found in a variety of fabrics (fig. 5.6).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Excavations at the Early Bronze Age Site «ST1» near Bisyah (Sultanate of Oman)

675

The most widespread decoration comprises paired wavy lines running between two pairs of horizontal lines, just below the rim (fig. 5.1–4). This pattern clearly recalls similar decoration on red-slipped vessels from Umm an-Nar (Frifelt 1991: fig. 96.BA, DF), and Hili (Cleuziou 1989: fig.6.6–8, 15–16 etc.; Cleuziou et al. 2011: fig. 153, top row; etc.), though the best parallels come, significantly, from closer sites in central Oman such as Bat (Frifelt 1975: fig. 30.a, e) and Maysar (Weisgerber 1980: Abbs. 42, 45). Ridged vessels are also attested (fig. 5.7). Compared with pottery, other local materials are poorly represented at ST1, with only a few stone vessels deserving mention (fig. 5.8–10). These all conform to the typical Early Bronze Age «Umm an-Nar» repertoire, as defined by David (1996: 34–38). Metal finds are limited to a couple of short pins/chisels, and two small rings made out of a coiled thin rod, which can be compared with examples, among others, from Tell Abraq (Weeks 1997: 44, fig 18), tomb A at Hili North (Cleuziou et al. 2011: 203, fig. 247), and Ras al-Jinz (Cleuziou/Tosi 2000: fig. 12.8). The mentioned parallels help to date ST1 occupation to the latter half of the 3rd millennium BC, likely after 2400BC. This is confirmed by the first carbon dating obtained so far, for a context linked to the final phase of the main ditch as discussed above. The calibrated range is 2457–2150 BC at 95.4% probability (fig. 6).

Imports and Imitations: ST1 in the Broader Context The pottery and softstone described above are mainly consistent with a local production, with the possible exception of the scarce fine red wares (Fabric C), which could indicate intra-regional exchange. There are, however, some finds that indicate contacts with the coast, and these contacts also acted as a conduit for the acquisition of more exotic items ultimately derived from Iran, the Indus and possibly Central Asia too. Contacts with coastal areas is indicated, for example, by a worked apex of conus shell (fig. 5.24), for which parallels are known from different sites and whose production is well attested in the excavation of Ras al-Jinz (Cleuziou/Tosi 2000: 67). Similarly, a bead made from an engina mendicaria shell was retrieved from a late context at ST1 (fig. 5.23), but parallels suggest it was probably an earlier Bronze Age import. Similar beads have been found in early 3rd millennium cairns on the Jebel Hafit (most recently Cleuziou et al. 2011: fig. 36), and in recent excavations at Bat (Possehl et al. 2009: fig. 49). A unique discovery is that of a bead obtained by piercing a fossil sea urchin (fig. 5.22). The foreign imports discovered at ST1 include numerous fragments of Indus Black Slipped Jars, which are found throughout the entire stratigraphic sequence. This further supports a date for the occupation of ST1 confined to the second half of the 3rd millennium BC.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

676

Degli Esposti

Fig. 7. The fabrics typology defined at ST1.

Other Indus-related pottery typologies were found; the variety in shapes is such that it finds parallel only at other coeval coastal sites in the region. In fact, while usually only one or two shapes are known from sites in the interior (Méry 2000: 236), at ST1 one finds footed cups, cooking pots, small bottles, and one richly decorated jar (fig. 5.11–17). So far, in South East Arabia, small bottles have been found mainly in burials excavated in the UAE (e.g. Méry 2000: 240–243). Fragments of a decorated jar were instead found at Ra’s al-Jinz RJ-2 (Cleuziou/Tosi 2000: fig. 8). This abundance of Indus-related vessels, and in particular the abundance of cooking pots, has led to speculation that Harappans could be dwelling at ST1 (Degli Esposti et al., n.d.). The fact that the fabric of many of the Indus-related fragments found at ST1 appears identical to that of the local pottery could support this hypothesis, or anyhow it could indicate local imitations rather than actual imports. Additional evidence for contacts between Oman and the Indus is provided by a grey softstone square stamp seal, which was found in one of the latest Bronze Age contexts at ST1 (fig. 7). The seal depicts a short horned bull facing a manger, surmounted by a line of Indus script. The stone used to make the seal, however, appears to be of local provenance, while carving details combines elements which could speak of a local imitation with other more likely suggesting that the seal was made by an Indus carver.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Excavations at the Early Bronze Age Site «ST1» near Bisyah (Sultanate of Oman)

677

Two other seals found in Oman provide a good parallel with the seal from ST1: one was found at Ras al-Jinz 2 (Cleuziou/Tosi 2000: 59–60, fig. 17) and the other one curiously came from a tomb not far from ST1 – Tomb 1, excavated several years ago by the Birmingham Archaeological Expedition. The latter appears to have been made from a stone similar to that used for the ST1 seal, but it lacks any inscription. The seal from Ras al-Jinz 2 does bear an inscription but is made of copper, a relatively unusual material for seals in the Indus Valley proper. Evidence for contacts with places other than the Indus is much thinner on the ground. A handful of sherds of unmistakeable incised grey ware, and a single painted gray ware fragment (fig. 5.18–21), indicate interactions with South East Iran, although local imitations are also known (cf. Méry 2000: 217). The provenance of a fragmentary alabaster/calcite vessel, could also be located in South East Iranian area or possibly Central Asia (e.g. Potts 1986). Although further work is necessary at ST1, its importance and potentials are already evident; planned geoarchaeological study of the sediments and archaeometric characterization of the pottery, among others, will hopefully help making of ST1 a reference site for the Early Bronze Age of the whole Oman Peninsula.

Bibliography Cleuziou, S., 1989. The chronology of protohistoric Oman as seen from Hili. P. M. Costa/M. Tosi (eds.), Oman Studies. Papers on the archaeology and history of Oman. Serie Orientale Roma LXIII. Roma, 47–78. Cleuziou, S./Tosi, M., 2000. Ra’s al-Jinz and the prehistoric coastal cultures of the Ja’alān. Journal of Oman Studies 11, 19–73. Cleuziou, S./Mery, S./Vogt, B. (eds.), 2011. Protohistoire de l’oasis d’al-Ain, Travaux de la Mission archéologique française à Abou dhabi (Emirats Arabes Unis). BAR International Series 2227. Oxford. David, H., 1996. Styles and evolution: soft stone vessel during the Bronze Age in the Oman peninsula. Proceedings of the Seminar for Arabian Studies 26, 31–46. De Cardi, B./Collier, St./Doe, B., 1976. Excavations and survey in Oman, 1974–1975. Journal of Oman Studies 2, 101–188. Degli Esposti, M., 2011. The excavation of an Early Bronze Age tower near Salut (Bisyah, Sultanate of Oman): the Iron Age leves. Egitto e Vicino Oriente XXXIV, 189–224. Degli Esposti, M./Kenoyer, M. J./Frenez, D., n.d. Indus Interaction with Interior Oman: New insights from the Early Bronze Age Stone Tower ST1 at Salut. Paper presented at the 22nd conference of the European Association for South Asian Archaeology and Art, Stockholm, 30th of June to 4th of July 2014 Eddisford, D./Phillips, C., 2009. Kalba in the third millennium (Emirate of Sharjah, UAE). Proceedings of the Seminar for Arabian Studies 39, 111–124.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

678

Degli Esposti

Frifelt, K., 1975. On prehistoric settlement and chronology of the Oman Peninsula. East and West 25, 359–424. –

1976. Evidence of a Third Millennium B.C. Town in Oman. Journal of Oman Studies 2, 57–74.

–

1991. The island of Umm an-Nar. Vol.1 – Third millennium graves. Jutland Archaeological Society Publications 26/1. Aarhus.

Hastings A./Humphries, J. H./Meadow, R. H., 1975. Oman in the Third Millennium BCE. Journal of Oman Studies 1, 9–55. Humphries, J. H., 1974. Harvard archaeological survey in Oman: II – Some late prehistoric sites in the Sultanate of Oman. Proceedings of the Seminar for Arabian Studies 4, 49–77. Mayewski, P. A., et al., 2004. Holocene climate variability. Quaternary research 62, 243–255. Méry, S., 2000. Les céramiques d’Oman et l’Asie moyenne. Une archéologie des échanges à l’Âge du Bronze. CRA monographies 23. Paris. Orchard, J. C./Orchard, J. J., 2006. The Third Millennium Bc Oasis Settlements of Oman and the First Evidence of their Irrigation by aflaj from Bahla. J. C. Orchard/J. J. Orchard (eds.), Proceedings of the International Symposium Archaeology of Arabian Peninsula through the Ages, 7th–9th May 2006. Muscat, 143–173. Orchard, J. C./Stanger, G., 1994. Third millennium oasis towns and environmental constraints on the settlement in the al-Hajar region. Iraq 56, 63–100. Possehl, G. L./Thornton, C. P./Cable, C. M., 2009. Bat 2009. A report from the American team. Philadelphia. Potts, D. T., 1986. The Booty of Magan. Oriens Antiquus XXV, 271–285. –

1994. Contributions to the agrarian history of Eastern Arabia II. The Cultivars. Arabian Archaeology and Epigraphy 5, 236–275.

–

1997. Rewriting the Late Prehistory of South-eastern Arabia: a Reply to Jocelyn Orchard. Iraq 59, 63–71.

Reimer, P. J. et al., 2009. IntCal09 and Marine09 radiocarbon age calibration curves, 0–50, 000 years cal BP. Radiocarbon 51, 1111–1150. al-Tikriti, W. Y., 1989. The Excavations at Bidya, Fujairah: the 3rd and 2nd millennia culture. Archaeology in the United Arab Emirates 5, 101–114. Weeks, L. R., 1997. Prehistoric metallurgy at Tell Abraq, U.A.E. Arabian Archaeology and Epigraphy 8, 11–85. Weisgerber, G., 1980. “… und Kupfer in Oman” – Das Oman-Projekt des Deutschen Bergbau-Museum, Der Anschnitt 32, 62–110.

Michele Degli Esposti, PhD, Dipartimento di Civiltà e Forme del Sapere – University of Pisa, and Italian Mission To Oman (IMTO).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 679–687

Stephanie Döpper

Expressions of Collective Memory – The Reuse of EBA Tombs in the Necropolis of Bāt, Sultanate of Oman This article deals with the reuse of two 3rd millennium tombs in the necropolis of Bāt, attested by burials and diverse objects from the Late Iron Age and other periods found within. This phenomenon can be interpreted as an expression of the collective memory of particular social groups.

1. The German Excavations in the Necropolis of Bāt Between 2004 and 2008 a team from the German Mining Museum Bochum, under the direction of Gerd Weisgerber, investigated and reconstructed several tombs from the Hafit period to the Iron Age at the necropolis of Bat (Weisgerber 2006; Weisgerber 2007; Böhme et al. 2008; Böhme 2012; Böhme/Ali Al-Sabri 2011). Among them was the Umm an-Nār (ca. 2700–2000 BC) Tomb 156 at its centre. In 2010, Conrad Schmidt from the University of Tübingen started a new archaeological project at Bāt (Döpper/Schmidt 2011, 2013; Schmidt/ Döpper 2014). His research includes the excavation of the interior of Tomb 156, which was not uncovered previously by the German Mining Museum, and of the Umm an-Nār Tomb 155 close by.

1.1 Tomb 156 Tomb 156 is one of a group of three Umm an-Nār tombs, Tomb 154, Tomb 155 and Tomb 156, in the core area of the necropolis of Bāt (fig. 1). It has an outer diameter of 8.10 m and is divided into two separate, semi-circular chambers, Room A and Room B, by an interior wall. Chamber A measures at its maximum 5.40 m in length and 2.30 in width, with chamber B featuring similar dimensions. The entrances for each tomb consist of triangular shaped openings in the outer wall. At the beginning of the German Mining Museum Bochum work, the outer facade of the tomb was preserved to a height of two metres in the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

680

Döpper

Fig. 1. Tomb 154, Tomb 155 and Tomb 156 in the core area of the necropolis of Bāt (photo: GGH Solutions in Geosciences).

Fig. 2. Tomb 156 before restoration (photo: German Mining Museum Bochum).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Expressions of Collective Memory – The Reuse of EBA Tombs in the Necropolis of Bāt

681

Fig. 3. Samad period burial on top of the outer wall of Tomb 156 (photo: German Mining Museum Bochum).

east, but only to a height of a few rows of stones in the other parts (fig. 2; Böhme et al. 2008: 68). The debris cone around the tomb yielded some finds, especially pottery and beads from the Umm an-Nār period, which were concentrated in the area around the two entrances (Böhme et al. 2008: 68). In front of the eastern entrance a small pit filled with burned animal bones of sheep or goat were discovered, along with Iron Age pottery sherds. After removing the debris cone around the tomb, the team from the German Mining Museum Bochum reconstructed its outer facade, in order to prevent it from further decay. In a collapsed part of the tomb’s exterior wall, a badly preserved skeleton was found and labelled Tomb 156 W (fig. 3). According to the excavators, it belongs to a male individual. This burial was associated with metal plates and at least three iron arrow heads, dating from the Late Iron Age/Samad period (ca. 300 BC–AD 300) (Böhme et al. 2008: 68). The photographs of the skeleton suggest that it was lying in a flexed position. In 2010 and 2011, the team from the University of Tübingen investigated the interior of the tomb that was unaffected by the previous restoration works of the German Mining Museum Bochum. The fill consisted of windblown earth and stone debris from the surrounding walls. No distinct stratigraphic layers could be determined. The fill contained

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

682

Döpper

Fig. 4. Achaemenid Cream Bowl BAT11A-i0744 from Tomb 156.

Fig. 5. Iron arrow head BAT11A-i0286 from Tomb 156.

very fragmented and badly preserved human bones, along with many small Umm an-Nar period beads found above the floor in Room A. Objects from later periods include: decorated chlorite vessel fragments from the Wadi Suq (2000–1300 BC) and Iron Age (1300–300 BC) periods (Döpper/Schmidt 2014), coarse mineral tempered Iron Age pottery, as well as a complete profile of a so-called Achaemenid Cream Bowl from present day Iran or a

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Expressions of Collective Memory – The Reuse of EBA Tombs in the Necropolis of Bāt

683

Fig. 6. Handmade pottery jar BAT12A-i0674 from Tomb 155.

local imitation of it (fig. 4; BAT11A-i0744). Comparable bowls are known on the Omani Peninsula, for example from the Iron Age III levels of Rumeilah in the United Arab Emirates (Tosi/Cleuziou 2007: 297). In addition, metal plate fragments with rivets and several iron arrow heads (fig. 5), whose original shape cannot be clearly determined any more due to severe corrosion, but which are generally comparable to those found together with the burial of Tomb 156 W, also originate from the fill of Tomb 156. A large, incised stone bead (BAT12A-i0633) that closely parallels an Iron Age pendant from Tell Abraq in the United Arab Emirates (Döpper/Schmidt 2013: 28 Abb. 4), was a special find from this deposit. Other smaller and mostly undecorated beads were discovered as well. As all finds were found loose in the fill, they cannot be connected clearly with the burial Tomb 156 W, but many of them seem to be of a comparable time range.

1.2 Tomb 155 Tomb 155 lies in the direct vicinity of Tomb 156 and was excavated between 2012 and 2013 by the University of Tübingen (fig. 1). This tomb has an outer diameter of 6.5 m. Its inner wall, unlike Tomb 156, does not separate the tomb into two chambers, but instead runs only part the length of the tomb, leaving the inside space by the entrance free. Along the outer wall, which is preserved to a maximum height of 0.72 m above floor level, some so-called sugar lumps of the original white façade of the tomb are preserved. Tomb 155 is paved with large, flat undressed stones. Most of them are local limestone slabs, while some white ‹sugar lumps› were occasionally integrated into the floor as spoil. Many finds were discovered in the fill, especially on the stone floor. Among them are 532 beads including a 7.6 cm long biconical carnelian bead from the Indus region (Frenez et al. 2014). Additionally, Umm

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

684

Döpper

Fig. 7. Samad period burial interred in the floor of Tomb 155 with associated grave goods.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Expressions of Collective Memory – The Reuse of EBA Tombs in the Necropolis of Bāt

685

an-Nār and Iron Age chlorite vessels, black on red Umm an-Nār pottery, some Umm an-Nār jars with very distinct carinations, and a complete handmade pottery jar with an incised wavy line decoration, were found (fig. 6; BAT12A-i0674). The chronological setting of the latter is not certain as no comparisons are known, but it seems to be of a much more recent date than the Umm an-Nār period. Along the inner dividing wall in the south-eastern part of Tomb 155, some stone floor slabs were missing. They leave a gap of 1.60 x 1.20 m, in which a skeleton in flexed position was discovered (fig. 7). It was determined by the anthropologist Steve Zäuner that this individual could be 16 to 25 years in age, and although the bones are in a very bad condition, there are some hints that it might be a woman. The skeleton is associated with several grave goods, which were distributed around it. These are two iron arrow heads (BAT13A-i0437 and BAT13A-i0436) found close to the feed of the skeleton, two more iron arrow heads (BAT13A-i0549 and BAT13A-i0575), which come from a spot at its lower back, and yet another one (BAT13A-i0561), that was lying in the area of the forearms. A metal plate with rivets (BAT13A-i0566) was discovered on the shoulder of the skeleton. All arrow heads date the burial to the Late Iron Age/Samad period. This, together with the orientation of the skeleton and the associated grave goods, makes it very akin to the Late Iron Age burial Tomb 156 W. The other Late Iron Age finds from the fill of Tomb 155 reminds one of the scattered Late Iron Age objects from the fill of Tomb 156.

2. Reuse of Tombs as Expressions of Collective Memory The material evidence makes it clear that the reuse of Tomb 155 and Tomb 156 was intentional and that both tombs again functioned as burial places in the Late Iron Age. Other Umm an-Nār tombs in the necropolis of Bāt, the Hafit tombs at Al-Ayn (Döpper 2014; Döpper 2015), and tombs at other places on the Omani Peninsula (Tosi/Cleuziou 2007: 109), show similar signs of reuse, although complete burials are not present in every case. Hence, reuse of Bronze Age tombs is no rare phenomenon in this area and has been known for quite a while. Nevertheless, it has not drawn much scholarly attention and when discovered, tomb reuse has often been dismissed as a singular case. So far, there has been no interest in the question of why Iron Age people reused old and already partly decayed tombs as a place to bury their loved ones. One answer may lie in actions connected with collective memory. Maurice Halbwachs, a French sociologist, explained in the 1920s that all memories are influenced by social processes and that not only individuals have individual memories, but also that social groups have collective memories (Halbwachs 1967, 1966). For him, collective memory always interacts with spaces (Halbwachs 1967: 142). Therefore, social groups that occupy a space, are at the same time influenced by it (Halbwachs 1967: 129). A similar

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

686

Döpper

argument comes from the German Egyptologist Jan Assmann. According to him, aspects of collective memory can be reflected in material culture, not just in pictures, buildings, monuments, and settlements, but also in rituals and landscapes (Assmann 1988: 11). Therefore, places that evoke special memories in people can become sites of collective memory for society (Wickholm 2008: 89). This is what might have happened to the Early Bronze Age tombs on the Omani Peninsula in the Late Iron Age. In this theory, they became a place where people generated and maintained the collective memory of their group through specific actions by their members. As collective memories are a key aspect in shaping group identities, the reused tombs in the necropolis of Bāt became the location where group identity was stored and kept for the future (Wickholm 2008: 95). Similar practices can be found, for example, at Finnish cremation cemeteries (Wickholm 2008: 95) and Swedish prehistoric graves (Thäte 2007: 82) demonstrating that cemeteries can become sites for group identity across multiple cultures. Halbwachs explains further that continuity is a key aspect in the practice of actions connected to collective memory (Halbwachs 1967: 157). Emphasising continuity would maintain, legitimate and stress group identity over time. Thus, the necropolis of Bat, that was perceived by ancient people as representing continuity from the Early Bronze Age to the Late Bronze Age, can be interpreted as places were present and past are bound together in a continuum.

Bibliography Assmann, J., 1988. Kollektives Gedächtnis und kulturelle Identität. J. Assmann/T. Hölscher (Hrsg.), Kultur und Gedächtnis, Frankfurt am Main, 9–19. Böhme, M., 2012. The restoration of Tomb 154 in Bat. D. T. Potts/P. Hellyer (eds.), Fifty years of Emirates archaeology. Proceedings of the Second International Conference on the Archaeology of the United Arab Emirates, Abu Dhabi, 112–119. Böhme, M./Ali Al-Sabri, B., 2011. Umm an-Nar burial 401 at Bat, Oman: architecture and finds. Arabian Archaeology and Epigraphy 22, 113–154. Böhme, M./Heckes, J./Weisgerber, G., 2008. BAT Research & Restoration Project/Sultanate of Oman. Ein Tätigkeitsbericht für 2006 bis 2008. Bochum. Döpper, S., 2014. On the Reuse of Early Bronze Age Tombs – the German Excavations at Bāt and Al-Ayn, Sultanate of Oman. P. Bieliński et al. (eds.), Proceedings of the 8th International Congress on the Archaeology of the Ancient Near East. 30 April–4 May 2012, University of Warsaw. Wiesbaden, 57–69. –

2015. The reuse of tombs in the necropolis of Bāt, Sultanate of Oman. Proceedings of the Seminar for Arabian Studies 45, 83–92.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Expressions of Collective Memory – The Reuse of EBA Tombs in the Necropolis of Bāt

687

Döpper, S./Schmidt, C., 2011. Die Grabtürme der Nekropolen von Bāt und Al-Ayn im Sultanat Oman. Bericht über die Vorkampagne 2010. Mitteilungen der deuschen Orient-Gesellschaft 143, 293–321. –

2013. Bericht über die Ausgrabungen 2011 und 2012 in Bāt und Al-Ayn, Sultanat Oman. Mitteilungen der deuschen Orient-Gesellschaft 145, 23–52.

–

2014. Chlorite Vessels from Tomb 155 and Tomb 156 in Bāt, Sultanate Oman. B. Cerasetti et al. (eds.), My life is like the summer rose. Maurizio Tosi e l’archeologia come modo di vita. BAR, Oxford, 187–192.

Frenez, D./Kenoyer, J. M./degli Esposti, M./Thornton, C./Cable, C./Schmidt, C., 2014. Indus Style Artifacts from Interior Oman: Long Distance Trade or Local Production? Paper presented at the 48th Seminar for Arabian Studies, London, 25–27 July 2014. Halbwachs, M., 1966. Das Gedächtnis und seine sozialen Bedingungen. Soziologische Texte 34, Berlin. –

1967. Das kollektive Gedächtnis, Stuttgart.

Schmidt, C./Döpper, S., 2014. German Expedition to Bāt and Al-Ayn, Sultanate of Oman: The 2010 to 2013 seasons. Journal of Oman Studies 18, 187–230. Thäte, E. S., 2007. Monuments and Minds: Monument Re-use in Scandinavia in the Second Half of the First Millennium AD. Acta archaeologica Lundensia Series 40, Lund. Tosi, M./Cleuziou, S., 2007. In the shadow of the ancestors. The prehistoric foundations of the early Arabian civilization in Oman. Muskat. Weisgerber, G., 2006. Dokumentations-, Restaurierungs- und Forschungsprojekt in Bat (Oman) 2006. –

2007. B A T World Heritage Monument. Springseason 2007 Report.

Wickholm, A., 2008. Reuse in Finnish Cremation Cemeteries under Level Ground – Examples of Collective Memory. F. Fahlander/T. Oestigaard (eds.), The Materiality of Death. Bodies, burials, beliefs. BAR International Series 1768. Oxford.

Stephanie Döpper, University of Tübingen, Institute for Ancient Near Eastern Studies (IANES).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 689–696

Conrad Schmidt

Mobile Pastoralists as Global Players: Excavations at Al-Zebah, Sultanate of Oman This paper presents the results of three seasons of recent excavation carried out at AlZebah, Sultanate of Oman. The site dates to the Umm an-Nar period and is interpreted as a campground for pastoral nomads. After several seasons excavating at Bāt, which is thought to be the largest 3rd millennium BC necropolis in Oman, work has almost exclusively focused on tomb and burial contexts (Döpper/Schmidt 2011, 2013, i. p. a–c; Schmidt i. p.; Schmidt/Döpper i. p.; Thornton/ Schmidt i. p.). The question of where the hundreds of people who were buried here lived has therefore become more and more relevant. Luckily, the late Gerd Weisgerber drew my attention to a small but very well preserved settlement called Al-Zebah in the vicinity of Bāt. In 2012 excavations began at Al-Zebah, which is completely different from what we know from sites like Bāt, Bisya or Hili, as both tombs and towers are absent. This paper presents the results of research conducted between 2012 and 2014. The site of Al-Zebah lies seven kilometres to the north-west of the necropolis of Bāt. 14C analyses of charcoal date the settlement to 2200–2150 BC. Al-Zebah is situated in a plain surrounded on three sides by hills of up to 200 metres in height. Directly to the north is the Wadi Shwoi’ai, which runs straight in the direction of Bāt and carries water after heavy rains. Al-Zebah is characterised by large rectangular enclosures, scattered across an area of approximately 150 by 200 metres. Three house-complexes have been studied in detail, one in the north, one in the east and one in the west (fig. 1). At the northern house-complex, House III, 9 rooms were excavated overall. Room G1 has a size of 8.90 by 9.30 m and features an interior wall, which runs to the centre of the room (fig. 6). The partly blocked door of this room is situated in the south-eastern corner. Between Room G and Courtyard V lies 1

For plans of the excavated houses with rooms labelled and other details see Döpper/Schmidt i. p. a: Beilage 2 and 3.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

690

Schmidt

Fig. 1. Referenced orthophoto of Al-Zebah (GGH Solutions in Geosciences, Freiburg).

