Interactions and New Directions in Near Eastern Archaeology

t a

s s

t e W

SISBA

West & East

Monografie,4 Direttore responsabile / Publisher Franco Zanini

Direttore scientifico /Editor Franco Zanini Ludovico Rebaudo Marco Iamoni

Comitato scientifico /Scientific Board Stephen Bourke Aaron A. Burke Filippo Maria Carinci Bojan Djurič Jan Driessen Mario Fales Robert Matijašić Augusta McMahon Maria Chiara Monaco Monika Verzár

Comitato editoriale / Editorial Board Daniele Morandi Bonacossi Simonetta Minguzzi Elisabetta Borgna Matteo Cadario

Segreteria di Redazione – Comunicazione editoriale e gestione Web / Assistant Editor – Editorial Assistant and Web-manager Elena Gagliano

Impaginazione Questo volume è integralmente disponibile online Gabriella Clabot a libero accesso nell’archivio digitale OpenstarTs, al link: https://www.openstarts.units.it/handle/10077/11831 © copyright Edizioni Università di Trieste, Trieste 2020

Proprietà letteraria riservata. I diritti di traduzione, memorizzazione elettronica, di riproduzione e di adattamento totale e parziale di questa pubblicazione, con qualsiasi mezzo (compresi i microfilm, le fotocopie e altro) sono riservati per tutti i paesi.

ISBN 978-88-5511-048-8 (print)

ISBN 978-88-5511-049-5 (online)

EUT Edizioni Università di Trieste via Weiss 21 – 34128 Trieste http://eut.units.it https://www.facebook.com/EUTEdizioniUniversitaTrieste Civilizations in Contact

Proceedings of the 5th “Broadening Horizons” Conference (Udine 5-8 June 2017) Volume 3 Interactions and New Directions in Near Eastern Archaeology

Edited by Costanza Coppini Francesca Simi

EUT EDIZIONI UNIVERSITÀ DI TRIESTE

Verità e Giustizia per Giulio Regeni, ricercatore appassionato Truth and Justice for Giulio Regeni, passionate researcher

TABLE OF CONTENTS

XI Preface

XIII Foreword

Civilizations in contact: current research and new approaches in Mediterranean and Near Eastern Archaeology

3 Elena Rova How wide is the Near East? Some reflections on the limits of “Near Eastern Archaeology” 23 Beatrice Barbiero A territorial model of animal husbandry for the southern Caucasus: some preliminary results 41 Eloisa Casadei The pottery repertoire at the end of the third millennium BC as chronological marker between southern Mesopotamia and the neighbouring regions. The case-study of the Syrian Jazirah 59 Francesca Chelazzi, Simone Bonzano Thinking data. Integrative big data approaches towards an ‘introspective’ digital archaeology in the ancient Mediterranean 77 Francesca Cioé, Marzia Merlonghi The Crisis Areas Archaeological Database (CAAD): a WebGIS for monitoring and safeguarding archaeological heritage 93 Costanza Coppini Settling the land: settlements pattern and ceramics in the land behind Nineveh from the Middle Bronze Age to the establishment of the Middle Assyrian State 111 Lorenzo Crescioli ‘Living burials’: reopening actions in kurgans and tombs between Central Europe and the Eurasian Steppe 137 Vittoria Dall’Armellina From the Caspian to the Aegean, from the Caucasus to the Carpathians. The circulation of the “long-sword” in the second millennium BC 153 Margherita Dallai The vaulted funeraryhypogea in Mesopotamia between the second and first millennium BC: localization and architectural features 171 Maurizio Fascitiello Rotary querns from Tell Barri (Syria): chronology, use and function 187 Soheila Hadipour Moradi Globalization in zoomorphic motifs during Iron Age in Iran and its neighbour

West & East IX Monografie, 4 205 Valentina Oselini Defining the MB-LB transition in northern Mesopotamia: some archaeological considerations on the new data from the Erbil Plain and neighbouring regions 221 Melissa Ricetti Matching Near Eastern seals and sealings: current issues and new perspectives 241 Sergio Giuseppe Russo Connecting cultures, dividing countries: a preliminary assessment of the Khabur Ware from Girnavaz Höyük, Southeastern Turkey 265 Kathrin Schmitt (Hi)storytelling: the ancient Near East in western historical novels and archaeological writing 279 Francesca Simi The Tell Gomel archaeological survey. Preliminary results of the 2015-2016 campaigns 293 Alessandro Canci, Hasan Ahmed Qasim The human burials from Şinduxa (Iraqi Kurdistan). A bioarchaeological and archaeothanatological approach

Marine connections: the Gulf and the interactions between the Arabian Peninsula, Mesopotamia, the Iranian world and beyond

313 Maurizio Cattani Adaptation and transformation of human setting from Middle Holocene to Early Bronze Age in south-eastern Arabian Peninsula 331 Massimiliano Ghiro, Enzo Cocca, Sabatino Laurenza A transitional model between Umm An-Nar and Hafit cultures: the case study of Grave 4 of Al Arid 343 Sabatino Laurenza, Marco Bianchi, Antonio Di Michele Graves, distribution and social memory: towards a new definition of funerary landscape in Oman 359 Silvia Lischi Decorated shell discs from Sumhuram, Oman 371 Mara Nicosia Christianity in the Gulf: vestiges of the East Syrian presence in late antiquity 385 Eleanor Lucy Preston The Ubaid in the Gulf: compositional analysis of ceramic material (sixth‑fifth millennium BC) 405 Cristiano Putzolu, Carlo Baione, Enzo Cocca, Sabatino Laurenza Rescue archaeology in the sultanate of Oman: methods and solution strategies

West & East X Monografie, 4 Preface

With the proceedings of the 5th edition of the “Broadening Horizons” Conference we have the pleasure to open the monograph series of the electronic journal “West & East” to the publication of international congresses and workshops whose topics fall within the main themes covered by the journal (for a presentation see https://www.openstarts.units.it/handle/10077/10886). We believe that West & East, although a mere youngster compared to other more renowned and prestigious journals, offers an appropriate (perhaps even ideal!) base for the promotion and diffusion of – in particular, though not only – the work of early career researchers. Thanks to the possibility of publishing with a “Gold Open Access” policy – i.e. immediate access to the articles or other features that can be downloaded from the website of the publisher EUT – and adhesion to the international rules of high quality standards for academic work – double peer review (which for West & East is also blind, i.e. the reviewers do not know who the author is) – we think that West & East is a scholarly journal well qualified to publish work that makes a significant contribution to the advancement of archaeological research in the Mediterranean basin and the Levant/Near East. The three volumes of the “Broadening Horizons” Proceedings undoubtedly belong to this category, and their publication opens the journal to stimulating future developments.

The editors of West & East

M. Iamoni L. Rebaudo F. Zanini

West & East XI Monografie, 4

Foreword

Broadening Horizons is a series of international East, and how these changes alter the “spectrum of conferences on the archaeology and history of our present field activities and research interests” the Ancient Near East, dedicated to students and (infra, p. 4). The diversity in terms of geographical early-stage researchers in these fields. The fifth environments, historical periods, and topics in Near edition was held in Udine from 5th to 8th June Eastern Archaeology is shown by the contributions 2017, hosted by the University of Udine, and to this session. Field research in relatively new organized by Marco Iamoni, Costanza Coppini, regions, i.e. the Caucasus, is reflected in papers Katia Gavagnin, Rocco Palermo and Francesca dealing with different aspects of archaeological Simi. It followed the conferences in Torino (Italy) investigations in these areas. Again with regard in 2011, Barcelona (Spain) in 2010, Lyon (France) to field research, the resurgence of archaeological in 2007 and Ghent (Belgium) in 2006. The main research in Iraq, especially in the Kurdistan region, topic Civilizations in Contact served to emphasize is one of the main topics tackled in this session, the importance of cross-cultural interactions concerning results from both archaeological survey in the Near East over time. In particular, the and excavation. Besides traditional research topics, present volume – Interactions and new directions although developed with original approaches, in Near Eastern Archaeology, edited by Costanza this edition of Broadening Horizons looks also at Coppini and Francesca Simi – is devoted to the contributions involving the use of computer science contributions from Session 7, “Civilizations in in archaeology, thus introducing the field of digital contact: current research and new approaches in humanities. Mediterranean and Near Eastern Archaeology”, Session 6 “Marine connections: the Gulf and and Session 6, “Marine connections: the gulf and the interactions between the Arabian Peninsula, the interactions between the Arabian Peninsula, Mesopotamia, the Iranian world and beyond” offers Mesopotamia, the Iranian world and beyond”. a wide-ranging view of archaeological and historical Session 7 “Civilizations in contact: current research carried out in the Arabian Peninsula, a research and new approaches in Mediterranean region often considered peripheral with respect and Near Eastern Archaeology” deals with new to Near Eastern Archaeology. The contributions research carried out throughout the Near East in collected in this volume do not reflect this tendency, various fields. The Near East’s wide-ranging cultural dealing with many epochs in the history of this diversity is addressed in detail by the session’s keynote region, from the Neolithic to the Islamic period. speaker, Elena Rova, who highlights the changes The opening keynote lecture of Maurizio Cattani under way in the archaeology of the Ancient Near focuses on the interactions of the Arabian Peninsula

West & East XIII Monografie, 4 with other regions, some beyond the traditional published here, without whom this volume would limits of the Near East such as cross-cultural have not been possible. contacts with Mesopotamia and Iran, speaking of The publication of this volume would have not “a wide-ranging history of territories, peoples and been possible without the invaluable support of culture” (infra, p. 315). These interactions can be the former (B. Callegher) and current (F. Zanini) readily observed with regard to pottery and the use director of the Inter-University School of Speciali of high-value materials that show links to Iranian zation in Archaeology (SISBA) and the EUT staff, traditions, as well as in the funerary practices in in particular G. Clabot. For the support in the cemeteries in Oman. organization of the conference, we are extremely These proceedings give a vivid picture of the grateful to the Friuli Venezia Giulia Regional exchanges and interactions that occurred during Authority and the Department of Humanities and the presentation and debate of specialist papers in Cultural Heritage of the University of Udine. A Udine at the conference. We would like to thank special thank goes to our friends and co-organizers the University of Udine, and its Department of Marco, Katia and Rocco for the time and the Humanities and Cultural Heritage Studies in commitment that we shared. particular, for having accomplished this successful Lastly, we would like to thank all contributors conference in a most stimulating environment. Our for their enthusiasm, collaboration and patience, and warmest thanks to the keynote speakers, especially for the scientific value of their papers. We hope to Elena Rova and Maurizio Cattani – whose papers have helped to broaden your archaeological horizons form part of this volume – for their invaluable – as ours too were broadened during the process of contributions and the example they set for younger organizing, managing and publishing the conference. researchers and students. During the initial organization and arranging Berlin and Oxford, December 2019 of the papers published here the support of the Scientific Committee was invaluable: we thank Costanza Coppini them all, as well as all peer-reviewers of the papers Francesca Simi

West & East XIV Monografie, 4 Civilizations in contact: current research and new approaches in Mediterranean and Near Eastern Archaeology

How wide is the Near East? Some reflections on the limits of “Near Eastern Archaeology”

ELENA ROVA

Ca’ Foscari University of Venice

Abstract

The paper considers how the traditional focus of Near Eastern Archaeology, which was centred on Mesopotamia and the surrounding areas of Syro-Palestine, Anatolia and Iran, has been challenged, in several ways and for several reasons, by recent developments of the discipline, whose geographical limits have thus become increasingly vague. Examples are given of how recent field activities, by revealing deep and in some cases unsuspected connections with areas which are traditionally the object of different disciplines and have developed a different scholarly tradition, encourage a renewed interest on long-distance circulation and diffusion of raw materials, artefacts and ideas, but at the same time require a deep re-adjustment of our theoretical frameworks and even of our scientific background. Special attention is devoted to the new perspectives about interconnections within the northern portion of the Near East and between this and other cultural macro‑areas (Central Asia, the Aegean, South-Western Europe and the Eurasian steppes) opened by field research in the highlands of the Southern Caucasus and Northern Turkey in the course of the last decades.

Keywords

Near Eastern archaeology, long-distance interactions, core/periphery, highlands and lowlands, raw materials, chronology, Anatolia, Iran, Caucasus, metallurgy

West & East 3 Monografie, 4 elena rova

1. Introduction1 results, my aim is, therefore, to provide a synthesis of recent trends of investigation. One of the questions a congress dedicated to “early-stage researchers and postgraduate students who work in various disciplines regarding the Ancient 2. Integrating new regions Near East and Eastern Mediterranean” should like- into Near Eastern Archaeology ly answer is: “Where are our disciplines heading to in the next future?”. As for Near Eastern Archae- Traditionally, Near Eastern Archaeology deals with ology, this, however, leads us to face another, more an area roughly corresponding to the present “Mid- basic question, namely: “What is the ancient Near dle East”, with Egypt being the object of a separate East now?” or, better said: “Are the scope and the discipline (Egyptology). Within this huge area, the limits of our discipline, as we traditionally con- main focus of research has long been the urban, lit- ceived them, still reflecting the spectrum of our pre- erate civilisations of “Greater Mesopotamia”, i.e. sent field activities and research interests?”. the Mesopotamian alluvium with its immediate This issue has a chronological dimension, as well, surroundings. Iran, Anatolia, Syria and the Levant as one could easily argue that the traditional focus have been progressively integrated into the disci- of our disciplines with the pre-classical civilisations pline, often as it turned out that they as well proof the “Near East” has also been challenged in the duced literate, urban cultures (e.g. Hittite Anatolia course of the last decades: let’s consider for instance or, more recently, Early Bronze Age Syria after the the inclusion of sessions devoted to Islamic Archae- discovery of Ebla in the 1970s) or, especially for the ology in the ICAANE congresses,2 or the presence, Levant, for their relevance to “Biblical studies”. in the very “Broadening Horizons 5” conference, of More often than not, field activities have been a session titled “West vs East: from Hellenism to the driven to these “peripheral” areas by factors which Roman expansion in the Near East”. are external to pure research questions, but depend I would like, however, to concentrate on the on the chances of modern international politics: in spatial dimension instead, and revise how the tra- plain words, archaeologists have tended to move to ditional geographical borders of “Near Eastern Ar- other regions either when Mesopotamia became in- chaeology” have been progressively challenged as accessible or difficult to access, or/and when individ- field research opened new areas of investigation, to ual countries offered particularly favourable condi- the point that we feel more and more uncomfort- tions to expeditions. One example is the ephemeral able with them, and that in some cases they have flourishing of Iranian archaeology in the 1970s, pri- even become an obstacle to new and promising or to the Islamic revolution, another one is the em- research approaches. Many of the issues I am go- phasis, between the 1960s and the early 2000s, on ing to discuss have recently been raised by other “Upper Mesopotamia” as a consequence of the nu- scholars; rather than illustrating original research merous projects of salvage archaeology connected with the constructions of dams on the main water- 1 The following text is an only slightly revised version of courses of the region. It is also worth observing how the paper presented at the “Broadening Horizons 5” confer- each of these “waves” of field activity created a gen- ence as a keynote lecture to the session: “Civilizations in contact: current research and new approaches in Mediterranean eration of specialists (many archaeologists who were and Near Eastern archaeology”. I would like to thank the two trained in the late 1970s and 1980s, for instance, are anonymous reviewers for their useful comments and sugges- specialists of Upper Mesopotamia and Syria, while tions, which I tried to integrate into the text. many of those belonging to the previous generation 2 “The International Congresses on the Archaeology of the Ancient Near East are organized every two years by the started as specialists of Iranian Archaeology), and scientific community of scholars working on and in the Near how, in its turn, it produces new research questions. East and studying therein material and environmental evi- Typically, intense field activities in new, or rel- dence from the most remote phases until the Islamic period within a multidisciplinary approach” (cit. from http://www. atively unexplored regions provoke an interest in icaane.net/). long-distance relations, “trade” and exchange of raw

West & East 4 Monografie, 4 How wide is the Near East? Some reflections on the limits of “Near Eastern Archaeology”

materials and artefacts, as opposed to the internal schedji, 2006-2011) for the third millennium,9 and development dynamics of each individual area. Suf- the SCIEM 2000 project headed by Manfred Bietak fice it to mention, here, the wave of studies about (1999-2011) for the second millennium.10 lapis lazuli,3 carneol and steatite/clorite,4 which ac- Due to the dramatic crisis which, starting in the companied the above-mentioned excavations in early 1990s with the First Gulf War and still con- Iran, or the debate about the “Uruk colonisation” tinuing today in Syria and in parts of Iraq, made originated by salvage archaeology in Northern Syria fieldwork virtually impossible over most of the and South-Eastern Turkey.5 Long-distance inter- Mesopotamian core area and to the worsening con- actions and “trade” are of course old and recurring ditions for foreign expeditions in other countries of themes of research, but what I would like to draw the Near East, we are now again in a phase in which attention on is how their popularity shows an “ebb new, or previously little explored regions have be- and flow” trend, whose peaks coincide with, or im- come the focus of intensive field research. One case mediately follow, periods when excavations prolif- at issue are the former republics (now independent erate in regions outside of the Mesopotamian core, states) of the Soviet Union, both in the Southern and are interspersed by periods in which this field of Caucasus (Georgia, Armenia and Azerbaijan) and research is less practiced. in the area of Central Asia (Turkmenistan, Kazakh- Another consequence of the multiplication stan, Uzbekistan), many of which have recently of archaeological excavations in the external sec- opened to international collaborations in the field tors of the Near East has been a renewed interest, of archaeology; another one is the states of the Gulf which still continues today, in relative chronologies and the Arabian Peninsula, where archaeological ac- and synchronisation of the regional periodisations, tivities have recently intensified. a pre-condition for any worthwhile analysis of in- The clearest evidence of the ongoing integration terregional relations. Between the late 1980s and into Near Eastern Archaeology of these regions, lo- the early 2000s, this resulted in a number of dedicated at the very limits of the Near, or Middle East, cated congresses6 and international projects, which or even outside of its borders (in the case of Central completely superseded traditional syntheses on the Asia) from a purely geographical point of view, is the topic.7 Among the most successful international appearance of dedicated sessions and the increasing ventures, we can mention the Santa Fe workshop number of papers dealing with related subjects in “Mesopotamia in the Era of State Formation” or- the last editions of the ICAANE international con- ganised by M. Rothman in 1998,8 which resulted in gress. 11 In spite of all this, large areas of the “Ancient the new, now widely accepted periodisation of the Near East” still remained, until very recently, only Late Chalcolithic period for the fourth millennium, poorly investigated, and are on the whole rather ne- the international EFS supported ARCANE project glected in international field literature as well. (principal investigators M. Lebeau, Pierre de Miro- Not only have the geographical boundaries of Near Eastern archaeology become wider and wid- 3 Tosi 1974, 1980a, 1980b. er, and increasingly blurred and vague; when inves- 4 Kohl 1974, 1975, 1978. tigating regions far from the Mesopotamian core 5 Algaze 1993; Stein 1999; Rothman (ed.) 2001; area, researchers are often confronted with cultures Postgate (ed.) 2002. and societies, which escape the traditional objects 6 Examples are the above mentioned congress “Arte- facts of Complexity” (Manchester 1998, Postgate (ed.) 9 Thepublication of the results of ARCANE project is 2002), and the congress “Chronologie des Pays du Caucase et still in progress. For a general presentation and the final perio- de l’Euphrate aux IVe-IIIe millénaires”(Istanbul 1998, Marro, disation, see the project’s website (http://www.arcane.uni-tue- Hauptmann (eds.) 2000). bingen.de/) and the introduction to the first published volume, 7 One such case was the long-expected new edition of which is dedicated to the Syrian Jezirah (Lebeau 2011). Ehrich’s Chronologies in Old World Archaeology (Ehrich 10 For the wide literature produced by the SCIEM 2000 (ed.) 1992), which became outdated very soon after it was first project, the reader is referred to the project’s website (http:// published. www.oeaw.ac.at/sciem2000/). 8 Rothman (ed.) 2001. 11 http://www.icaane.net/.