Room U, which is connected to Room G by two ramps leading up and down the wall. The Courtyard V is 8.50 m long and features a multitude of simple hearths made up of palmsized stones and heaps of ash. Adjacent to the courtyard are two small rooms, one in the north and one in the south. Both have wide openings into the courtyard. Additionally, Room L, which measures only 3.20 by 1.80 m, contains a large hearth made of neatly placed stones. Outside the same room lower parts of three pottery vessels have been found in situ (fig. 2). In general, almost all of the excavated walls at Al-Zebah are double-faced and made of undressed stones of white limestone. They rest on natural gravel, which simultaneously functions as the rooms’ floor. The walls are on average 60 centimetres wide and often consist of two layers of stone. They are preserved to about 45 to 50 centimetres in height. In most cases, their upper edges reach only few centimetres above modern ground level.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Mobile Pastoralists as Global Players: Excavations at Al-Zebah, Sultanate of Oman

Fig. 2. House III: Room L with large hearth inside and three pottery vessels outside the room.

Fig. 3. House III: semi-circular retaining wall at the north-western corner of Room I.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

691

692

Schmidt

All the walls were presumably never built any higher, at least not the part constructed of limestone. In the northern part of House III, a doorway connects Room O with Room I to the west of it. This room is interesting because its layout was changed after the Wadi washed away its northern part (figs. 1, 6). Presumably, this room was originally square in shape, before its northern wall was rebuilt along the Wadi-line and the room became triangular. This new northern wall was further strengthened with a stone glacis and with two semi-circular retaining walls at the north-western corner of the room (fig. 3). At the same time, a new, small room was created in the corner. It only measures 3.40 by 2.30 m and has a lot in common with Room L, for instance several pottery vessels were outside the room along with some simple fireplaces, both inside and outside, indicating storage and food processing. To the west of this room lies Courtyard BA, which has not been excavated but which features a heavily blocked doorway. To the same wall an extra row of stones is attached, a common feature at Al-Zebah. Finally, there is Room EA, which measures 6.60 by 4.20 m and has an interior wall similar to that of Room G. The other excavated house-complex, House VI, lies in the east of the settlement (fig. 1). It also contains a large, square room, Room AA, but in this case the room has been separated into several smaller rooms in a later phase, which most likely also falls within the Umm an-Nār period (fig. 4). To these smaller rooms in the west of the house a circular structure in the centre of Room AA can be connected, which has not yet been excavated. Such circular structures are very typical for Al-Zebah, for instance at House IV in the west part of the site. It is believed that they indicate a functional change within the settlement. House VI was further modified by an annex in the north-east, Room GA. The walls of this room curve slightly and are constructed with upright stones that are different to any other excavated wall at Al-Zebah. Fireplaces with an outer circle of red burned clay and filled with ashes and small stones have been found here. Furthermore, stone collapse, containing many pottery sherds and a grinding stone, superimposes a large hearth, which is similar to the hearth of Room L in House III. Another parallel to House III is the extra row of stones in front of the northern wall of Room AA (fig. 5), which shows that the original layout of House VI is comparable to House III. Compared to, for instance, the large sites of Hili in the United Arab Emirates and Qumayrah, Bāt, and Bisyah in the Sultanate of Oman, the structure of Al-Zebah is completely different. Instead of tombs and towers, Al-Zebah has exclusively domestic architecture, which is almost absent at the so-called centres of the same period. Also, when looking at the pottery of Al-Zebah (Döpper/Schmidt 2013: Abb. 22; Döpper/Schmidt i. p. a: Abb. 23; Schmidt/Döpper i. p.: fig. 17), there is little correspondence to the pottery from the burial pits in Bāt, such as in the fine red-on-black Umm an-Nār ware and lugged jars. Better parallels can be found at the settlements of Umm an-Nār island, Tell Abraq, and Maysar. This

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Mobile Pastoralists as Global Players: Excavations at Al-Zebah, Sultanate of Oman

Fig. 4. House VI: overview.

Fig. 5. House VI: northern wall with extra row of stones.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

693

694

Schmidt

Fig. 6. Overview of the core area of Al-Zebah with House III in the foreground.

shows that architecture, as well as movable goods, militate in favour of different functions of Al-Zebah and the large centres of the Umm an-Nār period. Al-Zebah’s function is clearly domestic. Activity zones for storage and food processing have been identified. The site’s economy was most likely predominantly based on animal husbandry, although this is not yet attested. The structure of the walls and the lack of clay in the house fill, which might be attributed to mudbrick debris, lead to the hypothesis that the rooms were covered with tents. Ethnographic parallels, for example the Bakhtiyari and the Luri pastoralists in Iran (Cribb 1991; Hole 2004; Potts 2014), show that stone walls similar to those of Al-Zebah can serve as substructures for black tents. For the large rooms of AlZebah, tents are the only possible way for roofing these buildings since there is no adequate wood. Additionally, the wide openings of some of the small rooms are a sign of tents rather than of fully built houses with solid doors. Such tents can serve as a shelter for both man and domesticated animals. In summary, the many fireplaces, hearths, storage vessels, and the layout of houses at AlZebah, as well as the interpretation of the stone walls as substructures for tents, speak for its seasonal use by pastoral nomads. Additionally, the loose arrangement of the houses (fig. 6) is typical for mobile pastoralists with seasonally alternating camps. The site might have been occupied only during certain times of the year, when the area was more favourable than

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Mobile Pastoralists as Global Players: Excavations at Al-Zebah, Sultanate of Oman

695

other places in Northern Oman. This may have been for climatic reasons, the availability of wild animals, or perhaps because of the growth period of certain plants. The location of AlZebah at the foothills of the Al-Hajar Mountains seems ideal for vertical movement between the uplands and lowlands. Finally, the question of whether nomads played an important part in supplying Mesopotamia with copper in the 3rd millennium BC must remain open for the moment. If the different functions of Al-Zebah on the one hand and the centres like Hili, Bisya, and Bāt on the other is correct, then it is coherent to say that the latter were used by the nomads for burials and as meeting or market places, while their campgrounds were elsewhere. The verification of this hypothesis is one of the major tasks for the future. Al-Zebah is surely not an unique case; other comparable sites will be found and investigated. Another aspect that the excavations at Al-Zebah have provided evidence of, is that the layout of the houses, installations and finds within the settlement are very homogeneous. This shows the existence of strong standards within society. Other material culture from the 3rd millennium BC in the Omani Peninsula reflects this state of affairs; it is extremely similar, which is very unusual for such a large region. There is no evidence for centralised political powers, administrative and religious buildings, or a settlement hierarchy on the Omani Peninsula in the 3rd millennium BC, in contrast to Mesopotamia at this time. Therefore, pastoral nomads, like those attested at Al-Zebah, might be one reason for the homogeneity of material culture in the Umm an-Nār period.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

696

Schmidt

Bibliography Cribb, R., 1991. Nomads in Archaeology, New Studies in Archaeology. Cambridge. Döpper, S./Schmidt, C., 2011. Die Grabtürme der Nekropolen von Bāt und Al-Ayn im Sultanat Oman: Bericht über die Vorkampagne 2010, Mitteilungen der Deutschen Orient-Gesellschaft 143, 293– 321. –

2013. Bericht über die Ausgrabungen 2011 und 2012 in Bāt und Al-Ayn, Sultanat Oman. Mitteilungen der Deutschen Orient-Gesellschaft 145, 23–52.

–

i. p. a. Bericht über die Ausgrabungen 2013 und 2014 in Bāt und Al-Ayn, Sultanat Oman. Mitteilungen der Deutschen Orient-Gesellschaft 146.

–

i. p. b. The Umm an-Nār burial pits of the necropolis of Bāt. S. Cleuziou/M. Tosi (eds.), In the Shadow of the Ancestors: The Prehistoric Foundation of the Early Arabian Civilization in Oman (2nd edition).

–

i. p. c. Chlorite Vessels from Tomb 155 and Tomb 156 in Bāt, Sultanate Oman. B. Cerasetti et al. (eds.), My Life is like the Summer Rose. Maurizio Tosi e l’archeologia come modo di vita, British Archaeological Reports. Oxford.

Hole, F., 2004. Campsites of the Seasonally Mobile in Western Iran. K. von Folsach et al. (eds.), From Handaxe to Khan: Essays presented to Peder Mortensen on the occasion of his 70th birthday, 67–85. Potts, D. T., 2014. Nomadism in Iran: From Antiquity to Modern Era. New York. Schmidt, C., i. p. Aus Alt mach Neu – Bestattungssitten in Bat, Sultanat Oman. R. Slotta (Hg.), Gedenkschrift für Gerd Weisgerber (Deutsches Bergbau-Museum Bochum). Schmidt, C./Döpper, S., i. p. German Expedition to Bāt and Al-Ayn, Sultanate of Oman: The 2010 to 2013 seasons, The Journal of Oman Studies 18. Thornton, C. P./Schmidt, C., i. p. The UNESCO World Heritage Site of Bat and Al-Ayn: Past, Present, and Future. H. M. A. al-Lawati (ed.), Proceedings of the International Symposium “The Archaeological Heritage of Oman. An Early Cradle of Arabian Civilization brought to the Forefront of Cultural Resource Management”. UNESCO headquarters, Paris, September 7, 2012.

Conrad Schmidt, University of Tübingen.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Azerbaijan – Georgia – Caucasus-Region

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9th ICAANE, Basel 2014, Vol. 3, 699–708

Jeyhun Eminli – Emil Iskenderov

Archaeological Investigations at Piboz Tapa Necropolis. Lerik, Southern Azerbaijan (Preliminary Report 2012–2013) Although the southern region of Azerbaijan is rich in antique monuments dating to the Achaemenid and Parthian periods, there have been few investigations in the area. For the purpose of studying antique monuments in the region, the Institute of Archaeology and Ethnography of the National Academy of Sciences of Azerbaijan founded the Lerik Antique Expedition in 2011–2012. The main objective of the expedition is to make a field survey and recheck sites registered in the Soviet period. In addition, we aim to study the Achaemenid, Parthian and Atropatenean influences on the region, and trace the impact on Caucasian Albania through the Mugan steppe. During investigations at Lerik, we investigated the sites of Piboz Tapa, Halasa Yeri and Boykendil. This paper presents the preliminary results of Piboz Tapa necropolis excavated 2012–2013.

Introduction Piboz Tapa is located in Gurdasar village of the Lerik region in southern Azerbaijan. The Lerik region is divided into 5 zones, with the village Gurdasar falling in the Dirikh zone. This area is comprised of low mountains and hills. The site is situated on the AliabadGurdasar road, at a distance of 18 km from Lerik centre (fig. 1). The existence of the site was first recorded during the 1980’s in the course of construction of a road. At that time, the north-east part of the hill was destroyed. Later on, the hill was used in agricultural works, and the soil of the hill was taken away to be used as a plaster material for local houses. Upon the discovery of human remains, the local people stopped digging at the site. The name Piboz in the local Talyish language means «the old area» or «the area out of use». In 2012, a wide-scale archaeological investigation was carried out to study this site. Investigations were conducted on the northeastern slope of the hill. The following year, along with the northeastern slope the highest parts were included in the investigation, in order

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

700

Eminli – Iskenderov

Fig. 1. Panorama of Gurdasar village and view of Piboz Tapa.

to study the areal borders and depth of the necropolis. During the course of the 2012–2013 excavation campaign, 72 graves were explored. Most burials belong to the first stage of the necropolis dating to the Achaemenid through Parthian periods. Later on, the site was occupied by both Christian and Muslim burials. The early stage of the necropolis was marked by burials in flexed and tightly flexed positions, lying on the left and right sides. Some of these graves were destroyed accidentally by the later period graves and during road construction.

Interpretation of Burials The inhumation burials contain both adults and infants. All the graves are dug into the ground at a depth between 50 and 120 cm. Most burials were placed in flexed position on the left or right side. The bodies lying on the right side are oriented southwest to northeast, with the head pointing to the southeast and the face to east. The bodies lying on the left side are oriented southwest to northeast with the head to the southwest and face to north or northeast. The hands are oriented towards the face. In most cases, the burials were covered by ash and coal. Sometimes ash and coal covered inside the graves, the hands and legs, and in front of face and arms. In general, small

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Archaeological Investigations at Piboz Tapa Necropolis. Lerik, Southern Azerbaijan

701

Fig. 2. Spear heads in grave no.36.

ceramic vessels, such as bowls, phials, and one-handle jars, were placed above or behind the skull. Large vessels, such as one or two handle jars and pitchers, were placed near the feet. The bowls and phials were placed together with iron spear-heads above or near the skull. In some graves the spear-heads number up to 5 (fig. 2). In addition to the spear-heads, iron knifes and swords are also attested. All the burials are rich in bronze bracelets, bronze and silver finger-rings and earrings, as well as glass, stone, paste and organic beads. Dog and deer-form bronze pendants were also discovered in these burials. Most graves were single inhumations. Only in two cases were pair burials attested. For example, in grave No. 41 a pair burial was discovered, provisionally mother and infant (fig. 3). The bodies are lying on the right side and are oriented southwest to northeast. The faces are oriented to the east. The infant was positioned close to the mother. Two bowls were placed above the skull of the adult. There was a sickle shape iron knife inside the bowl, and an iron spear-head near the bowl. Large vessels in damaged condition were identified near the feet. The burials were rich in bronze and silver adornments and glass beads. A layer of ash and coal covered the feet and body of the infant. The only difference among the flexed burials is whether the skeleton is lying on its left or right side. There are many typical features in these burials such as the basic position of the body, rituals and burial assemblages. Burial No. 24 is an example of this (fig. 4). Here we

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

702

Eminli – Iskenderov

Fig. 3. Pair burial No.41.

Fig. 4. Burial No.24 Flexed position on the right side.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Archaeological Investigations at Piboz Tapa Necropolis. Lerik, Southern Azerbaijan

Fig. 5. Headdress remains. Burial no. 24.

Fig. 6. Burial No. 22. Tightly flexed position on the left side.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

703

704

Eminli – Iskenderov

Fig. 7. Dismembered bural no. 63.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Archaeological Investigations at Piboz Tapa Necropolis. Lerik, Southern Azerbaijan

705

find the same skeleton position, an ash layer over the body and personal belongings. Some textile remains and headdress were also found here (fig. 5). The same features are typical for both adult and infant burials. During excavations, several tightly flexed inhumation burials were excavated. This form of burials is a minority, and has not yet been discovered in a wide area. The position of these burials is both on left and right side. Although the bodies are in a tightly flexed position, the sepulchres themselves are 3 meters or more in length. The grave belongings are in the same position as above, with small vessels above the skull and large vessels near the feet (fig. 6). In contrast to the other burials, more vessels and animal bones were found here. Additionally, a jar was placed in the arms close to the body. In all these cases, clay or bone spindles were placed in the jars. A less-attested type of grave contains dismembered burials. An ash coat traces the inside these graves, while vessels, adornments and iron spear-heads are the same as in the flexed burials. In one case, and upper jaw was turned up and placed in the middle of the sequestrum (fig. 7). 15 damaged burials were recorded. These burials were generally disturbed during the digging of later graves, as demonstrated by their proximity to later-stage burials. During internment of the later-stage bodies, the earlier remains were accidentally damaged, and the skeletons and grave belongings were collected and reburied compactly. It was therefore impossible to define the position of the burials.

Pottery Assemblages in the Burials Pottery comprised the largest class of burial belongings, in comparison with other artefact types. Most common were reddish one- or two-handle large vessels coated with yellow slip. Reddish-brown one handle small jars are typical for all burials (fig. 8a–b). The form of these jars is a round body and narrow neck. Sometimes, they were coated with brown slip, applied asymmetrically on the body of the jars. The small jars were placed mostly behind or above the skull. Near the small jars were also placed bowls and phials. The bowls are of high quality, and for the most part imitations of toreutics (fig. 8c). The indent at the bottom of the bowls is an omphalos. Pottery with an omphalos includes phials as well. Together with high quality phiales, there were also many crude pieces. The crude phials differ in their coarse-grained matrix. In some cases, ancient restorations were noted on the well-formed phials (fig. 8d).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

706

Eminli – Iskenderov

Fig. 8. Pottery samples from the burials.

Zoomorphic rhyta in the form of a horse, a bird and a wild boar were also noted. The horse-shape rhyton was highly burnished and black polished. A harness was moulded on the rhyton (fig. 9a). Another zoomorphic rhyton is in the shape of a wild boar (fig. 9b). The mouth, eyes and horn or fangs are formed in brown and the vessel is polished. All the vessels were made without potter’s wheel.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Archaeological Investigations at Piboz Tapa Necropolis. Lerik, Southern Azerbaijan

Fig. 9. Zoomorphic rhyta.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

707

708

Eminli – Iskenderov

Conclusion In conclusion, 72 burials were investigated, grouped below in the following forms: Early-stage burials • 17 burials in a flexed position on the right side • 11 burials in a flexed position on the left side • 2 burials in a tightly flexed position on the right side • 3 burials in a tightly flexed position on the left side • 4 burials that had been dismembered • 15 disturbed burials Later-stage burials (Medieval) • 9 burials on the right side • 3 burials on the right side covered with stones • 8 burials on the left side The scene demonstrates that Piboz Tapa was a necropolis from the 5th century BC until the Medieval period. Chronologically, the first stage of the necropolis belongs to the Achaemenid and Parthian periods. Although there were different types of burials in this period, there were characteristic features for the all burials described above. In the early stage, both adults and infants were buried across the entire area. But in the later stage, only infants were buried at the bottom of the hill. During the 2013 excavation campaign we also defined a settlement at the bottom of the necropolis. We hope that future investigation will allow to study necropolis and settlement as a single complex.

Jeyhun Eminli, Leading researcher, Institute of Archaeology and Ethnography, ANAS. Head of Lerik Antique Expedition. Emil Iskenderov, Senior scientific researcher, Institute of Archaeology and Ethnography, ANAS.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 709–722

Seiji Kadowaki – Farhad Guliyev – Yoshihiro Nishiaki

Chipped Stone Technology of the Earliest Agricultural Village in the Southern Caucasus: Hacı Elamxanlı Tepe (the Beginning of the 6th Millennium BC) Hacı Elamxanlı Tepe provides new evidence for the Neolithization in the southern Caucasus as it represents one of the oldest agricultural villages in this region, corresponding to the initial phase of the Shomutepe-Shulaveri culture. This paper reports chipped-stone technology of this site and discusses its possible indigenous and foreign links.

Neolithic Agricultural Villages in the Southern Caucasus The southern Caucasus is located at a northern fringe of the Fertile Crescent, where early management of cereal plants and major livestock species started more than 10,000 years ago. Such a geographic position raises the question of whether the domestication of indigenous plants and animals took place also in the southern Caucasus, or the domestication process in this region occurred secondarily under the influences from the advanced centers to the south. However, our current knowledge to answer this question is limited due to the scarcity of data on chronology, material cultures, settlement patterns, and subsistence practices for the early part of the Neolithic. In the southern Caucasus, currently known evidence for the oldest agricultural settlements comes from Pottery Neolithic sites grouped under a chrono-cultural unit, called the «Shomutepe-Shulaveri» or «Shulaveri-Shomu» culture, which are dated to the 6th millennium cal. BC. These early agricultural sites are formed in anthropogenic mounds (tepe) mainly distributed in the middle course of the Kura and Araxes Rivers (fig. 1). Such tepe sites resulted from repeated constructions of mud-walled buildings for residence and various domestic activities including tool production, food processing, and cereal/tool storage (Kadowaki et al. 2015). Botanical and faunal remains from these sites are dominated by domestic species, such as wheat, barley, legumes, sheep, goats, and cattle (Badalyan et al. 2010; Lyonnet et

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

710

Kadowaki – Guliyev – Nishiaki

Fig. 1. Map of the southern Caucasus, showing the location of Neolithic sites affiliated with the Shomutepe-Shulaveri culture. (1) Hacı Elamxanlı Tepe and Göytepe, (2) Guseingulutepesi, (3) Shomutepe, Gargalartepesi, Toiretepe, (4) Imiris Gora, Shulaveri Gora, Khramis Didi Gora, (5) Aruchlo I, (6) Aratashen, (7) Aknashen-Khatunarkh.

al. 2012). Lithic assemblages include abundant sickle elements as well as a large size and quantity of milling tools, indicating frequent practices of cereal harvesting and processing (Lyonnet et al. 2012). Bone tool industry is well developed with a variety of types (e.g., awls, spatulas, and hammers) made of domestic animals (Narimanov 1992). Pottery is also a unique feature of the Shomutepe-Shulaveri culture, whose chronological and geographic variations have been suggested by the kinds of temper and decoration patterns (Chataigner 1995). In addition, a few painted fine wares resembling those of the Halaf culture indicate a foreign cultural link to Upper Mesopotamia (Narimanov 1992). In this way, the Shomutepe-Shulaveri culture can be considered a kind of terminus ante quem for the development of agricultural socio-economy in the southern Caucasus. Thus, the question of when and how this cultural entity emerged is a major concern in studying the Neolithization in the southern Caucasus. From this perspective, notable recent discoveries are the initial phases of the Shomutepe-Shulaveri culture at lower levels of Aratashen, Aknashen-Khatunarkh, and Masis Blur in the middle Araxes region (Arimura et al. 2010; Badalyan/Harutyunyan 2014; Hayrapetyan et al. 2014). Although these levels represent fully agricultural settlements, they have yielded very few pottery sherds, one of the hallmarks of the Shomutepe-Shulaveri culture; they dated to early 6th millennium cal. BC, indicating that

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Chipped Stone Technology of the Earliest Agricultural Village in the Southern Caucasus

A

711

B

Fig. 2. A. Topographic map of Hacı Elamxanlı Tepe, showing the excavation squares (shaded areas) in the 2012 and 2013 seasons. B. Plan of mud-brick buildings and other architectural features in Level 3 of Hacı Elamxanlı Tepe.

the Shomutepe-Shulaveri culture is preceded by a pre-pottery phase, which should provide new insights into the Neolithization process in the southern Caucasus. This paper reports a new discovery of such an initial phase of the Shomutepe-Shulaveri culture from Hacı Elamxanlı Tepe, west Azerbaijan. The site is located in the middle Kura Valley, where many Neolithic tepe sites are situated (fig. 1). The excavations at Hacı Elamxanlı Tepe in the 2012–2014 seasons by the Azerbaijani-Japanese mission revealed that the site mostly consists of Neolithic deposits dated to the beginning of the 6th millennium cal. BC, providing a rare opportunity to focus on the investigation of the initial phase of the Shomutepe-Shulaveri culture.

Haci Elamxanli Tepe (Azerbaijan) The site was discovered in 2011 during a survey around Göytepe (ca. 145m in diameter with 11m-thick Neolithic deposits), one of the largest Neolithic sites in the southern Caucasus, where the excavations by the Azerbaijani-Japanese mission since 2008 revealed successive agricultural settlements dated to the mid 6th millennium cal. BC and material cultures affiliated with the Shomutepe-Shulaveri culture (Guliyev/Nishiaki 2012, 2014; Kadowaki et al. 2015). Göytepe is located only 1.5 km apart from Hacı Elamxanlı Tepe, which is a small

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

712

Kadowaki – Guliyev – Nishiaki

mound, measuring 60m x 80m with a height of 1.5m (fig. 2A). The excavations of a 10m x 10m square near the top of the tepe uncovered Neolithic deposits of ca. 1.5m thickness that are divided into four levels according to the stratigraphic sequence of mud-brick buildings. Twelve radiocarbon dates from Levels 1–4 range approximately between 5,950 and 5,800 cal. BC, closely predating the earliest occupation at Göytepe (Nishiaki et al. 2015a). Thus, these two neighboring sites are likely to represent successive occupations by Neolithic inhabitants, i.e., from Hacı Elamxanlı Tepe to Göytepe. Given such a chronological and geographical proximity between the two sites, it is not surprising that the two sites share basic characteristics of material cultures and agro-pastoral economy. The settlements at both sites consist of round mud-brick buildings, around which many architectural features (e.g., storage bins and hearths) as well as domestic refuse (e.g., lithics, bones, charred plants, and ash) are densely distributed (fig. 2B). However, a noticeable characteristic of Hacı Elamxanlı Tepe is the extreme scarcity of pottery, corresponding to the initial phase of the Shomutepe-Shulaveri culture. In each of the levels, only a few sherds have been found in comparison to 1,300–3,300 pieces of chipped stones, and such a small sherd assemblage notably includes two fine wares painted in geometric patterns reminiscent of the Samarra or early Halaf wares (Nishiaki et al. 2013, 2015a). The scarcity of pottery, however, is not the only chrono-cultural marker of the initial phase of the Shomutepe-Shulaveri, but its unique features encompass other material cultures, such as architecture, lithics, and bone artifacts, as well as subsistence practices represented by faunal and botanical remains (Nishiaki et al. 2013, 2015a, 2015b). The following describes chipped stone assemblages from Hacı Elamxanlı Tepe (ca. 8, 400 pieces excavated from Neolithic Levels 1–4 in the 2012 and 2013 seasons), and discusses some techno-typological aspects that may characterize the initial phase of the ShomutepeShulaveri culture.

Chipped Stone Technology Raw Materials Obsidian is the major raw material for chipped stone artifacts, followed by red brown flint, green tuff, red dacite/rhyolite among other kinds. The proportions of obsidian is 45%–57% in the assemblages from Square M10, where excavated sediments were sieved, while the proportion of obsidian increases to 57%–81% in the non-sieved assemblages from Squares L10, L11, and M11.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Chipped Stone Technology of the Earliest Agricultural Village in the Southern Caucasus

713

Among these raw material types, obsidian provides the best flaking quality, and its products occur in a wide range of size (i.e., from large blades or prismatic cores to chips). Obsidian was most likely imported from sources in the southern Caucasus, and its geochemical analyses are in progress for identifying specific sources. Red brown flint and green tuff are fine- to medium-grained. Red brown flint was probably obtained from outcrops in the upstream of the Aghstafa River, ca. 40–60 km away from the site. Various volcanic rocks and green tuff are locally available from river beds or outcrops within the range of 10km from the site.

Core Reduction Methods The production of blades and bladelets by unidirectional flaking is indicated by a large number of blade/bladelet blanks (more than 900 pieces), single-platform blade/bladelet cores (n=15; fig. 3), and numerous core tablets (n=58), some of which retain blade/bladelet scars. Obsidian is mainly used for the production of regular blades/bladelets, while bladelets of irregular forms are made also of other raw material types, such as green tuff and red brown flint (fig. 3.3). Notably, the prismatic blade cores include three complete specimens measuring 95–170 mm in length. The largest core (fig. 3.7) is circular in the cross-section and shows irregular blade or flake scars on the working surface. On the other hand, the other two blade cores (fig. 3.5–6) are flattened in the cross section and show regular, parallel blade scars on their working surfaces that extend about half of the periphery of their striking platform. The two cores have ridges abraded or slightly battered near the periphery of the working surfaces or the back sides. These rounded ridges may have resulted from the stabilization of the cores with vices during their reduction for detaching blades. Such a technique along with regular blade scars indicates the employment of pressure flaking, as suggested for the assemblages from Aratashen and Aknashen-Khatunarkh (Badalyan et al. 2010). There are also several complete cores for bladelets (fig. 3.1–3). Because they are distinctively smaller than the blade cores, the bladelet cores are likely to have been separately set up rather than exploited continuously from larger blade cores. Such a purposeful production of bladelets is probably linked with their selective use for trapezes, as mentioned below. The assemblages also include numerous flakes of various raw materials. Non-obsidian raw materials are mainly exploited for producing flakes of irregular or elongated forms, some of which served as blanks of retouched tools, such as sickle elements and scrapers. Obsidian flakes are also used as blanks of some retouched tools, such as steep round scrapers and splintered pieces. Obsidian flakes should have occurred as byproducts in the preparation or maintenance of prismatic cores. In addition, flake removals are observable on exhausted blade/bladelet cores made of obsidian (fig. 3.4).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

714

Kadowaki – Guliyev – Nishiaki

Fig. 3. Chipped stone artifacts from Hacı Elamxanlı Tepe. (1–3) Bladelet cores, (4) Blade core re-used for flake production, (5–7) Blade cores. (1, 2, 4–7) Obsidian, (3) Red brown flint.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Chipped Stone Technology of the Earliest Agricultural Village in the Southern Caucasus

715

Retouched Tools Tables 1 and 2 show the frequencies of raw material types and blank forms by retouched tool types. Retouched tools are primarily made on obsidian blades/bladelets, followed by obsidian flakes and blanks of other raw materials. Tool types are characterized by high proportions of burins and retouched/nibbled pieces in addition to several unique types, such as trapezes, sickle elements, and steep round scrapers. Trapezes constitute a microlithic component as they are made on bladelets segmented by snapping or oblique truncation (fig. 4.1–19). They are mostly made of obsidian but include a few pieces of red-brown flint. From snapped ends, flat retouch is sometimes made on the dorsal surface. These retouch methods are observable on some bladelet segments (that are not exactly trapezoidal in shape) recovered together with trapezes in the same contexts or in close proximity. The bladelet segments with such unique retouch and recovery contexts are classed here as unfinished trapezes. Burins, the most abundant tool type, are mostly made on obsidian blades, and exhibit various types of facets, including angle, dihedral, and transversal types (fig. 4.21–25). Angle burin facets occur on (often obliquely) truncated ends or snapped ends. Notably, many burin facets are overshot, removing the other end. We recovered more than 200 burin spalls, some of which show alternate flaking scars at their tips that may have been used as borers. Sickle elements are blade segments or flakes with glossed edges (fig. 5.1–8). Although blades are the main blank form, the use of flakes (particularly of non-obsidian raw material) is more frequent than other tool types. Glossed edges are either plain or slightly retouched, while non-glossed sides are often backed or retouched. Retouched or nibbled pieces include blades and flakes with a varying degrees and patterns of retouch on the edges (fig. 5.10–12). Those with smaller, marginal retouch (i.e., nibbling) are classed as nibbled pieces. When retouched edges show denticulated or notched forms, they are classed as denticulates or notches (fig. 5.9). Scrapers include a distinct subtype, i.e., steep round scrapers made on thick flakes of various raw material types, such as obsidian, green tuff, red dacite/rhyolite among others (fig. 5.13–14). Their round forms are created by steeply retouched edges, and the edges of several pieces are partly rounded due to use-wear. Another possible use-wear is matted surfaces with many scratches on the ventral side of obsidian steep scrapers.

Discussion and Conclusion Chipped stone technology of Hacı Elamxanlı Tepe, as described above, is broadly similar to those of Göytepe and other sites affiliated with the Shomutepe-Shulaveri culture. They are primarily characterized by the production of obsidian blades from single-platform prismatic

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

716

Kadowaki – Guliyev – Nishiaki

Fig. 4. Chipped stone artifacts from Hacı Elamxanlı Tepe. (1–19) Trapezes including broken or unfinished pieces, (20) Borer, (21 and 22) Dihedral burins, (23–25) Angle burins on truncation. All are made of obsidian except for No. 8 that is made of red brown flint.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Chipped Stone Technology of the Earliest Agricultural Village in the Southern Caucasus

717

Fig. 5. Chipped stone artifacts from Hacı Elamxanlı Tepe. 1–8: Sickle elements with glossed edges (marked in grey). 9: Denticulated blade. 10/12: Retouched blades. 11. Nibbled blade. 13–14: Steep round scrapers. 1–3, 5, 9–12: Obsidian. 6–8: Red brown flint. 13: Red dacite/rhyolite. 4/14: Unidentified raw materials (non-obsidian).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

718

Kadowaki – Guliyev – Nishiaki

cores, some of which indicate the employment of pressure flaking. Obsidian blades are the major blanks for various retouched tools. There are, however, several characteristics of Hacı Elamxanlı Tepe in the use of raw materials, the core reduction methods, and the retouched tools, in comparison with other sites of the Shomutepe-Shulaveri culture. First, the proportion of obsidian at Hacı Elamxanlı Tepe (45%–57% for sieved assemblages, 57%–81% for non-sieved assemblages) is noticeably smaller than those in the later phase of the Shomutepe-Shulaveri culture, which reportedly accounts for 84%–87% (Korobkova 1996: 74). This suggests that local raw materials are more frequently used at Hacı Elamxanlı Tepe. Second, the chipped stones of Hacı Elamxanlı Tepe clearly include a microlithic component in the core reduction technology (i.e., the production of blades as well as bladelets) and the retouched tools (i.e., trapezes made on bladelets). Although a few trapezes are also present at Göytepe, they occur more frequently at Hacı Elamxanlı Tepe in association with many bladelet cores and blanks. Another characteristic tool type at Hacı Elamxanlı Tepe is steep round scrapers made on thick flakes of various raw material types including both obsidian and locally available rocks. The above techno-typological aspects of chipped stones from Hacı Elamxanlı Tepe may characterize the initial phase of the Shomutepe-Shulaveri culture. This proposal can be further examined by more detailed comparisons with the later assemblages, such as Göytepe, or with other assemblages of the initial phase, such as lower levels at Aratashen, AknashenKhatunarkh, and Masis Blur. For such inter-site comparisons, it is important to consider whether sieving was employed to collect artifacts from excavated sediments because it should significantly affect the recovery rate of microlithic components (e.g., trapezes and bladelets) that may characterize the initial phase of the Shomutepe-Shulaveri culture. If the initial phase of the Shomutepe-Shulaveri can be diagnosed by the occurrence of microlithic technology, it may help us examine the origin of the Shomutepe-Shulaveri culture through the comparison with assemblages of earlier time periods or of other regions. For example, microlithic technology may have persisted from the late Pleistocene to the early Holocene in the southern Caucasus, according to the assemblages from Kotias Klde and Kmlo-2. The Mesolithic assemblage of Kotias Klde Layer B is characterized by microliths, including both non-geometric (e.g., backed or truncated bladelets) and geometric forms (e.g., scalene or isosceles triangles) (Meshveliani et al. 2007). The early Holocene assemblages at Kmlo-2 also include a high proportion (30%) of microliths dominated by backed bladelets and scalene bladelets with some geometric forms (Arimura et al. 2009, 2010). In comparison with these earlier assemblages, the proportion and variety of microliths at Hacı Elamxanlı Tepe are quite limited because the microlithic tool type includes only trapezes that account for less than 5% even in the sieved assemblages. In addition, the retouching technique of trapezes differs between Hacı Elamxanlı Tepe and Kmlo-2. Trapezes at Hacı Elamxanlı are made by segmenting bladelets either by snapping or oblique truncation on the dorsal

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Chipped Stone Technology of the Earliest Agricultural Village in the Southern Caucasus

719

surface. The trapezes are sometimes thinned by flat invasive retouch from segmented ends. These methods appear different from those at Kmlo-2, where bifacial retouch is observable on illustrated trapezes (Arimura et al. 2009: 18). In addition, Kmlo-2 is also characterized by «Kmlo tools» that are defined by continuous, parallel retouch by pressure flaking on lateral sides and abrasion traces on the ventral surface (Petrosyan et al. 2014). We have not identified this tool type at Hacı Elamxanlı Tepe although some of the retouched blades show continuous squamous retouch (fig. 5.10). As for subsistence practices at Kmlo-2, all the identifiable faunal and botanical remains are wild (Arimura et al. 2010). On the other hand, the trapeze dominant microliths at Hacı Elamxanlı Tepe may be more comparable to the assemblage from Layer A2 at Kotias Klde, western Georgia. A brief remark on this layer mentions transverse arrowheads/trapezes, ventrally retouched denticulates, and flake scrapers (Meshveliani et al. 2007). Another possible comparable example is the collection from Anaseuli I, also in western Georgia, which includes trapezes, abundant burins, and steep scrapers (Korobkova 1996; Meshveliani 2013). Although these sites are reported to be «Neolithic» or «aceramic Neolithic», their chronological or subsistence records are too scarce to determine whether they represent a transitional stage from the Mesolithic to the Neolithic. In this way, currently available data on the local cultural sequence are too limited for us to determine whether the trapeze dominant microliths at Hacı Elamxanlı derived indigenously from the more frequent and varied microliths during the Mesolithic or the early Holocene. On the other hand, trapezes at Hacı Elamxanlı may represent part of regional cultural phenomena contemporary with Pottery Neolithic sites in the Middle East, where similar trapezes occur (Nishiaki et al. 2013). In the Levant, Mesopotamia, and Zagros regions, these Neolithic trapezes (along with other geometrics) are considered to have no continuous link with the microliths during the Epipalaeolithic (Nishiaki 1993). Notably, the retouching methods for producing trapezes also indicate this foreign cultural link. As described above, the retouch methods at Hacı Elamxanlı Tepe (i.e., segmenting blades with snapping or oblique truncation with occasional flat retouch on the dorsal surface) are different from those at Kmlo-2 but find some parallels in the trapezes at Pottery Neolithic sites in Upper Mesopotamia, including those contemporaneous with Hacı Elamxanlı, such as Levels 4–10 at Sabi Abiyad (Copeland 1996). In this way, future studies are necessary to determine whether the local Mesolithic or the foreign Pottery Neolithic, or both, have cultural links to the chipped stone technology, particularly its microlithic components, at Hacı Elamxanlı Tepe because this interpretation has direct relevance to our assessment on the roles of indigenous hunter-gatherers or demic/ cultural diffusions from Anatolia or Upper Mesopotamia in the formation of Neolithic socio-economy in southern Caucasus.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

720

Kadowaki – Guliyev – Nishiaki OB

RB flint

Green tuff

Red dacite

Others (non-OB)

Total

Trapeze (n=21)

86%

14%

0%

0%

0%

100%

Unfinished trapeze (n=23)

96%

4%

0%

0%

0%

100%

Sickle element (n=56)

30%

30%

13%

0%

27%

100%

Scraper (n=44)

30%

0%

11%

18%

41%

100%

Burin (n=671)

98%

1%

0%

0%

1%

100%

Denticulate (n=72)

90%

1%

0%

4%

4%

100%

Notch (n=32)

94%

0%

0%

3%

3%

100%

Retouched piece (n=319)

89%

2%

1%

5%

4%

100%

Nibbled piece (n=180)

98%

0%

0%

0%

2%

100%

Borer (n=13)

92%

0%

0%

0%

8%

100%

Splintered piece (n=73)

99%

0%

0%

0%

1%

100%

Truncation (n=29)

93%

0%

0%

0%

7%

100%

Others (n=1)

100%

0%

0%

0%

0%

100%

Table 1. Frequency of raw material types by retouched tool types at Hacı Elamxanlı Tepe. OB: obsidian, RB flint: Red brown flint. Blade/let

Flake

Others or unknown

Total

Trapeze (n=21)

90%

0%

10%

100%

Unfinished trapeze (n=23)

100%

0%

0%

100%

Sickle element (n=56)

61%

36%

4%

100%

Scraper (n=44)

5%

86%

9%

100%

Burin (n=671)

92%

5%

3%

100%

Denticulate (n=72)

82%

17%

1%

100%

Notch (n=32)

69%

31%

0%

100%

Retouched piece (n=319)

71%

29%

0%

100%

Nibbled piece (n=180)

96%

4%

0%

100%

Borer (n=13)

38%

15%

46%

100%

Splintered piece (n=73)

41%

52%

7%

100%

Truncation (n=29)

97%

3%

0%

100%

Others (n=1)

0%

0%

100%

100%

Table 2. Frequency of blank forms by retouched tool types at Hacı Elamxanlı Tepe.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Chipped Stone Technology of the Earliest Agricultural Village in the Southern Caucasus

721

Bibliography Arimura, M./Badalyan, R./Gasparyan, B./Chataigner, C., 2010. Current Neolithic Research in Armenia. Neo-Lithics 1/10, 77–85. Arimura, M./Chataigner, C./Gasparyan, B., 2009. Kmlo 2. An Early Holocene Site in Armenia. NeoLithics 2/09, 17–19. Badalyan, R./Harutyunyan, A., 2014. Aknashen–The Late Neolithic Settlement of the Ararat Valley: Main Results and Prospects for the Research. B. Gasparyan/M. Arimura (eds.), Stone Age of Armenia: A Guide-book to the Stone Age Archaeology in the Republic of Armenia. Center for Cultural Resource Studies, Kanazawa University, 161–176. Badalyan, R./Harutyunyan, A. A./Chataigner, C./Le Mort, F./Chabot, J./Brochier, J. E./Balasescu, A./ Radu, V./Hovsepyan, R., 2010. The Settlement of Aknashen-Khatunarkh, A Neolithic Site in the Ararat Plain (Armenia): Excavation Results 2004–2009. Tüba-ar 13, 185–218. Chataigner, C., 1995. La Transcaucasie au Néolithique et au Chalcolithique. BAR International Series 624. Oxford. Copeland, L., 1996. The Flint and Obsidian Industries. P. M. M. G. Akkermans (ed.), Tell Sabi Abyad: The Late Neolithic Settlement. Volume II. Nederlands Historisch-Archaeologisch Instituut te Istanbul. Istanbul, 285–338. Guliyev, F./Nishiaki, Y., 2012. Excavations at the Neolithic Settlement of Göytepe, the Middle Kura Valley, Azerbaijan, 2008–2009. R. Matthews/J. Curtis (eds.), Proceedings of the 7th International Congress of the Archaeology of the Ancient Near East, Vol. 3: Fieldwork and Recent Research. Wiesbaden, 71–84. Guliyev, F./Nishiaki, Y., 2014. Excavations at the Neolithic Settlement of Göytepe, West Azerbaijan, 2010– 2011. P. Bieliński et al. (eds.), Proceedings of the 8th International Congress of the Archaeology of the Ancient Near East, Vol. 2: Excavation and Progress Reports, Posters. Wiesbaden, 3–16. Hayrapetyan, A./Martirosyan-Olshansky, K./Areshian, G. E./Avetisyan, P., 2014. Preliminary Results of the 2012 Excavations at the Late Neolithic Settlement of Masis Blur. B. Gasparyan/M. Arimura (eds.), Stone Age of Armenia: A Guide-book to the Stone Age Archaeology in the Republic of Armenia. Center for Cultural Resource Studies, Kanazawa University, 177–190. Kadowaki, S./Maher, L./Portillo, M./Albert, R. M./Akashi, C./Guliyev, F./Nishiaki, Y., 2015. Geoarchaeological and Palaeobotanical Evidence for Prehistoric Cereal Storage in the Southern Caucasus: the Neolithic Settlement of Göytepe (Mid 8th Millennium BP). Journal of Archaeological Science 53, 408–425. Korobkova, G. F., 1996. The Neolithic Chipped Stone Industries of the Southern Caucasus. S. K. Kozlowski/H. G. K. Gebel (eds.), Neolithic Chipped Stone Industries of the Fertile Crescent, and Their Contemporaries in Adjacent Regions. Berlin, 57–89.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

722

Kadowaki – Guliyev – Nishiaki

Lyonnet, B./Guliyev, F./Helwing, B./Aliyev, T./Hansen, S./Mirtskhulava, G., 2012. Ancient Kura 2010– 2011: The First Two Seasons of Joint Field Work in the Southern Caucasus. Archäologische Mitteilungen aus Iran und Turan 44, 1–190. Meshveliani, T., 2013. On Neolithic Origins in Western Georgia. Archaeology, Ethnology, and Anthropology of Eurasia 41.2, 61–72. Meshveliani, T./Bar-Oz, G./Bar-Yosef, O./Belfer-Cohen, A./Boaretto, E./Jakeli, N./Koridze, I./Matskevich, Z., 2007. Mesolithic Hunters at Kotias Klde, Western Georgia: Preliminary Results. Paléorient 33.2, 47–58. Narimanov, I. G., 1992. The Earliest Agricultural Settlements in the Territory of Azerbaidzhan. Soviet Anthropology and Archeology 30.4, 9–66. Nishiaki, Y., 1993. Anatolian Obsidian and the Neolithic Obsidian Industries of North Syria: A Preliminary Review. H.I.H. Prince T. Mikasa (ed.), Essays on Anatolian Archaeology. Wiesbaden, 120–140. Nishiaki, Y./Guliyev, F./Kadowaki, S., 2015a. Chronological Contexts of the Earliest Pottery Neolithic in the Southern Caucasus: Radiocarbon Dates for Göytepe and Hacı Elamxanlı Tepe, West Azerbaijan. American Journal of Archaeology 119.3, 279–294. Nishiaki, Y./Guliyev, F./Kadowaki, S./Alakbarov, V./Miki, T./Salimbayov, S./Akashi, C./Arai, S., 2015b. Investigating Cultural and Socioeconomic Change at the Beginning of the Pottery Neolithic in the Southern Caucasis: The 2013 Excavations at Hacı Elamxanlı Tepe, Azerbaijan. Bulletin of the American Schools of Oriental Research 374 (i. p.). Nishiaki, Y./Guliyev, F./Kadowaki, S./Arimatsu, Y./Hayakawa, Y./Shimogama, K./Miki, T./Akashi, C./ Arai, S./Salimbeyov, S., 2013. Hacı Elamxanlı Tepe: Excavations of the Earliest Pottery Neolithic Occupations on the Middle Kura, Azerbaijan, 2012. Archäologische Mitteilungen aus Iran und Turan 45 (i. p.). Petrosyan, A./Arimura M./Gasparyan, B./Nahapetyan, S./Chataigner, C., 2014. Early Holocene Sites of the Republic of Armenia: Questions of Cultural Distribution and Chronology. B. Gasparyan/M. Arimura (eds.), Stone Age of Armenia: A Guide-book to the Stone Age Archaeology in the Republic of Armenia. Center for Cultural Resource Studies, Kanazawa University, 135–159.

Seiji Kadowaki, Nagoya University Museum, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, 4648601, Japan. Farhad Guliyev, Museum of Archaeology and Ethnology, The Institute of Archaeology and Ethnography of the National Academy of Sciences, Baku, Azerbaijan. Yoshihiro Nishiaki, The University Museum, The University of Tokyo, Hongo 7-3-1, Bunkyo, Tokyo, 113-0033, Japan.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 723–733

René Kunze

Interdisciplinary Studies on the Small Finds from the Late Bronze/Early Iron Age Settlements of Udabno I–III (Eastern Georgia) The goal of this paper is to show the results of the analysis of more than 1400 small finds from the settlements of Udabno which is situated in the East Georgian region of Kakheti and date to the transition period from the 2nd to 1st millennium BC, i.e. the Late Bronze/Early Iron Age in the South Caucasus.

Introduction This paper about the small find inventory of the Late Bronze/Early Iron Age settlements of Udabno in eastern Georgia presents a part of the author’s doctoral thesis (Kunze 2012) and is based on the activities of the University of Tübingen at Troy and their implications for research related to Black Sea archaeology, going back to the 1990s (Korfmann et al. 2005). The find spot of Udabno is situated about 40 km southeast from the Georgian capital, Tbilisi, in the Georgian region of Kakheti (coordinates 41°36´54.65´´N; 45°17´54.97´´E). It is located in the mountainous David Gareji Desert between the Iori River to the north and the Mtkwari (Kura) River to the south (fig. 1). Since 2000 eight archaeological campaigns were carried out at three large settlements (Udabno I, II, and III) under the direction of Manfred Korfmann, Ernst Pernicka and Jan-K. Bertram together with the Georgian Academy of Sciences represented by Konstantin Pizchelauri. The areas of Udabno I–III were excavated comprehensively based on the results of aerial images and magnetic prospection. It revealed single-phase settlements. The chronology of the aforementioned sites is based on dozens of 14 C-analyses (Bertram/Ilgezdi-Bertram 2012: 95–97). On this basis the three settlements date to the Late Bronze/Early Iron Age, i.e. the transition from the 2nd to 1st millennium BC.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

724

Kunze

Fig. 1. Aerial image. Settlements of Udabno I–III (modified, after Korfmann et al. 2005: 177, fig. 2; overview map designed by stepmap).

Ground-level structures with dry masonry walls rest within the fortified citadel area in the northwest of Udabno I (fig. 2a). From there to southeast plenty of pit dwellings have been found. The dwellings are rectangular with an area about 40 to 60 square metres (fig. 3). Remains of 18 houses have been excavated in Udabno I. The dwellings have walls sometimes strengthened lightly with orthostats (fig. 4). Same types of pit dwellings were also found at the two other settlements Udabno II and III situated about 1000 m at the north of Udabno I (fig. 2b/c). Remains of two houses in Udabno II and five houses in Udabno III were explored. With exception of the citadel in Udabno I, intensive traces of burning have been found on the floors of houses in most cases. According to current research results, a conflagration was the reason for the abandonment of the settlements after a relatively short duration (at least in Udabno I and II). The majority of the houses were destroyed. The inventories in these burnt houses therefore were preserved in situ.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Interdisciplinary Studies on the Small Finds from … Udabno I–III (Eastern Georgia)

725

Fig. 2. Udabno I–III. Magnetograms on aerial images with marks (black dots) of excavated houses (modified, after Hübner et al. 2002).