West & East 5 Monografie, 4 elena rova

of the discipline: urban and/or literate civilisations the landscape, than those archaeologists working in and lowland agricultural societies. Increasing atten- Greater Mesopotamia are used to, a circumstance tion is being focused, in particular, on the specific which often creates the impression that one is deal- features of cultures which developed in so far ne- ing with vast “empty” spaces with little or no human glected natural environments, with consequences occupation. While totally artificial,tell -like mounds which begin to call into question traditional as- are rare and mounded settlements tend to be found- sumptions and research parameters of the disci- ed on natural hills and do not show a considerable pline. Although I will especially concentrate, for il- depth of anthropic layers, many other archaeolog- lustrating this point, on recent research concerning ical sites are representative of “lighter”, rather un- the highland societies of the northern sector of the stable forms of occupation: small ephemeral villages Near East, these represent by no means the only case built in perishable materials, seasonal campsites etc. at issue, another one being represented, for instance, All of them are not easily detectable not only by tra- by recent developments in the archaeology of the ditional survey, but also by modern remote sensing Arabian Peninsula.12 techniques, on the diverse topography which characterises these regions, which are often covered by vegetation for most of the year. On the other hand, 3. Integrating “Highland societies” into other categories of sites which are frequently met Near Eastern Archaeology on the Highlands (e.g. stone forts, military outposts and terracing systems, ancient mines, rock art sites, If one looks at a physical map of the Middle East, or monumental funerary barrows) are more easily it becomes obvious that most of its territory actu- visible on the surface, but consist of features which ally consists of mountains, highlands and plateaus. are quite difficult to date, as they typically yield little These regions rarely produced urban centres. Even diagnostic finds and often show a high continuity of when they did produce cities, these were generally occupation. of smaller dimensions than those of the Mesopota- In short, in order to be discovered and proper- mian alluvium and showed a lower level of occupa- ly evaluated, Highland anthropised landscapes need tional continuity. Their economy relied less heavily, special attention and appropriate methods, which and frequently not at all, on irrigation agriculture. archaeologists used to work in the “core areas” of On the other hand, the role of activities such as pas- the Near East are often not trained for, nor accus- toralism, mining, etc., in the economy of these high- tomed to. In particular, techniques developed in land cultures was much higher, as was the interac- the last decades in the fields of European Pre- and tion of their “cities” with groups practicing a more Protohistory and Ancient Topography may be prof- mobile style of life. Central governments in these re- itably applied to these regions. One especially suc- gions had to negotiate their authority with the lead- cessful example is the recent work on metal and ers of tribal groups, a fact which generally resulted salt mines between the Southern Caucasus and into unstable political entities, which often took North-Western Iran, much of which was accom- the form of tribal confederations. Let’s think, just plished by scholars trained in European prehistory to make the most obvious examples, to the Elamite or mining archaeology (such as the specialists of the 15 kingdom,13 or to the kingdom of Nairi/Urartu.14 Deutsches Bergbau Museum in Bochum). In addition, the inhabitants of the Highlands More in general, the growing attention to the generally left less evident traces of their presence on archaeological indicators of transhumant/mobile pastoralism16 and the development of the debate

12 This was the object of Session 6 (“Marine connec- 15 Gambashidze, Stöllner (eds.) 2016 (gold mines tions”) of the Broadening Horizons conference, to whose pro- in Georgia); Bobokhyan et Al. 2017 (gold mines in Arme- ceedings I refer for further discussion. nia); Marro et Al. 2010 (salt mines in Azerbaijan). 13 Potts 1999. 16 Although analysis of animal bones is increasingly inte- 14 Kroll et Al. (eds.) 2012. grated in excavation reports, specific studies focusing on their

West & East 6 Monografie, 4 How wide is the Near East? Some reflections on the limits of “Near Eastern Archaeology”

about its role in Near Eastern societies17 reflect the can as well represent the expression of egalitarian intensification of field research in countries (Anato- societies. N. Laneri however recently argued that lia, Iran, the Southern Caucasus, etc.) where these Highland ceremonial sites of later periods may rep- practices traditionally play an important economic resent a different type of social complexity, which role. However, non-urban societies were also living may be conceptualised in terms of heterarchical sys- in close proximity to, and lively interacted with, the tems where leadership is not centralised, but shared sedentary societies of the Mesopotamian alluvium, or dispersed among different agents.20 whose dynamics cannot be fully understood with- In order to evaluate the intellectual achieve- out taking them into consideration, as recent exca- ments of Highland populations, attention may be vations in Iraqi Kurdistan have clearly shown.18 focused on their capacity to control the procure- From a more theoretical point of view, research ment and the exchange of rare raw materials, and in the Highlands requires an attention to indica- their being the seat of specialised handicrafts, or tors of social complexity and cultural achievements the point of origin of technological innovations. In which are different from those of lowland societies. particular, it is evident that by their very location In fact, highland societies are inevitably deemed to and geology highland territories played a crucial appear as “primitive” (thus following the stereo- role in the procurement and circulation of metals, typed description which Mesopotamian textual whose leading role in the economy of Near Eastern sources conveyed of them) if we measure their so- Bronze Age societies is evident even from textu- cial complexity by largest settlement size, presence al sources, and had therefore long been recognised of palatial and temple architecture and other indica- by scholars. Not by chance, therefore, the intensi- tors of a central power, and their intellectual perfor- fication of field activities in the Highland regions mances by the presence of written texts, etc. of Anatolia, North-Western Iran and the Caucasus In fact, monuments whose construction in- was matched by a renewed wave of research about volved a high degree of communal work are not metals and metallurgy.21 missing in these areas, but take forms to which Near Eastern archaeologists are less accustomed, such as huge barrow graves (kurgans) and sanctuar- 4. New perspectives on long-distance ies positioned in rather inaccessible locations. The interactions presence of monumental burials and the accumulation in burial contexts of wealth in the form of I would now like, however, to elaborate on another mobile objects of precious and/or, exotic materials theme, that is how intensive field research in pre- (metals, semi-precious stones, etc.) have long been viously scarcely explored areas located at the geo- recognised to represent appropriate indicators of graphic periphery of the Near East has revealed social complexity for Highland societies. (or confirmed) deep and in some cases unsuspect- More recently, scholars’ attention was also fo- ed connections with areas which are traditionally cused on extra-urban sanctuaries. As clearly demon- the object of other archaeological disciplines, a fact strated by the case of the aceramic Neolithic com- which challenges our traditional view of the Near 19 plex of Göbekli Tepe , these are not necessarily a East as consisting of a “core” and of a periphery, manifestation of state-like hierarchical societies, but whose main role was to provide the core with nat- contextual analysis and on their role as indicators of different 20 Laneri 2014. models of animal exploitation are still rare (see, e.g., Piro 2009; Berthon 2013; Siracusano, Bartosiewicz 2012; Sira- 21 Relevant literature is too wide to be analytically listed cusano, Palumbi 2014), and the use of isotopic analysis of here, but Yener 2000; Courcier 2010; Gambashidze et both bones and seeds for tracking population movements is still Al. 2010; Hansen et Al. (eds.) 2010, Vatandoust et Al. in its beginnings (Messager et Al. 2015). (eds.) 2011 can provide a first overview of recent research in different regions. For earlier decades, Chernykh 1992 provides 17 Marro 2004; Palumbi 2010, 2012. the most complete synthesis of Soviet research on the topic, fo- 18 Kopanias, MacGinnis (eds.) 2016. cusing on the northern half of Eurasia, but including the Cau- 19 Schmidt 2006. casian region as well.

West & East 7 Monografie, 4 elena rova

Figure 1 A traditional model of interregional relations in the ancient Near East (adapted from Wilkinson 2016; Sherrat 2004)

ural resources (again, as clearly demonstrated many belong to the near eastern periphery, play a crucial years ago by M. Liverani, following the view provid- role in connecting the Near East with the different ed by written Mesopotamian sources) 22 (fig. 1). cultural macro-areas (the Indus Valley and Cen- In fact, less monolithic models of interconnec- tral Asia and, respectively, Egypt and the Aegean) tions within and outside of the ancient Near East which surround it. These are traditionally the object had already emerged in the 1970s, following the im- of other archaeological disciplines; when seen in a portant discoveries in Iran and Central Asia. wider perspective, however, they represent parts Fig. 2, for instance, shows a map, by the late of a wider continuum of ancient civilisations and Maurizio Tosi, of the Early Bronze Age civilisations cultures, whose frontiers appear more and more of Middle Asia in the later third millennium BC, in blurred and permeable as research progresses. which Mesopotamia occupies a relatively periph- The old core/periphery paradigm is thus becom- eral position. In a similar way, Mesopotamia plays ing less and less satisfactory while, in its place, the a rather marginal role in the circuit of Late Bronze model of a polycentric network is gaining momen- Age international trade as exemplified by the mate- tum. This not only does not require drawing fixed rials from the famous Uluburun shipwreck, excavat- boundaries between civilisations and cultures, but ed between 1984 and 1994, which rather focuses on allows for different areas (or policies) to act as a cen- the Eastern Mediterranean and the Levantine area tre for their immediately surrounding regions and (fig. 3). to occupy, at the same time, a peripheral position These two examples highlight how regions (Iran with respect to other regions. More remarkably, and the Levant), which are usually considered to it also allows to highlight the importance of some areas/policies – or human groups –, which at first 22 See, most recently, Liverani 2017. On traditional models and techniques for visualising highland-lowland inter- sight play a minor role in this cultural continuum, dependence, see also Wilkinson 2016. in connecting with each other different foci of civil-

West & East 8 Monografie, 4 How wide is the Near East? Some reflections on the limits of “Near Eastern Archaeology”

Figure 2 Cultural areas of Middle Asia in the 3rd millennium BC (after Tosi 1980b, p. 1796)

Figure 3 Possible route of the Uluburun ship (after Aruz et Al. (eds.) 2008, fig. 97)

West & East 9 Monografie, 4 elena rova

isation. The recent popularity of network analysis in Caucasus region and the traditional Mesopota- archaeology may provide a better fitting paradigm mian core are very few, though not totally nonex- and a more appropriate terminology to this situa- istent. On the other hand, however, connections tion, by helping to define, e.g., “nodes” and “links” with regions belonging to the northern portions of within this huge interconnected area23. the Near East – Northern Iran and the Southern The intensification of field research in areas lo- Caspian area, Anatolia, parts of Northern Mesopo- cated at the very limits of the “Near East” also in- tamia – and, beyond these, with the Aegean to the vites us to explore long-distance routes, or corridors, west and with Central Asia to the east are becom- and networks of interconnections which are rather ing increasingly evident as field research progresses. remote from the traditional Mesopotamian core Finally, the position of the region at the crossroads and even, to a certain extent, independent from it. between the Near East and the huge world of the An approach to ancient trade which cross-cuts tra- Eurasian steppes which extends beyond the Cauca- ditional borders between Near Eastern Archaeolo- sus opens up the possibility of connections with ar- gy and other disciplines is, for instance, evident in eas (the Northern Black Sea, the Balkan peninsula) various contributions to the “ArchAtlas Project”24 which are traditionally the object of European pre- founded by the late Andrew Sherrat at the Univer- history, and even, through the intermediary of the sity of Sheffield, first of all in the attempts (2004) Northern Caucasus, with the steppes of Northern by Sherrat himself at modelling the long-term de- Central Asia. velopment of trade routes in Eurasia and beyond.25 In considering the origins and affiliation of More recently, a similar approach characterises cultural phenomena, Near Eastern archaeologists the volume Tying the Threads of Eurasia by Toby working in the Southern Caucasus are naturally in- C. Wilkinson,26 which explores the flows of differ- clined to look at the areas they know better, i.e. in ent materials and objects (metals, stones, textiles southern direction. This is certainly appropriate in etc.) between Anatolia, the Southern Caucasus and many cases: examples are, for the late fifth-earlier Western Central Asia by mobilising a wide array of fourth millennium BC, the diffusion in the region data and methods, and other recent research by the of Chaff-faced wares with North Mesopotamian af- same author27. finities29 or, for the early second millennium Middle Bronze Age, the Trialeti culture, the connections of which with Anatolia and even with Southern Mes- 5. Northern connections? opotamia have long been recognised.30 However, a look toward the north may some- In the following, I will concentrate, as an example, times be equally illuminating: for instance, the pro- on the perspectives opened by excavations in the fusion of gold in the graves of the Late Chalcolith- Southern Caucasus. 28 If we consider the Greater ic North Caucasian Maikop culture and, later on, Caucasus range as the northern geographical limit in those of the South-Caucasian mid-third mil- of the Near East, the countries of the Southern Cau- lennium Early Kurgan cultures, may be explained, casus are undoubtedly part of the latter, although and has actually been explained by some, through with an apparently rather marginal role within it. a distant familiarity with the early achievements of In fact, direct connections between the Southern Balkan-Carpathian metallurgy. 31 To make another example, the diffusion, in the Late Bronze Age, of stone architecture and cyclopic fortresses in the 23 Knappet (ed.) 2013. Southern Caucasus after a long period of apparent- 24 http://www.archatlas.org/Home.php. ly non sedentary occupation is sometimes explained 25 Sherrat 2004. 26 Wilkinson 2014. 29 27 E.g., Wilkinson 2016. Marro 2011. 30 28 The most recent and updated synthesis on the archae- Rubinson 2003; Puturidze 2005. ology of the Caucasus is provided by Sagona 2018. 31 Kohl 2007; Hansen 2010, 2014a.

West & East 10 Monografie, 4 How wide is the Near East? Some reflections on the limits of “Near Eastern Archaeology”

with the emergence in the area of complex policies lennium BC, of comparable élite “wagon graves” under stimuli from – or as a reaction to – the north- (a generally northern tradition) from the Southern ern expansion of the contemporary Near Eastern Caucasus (Martqopi and Bedeni cultures) to South- empires. Sabine Reinhold, however, notices that a ern Mesopotamia and Elam (Ur, Susa). similar spreading of stone architecture is attested, I am not proposing here a simple revival of old dif- at the same time, in the Northern Caucasus, where fusionist theories (e.g. Gimbutas’ kurgan theory36), she connects it with the emergence of settled pasto- but I argue that the fact that these theories have been ralism, a phenomenon that, starting with the early abused or misapplied in the past shouldn’t restrain second millennium BC, covered an immense area in one from considering such widespread phenomena Western Eurasia to the north32. as a worthwhile object of study and that, in fact, data More in general, Near Eastern archaeologists from recent excavations and new methods of analy- working in the Southern Caucasus are confronted sis may contribute to shed new light on them. For in- with new evidence pertaining to long-debated phe- stance, high-precision radiocarbon chronology may nomena which have a much wider diffusion than the make a significant contribution, in the future, to old Near East and in connection to which the Near East disputes about the priority of one or another region plays a rather marginal role. This is the case, for inin the introduction of these innovations.37 stance, of the spreading of barrow graves (kurgans),33 The northern connections of the Southern Cau- or of the diffusion of anthropomorphic stelae with casus, which go back at least to the Late Chalcolithic warrior-like features. The latter appear, between the period, thus urge us to reconsider the “urban revolu- fourth and the third millennium BC, over an area tion” and its pivotal role as the foundation of Near spanning from the Caucasus as far as Portugal, and Eastern civilisations as a special manifestation of a subsequently spread over the vast Central Asiat- “global revolution” which takes place, in the fourth ic steppes as well. Sven Hansen (2013) has recent- millennium BC, over a much wider area compris- ly connected their first appearance with the emer- ing large portions of Asia and Europe. 38 Within this gence in Eurasia of a new social type: a “heroic ruler”, large interconnected area, which crosscuts our tra- whose presence would be also signalled by the em- ditional disciplinary limits and which we can visu- phasis on metal weaponry (especially daggers and alise as a network with many centres and peripher- axes) in contemporary élite burials. As Hansen notes, ies, in which flows of raw materials, objects, people, it is probably no chance that this “northern” type of but also of technologies, information and ideologies heroic ruler appears at approximately the same time travelled along multiple and not mutually exclusive as the Uruk “priest-king”, another symbolic expres- routes, each area and each human group developed sion of emerging kingship, the iconography of which its own combination, which we should certainly try spreads, in the second half of the fourth millennium, to analyse in its peculiarities, but without forgetting from Southern Mesopotamia to Western Iran and the general framework. the Upper Euphrates, and even reaches Egypt.34 This requires, in Near Eastern archaeologists Other cases of still to be explored possible who work in this area at least, an intensified dia- long-distance connections are the appearance, again logue with specialists of disciplines (European pre- in the fourth millennium BC, of wheeled vehicles history in this case), and an effort in familiarising over an area spanning from Southern Mesopota- with concepts, periodisation systems, 39 materials mia, through the Northern Caucasus, to Northern 36 Gimbutas 1997; more recently, see also Anthony 35 Europe, and the presence, in the mid-third mil- 2007. 37 For a synthesis of problems and perspectives of radio- 32 Reinhold 2017. carbon chronology in the Southern Caucasus, see, among oth- 33 Gimbutas 1997. For a recent overview of research on ers, Passerini et Al. 2018. barrow graves, cf. Borgna, Müller Celka (eds.) 2012. 38 Thus, the fourth millennium BC has recently been 34 Schmandt-Besserat 1993. defined by S. Hansen (2014b) “a watershed in European pre- 35 Fansa, Burmeister (eds.) 2004; Petrequin et history”. Al. (eds.) 2006; Burmeister 2011. 39 On the synchronisation of chronologies and periodisa-

West & East 11 Monografie, 4 elena rova

etc. which may not be so familiar to them. On the sised how it represents an over-simplification of a other hand, this attention to the wider framework much more complex collection of interconnected should be matched by a deepened analysis of each local, regional and interregional communications individual region and even of each category of ar- networks. This is certainly a consequence of west- tefacts, 40 in a continuous feedback between general ern field research of the second half of the twenti- and particular, which is the only way to avoid exces- eth century focusing on specific regions; however, sive generalisations and to guarantee a real progress it also reflects the special attention paid by scholars of knowledge. interested in long-distance connections to classes of artefacts recovered in earlier excavations in Meso- potamia and, ultimately, depends on the narrative 6. Anatolia and the Southern Caucasus transmitted to us by Mesopotamian textual sources as part of a Northern Near Eastern as well. connection network Thus, the role of Anatolia in most general models of Near Eastern trade is limited to a few exceptional sites (e.g. Kültepe in Central Anatolia, or An example of such promising topics of research is Troy on the Aegean coast) which yielded Mesopo- the attempt at better defining the flows of materi- tamian-related finds. In the case of Kültepe, in par- als, information etc. which connected, during the ticular, the discovery of the archives of the Assyrian Bronze Age, Anatolia and the Southern Caucasus merchants contributed to direct scholars’ attention with the rest of the Near East, the Aegean and Cen- on the site’s connections with Syro-Mesopotamia, tral Asia. whereas possible northern connections were rather It is evident that in most of the past models of disregarded. Fig. 5 shows two reconstructions of trade routes and interconnections between the Kültepe’s connection network (around 180043 and, Near East and either Central Asia or the Aegean respectively, in the mid-third millennium BC44); it (see, for instance figs 2 and 3, above), the Southern is interesting to observe how on both maps North- Caucasus is either not considered, or given a total- ern, and especially North-Eastern Anatolia, reply marginal role. The same is true for large parts of resents a sort of “empty space”. Anatolia, as well. Thus, the main third-early second Recent research has however begun, through a millennium BC land and sea routes (or, better said, multiplicity of different approaches, to unravel the corridors, or “highways”)41 crossing the Near East, complexity of exchange networks within ancient as reconstructed from the distribution of materials Anatolia. Michele Massa and Alessio Palmisano45, like lapis lazuli, carneol or chlorite vessels, are those for instance, have better characterised the impor- connecting Central Asia, the Indus Valley and Iran, tance of Central-Western Anatolia within different through Mesopotamia and Syria, with the Levan- exchange networks, and the role played by North- tine coast, and from there with Egypt. The result- ern Mesopotamia and the Northern Levant in con- ing general model (fig. 4) is somehow similar to a necting this area with the Mesopotamian alluvium. version of the classical Medieval “Silk Road”.42 Al- As their analysis focuses on exchanges between Ana- though the importance of this main axis of commu- tolia and Mesopotamia, however, North-Eastern nication should not be denied, it should be empha- Anatolia remains an empty space on their maps, too. On the other hand, other contributions46 tried tion systems between Caucasia and the Near East, see Rova in press. to reconstruct the network of interactions in 40 E.g. Helwing 2012, 2017 argues for different, over- Bronze Age Anatolia by comparing it with net- lapping networks of ceramic versus metal manufacture traditions and artefacts circulation and exchange in the Late Chal- 43 colithic period. Kulakoğlu, Kangal (eds.) 2010. 44 41 Wilkinson 2014. Tonussi 2007. 45 42 As recently formulated by T. Wilkinson (2014), it may Massa, Palmisano 2018. be defined as “one of many possible ‘Silk routes’”. 46 Wilkinson 2014; Massa 2010.

West & East 12 Monografie, 4 How wide is the Near East? Some reflections on the limits of “Near Eastern Archaeology”

Figure 4 Satellite map of the Near East with indication of the main axe of Bronze Age interregional communications passing through Syro-Mesopotamia (based on NASA, Visible Earth, Bluemarble images)

Figure 5 Maps showing the network of interconnections of the Kültepe site around 1800 (above) and in the mid-3rd millennium BC (below) (after Kulakoğlu, Kangal (eds.) 2010, map on pp. 170‑171; Tonussi 2007, map 22)

West & East 13 Monografie, 4 elena rova

works of routes and roads from different historical excavations, the publication of artefacts from old periods, like, for instance, Roman roads or pilgrim- museum collections, renewed attention to specific age routes. Toby Wilkinson in particular, following categories of materials (metals in particular) and a line of research inaugurated by Catherine Marro47 through comparative study of the evidence from some years before, drew special attention to the im- the two regions. portance, in this connection, of less formal routes Discoveries at third millennium sites in the (“pathways” as he calls them) like for instance those northern part of central Turkey (especially graves or of seasonal migration by nomadic and transhu- hoards with rich collections of metal burial goods, mant groups (fig. 6). It is interesting to observe that like e.g. Resuloğlu e Kalınkaya-Toptaştepe, Bekaroğ- the “empty space” represented by Northern and lu Köyü, but also settlements, such as Ikiztepe) have North-Eastern Anatolia, the regions which presum- allowed Th. Zimmermann49 and others to re-evalu- ably had the strongest connections with the South- ate well-known finds like those from Alaca Höyük ern Caucasus in the interregional communication and Horoztepe in the context of a mainly east-west network, is slowly starting to get filled in the maps oriented Early Bronze Age network of intercon- attached to these contributions. nections. This northern corridor can be assumed to Another, more traditional but complementary continue not only in western direction towards the way to tackle the same question is to look at the Aegean, the Black Sea and South-Eastern Europe, distribution of sites and archaeological materials. but also, in eastern direction, toward the Southern Earlier research, mainly by archaeologists work- Caucasus50 and even beyond this, to the South-Cas- ing in Western Turkey and by specialists in Aege- pian region51 and Northern Iran and, through the an prehistory, had already made clear how Central latter, until Central Asia52. Finally, it also provided Anatolia was part of a network of communications access, through the coastal plains and the few pass- involving the Aegean and South-Eastern Europe, es which cross the Greater Caucasus barrier, to the which concerned different artefacts, and was par- Northern Black Sea and the steppes of the North- tially independent from, though interconnected ern Caucasus (fig. 7). with, the “mainstream” route heading to the Med- That this northern network of connections – iterranean coast through Syro-Mesopotamia. A or at least some portions of it – were already exist- good example is represented by the distribution, in ing in the fourth millennium BC is proved, among the mid-third millennium BC, of depas and tan- others, by the presence of lapis lazuli in the kurgans kard cups,48 which barely touches North-Western of the Maikop and Novosvobodnaya cultures in the Syria and the Middle-Upper Euphrates, but does Northern Caucasus53 and in the roughly contem- not extend beyond these limits. North-Eastern porary Late Chalcolithic kurgans of Soyuk Bulak in Turkey, on the other hand, probably lies outside Azerbaijan. 54 While lapis lazuli is known, at ca the the dissemination area of such ceramic types, and same time, in different areas of the Near East (e.g. at therefore appears as an “empty space” on these dis- Tepe Gawra in Northern Mesopotamia) and even tribution maps as well. in Egypt, and may therefore have travelled, like in Only in the course of the last decade, intensification of research both in Northern Turkey and in the Southern Caucasus has resulted in the expan- 49 Among the numerous contributions by this author, see sion in north-eastern direction, in scholars’ percep- especially Zimmermann 2007, 2009, 2011. tion, of the Early Bronze Age Anatolian connection 50 On the Caucasian connections of the Alaca Höyük network. Gaps in our knowledge are being progres- graves, see already Mansfeld 2001. 51 sively filled through the infusion of data from new For some preliminary considerations about this still poorly explored region, see Piller 2012. 52 47 Marro 2004. On third millennium connections with Central Asia, see especially Wilkinson 2014. 48 For a recent visualisation of the distribution of de- 53 pas and tankards, see Massa 2010; see also Tonussi 2007, Ivanova 2012. pp. 261-272 et passim. 54 Lyonnet et Al. 2008.

West & East 14 Monografie, 4 How wide is the Near East? Some reflections on the limits of “Near Eastern Archaeology”

Figure 6 Traditional routes of seasonal migration in Anatolia (after Wilkinson 2014, fig. 2.9)

Figure 7 Satellite map of the Near East with indication of the supposed “northern corridor” and of its connections with other main axes of interregional communications (based on NASA, Visible Earth, Bluemarble images)

West & East 15 Monografie, 4 elena rova

later times, along multiple, different routes,55 other just by comparing a map of the archaeological finds finds from the Soyuk Bulak kurgans support these in the involved regions with a map of the location “northern connections”. This is the case, for in- of the main metal sources, that it is also connect- stance, of the animal-headed sceptre, which has a ed with the search, control and trade of metal ores. precise parallel at Sé Girdan in North-Western Iran, Hence, future advances in research concerning min- and is also vaguely reminiscent of sceptres from the ing activities and metallurgical technology, along northern pontic steppes, but especially of the metal the different lines of enquiry that have been men- finds, which include a “copper” dagger and a num- tioned above,60 are also likely to produce relevant re- ber of golden beads. sults on the specific topic of “northern interconnec- Connections of the Maikop culture of the tions” in the next future. Northern Caucasus with Northern Iran and Cen- On the other hand, the emphasis placed, since tral Asia have been repeatedly pointed out by Maria the very beginning, on metal weapons within this Ivanova in the course of the last few years.56 In a re- circuit of exchanges brings us back to another issue cently published contribution she focuses in partic- which was briefly commented upon in a previous ular on the distribution of a series of metal artefacts section of this paper, namely that of a specifically of Central Asian and South Caucasian origin (shaft- northern type of heroic or warlike rulership, which hole axes and grip-tongued daggers) which, in her possibly travelled along the same routes.61 opinion, were introduced in the fourth millenni- This early tradition of “northern connections”, um BC to the Northern Caucasus and from there which can be traced back at least to the fourth mil- spread, in the early third millennium, to Eastern lennium BC, represents a possible background Europe as well as to Anatolia and the Aegean.57 The against which specific transfers of objects, rituals origin of fourth millennium shaft-hole axes is con- and ideologies between the Near East and the Aege- troversial, as some authors regard them as a belat- an area during the third and second millennia might ed transfer of the East-European tradition of heavy be, in the future, evaluated independently from, and copper axes;58 be that as it may, it is clear that they as an alternative to, the traditional route following mainly circulate along routes, and in ideological cir- the “Syro-Levantine Corridor”. One among many cuits, which are different, though to a certain extent cases at issue is the distribution, in the second mil- in communication with, those of the contemporary lennium, of long swords (so-called “rapiers”),62 emerging Syro-Mesopotamian urban centres. which stretches from the Southern Caucasus to the The same is true for the earliest swords: these are Aegean islands and Inland Greece, thereby com- first attested at approximately the same time (the pletely skipping Syro-Mesopotamia. late fourth millennium BC) both in the Turkish Upper Euphrates (at Arslantepe) and in the North- ern Caucasus (in kurgan no. 31 at Klady) and even 7. Some final remarks in the third millennium BC they are not attested south of the Anatolian region (Alaca Höyük).59 To conclude, recent excavations in less known, geo The mere number of metal objects which are dis- graphically marginal areas of the Near East have tributed throughout this northern network of in- opened up the view of a precociously interconnect- terconnections is a clear hint to the fact that the lated, global ancient world which widely exceeds the ter is associated with the circulation of metals and traditional limits of Near Eastern archaeology as an metallurgical expertise. It becomes equally evident, institutional discipline (fig. 8). Research in these areas is beginning to provide the hitherto missing 55 Cf. Wilkinson 2014, 123-133, fig. 4.3 for a map. 56 Ivanova 2012, 2013. 60 See supra, § 3. 57 Ivanova 2016. 61 See supra, § 5. 58 For a recent discussion, see Helwing 2017. 62 Abramishvili 2001; most recently, also Dall’Ar- 59 Piller 2015; see also Dall’Armellina 2017. mellina 2017; Dall’Armellina, this volume.

West & East 16 Monografie, 4 How wide is the Near East? Some reflections on the limits of “Near Eastern Archaeology”

Figure 8 Satellite map of the Near East with indication of the interrelations of the different peri-Mesopotamian regions with each other and with the other contemporary cultural macro‑areas (based on NASA, Visible Earth, Bluemarble images)

links between the Near Eastern civilisations and in order to analyse the single threads of this intercon- their contemporary neighbours, and will allow a nected network and their evolution in the course of fresh look at old-debated questions concerning long time, we have to focus our research more narrowly, distance circulation of raw materials, objects, tech- and less episodically, on each individual area, as well nologies, and ideas. If on the one side this may lead as on each period, material, technology, category of to somehow relativise some achievements of the tra- objects, etc. For this, we have to broaden our set of ditional Mesopotamian core area, on the other one methodological instruments to all nowadays avail- it can definitely contribute to better contextualise able techniques of analysis. At the same time, how- them in the framework of a multicentric network of ever, it is necessary for us to “broaden our horizons” exchanges which connects the Near East with other to the other civilisations of the ancient world, either cultural macro‑areas: Central Asia, the Indus Val- by gaining ourselves new competencies in disciplines ley, the Eurasian steppes, South-Eastern Europe, the which may not be part of, or may not be central in, the Eastern Mediterranean, and Egypt. university training of Near Eastern archaeologists, or In order to fully exploit the new wealth of data at least by promoting a continuous scholarly dialogue which field research is continuing to offer us, it is crosscutting traditional disciplinary borders. necessary for us to swing between two opposite, but Definitely, the developments described in this not mutually exclusive approaches. On the one hand, paper are not totally devoid of risks. It might be

West & East 17 Monografie, 4 elena rova

feared, for instance, that the same identity of “Near Near East as a specific cultural macroarea, as vague as Eastern archaeology” as a separate discipline may its borders may become. This undoubtedly requires be threatened by the constant expansion of its geo- that future generations of Near Eastern archaeolo- graphical field of activities and by the overlap, in gists, besides focusing their research on a specific re- scope as well as in required competencies, with dif- gion within the Near East and considering it in its ferent archaeological disciplines. This could be overall network of connections – including those considered as a special case of a very general trend, with other cultural macro‑areas –, continue to pos- whereby traditional university curricula and aca- sess a general knowledge about the history of the demic disciplinary borders are confronted with in- Near East, its natural environment and its and past creasingly interdisciplinary research practices. and present social dynamics, which is not requested In fact, the challenge for Near Eastern to the specialists of other archaeological disciplines, archaeology in the next future may well be to main- and thus represents an essential part of their scientif- tain its traditional focus on the civilisations of the ic identity.

West & East 18 Monografie, 4 How wide is the Near East? Some reflections on the limits of “Near Eastern Archaeology”

BIBLIOGRAPHY

Abramishvili M. 2001, Transcaucasian Rapiers and Chernykh E.N. 1992, Ancient Metallurgy in the USSR. the Problem of their Origin, in: Boehmer R.M., The Early Metal Age, Cambridge. Maran J. (eds.), Lux Orientis, Archäologie zwischen Courcier A. 2010, Les débuts de la métallurgie dans Asien und Europa, Stellerloh, pp. 1-8. les pays du Caucase, de la fin du VIe au milieu du IIIe Algaze G. 1993, The Uruk World System. The Dynamics millénaire avant notre ère (Chalcolithique et début de of Expansion of Early Mesopotamian Civilization, l’Age du Bronze): recherches sur leur origine d’après le Chicago. potentiel métallifère, l’analyse d’objets et les relations Anthony D.W. 2007, The Horse, the Wheel, and observées avec les régions voisines (unpublished Ph the- Language: How Bronze-Age Riders from the Eurasian sis), Paris. Steppes Shaped the Modern World, Princeton. Dall’Armellina V. 2017, Power of Symbols or Symbols Aruz J., Benzel K., Evans J.M. (eds.) 2008, Beyond of Power? The ‘Long Sword’ in the Near East and the Babylon. Art, Trade and Diplomacy in the Second Aegean in the Second Millennium BC, «Ancient Millennium B.C., New Haven. Near Eastern Studies» 54, pp. 143-182. Berthon R. J.-P. 2013, Ehrich R.W. (ed.) 1992, Chronologies in Old World New Data on the Exploitation rd of Animal Resources in the Upper Tigris River Area Archaeology (3 edition), Chicago. (Turkey) during the Second and First Millennia Fansa M., Burmeister S. (eds.) 2004; Rad und BC, in: De Cupere B., Linseele V., Hamilton- Wagen. Der Ursprung einer Innovation im Vorderen Dyer S. (eds.), Archaeozoology of the Near East X. Orient und Europa (Wissenschaftliche Beischrift zur Proceedings of the Tenth International Symposium Sonderausstellung vom 28. März bis 11. Juli 2004 im on the Archaeozoology of South-Western Asia and Landesmuseum für Natur und Mensch in Oldenburg Adjacent Areas, Paris, pp. 145-162. / Beiheft der Archäologischen Mitteilungen aus Bobokhyan A., Kunze R., Meliksetian K., Nordwestdeutschland 40), Mainz. Pernicka E. 2017, Society and Metal in Bronze Age Gambashidze I., Mindiaschwili G., Armenia, in: Rova E., Tonussi M. (eds.), At the Gogotschuri G., Kachiani K., Dschaparidze Northern Frontier of Near Eastern Archaeology: Recent I. 2010, Udzvelesi metalurgia da samto saqme Research on Caucasia and Anatolia in the Bronze Age Saqartveloshi dzv. ts. VI – III atastsleulebshi (Alte / An der Nordgrenze der vorderasiatischen Archäologie: Metallurgie in Georgien in 6.-3. Jt. v. Chr.), Tbilisi. Neue Forschung über Kaukasien und Anatolien Gambashidze I., Stöllner Th. (eds.) 2016, The Gold (Publications of the Georgian-Italian Shida Kartli of Sakdrisi. Man’s First Gold Mining Enterprise. Das Archaeological Project II / Subartu 38), Turnhout, Gold von Sakdrisi. Die erste Goldmine der Menschheit, pp. 501-523. Rahden/Westf. Borgna E., Müller Celka S. (eds.) 2012, Ancestral Gimbutas M. 1997, The Kurgan Culture and the Indo- Landscapes: Burial Mounds in the Copper and Bronze Europeanization of Europe. Selected Articles from Ages (Central and Eastern Europe, Balkans, Adriatic, 1952 to 1993, (Robbins Dexter M., Jones-Bley th nd Aegean, 4 -2 millennium BC). Proceedings of the K. eds.), Washington. International Conference held in Udine, May 15th‑18th Hansen S. 2010, 2008 (Travaux de la Maison de l’Orient et de la Communication and Exchange betwe- Méditerranée. Série recherches archéologiques 58), en the Northern Caucasus and Central Europe Lyon. in the Fourth Millennium BC, in: Hansen S., Hauptmann A., Motzenbäcker I., Pernicka E. Burmeister S. 2011, Innovationswege – Wege der (eds.), Von Maikop bis Trialeti. Akten des Symposiums Kommunikation. Erkenntnisprobleme am Beispiel Berlin 1.-3. Juni 2006, Bonn, pp. 297-313. des Wagens im 4. Jt. v. Chr., in: Hansen S., Hansen S. 2013, Müller J. (eds.), Sozialarchäologische Perspektiven: The Birth of the Hero. The Emergence of th , in: Starnini Gesellschaftlicher Wandel 5000-1500 v. Chr. zwischen a Social Type in the 4 Millennium BC E. (ed.), Atlantik und Kaukasus (Internationale Tagung 15.- Unconformist Archaeology. Papers in honour 18. Oktober 2007 in Kiel), Bonn, pp. 211-240. of Paolo Biagi (BAR IS 2528), Oxford, pp. 101-112.

West & East 19 Monografie, 4 elena rova

Hansen S. 2014a, Gold and Silver in the Maikop Culture, Third Millennium (unpublished Ph.D. dissertation, in: Meller H., Risch R., Pernicka E. (eds.), Department of Archaeology, Harvard University). Metalle der Macht – Frühes Gold und Silber, Metals Kohl Ph.L. 1975, Carved Chlorite Vessels: A Trade in of Power – Early Gold and Silver (6. Mitteldeutscher Finished Commodities in the Mid-Third Millennium, Archäologentag vom 17. bis 19. Oktober 2013 in «Expedition» 18(1), pp. 18-31. Halle (Saale)), Halle (Saale), pp. 389-410. Kohl Ph.L. 1978, The Balance of Trade in Southwestern th Hansen S. 2014b, The 4 Millennium: A Watershed Asia in the Mid‑Third Millennium B.C., «Current in European Prehistory, in: Horejs B., Mehofer Anthropology» 19, 1978, pp. 463-92. M. (eds.), Western Anatolia before Troy. Proto- Kohl Ph.L. 2007, , Urbanisation in the 4th Millennium BC? (Proceedings The Making of Bronze-Age Eurasia of the International Symposium held at the Cambridge, New York, Merlbourne. Kunsthistorisces Museum Wien, Austria, 21-24 Kopanias K., MacGinnis J. (eds.) 2016, Archaeological November 2012), Vienna, pp. 243-259. Research in the Kurdistan Region of Iraq and the Hansen S., Hauptmann A., Motzenbäcker I., Adjacent Areas, Oxford. Pernicka E. (eds.) 2010., Von Maikop bis Trialeti Kroll S., Gruber C., Hellwag U., Roaf M., (Akten des Symposiums Berlin 1.-3. Juni 2006), Zimansky P. (eds.) 2012, Biainili-Urartu: the Bonn. Proceedings of the Symposium held in Munich 12-14 (Acta Iranica 51), Leuven. Helwing B. 2012, Late Chalcolithic Craft Traditions at October 2007 the North-Eastern ‘Periphery’ of Mesopotamia: Potters Kulakoğlu F., Kangal S. (eds.) 2010, Anatolia’s vs. Smiths in the Southern Caucausus, «Origini» 34, Prologue, Kültepe Kanesh Karum, Assyrians in pp. 201-220. Istanbul, Istanbul. Helwing B. 2017, Networks of Craft Production and Laneri N. 2014, Ritual Practices and the Emergence of Material Distribution in the Late Chalcolithic: Social Complexity in the Upper Tigris Region at the Metallurgical Evidence from Iran and the Southern Beginning of the Second Millennium, in: Bonatz Caucasus, in: Rova E., Tonussi M. (eds.), At the D. (ed.). The Archaeology of Upper Mesopotamian Northern Frontier of Near Eastern Archaeology: Piedmont in the Second Millennium BC (Topoi Recent Research on Caucasia and Anatolia in the – Berlin Studies of the Ancient World), Berlin, Bronze Age / An der Nordgrenze der vorderasiati- pp. 119-130. schen Archäologie: Neue Forschung über Kaukasien Lebeau M. (ed.) 2011, Associated Regional Chronologies und Anatolien (Publications of the Georgian-Italian for the Ancient Near East and the Eastern Shida Kartli Archaeological Project II / Subartu 38), Mediterranean. Vol. 1: Jezirah, Turnhout. Turnhout, pp. 51-78. Liverani M. 2017, Assiria: La preistoria dell’imperiali- Ivanova M. 2012, Kaukasus und Orient: Die Entstehung smo, Bari, Roma, 2017. , des „Maikop-Phänomens“ im 4. Jahrtausend v. Chr. Lyonnet B., Akhundov T., Almamedov K., «Praehistorische Zeitschrift » 87, pp. 1-28. Bouquet L., Courcier A., Jellilov B., Ivanova M. 2013, The Black Sea and the Early Huseinov F., Loute S., Makharadze Z., Civilizations of Europe, the Near East and Asia, Reynard S. 2008, Late Chalcolithic Kurgans Cambridge. in Transcaucasia. The Cemetery of Soyuk Bulaq Ivanova M. 2016, Stop and go: die Ausbreitung kauka- (Azerbaijan), «Archäologische Mitteilungen aus sischer Metallformen in Osteuropa in der ersten Hälfte Iran und Turan» 40, pp. 27-44. des 3. Jt. v. Chr., in: Hänsel B., Schier W. (eds.), Mansfeld G. 2001, Die ›Königsgräber‹ von Alaca Der Schwarzmeerraum vom Neolithikum bis in Höyük und ihre Beziehungen nach Kaukasien, die Früheisenzeit (6000–600 v. Chr.). Kulturelle «Archäologische Mitteilungen aus Iran und Turan» Interferenzen in der Zirkumpontischen Zone und 33, pp. 19-61. Kontakte mit ihren Nachbargebieten (Prähistorische Marro C. 2004, Itinéraires et voies de circulation du Archäologie in Südosteuropa 30), Rahden/Westf., Caucase à l‘Euphrate: le rôle des nomades dans le pp. 403-415. système d‘échanges et l‘économie protohistorique des Knappet C. (ed.) 2013, Network Analysis in Archaeology: IVème-IIIème millénaires av. notre ère, in: Nicolle New Approaches to Regional Interaction, Oxford. C. (ed.), Nomades et sédentaires dans le Proche- Kohl Ph.L. 1974, Seeds of Upheaval: The Production of Orient ancien. Compte rendu de la XLVIe rencontre Chlorite at Tepe Yahya and an Analysis of Commodity Assyriologique internationale (Paris, 10-13 Juillet Production and Trade in Southwest Asia in the Mid- 2000), Paris, pp. 51-62.