Small Find Inventory The goal of this report is to give a short overview about more than 1400 small finds which were documented stratigraphically during the campaigns that were started in 2000 and finally completed in 2007. A special focus involved drawing conclusions about the everyday life and organisation in the settlements and the reconstruction of domestic activity areas and lifestyle generally as well as some facts about the socio-economic structures during the transition from the 2nd to the 1st millennium BC in the eastern central Caucasus region.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

726

Kunze

Fig. 3. Example of a «typical» pit dwelling in Udabno (middle part of Udabno I, house J).

The small finds were grouped in stone (70%), metal (15%), obsidian (2%), bone and clay (13%) artefacts (fig. 5). Among other finds, a large number of items made of stone were analysed, including mortars and pestles, grinders and rubbers, pounders and stone axes. Beside these artefacts, the assemblage includes metal finds in the form of spearheads, knives, chisels, pins and punched sheets. Almost all the metal artefacts are made of arsenic copper or tin bronze. Iron is used only sporadically. In addition, there are some small finds made of bone like combs and made of potter’s clay, like spindle whorls and (bread) stamps. Obsidian artefacts in form of sickles, scrapers and weapons, i.e. arrow heads, were rare. Often lithic debitages, consisting of broken pieces, were found in the houses. Archaeometrical analyses of the obsidian by using neutron-activation analyses (NAA) have revealed an import from the geological source of Chikiani, which is situated at the northeastern shore of Lake Paravani in the region of Kvemo Kartli and Samtskhe-Javakheti in southeastern Georgia. Frequently, a complete inventory of small finds as well as pottery provided an insight into the households. Thus, it is possible to reconstruct the household organisation in great detail. Many small finds are closely packed together what allows a detailed analysis of the activities in these houses and the building-structure. Upon the analysis of small finds, combined

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Interdisciplinary Studies on the Small Finds from … Udabno I–III (Eastern Georgia)

727

Fig. 4. Northern wall of a dwelling. Strengthened lightly with orthostats (middle part of Udabno I, house K).

with the pottery (Brodbeck-Jucker in prep.) four activity zones could be constructed within the dwellings: areas for cooking/food preparation, working, living, and for storage (fig. 6). It seems that the houses were used for multi-functional activities, for eating, sleeping, working and all other social engagements. Most of the activities were domestic. Different working areas could be identified by the distribution of the small finds: production of textiles (weaving combs, spindle whorls, loom weights), bone tools (awls), jewellery (carnelian beads, and stone pendants), and leather and obsidian-working (sickle elements and scrapers). The documented oven or stove structures in the northwestern corner of most houses point to a decentralised production of food, i.e. an autonomous domestic production of basic foodstuffs. Around the hearths, implements of food preparation such as grinders and pestles often were recovered. Some activities for foodstuffs seem to have taken place outside of the houses, in front of the door or probably also on the roofs. A significant concentration of grinding stones and (bread) stamps inside the citadel and bordering to its walls (Udabno I) can only be interpreted as an indication of the citadel’s function as storage place and for

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

728

Kunze

Fig. 5. Assembling of some small finds found in the settlements of Udabno I–III. a. mortar and pestle. b. (bread) stamp. c. spearhead (tin bronze). d. decorative punched and perforated sheets (jewellery). e. pendants (stone). f. knife (arsenic copper/bronze).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Interdisciplinary Studies on the Small Finds from … Udabno I–III (Eastern Georgia)

729

Fig. 6. Activity zones upon the analysis of small finds using the example of house UIII-C.

workmanship and not as a «bakery» or something else. There is no evidence for a central management of surplus foods in case of famine based on the small finds. The type and quality of the small finds point to a rather low level of prosperity. Animal husbandry, without signs of transhumance (Knipper et al. 2008: 168), seems to be the most important activity of the residents. According to an examination of the animal bones, sheep dominate the bone material, followed by horses and pigs (Uerpmann 2006). Secondary evidence for agricultural activity includes remains of a threshing sledge as well as many (bread) stamps and grinding stones. Finds such as spindle whorls, loom weights, and weaving combs in the houses are evidences for textile production. The quantity of these finds shows the importance of textile trade at Udabno. As already mentioned, the highest percentage in the bone material could be assigned to sheep. However, for food production the residents often used the meat of pigs and horses. This investigation also supports the interpretation of using the lamb’s wool for textile production.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

730

Kunze

In the range of weapons only one spearhead (house K, Udabno I), and some arrow heads made of obsidian and bone, as well as some sling shots were documented and classified. The latter of them could be used also as hunting gear, inasmuch bones of hares and birds or deers have been found. Jewellery includes metal rings, pins, necklaces and punched sheets as well as carnelian beads and pendants made of stone and snails. More than 60 beads as well as 50 pendants were found in each case together as a deposit in two pits in Udabno I (house K) and III (house B). Due to the location of the tools (e. g., chisels) around these deposits and the abundant semi-products, it is interpreted as jewellery-craftsman’s storages. As a result of the investigation of the small finds and their architectural context it can be said that there is no compelling evidence for specific differences within the settlements that indicate social stratification. The almost uniform distribution of stone artefacts (grinding and construction tools) gives no direct indication of different occupational groups. It is probable that the development of social stratification was prevented by the short duration of the settlement activity in this region.

Archaeometrical Results of the Metal Finds A special study was dedicated to the issues of copper-bronze artefacts from Udabno. 31 objects obtained by the Udabno excavation team have been studied as well. The material, as has already been mentioned, consists of a spearhead, many kinds of pins and rings, some decorative and perforated sheets, chisels, bracelets and a knife. Chemical compositions of alloys were defined and metallographic examination was performed. The data of x-rayfluorescence analysis (XRFA) at the Curt Engelhorn Centre for Archaeometry (Mannheim, Germany) and electron probe microanalysis (EPMA) at the Institute of Geochemistry, University of Göttingen (Germany), is able to classify into two main alloy types. In the material of the same period arsenic copper and tin bronze were used in equal quantities (fig. 7). There are no differences between artefact groups; so examplarily some pins which are equal in typology were made of arsenic copper, some others of tin bronze. Unfortunately it was not possible to make a real statement for metal treatment. For sure, the tin ores are not from the Greater or Lesser Caucasus region. Maybe they are from the regions of Central Asia. The chemical composition of the arsenic copper artefacts is comparable with all other items from the Central Caucasus. So there is a special fingerprint for this ores: a relatively high content of arsenic and a very low content of nickel (e.g. Kavtaradze 2001: 136–138; Badalyan et al. 2007: 52). For that reason we can assume regional copper sources for the bronzes from Udabno – probably from the eastern part of the Caucasus Mountains. For this purpose, there are just ongoing survey activities by the author in the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Interdisciplinary Studies on the Small Finds from … Udabno I–III (Eastern Georgia)

731

Fig. 7. Plotting of the comparison of tin and arsenic of the investigated metal finds.

northern part of Kakheti. First results show some indications of metallurgy at least for the pre-Christian period (Kunze 2015). Based on metallographical and further analysis most of the objects show a very simple kind of production, e.g. the mentioned spearhead which was found at the floor level in house K in Udabno I. Many metallurgical defects in the microstructure, like blowholes, can be shown, as well as the lack of rework after casting. Some other items, especially jewellery, were produced by recycled materials.

Origin of the Raw Material for (Some) Stone Artefacts Another category of small finds can help to understand trade routes during the end of 2nd and beginning of 1st millennium BC in eastern Georgia. In the inventory there was a large number of grinding stones consisting of a very coarse-grained porous rock, apparently basalt and andesite. The grinding stones also had a secondary function as building bricks or stones and may have been used to secure the roofing of houses. Traversals and geological investigations

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

732

Kunze

of the vicinity of Udabno, as well as geological maps, showed no basaltic outcrops. Grinding stones were, however, an important commodity as trade items. The chemical analysis gives an indication of a common area of origin. The Udabno artefacts correspond to geological samples of eruptive basalt from the Southeast-Georgian and North-Armenian area around Bolnisi in the Kvemo Kartli region and Lori in Armenia in the Lesser Caucasus. This mineralogical evidence for the sources of the material used for producing grinding stones in Udabno as well as the obsidian analyses provide the basis for a reconstruction of trade routes and intra-regional relationships during the Bronze to Iron Age transition. Based on such data, it is not possible to make an unambiguous assignment to a particular locality – however, the samples show a regional unity.

Conclusion The activities of the residents of Udabno have been divided into two main classes: subsistence and manufacturing. Among the manufacturing activities, small finds/tools for the production of textiles and jewellery were found very often. Several studies as part of the author’s doctoral thesis have addressed the question of origin (basalts, metals, obsidian). Unfortunately, for chronology small finds are not usable because of their «timelessness». The picture of the agricultural and «industrial» activities that was gained from our analyses shows us the daily life of a rural population in the Late Bronze/Early Iron Age which had individual trade links to the southeast of Georgia, northern Armenia and probably also to eastern Anatolia. The excavations in Udabno revealed an interesting example of a characteristic rural settlement between the Greater and Lesser Caucasus Mountains. The magnetic prospection and the results of the excavations show that we are dealing with planned settlements from the late 2nd to the early 1st millennium BC containing varying structures that may be considered. However, a final assessment will only be possible by combining all the results of the current investigations, particularly the pottery and the architecture.

Acknowledgment This research about the small find inventory has been carried out within the context of the author´s doctoral thesis. I would like to express my special thanks to my supervisors Ernst Pernicka (Mannheim/Heidelberg) and Jan-K. Bertram (Ankara) as well as to all persons who actively support me in this often difficult and challenging work.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Interdisciplinary Studies on the Small Finds from … Udabno I–III (Eastern Georgia)

733

Bibliography Badalyan, R./Lombard, P./Avetesyan, P./Chataigner, C./Chabot, J./Vila, E./Hovsepyan, R./Willcox, G./ Pessin, H., 2007. New data on the Late Prehistory of the Southern Caucasus: The excavations at Aratashen (Armenia). B. Lyonet (ed.), Les cultures du Caucase (VI.-III. Millénaires avant notre ère): Leurs relations avec le Proche-Orient. Paris, 37–61. Bertram, J. K./Ilgezdi-Bertram, G., 2012. Eine erste Zusammenfassung der Ausgrabungs- und Prospektionsergebnisse nach Abschluss der Feldarbeiten. A. Mehnert et al. (eds.), Austausch und Kulturkontakte in der Spätbronze-/Früheisenzeit im Südkaukasus und seinen angrenzenden Regionen. Schriften des Zentrums für Archäologie und Kulturgeschichte des Schwarzmeerraumes, 22. Langenweißbach, 87–121. Hübner, C./Giese, S./Grubert, A., 2002. Geomagnetische Prospektionskampagne 2000 in Udabno, Kachetien, Georgien. Studia Troica 11, 427–436. Kavtaradze, G. L., 2001. Die frühesten Metallobjekte in Zentral-Transkaukasien. I. Gambaschidze et al. (eds.), Georgien – Schätze aus dem Land des Goldenen Vlies. Ausstellungskatalog des Deutschen Bergbaumuseums Bochum 2001–2002. Bochum, 136–141. Knipper, C./Paulus, S./Uerpmann, M./Uerpmann, H. P., 2008. Seasonality and land use in Bronze and Iron Age Kakhetia (Georgia). Oxygen and strontium isotope analyses on horse and cattle teeth. Archäologische Mitteilungen aus Iran und Turan 40, 149–168. Korfmann, M./Pizchelauri, K./Bertram, J.K./Kastl, G., 2005. Erster Vorbericht zu den Vermessungs- und Ausgrabungsarbeiten in Udabno (Ostgeorgien) 2000–2004. Archäologische Mitteilungen aus Iran und Turan 35/36, 275–224. Kunze, R., 2012. Interdisziplinäre Studien zu den Kleinfunden der Siedlungen Udabno I–III (Ostgeorgien), dissertation at the University of Tübingen, online publication: http://hdl.handle.net/10900/49922. –

2015. Living and working in Late Bronze/Early Iron Age Georgia – The settlements of Udabno in Kakheti (eastern Georgia) and a contribution to metallurgy based on a field survey in the upper Alazani River basin. S. Makalatia (ed.), Studies in Caucasian Archaeology 3. Gori, in press.

Uerpmann, M., 2006. Tierknochenfunde aus Udabno I (Georgien). Archäologische Mitteilungen aus Iran und Turan 38, 197–211.

René Kunze, Collaborative Research Center “ResourceCultures” (SFB 1070) Eberhard Karls University of Tübingen, Germany.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

General Topics

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 737–739

Silvia Alaura – Davide Nadali

Researching the Archives: A Shared Past for the Future. The GRISSO Project GRISSO was founded in 2013 in Rome. GRISSO’s current research centers around the life and scientific activities of archaeologists, assyriologists and hittitologists and around the history of museums and academic institutions they belonged to. Further studies deal with the historical, political and socioeconomic contexts in which the archaeological explorations took place and their reception by the public. The Gruppo di Ricerca Interdisciplinare di Storia degli Studi Orientali (GRISSO) – Group for Interdisciplinary Research on the History of Oriental Studies – was founded in 2013 by Silvia Alaura as a permanent initiative at the Istituto di Studi sul Mediterraneo Antico (ISMA), CNR Roma, in collaboration with the Sapienza Università di Roma, as well as with other Italian and foreign universities and research centres (Accademia Nazionale dei Lincei – Fondazione Caetani, Roma; Università di Torino; Centro Ricerche Archeologiche e Scavi di Torino per il Medio Oriente e l’Asia; Università di Firenze; Freie Universität Berlin; Albert-Ludwigs-Universität Freiburg; UMR 8546 «AOROC», CNRS – École normale supérieure, Paris; Yale University, USA). GRISSO focuses on the history of archaeological, philological and historical studies of the ancient Near East and their reception in contemporary and modern society. GRISSO aims at establishing cross-cultural academic dialogue and worldwide cooperation in the field of the history of Oriental studies, which is still in its earliest stages. The GRISSO initiative was presented at CNR, Rome, on 5 November 2013 during an international launch workshop «Scavi d’archivio: dalla storia dell’Orientalistica alla storia delle idee» organised by Silvia Alaura with the participation of Mario Liverani, Jörg Klinger, Stefano de Martino, Marco Bonechi, Davide Nadali, Marina Pucci and Maria Gabriella Micale. The proceedings of the workshop will be published in a forthcoming volume (Alaura 2015). The volume will include not only the contributions of the participants in the conference but also articles from other scholars in order to illustrate all the fields of interest and

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

738

Alaura – Nadali

activity of GRISSO (the practice of archaeology, between institutions and excavations; the making of philologies, from unity to specialization; integrating oriental studies and social life). One of the most important aspects of GRISSO’s work is the identification, study and evaluation of unpublished or insufficiently exploited archive materials. Archives not only provide material documents (correspondence, photographs, drawings, cards, field notes and journals), but also make it possible to retrace the ideas and purposes lying behind an archaeological excavation and philological activity. Archives are a resource which gives the whole of society access to the evidence of our shared past. Awareness of the role of archival documentation in historical and archaeological research has already led some of the GRISSO’s members to work individually in this field (Alaura 2006, Amadasi 2009, Fenet 2013, Foster 2006, Ghione/Sagaria Rossi 2004, Liverani 2013, Nadali 2010–2011, Micale 2008, Petersen 2008, Pucci 2007–2008). However, GRISSO’s distinguishing characteristic is its interdisciplinary approach: its members represent various disciplines such as archaeology, philology, ancient and modern history, the history of ideas, philosophy, anthropology and archival science. The GRISSO members working in such diverse cultural fields are organised into three internationally composed research teams: 1) The key figures in Oriental studies and their universities and institutes. Ongoing research projects: Otto Puchstein (S. Alaura, L. Petersen); Leone Caetani (M. Liverani, V. Sagaria Rossi, S. Alaura, D. Nadali); Giulio Cesare Teloni, Gerardo Meloni, Giorgio R. Castellino (M. Bonechi, D. Baldi); the Beginning of Assyriology in the USA (B. R. Foster); Hans Gustav Güterbock (S. Alaura); Giorgio Levi Della Vida (M. G. Amadasi); Giorgio Gullini and the Centro Scavi di Torino (S. de Martino, C. Lippolis); the Orient-Comité Excavations at the site of Zincirli, Turkey (M. Pucci, N.-A. Peek); the Oriental Institute Excavations at Chatal Höyük in the Amuq Plain, Turkey (M. Pucci); the Formative Period of the Oriental Studies in France (A. Fenet). 2) Oriental studies in the epoch of ideologies. Ongoing research project: Studien zur Entwicklung der Hethitologie in Deutschland in den 20er und 30er Jahren des 20. Jahrhunderts (S. Alaura, J. Klinger). The project was financed in 2012 by a DAAD (Deutscher Akademischer Austausch Dienst) Grant. 3) The acceptance of Oriental studies in society: (ab)use, reuse and adaptation. Ongoing research projects: the use of photography in archaeology (C. Casero, M. G. Micale, D. Nadali); the archaeology of images and words (D. Nadali, M. G. Micale); methodology in classifying archive materials: photographs, drawings, field notes (D. Nadali, M. G. Micale, M. Pucci); philately and the Ancient Near East (A. Ercolani, U. Livadiotti). Next steps for the GRISSO initiative may be presented as a one-day seminar on a case study, or in a more interactive workshop format focusing on selected topics, process improvement or work strategies.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Researching the Archives: A Shared Past for the Future. The GRISSO Project

739

Head Office and Contact Information GRISSO – Istituto di Studi sul Mediterraneo Antico (ISMA) Consiglio Nazionale delle Ricerche (CNR) Area della Ricerca di Roma 1 Via Salaria km 29, 300 – C.P. 10 I – 00015 Monterotondo Scalo (ROMA) Email: grisso (at) isma.cnr.it

Bibliography Alaura, S., 2006. “Nach Boghasköy!” Zur Vorgeschichte der Ausgrabungen in Boğazköy-Ḫattuša und zu den archäologischen Forschungen bis zum Ersten Weltkrieg. Darstellung und Dokumente. 13. Sendschrift DOG. Berlin. Alaura, S. (ed.), 2015. Digging in the Archives: From the History of Oriental Studies to the History of Ideas. Documenta Asiana XI. Roma, forthcoming. Amadasi, M. G., 2009. Giorgio Levi Della Vida. Perché il ricordo?. Civiltà del Mediterraneo 15, 105–115. Fenet, A., 2013. La Société Asiatique: histoires et collections. II. Genèse d’une bibliothèque orientaliste, de la Restauration au début du Second Empire. Journal asiatique 301, 279–322. Foster, B. R., 2006. The Beginnings of Assyriology in the United States. S. W. Holloway (ed.), Orientalism, Assyriology and the Bible, Sheffield, 44–73. Ghione P./Sagaria Rossi V., 2004. L’Archivio Leone Caetani all’Accademia Nazionale dei Lincei. Studi e documenti d’archivio. Fondazione Camillo Caetani 11. Roma. Liverani, M., 2013. Immaginare Babele. Due secoli di studi sulla città orientale antica. Roma/Bari. Micale, M. G., 2008. European Images of the Ancient Near East at the Beginnings of the 20th Century. J. Nordbladh/N. Schlanger (eds.), Archives, Ancestors, Practices. Archaeology in the Light of its History. Goteborg, 191–203. Nadali, D., 2010–2011. L’archeologia di Nabucco: l’Oriente antico in scena. Studi Verdiani 22, 73–88. Petersen, L., 2008. Die Orientreise des deutschen Kaisers Wilhelms II. und der Beginn der Ausgrabungen in Baalbek 1898. Ch. Trümpler (ed.), Das große Spiel. Archäologie und Politik im Zeitalter des Kolonialismus. Köln, 398–409. Pucci, M., 2007–2008. The Chatal Höyük Publication Project, The Oriental Institute Annual Report 2007– 2008, 19–25.

Silvia Alaura (ISMA, CNR, Roma). Davide Nadali (Sapienza Università di Roma).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 741–749

Gino Caspari

Ein Inventar der Hügelgräber im Altaigebirge. Kulturgüterschutz durch Fernerkundung The frozen tombs of Iron Age civilizations in the Altai Mountains are threatened by climatic change. Using high-resolution optical data and algorithm-based automatic detection, we propose a method for tagging areas of paramount importance for future intensive surveys and preservation measures, thus rendering a time intensive manual process obsolete.

1. Einleitung Die Eisgräber der eisenzeitlichen Kulturen des Altaigebirges laufen Gefahr, durch die voranschreitende Klimaerwärmung aufzutauen und einem unumkehrbaren Zersetzungsprozess anheimzufallen. Dadurch wird dieses einzigartige kulturelle Erbe, welches mehr als 2000 Jahre unbeschadet überdauert hat, unwiederbringlich zerstört (Bourgeois et al. 2007: 458–459). Im letzten Jahrzehnt wurden Anstrengungen unternommen, diese Quelle archäologischen Wissens mittels internationaler Surveyprojekte zu sichten und zu katalogisieren, um durch ein repräsentatives Inventar und genaue Karten eine Übersicht über die cultural landscape des Altaigebirges zu erhalten (siehe z.B. Bourgeois et al. 2010). Obwohl diese Erkundungen unser Wissen erweiterten und durch minutiöse Kartierung einer Vielzahl von Monumenten erstmals einen Einblick in die räumliche Organisation der archäologischen Hinterlassenschaften boten, so blieben die untersuchten Flächen im Vergleich zu dem Gebiet, in dem potenziell Eisgräber vorhanden sind, doch stark begrenzt. Begehungen vor Ort haben klare Limitationen und sind nicht für eine überregionale Kartierung und Inventarisierung im Hinblick auf die Implementierung eines nachhaltigen Kulturgüterschutzes geeignet.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

742

Caspari

2. Nachhaltiger Schutz der Eisgräber Die beschränkte Zugänglichkeit der Gebirgsregionen und die weitverstreuten archäologischen Sites in einem sich über tausende von Quadratkilometern erstreckenden Gebiet machen eine komplette Kartierung mittels traditioneller Methoden schlichtweg unmöglich. Monetäre und zeitliche Ressourcen sind äusserst begrenzt, selbst wenn es sich um ein solch einzigartiges kulturelles Erbe wie die Eisgräber in den Permafrostgebieten des Altai handelt. Feldprojekte werden weiter kompliziert durch den Umstand, dass das Gebirge auf dem Territorium vierer Staaten mit unterschiedlichen Regulationen bezüglich archäologischer Forschungsvorhaben liegt. Teilweise militarisierte Grenzzonen können Lizenzbeantragungen verunmöglichen oder so einschränken, dass beispielsweise die Verwendung von GPS-Technologie nicht erlaubt wird. Schliesslich besteht besonders im südlichen, chinesischen Teil ein latentes politisches Risiko, das die Durchführung von Projekten zum Vabanquespiel werden lässt. Die Durchführung eines regionalen oder gar internationalen Schutzprojektes ist somit eine äusserst komplizierte Angelegenheit. Der Plan zum Schutz der altaiischen Eisgräber, welcher von einem Team rund um J. Bourgeois entwickelt wurde, schlägt vier – teilweise parallel ablaufende – Implementierungsphasen vor (Bourgeois et al. 2007: 463): 1. Die Etablierung eines Inventars der archäologischen Hinterlassenschaften und dessen Integration in ein Geografisches Informationssystem als Planungswerkzeug. 2. Die Erforschung der Permafrostgrenzen, sowie Temperaturschätzungen und Trendanalysen, um einzelnen Monumenten Risikokategorien und Prioritäten zuweisen zu können. 3. Die Entwicklung von Methoden zur in situ Erhaltung. 4. Das Ausgraben und Auswerten von Gräbern, welche sich nicht anders vor dem Auftauen bewahren lassen. Da sich wie oben erläutert reine bodenbasierte Surveys der Landschaft nicht als Methode für eine grossräumige Inventarisierung eignen, müssen Fernerkundungsmethoden diese Lücke schliessen.