West & East 20 Monografie, 4 How wide is the Near East? Some reflections on the limits of “Near Eastern Archaeology”

Marro C. 2011, Where Did Late Chalcolithic Chaff- Petrequin P., Arbogast R.M., Petrequin A.M., Faced Ware Originate? Cultural Dynamics in Van Willigen S., Bailly M. (eds.) 2006, Premiers Anatolia and Transcaucasia at the Dawn of Urban chariots, Premiers araires. La diffusion de la traction Civilization (ca 4500-3500 BC), «Paléorient» 36.2 animale en Europe occidentale pendant les IVe et IIIe (2010), pp. 35-55. millénaires avant notre ère (Monographies du CRA, Marro C., Bakhshaliyev V., Sanz S. 2010, 29), Paris. Archaeological Investigations on the Salt Mine of Piller C.K. 2012, Landscape Archaeology South of the Duzdagi (Nakhchivan, Azerbaïdjan), «TÜBA-Ar» Caspian Sea: Some New Insights, in: Matthews R., 13, pp. 233-246. Curtis J., Seymour M., Fletcher A., Gascoigne A., Marro C., Hauptmann H. (eds.) 2000, Chronologies des Glatz C., Simpson St. J., Taylor H., Tubb J., Chapman th Pays du Caucase et de l’Euphrate aux IV-IIIe millénai- R. (eds.), Proceedings of the 7 International Congress res/From the Euphrates to the Caucasus: Chronologies on the Archaeology of the Ancient Near East 12 April for the 4th-3rd. Millennium B.C./Vom Euphrat in den – 16 April 2010, the British Museum and UCL, Kaukasus: Vergleichende Chronologie des 4. und 3. London, Vol. 3. Fieldwork & Recent Research, Posters, Jahrtausends v. Chr. (Varia Anatolica 11), Paris. Wiesbaden, pp. 119-133. Massa M. 2010, Networks of Interaction in Early Bronze Piller C.K. 2015, Die Ältesten Schwerter der Welt: die Age Anatolia, «ArchAtlas, Version 4.1», http:// Funde von Arslantepe und Klady, in: Meller H., www.archatlas.org/workshop09/works09-massa. Schefzik M., Ettel P. (eds.), Krieg – eine archäolo- php, accessed: 29 December 2017. gische Spurensuche, Darmstadt, pp. 283-284. Massa M., Palmisano A. 2018, Change and Continuity Piro J. J. 2009, Pastoralism in the Early Transcaucasian in the Long-distance Exchange Networks betwe- Culture: The Faunal Remains from Sos Höyük en Western/Central Anatolia, Northern Levant and (Unpublished PhD Thesis), New York University. Northern Mesopotamia, c. 3200–1600 BCE, « Journal Potts D.T. 1999, The Archaeology of Elam. Formation of Anthropological Archaeology» 49, pp. 65-87. and Transformation of an Ancient State, Cambridge Messager E., Herrscher E., Martin L., Kvavadze 1999. E., Martkoplishvili I., Delhon C., Kakhiani Postgate J.N. (ed.) 2002, Artefacts of Complexity: K., Bedianashvili G., Sagona A., Bitadze Tracking the Uruk in the Near East, Oxford. L., Poulmarch M., Guy A., Lordkipanidze Puturidze M. 2005, About the Problem of South D. 2015, Archaeobotanical and Isotopic Evidence of Caucasian – Near Eastern Cultural Relations Early Bronze Age Farming Activities and Diet in the According to the Items of Artistic Craft During the Mountainous Environment of the South Caucasus: First Half of the II Millennium BC, «Metalla» 12(1- A Pilot Study of Chobareti Site (Samtskhe-Javakheti 2), pp. 8-22. Region), «Journal of Archaeological Science» 53, pp. 214-226. Reinhold S. 2017, Late Bronze Age Architecture in Caucasia and Beyond: Building a New Lifestyle for Palumbi G. 2010, Pastoral Models and Centralised a New Epoch, in: Rova E., Tonussi M. (eds.), At Animal Husbandry. The Case of Arslantepe, in: the Northern Frontier of Near Eastern Archaeology: Frangipane M. (ed.), Economic Centralisation in Recent Research on Caucasia and Anatolia in the Formative States. The Archaeological Reconstruction Bronze Age / An der Nordgrenze der vorderasiati- th of the Economic System in 4 Millennium Arslantepe, schen Archäologie: Neue Forschung über Kaukasien Roma, pp. 149-163. und Anatolien (Publications of the Georgian-Italian Palumbi G. 2012, Bridging the Frontiers. Pastoral Shida Kartli Archaeological Project II / Subartu 38), Groups in the Upper Euphrates Region in the Early Turnhout, pp. 337-366. Third Millennium BCE, «Origini» 34, pp. 261-278. Rothman M.S. (ed.) 2001, Uruk Mesopotamia & Its Passerini A., Rova E., Boaretto E. 2018, Revising Neighbors: Cross-cultural Interactions in the Era of the Absolute Chronology of the 4th and 3rd Millennium State Formation, Santa Fe. BCE in the Southern Caucasus, in: Horejs B., Rova. E. in press, A Unified Terminology for the South- Schwall Ch., Müller V., Luciani M., Ritter M., Caucasian „Early Bronze Age“: A Worthy and Giudetti M., Salisbury R. B., Höflmayer F., Achievable Target?, in: Kashuba M., Reinhold th Bürge T. (eds.), Proceedings of the 10 International S. (eds.), Der Kaukasus im Spannungsfeld zwischen Congress on the Archaeology of the Ancient Near East Osteuropa und Vorderem Orient: Dialog der (25 -29 April 2016, Vienna), vol. I, Wiesbaden, Kulturen, Kultur des Dialoges (im Gedenken an den pp. 161-172. 140. Geburtstag von Alexander A. Miller).

West & East 21 Monografie, 4 elena rova

Rubinson K. 2003, Silver Vessels and Cylinder Sealings: (ed.), Gururajamanjarika. Studi in Onore di Giuseppe Precious Reflections of Economic Exchange in the Early Tucci 1, Napoli, pp. 1-20. Second Millennium BC, in: Rubinson K., Smith A.T. Tosi M. 1980a, Il crocevia dell’Asia, Roma, in: Le grandi (eds.), Archaeology in the Borderlands: Investigations in avventure dell’archeologia. I misteri delle civiltà scom- Caucasia and Beyond, Los Angeles, pp. 128-143. parse, Vol. V., Roma, pp. 1754-1795. Sagona A. 2018, The Archaeology of the Caucasus. From Tosi M. 1980b, Il viaggio del lapislazzuli, in: Le grandi Earliest Settlements to the Iron Age, Cambridge. avventure dell’archeologia. I misteri delle civiltà scom- Schmandt-Besserat D. 1993, Images of Enship, in: parse, Vol. V, Roma, pp. 1796-1832. Frangipane M., Hauptmann H., Liverani Vatandoust A., Parzinger H., Helwing B. (eds.) M., Matthiae P., Mellink M. (eds.), Between 2011, Early Mining and Metallurgy on the Western the Rivers and Over the Mountains. Archaeologica Central Iranian Plateau: The First Five Years of Work Anatolica et Mesopotamica Alba Palmieri Dedicata, (Archäologie in Iran und Turan 9), Mainz am Rhein. Roma, pp. 201-219. Wilkinson T.C. 2014, Tying the Threads of Eurasia: Schmidt K. 2006, Sie bauten die ersten Tempel. Das rät- Trans-regional Routes and Material Flows in selhafte Heiligtum der Steinzeitjäger, Munich. Transcaucasia, Eastern Anatolia and Western Central Sherrat A. 2004, Trade Routes: Growth of Global Asia, c. 3000-1500BC, Leiden. Trade, «ArchAtlas, Version 4.1», http://www. Wilkinson T.C. 2016, Steps Toward the Study of archatlas.org/Trade/WEurasia.php, accessed: 18 Seasonality and Trade, «ArchAtlas, Version 4.1», December 2017. http://www.archatlas.org/occpaper/Wilkinson.php, Siracusano G., Bartosiewicz L. 2012, Meat accessed: 18 December 2017. Consumption and Sheep/Goat Exploitation in Yener K.A. 2000, The Domestication of Metals: The Rise Centralised and Non-Centralised Economies at of Complex Metal Industries in Anatolia (Culture and Arslantepe, Anatolia, «Origini» 34, pp. 111-123. History of the Ancient Near East 4), Leiden. Siracusano G., Palumbi G. 2014, “Who’d be Happy, Zimmermann Th. 2007, Anatolia as a Bridge from Let Him Be so: Nothing’s Sure about Tomorrow”. North to South? Recent Research in the Hatti Discarded Bones in an Early Bronze I Elite Area at Heartland, «Anatolian Studies» 57, pp. 65-75. Arslantepe (Malatya, Turkey): Remains of Banquets?, in: Bieliński P., Gawlikowski M., Koliński Zimmermann Th. 2009, Frühmetallzeitliche Eliten R., Ławecka D., Sołtysiak A., Wygnańska Z. zwischen Ostägäis und Taurusgebirge im 3. Jahrtausend , (eds.), Proceedings of the 8th International Congress on v. Chr. – Versuch einer kritischen Bestandsaufnahme in: Egg M., Quast D. (eds.), the Archaeology of the Ancient Near East (30 April, 4 Aufstieg und Untergang. May 2012), Vol. 2: Excavation and Progress Reports, Zwischenbilanz des Forschungsschwerpunktes „Studien Posters, Wiesbaden, pp. 349-365. zu Genese und Struktur von Eliten in vor- und frühge- schichtlichen Gesellschaften„ (Monographien des Stein G.J. 1999, Rethinking World-Systems: Diasporas, Römisch-Germanischen Zentralmuseums 82), Mainz, Colonies, and Interaction in Uruk Mesopotamia, pp. 1-29. Tucson. Zimmermann Th. 2011, Legal Aliens on Hattian Tonussi M. 2007, Dall’Eufrate allo Scamandro: contat- Grounds? – Tracing the Presence of ‘Foreigners’ in 3rd ti e scambi nel III millennio a.C., Padova. Millennium Central Anatolia, «Anodos» 10 (2010), Tosi M. 1974, The Lapis Lazuli Trade across the Iranian pp. 335-340. Plateau in the 3rd Millennium, in: Di Napoli F.

Webliography http://www.arcane.uni-tuebingen.de/, ARCANE project: Associated Regional Chronologies for the Ancient Near East and the Eastern Mediterranean (accessed: 10 December 2017). http://www.oeaw.ac.at/sciem2000/, SCIEM 2000: The Synchronization of Civilizations in the Eastern Mediterranean in the 2nd Millennium BC (accessed: 10 December 2017). http://www.icaane.net/, International Congress on the Archaeology of the Ancient Near East (ICAANE) (accessed: 10 December 2017). http://www.archatlas.org/Home.php, ArchAtlas, Version 4.1 (accessed: 18 December 2017).

West & East 22 Monografie, 4 A territorial model of animal husbandry for the southern Caucasus: some preliminary results

BEATRICE BARBIERO

Ca’ Foscari University of Venice

Abstract

Drawing on the first-hand analysis of faunal data from recent excavations at Aradetis Orgora (Georgia) and on their comparison with previously published data from selected sites in Georgia, Armenia and Eastern Turkey, the paper attempts at providing an insight of animal husbandry during the Bronze Age (with a special attention to the Middle and Late Bronze Ages) in the Southern Caucasus. The huge amount of animal bones recovered from those sites includes in each of them Ovis/Capra, Bos and Sus, but they apparently show different kill-off pattern and different forms of animal exploitation. The importance of faunal remains in identifying different types of pastoralism and herding strategies and, more in general, in understanding behaviours and decision- making process of prehistoric cultures, but also the need for more comprehensive sampling and better contex- tualisation of the finds are emphasised.

Keywords

Southern Caucasus, Bronze Age, zooarchaeology, Georgia, animal breeding

West & East 23 Monografie, 4 Beatrice Barbiero

1. Introduction by. These people are usually engaged in agriculture.5 Transhumant pastoralists use specifically selected In the Ancient Near East, during the Bronze Age, herdsmen for this task, to migrate with herds in or- different forms of pastoralism were performed, der to exploit seasonal pastures at their productive which were the expression of, or were expressed by, peak. Usually these people stay in the mountains different kinds of societies. Despite the importance during spring and summer, and in the lowlands dur- of this topic for defining the socio-economic foun- ing fall and winter.6 There are also cases, however, dations of ancient societies, researches based on where the seasonal migration involves the whole a first-hand analysis of the most relevant data (i.e. community. Finally, there is nomadic pastoralism7: animal bones remains from archaeological excava- this is a highly specialised activity practiced by pure- tions) are still relatively few and not homogeneous- ly nomadic societies. All or almost all the popula- ly distributed. Zeder (1991), Vila (1998), Arbuckle tion moves with the herds from place to place in (2006) and Berthon (2011) are worth mentioning search of pasture throughout the seasons, living in among the main archaeozoological syntheses contemporary encampments and often travelling very cerning wider regions within the Ancient Near East. far over the course of a year. Usually these groups Following the recent development of international don’t practice agriculture.8 field research in Georgia, Armenia and Azerbaijan, Many factors make communities choose one the Southern Caucasus has also been fully involved kind of pastoralism above others, for instance: the in this field of research.1 location of the settlement, the presence or absence In this paper, the results of the study carried out of a central power, and the specific strategies and by the author on the occasion of her MA thesis2 on aims of animal husbandry.9 The latter are to be in- part of the faunal remains from the site of Aradetis tended in the sense of the prevalent use of animals: Orgora (Georgia) are presented and compared with as food resource, as sources of raw materials (wool, the data from selected sites of the Southern Cauca- leather, horns) for handicraft, for traction (in the sus and Upper Mesopotamia, such as Didi Gora in case of cows, donkeys or horses), for hunting (in the Georgia, Gegharot, Tsaghkohovit in Armenia, Ar- case of dogs and horses), or as value.10 slantepe and some settlements of the Upper Tigris area in Turkey, in a preliminary attempt to understand the animal economy choices of this highland 2. Archaeological context region of the Near East and their development. and taxonomic composition The term “pastoralism” in its most basic sense of the analysed sample means the herding of domestic animals such as cat- 3 tle, sheep, goats or pigs. Anthropologists generally The archaeological area of Aradetis Orgora in the further categorise pastoralist societies according to Shida Kartli province of Georgia includes three two criteria: how settled they are, and how much mounds and a burial ground (fig. 1). The main 4 they engage in agriculture in addition to herding. mound (Dedoplis Gora) dominates the Kura Riv- There are sedentary pastoralists, who live in per- er valley from up 34 m on the present level of the manent settlements, using stables to keep animals, Western Prone, one of the main tributaries of the and feed their herds year-round on pastures near- latter. On the top of the hill an impressive fortified

5 Piro 2009, p. 4; Abdi 2003, p. 400. 1 Gilbert 1979; Uerpmann, Uerpmann 2008; 6 Piro 2009; Badalyan et Al. 2008, 2016; Monahan 2012; Piro 2009, p. 5. Berthon 2013, 2014. For a synthesis of Soviet-period re-� 7 Cribb 1991; Honeychurch, Makarewicz 2016. search of the topic, see also Kushnareva 1997. 8 Bar-Yosef, Khazanov 1992, p. 2; Marshall, 2 Barbiero 2017. Capriles 2014, pp. 249-251. 3 Bar-Yosef, Khazanov 1992, p. 2. 9 Piro 2009, p. 62. 4 Piro 2009, p. 4; Marshall, Capriles 2014, p. 249. 10 Vila 1998, pp. 126-127.

West & East 24 Monografie, 4 A territorial model of animal husbandry for the southern Caucasus: some preliminary results

Figure 1 Map of Georgia with location of the Aradetis Orgora site (afterGagoshidze, Rova 2015, fig. 1, p. 5)

palatial building of the Hellenistic/Early Roman Field A, from which the analysed corpus of ani- period defines one of its most recent phases of oc- mal bones derives, is located on the south-western cupation. In fact, the total depth of the archaeologi- slope of the mound. The recovered stratigraphical cal levels measures at least 14 metres and consists of sequence reached the base of Middle Bronze Age an almost continue sequence from the late fourth occupation and was preliminarily divided into six- millennium BC to the Early Medieval period (sixth teen main levels, where the Late Bronze/Early Iron century AD). to Middle Bronze Age sequence is represented by In order to clarify the importance and the exten- Levels 8 to 16.12 sion of the Aradetis Orgora settlement in its differ- The assemblage presented in this study is com- ent periods of occupation and to obtain a stratified posed by 6801 fragments of animal bones, which sequence of artefacts and ecofacts, two stratigraph- were recovered in two different groups ofloci . The ic soundings (Field A and Field B) were opened first group represents the filling of open areas of the in 2013-2016 on the opposite sides of the Main Late Bronze period (spaces 1630/1776 of Levels Mound by the Georgian-Italian Shida Kartli Ar- 9 and 10) (fig. 2), which were presumably used as a chaeological expedition of Ca’ Foscari University 12 of Venice in collaboration with the Georgian Na- I refrain from attributing specific levels to precise cultural phases as there is no general agreement, until now, on tional Museum of Tbilisi headed by I. Gagoshidze the archaeological periodisation of the second half of the sec- and E. Rova.11 ond/early first millennium BC in the Southern Caucasus and particularly in Georgia (see most recently Sagona 2018, 11 Gagoshidze, Rova 2015, 2018, in press. pp. 378‑422).

West & East 25 Monografie, 4 Beatrice Barbiero

Figure 2 Field A, plan of Level 10 occupation, earlier stage (Georgian-Italian Shida Kartli Archaeological Project)

junkyard mainly for the disposal of butchering activi- The second group represents finds from the deep ties.13 These spaces had no formal floor and were filled sounding (Levels 11-16) (fig. 3), which date to the with successive layers of grey to dark brown sediments Late/Middle Bronze transition and to the Middle including charcoals and ashes, which contained large Bronze Age. In particular, bones from the Middle animal bones and other mixed waste: pottery sherds, Bronze levels derive from different levels of ephem- fragments of flint and obsidian, and wood remains. eral occupation, characterised by the presence of post-holes, remains of thin wattle-and-daub walls, 13 Barbiero, Rova, in press. pits and other cut-in features (possible firing instal-

West & East 26 Monografie, 4 A territorial model of animal husbandry for the southern Caucasus: some preliminary results

Figure 3 Field A, east section of the deep sounding, from W (Georgian-Italian Shida Kartli Archaeological Project)

lations, bases of cut-in huts etc.), and from thick ac- (26.7%) are only recognisable either anatomically cumulations of muddy soil. These levels were also or taxonomically. 53,8% of the fragments are non- quite rich in disposed animal bones, which makes identified, a category which includes both very small the comparison with the Late Bronze assemblages and poorly preserved fragments. mentioned above especially interesting for the aim The following species were encountered:Ovis of highlighting possible diachronic changes in hus- aries (domestic sheep), Capra hircus (domestic bandry practices. goat)14, Sus scrofa (domestic pig), Equus sp., Canis Within the whole assemblage, 1319 bones sp., Cervus elaphus (red deer), Capreolus capreolus (19.3%) were determined both from the taxonomic 14 Ovis aries and Capra hircus are included in the mac- and from the anatomic point of view, whereas 1816 ro-category of caprine if it was not possible to determine the taxa.

West & East 27 Monografie, 4 Beatrice Barbiero

Table 1 – Number of identified, recognisable and un-determined fragments in the analysed assemblage

Middle Bronze Transition Late Bronze

Identified 631 31 657

Recognizable Aves indet. 17 0 9 Pisces indet. 6 0 2 Small size 60 0 117 Medium size 466 27 461 Big size 387 18 279 Total rec. 914 45 857

Non-identified 1994 220 1446

Total 3545 296 2960

(roe deer), Rupicapra rupicapra (chamois), Castor share of remains belongs to domestic animals, with fiber(beaver), Lepus europeus (hare), and Rattus/Ar- a significant majority ofSus scrofa, Bos taurus and vicola15 (mouse).16 caprines, whereas wild animals are present in small Tab. 1 shows all recognisable fragments in the proportion and in a casual way. Hunted taxa in- assemblage, while tab. 2 develops taxonomic com- clude Cervus elaphus as the main game on the site. position at Aradetis Orgora based on NISP (Num- The main skeletal elements are parts of the cra- ber of Identified Specimen) and Tab. 3 does the nium and limb extremities in varying proportions. same but based on MNI (Minimum Number of This suggests a first work on carcasses on site and a Individuals).17 In both groups of loci, the highest following transport of the entire animals to other places for proper butchery work on them focusing 15 The small assemblage of mouse bones derives from se- on the meatier skeletal elements. This is true for the lected loci, whose sediments were dry-sieved or floated. whole triad of domestic animals found at Aradetis 16 Barone 1995a, b; Schmid 1972. Orgora and over the whole stratigraphical sequence 17 In this study both the NF (Number of Fragments) and (which guarantees a good comparability of the two the MNI (Minimum Numbers of Individuals) were used, fol- groups of finds), but especially in contexts belong- lowing the studies of White 1953, Ringrose 1993, Lyman 2008, and De Grossi Mazzorin 2008. ing to its Late Bronze portion.

West & East 28 Monografie, 4 A territorial model of animal husbandry for the southern Caucasus: some preliminary results

Table 2 – Taxonomic composition of the analysed assemblage based on NISP

Taxonomic composition of the analysed assemblage based on NISP

Middle Bronze % Transition % Late Bronze %

Domestic animals

Bos taurus 156 24,5% 8 25,8% 271 41%

Caprines 299 11 119

Capra hircus (20) 47,3% (0) 35,4% (13) 18,1%

Ovis aries (55) (4) (32)

Sus scrofa fd 72 11,3% 9 29% 226 34,3%

Equus sp. 16 2,5% 0 8 1,2%

Canis sp. 3 0,6% 0 3 0,4%

Subtotal 545 86,3% 28 90,3% 626 95,2%

Wild animals

Cervus elaphus 11 1,8% 0 25 3,9%

Capreolus capreolus 0 1 3,2% 1 0,1%

Rupicapra rupi- 0 0 1 0,1% capra Castor fiber 1 0,1% 0 0

Lepus europeus 0 1 3,2% 0

Rattus/Arvicola 73 11,5% 1 3,2% 3 0,4%

Subtotal 86 13,6% 3 9,6% 31 4,7%

Total 631 100% 31 100% 657 100%

West & East 29 Monografie, 4 Beatrice Barbiero

Table 3 – Taxonomic composition of the analysed assemblage based on MNI

Taxonomic composition of the analysed assemblage based on MNI

Middle Bronze % Transition % Late Bronze %

Domestic animals

Bos taurus 3 12% 1 25% 9 21,4%

Caprines 10 1 8

Capra hircus (4) 40% (0) 25% (4) 19%

Ovis aries (6) (1) (4)

Sus scrofa fd 5 20% 1 25% 13 30,9%

Equus sp. 2 8% 0 1 2,3%

Canis sp. 1 4% 0 1 2,3%

Subtotal 21 84% 3 75% 32 76,1%

Wild animals

Cervus elaphus 1 4% 0 2 4,7%

Capreolus capreolus 1 4% 0 0

Rupicapra rupi- 0 0 1 2,3% capra Castor fiber 1 4% 0 0

Lepus europeus 0 1 25% 0

Rattus/Arvicola 1 4% 0 7 16,6%

Subtotal 4 16% 1 25% 10 23,8%

Total 25 100% 4 100% 42 100%

West & East 30 Monografie, 4 A territorial model of animal husbandry for the southern Caucasus: some preliminary results

Figure 4 Histogram showing mortality data based on dental and bones ages during Middle and Late Bronze

West & East 31 Monografie, 4 Beatrice Barbiero

The general trend in relative frequency sug- 3. Aradetis Orgora in a wider context gests a gradual decrease, in the course of time, in the proportion of sheep and goats, matched by a In the following, I will present the results of a first corresponding increase of cattle and, more impor- comparison of the data from Aradetis Orgora with tant, of pigs. those from other Bronze Age sites from the South- Mortality data18 (fig. 4), instead, yield two rath- ern Caucasus and from Upper Mesopotamia. The er similar results, which may indicate that animals sites were selected for the availability of good fau- were managed in a similar way in both periods. nal analyses, for the proximity, in time and space, to Ovis/Capra mortality has a peak between the first Aradetis Orgora, or as they provide a good example and the second year and, mostly, after the third of the cultural dynamics between highland and low- year. Cattle were killed after the third year in 80% land populations. In very general terms, all of them of the cases. Although during the Late Bronze this had a similar animal husbandry economy, dominat- percentage slightly decreases (to 70%), this picture ed by the triad of animals composed of cattle, pigs suggests for both periods a use of these animals for and caprines. dragging wagons or in agricultural works. Other- wise, both Ovis/Capra and cattle were slaughtered in adult age, thus showing both interest in prod- 3.1. Didi Gora (Eastern Georgia) ucts such as milk and wool, and in meat of good quality. However, during the Late Bronze the Didi Gora is a settlement in the Alazani valley of community also consumed the fat meat of pigs, Eastern Georgia.20 According to the excavators, which were killed between the first and the second the site was occupied from the Early Bronze Age year of life. to the Iron Age, probably by a transhumant popu- However, what changes with time is the pres- lation in the earlier stage, and then by a more sed- ence or absence of some specific ranges of age: in entary community. the Middle Bronze there are just a few juveniles or Animal bones used in this comparison are dat- newborns animals, thus suggesting the absence of ed to the Middle and Late Bronze, and therefore people at Aradetis during some seasons19 (a picture compare well, from the chronological point of view, which may be compatible with transhumant pas- with those from Aradetis Orgora. According to the toralism). In general terms, the results of the fau- analysis results, husbandry at Didi Gora was mostly nal analysis of the Aradetis Orgora Late Bronze focused on cattle (72.7% during the Middle Bronze, sample, showing a growing number of pigs and the 55.5% in Late Bronze), breeding animals both presence of caprines during the whole year, would for milk (15.8%, with 10% of new-borns slaugh- thus support traditional interpretations of the de- tered) and meat. The same is valid for Ovis/Capra velopment of South-Caucasian society from the husbandry (with 20% of animals killed in juvenile Middle to the Late Bronze Age, which would be age). Pigs, instead, are very few on this site (2.1 % in characterised by a renewed trend toward seden- the Middle Bronze, and 1.2% in the Late Bronze), tarisation and reduced mobility. which was tentatively explained with the effort of raising cows as the main meat resource and with the fact that pigs are very harmful to agriculture, which was an important economic activity at the site. The number of pigs is actually an element of 18 Age was determined by confronting age fusion of an- strong contrast with Aradetis Orgora, where they cient bones with the ones reported on Barone 1995 studies. represent more than one third of the faunal assem- For dental remains, age classes were verified following pro- blage, while at Didi Gora their amount is just above cedures described by different authors:Grigson 1982 and Grant 1982 for bovines (MWS, Mandibular Wear Stage), 1%. The difference between these two nearly con- Payne 1973 for caprines, and Bull, Payne 1982 for pigs. 19 Connor 1998. 20 Uerpmann, Uerpmann 2008.

West & East 32 Monografie, 4 A territorial model of animal husbandry for the southern Caucasus: some preliminary results

temporary settlements appears, therefore, to lie in Sus scrofa at the two Armenian sites. This fact could the choice of the main source of meat: pigs in one be explained with the mortality data of cattle: in case, and cattle in the other one. It remains diffi- Tsaghkahovit, cows were butchered more in juve- cult to explain since that both settlements are lo- nile than in adult age, showing a prevalent choice cated in roughly comparable natural environments in primary products for theseanimals. Cattle was in and in the same general region. Otherwise, also at fact the main source of meat here, as it happened Didi Gora cattle and caprines were killed mostly in at Didi Gora. Another interesting point is the high adult age, showing an economy choice for secondary percentage of Ovis/Capra at the two Armeniam product, such as wool, milk and the use of cattle for sites, which contrasts with both Aradetis Orgora traction activities. and Didi Gora, both of which show constantly diminishing percentage of caprines from the Middle to the Late Bronze Age. According to the excava- 3.2. Gegharot and Tsaghkahovit (Armenia) tors, as we previously mentioned, it would be possible to relate this high percentage of caprines to Gegharot and Tsaghkahovit are located atop two the contemporary development of central political hills, both with evidence of occupation since the authorities in Armenia, heralded through the con- Early Bronze Age, in the Tsaghkahovit Plain in cen- struction of fortresses.23 tral western Armenia.21 During the Late Bronze, both are stable agropastoral settlements with fortified hilltops and an adjacent lowland settlement. 3.3. Upper Mesopotamian sites They yielded evidence for ritual activities and met- allurgic production, and have been associated with Only few remarks can be made concerning the Up- emergent institutions. per Tigris and Upper Euphrates sites in South-east- Faunal remains were recovered from the cita- ern Turkey. Most faunal analyses from this area are dels as well as from the low settlements: all of them from Early Bronze settlements at sites which had were analysed by B. Monahan.22 Over three thou- previously been subjected to the influence of the sand bones were identified at each site for the Late South-Mesopotamian Uruk culture, and any com- Bronze Age, with only negligible differences be- parison will therefore be biased by this chronologi- tween them. Both yielded a large majority of sheep cal difference. In spite of this shortcoming, they still and goat (between 53.4% and 56.1%,), between can offer a useful comparative perspective. 31.8% and 33.7% of cattle and around 1% of pigs. The well-known site of Arslantepe has excellent Data on mortality shows animals being killed most faunal data, based on the studies by Bökönyi, Bar- of the time in adult age (around 30%), thus expos- tosiewicz and Siracusano24, which reflect the socio- ing a focus on wool, milk and the use of the animals political and economic transformations that took in agriculture. This husbandry choice was connect- place during the second half of the fourth and the ed by the excavators both with transhumant pas- beginning of the third millennium BC. The settle- toralism and with a central power, that controlled ment is located in the Turkish Upper Euphrates ba- the economies of the surrounding sites. The similar- sin, at the crossroads of different highland and low- ity between Gegharot and Tsaghkahovit can be ex- land “worlds” – the South-Caucasian, the Central plained through their proximity and the similar set- Anatolian and the Mesopotamian –, and is there- tlement structure. fore open to influences from all of them. As it happened with Didi Gora, the main dif- During phase VIA (3350-3000 BC), when the ference in breeding choices with Aradetis Orgora site was under strong influence by the southern concerns pigs, with a constantly small number of 23 Badalyan et Al. 2008, p. 94. 21 Smith 2012, p. 47; Badalyan 2008, p. 50. 24 Bökönyi 1983; Palumbi 2010, 2012; Siracusano, 22 Badalyan et Al. 2008, 2016; Monahan 2012. Bartosiewicz 2012; Siracusano, Palumbi 2014.

West & East 33 Monografie, 4 Beatrice Barbiero

Uruk civilisation, the emphasis on sheep and goat 4. Discussion herding (75%, particularly sheep) reflects a kind of herding that took place under centralised con- From the comparison between the settlement of trol. Despite a slight decrease, high proportions Aradetis Orgora and the selected contemporary of caprines are still evident in the subsequent VIB sites it is not possible to distinguish clear territori- phase (3000 – 2800 BC), which shows a large Kura- al trends concerning animal husbandry choices or Araxes component instead. This suggested that the clear chronological trends valid for all the studied focus on caprine herding was not easily reversed af- highland regions. ter the collapse of the central power. Besides, the As anticipated, the triad of domestic animals mortality profile ofOvis/Capra showed a high per- leads the economy of all the analysed settlements, centage of sub-adult and adult kill-off, which was but their relative frequencies, besides being proba- interpreted as indicative of a meat-maximising and bly influenced by local factors, appear to reflect dif- wool production strategy.25 ferent models of animal management, where, for Data from the Upper Tigris sites dated to the instance, an emphasis on caprines may be tenta- Bronze and Iron Age (Giricano, Bashur Höyük or tively associated with more centralised economies Kenan Tepe for example) analysed by Berthon26, in- and influences from the south. Research has in fact stead, show a considerable variety in the percentages shown that a dominance of caprines in husbandry of animal presence, which are different from place choices is quite characteristic, since the fourth mil- to place and from time to time, and suggest the ex- lennium BC, of the centralised economies of the ur- istence of diversified husbandry economy choic- ban civilisations of Greater Mesopotamia28, where- es. Global skeletal representations tend to suggest as northern highland communities characterised by that the studied material comes from a collection of less hierarchical forms of socio-political organisa- butchery, kitchen and consumption refuse contexts. tion generally tend to adopt more balanced strate- Concerning the “triad” (caprine, cattle and pigs), gies of animal management, unless under stimuli all taxa are almost equally represented at every site from their southern neighbours.29 However, huge analysed by the author, with cattle and ovis/capra numbers of cattle could also be indicative of the slaughtered mostly after two years, thus likely with presence of a centralised power or authority, as for aims other than meat (i.e. milk and wool). Due to instance in Bronze Age Anatolia.30 the inter-site variability within each chronological The leading role of pig husbandry at Late Bronze period, it was difficult to estimate a precise pattern Age Aradetis Orgora, which is unparalleled at the of animal exploitation. However, as the author ex- other sites, then, probably reflects the choices of plains, differences between the Middle and the Late groups, whose economy was not under strong in- Bronze Age (i.e. increasing amount of caprines) may stitutional control. In fact, it has been argued that suggest that during the former period every settle- a relative abundance of pigs may reflect an animal ment was politically and economically independ- economy beyond institutions because, among oth- ent, whereas during the latter political control was ers, these animals reproduce rapidly and provide a stronger on the area, focusing on textile produc- large amount of meat and fat.31 tion and therefore on caprines husbandry management. Otherwise, each site appears to have been influenced also by the proximity of agricultural fields, hilly areas and by cultural considerations.27

28 Zeder 1991. 29 See Frachetti 2008; Hammer 2014; Zeder 1991 and, with specific reference to the case of Zeytinli Bahçe (Urfa 25 Frangipane, Siracusano 1998, p. 243. region, Turkey), Siracusano 2005. 26 Berthon 2011. 30 Arbuckle 2014, p. 286. 27 Berthon 2011, p. 190. 31 Price et Al. 2017, p. 57.

West & East 34 Monografie, 4 A territorial model of animal husbandry for the southern Caucasus: some preliminary results

5. Conclusions tle and caprine husbandry.33 For sure, at all the sites researched for this study, Aradetis Orgora included, In general terms, the results of the faunal analy- cattle and caprine may have played an important sis of the Aradetis Orgora sample would be com- role in providing secondary products such as wool, patible with traditional interpretations of the de- dairy products and probably traction, while pigs velopment of South-Caucasian society from the mainly yielded primary products.34 In the analysed Middle to the Late Bronze Age. In fact, according assemblage from Aradetis Orgora, for instance, to the most widespread theories, during the Mid- mortality based on epiphyseal fusion and tooth wear dle Bronze, South-Caucasian societies should have data seems to be compatible with a secondary prod- been transhumant or nomadic32, whereas the Late uct animal management strategy, with cattle and Bronze Age would be characterised by a renewed caprine slaughtered in adult age, focusing on their trend toward sedentarisation and reduced mobility. capacity for traction and dairy production, while a This seems to be true for Aradetis Orgora as well, meat-oriented strategy was adopted for pigs, killed based both on the lack of certain ranges of animals before the third year. This husbandry choice seems age, such as new-borns and juveniles, an important to be rather common among non-centralised com- amount of parts of the cranium and limb extremi- munities of the highland regions of the Near East.35 ties that may suggests only a first work on carcasses It is known that cattle, pigs and caprine require on site, and on the archaeological evidence, which different herding strategies and are raised for par- shows cut-in features and flimsy remains of wat- tially different aims. Factors such as mobility, as tle-and-daub constructions, i.e. traces of a rather well as availability, accessibility, and predictability ephemeral occupation, in the Middle Bronze levels. of high-quality grazing pastures, water and fodder During the Late Bronze Age, in contrast, the Ara- must have impacted the degree and scale of cattle detis community appears to have become more set- herding.36 A similar strategy is needed for an opti- tled, with a growing number of pigs and cattle and mal caprine herding, that involves the movement of with more substantial architecture on site. On the animals between seasonally available grazing areas.37 other hand, to the present stage of research, avail- On the other hand, pigs provide substantially more able data from other sites of the Southern Caucasus meat than sheep and goats, but none of the dairy and Upper Mesopotamia provide a much more crit- products cattle and caprine give. Plus, pigs require ical insight into the multifaced nature of pastoral- plenty of water, and an open wooded area with shel- ism, which does not suggest general chronological ter from the sun.38 trends. In fact, analysis data do not always coincide Though, there are still too few good faunal with expectations and previous hypothesis about analyses concerning larger faunal assemblages animal exploitation: even if the triad of animals from other sites, such as would be needed for an in- chosen (cattle, pigs and caprine) is always the same depth analysis of herd mortality and demographic at all examined sites, and so are the chosen aims of patterns, and these are often far away from each animal husbandry, herd management and husband- other in space and/or time, a fact which makes it ry choices at individual sites are more complex and difficult to appreciate the relative importance of heterogeneous than one could think at first sight. environmental versus chronological factors in ex- First of all, it must be emphasised that the num- plaining the differences between them. Plus, more bers of cattle and sheep bones do not necessarily attention to spatial analysis is needed in the fu- reflect the true economic value of these species or their caloric contribution to the ancient diet. As 33 Price 2017, p. 46. Price argues, zooarchaeological evidence has been 34 Sherratt 1983; Charles et Al. 1998. often concealed by pastoral biases due to the fact 35 O’Connor 1998, p. 5. that ancient records gave more importance to cat- 36 Atici et Al. 2017, p. 23. 37 Zeder 1991, p. 25. 32 Sagona 2018, p. 298. 38 Piro 2009, p. 265.