3. Ziele und Ausgangslage Ziel der hier präsentierten Untersuchungen war es, ein effizientes Werkzeug zur automatischen Analyse grosser Datenmengen zu schaffen, welches zuverlässig Hotspots in optischen Daten erkennt. Diese sollten für eine genauere optische Analyse markiert werden, um schliesslich – bei entsprechendem Resultat – bei der vorbereitenden Untersuchung und Planung von Surveys von Nutzen zu sein.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Ein Inventar der Hügelgräber im Altaigebirge

743

Die meisten frei verfügbaren optischen Daten weisen eine zu geringe Auflösung auf, um eine umfassende Detektion auch kleinerer Gräber zu ermöglichen. Die überwiegende Mehrzahl der Gräber verfügt über einen Durchmesser von 5–18 m (Gheyle 2009: 180). Somit bieten beispielsweise KH-6 CORONA-Daten mit einer Auflösung von 1.8 m zu wenig Information für einen verlässlichen algorithmusbasierten Ansatz. Aus diesem Grund wurden für Adaption und Analyse des Algorithmus hochauflösende Daten verwendet –im vorliegenden Falle IKONOS-2 Daten mit einer Auflösung von 0.8 m panchromatisch und 3.2 m multispektral. Das Erscheinungsbild eisenzeitlicher Grabhügel ist stark von Grösse, Erhaltungszustand und Bewuchs abhängig, wobei sich die drei Parameter nicht klar voneinander trennen lassen. Die grössten Hügel weisen Durchmesser von über 50 m auf, während die kleinsten noch detektierbaren einen Durchmesser von etwa 5 m haben. Selbstverständlich gibt es noch kleinere Aufschüttungen, dort stossen wir aber sehr schnell an die Grenzen der Datenauswertung. Viele der Strukturen sind von Vegetation bedeckt. Nicht nur Gras und kleinere Sträucher wachsen zwischen den Steinen, sondern in vielen Fällen verändern Nadelbäume das Erscheinungsbild der Gräber bis zur Unkenntlichkeit. Zudem sind die tiefergelegenen Täler des Altai durchwegs landwirtschaftlich erschlossen. Ackerbau findet primär auf den Flussterassen statt, auf welchen auch die Hügelgräber liegen. Kleinere Gräber werden häufig zerpflügt und sind danach lediglich noch als verwischte, ovale Bodenmerkmale identifizierbar, wohingegen auf grössere Hügel häufig Rücksicht genommen wird. Die umfassende Beraubung der Strukturen verändert deren Form durch Umschichtung von Material. Ironischerweise kann die zumeist zentral gelegene Raubgräbermulde mithelfen zu bestimmen, ob es sich wirklich um ein eisenzeitliches Grab handelt. Schliesslich tragen natürliche Erosionsprozesse das ihrige zur Veränderung der originalen Form der Strukturen bei. Trotz aller Veränderungen bilden die eisenzeitlichen Hügelgräber eine relativ uniforme Objektklasse. Dieser Umstand macht ihre Detektion vergleichbar mit anderen wohlbekannten Herausforderungen im Forschungsfeld der computer vision. Die Detektion und Verfolgung von Fussgängern oder das Auszählen von Fahrzeugen in Videomaterial stellen vor ähnliche Probleme. Aus diesem Grunde sind bereits eine Anzahl von gut dokumentierten Methoden vorhanden, welche adaptiert und auf diese archäologische Fragestellung angewendet werden können (Caspari et al. 2014: 906).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

744

Caspari

4. Automatische Detektion Die Wahl fiel nach ausgiebigen Tests mit sowohl sofistizierten als auch simpleren computer vision Algorithmen auf objektklassenspezifische Hough Forests, da letztere aufgrund ihres im Vergleich klar besseren Abschneidens bezüglich Robustheit und Detektionsrate hervortraten. Da die Funktionsweise des Algorithmus bereits andernorts ausführlich besprochen wurde (siehe Caspari et al. 2014) wird hier lediglich eine kurze Übersicht gewährt: Hough Forests sind randomisierte Entscheidungswälder, welche über ein Stimmenabgabesystem Objekte detektieren (Gall et al. 2011: 2188). Zur Erstellung der Entscheidungsbäume, aus denen sich die Hough Wälder zusammensetzen, müssen zunächst Trainingssamples aus den Daten extrahiert werden. Dies geschieht, indem positive und negative Samples aus den verwendeten Rasterdaten ausgeschnitten und als Bilddateien gespeichert werden. Bei der Verarbeitung werden die positiven Trainingssamples (die Objektklasse) auf eine einheitliche Grösse skaliert. Negative Trainingssamples werden vom Hintergrund erstellt, wobei es sinnvoll ist darauf zu achten, dass möglichst alle vorhandenen Hintergrundstrukturen abgedeckt werden (Wald, Felsen, Grasland, Acker ect.), da dies die Anzahl der falsch-positiven Detektionen vermindert. Aus den Trainingsbildern werden sodann sogenannte Patchs kreiert, aus denen während des Trainings Punkte extrahiert werden, die die Basis für simple binäre Tests bilden. Da bekannt ist, welche Patchs aus positiven und welche aus negativen Trainingssamples stammen, können die zufällig generierten Tests bewertet werden. Jeder binäre Test teilt das Trainingssample in zwei Hälften auf. Damit ist die erste Verzweigung des Entscheidungsbaumes gebildet. Es entstehen zwei neue Ausgangspunkte, an denen wieder binäre Test durchgeführt und neue Äste des Baumes konstruiert werden. So wird fortgefahren, bis jeder Patch korrekt zugeordnet ist. In den Blättern des Baumes werden die Patchs gespeichert. Auf diese Weise wird eine beliebige Anzahl Entscheidungsbäume geschaffen, welche zusammen den Entscheidungswald bilden (Gall et al. 2011: 2191–2196). Während der Durchführung des Detektionsprozesses gibt jeder Entscheidungsbaum seine Stimme ab, wenn ein Patch für zu einer gesuchten Struktur zugehörig befunden wurde. Mittels thresholding wird danach bestimmt, wie viele Stimmen ein Objekt benötigt, um als Detektion zu gelten. Für einen maximierte Detektionsrate müssen verschiedene Parameter optimiert werden: Dies sind die Anzahl der Entscheidungsbäume, die Grösse der extrahierten Patches, die Skalierung der Testbilder, die Anzahl der Trainingsbilder, der Schwellenwert, der die Stimmenanzahl für die Markierung als Detektion bestimmt, die Scanfenstergrösse, sowie die Abgrenzung, welche bestimmt, wie nahe zwei detektierte Objekte beeinanderliegen können.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Ein Inventar der Hügelgräber im Altaigebirge

745

5. Resultate Für die Auswertung und Anpassung des Algorithmus wurden zwölf Testgebiete definiert, von denen eines für das Training des Algorithmus genutzt wurde und die anderen zur Evaluierung dienten. Für die Evaluierung wurden drei Kategorien von Grabhügeln anhand ihrer Erkennbarkeit definiert. Mit Kategorie I wurden solche Monumente taxiert, welche sich sowohl durch den starken Kontrast zum Hintergrund als auch durch ihre kreisrunde und wenig durch äussere Einflüsse – wie Bebauung, Vegetation und landwirtschaftliche Aktivitäten – veränderte Form auszeichnen. In Kategorie II befinden sich alle Grabhügel, welche von Auge einwandfrei als solche erkannt werden können, jedoch durch viel Bewuchs und anthropogene Einflüsse gegenüber ihrer ursprünglichen Form Veränderungen aufweisen. Kategorie III besteht aus Monumenten, welche von Auge zwar noch erkennbar sind, sich aber nur aufgrund ihres Kontextes (beispielsweise Einreihung in eine lineare Folge anderer Kurgane) als solche Strukturen einordnen lassen. Zudem werden alle Strukturen, die sich lediglich über die zentrale Absenkung als eisenzeitliche Grabhügel identifizieren lassen, sich sonst aber kaum von der Umgebung unterscheiden, respektive stark durch äussere Einflüsse in ihrer Form verändert wurden, dieser letzten Kategorie zugeordnet. Die Detektion von Monumenten der Kategorie I funktioniert sehr gut. Selbst unter eingeschränkten Parametern werden über 80% der Grabhügel detektiert. Die optisch einfach zu detektierenden Grabhügel werden von den objektklassenspezifischen Hough-Entscheidungswäldern so aufgespürt und gekennzeichnet, dass ein repräsentatives Bild entsteht (fig. 1). Die Grabhügel der Kategorie II werden erheblich schlechter aufgefunden. Bereits bei einem leicht erhöhten Schwellenwert werden nur noch knapp 60% erkannt. Die Detektion von stark gestörten oder überwachsenen Grabhügeln ist als nicht erfolgreich zu werten. Von den über hundert vorhandenen Strukturen wurden lediglich 13.5% aufgefunden. Manche Trainingssets zeigen eine vergleichsweise gute Leistung bei Grabhügeln der Kategorie III, während weder Monumente der Kategorie I noch der Kategorie II sinnvoll detektiert werden. Die Qualität der Detektion hängst stark von den Trainingsdaten ab. Die besten Resultate werden erwartungsgemäss bei der Detektion von Monumenten der Kategorie I erreicht: Bei einem tiefen Schwellenwert konnten über 96% der Grabhügel aufgefunden werden. Für die Grabhügel der Kategorie II bleiben die Detektionsraten relativ konstant, Monumente der Kategorie III werden de facto nicht detektiert. Ceteris paribus zeigt sich sehr deutlich der Einfluss des Schwellenwertes auf die Anzahl der falsch-positiven Detektionen. Die Einschränkung der Anzahl falsch-positiver Detektionen ist zunächst eine Abwägung zwischen gewünschtem Arbeitsaufwand (dem Aussortieren der Fehldetektionen) und der Vollständigkeit der Untersuchung. Auffällig ist, dass die Festlegung des Schwellenwertes stark von der Art und Anzahl der positiven Trainingsbilder abhängig ist. Werden ausschliesslich Monumente der Kategorie I und II

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

746

Caspari

Fig. 1. Automatische Detektion von Hügelstrukturen in IKONOS-2 Daten.

als Trainingssample verwendet, dann ist der Schwellenwert relativ robust. Werden Monumente der Kategorie III ins Trainingssample eingeschlossen, erhöht sich in erster Linie die Anzahl der Fehldetektionen. Es scheint daher ratsam zu sein, in erster Linie Grabhügel mit hohem Kontrast zum Hintergrund, wenig Bewuchs und geringer Formveränderung in die Trainingsdaten einzubeziehen. Teilweise zerstörte oder von Bewuchs überzogene Monumente entsprechen in etwa der Problematik von verdeckten Personen bei einer Zählung von Fussgängern in Videos und können ähnlich behandelt werden (dass heisst unklare Repräsentationen der Objektklasse werden aus den Trainingsdaten ausgeschlossen, die Detektion von verdeckten Objekten erfolgt mittels tieferer Schwellenwerte). Grundsätzlich gilt: Je grösser das Trainingsset, desto robuster der Detektionprozess.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Ein Inventar der Hügelgräber im Altaigebirge

747

Fig. 2. Optische Analyse der IKONOS-2 Daten in ArcGIS.

6. Diskussion Durch die Anwendung des Algorithmus konnten wir in den Testdaten jedesmal zuverlässig innert Sekunden die wichtigsten archäologischen Sites detektieren und damit einen Überblick über die Gebiete mit den höchsten Grabhügelkonzentrationen gewinnen (fig. 2). Für kleinräumige Untersuchungen, die eine möglichst vollständige Abdeckung des Untersuchungsgebietes zum Ziel haben, ist der Algorithmus nicht geeignet. Würde allerdings ein Survey eines Gebietes von mehreren tausend Quadratkilometern durchgeführt und das Hauptaugenmerk nicht auf die einzelnen Grabhügel, sondern auf Sites gelegt, wie im hier diskutierten Fall notwendig, dann ergäbe sich eine starke Verringerung des Aufwandes durch die automatische Detektion. Das gesetzte Ziel, ein Werkzeug für die effiziente, automatischen Detektion von Regionen mit hoher Grabhügeldichte zu kreieren, wurde somit erreicht.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

748

Caspari

7. Ausblick Da sich die eisenzeitlichen Kurgane im Altai grösstenteils zu Linien mit ungefährer nordsüdlicher Ausrichtung formieren und zudem nur in flachem Gelände (das heisst primär auf Flussterassen) zu finden sind, bestehen weitere Ansatzpunkte für eine effizientere Gestaltung des Verfahrens. Durch Einbezug eines digitalen Höhenmodells können alle Gebiete mit einer Steigung von mehr als 10% ausgeschlossen werden; dies verringert die zu bearbeitende Fläche, sowie die Anzahl der Fehldetektionen stark, ohne dabei das Resultat zu verschlechtern. Durch die Programmierung von Suchsektoren nach Süden und nach Norden könnten zudem Grabhügelnekropolen mit nord-südlicher Ausrichtung einwandfrei identifiziert werden. Der Einbezug dieser beiden Parameter würde zu einer weitgehend verlässlichen Detektion von eisenzeitlichen Sites im Altai führen. Eine Kombination mit den Forschungsresultaten zu Permafrostgebieten im russischen Altai könnte zudem die Suche nach Eisgräbern erheblich erleichtern (Goossens et al. 2005: 596–601). Solchermassen generierte Daten müssen das Rahmenwerk für die zukünftige Planung und Ausführung weiträumiger, grenzübergreifender Schutzprojekte bilden, damit die gefrorenen Schätze des Altaigebirges weiterhin erhalten bleiben.

Dank Der Autor dankt der Digital Globe Foundation für die Bereitstellung der ansonsten kostenpflichtigen IKONOS-2 Daten und der Universität Wuhan, insbesondere Prof. Dr. T. Balz, für die Unterstützung während des Forschungsaufenthaltes in China.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Ein Inventar der Hügelgräber im Altaigebirge

749

Bibliographie Bourgeois, J. et al., 2007. Saving the frozen Scythian tombs of the Altai Mountains (Central Asia). World Archaeology 39:3. London, 458 – 474. –

2010. Report on the Belgian-Russian Expedition in the Altay Mountains. Ghent.

Caspari, G. et al., 2014. Application of Hough Forests for the detection of grave mounds in high-resolution satellite imagery. Proceedings IGARSS 2014. Quebec, 906–909. Gall, J. et al., 2011. Hough Forests for Object Detection, Tracking, and Action Recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence. New York, 2188–2202. Gheyle, W., 2009. Highlands and Steppes – An Analysis of the Changing Archaeological Landscape of the Altay Mountains from the Eneolithic to the Ethnographic Period. Ghent. Goossens, R. et al., 2005. Estimation of the permafrost area in the Altai Mountains (Russia) in the framework of the preservation of the Frozen Tombs of the Altai Mountains. A. Marcal (ed.), Global Developments in Environmental Earth Observation from Space. Proceedings of the 25th Symposium of the European Association for Remote Sensing Laboratories. Rotterdam, 595–601. Gino Caspari, University of Hamburg, Institute of Archaeology, Hamburg, Germany, [email protected].

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9th ICAANE, Basel 2014, Vol. 3, 751–762

Fabrice De Backer

La construction d’un char de guerre néo-assyrien This paper displays the bases of the author’s attempts to built a life-size Neo-Assyrian war chariot. The sources and main components are investigated in order to target a deepened study of each component. The author also provides first hand results of his endeavouring experiments and references to the forthcoming monograph.

1. Introduction Du 9ème au 7ème siècle av. J.-C., l’iconographie assyrienne fournit une profusion de données sur la construction et la conduite des chars de guerre utilisés pour l’expansion du royaume (fig. 1). Celles-ci permettent de tracer la base des plans employés pour la réalisation de ces voitures, les premières lignes du manuel de l’utilisateur, et les premières pages du programme d’entraînement des combattants embarqués (figs. 2–3). Seuls les chars de guerre néo-assyriens représentés ou mentionnés entre le 9ème et le 7ème siècle av. J.-C. feront l’objet de cette étude, ce qui en exclut les représentations de trône roulant et les charrettes. Ces véhicules devaient répondre aux contraintes imposées par l’armée du Roi d’Assyrie à cette époque: la vitesse de dé- ou de re-montage pour les passages montagneux, la facilité de réparation ou de remplacement des pièces et la légèreté du poids pour la traversée des rivières. L’iconographie d’Assurnasirpal II et de Salmanazar III démontre clairement que les chars de guerre assyriens répondaient clairement à toutes ces nécessités essentielles. Les chars devaient pouvoir être remisés en pièces détachées, ou pas, dans les arsenaux, pendant la mauvaise saison, durant les trajets à bord des navires utilisés pour la visite de Sargon II et d’Assarhaddon à Chypre, ou ceux que Senachérib employa pour son débarquement en Élam.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

752

De Backer

Fig. 1. Le char de guerre d’Assurnasirpal II. Dessin de E . Dehenin (d’après Wallis-Budge 1914: pl. XIV, a).

Figs. 2–3. Schémas représentants un char de guerre assyrien du 9ème siècle av. J.-C. Dessins de l’auteur. Fig. 2. Vu du haut. Fig. 3. Vu de profil.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

La construction d’un char de guerre néo-assyrien

753

Fig. 4. L’échelle des roues du char royal assyrien durant les 8ème et 7ème siècles av. J.-C. Dessin de l’auteur (d’après Spruytte 1994: 43, fig. 8).

2. Sources Malgré les précautions d’usage à respecter pour toute étude de matériel essentiellement documenté par l’iconographie, les normes employées par les artistes assyriens pour la représentation du char royal permettent d’estimer les dimensions d’un char de guerre basique à l’échelle réelle (fig. 4). À l’heure actuelle, les sources écrites néo-assyriennes donnent plus d’informations sur l’organisation des spécialistes qui maintenaient ces véhicules de guerre ou sur le programme d’entraînement des attelages qui les tiraient, comme celui de Kikkuli qui fut écrit pendant l’Âge du Bronze Récent.

3. Le timon Le timon formait l’épine dorsale des chars de guerre néo-assyriens, situé entre la traction et le train de roulage fixé au châssis qui supportait le poids de l’équipage. L’iconographie contemporaine montre que cette longue pièce de bois subissait une déformation particulière qui permettait de la courber selon l’angle voulu, sans doute grâce à une exposition à l’eau bouillante, puis à son insertion dans un moule fait de poteaux de bois.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

754

De Backer

Sur l’extrémité distale, située le plus loin du châssis, l’objet comportait une série de tenons et de mortaises qui permettait d’y fixer un objet décoratif sur base cylindrique autour du mât, ainsi qu’un emblème divin à partir du 8ème siècle av. J.-C. Cette extrémité comportait aussi un emplacement réservé pour le joug auquel les chevaux de l’attelage étaient harnachés. Une voilure reliant l’habitacle et l’extrémité distale du timon renforçait la solidité du timon, sans doute fixée par un système similaire à une voilure de cuir, peut-être assemblée grâce à des olives et à des lanières de cuir, comme c’était le cas des machines de siège contemporaines. Cette voilure cache sans doute une deuxième partie du timon qui permettait de le rigidifier et de le solidifier, afin de prévenir tout risque de rupture pendant les mouvements brusques causés pendant les charges et les déplacements de combat. Ce principe permettait aussi de renforcer le timon des chars de guerre grecs au moyen d’un renfort et d’un chaînage pendant la période archaïque, comme de multiples exemples le montrent très visiblement (De Backer forthcoming). L’extrémité proximale du timon recevait sans doute une série de mortaises destinées à l’associer au châssis par des tenons verticaux. L’un des segments de cette extrémité pouvait aussi comporter quelques encoches nécessaires pour y attacher les renforts antérieurs qui renforçaient la solidité de la caisse et du timon. La première série des renforts antérieurs reliait les angles latéraux de la partie supérieure du cadre de la caisse au timon, tandis que la deuxième se situait à mi-hauteur de la face antérieure de l’habitacle. Sous le châssis, deux arbalétriers, fixés par quelques tenons dans les mortaises correspondantes du timon, pouvaient aussi renforcer la résistance du timon au poids de la caisse et équilibrer le poids de la masse embarquée.

4. Le joug Le joug représentait la jonction entre la force de traction et la masse tractée, ce qui en faisait l’emplacement qui devait être le plus solide et le plus souple. Cette pièce de bois, travaillée en ondulations, comportait un nombre d’espaces de tractage proportionnel au nombre d’animaux destinés à y être attelés: de trois à quatre entre le 9ème et le 7ème siècle av. J.-C. Une deuxième pièce, qui adoptait la forme d’un «Y» renversé et surmonté d’un anneau passe-guide, venait compléter le système y étant attachée par le biais de sangles de cuir sur la face postérieure du joug. Le choix de cet emplacement pour le rattachement des armatures en question réside dans l’avantage que l’endroit offre en termes de traction des sangles et de propulsion du joug à la fois, ce qui optimisait le résultat des efforts fournis par l’attelage pendant les déplacements du véhicule.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

La construction d’un char de guerre néo-assyrien

755

Les charrons façonnaient parfois les extrémités distales du joug en forme de protomé de serpents ou d’autres motifs décoratifs, selon le statut et le goût des membres de l’équipage et les époques.

5. Le châssis Le châssis se composait du train de roulage et de la caisse qui recevait l’habitacle. Le train de roulage comportait une pièce rectangulaire, aplatie et réalisée dans un bois extrêmement solide pour résister aux contraintes imposées par le poids de la tare, et par les mouvements de la force de traction. Pour des raisons de facilité et de rapidité de remplacement, le train de roulage comportait sans doute une série de mortaises et de tenons qui permettaient d’en changer les roues sans démonter l’entièreté du véhicule. Les axes fixes comportaient deux sabots qui ralentissaient l’usure du bas de caisse et de l’essieu pendant le roulement, et permettaient peut-être même de régler la géométrie des roues. L’extrémité proximale du châssis recevait sûrement une série de mortaises qui permettait de la fixer au timon par quelques tenons.

6. Le cadre Le cadre de l’habitacle devait être réalisé de la même manière que le plancher, ce qui en facilitait le remplacement par la production de panneaux interchangeables, et maintenu par une série de tenons, de mortaises et de saignées dans le châssis qui en assuraient la verticalité des parois. Les montants du plancher pouvaient être fixés au timon et au châssis par une série de tenons et de mortaises.

6.1. L’habitacle L’habitacle d’un char repose sur un cadre fabriqué à partir de montants dotés de tenons et de mortaises qui en assurent la solidarité et la solidité. Le plancher pouvait être réalisé en cannes de bois ou d’osier tressé, ce qui renforçait l’adhérence des pieds de l’équipage, nus ou pourvus de talonnières, à la caisse du véhicule pendant les mouvements, contrairement à un tapis ou à un treillis de cuir.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

756

De Backer

6.2. Le plancher Du 9ème au 7ème siècle av. J.-C. le plancher de certains chars de guerre comportait une mortaise supplémentaire destinée à recevoir le saurochtone des enseignes militaires que les équipages embarquaient (fig. 2).1 À ce jour, la répartition du nombre d’enseignes militaires au sein de la charrerie néo-assyrienne entre le 9ème et le 7ème siècle av. J.-C. reste vague et mériterait de plus amples recherches. Un ruban de tissu pouvait décorer ces emblèmes, comme cela apparaît clairement sur les exemplaires retrouvés en fouilles pour les modèles du 9ème siècle av. J.-C. et permettait aussi de différencier certaines unités des autres, fût-ce seulement selon la couleur du tissu employé.

6.3. La caisse L’iconographie montre également qu’une couverture de cuir pouvait recouvrir le cadre de l’habitacle, sans doute composée de segments dotés d’olives et de lanières qui en permettaient l’attachement et la solidarité. La tension de cette couverture dépendait sans doute d’une série de tenons fixés à l’intérieur de l’habitacle, sous le rebord formé par le sommet de la rambarde de la caisse, et sur la partie inférieure du châssis. L’arrière de la caisse devait sans doute comporter quatre poignées de préhension de cuir destinées à faciliter l’embarquement du personnel: deux sur la partie postérieure des montants verticaux extérieurs de l’habitacle, et deux autres pour l’équilibre des porteurs de boucliers pendant les déplacements, situées sur le sommet de ces mêmes montants. La face antérieure de l’habitacle comportait certainement aussi deux poignées de préhension au niveau du timon pour que les membres de l’équipage puissent se maintenir pendant les déplacements du véhicule.

6.4. Les marches d’accès L’accès à l’habitacle des véhicules pouvait être facilité par une ou deux marches situées à l’arrière de la caisse et représentées de profil dans l’iconographie contemporaine. Compte tenu de l’usure intense que ces éléments devaient subir, les charrons utilisaient certainement des éléments de cuir de différentes qualités pour façonner ces proto-étriers. Pour faciliter et accélérer l’embarquement de trois charristes, du 9ème au 8ème siècle av. J.-C., et de quatre personnes, durant le siècle suivant, ces appuis se répartissaient sur chaque flanc du véhicule et par nombre d’individus situés de chaque côté du timon. La présence de ces éléments réduisait également la pression exercée à chaque embarquement ou débarquement sur le timon qui supportait le plancher du char, et cela en réduisait la fragilisation.

1

Le saurochtone est un petit embout de métal façonné en pointe et destiné à ficher une hampe dans le sol afin de l’y maintenir en position verticale.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

La construction d’un char de guerre néo-assyrien

757

Fig. 5. Schéma représentant le mode de suspension des armes à la caisse du char de guerre assyrien durant le 9ème siècle av. J.-C. Dessin de l’auteur.