West & East 35 Monografie, 4 Beatrice Barbiero

ture in order to highlight functional differences in sites will allow to uncover the full range of pasto- animal use and animal bones deposition patterns ral activities practiced in the Southern Caucasus, within each settlement39, a fact which may heavily Eastern Anatolia and Upper Mesopotamia during bias the analysis results. the Bronze Age, to distinguish chronological from In fact, only the availability of a much larger se- geographical and functional aspects, and to sketch lection of bone assemblages from archaeological a general history of pastoral practices in the region.

39 Barbiero, Rova in press.

West & East 36 Monografie, 4 A territorial model of animal husbandry for the southern Caucasus: some preliminary results

BIBLIOGRAPHY

Abdi K. 2003, The Early Development of Pastoralism in Berthon R. J.-P. 2011, Animal Exploitation in the the Central Zagros Mountains, «Journal of World Upper Tigris River Valley (Turkey) between the 3rd Prehistory » 17/4, pp. 395-448. and the 1st Millennia BC, Kiel. Arbuckle B. 2006, The Evolution of Sheep and Goat Berthon R. J.-P. 2013, New Data on the Exploitation Pastoralism and Social Complexity in Central of Animal Resources in the Upper Tigris River Area Anatolia, Harvard. (Turkey) during the Second and First Millennia Arbuckle B. S. 2014, The rise of cattle cultures in Bronze BC, in: De Cupere B., Linseele V., Hamilton- Age Anatolia, «Journal of Eastern Mediterranean Dyer S. (eds.), Archaeozoology of the Near East X. Arcaheology and Heritage Studies» 2/4, pp. 277-297. Proceedings of the Tenth International Symposium on the Archaeozoology of South-Western Asia and Atici L., Birch S.E.P., Erdoğu B. 2017, Spread of Adjacent Areas, Paris, pp. 145-162. Domestic Animals across Neolithic Western Anatolia: New Zooarchaeological Wvidence from Uğurlu Höyük, Berthon R. J.-P. 2014, Past, Current and Future the Island of Gökçeada, Turkey, «Porta Linguarum Contribution of Zooarchaeology to the Knowledge Orientalium. Neue Serie» 12, pp. 1-30. of the Neolithic and Chalcolithic Cultures in South Caucasus, in: Mindiashvili, G. (ed.), Studies in Badalyan R., Smith A. T., Lindsay I., Caucasian Archaeology, Tbilisi, pp. 4-33. Khatchadourian L., Avetisyan P. 2008, Village, Fortress, and Town in Bronze and Iron Age Bökönyi S. 1983, Late Chalcolithic and Early Bronze I Southern Caucasia: a Preliminary Report on the Animal Remains from Arslantepe (Malatya), Turkey: 2003–2006 Investigations of Project ArAGATS on the A Preliminary Report, «Origini» 12/2, pp. 581-598. Tsaghkahovit Plain, Republic of Armenia, «AMIT» Bull, G., Payne, S., 1982, Tooth Eruption and 40, pp. 45-105. Epiphysial Fusion in Pigs and Wild Boar, in: Wilson, Badalyan R., Smith A.T., Lindsay I., B., Grigson, C., Payne, S. (eds.), Ageing and Sexing Harutyunyan A., Greene A., Marshall M., Animal Bones from Archaeological Sites, Oxford, Monahan B., Hovsepyan R. 2016, A Preliminary pp. 55-71. Report on the 2008, 2010, and 2011 Investigations Charles M. Halsted P., Jones G. 1998, The Archaeo of Project ArAGATS on the Tsaghkahovit Plain, logy of Fodder, «Environmental Archaeology» 1, Republic of Armenia, « AMIT» 46, pp. 149-222. pp. 1-15. Bar-Yosef O., Khazanov A. 1992, Pastoralism in the Cribb R. 1991, Nomads in Arcaheology, Cambridge. Levant. Archaeological Materials in Anthropological De Grossi Mazzorin J. 2008, Archeozoologia: lo stu- Perspectives, Madison. dio dei resti animali in archeologia, Bari Barbiero B. 2017, Scelte di sussistenza nella Frachetti M. D. 2008, Pastoralist Landscapes and Transcaucasia dell’Età del Bronzo: analisi paleofau- Social Interaction in Bronze Age Eurasia, London. nistica di loci selezionati del sito di Aradetis Orgora, Frangipane M., Siracusano G. 1998, Georgia, unpublished MA dissertation, Università Changes Ca’ Foscari Venezia, a.a. 2016-2017. in Subsistance Strategies in East Anatolia during the 4th and 3rd Millennium BC, in: Anreiter P. Barbiero B., Rova E. (in press), Remains from Butchery Bartosiewicz L., Jerem E., Meid W. (eds.), Man Activities from Late Bronze Age Contexts at the and the Animal World. Studies in Archaeozoology, Aradetis Orgora Site (Georgia, Southern Caucasus), Archaeology, Anthropology and Palaeolinguistics in «Origini» (under revision). Memoriam Sándor Bökönyi, Budapest, pp. 237-246. Barone R. 1995a, Anatomia comparata dei mammiferi Furtwängler A., Gagoshidze I. 2008, Iberia and domestici. Vol. 1 Osteologia, Bologna. Rome: The Excavations of the Palace at Dedoplis Gora Barone R. 1995b, Anatomia comparata dei mammiferi and the Roman Influence in the Caucasian Kingdom of domestici. Vol. 3 Splancnologia, Bologna. Iberia, Langenweißbach.

West & East 37 Monografie, 4 Beatrice Barbiero

Gagoshidze I., Rova E. 2015, Two Seasons of Late Bronze Age Southern Caucasia: Preliminary Georgian-Italian Excavations at Aradetis Orgora Results from a Pottery Provenance Study in (Georgia), «Rivista di Archeologia» 39, pp. 5-28. Northwestern Armenia, «Journal of Archaeological Gagoshidze I., Rova E. (2018), 2013-2015 Activities Science» 35, pp. 1673-1682. of the Georgian-Italian Shida Kartli Archaeological Marshall F., Capriles J. M. 2014, Animal Project at Aradetis Orgora (Georgia), in: Horejs B., Domestication and Pastoralism: Socio-Environmental Schwall C., Müller V., Luciani M., Ritter Contexts, in: Smith C. (ed.), Encyclopedia of Global M., Giudetti R., Salisbury B., Höffmayer F., Archaeology, New York, pp. 249-258. th Bürge T. (eds.), Proceedings of the 10 International Monahan B. 2012, Beastly Goods: Pastoral Production Congress on the Archaeology of the Ancient Near East in the Late Bronze Age Tsaghkahovit Plain, in: 25–29 April 2016, Vienna, Volume 2, Wiesbaden, Hartley C., Bike Yazicioğlu G., Smith pp. 497-510. A.T. (eds.), The Archaeology of Power and Politics Gagoshidze I., Rova E. (in press a), New Investigations in Eurasia: Regimes and Revolutions, Cambridge, at Aradetis Orgora, a Multiperiod Centre of the pp. 337-347. Shida Kartli Region in Georgia, in: Batmaz A., O’Connor T. P. 1998, On the Difficulty of Detecting Bedianashvili G., Michalewicz A., Robinson Seasonal Slaughtering of Sheep, «Environmental A. (eds.), Context and Connection: Essays on the Archaeology» 3, pp. 5-11. Archaeology of the Ancient Near East in Honour of Palumbi G. 2010, Antonio Sagona (Orientalia Lovaniensia Analecta), Pastoral Models and Centralized Leuven. Animal Husbandry. The Case of Arslantepe, in: Frangipane M. (ed.), Economic Centralisation in Gilbert A.S. 1979, Urban Taphonomy of Mammalian Formative States. The Archaeological Reconstruction Remains from the Bronze Age of Godin Tepe, Western of the Economic System in 4th Millennium Arslantepe, Iran, unpublished Ph.D. dissertation, Columbia Roma, pp. 149-163. University NY. Palumbi G. 2012, Bridging the Frontiers. Pastoral Grant, A. 1982, The Use of Tooth Wear as a Guide Groups in the Upper Euphrates Region in the Early to the Age of Domestic Ungulates, in: Wilson, B., Third Millennium BCE, «Origini» 34, pp. 261-278. Grigson, C., Payne, S. (eds.), Ageing and Sexing Payne, S., 1973, Animal Bones from Archaeological Sites, Oxford, Kill-off Patterns in Sheep and Goats: the pp. 91-108. Mandibles from Asvan Kale , «Anatolian Studies» 23, pp. 281-303. Grigson, C. 1982, Sex and Age Determination of Some Piro J. J. 2009, Bones and Teeth of Domestic Cattle: A Review of the Pastoralism in the Early Transcaucasian , Literature, in: Wilson, B., Grigson, C., Payne, Culture: The Faunal Remains from Sos Höyük Unpublished PhD Thesis, New York University. S. (eds.), Ageing and Sexing Animal Bones from Archaeological Sites, Oxford, pp. 7-23. Price M., Grossman K., Paulette T. 2017, Pigs Hammer E. 2014, Local Landscape Organization of and the Pastoral Bias: The Other Animal Economy in , «Journal of Anthropological Mobile Pastoralists in Southeastern Turkey, «Journal Northern Mesopotamia Archaeology» 48, pp. 46-62. of Anthropological Archaeology» 35, pp. 269-288. Sagona A. 2018, Honeychurch W., Makarewicz C. A. 2016, The The Archaeology of the Caucasus. From Cambridge. Archaeology of Pastoral Nomadism, «Annual Review Earliest Settlements to the Iron Age, of Anthropology» 45, pp. 341-359. Schmid E. 1972, Atlas of Animal Bones. For Khazanov A. M. 1984, Nomads and the Outside World, Prehistorians, Archaeologists and Quaternary Cambridge. Geologists, Amsterdam. Sherratt A. 1983, Kushnareva K.Kh. 1997, The Southern Caucasus The Secondary Exploitation of , «World Archaeology» in Prehistory. Stages of Cultural and Socioeconomic Animals in the Old World 15/1, pp. 90-103. Development from the Eight to the Second Millennium B.C., Philadelphia. Siracusano G. 2005, La fauna dei livelli preistorici di , in: Malerba G., Lyman R.L. 2008, Quantitative Paleozoology, Zeytinli Bahçe (Urfa, Turchia) Cambridge. Visentini P. (eds.), Atti del 4° Convegno Nazionale di Archeozoologia (Pordenone, 13-15 Novembre Lindsay I., Minc L., Descantes C., Speakman 2003), Pordenone, pp. 205-213. R. J., Glascock M. D. 2008, Exchange Patterns, Siracusano G., Bartosiewicz L. 2012, Boundary Formation, and Sociopolitical Change in Meat Consumption and Sheep/Goat Exploitation in

West & East 38 Monografie, 4 A territorial model of animal husbandry for the southern Caucasus: some preliminary results

Centralized and Non-Centralized Economies at Smith A. T. 2012, The Caucasus and the Near East, Arslantepe, Anatolia, «Origini» 34, pp. 111-123. in: Potts, D. (ed.), Blackwell Companion to the Siracusano G., Palumbi G. 2014, ’Who’d be Happy, Archaeology of the Near East, Oxford, pp. 668-686. let him be so: Nothing’s Sure about Tomorrow’. Uerpmann M., Uerpmann H.-P. 2008, Bronze and Discarded Bones in Early Bronze I Elite Area at Iron Age Animal Economy at Didi-gora and Tqisbolo- Arslantepe (Malatya, Turkey): Remains of Banquets?, gora (Kakhetia, Georgia), «AMIT» 40, pp. 167-264. in: Bieliński P., Gawlikowski M., Koliński Vila E. 1998, L’exploitation des animaux en Mésopotamie R., Ławecka D., Sołtysiak A., Wygnańska Z. aux IVe et IIIe Millénaires avant J.-C., Paris. (eds.), Proceedings of the 8th International Congress White T. E. 1953, on the Archaeology of the Ancient Near East 30 A Method of Calculating the Dietary April – 4 May 2012, University of Warsaw. Volume Percentage of Various Food Animals Utilized by , «American Antiquity» 18/4, 3: Archaeology of Fire, Conservation, Preservation Aboriginal Peoples pp. 386-398. and Site Management, Bioarchaeology in the Ancient Near East, Islamic Session, Wiesbaden, pp. 349-364. Zeder M. A. 1991, Feeding Cities. Specialized Animal Economy in the Ancient Near East, Washington.

Webliography http://venus.unive.it/erovaweb/ShidaKartliProject.html (accessed on 29.01.2018).

West & East 39 Monografie, 4

The pottery repertoire at the end of the third millennium BC as chronological marker between southern Mesopotamian and the neighbouring regions. The case-study of the Syrian Jazirah

ELOISA CASADEI

Sapienza, University of Rome

Abstract

The end of the third millennium BC in Southern Mesopotamia is characterized by a series of political events that, apparently, determined marked changes in the social structure of the area. Archaeologically, this period is still poorly represented, and a general reassessment of the available data is necessary. In the present paper, the pottery repertoire is taken as a case study for the identification of chronological markers of this period. The analysis of key pottery sequences allowed the identification of a homogenous ceramic horizon that characterized the timespan between the late Akkadian to the end of the Ur III periods. Two sub-phases are well distinguished by the presence/absence of types. The cross-dating between southern contexts and the well-know Syrian Jazirah helped in the better definition of the chronological limits of the two sub-phases and the general pottery phase as a whole. Nippur, Tell Asmar, Tell Brak and Tell Mozan have been used as key-sites.

Keywords

Pottery, chronology, post-Akkadian period, Ur III, Mesopotamia, Khabur region

West & East 41 Monografie, 4 Eloisa Casadei

1. The end of the third millennium BC with other regions. The high quantity of textual as an interregional problem sources had driven historians in the reconstruction of the chain of events from the decline of the Ak- In the last few years, a reconsideration of the much- kadian dynasty (2200-2120 BC) to the collapse of discussed problem of the transition between third the Ur III state. On the contrary, the archaeologi- and second millennium BC generated new signifi- cal contexts related to this period appear scattered cant interpretations and theories, pointing to a con- on the territory, and sometime affected by issues of comitance of different anthropic and climatic fac- past excavations. tors.1 The historical events that characterized this This paper focuses on the problem of the archae- period are marked by two main political phenom- ological markers that can be considered the reflec- ena, that developed starting from the second half of tion of connections linking the Southern Mesopo- the third millennium BC and resulting in a wide- tamia and the neighbouring regions, in particular spread interconnection between Southern Meso- the Syrian Jazirah (fig. 1). As one of the most ubiq- potamia and the neighbouring regions. On the one uitous elements of material culture, pottery will be hand, the political and economic expansion of the treated as a key element for recognizing interregion- Akkadian empire (2350-2200 BC) left evident trac- al interactions and investigating social dynamics es in the archaeological record.2 On the other hand, and changes. Such a perspective can help for under- the establishment of the Ur III territorial state and standing the historical significance of the intercon- the development of commercial connections connections between the different regions of the An- trolled by the Third dynasty of Ur (2120-2000 BC) cient Near East. In particular, the main focus will is testified by a wealth of epigraphic data.3 It is pos- be the identification of chronological markers that sible that the Alluvial Plain had been affected by the could help the correlation between the stratigraphy 5 gradual decline in the climate condition and agri- of sites spread in a wide territory. cultural productivity documented in the texts, and by the general reduction of the site area registered in the other Near Eastern regions.4 2. Ceramics repertoires of southern Archaeological and epigraphic data across Mesopotamia: status quaestionis Greater Mesopotamia has suggested, in many in- stances, a straight connection between the different Unfortunately, the pottery sequence of Southern regions. Nevertheless, though Southern Mesopo- Mesopotamia represents an open problem. The tamia should represent the focus of many political first attempt to study the region’s pottery chrono- events, the lack of reliable archaeological data im- typology was made by Delougaz, with his substan- pacts the reconstruction of individual site sequenc- tial work on the pottery from the excavation of the es, hampering an accurate chronological reconstruc- Oriental Institute of Chicago in the lower Diyala tion and the establishment of cross-dating elements basin during the 1930s.6 This work remained isolated for a long time, and the massive catalogue of 1 Sallaberger, Westenholz 1999; Schmidt pottery presented in Woolley’s report dealing with 2005; Kuzucuoğlu, Marro 2007; Orsi 2011; Laneri, the cemetery of Ur did not provide sufficient chro- Pfälzner, Valentini 2012; Weiss 2012; Riehl et Al. 7 2013. no-stratigraphic information. Only starting from 2 In the present paper, absolute dates follow the Middle the 1980s renewed attention on the southern Mes- Chronology as proposed lastly by Sallaberger, Schrakamp opotamian pottery provided fresh well-published 2015. For a recent reassessment of the Akkadian period see Foster 2016; Schrakamp 2017. About the Akkadian influence on the Siryan Jazira see Weiss 2012. 5 On the use of ceramic repertoire and the spread of pot- 3 For the subject see Steinkeller 1987; Schmidt tery types, several authors developed this concept (see among 2005. others Mazzoni 2000; Orsi 2011, pp. 385). 6 4 On the economic crisis at the time of Ibbisuen, see Delougaz 1952. Jacobsen 1953. 7 Woolley 1934.

West & East 42 Monografie, 4 The pottery repertoire at the end of the third millenniumBC as chronological marker…

Figure 1 Map of the area and the sites cited in the texts. The Akkadian territory has been re-elaborated by the author fromLiverani 2014, fig. 39. The Ur III territory has been re-elaborated fromSteinkeller 1987, fig. 6. The chronological table follows the Middle Chronology proposed by Sallaberger, Schrakamp 2015, tab. 39

data, thanks to the work at Abu Salabikh.8 A first fers from the paucity of good parallels with other general reassessment of the ceramic horizon of the regional sequences.10 late third to early second mill. BC was proposed by In this regard, the method proposed by Arm- Ayoub, but it suffers from the inaccuracy of the ex- strong and Gasche in 2014 for the second mill. BC cavation methods.9 Up to now, the long interrup- pottery represents a new possibility and a starting tion of the archaeological excavations in the region point for the future research. The book takes into impacted the creation of a sound ceramic periodiza- consideration all excavated sites with reliable strati- tion. Moreover, the paucity of data collected in the graphic sequences, located in a wide area encom- previous excavations make a quantitative analysis passing Iraq, western Iran and Syria.11 almost impossible. New impulses derived from the well-published materials from the WF sounding at 10 McMahon 2006. Nippur, but it still remains an isolated case and suf- 11 Armstrong, Gasche 2014. Even if the main area comprehends Southern, Central and Northern Iraq, also the 8 Moon 1987. Susiana, the Syrian Jazirah, and the Central and Western Syria 9 Ayoub 1982. are taken as parallels in the general comment.

West & East 43 Monografie, 4 Eloisa Casadei

On the contrary, the proliferation of the archae- material has been arranged, in order to identify the ological activity in the Upper Khabur region and the relative chronology of the ceramic repertoire dated Syrian Jazirah has provided a considerable advance- to the time-span between the Late Akkadian peri- ment in the definition of the chrono-stratigraphic od to the end of the Ur III period. The systematic sequences and an abundant amount of materials so analysis of all the ceramic materials had been pre- that the region can be considered a good case-study. ceded by an accurate selection of the stratigraphic A new method of analysis has been developed by contexts reliable for the purpose, according to the the international research project «Associated Re- availability of the pottery repertoire, the nature of gional Chronologies for the Ancient Near East» the contexts, and the state of publications. Table 1 (ARCANE).12 The aim of the ARCANE project is indicates sites and sequences selected in the present to synchronize the chronologies, and therefore the analysis. All the published materials from these sites histories, of the various regions of the Ancient Near have been considered in order to build a pottery ty- East and the Eastern Mediterranean.13 Such per- pology and to elaborate a chrono-stratigraphic se- spective can help in a better definition of the his- quence based on the occurrence of pottery types in torical significance of the interconnections between stratigraphic contexts. From the southern contexts, the different regions of the Ancient Near East. ca. 1200 published sherds have been considered in the analysis and organized according to a morphological typology. A cross-occurrence seriation table 3. The southern ceramic horizon was constructed for each selected site, and then they have been correlated to each other in order to identify similar trend in the development of the ceramic 3.1. Methodological premise repertoire. The final goal is the recognition of a precise ceramic horizon which describes the late third The scope of the paper is to analyze the pottery mill. Southern Mesopotamia and its inner varia- markers dated at the end of the third mill. BC at- tions. The description of the whole ceramic reper- tested both in Southern Mesopotamia and in the toire of this phase is far beyond the limit of the pre- Syrian Jazirah, to better understand the inner rel- sent paper. Here, only the chronological markers of ative chronology of that period. This objective is such phase will be described in detail, and they will part of a wider research that aims to identify the ce- be used as parallels between Southern Mesopotamia ramic traditions that characterized the final stage of and the Syrian Jazirah (fig. 1). the third mill. BC and the beginning of the second mill. BC in Southern Mesopotamia, from the lower Dyiala region to the shoreline of the Gulf. Before 3.2. Southern chronological markers of the late the analysis of the data and the presentation of the third millennium BC results, a short description of the Southern Meso- potamian pottery repertoire and the main pottery The timespan corresponding to the late Akkadian pe- markers is necessary. Here, a short note about the riod to the end of the Ur III and the very beginning methodology employed is useful to clarify the selec- of the second mill. BC is characterized by a homoge- tion of the types used in the present paper as fossils neous ceramic horizon during which the new types of this phase.14 A typological seriation of the pottery gradually appeared and disappeared without clear breaks in the relative pottery sequences.15 While the 12 http://www.arcane.uni-tuebingen.de/. 15 The problem of the definition of the chronological phase 13 Lebeau, de Miroschedji 2013, p. ix. in Southern Mesopotamia is still open and, unfortunately, still 14 I would like to thank dr. Marta D’Andrea and dr. Ag- marginal. The use of historical terms such as Early Dynastic, Ak- nese Vacca for their reading a draft of this paper and provid- kadian or Neo-Sumerian impact the construction of a diachron- ing fundamental methodological suggestions during the entire ic periodization strictly based on archaeological elements. More- course of the research. They are in no way responsible for errors over, it is difficult to compare Southern Mesopotamia to the oth- or omissions. er regions (Sallaberger, Schrakamp 2015, p. 4).

West & East 44 Monografie, 4 The pottery repertoire at the end of the third millenniumBC as chronological marker…

Table 1 Chrono-stratigraphic correlation of the Southern sites analysed in the present research, and attribution of the levels at the respective sub-phases A and B

Epigraphic Data Nippur Asmar Der Uruk

WF TB North Q Palace Acropolis Ensamble graves Sinkashid

Namarsuen/ A Sharkalisharri XIV-XI XII-IX Va-IVa Main lev. – – Grave 10 – to Shudurul

A/B X-IX IVb – – – –

Amarsuen Houses Graves B VIII-VI (?) VIII-V III-I Before B Ens IV 1 to Ibbisuen above 12 – 13

great majority of the types recognized are attested for the Ur III/IL building at Tell ed-Der nor from the the entire period, a small group helps in the identi- Sinkashid area at Uruk. A heart-shaped double ridge fication of an inner relative periodization. Two pot- rim beaker was found in Grave 10 at Uruk, dated by tery sub-phases can be identified (here called A and the excavator to the Akkadian period.17 B), each characterized by the occurrence of exclu- During sub-phase B, the types of the previous sive types. Sub-phase A is represented by the dou- groups disappeared, while others appeares now ble carinated flasks (fig. 4: 10-11), the conical beak- for the first time. Among them, the heart‑shaped ers with horizontally expanded rim (fig. 3: 4-6), beakers are now characterized by a triple-ridge rim the heart‑shaped beakers with double ridged rim (fig. 4: 7-9), and a new type of medium size storage (fig. 4: 4-6), and the triangular rim, carinated jars vessel appear, characterized by high sinuous shoul- with plain shoulder (fig. 5: 4-5). These types are at- der, everted horizontal rim and droop below the lip tested at Nippur in area WF from level XIV to VIII, (fig. 5: 8-9). These two types are attested at Nippur, area TB from XII to XI, at Tell Asmar from level Va area WF levels XI-VI and TB levels IX-III, at Tell to IVa of the Northern quarter and at the Main Level Asmar, Northern Quarter levels IVa-III and from of the Northern Palace.16 None of them is attested at the second mill. levels from the acropolis.18 They are also well attested at Tell ed-Der, Ensamble levels 16 For area WF at Nippur see McMahon 2006, types C-19, O-14, C-13b, C-16a respectively. Type C-16a is de- IVb-II, at Uruk, from level 1 of the Sinkashid palace scribed as contemporaneous of its respective C-16b – with cor- and from graves 12 and 13 of the same site.19 rugated shoulder – even if C-16a decrease significantly starting to lev. XIB (McMahon 2006, p. 73). For area TB at Nippur (http://diyala2.uchicago.edu/pls/apex/f?p=102:111::::::). see McCown, Haines 1967, pl. 81:1, pl. 80: 5, 7 (type 3); 17 Bohemer 1995, pl. 5:a. pl. 80:18 (type 5). About type 3, the one cited by the authors 18 with outward expanded and oblique rim and disk instead of flat For the parallels from Nippur see McMahon 2006, base (from a level IV drain) in not included in this type. No types C-28 and C-29; McCown, Haines 1967, pl. 84: 3-4; sherds related to the carinated jar and plain shoulder are pub- pl. 85:16. For parallels from Tell Asmar see Delougaz 1952, lished in this context. For Tell Asmar see Delougaz 1952, types B.645.540a, C.656.340. types B.633.570a, B.064.210, B.556.540, C.466.370 respec- 19 For parallels from Tell ed-Der see De Meyer 1984, pl. tively. The contexts are taken from the Dyiala project database 14:14; pl. 8: 9-11. For parallels from Uruk see Van Ess 1988,

West & East 45 Monografie, 4 Eloisa Casadei

Other specific markers are attested in both sub- date is possible for the present sub-phase A, point- phases A and B. These are tall vessels with cylindri- ing to a late/post Akkadian date for this repertoire. cal body and flaring rim, known also from the ico- Moreover, textual data from Nippur suggest the nography (fig. 4: 1-3), the carinated rim jar with same date. From level XIII of the WF sounding at corrugated shoulder (fig. 5: 6-7), and the tall col- Nippur, four tablets are dated by Biggs at the time lared jars (fig. 5: 1-3). Finally, a series of new types of Naramsuen or Sharkalisharri.22 From the adja- start at the beginning of sub-phase A and continues cent sounding WA50C, late Akkadian tablets and after the end of sub-phase B. They are the carinat- a brick stamp bearing the name of Narasuen were ed rim bowls with modelled rim, both in the small- found in level X, stratigraphically related by McMa- er (fig. 3: 7-12) and larger (fig. 3: 13-17) versions, hon to the levels WF XIII-XII, here dated to the the hemispherical bowls with upturned plain rim same sub-phase A.23 (fig. 3: 1-3), and the deep combed decorated basins The first change corresponds to level VIII in the (fig. 5: 10-12). TB sequence. At that time, the administrative build- Table 1 summarizes the relative chronologi- ing was built and, even though the stratigraphic loca- cal sequences at Nippur, Tell Asmar, Tell ed-Der tion of the tablets is not precise, we can assume that and Uruk. The cross-occurrence of the types (fig. 2) the building was contemporaneous to the Ur dynas- highlights the gradual changes between the two ty, at least for a certain point. Looking at the year’s sub-phases A and B, particularly evident in the WF name of the kings written on the tablets, Amarsuen’s sounding at Nippur.20 This fact is particularly rele- year 4 is the terminus post quem for levels VIII-VII, vant if compared to the Syrian contexts. while Ibbisuen’s 2 is the terminus for level V.24 Final- ly, level IV1 have to be dated to the Isin-Larsa period, with tablets dated to Shuilishu, Iddin Dagan and 4. Historical chronology Ishme Dagan.25 These epigraphic data suggest a late date for our sub-phase B in the Ur III period, lacking There is a strong correspondence of types between levels that clearly date to the Ur Nammu and Shul- level IVa of the Houses at Tell Asmar and the Main gi period, and it does not continue in Isin/Larsa lev- Level of the Northern Palace. Chronological ele- els, because of the clear break that occur at TB level ments such as the Akkadian tablets from the Main IV. Looking at the comparisons with the Nippur se- Level of the Northern Palace and the seal impres- quence, the level 1 in the Sinkashid area at Uruk is sion of Shudurul from the Houses level IVa allow attributed to sub-phase B and so dated at the second to date these contexts to the late Akkadian peri- half of the Ur III period, as well as level IVb of the En- od.21 If the ceramic parallels are accepted, the same samble area at Tell ed-Der. pl. 12:102; pl. 17: 128; Boehmer 1995, pl. 6: a2, a1; pl. 9: d-f; pl. 16: b, g-h. Schrakamp 2015, p. 110). 20 Here, the intermediate period X-IX is determined by 22 Biggs 2006, p. 165. According to the Middle a phase of partial abandon of the buildings (McMahon 2006, Chronology, Naramsuen reigned between 2261-2206±30 BC pp. 24-26). It is possible that a more precise seriation of the and Sharkalisharri between 2205-2181±30 (Sallaberger, sherds from the other contexts will provide a better definition Schrakamp 2015, tab. 39). of the transition between the two phases here recognized. 23 McMahon 2006, fig. 5. About the tablets found 21 Delougaz 1967, p. 144 (note 3). Unfortunately, the in WA50C, see Gibson 1975, p. 72. For the stamp brick of Akkadian tablets from the Main Level of the northern Palace Naramsuen found in the same context see ib., fig. 69.3. were found out of context (Delougaz, Hill, Lloyd 1967, 24 p. 196). The akkadian king Shudurul ruled immediately before According to Middle Chronology, the absolute the so-called Gutian period, and his reign is estimated to be 15 chronology of the last three kings of the Third Dynasty of Ur years long in a span of time dated to 2156-2142 ±30 (Middle are as follow: Amarsuen 2044-2036 BC; Shusuen 2035-2027 Chronology) or 2148-2134 ±30 (Middle Chronology II) (see BC; Ibbisuen 2026-2003 BC (Sallaberger, Schrakamp Sallaberger, Schrakamp 2015, tab. 39). Following the 2015, tab. 39). definition proposed by the ARCANE III group, the late Akka- 25 McCown, Haines 1967, pp. 74-76. On the same dian period roughly correspond to the end of the Sharkalishar- subject, and especially related to TB level IV date, see Zettler ri’s reign until the end of the Shudurul’s reign (Sallaberger, 2008.

West & East 46 Monografie, 4 The pottery repertoire at the end of the third millenniumBC as chronological marker…

Trying to date the relative sequence identified in A particular focus on the two main sites Tell Brak the previous paragraph, sub-phase A begins in the and Tell Mozan will be taken as case-studies. late Akkadian period, probably between the reign of The carinated storage jar with high and – some- Naramsuen and Sharkalisharri, and continues until time – corrugated shoulder is a very common fea- the reign of Shulgi, with gradual transformation ture in the southern contexts (fig. 5: 4-5). This jar is that occur but without general breaks. Sub-phase B rather frequent at Tell Brak only, where it is dated points to a short span of time dated to the second to the late and post-Akkadian period, phase M-N.28 half of the Ur III reign, between Amarsuen until the On the contrary, it does not be present at other very beginning of the Isin predominance. EJZ 5 sites. According to McMahon, its introduction from south to north during the Akkadian period might visually materialize southern management 5. Analysis of the ceramic connections of agricultural products, even if its persistency after the Akkadian withdrawn could signify the loss of One of the elements that puts light on the inter- political-symbolic meaning.29 connection between different chronological and Among the types selected in the ARCANE geographical entities is the ceramic repertoire. The Interregional I volume, the carinated bowls with detection of similar technical or stylistic ceramic profiled rim (fig. 3: 7-12) represent one of the features in a wide territory can be interpreted as el- main chronological markers of this phase.30 As ement of chronological correspondence, as well as shown in the previous paragraph, the earliest geographical-cultural relations.26 Southern Mesopotamian contexts in which these The identification of a post-Akkadian phase bowls occur are dated to the late-Akkadian pe- (EJZ 4c-5), characterized by changes in settlement riod, being a type attested in both sub-phases A patterns in the Jezirah region allowed to identi- and B, as demonstrated at Nippur and Tell As- fy some parallels between the pottery repertories mar.31 This type is ubiquitous in Ur III contexts of Northern and Southern Mesopotamia, making in the South, and it seems to be well attested in the Syrian Jazirah one of the best key-studies for a a wider territory. This type is associated with the better definition of the period. Thanks to the well- ARCANE type EJZ117, and it is well attested at defined sequence, it is possible to highlight the im- Tell Mozan during EJZ 5, and not earlier.32 The portance of the introduction of new ceramic types, type is also attested at Chagar Bazar (EJZ 4c), and both from a chronological and cultural perspective.27 Tell Barri (EJZ 5), while the earliest examples were found at Tell Brak, area SS (phase N, post- 26 On the subject see Hegmon 1998. In particular, the Akkadian period). According to Armstrong and author explains the concepts of «style as function» and «technology as style» to better define the «social boundaries» between societies and he introduces the concept of «overlapping identities» between groups (ib., p. 274). The idea of ceramic el- 28 This type is particularly common in both versions, with ements as social symbols for political relations is well expressed and without corrugation. See: Matthews 2003, fig. 6.55:1 in Bunimovitz, Greenberg 2004. See also Mazzoni 2000; (area HS3/5 level 5, late phase L, EJZ 3b), fig. 6.57: 2 (area Orsi 2011, p. 385; D’Andrea, Vacca 2015, p. 47; Vacca HS3/5 level 3, phase M, EJZ 4a), maybe fig. 6.58: 18 (area et Al. 2018. The political relationship between the Southern HS3/5, level 2, phase M, EJZ 4b); Oates, Oates, McDon- Mesopotamia and the Syrian Jazirah during the last centuries ald 2001, fig. 425: 813-819 (areas FS levels 2-1, ER level 1, of the third mill. BC is well documented by epigraphic data (see both Brak phase N/EJZ 4c-5). above, note 2 and 3). It is possible that some of the types repre- 29 McMahon 2012, p. 31. sent the reflection of social and political dynamics between the 30 two regions. For a general discussion about this shape see Schmidt 2014. 27 In the collective volume about pottery published in the 31 ARCANE Interregional Series (Lebeau 2014), some types had See § 3. Schmidt (2014, p. 368) highlights that this already been detected that spread from the Southern region to type is attested from two houses above the Main level of the the North and North West. Some other types have been indi- Northern Palace. Nevertheless, the type is already attested in cated by McMahon (2012), as representative of a persistence level IV of the Houses quarter of the same site (unpublished of the Akkadian element at Tell Brak even after the urban col- datum from the Dyiala Project database. See § 2). lapse (EJ 4c phase). 32 Rova 2011, pl. 21: 10-15.

West & East 47 Monografie, 4 Eloisa Casadei

Gasche, this type (type 20E) is no longer predomi- highly attested in funerary contexts.39 In Southern nant after the eighteenth century.33 Mesopotamia, many sherds were found in post-ED The heart-shaped small jars have been recently se- layers at Abu Salabikh and at Nippur.40 Notewor- lected by the ARCANE project, and called «Akka- thy is the fact that, until now, this particular type dian» beakers and bottles.34 They are attested across of combed decoration is completely absent in earlier a wide geographical area that encompasses the whole southern contexts, and it seems to be characteristic area of the modern Iraq and part of the south-west- of the MB I Syrian repertoire, with a higher variety ern Iran and the Jazirah. The first sub-type with dou- of rim profiles. This form is well attested at Tell -Mo ble-ridge rim (fig. 4: 4-6) is present at Mozan in one zan from area C2 (stratum C7) and area A (stratum single specimen dated to the EJZ 5 (fig. 4: 6).35 An- 4b) both dating to EJZ5. They seem to be rare in other example of this type comes from the second this early phase, but their percentage growth pro- phase of the Main Building at Tell Arbid, dating at gressively in the following phases.41 It is less com- the midst of the post-Akkadian sequence at the site.36 mon at Tell Brak where almost only rims of this The second sub-type, characterized by a triple or mul- kind were found in layers related to phase N.42 tiple ridge rim (fig. 4: 7-9), is attested at Brak, but Besides these very specific types, other simi- only three specimens are published, one from area larities in the repertoire can be detected, especially HS3/5, level 3, one from area SS level 3 (fig. 4: 9) and thanks to the great number of sherds from the WF the third one from area FS level 2.37 sounding at Nippur.43 The combed-decorated crater is represented Some types are rather identical between the north- by large open vessels, quite deep, with vertical or ern and southern exemplars, and are dated at EJZ 4b-c slightly everted rim sometime modelled with a se- and sub-phase A. They are the small globular flasks ries of ridges on the exterior side, more common in with double carinated body (WF type C-19, EJZ type the South. A combed decoration on the shoulder 99, fig. 4: 11), the elongated rim lids (WF type O-16, is characterized by a wavy motive bounded by two EJZ type 77) and the collared jars with a globular body horizontal bands.38 Armstrong and Gasche pointed found at Tell Brak, even if in a very small quantity out a progressively simplification of the rim mould- (fig. 5: 3; WF type C-20, EJZ type 102).44 The beak- ing, particularly visible at Ur, where this type is ers with a truncated conical shape (WF type O-14, EJZ type 84, fig. 3: 5-6) are characteristic of the EJZ 4b-c and 5. The rim can be plain or thinner than the 33 From Chagar Bazar, parallels come from Batiment 1 in area D, phase II (Schmidt 2014, fig. 6; McMahon, Quen- southern version, but a small group from Tell Brak et 2007, pl. 3.13, figs 71-73). At Tell Mozan, parallels are at- shows the identical internal expanded rim typical of tested from phase O and P (Schmidt 2014, fig. 6; Orsi 2011, the southern exemplars.45 They firstly appeared in the tav. 181: 274 – phase P; tav. 191: 442 – phase O). See also Armstrong, Gashe 2014, p. 26, pl. 33-34. late Akkadian period in both regions, but while this 34 See Arrivabeni 2014.The beakers are described to type is characteristic of the southern phase A only, it have a sinuous/squat bodies («heart-shaped») ending in a pointed or rounded base. The bottles (shape T) have more elon- 39 Armstrong, Gasche (2014, p. 69, type 265 A3). gated bodies ending in a rounded base, and slightly flaring necks 40 with out-turned outside profiled or ledge rims (Arrivabeni Postgate, Moon 1987, nn. 21-24; McMahon 2014, p. 237). 2006, type O-24. 41 35 Schmidt 2013, pl. 68: K 480, type 2-34.101. At- Schmidt 2013, type 40. tributed to the same type by the author is also Schmidt 2013, 42 See also Oates, Oates, McDonald 2001, fig. 407. pl. 68: K 481. 43 For the detailed discussion of the EJ pottery and its 36 Koliński 2012, fig. 9a. chronological and geographical distribution, see Rova 2011. 37 Matthews 2003, fig. 6.57: 12 (HS3/5 lev. 3, phase For a detailed counting of the shared types found at Tell Brak, M, EJZ 4a); Oates, Oates, McDonald 2001, fig. 424: 793- see McMahon 2012. References of the WF types from Nip- 794, Brak phases M and N respectively. pur are taken from McMahon 2006. 44 38 TheARCANE group had focused on the spread of The elongated rim lid is slightly earlier in the Jazirah, combed decoration on several types of vessels as chronological dating at EJZ 3b to EJZ 4a. For the collared jar from Tell Brak marker for the last centuries of the third mill. BC (Sconzo, see McMahon 2012, fig. 6. Bianchi 2014, p. 383). 45 McMahon 2012, fig. 4.