7. Les roues Les roues du char royal néo-assyrien firent l’objet d’un excellent article de J. Spruytte qui en fabriqua quelques répliques à échelle réduite (Spruytte 1994: 37–48). L’auteur y démontre que ces éléments essentiels aux chars de guerre se composaient d’une série de rayons formés d’entretoises de bois, articulées autour d’un moyeu oblong et solidifiés par une série de tenons en métal. Ceux-ci permettaient également d’assembler les segments arrondis des bandages de la roue, qui faisaient donc office de pneumatiques, protégés par des renforts de métal sur leur face extérieure. Une série de clous à large tête hémisphérique renforçait l’adhérence de la bande de roulement au sol et servait également à maintenir les bandages de métal en place.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

758

De Backer

8. L’équipement L’équipement d’un char de guerre néo-assyrien se déclinait en trois types de dotation associée à la proximité du danger que les équipages auraient eu à combattre: les armes de trait pour la longue distance, les armes d’hast pour la distance moyenne et les armes du corpsà-corps. Malgré le temps, peu de changements apparaissent pour la dotation des chars dans l’iconographie assyrienne entre le 9ème et le 7ème siècle av. J.-C. (fig. 5). L’un des chars de guerre découvert dans une tombe de Chypre et daté des 8ème –7ème siècle av. J.-C. contenait une grande rapière, qui apparaît souvent représenté comme l’arme de prédilection des guerriers assyriens entre le 9ème et le 7ème siècle av. J.-C. (Karageorghis 1965: 286, figs. 282–283).. Ce choix s’explique aisément par les motifs suivants: une longue et fine lame permet d’estoquer à distance raisonnable d’un ennemi plus ou moins revêtu d’armure, comme un porteur de pavois avec sa lance et dans le cas d’un corps à corps, de le garder à distance par de grands moulinets de taille. Cette dernière méthode permet également d’éloigner les cavaliers en menaçant la gorge et les jambes antérieures de leurs montures. Une rapière demande également une certaine dextérité ainsi qu’un entraînement constant et assez conséquent pour la conserver à un bon niveau d’efficacité. En plus de cette triple dotation, le véhicule emportait les boucliers des charristes, ce qui renforçait leur armure, leur permettait de fonctionner de manière pluridisciplinaire en tant qu’infanterie mécanisée, de cavalerie lourde, voire seulement de pouvoir se défendre en attendant les renforts si leur situation au combat devenait critique.2 L’extrême compétence guerrière des charristes se révèle par la présence de ces quatre carquois, de ces quatre masses de guerre et haches d’armes, de ces deux lances, ou quatre si les artistes assyriens les ont rabattues sur les reliefs, et de ces quatre boucliers.

8.1. La dotation primaire Les armes de trait et de jet formaient la dotation primaire des chars de guerre néo-assyriens. Au 9ème siècle av. J.-C., celle-ci se composait généralement de deux carquois croisés sur chaque flanc de la caisse, chacun comportant une quarantaine de saguettes3 ainsi qu’un arc composite. Les carquois pouvaient aisément et rapidement être remplacés pendant les batailles en détachant leurs baudriers des tenons auxquels ils étaient fixés à l’intérieur de 2

3

Dans le cadre de cette communication, l’infanterie mécanisée désigne une infanterie embarquée dans un véhicule, qui combat, se déplace et interagit autour de ses véhicules. La cavalerie lourde désigne, quant à elle, une force de frappe lourdement cuirassée, très rapide et dotée d’un grand potentiel offensif. Les saguettes, décochées par un arc, se composent d’un empennage, qui en assure la stabilité en vol, d’une hampe, qui forme le corps du missile, et d’une armature ou d’une flèche, qui constitue la tête du missile.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

La construction d’un char de guerre néo-assyrien

759

Fig. 6. Dessin représentant Halzi utilisant le point faible des chars ennemis pour en tuer les équipages. Dessin de l’auteur (d’après Konaci 2009: 170, fig. 1).

l’habitacle. Pour éviter que ceux-ci ne gênent l’équipage dans ces mouvements, ils étaient sans doute placés sous un rebord vertical situé au sommet du cadre de la caisse. Entre le 8ème et le 7ème siècle av. J.-C., les carquois sont fixés verticalement sur l’avant de la caisse du véhicule, grâce à une courroie de cuir que l’équipage fixait sur la face externe des flancs de la caisse. Néanmoins, aucune arme de la dotation primaire n’est représentée sur l’iconographie contemporaine.

8.2. La dotation secondaire Les armes d’hast, comme les lances de la dotation secondaire, étaient glissées dans un étui placé au-dessus du premier carquois, placé contre la caisse, et formaient une saillie vers l’arrière de l’habitacle, ce qui permettait de les utiliser pour repousser les lions et les cavaliers qui auraient tenté d’éliminer l’équipage en attaquant sur ce point. Dans certains cas, la bague de soutien des lances pouvait recevoir une décoration particulière plus raffinée, comme la forme d’un protomé anthropomorphe sur le char d’Assurnasirpal II. Le nombre de ces armes suivait sans doute celui des membres de l’équipage, même si les lances étaient rangées par deux sur chaque flanc de la caisse. Les rubans et/ou les houpettes qui ornaient ces armes devaient sans doute également pouvoir servir à distinguer certaines unités de l’armée assyrienne, fût-ce seulement par leur couleur distinctive. À partir du 8ème siècle av.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

760

De Backer

Fig. 7. Schéma représentant le mode probable de la suspension des boucliers de l’équipage au dos du char de guerre assyrien durant le 9ème siècle av. J.-C. Dessin de l’auteur.

J.-C., les armes de la dotation secondaire apparaissent en saillie à l’arrière de la caisse du véhicule, le fer tourné vers le bas.

8.3. La dotation tertiaire La dotation tertiaire de ces véhicules constituait le dernier recours de l’équipage, souvent composée d’une hache de guerre et d’une masse d’armes qui servaient surtout pour le corpsà-corps quand le char était immobilisé ou pour la poursuite d’un grand nombre de fantassins en déroute.

8.4. Les boucliers Les boucliers emmenés par les charristes forment une dotation particulière, car ils leur permettaient ce combattre comme infanterie débarquée ou de protéger l’aurige et le tireur du véhicule (De Backer 2010: 1–25). Les boucliers de ces deux derniers combattants étaient suspendus à l’arrière de l’habitacle, afin d’y former une chicane qui limitait la menace d’agressions sur ce point, tout en constituant une réserve d’arme pour le ou les porteurs de boucliers embarqués, selon les époques (fig. 6). Ces armes devaient pouvoir être suspendues à un système de montants amovibles dans le plancher et rabattables sur les flancs intérieurs de la caisse, maintenus en place par des lanières de cuir ou des chaînes, beaucoup plus solides (fig. 7). Une série de chaînes aurait certainement permis de limiter les risques de projection vers l’arrière pour l’équipage pendant les cahots d’une charge.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

La construction d’un char de guerre néo-assyrien

761

Un char de guerre pouvait certainement aussi emporter un pavois en osier pendant certaines opérations, comme certaines sources peuvent le laisser penser (De Backer 2013: 69–78).

9. Conclusion L’expérimentation de terrain supporte aisément l’étude croisée des sources visuelles, matérielles et textuelles pour l’analyse des chars de combat néo-assyriens. Déjà à l’époque assyrienne, ce véhicule représentait l’aboutissement et la poursuite de nombreuses recherches entreprises sur les véhicules à roues et l’entraînement des équipages depuis l’Âge du Bronze Ancien. La poursuite de ces recherches au 21ème siècle repose également, même beaucoup sur l’analyse des petites découvertes, comme les bouterolles d’essieu ou les clous de bande de roulement anciens. Les comparaisons ethnographiques avec des véhicules contemporains et utilisés à des fins similaires dans des régions voisines montre de même la grande importance de l’expérimentation et de la course à l’armement représentée par ce type de machines. L’auteur prépare en ce moment une étude complète du char de guerre néo-assyrien, incluant de nombreux plans, de multiples mesures, ainsi qu’un manuel de’utilisateur et de conduite, pour laquelle la maison d’édition Brill lui a montré le plus grand intérêt.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

762

De Backer

Bibliographie Albenda, P., 1986. Le palais de Sargon d’Assyrie. «Synthèse» no 22, Recherche sur les civilisations. Paris. Barnett, R. D./Falkner M., 1962. The Sculptures of Aššur-Nasir-Apli II (883–859 B. C.), Tiglath-Pileser III (745–727 B.C.), Esarhaddon (682–669 B.C.) from the Central and South-West Palaces at Nimrud, Trustees of the British Museum. Londres. Barnett, R. D./Bleibtreu, E. et al., 1998. The Sculptures from the South-West Palace of Sennacherib at Nineveh. London. Barnett, R. D., 1976. Sculptures from the North Palace of Ashurbanipal at Nineveh (668–627 B.C.), Londres. De Backer, F., 2009–2010. Evolution of War Chariot Tactics in the Ancient Near East. Ugarit Forschungen 41, 29–46. –

2010. Neo-Assyrian Siege – Redoubts: Some Issues. Historiae 7, 1–25.

–

2013. Notes on the Neo-Assyrian Siege-Shield and Chariot. J. Sanmartin (éd.), Proceedings of the 56th Rencontre Assyriologique Internationale, Barcelona, 26th – 30th July 2010, Winona Lake, 69–78.

–

forthcoming. Assyrians in Edeli: Private Life and State Activities. Proceedings of the 58th Rencontre Assyriologique Internationale, Leyde, 16th–20th July 2012.

–

forthcoming. La construction d’un char de guerre néo-assyrien.

Karageorghis, V., 1965. Chronique des fouilles et découvertes archéologiques à Chypre en 1964. Bulletin de Recherche Hellénique 89.1, 231–300. King, L., 1915. Bronze Reliefs from the Gates of Shalmaneser King of Assyria B.C. 860–825. Londres. Konaçki, E., 2009. Military and Militia in the Urartean State. Ancient West and East 8, 169–201. Littauer, M./Crouwel, J., 2001. Selected Writings on Chariots and Other Early Vehicles, Riding and Harness. Culture and History of the Ancient Near East 6. Leyde. Nyland, A., 2009. The Kikkuli Method of Horse Training, Revised Edition. Sydney. Spruytte, J., 1994. La roue du char royal assyrien. Revue d’Assyriologie 88, 37–48. Wallis-Budge, E., 1914. Reign of Ashur-Nasir-Pal, 885–860 B.C. Londres.

Dr. Fabrice De Backer, Chercheur Associé, Centre Aloïs Musil pour l’Archéologie Orientale, Université de Vienne (Autriche).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 763–768

Rita Gautschy

Astronomical Data and their Potential for Chronological Purposes To derive absolute dates of pharaohs and Mesopotamian kings records of astronomical observations have been frequently used. For the utilization of astronomical data and their reliability different rules of thumb hold in Egypt and Mesopotamia due to the varying nature of the solar Egyptian and the lunisolar Mesopotamian calendar. I will address the possibilities and limits of astronomical data for chronological purposes in Egypt and discuss the agreements and disagreements with C14 determinations.

1. Egypt in the 2nd Millennium BC Already in the early parts of the 3rd millennium BC a solar calendar with 365 days was in use in Egypt. The special feature of this calendar is that never any intercalation was attempted until Ptolemaic times although being short a quarter of a day each year in comparison to astronomical circumstances. Obviously Egyptians were not disturbed by the fact that their seasons – named by given factors like inundation or sowing – did not correspond to reality anymore at all after about 500 years. However, such a calendar with 365 days and no intercalations is the perfect tool for astronomical calculations where distances between two dates are involved. From Middle Kingdom times a set of 42 lunar feast data is preserved in the temple archive of Illahun (Luft 1992). Concerning the New Kingdom about 10 lunar data are known (Gautschy 2014). In a few cases it is preserved on which lunar day a certain feast was celebrated, e.g. for the feast mnḫt Pap. Berlin P. 10248 rt II tells us indirectly but unambiguously that it happened on a fourth lunar day (Luft 1992: 110–111). But for most of the feasts the corresponding lunar day has to be calculated by means of distances to other feasts. Thus, a data reduction is necessary to extract the needed information for a comparison with calculations. The uncertainty of the resulting attribution of a lunar feast to a specific lunar day amounts to one day (see Krauss 2006: 424–427, Luft 1992: 202 and Gautschy 2011a: 3

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

764

Gautschy

for different assignments of the Wagi-feast). Besides these lunar data two complete dates of a heliacal rising of the star Sirius are known from the 2nd millennium BC. Unfortunately, the reference point of these observations or predictions are unknown.

2. Mesopotamia In Mesopotamia a lunar calendar was used with a new month starting with first visibility of the lunar crescent after new moon. In order to keep such a calendar in line with the solar year, an intercalary month approximately every third year is needed. However, although a synchronization of the lunar and the solar year was attempted for, intercalary months were fitted in on a rather irregular basis. Only very late in history, namely from the 7th century BC onwards, a regular intercalary pattern was established. This crude fact leaves us with the uncomfortable situation that although from the day number of a Mesopotamian calendar date it is immediately clear which day of a lunar month it concerns, nonetheless it is not clear to which month in our Gregorian calendar it belongs during the 2nd millennium BC. The uncertainty can amount to several months and one can never be sure how many months between two Mesopotamian calendar dates elapsed as long as the sequence of intercalary months cannot be reconstructed from other documents like e.g. economic texts. The recorded sequence of intercalations during the reign of Hammurapi nicely illustrates the problem (Huber 1982: 57). The most famous set yet is the so-called Venus tablet of Ammiṣaduqa containing observed as well as calculated dates of heliacal risings and settings of the planet Venus. Together with the available month length data and the known intercalary pattern during the reigns of kings Ammiṣaduqa and Ammiditana these Venus data provide a means of dating the Hammurapi dynasty. Since astronomical data are cyclic events there always exist multiple possibilities for absolute dating. Based on the Venus data three different chronologies with a number of sub-variants have been proposed during the last decades: High Chronology (Huber 2000), Middle Chronology (Veenhof 2001: 311) and Low Chronology (Mebert 2010).

3. Egyptian Middle Kingdom All lunar feast data are preserved on papyri from Illahun in the Fayum. The city was founded by pharaoh Senusret II and occupied for about 80 years, in later times it was reused (Quirke 2005). Thus, all papyri must date to the reigns of Senusret II, Senusret III, Amenemhet III and maybe Amenemhet IV. Distance calculations between lunar dates on papyri allow to assign most papyri definitely to either Senusret III or Amenemhet III (Luft 1992). The

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Astronomical Data and their Potential for Chronological Purposes

765

temple diary on papyrus Pap. Berlin P. 10012 A rt II documents the arrival of a letter on III Peret 25 containing the prediction of a heliacal rising of the star Sirius in year 7 of Senusret III on IV Peret 16 (Luft 1992: 54–57; Gautschy 2011b: 124–125). A second scrap of the same temple diary (Pap. Berlin P. 10012 B rt) proves the delivery of supplies for the feast of the heliacal rising of Sirius on IV Peret 17 (Luft 1992: 57–58). Assuming different reference points for this prediction of the heliacal rising of Sirius and accounting for the uncertain day number (16 or 17), the following time slots for year 7 of Senusret III remain: • • •

Memphis: 1881–1864 BC Thebes: 1863–1846 BC Elephantine: 1856–1839 BC

Thus, year 7 of Senusret III has to be allocated between 1881 BC and 1839 BC according to the documented Sirius phenomenon. With the help of the about 40 lunar feast data discrete solutions can be picked out. Finally, two possible chronologies remain, the High Chronology with year 7 of Senusret III in 1866 BC and a reference point in Memphis (Luft 1992; Gautschy 2011a, 16–18; Huber 2011: 211–221) and the Low Chronology with year 7 of Senusret III in 1841 BC and a reference point in Elephantine (Krauss 2006: 424–427). C14 determinations may help to further narrow down the possibilities. Three papyri from Illahun which can be astronomically dated (Pap. Berlin P. 10092, 10009 and 10248) have also be dated by means of C14 determinations (Bronk Ramsey et al. 2010: Online Supporting Material). For the High Chronology year 1 of Senusret III would fall in 1872 BC and year 1 of his successor Amenemhet III in 1842 BC based on the astronomical data, the corresponding years for the Low Chronology are 1847 BC and 1817 BC. C14 data suggest a placement of year 1 of Senusret III between 1889 and 1836 BC in the 95% confidence range, and of year 1 of Amenemhet III between 1851 and 1798 BC. Thus, C14 and astronomical dating agree well with each other and unfortunately none of the two possible astronomical chronological options can be dismissed currently. However, it has to be stressed that at present the assumed reign length of Senusret III is still a matter of debate. His highest securely attested date is year 39 and papyrus Turin mentions 30+ years. But Egyptologists agreed on that only 19 years should be counted for this king, presuming a very long coregency of at least 20 years with his successor. Items cited opting for a limitation include the lack of any dated monument after his 19th regnal year, the evidence of the so-called ANOC group 1 stelae where a transition from a year 19 to a year 1 is preserved (Simpson 1974: 16–17), and last but not least papyrus Pap. Berlin P. 10055 in the Illahun archive where a year 19 is followed by a year 1. However, the lunar data suggest a reign length of 30 years of Senusret III, still supporting a comparably long phase of co-regency (Gautschy 2011a: 4–18). The possible chronological options presented in this paper assume a reign length of

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

766

Gautschy

30 years of Senusret III while the C14 determinations were modeled with the assumption that he reigned only 19 years.

4. Egyptian New Kingdom Merely two lunar dates exist where the text explicitly states that the reported day was a first lunar day, namely the so-called Piramesses date of Ramesses II (Jansen 1961: 12 and 33) and the reported date of the battle of Megiddo of Thutmose III (Urkunden IV: 657.2). Alongside there exist a couple of lunar feast data referring to the Theban Feast-of-the-Valley. Sirius data from the New Kingdom are problematic: one from the time of Thutmose III is incomplete, lacking the year of the king, and the other one stems from the Papyrus Ebers. The latter belongs to year 9 of Amenhotep I. However, it is highly suspicious that all months on Papyrus Ebers begin on day 9, hinting at that we are dealing with a schematic calendar with 30 days for each month instead of with real observations. Additionally, the five epagomenal days are not considered. Thus, it seems safest not to put too much weight on this Sothic date. Assuming that it was indeed observationally determined on III Shemu 9, Amenhotep’s 9th regnal year would fall between 1550 and 1537 BC if Memphis was the reference point, or between 1532 and 1515 BC if Thebes was the reference point. In addition to these Egyptian astronomical data, an allusion to a solar eclipse is preserved in a fragmentary Hittite text of a prayer of Muršili II. Besides these astronomical data documented direct synchronisms – mainly from the Amarna period and from the time of Ramesses II – enable to link information from different regions and thus to bridge gaps in local chronologies (Gautschy 2014). For the epoch of the Egyptian New Kingdom no king list with regnal years of pharaohs is preserved. Nevertheless, due to dated wine jars, ostraca, documents and the admittedly partly corrupted information from Manetho the series of successive kings and their corresponding regnal years can be fairly well established (Aston 2014; von Beckerath 1994: 35–63). Questionable reign lengths mainly concern Horemheb, Seti I and Semenkhkare. Here, 15 years are adopted for Horemheb, 9 for Seti I and 3 for Semenkhkare together with Neferneferuaten (Aston 2014: 288–291). Astronomical data, like the archaeological evidence, are in good agreement with a short reign length of only 15 years for Horemheb contrary to the 27 years usually adopted in standard chronology (e.g. Shaw 2000: 481). I have recently presented the three possible chronological options for the New Kingdom listed in Table 1 taking into account astronomical data from Egypt and Mesopotamia and documented synchronisms (Gautschy 2014). Once again different assumptions about reign lengths had been adopted for the modeling concerning the C14 determinations (Bronk Ramsey et al. 2010: Online Supporting Material). The influence of the reign length on the modeled C14 dates manifests itself in the fact that for none of the three chronological options there is perfect

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Astronomical Data and their Potential for Chronological Purposes

767

agreement between astronomical dating and C14 determinations for all kings. However, the deviations are very small, amounting only a few years. Hence, currently none of the three astronomically determined chronological options of the New Kingdom can be excluded. It should be mentioned that these chronological options have been established under the assumption that our current interpretation of the Assyrian king list as a sequence of succeeding kings is correct with the implication that year 1 of Ramesses I cannot be lower than 1279 BC. Recently, doubts have been raised whether this treatment of the Assyrian king list is the correct approach (Dodson 2012: 181–189). pharaoh

P1

P2

P3

C14

1 Ahmose

1539 BC

1550 BC

1564 BC

−

1 Thutmose III

1468 BC

1479 BC

1493 BC

1498–1474 BC

1 Amenhotep II

1414 BC

1425 BC

1439 BC

1445–1423 BC

1 Amenhotep III

1378 BC

1378 BC

1403 BC

1408–1386 BC

1 Akhenaten

1339 BC

1339 BC

1364 BC

1370–1348 BC

1 Tutankhamun

1319 BC

1319 BC

1344 BC

1353–1331 BC

1 Ramesses II

1279 BC

1279 BC

1304 BC

1297–1273 BC

1 Twosre

1192 BC

1192 BC

1217 BC

−

1 Ramesses III

1187 BC

1187 BC

1212 BC

−

1 Ramesses VI

1145 BC

1145 BC

1170 BC

−

1 Ramesses XI

1106 BC

1106 BC

1131 BC

−

Table 1: Astronomically determined chronological options for the Egyptian New Kingdom and C14 determinations (95% range). Where no C14 dates are given no material was available for C14 dating.

5. Summary C14 and astronomical determinations agree well for the Egyptian Middle Kingdom and the New Kingdom. Two chronological options of the Middle Kingdom and three of the New Kingdom seem to be feasible. Unfortunately the C14 measurements currently cannot help to eliminate the one or the other astronomically determined option. It is necessary to model the C14 dates with the same underlying assumptions about reign lengths of kings in order to being able to narrow down the possibilities. This is planned for the future.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

768

Gautschy

Bibliography Aston, D., 2014. Radiocarbon, wine jars and New Kingdom Chronology. Ägypten & Levante XXII/ XXIII, 285–311. Bronk Ramsey, C./Dee, M. W./Rowland, J. M./Higham, T. F. G./Harris, S. A./Brock, F. /Quiles, A./Wild, E. M./ Marcus, E. S./Shortland, A. J., 2010. Radiocarbon-Based Chronology of Dynastic Egypt. Science 328, 1554–1557. Online Supporting Material: www.sciencemag.org/cgi/content/full/328/5985/1554/DC1. Dodson, A., 2012. Afterglow of Empire. Egypt from the Fall of the New Kingdom to the Saite Renaissance. Cairo. Gautschy, R., 2011a. Monddaten aus dem Archiv von Illahun: Chronologie des Mittleren Reiches. ZÄS 138/1, 1–19. –

2011b. Der Stern Sirius im Alten Ägypten. ZÄS 138/2, 116–131.

–

2014. A reassessment of the absolute chronology of the Egyptian New Kingdom and its ‛brotherly’ countries. Ägypten & Levante XXIV, 143–160.

Huber, P. J./Sachs, A./Stol, M./Whiting, R. M./Leichty, E./Walker, C. B. F./van Driel, G., 1982. Astronomical Dating of Babylon I and Ur III. Occasional Papers on the Near East 1.4, Malibu. Huber, P. J., 2000. Astronomy and Ancient Chronology. Akkadica 119–120, 159–176. –

2011. The Astronomical Basis of Egyptian Chronology of the Second Millennium BC. JEgH 4, 172–227.

Jansen J. J., 1961. Two Ancient Egyptian Ship’s Logs. Leiden. Krauss, R., 2006. Lunar Dates. E. Hornung et al. (eds.), Handbook of Egyptian Chronology. Leiden, 395–431. Luft, U., 1992. Die chronologische Fixierung des ägyptischen Mittleren Reiches nach dem Tempelarchiv von Illahun. Vienna. Mebert, J., 2010. Die Venustafeln des Ammi-ṣaduqa und ihre Bedeutung fur die astronomische Datierung der altbabylonischen Zeit. Archiv für Orientforschung Beiheft 31, Vienna. Quirke, S. G. J., 2005. A town in Egypt 1800 BC, and the history of its landscape. London. Shaw, I. (ed.), 2000. The Oxford History of Ancient Egypt. Oxford. Simpson, W. K., 1974. The Terrace of the Great God at Abydos: The Offering Chapels of Dynasties 12 and 13. New Haven. Veenhof, K. R., 2001. Geschichte des Alten Orients bis zur Zeit Alexanders des Grossen. Göttingen. von Beckerath, J., 1994. Chronologie des ägyptischen Neuen Reiches. Hildesheim.

Rita Gautschy, University of Basel.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 769–779

Krzysztof Hipp

Šamšī-Adad V’s Campaigns into the Zagros Revisited This paper aims to take a closer look at the military campaigns of Šamšī-Adad V onto the Iranian Plateau, which so far has not been a subject of a broader analysis. I understand that in some cases the reconstruction remains hypothetical and open to discussion. The reign of Šamšī-Adad was the last attempt to contain the Urartian expansion to the south of Urmia Lake and to regain control over the northern Zagros. L. D. Levine (1977: 178 n. 5) has rightly pointed out that the campaigns of Šamšī-Adad in this area are – «very confused and seem to follow no geographical order, but the places mentioned span all of Iranian Kurdistan». There are not many royal inscriptions of Šamšī-Adad. The basic, relevant source is the Calah Stele (Grayson 1996: A.0.103.1). According to the king’s annals, the Assyrians campaigned three times in the Iranian highlands. The absolute chronology of the campaigns is problematic, because Šamšī-Adad neither applied the practice of dating successive years with the names of the limmu nor continued his father’s tradition of using the term «palû». Therefore, we must only correlate the events described in the annals with entries in the Eponym Chronicle (Millard 1994: 57). From 822 to 820 B.C.E. a revolt instigated by Aššūrda’’in-aplu is reported and not until 819 the Chronicle mentions a campaign against the land of Mannea. This is the only unquestionable entry which refers to Šamšī-Adad’s activities in north-western Iran. The next year lists an unidentifiable toponym - […]šumme. Therefore, the campaign of 819 is likely to be the last one conducted in the east. A. Fuchs (2012: 136) accepts such a correlation. He claims that the two earlier campaigns took place during the revolt. However, K. Radner (2003: 42) points to a fragmentary preserved entry from 820. She claims that the Eponym Chronicle B4 reports the name of the Median city Sikris known from the reigns of Tiglath-pileser III and Sargon II. She reconstructs the name by adding the cuneiform –ik- in front of the two preserved signs ri-is. But this sign is illegible, so I dare to say that it is an over-interpretation. Besides, K. Radner combines two entries from 820 and 819 and as a result, she dates the third campaign of Šamšī-Adad to 820 and/or 819. If her reconstruction is accepted, then we will have the succession of the events reordered. Finally,

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

770

Hipp

I would like to refer to the dating proposed by S. C. Brown (1980: 42–47). He correlates the first campaign with the year 819. The next one is dated to the year 818, and the last to 815. Only the entry from 819 is unquestionably connected with the East. The year 815 refers to the city of Zarāte. The Prism of Sargon II from Calah (Gadd 1954: 186–187) mentions Zarāti, a Babylonian city, so we can assume that Šamšī-Adad campaigned in Babylonia in that year (Brinkmann 1968: 208). Let us now proceed to the first campaign. The Calah Stele does not describe its successive stages (A.0.103.1: i 53b–ii 16a). The text limits itself to define the frontiers of Assyria’s influence in the East. It does not necessarily mean that Šamšī-Adad did not campaign in the Zagros, because he received tribute from local rulers of the lands of Nairi. They are not specified by name, gentilicium or location of their realms. Nairi itself poses a problem. In this context it may denote: 1. a generic geographical term; 2. a more specific toponym, a synonym of the land of Ḫubuškia; 3. a synonym of the territory under Urarṭian control. Since the king boasts himself: «At that time I spread over the entire land Nairi like a net» it suggests a broader geographical territory even more so because Paddira is part of the land Nairi and defines the eastern frontier of the Assyrian sphere of influence, as opposed to Kār-Shalmaneser in the West. The location of Paddira/Šurdira/u is also controversial. It merits some comments here. Paddira was the capital of Allabria. During the 30th regnal year of Shalmaneser III, its ruler, Arta-sari, surrendered to the Assyrians and paid tribute (Grayson 1996: A.0.102.14-171). In all likelihood he is the same person named Arta-sirari from Šamšī-Adad’s 3rd campaign. R. Heidari (2010: 147–150) has suggested that Rabat Tepe, an archaeological site near modern Sardašt, can be identified with Paddira. He discovered a few inscribed bricks in situ. One of them is signed with the name Atâ/Adâ. R. Heidari claims that Adâ might have been a ruler of Paddira in the times of Sargon II. This is a misinterpretation. Adâ was not a ruler of Allabria, but of the small land of Šurda/Patta (Fuchs 1998: 20). According to the Najafehabad stele Šurda was situated between Karalla and Nikkur, while the Assyrian sources describing the 16th campaign of Shalmaneser III and the 8th campaign of Sargon II locate Allabria between Parsua and Mannea (Levine 1972: 38 – 33’; Grayson 1996: A.0.102.6 iii 58–iv 6; Mayer 1983: 70–71 37’–38’). If we accept the location of Parsua in the vicinity of Sanadaj, the territory of Allabria stretched from the area to the north-west of this city and to the north-east of the Lake Zeribor (Reade 1978: 139–140). Therefore, the identification of Rabat Tepe with Paddira should be rejected.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Šamšī-Adad V’s Campaigns into the Zagros Revisited

771

The next toponym defining the boundary of the Assyrian influence was the city of Aridu, the capital of the Assyrian province of Kirruri/Ḫabruri. This province was in modern Dašt-i-Harir (Liverani 1992: 22; Zadok 2002: 87). Ḫabruri was situated along an important northern road leading from Erbil to the Ušnu-Solduz valley on the south-eastern shore of Lake Urmia. Muṣaṣir/Ardini located near Sidikan was along the road (Radner 2012: 248–251). It was separated by Mount Andaruttu (Baradost chain) and the passes of Simesi (Dašt-i-Diana). If we compare the route of the 31st regnal year campaign of Shalmaneser III (Russell 1984: 195; Grayson 1996: A.0.102.14 174b–190) to the range of Assyrian influence in the region under Šamšī-Adad, we can see that the king did not control the northern road after the revolt of his brother. During the 31st regnal year of Shalmaneser III the Assyrians campaigned against cities in the land of Muṣaṣir and against Gilzānu. After that campaign the latter definitely disappears from the sources. Šamšī-Adad’s control was confined only to Ḫabruri. This dramatic change was undoubtedly caused by the Urarṭian expansion in the region. Its first stage was the annexation of the Ušnu-Solduz valley and control of Muṣaṣir. The Qalatgāh inscription (Salvini 2008: A 3–10) attests to the firm presence of Urarṭians in the south-western corner of the Lake Urmia, while the Kelišin stele (Salvini 2008: A 3–11) refers to a cult visit of Išpuini and Menua in the sanctuary of the god Ḫaldi. I think that Urarṭu may have subdued Muṣaṣir around that time, since as I mentioned, Aridu is the farthest point controlled by Assyrians along the route just after the revolt. Besides, M. Salvini (2001: 350–351) pointed out that the mention in the list of Meher Kapısı (Salvini 2008: A 3-1) of such deities as Adaruta (Mount Andaruttu), Šebitu or Artu’arasau, which are associated with the Urmia region, attests to the early stage of the expansion. The second campaign was led by the eunuch Mutaqqin-Aššūr (A.0.103.1: ii 16b – 34a). The starting point is the land not mentioned by name of a certain Šarṣina, son of Meqdiara. The latter is believed to be the same person as Nikdēra/Nikdiara who appears during the 4th campaign of Shalmaneser III (Grayson 1996: A.0.102.6 ii 10–15). Nikdēra was the ruler of the land Idu. It was part of Inner Zamua at that time. Idu was located along the Lower Zab in front of the Rania plain, which is confirmed by the recent excavations at Satu Qala (van Soldt et al. 2013). Therefore, I think that the plain of Rania may have been identified with Inner Zamua. It is also reasonable to suggest that the Assyrians proceeded along that river through the Rania and Pizhder plains during this campaign. This reconstruction is hypothetical, but it makes sense. After crossing the Zagros Mountains, probably through one of the passes of the chaȋne magistrale, Mutaqqin-Aššūr directed towards the cities of Ušpina. Ušpina is identified with Išpuini (Arutyunyan 1970: 138; Salvini 1995: 38), and his cities may have been within the boundaries of Urarṭu itself or just invaded and controlled by the Urarṭians. Anyway, the Assyrians probably took a route northwards into the Ušnu-Solduz valley. The army may

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

772

Hipp

Fig. 1. The second campaign against Nairi (dotted line – frontier of the Urarṭian expansion; white arrows – hypothetical route of the 2nd campaign).

have been advancing in the narrow valley along the upper reaches of Lower Zab, just to the east of the chaȋne magistrale. In the course of the campaign the Assyrians boast that they reached the Sea of the West. It is a bit confusing in this context, because the Assyrians used this term for the Mediterranean. It seems to be a scribal error, but I. M. Diakonoff (1985: 66) claimed that it might be indentified with the Caspian Sea. He argued that until the 4th century B.C.E. both the Caspian and Black Sea had been perceived as part of the Mediterranean. It is doubtful, however, that during this campaign the Assyrians reached so far to the East. They campaigned against the Urarṭians and we can suggest that the military actions took place in the Urmia Lake basin. Sunbu is the last toponym on the way back to Assyria. It was presumably part of Inner Zamua (Zadok 2002: 134). If we separate Zamua and Inner Zamua at least for the 9th century (during the reign of Sargon II the two entities merged), it is reasonable to locate this land somewhere in the Pizhder or in the Lower Zab valley in the vicinity of Šardašt. Note that one fragmentary preserved piece of the annals, which A. K. Grayson (1996: A.0.102.1002) dates to the reign of Shalmaneser III, while R. Zadok (2002: 134) accepts the possibility of dating it to the times of Adad-nārārī III, mentions Sunbu followed by Ḫubuškia. It may suggest that both toponyms are contiguous provided it is an itinerary of the campaign. The proposed route of the second campaign is presented on the map (fig. 1). Politically, the second campaign may be considered as an Assyrian attempt to regain control over the more southerly road leading along the Lower Zab into the Zagros by trying to stop Išpuini’s penetration towards the south along the chaȋne magistrale.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Šamšī-Adad V’s Campaigns into the Zagros Revisited

773

Toponym

Proposed location

Mount Kullar

Kolara Dagh

Ḫubuškia

Pirānšahr

Parsua

Sanandaj area

Mannea

Rūd-e-Tātā’ū and Rūd-e-Zarineh

Mēsu

in the vicinity of modern Takāb

Gizilbunda

Zanjān-Čāy valley

Sibara

between Zanjan and Qazvin ?

Sagbita = Bīt Sagbat

Hamadān

Araziaš

Harsin area

Fig. 2. The hypothetical itinerary of the 3rd campaign.

The last campaign is well documented (A.0.103.1: ii 34b – iii 70a). After crossing the River Zab and the pass of Mount Kullar, the Assyrians reached the land Nairi, where Šamšī-Adad received tribute from the following locals rulers: Dādî of Ḫubuškia, already mentioned Šarṣina and from people of Sunbu, Mannea, Parsua and Taurla. The next stage was the land of Mēsu. From there, they went to Gizilbunda. Its king Pirišāti was captured. At this stage of the campaign Šamšī-Adad encountered other local rulers whose names and origins are reported on the Stele, namely – Titamaška from Sassiašu and Kiara from the city of Karsibuta. They both escaped to the fortress of Uraš, which was eventually conquered. This stage of the campaign ended with the receipt of tribute from Bēl-āli, the ruler of the city of Ṣibara, and with setting up a monumental statue with a commemorative inscription praising Šamšī-Adad’s deeds. Here, it is noteworthy to mention that some scholars claim that Bēl-āli is a proper name, while others tend to see it as a title of the ruler, which was used erroneously (Lanfranchi 2003: 93–94). The next stage is the territory occupied by the Medes. For the first time the Assyrians encounter a Median king. Šamšī-Adad defeated a certain Ḫanaṣiruka, who ruled from his royal city of Sagbita. The etymology of his name is uncertain. After crossing the mountain of mūṣu-stone, another adversary, Munirsuarta, the ruler of the city of Araziaš, is reported. In the end, the annals list 28 local rulers who paid tribute to the king. Some of them bear Iranian names, but the etymology of the others is unknown. Let me now proceed to the reconstruction of the route. Its itinerary is presented on figure 2. The location of Mount Kullar depends on the position of Inner Zamua. I. N. Medvedskaya (2000: 429–443) convincingly separates Inner Zamua from Zamua and as a consequence she rejects the identification of Kullar with Bazian Dagh (Levine 1973: 17–18). During the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

774

Hipp

Fig. 3. The First Stage of the 3rd campaign (wide white lines - route of the campaign; dotted white lines – roads; dark triangles – major towns).

8th palû of Sargon II the Assyrian army went to the land of Sumbi, which I propose to identify with the Pizhder plain or the region around Sardašt. If so, we can accept J. Kinnier Wilson’s (1962: 111) identification of Mount Kullar with modern Kolara Dagh. Note that it is quite possible that Argišti I campaigned in the vicinity of this mountain since the Ḫorḫor Chronicle (Arutyunyan 2001: KUKN 173 III; Salvini 2008: A 8-3 III) mentions under his 5th regnal year the toponym Kulaše. I. Diakonoff and S. M. Kashkai (1981: 49) suggested the identification with Kullar. Interestingly, during that campaign the Urarṭian king encounters a certain Dadani, whose name bears a striking resemblance to the name of a Ḫubuškean ruler. In the land of Nairi the Assyrians received horses from the already mentioned Šarṣina of Idu, Dādî of Ḫubuškia, as well as from people of Mannea, Parsua and Taurla. Although it is not stated that the Assyrians traversed all these lands, their location is important, because they show how Šamšī-Adad wanted to secure the political environment in the region by at least neutralizing the lands along his flanks before going deeper onto the Iranian Plateau. The location of Ḫubuškia is still a subject of debate depending on the reconstruction of the VIII palû of Sargon II. Some scholars place it far to the north and north-east of Muṣaṣir (see the discussion on its location – Maniori 2010: 183–186). But if we accept this location Ḫubuškia was to be approached from Assyria through exceptionally difficult mountainous terrain. Besides, there are good reasons to locate this land closer to Mannea. In the 30th campaign of Shalmaneser III, his turtānu, Daiān-Aššūr went from Ḫubuškia through the city of Madahisa to Mannea (Grayson 1996: A.0.102.14 - 163), while Esarhaddon in his query to the

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Šamšī-Adad V’s Campaigns into the Zagros Revisited

775

Fig. 4. The Second Stage of the 3rd campaign (along the Great Khorasan Road).

sun god (Star 1990 = SAA 4 23) asks whether a Scythian force is going to invade Assyria, from Mannea through the passes of Ḫubuškia. If we assume that Mannea occupied the territory to the south of the Lake Urmia and Gilzānu was situated in the Ušnu-Solduz valley (Reade 1995: 34, fig. 2 – 35, fig.3), then the area around Pirānšahr is a candidate for the location of Ḫubuškia. Among the lands paying tribute to Šamšī-Adad we encounter Parsua. Its location is controversial, but I tend to accept J. Reade’s (1978: 139) proposal to place it in the vicinity of Sanandaj. The next toponym, Mannea, lay between two rivers – Rūd-e-Tātā’ū and Rūd-e-Zarineh (Maniori 2010: 186–187). It consisted of five provinces including Missi. Šamšī-Adad undoubtedly traversed Mēssi/Missi on the way to Gizilbunda (fig. 3). From the Assyrian sources we learn that it was the southernmost Mannean province (Zadok 2002: 94–95; Mayer 1983: 72–73, 51’). Its strategic position resulted from the fact that it was located along the corridor through which the Assyrians reached the eastern Mannean provinces of Andia and Zikirtu, the land of Gizilbunda, and Media. This toponym is also known from the annals of Shalmaneser III (24th palû) and the «Letter to the God Aššūr» of Sargon II. Both kings reached Mēssi/Missi from the land of Parsua. If one locates Parsua to the north of Sanandaj, Mēssi/Missi may have occupied the area in the upper reaches of the Rūd-e-Zarineh, near modern Takāb (Chamaza 1994: 115). There is a road linking this town through Bijār with Sanandaj.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

776

Hipp

The next stage was Gizilbunda, probably located in the valley of the Zanjān River and stretching to the Caspian Sea (Reade 1995: 41). In Gizilbunda Šamšī-Adad received tribute from a few local city rulers, whose realms are difficult to locate. The city of Sassiaš appears later during the campaign of Tiglath-pileser III in 744 B.C.E., but in a fragmentary context (Tadmor/Yamada 2011: Text 6, 12’). Kinaki, the first city razed and burned in Gizilbunda, as well as Karsibuta and Uraš appear exclusively in this campaign. The only toponym of this group which we are able to locate approximately is Sibara. It was situated on the boundary of Gizilbunda. It also occurs during the campaign of Tiglath-pileser III. The Assyrian king defeated a certain Upaš, ruler of Bit-Kapsi. The next stage comprises three cities including Ṣibur. J. Reade (1979: 180) identified Bit-Kapsi with modern Qazvin, but we do not know its western boundaries. Nevertheless, comparing the routes of ŠamšīAdad V and Tiglath-pileser III, it is obvious that the former advanced from the north-western direction from the lake Urmia Basin down into the Median realms, while Tiglath-pileser III reached Media from the West. Ṣibur may have been located on the modern road between Bijār and Hamadān (Medvedskaya 2010: 113). From there Šamšī-Adad perhaps went along the Great Khorasan Road and turned against Ḫanaṣiruka, the first Median king (fig. 4). He resided in Sagbita, his royal city. If Sagbita is identical with Bīt Sagbat attested in the annals of Tiglath-pileser III and Sargon II (Tadmor/Yamada 2011: Text 17, 2’ and Text 47, obv. 31; Fuchs 1994: 103 and 318, Ann. 95), then I. N. Medvedskaya’s (2010: 106–107) identification with classical Ecbatana (modern Hamadān) seems acceptable. Compared to the other rulers upon whom Šamšī-Adad imposed tribute, Ḫanaṣiruka stands out. He commanded quite a big army. His territory comprised 1200 settlements, a number which should, of course, be taken with caution. The last stage was Araziaš, located to the south of Bisutun in the vicinity of modern Harsin (Parpola/Porter 2001: map 11 and page 6; Zadok 2002: 74–75). The Assyrian sources point to its location between Bīt Ḫamban, Ḫarḫar, and Bīt-Barrua (part of Ellipi). Two reasons made Šamšī-Adad campaign far away onto the Iranian Plateau. Firstly, he undoubtedly realized that during the revolt of his elder brother he had lost the control over the northern route and therefore he had to secure a corridor – the route leading from Assyria through Mount Kullar to Parsua and Gizilbunda (the same which Sargon II partially used in 714 B.C.E.). Šamšī-Adad managed to do it by receiving tribute from the lands lying along the road. This task was of paramount importance, because according to the account of Karagündüz Stele (Salvini 2008: A 3-9), one of the objectives of Išpuini and Menua was the land of Paršuā to the south. The Karagündüz inscription mentions the mid-stages of the campaign: Qua, Šāritu and Nīgibe. I. M. Diakonoff and S. M. Kashkai (1981: 60) identify the Urarṭian Nīgibe with Mount Nikippa, which occurs during the 8th palû of Sargon II and which was close to the land of Sumbi (Mayer 1983: 110–111, 418’). Anyway, in the face of the Urarṭian pressure so far to the south, Šamšī-Adad’s action may be seen as a countermeasure

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Šamšī-Adad V’s Campaigns into the Zagros Revisited

777

against the Urarṭians threatening the road to Gizlibunda and consequently to Media. It is important to notice that not long ago a badly damaged inscription was discovered not far from Bukān at Taraqeh (www.cais-soas.com – 2008; Muscarella 2012: 267). Unfortunately, to my knowledge, it has not been published yet and we do not know to whom it dates. As the further most Urarṭian inscription in the Urmia basin it could tell us more about the historical geography of the region. Secondly, the Uraṭian expansion into the Urmia basin cut off the vital supply of horses from Gilzānu and Mannea. The opening and maintaining of the more southerly route to Media compensated for this loss. The struggle for maintaining control over the road continued during the reign of Šamšī-Adad’s successor, Adad-nārārī III, which is reflected by the entries in the Eponym Chronicle (Millard 1994: 57–58). Adad-nārārī III campaigned four times in Ḫubuškia, four times in Mannea and eight times in Media. Unfortunately, only «summary inscriptions» have been preserved from his reign.

Bibliography Arutyunyan, N. V., 1970. Biajnili (Urartu): Voenno-političeskaja istorija i voprosy toponimiki. Erevan (in Russian). –

2001. Korpus urartskich klinoobraznych nadpisej, Erevan (in Russian).

Brinkmann, J. A., 1968. Political History of Post-Kassite Babylonia (1158–722 B. C.). Analecta Orientalia 43. Roma. Brown, S. C., 1980. Kingship to Kingship: Archaeological and Historical Studies in the Neo-Assyrian Zagros. Ph.D. dissertation, Department of Near Eastern Studies, University of Toronto. Chamaza, G. W. V., 1994. Der VIII. Feldzug Sargons II. Eine Untersuchung zu Politik und historischer Geographie des späten 8. Jhs. v. Chr. (I). Archäologische Mitteilungen aus Iran 27, 91–118. Diakonoff, I. M., 1985. Media. I. Gershevitch (ed.), The Cambridge History of Iran, 2, The Median and Achaemenian Periods. Cambridge, 36–148. Diakonoff, I. M./Kashkai, S. M., 1981. Geographical Names According to Urartian Texts. Répertoire Géographique des Textes Cunéiformes 9. Wiesbaden. Fuchs, A., 1994. Die Inschriften Sargons II. aus Khorsabad. Göttingen. –

1998. Adâ. K. Radner (ed.), The Prosopography of the Neo-Assyrian Empire, 1/I: A. Helsinki.

–

2012. Urarṭu in der Zeit. Kroll et al. 2012: 135–161.

Gadd, C. J., 1954. Inscribed Prisms of Sargon II from Nimrud. Iraq 16/2, 173–201. Grayson, A. K., 1996. Assyrian Rulers of the Early First Millennium BC II (858–745 BC). The Royal Inscriptions Of Mesopotamia, Assyrian Periods, 3. Toronto.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

778

Hipp

Heidari, R., 2010. Hidden Aspects of the Mannean Rule in Northwestern Iran. ARAMAZD, V/2. Yerevan, 147–151. Kinnier Wilson, J., 1962. The Kurba’il Statue of Shalmaneser III, Iraq 24, 90–115. Kroll. S. (et al.), 2012. Acta Iranica 51. Biainili-Urartu. The Proceedings of the Symposium held in Munich, 12–14 October 2007. Leuven Lanfranchi, G. B., 2003. The Assyrian expansion in the Zagros and the local ruling elites. G. B. Lanfranchi et al. 2003: 79–118. Lanfranchi, G. B./Roaf, M./Rollinger, R. (eds.), 2003. Continuity of Empire (?): Assyria, Media, Persia. History of the Ancient Near East. Monographs V. Padova. Levine, L. D., 1972. Two Neo-Assyrian Stelae from Iran. Toronto. –

1973. Geographical Studies in the Neo-Assyrian Zagros. Iran 11, 1–29.

–

1977. East-west trade in the late iron age: A view from the Zagros. J. Deshayes (ed.), Le plateau iranien et l’Asie centrale des origines à la conquête islamique, Colloques intemationaux de C.N.R.S. 567. Paris, 171–182.

Liverani, M., 1992. Studies on the Annals of Ashurnasirpal II 2: Topographical Analysis. Roma. Maniori, F., 2010. Le campagne Assire contro l’Urartu del 715 E 714 A. C. Studi Micenei ed Ego Anatolici, 177–251. Mayer, W., 1983. Sargons Feldzug gegen Urartu – 714 v. Chr. Text und Übersetzung. Mitteilungen der Deutschen Orient Gesellschaft 115, 65–132. Medvedskaya, I. N., 2000. Zamua, Inner Zamua and Mazamua. R. Dittmann et al. (Hgg.), Variatio Delectat. Iran und der Westen. Gedenkschrift für Peter Calmeyer, AOAT 272. Münster, 429–445. –

2010. Ancient Iran on the Eve of Empires (9–6 В. С.) The History of the Median Kingdom. St. Petersburg (in Russian).

Millard, A., 1994. The Eponyms of the Assyrian Empire 910–612 B.C. State Archives of Assyria Studies 2. Helsinki. Muscarella, O. W., 2012. Hasanlu and Urartu. Kroll et al. 2012: 265–279. Parpola, S./Porter, M., 2001. The Helsinki Atlas of the Near East in the Neo-Assyrian Period. Helsinki. Radner, K., 2003. An Assyrian view on the Medes. Lanfranchi et al. 2003: 37–64. –

2012. Between a Rock and a Hard Place: Muṣaṣir, Kumme, Ukku and Šubria – The Buffer States Between Assyria and Urarṭu. Kroll et al. 2012: 243–264.

Reade, J., 1978. Kassites and Iranians in Iran. Iran 16, 137–143. –

1979. Hasanlu, Gilzanu, and Related Considerations. Archäologische Mitteilungen aus Iran 12,

–

1995. Iran in the Neo-Assyrian Period. M. Liverani (ed.), Neo-Assyrian Geography. Roma, 31–42.

175–181. Russell, H. F., 1984. Shalmaneser’s Campaign to Urarṭu in 856 B.C. and the Historical Geography of Eastern Anatolia According to the Assyrian Sources. Anatolian Studies 34, 171–201. Salvini, M., 1995. Geschichte und Kultur der Urartäer. Darmstadt.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Šamšī-Adad V’s Campaigns into the Zagros Revisited –

779

2001. Die Einwirkung des Reiches Urartu auf die politischen Verhältnisse auf dem Iranischen Plateau. R. Eichmann/H. Parzinger (Hgg.), Migration und Kulturtransfer. Bonn, 343–356.

–

2008. Corpus dei Testi Urartei, Volume 1, Le Iscrizioni su Pietra e Roccia – I testi. Documenta Asiana VIII, Roma.

Star, I., 1990. Queries to the Sungod: Divination and Politics in Sargonid Assyria. State Archives of Assyria IV. Helsinki. Tadmor, H./Yamada, S., 2011. The Royal Inscriptions of Tiglath-pileser III (744–727 BC), and Shalmaneser V (726–722 BC), Kings of Assyria. The Royal Inscriptions of the Neo-Assyrian Period, 1. Winona Lake, Indiana. van Soldt, W. H. et al., 2013. Satu Qala: A Preliminary Report on the Season 2010–2011. Anatolica 39, 197–239. Zadok, R., 2002. The Ethno-Linguistic Character of Northwestern Iran and Kurdistan in the Neo-Assyrian Period. Iran 40, 89–151.

Krzysztof Hipp, Jagiellonian University, Kraków, Poland.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Proceedings, 9 th ICAANE, Basel 2014, Vol. 3, 781–789

Chamsia Sadozaï – David Gandreau

Conserver après la fouille: la question des structures en terre crue à partir d’exemples en Asie centrale Nowadays, the archaeologist can not leave an excavation without thinking of the possibilities for its preservation. Our experience is showing that in the case of earthen architectural remains, even if it seems difficult to preserve, it is possible most of the time. It needs methodology and a strategy thought out in advance. Decay process has to be observed and analysed before any implementation. Then, many strategical and technical approaches are acceptable as we can see with the example of Merv in Turkmenistan, Sarazm in Tadjikistan and Fayaz Tepa in Uzbekistan. Tout comme les fouilles ne se limitent plus à la recherche d’objets d’art, l’archéologie d’aujourd’hui ne reste pas indifférente face aux valeurs patrimoniales des architectures découvertes. La question globale de leur conservation se pose d’autant plus qu’un grand nombre de sites qui ont alimenté notre imaginaire et fondé notre compréhension des grandes civilisations sont situés dans des contextes de conflit armé et de revendication identitaire. L’archéologue, confronté à la responsabilité des vestiges qu’il exhume, est souvent bien démuni. Que faire pour protéger ce qui ne peut être déplacé? Que choisir parmi les kilomètres de murs et à quelles fins? Comment être efficace sans altérer la structure d’origine ni perdre son information scientifique? Comment ne pas considérer les vestiges comme «perdus» après le départ des chercheurs? Au manque de connaissances, d’exemples réussis, de temps et souvent de budget consacré à la conservation, vient s’ajouter la spécificité du matériau. L’abondance des structures en terre crue dans le monde est évidente et bien attestée en contexte archéologique dans les régions orientales qui nous intéressent ici. En effet, nombreuses sont les constructions anciennes et même contemporaines faites de terre mise en œuvre selon diverses techniques dont les plus répandues sont l’adobe, la bauge, le torchis et le pisé.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

782

Sadozaï – Gandreau

Fig. 1. Frise des techniques de conservation pouvant être utilisées pour la conservation des vestiges en terre © CRAterre - C. Sadozaï.

La terre crue est le matériau commun à l’architecture domestique, cultuelle ou prestigieuse que l’on retrouve en fouille. Entre deux campagnes ou en fin de programme, les processus de dégradation des structures, stabilisés par l’enfouissement et le temps, sont réactivés. L’altération des murs, enduits, sols et couvertures quand elles subsistent, est alors incontestable. Les vestiges subissent un ensemble de facteurs naturels et anthropiques qui les rendent vulnérables si aucune mesure de conservation n’est pensée. Ces mécanismes, rarement simples, sont le résultat de multiples causes telles que l’érosion éolienne, le ruissellement, les remontées d’eau par capillarité, les piétinements des visiteurs, l’utilisation de matériaux inappropriés pour la conservation, etc… qui entraînent des réactions de détérioration en chaîne (Joffroy/Moriset 1996). Tel est le constat que l’on retient souvent: l’architecture de terre est fragile. On oublie alors que les constructeurs étaient fins connaisseurs des propriétés et des faiblesses de leurs bâtisses et que, par conséquent, des systèmes étaient pensés dès la construction pour permettre leur longévité. Pour bien comprendre ces systèmes, élaborés à force d’expériences et d’ajustements, l’observation des techniques constructives et des dommages survenus durant les phases d’utilisation du bâtiment sont aussi importantes que l’analyse des pathologies visibles après fouilles. L’interprétation de l’archéologue et l’expertise du conservateur-architecte sont en cela deux approches complémentaires, indispensables au choix des solutions stratégiques et techniques utilisées pour la conservation d’un site.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Conserver après la fouille: la question des structures en terre crue …

783

Dans un premier temps, il convient de déterminer l’objectif de la conservation afin d’assurer une protection efficace des structures en terre crue. Il peut s’agir par exemple d’une stabilisation des vestiges pour maintenir leur teneur scientifique entre deux campagnes de fouilles, ou bien d’une mise en valeur en vue d’accueillir du public. Une série de conséquences découlent de la définition de ce but, entre autres la durabilité de la conservation (court, moyen, long terme), la portée du message (scientificité, vulgarisation, identité, universalité, etc…), le public concerné (enfants/adultes, touristes locaux/internationaux, grand public/scientifiques, etc…). Pour y répondre, de nombreuses techniques plus ou moins intrusives existent (fig. 1) et s’utilisent de manière combinatoire. Le drainage de surface et le drainage périphérique sont par exemple deux systèmes simples d’évacuation des eaux ruisselantes au niveau du sol. Ils consistent à créer de légères pentes à la base des murs à partir de terre damée en lits successifs, vers le centre d’une pièce, l’extérieur d’un bâtiment ou un puits de récupération. L’eau de pluie ainsi éloignée et gérée ne pose plus de problèmes de stagnation ou de capillarité. A l’échelle d’un site, le drainage périphérique peut entraîner d’importants aménagements qu’il convient de concevoir selon un plan d’ensemble. En outre, il existe plusieurs types de consolidation par ajout de matière sacrificielle. Au sommet des murs ou sur la totalité de la surface du vestige, un enduit en terre respectant la composition d’origine est appliqué uniformément afin de subir les dégradations à la place de l’original. Les trous et les fissures sont traités de la sorte et évitent un creusement plus important qui peut entrainer une fragilité structurelle. En base de mur notamment, la pathologie la plus fréquente est le creusement dit du «sillon destructeur»1. L’une des réparations possibles consiste à ajouter des briques de terre crues maçonnées (si telle est la technique d’origine) et/ou plusieurs couches d’enduit vouées à se dégrader, et ainsi conserver la construction originale intacte. Une autre technique souvent utilisée procède par recouvrement du vestige. Le ré-enfouissement sous des lits de terre damée est en effet une méthode de conservation efficace. Le recouvrement peut aussi se faire au moyen d’une enveloppe de briques maçonnées qui respecte les volumes découverts lors de la fouille (fig. 2). Il se distingue de la reconstruction, autre technique possible, par le fait que cette dernière propose une interprétation des données et un choix quant à ce qui devait exister mais n’a pu être identifié. D’un autre ordre, la pose d’une couverture au-dessus des vestiges exhumés est une solution souvent envisagée malgré ses effets parfois dévastateurs. Nombreux sont les exemples de détérioration aggravée par la présence d’une toiture, lorsque celle-ci ne résout pas entièrement les problèmes d’érosion directe ou indirecte par la pluie ou qu’elle accentue

1

En référence aux observations de J.-C. Margueron à Mari en Syrie (Margueron 1985).

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

784

Sadozaï – Gandreau

Fig. 2. Citadelle d’Ulug Dépé, Turkménistan © MAFTUR – P. Hamouda.

des phénomènes tels que l’évaporation, l’occupation par des animaux/oiseaux ou encore l’abrasion par le vent. Cependant, cette technique se révèle efficace dans certains cas (fig. 3). Enfin, deux procédés relativement exceptionnels peuvent être employés pour protéger la valeur du site en dehors de celui-ci, à savoir le déplacement total ou partiel et la reconstitution réelle ou virtuelle hors site. Dans ce dernier cas, tout comme pour la reconstruction, une interprétation est choisie en fonction des connaissances à l’instant T de la conservation. Quels qu’ils soient, tous ces travaux nécessitent une maintenance régulière pour assurer une efficacité optimale et durable. Plusieurs fois par an, souvent après des épisodes pluvieux importants, une équipe doit venir observer, analyser et intervenir sur les zones de conservation endommagées. À titre d’exemple, citons tout d’abord le cas de l’oasis de l’ancienne Merv au Turkménistan. Le site classé au patrimoine mondial de l’UNESCO en 1999 couvre une superficie de 1200 hectares et enregistre une centaine de bâtiments allant du premier millénaire avant Jésus Christ au 15ème siècle de notre ère. Devant une telle multitude et diversité de structures (et par conséquent de pathologies), la stratégie de conservation est de réaliser des interventions continues de stabilisation sur les zones prioritaires, en prenant appui sur les capacités locales existantes. Ces capacités sont renforcées par des formations spécifiques et une

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Conserver après la fouille: la question des structures en terre crue …

785

Fig. 3. Vue d’ensemble d’un des chantiers couvert de Sarazm, Tadjikistan © CRAterre - C. Sadozaï.

expertise internationale depuis 1999, dans le cadre d’un partenariat impliquant le Ministère de la Culture turkmène (National Department for the Protection, Study and Restoration of Historical and Cultural Monuments), l’University College of London, et CRAterre-ENSAG (Gandreau/Moriset 2011). La configuration de Merv offre aux archéologues et aux conservateurs une panoplie complexe et variée de structures, certaines étant juste affleurantes (Shahryar ark), d’autres en élévation sur plusieurs mètres de haut (Kyz Kala/Ice House/pavillon timouride). Le projet se traduit par l’utilisation de quelques-unes des techniques réversibles et peu intrusives citées ci-dessus, choisies pour assurer continuité et qualité de la conservation, en respectant les valeurs sur l’ensemble du site, ainsi que les normes nationales et internationales en vigueur. Les partenaires mettent ainsi en œuvre des réponses techniques adaptées à chaque situation: la production de matériaux compatibles pour la conservation (briques crues et enduits de terre), la consolidation des sommets de mur, la réparation des bases de mur, des fissures et des trous, le drainage de surface et le drainage périphérique.2 Cette approche multidisciplinaire permet d’expérimenter des procédés innovants pour la gestion du bien, pour la documentation des fouilles et de la conservation, et pour la mise en valeur pour le public. 2

Voir http://www.ucl.ac.uk/merv/our_research/conservation.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

786

Sadozaï – Gandreau

Fig. 4. Greater Kyz Kala, Turkmenistan © CRAterre - C. Sadozaï.

Fig. 5. Expérimentation sur le chantier 5 de Sarazm, Tadjikistan © CRAterre - C. Sadozaï.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Conserver après la fouille: la question des structures en terre crue …

787

L’expérience acquise à Merv envisage des interventions d’envergure, avec notamment le projet de documenter et conserver l’un des édifices les plus remarquables du site, le «greater Kyz Kala» (fig. 4). Cet édifice, encore en élévation sur plus de dix mètres, nécessite une étude approfondie et des consolidations importantes permettant d’assurer la stabilité aussi bien à la base qu’au niveau des parements et des sommets des murs. Le deuxième exemple concerne le site de Sarazm au Tadjikistan. Ce site chalcolithique, inscrit lui aussi sur la Liste du patrimoine mondial en 2010, est le témoin d’un système de conservation novateur. Fouillé dans les années 1980, partiellement abandonné puis protégé par des couvertures métalliques avant d’être classé, il fait aujourd’hui l’objet de nouvelles campagnes de fouilles menées en partenariat avec les archéologues tadjiks et le CNRS. Un appui pour la gestion et la conservation du bien est assuré par CRAterre-ENSAG. Depuis le classement, les opérations de conservation sont menées conjointement entre archéologues et conservateurs-architectes, conformément aux directives du plan de gestion du bien. Les consolidations se concentrent sur trois problématiques. La première concerne les vestiges exhumés, abandonnés puis couverts par une toiture. Dans un état de dégradation avancé, quoique ralenti par la couverture, ils ont fait l’objet de drainages de surface, de consolidations sommitales voire entières des murs, de réparations et de ré-enfouissement des structures remarquables. Le deuxième axe concerne les vestiges fouillés puis abandonnés sans mesure de protection depuis une trentaine d’années. Une expérimentation menée en 2013 a permis de mesurer et d’analyser la dégradation en effectuant une nouvelle fouille des murs (déjà découverts) avant d’opérer un drainage de surface et l’application d’un enduit sacrificiel (fig. 5). De nouvelles mesures sont prises régulièrement pour statuer de la vitesse de détérioration avec ou sans protection. Enfin, le dernier domaine d’action conserve préventivement les structures nouvellement fouillées, particulièrement entre deux campagnes. Il s’agit de ré-enfouir des structures fragiles, de drainer les pièces mises au jour et d’enduire les murs. Cette expérience, du chantier 16, a permis de conserver en recyclant les déblais de fouille une zone de 75 m2, sur une hauteur de vestiges comprise entre 20 et 80 cm, en sept heures de travail à l’aide de quatre personnes. L’expérimentation à Sarazm est donc primordiale. La formation d’étudiants en archéologie et de praticiens tadjiks est également une composante importante de la plupart des actions menées ici, de même que la sensibilisation des différents intervenants liés au site. La volonté de conservation consiste à ce stade à des consolidations en l’état de vestiges protégés par une toiture et à une surveillance régulière. Un projet de mise en valeur et d’amélioration du circuit de visite est en cours d’étude. À Fayaz Tepa en Ouzbékistan, l’ouverture du site bouddhique au grand public a engagé la mise en œuvre d’un plan de drainage complet du site, d’une protection de type «enveloppe» du stupa, inséré dans une gaine de briques nouvellement façonnées, d’une reprise

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

788

Sadozaï – Gandreau

des sommets de murs, systématiquement rehaussés de plusieurs assises recouvertes d’un enduit sacrificiel, ainsi que la création d’un musée de site en briques crues reprenant les systèmes constructifs du monastère (Joffroy/Bendakir 2006). On le voit, au sein du même site, la conservation a recours à l’utilisation de différentes techniques, et au sein d’un ensemble régional fort tel que l’Asie centrale, les configurations sont multiples. Il en va de même des possibilités techniques pouvant apporter des solutions pérennes à la conservation des vestiges archéologiques. La recette miracle n’existe pas mais une stratégie de conservation pensée en amont peut s’appliquer à n’importe quel site, tant qu’elle s’appuie sur une méthodologie solide. De concert entre archéologues, conservateurs, architectes, institutions patrimoniales et laboratoires de recherche, cette méthode se définit par une phase préparatoire d’analyse à l’échelle globale du site qui permet une planification concertée de la conservation, une mise en œuvre documentée et une évaluation régulière (Gandreau/Sadozaï 2014). L’analyse comprend notamment l’histoire du site autant que des fouilles, l’environnement naturel, le contexte culturel, économique, social, législatif, les traditions et savoir-faire constructifs, les pathologies, etc… Tous ces éléments donnent une compréhension d’ensemble et permettent la mise en place d’une équipe pluridisciplinaire pour la gestion du bien. Celle-ci peut alors passer à la seconde phase, celle de planification de la conservation, qui se construit sur les valeurs que l’on prête aux vestiges et l’élaboration de critères permettant de choisir quels vestiges conserver, pourquoi, pour qui et comment. L’étude de faisabilité des différentes options envisagées, selon les plans de gestion, de conservation et d’entretien, facilite le passage à la phase de mise en œuvre. Cette étape dense nécessite une documentation minutieuse dans le but de comprendre, de former et de sensibiliser les interlocuteurs internes et externes au projet. Ainsi, la diffusion des résultats promet un effet «boule de neige» pour une vulgarisation de la conservation des vestiges en terre crue, difficile mais possible et réalisable à moindre coût. Le processus ne reste pas figé pour autant, car le suivi constant des opérations, la maintenance et la veille archéologique sont nécessaires à son bon fonctionnement dans le temps. La conservation des vestiges archéologiques en terre n’est pas une question simple. Certains exemples ont montré qu’elle pouvait parfois se résumer à des opérations de bon sens, faciles à mettre en œuvre et peu coûteuses. Plus largement, la méthode de mise en place du programme et sa planification sont les clés de réussite d’un projet de conservation. L’archéologue seul ne peut y parvenir, et il ne sait bien souvent pas à qui s’adresser pour mener un tel chantier. Avec l’appui des institutions locales et internationales, des conservateurs-architectes, des praticiens artisans, des universitaires et plus largement de la communauté locale, des solutions acceptables pour chacun peuvent être envisagées. La responsabilité est donc

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Conserver après la fouille: la question des structures en terre crue …

789

partagée et le scientifique n’intervient plus uniquement en tant que découvreur d’un passé oublié mais également comme garant de sa mémoire.

Bibliographie Gandreau, D./Moriset, S., 2011. Preventive conservation of the monuments in Merv, Turkmenistan. TerrAsia 2011, Proceedings of the 2011 international conference on earthen architecture in Asia. Ed. TERRAKorea. Mokpo, 51–61. Gandreau, D./Sadozaï, C., 2014. Conservation des architectures de terre sur les sites archéologiques. Nouvelles pratiques et perspectives. Ed. CRAterre-ENSAG. Grenoble. Joffroy, T./Bendakir, M., 2006. Fayaz Tepa, Surkhan Darya Region, Uzbekistan. Ed. CRAterre-ENSAG/ UNESCO/Boards of Monuments of Uzbekistan. Villefontaine. Joffroy, T./Moriset, S., 1996. Palais royaux d’Abomey. 1. Circonstances et processus de dégradation. (Projet PREMA-Bénin). Ed. ICCROM/CRAterre-EAG/UNESCO. Grenoble. Margueron, J.-C., 1985. Notes d’archéologie et d’architecture orientales [4. Propos sur le sillon destructeur (étude de cas)]. Syria 62.1–2, 1–20.

Chamsia Sadozaï et David Gandreau, Laboratoire CRAterre – ENSAG, Grenoble.

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

P. Bieliński, M. Gawlikowski, R. Koliński, D. Ławecka, A. Sołtysiak, Z. Wygnańska (Eds.)

Proceedings of the 8th International Congress on the Archaeology of the Ancient Near East Vol. 1: Plenary Sessions, Township & Villages, High & Low – The Minor Arts for the Elite and for the Populace Vol. 2: Excavation & Progress Reports, Posters Vol. 3: Archaeology of Fire, Conservation, Preservation & Site Management, Bioarchaeology in the Ancient Near East, Islamic Session Volume 1: 2014. XXXI, 705 pages, hc ISBN 978-3-447-10132-5 € 128,– (D) Volume 2: 2014. XXXI, 795 pages, hc ISBN 978-3-447-10174-5 € 148,– (D) Volume 3: 2014. XXXI, 605 pages, hc ISBN 978-3-447-10175-2 € 118,– (D)

The 8th International Congress on the Archaeology of the Ancient Near East (VIII ICAANE) was held in Warsaw in spring 2012, jointly organized by the University of Warsaw, the K. Michałowski Foundation and the Institute of Prehistory of the Adam Mickiewicz University in Poznań. The three-volume Proceedings contain the latest information and analyses reported at the VIII ICAANE. The first volume contains papers presented during plenary sessions as well as two of the main themes: Township and Villages and High and Low – The Minor Arts for the Elite and for the Populace. The second volume includes the most extensive session dedicated to the results of field research, Excavations and Progress Reports, covering the entire Near East, the Gulf countries and the Caucasus. The volume is completed by contributions presented during the poster session. The third volume is a collection of four themes: Archaeology of Fire, Bioarchaeology – an innovation of the 8th ICAANE, introduced to bring together scholars working within interdisciplinary projects on various aspects of the ecology of the ancient populations of the Near East –, Conservation, Preservation and Site Management and the Islamic Session. Anyone with an interest in the ancient Near East will find much to enjoy and appreciate in these three impressive volumes. Also available: ·–·–·–·–·–·–·–· 4th ICAANE (Berlin):

1: ISBN 978-3-447-05703-5 € 98,– (D) 2: ISBN 978-3-447-05757-8 € 88,– (D)

6 th ICAANE (Rom):

1: ISBN 978-3-447-06175-9 € 128,– (D) 2: ISBN 978-3-447-06216-9 € 98,– (D) 3: ISBN 978-3-447-06217-6 € 68,– (D)

7th ICAANE (London):

1: ISBN 978-3-447-06684-6 € 98,– (D) 2: ISBN 978-3-447-06685-8 € 118,– (D) 3: ISBN 978-3-447-06686-0 € 98,– (D)

HARRASSOWITZ VERLAG • WIESBADEN

www.harrassowitz-verlag.de • [email protected]

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2

Qaṭna-Studien Forschungsergebnisse und vergleichende Untersuchungen des deutsch-syrischen archäologischen Projekts auf dem Tall Misrife Herausgegeben von Peter Pfälzner und Michel al-Maqdissi 3: Thomas Richter, Sarah Lange

4: Elisa Roßberger

Das Archiv des Idadda

Schmuck für Lebende und Tote

Die Keilschrifttexte aus den deutschsyrischen Ausgrabungen 2001–2003 im Königspalast von Qaṭna Mit einem Beitrag von Peter Pfälzner

Form und Funktion des Schmuckinventars der Königsgruft von Qaṭna in seinem soziokulturellen Umfeld

2012. XXXII, 295 Seiten, 17 Abb., 2 Farbtafeln, 1 Faltkarte, 5 Tabellen, 57 Tafeln, gb ISBN 978-3-447-06709-6 E-Book: ISBN 978-3-447-19254-5 je € 88,– (D)

Im Sommer 2002 wurden im syrischen Mišrife, dem alten Qaṭna, 71 Tontafeln und Tontafelfragmente entdeckt. Unter den Tafeln befinden sich Briefe an König Idadda, sodass das Archiv insgesamt auf die Zeit um 1340 datiert werden kann. Das Archiv des Idadda beschließt die Publikation des Tontafelarchivs mit einer interdisziplinären Zusammenarbeit von philologischer Bearbeitung der Texte und archäologischer Auswertung des Fundkontextes. In den archäologischen Kapiteln wird anhand einer detaillierten Untersuchung der Versturzsituation die ursprüngliche Anordnung und Lagerung der Tafeln rekonstruiert. Unter Berücksichtigung der Fundlage und anhand von Vergleichen mit den Archiven aus Ugarit, Mari und Alalaḫ wird die Systematik und Organisation des Archivs erörtert und es werden Rückschlüsse gezogen auf die Art des Archivs und seine Einbindung in das Palastgefüge. Die Publikation der Tontafeln erschließt die Texte in sprachlicher Hinsicht. Dabei gilt besonderes Augenmerk einer ungewöhnlichen Vermischung akkadischen und hurritischen Sprachmaterials. Die Einordnung der den Texten zu entnehmenden Informationen in den geschichtlichen Kontext zeigt zudem, dass sie einem relativ kurzen, genau einzugrenzenden Zeitraum innerhalb der Amarna-Zeit entstammen.

2015. XXXVI, 482 Seiten, 221 Abb., 15 Diagramme, 44 Farbtafeln, 118 Tabellen, gb ISBN 978-3-447-10439-5 € 198,– (D)

In Band 4 der Reihe Qaṭna-Studien werden die über tausend in den Felskammern gefundenen Schmuckobjekte einer seit mehr als drei Jahrtausenden unberührten, reich ausgestatteten königlichen Grabanlage in Tall Mišrife erstmals umfassend vorgelegt. Aus der Fülle an archäologischen Daten ergeben sich grundlegende Fragen nach Bedeutung und Wert von Schmuck im konkreten soziokulturellen Kontext. Für ihre Beantwortung ist das Hinzuziehen von Keilschriftquellen, die den Umgang mit Schmuck an Königshöfen im mittel- und spätbronzezeitlichen Syrien dokumentieren, unumgänglich. Dazu gehört eine Gruppe von Inventartexten aus dem Königspalast von Qaṭna selbst, in denen die Schätze der Palastgottheiten notiert wurden. Diese werden in Elisa Roßbergers Arbeit detailliert ausgewertet und mit der archäologischen Evidenz aus der Königsgruft verglichen. Das Wissen über die Herstellung, Verwendung und Bedeutung von Schmuck in der Welt der Lebenden verhilft auch zu einem besseren Verständnis seiner Funktion im Totenritual. Die in der Königsgruft Bestatteten wurden nur für einen begrenzten Zeitraum mit farbenprächtigen Textilien und wertvollem Schmuck in Bilder ihrer selbst verwandelt; bei der Umbettung in sekundäre Ruhestätten wurde diese Ausstattung entfernt. Schönheit und Glanz dienten den Lebenden, aber sie spielten in der Welt der Toten keine Rolle.

HARRASSOWITZ VERLAG • WIESBADEN

www.harrassowitz-verlag.de • [email protected]

© 2016, Otto Harrassowitz GmbH & Co. KG, Wiesbaden ISBN Print: 978-3-447-10615-3 — ISBN E-Book: 978-3-447-19479-2