West & East 48 Monografie, 4 The pottery repertoire at the end of the third millenniumBC as chronological marker…

Figure 2 Occurrence table of the types presented in the analysis (drawings of vessels not in scale). SM: Southern Mesopotamia; EJ: Early Jazirah; N: Nippur; AS: Tell Asmar; DR: Tell ed-Der

continues in EJZ 5 but in a coarser fabric and with a 6. Interpreting the relative ceramic plain or pointed rim.46 chronology of the late third millennium Among the open forms, the hemispherical, in- BC turned-rim bowl – attested in the south for the first time later in sub-phase B – can be paralleled This preliminary work has isolated a homogeneous with the bowl of curved profile of the Jazirah (EJZ ceramic horizon that characterized the late third type 91, WF type O-13) characteristic of the EJZ mill. BC, and that determined a single pottery phase 4b-c and 5 phases (fig. 3: 1-2). The series of large dated from the late Akkadian to the end of the Ur bowls with elaborated rim (EJZ types 93, 94 and III period (ca. 2200-2000 BC according to the Mid- 95; fig. 3: 16‑17) find a perfect parallel with sev- dle Chronology). This phase is characterized by a eral types from the WF area at Nippur (type O-1 gradual internal change in the pottery types, and and O-21). Both these open types are attested for for this reason two sub-phases can be recognized, the entire period (sub-phases A and B). The same here called A and B. Unfortunately, the transition chronological range is attested for a type of globular between these two sub-phases is well documented pot with ridged rim (EJZ type 107).47 only at Nippur, area WF, where the proper method Finally, the necked jars with everted rim and of excavation and the accurate publication of the ce- droop below are characteristic of the southern sub- ramic repertoire hallow a more detail chrono-strati- phase B only. Usually they have a globular body graphic analysis. with straight, vertical or slightly everted narrow The ceramic parallels between the Southern neck, and short shoulder decorated with horizon- Mesopotamian and the Syrian Jazirah sites high- tal or wavy combing (fig. 5: 10-12).48 The rim is ex- light a certain level of connection between the ternally expanded and it is characterized by a droop two regions that is directly related to the Akkadi- on the lower side (WF type C-29). This type is well an influence, since almost all the types start from attested in both the Jazirah (EJZ5 repertoire, type phase A – in the South – and phase EJZ 4b – in EJZ type 120) and Southern Mesopotamia, always the Jazirah. dating to the post-Akkadian and Ur III phases. According to the date proposed, a few types seem 46 Rova 2011, 75. to continue on a certain degree until the end of sub- 47 This type is rare in the South and apparently absent at phase A and EJZ 4c, post-dating the climax of the Nippur. It is attested in later contexts at Abu Salabikh, where it Akkadian period, especially in the phase N at Tell is characterized by wavy and geometric decoration known from the Syrian contexts (Postgate, Moon 1984, n. 29, 52, 53). Brak. As pointed out by McMahon, one could ex- 48 For a general discussion of this type see Casadei pect that typical southern Akkadian shapes do not 2016, 38-40; fig. 6.

West & East 49 Monografie, 4 Eloisa Casadei

occur in post-Akkadian levels.49 As we already seen, other hand, a general trend in the decorations could the southern sub-phase A is chronologically related suggest a different level of connections. This could to the period immediately post-dating the Naram- be testified by the spread of pattern combed decora- suen/Sharkalisharri reigns.50 It is possible that the tion as well as painted geometric designs that appear north-south connections visible in the ceramic rep- in the late third mill. BC horizon, both in Southern ertoire reflect a persistency of the Akkadian influ- Mesopotamia and Jazirah.53 ence at least at the site of Tell Brak. Finally, the two sub-phases identified, A and B, During the following EJZ 5, until now well at- are defined by exclusive types, but an intermediate tested only at Tell Mozan, Tell Barri and Tell Ar- transition exists where markers of the two phases bid, the ceramic horizon testifies a range of chang- coexisted for a short time. This transition is visible es that can be followed in the southern sub-phase B particularly at the WF sounding at Nippur, be- repertoire. Nevertheless, the appearance of new ele- tween levels X-IX, interpreted by the excavators as ments in concomitance with the Ur III dynasty (in a short period of abandon.54 In area TB, the pot- the South) and the Pusham house at Tell Mozan still tery parallels suggest this period correspond to lev- represents an open problem. Phase EJZ 5 is well at- els X-IX, that immediately preceded the Amarsuen tested in this context, and according to Volk almost tablets. Being WF XIII dated at the late Akkadian certainly epigraphically dated between Shulgi and period, this transition in the ceramic sequence Amarsuen.51 Parallels between the core area of the could be dated between Sharkalisharri and Shulgi Ur III dynasty and the Khabur area are less evident, (ca. 2205-2045 BC). In terms of historical chronol- and the main chronological markers of sub-phase B ogy, this period rather corresponds to the timespan in the South are less attested in the EJZ 5 contexts. between EJZ4c and the main level of the Pusham Nevertheless, the correspondence of the two reper- house at Tell Mozan. toires is represented by several types that have a low- Looking closer to the WF sequence, the exclu- er chronological significance. At Mozan C7, the cari- sive types of the VIII-VI perfectly match to the Mo- nated rim bowls represent the majority of the small zan C7 repertoire, and that could suggest the onset open shapes, and a high variety is attested. Some ex- of the Ur III period for that context, as say between amples of small bottles, hemispherical bowls with in- Ur Nammu and Shulgi reign. turned bevelled rim, and the large combed decorated To conclude, the strong continuity of the reper- basins found many parallels in the late Ur III/Isin- toire and the lack of extensive repertoire published Larsa repertoire, pointing to a closer date to the final determined the preliminary character of the present stage of the ceramic phase here identified.52 On the analysis. In the future, a greater number of pottery sherds from well stratified and well controlled con- 49 According to McMahon, their presence in a wider texts will help in a better clarification of chrono-ty- area proves a sort of persistence of shared fashion, that lower pology of the Southern Mesopotamian region. the symbolic political-related value of these southern elements (McMahon 2012, 31). In this sense, simultaneous changes in the ceramic horizon of the two regions at the midst of the Ak- kadian period might suggest a trend of connections that continues through time, even if at different scale. 50 Sallaberger, Shrakamp, 2015, tab. 39. 53 Combed decorations: Schmidt 2013, pl. 133:1105- 51 The date proposed by the author is based on toponym, 1110; pl. 134:1113, 1116; pl. 135:1120; pl. 136:1124,1126; paleography and onomastic of the inscription in the Pusham’s Rova 2011, pl. 23:1. For the geometric designs see Rova 2011, seal (Volk 2004, 93). According to Sallaberger (2011, pl. 23:2-9. 341), the tablet found in the same context can be attributed to 54 The fact that most of the types continue until level VI the Ur III or Isin period. The possibly Amarsuen year 3 as ter- could be probably related to a stronger tradition in the ceramic minus for the C7 phase at Mozan depends on the absence of the repertoire at domestic level in respect to a high-status building toponym Urkesh from the Ur III textual source (ibid.). of area TB. On the other hand, it could also suggest that the 52 Small bottles: Schmidt 2013, pl. 67: K460; pl. 68: entire stratigraphic range between level XI to VI has to be at- K475-6; pl. 121: K199. Bowls with inturned beveled rim: Id., tributed to such transitional period. If it is correct, level VIII-VI pl. 31: K47-49. Combed decorated basins: Id., pl. 132: K1091; in the WF sequence are earlier that level VII-V of area TB, as say pl. 46: K253. earlier of the Amarsuen reign.

West & East 50 Monografie, 4 The pottery repertoire at the end of the third millenniumBC as chronological marker…

Table 2 References for vessels in figs. 3-5

No Site Area Level/Phase Reference

3.1 Brak SS 1 EJZ 5 Oates et al. 2001, fig. 417:582 (type O-13)

3.2 Chagar Bazar B IIb EJZ 4c McMahon, Quenet 2007, n. 13 (type O-13)

3.3 Nippur WF X McMahon 2006, pl. 88:5 3.4 Asmar northern palace main level Delougaz 1953, B.024.210 3.5 Brak SS 3 EJZ 4b Oates et al. 2001, 418:610 3.6 Brak FS 2a EJZ 4c Oates et al. 2001, 421:14 3.7 Asmar acropolis sounding, O29:5 Delougaz 1953, B.151.210 3.8 Umm el Jir Area D level 4 Gibson 1972, fig. 42:c type A122 3.9 Warka Area P13 grave 12, level VI Van Ess 1988, pl. 3: form 6a, 37 3.10 Mozan C 7 EJZ 5 Schmidt 2013, pl. 44: K 205. 3.11 Brak SS 2 EJZ 4c Rova 2011, pl. 21: JZ005-P031 3.12 Brak SS 1 EJZ 5 Oates et al. 2001, 418:609 3.13 Nippur TB level VI McCown, Haines 1967, pl. 83:1, type 9 3.14 Asmar acropolis Bilalama Delougaz 1953, C.142.310 3.15 Nippur WF XIIIB McMahon 2006, pl. 89:4, type O-15 3.16 Brak SS 2 EJZ 4c Oates et al. 2001, fig. 432:941 (type O-21) 3.17 Chagar Bazar D IIb EJZ 4b-c McMahon, Quenet 2007, n. 78 (type O-18) 4.1 Asmar northern palace main level Delougaz 1953, B.256.200 4.2 Umm el Jir Area D phase V Gibson 1972, fig. 42:m 4.3 Wilaya Hussein et al. 2009, fig. 20:j 4.4 Asmar northern quarter IVa Delougaz 1953, B.556.540 4.5 Nippur WF IX McMahon 2006, pl. 127:1, type C-13 4.6 Mozan C 7 EJZ 5 Schmidt 2013, pl. 68: K 480 4.7 Umm el Jir area B level 4 Gibson 1972, fig. 43:l 4.8 Asmar northern palace earlier level Delougaz 1953, B.703.560 4.9 Brak SS 3 EJZ 4b Oates et al. 2001, 421:11 4.10 Asmar northern palace earlier level Delougaz 1953, B.703.560 4.11 Brak FS 2 EJZ 4c Oates et al. 2001, fig. 424:793 5.1 Nippur WF XIA McMahon 2006, pl. 115:4, type C.20 5.2 Wilaya Hussein et al. 2009, fig. 29:e

West & East 51 Monografie, 4 Eloisa Casadei

No Site Area Level/Phase Reference

5.3 Brak FS 3 EJZ 4b Oates et al. 2001, fig. 626:821 (type C-20) 5.4 Wilaya W1 2 Hussein et al. 2009, fig. 28:c 5.5 Brak HS3/5 5 EJZ 4a Oates et al. 2001, fig. 468:1708 5.6 Nippur TB level XI McCown, Haines 1967, pl. 81:9 5.7 Brak DH 1b EJZ 4c Oates et al. 2001, fig. 425:816 5.8 Nippur TB VIII McCown, Haines 1967, pl. 85:16 5.9 Mozan A 4b EJZ 5 Rova 2011, pl. 22:JZ005_I001 (type C-29) 5.10 Nippur TB VII2 McCown, Haines 1967, pl. 84:21 5.11 Nippur WF XIA McMahon 2006, pl. 91:4 5.12 Brak SS 2 EJZ 4c Oates et al. 2001, pl. 423:758

Figure 3 1-3: hemispherical bowls with vertical plain rim; 4-5: conical beakers with inner expanded rim; 7-12: small carinated bowls; 13-17: large bowls with profiled rim

West & East 52 Monografie, 4 The pottery repertoire at the end of the third millenniumBC as chronological marker…

Figure 4 1-3: high vessels with flaring termination; 4-6: heart-shaped beakers with double ridge rim; 7-9: heart-shaped beakers with triple ridge rim; 10-11: double carinated flasks

West & East 53 Monografie, 4 Eloisa Casadei

Figure 5 1-3: collared jars; 4-5: carinated jars with plain shoulder; 6-7: carinated jars with corrugated shoulder; 8-9: sinuous profiled jars with out-flaring rim and shallow droop below the lip; 10-12: basins or craters with elaborated rim and combed decoration on the upper external side

West & East 54 Monografie, 4 The pottery repertoire at the end of the third millenniumBC as chronological marker…

BIBLIOGRAPHY

Ayoub S. 1982, Die Keramik im Mesopotamien und in Gibson M. 1972, Umm el-Jir: A Town in Akkad, den Nachbargebieten: von der Ur III-Zeit bis zum «Journal of Near Eastern Studies» 31, pp. 237-294. Ende der kassitischen Period, München. Hegmon M. 1998, Technology, Style and Social Practices: Armstrong J.A., Gasche H. 2014, Mesopotamian Archaeological Approaches, in: Stark M.T. (ed.), The Pottery. A Guide to the Babylonian Tradition in the Archaeology of Social Boundaries (Smithsonian Series Second Millennium BC, Chicago. in Archaeology Inquiry), Washington-London, Arrivabeni M. 2014, Heart-Shaped «Akkadian» pp. 264-279. Beakers and Bottles, in: Lebeau M. (ed.), ARCANE Hussein S.Y., Altaweel M., Rejeb Z. 2009, Report Interregional I: Ceramic, Tutnhout, pp. 237-243. on excavations at Tell al-Wilaya, Iraq. Further infor- Biggs R.D. 2006, Appendix 3: The Inscription, in: mation on the 1999 and 2000 Seasons, «Akkadica» McMahon A., pp. 165-173. 130, pp. 3-42. Bohemer R.M., Pedde F., Salje B. 1995, Uruk: Die Jacobsen T. 1953, The Reign of Ibbi-Suen, «Journal of Gräber, Mainz. Cuneiform Studies» 7, pp. 36-47. Bunimovits S., Greenberg R. 2004, Revealed in Koliński R. 2012, Generation Count at Tell Arbid, their Cups: Syrian Drinking Customs in Intermediate Sector P, in: H. Weiss (ed.), pp. 109-126. Bronze Age Canaan, «Bulletin of American School Kuzucuoglu C., Marro C. 2007, Sociétés Humaines et of Oriental Research» 334, pp. 19-31. changement climatique à la fin du Troisième millénaire: Casadei E. 2016, Southern Mesopotamian Contexts and une Crisis a-t-elle eu lieu en haute Mésopotamie? (Actes a Reevaluation of the Ur III Pottery. Reconsidering du Colloque de Lyon, 5-8 décembre 2005), Paris. a Chronological Periodization, in: Stucky R.A., Laneri N., Pfälzner P., Valentini S. (eds.) 2012, Kaelin O., Mathys H.P. (eds.), Proceedings of the Looking North: the Socioeconomic Dynamics of 9th International Congress on the Archaeology of Ancient Northern Mesopotamia and Anatolia regions during Near East. Vol. 3: Reports, Wiesbaden, pp. 31-44. the Late Third and Early Second Millennium BC, D’Andrea M., Vacca A. 2015, The Northern and Wiesbaden. Southern Levant during the Late Early Bronze Age: Lebeau M. (ed.) 2014, ARCANE Interregional Serie I: A Reappraisal of the «Syrian Connection», «Studia Ceramics. Turnhout. Eblaitica» 1, pp. 44-73. Lebeau M., De Miroschedji P. 2013, Foreword, in: De Meyer L. 1984, Tell ed-Dēr IV, Leuven. Lebeau, M. (ed.), pp: ix-x. Delougaz P. 1952, Pottery from the Diyala Region Madhloum T.A. 1960, Excavation at Tell al-Wilaya, (Oriental Institute Publications 63), Chicago. «Sumer» 16, pp. 62-92. Delougaz P., Hill H.D., Lloyd S. 1967, Private Matthews R. 2003, Excavation at Tell Brak. Vol. 4: Houses and Graves in the Diyala Region (Oriental Exploring an Upper Mesopoptamian regional centre, Institute Publications 88), Chicago. 1994-1996, Oxford. Foster B.R. 2016, The Age of Agade: Inventing Empire Mazzoni S., 2000, Pots, People and Cultural Borders in in Ancient Mesopotamia, New York. Syria, in: Milano L., Fales F.M., Lanfranchi Frankfort H. 1940, Objects, in: Frankfort H., G.B. (eds.) Landscapes. Territories, Frontieres and Lloyd S., Jacobsen T., The Gimilsin Temple and Horizons in the Ancient Near East. Papers presented the Palace of the Rulers at Tell Asmar (Oriental to the XLIV Rencontre Assyriologique Internationale, Institute Publications 43), Chicago. Venezia, 7-11 July 1997, Padova, pp. 139-152. Gibson MG. 1972, Excavation at Nippur. Eleventh Season McCown D.E, Haines R.C. 1967, Nippur I: Temple (Oriental Institute Communications 22), Chicago. of Enlil, Scribal Quarter, and Soundings (Oriental Institute Publications 78), Chicago. Gasche H. 1984, Tell ed-Der 4: Le Soundage A: les Ensamble II-IV, Leuven.

West & East 55 Monografie, 4 Eloisa Casadei

McMahon A. 2006, The Early Dynastic to Akkadian W., Schrakamp I. (eds.), ARCANE: History and Transition. The Area WF Sounding at Nippur, Philology. Turnhout. (Oriental Institute Publications 129), Chicago. Sallaberger W., Westenholz A. 1999, McMahon A. 2012, Post-Akkadian ceramic assemblage Mesopotamien: Akkade-Zeit und Ur III-Zeit, «Orbis of the central Upper Khabur: What can pottery tell Biblicus et Orientalis» 160/3, Berlin. us about political and climate change?, in: Weiss H. Schmidt C. 2005, Überregionale Austauschsysteme (ed.), Wiesbaden, pp. 25-44. und Fernhandelswaren in der Ur III-Zeit, McMahon A., Quenet F. 2007, A Late Third «Baghdader Mitteilungen» 36, pp. 7–151. Millennium BC Pottery Assemblage from Chagar Schmidt C. 2013, Die Keramik der Früh-Gazīra V- bis Bazar (Area D, Phase II), in: Tunca Ö., McMahon Alt-Gazīra II-Zeit, Wiesbaden. A., al-M. Bagdo A. (eds.), Chagar Bazar (Syrie) II. Schmidt C. 2014, rd Les vestiges ‘post-akkadiens’ du chantier D et études di- Late 3 Millennium «Ur III» cari- , in: Lebeau M. (ed.). pp. 409-415. verses (Series of the Association pour la Promotion de nated bowls l’Histoire et de l’Archeologie orientales, Pubblications Schrakamp I. 2017, Ressourcen und Herrschaft: de la Mission archeologique de l’Université de Liege en RessourcenKulturen im Reich von Akkad (2300- Syrie), Leuven-Paris-Dudley, pp. 69-244. 2181 v. Chr.), in: Scholz A.K., Bartelheim M., Hardenberg R., Staecker J. (eds.), Moon J. 1982, The Distribution of Upright-Handled Jars Resource and Stemmed Dishes in the Early Dynastic Period, Cultures. Sociocultural Dynamics and the Use of , Tübingen, «Iraq» 44, pp. 39-70. Resources – Theories, Methods, Perspective pp. 81-132. Moon J. 1987, Abu Salabikh Excavation vol. 3: Sconzo P. 2014, , in: Lebeau M. (ed.), Catalogue of Early Dynastic Pottery, London. Syrian Bottles pp. 215-235. Oates D., Oates J., McDonald H. 2001, Excavation Sconzo P., Bianchi A. 2014, at Tell Brak. Vol. 2: Nagar in the Third Millennium North Mesopotamian , in: BC, Oxford. Comb-Incised and Comb-Impressed Pottery Lebeau M. (ed.), pp. 379-407. Orsi V. 2010, The Passage from the Early to the Middle Steinkeller P. 1987, Bronze Age in Jezirah: a Parallel Between Tell Mozan The Administrative and Economic and Tell Barri Ceramic Sequences, in: Matthiae Organization of the Ur III State: The Core and the P., Pinnock F., Nigro L., Marchetti N. (eds.), Periphery, in: Gibson MG, Biggs R.D. (eds.) The Proceedings of the 6th International Congress on the Organization of Power. Aspects of Bureaucracy in (Studies in Ancient Oriental Archaeology of the Ancient Near East, May, 5th-10th 2008, the Ancient Near East Civilization 46), Chicago, pp. 19-42. “Sapienza” – Università di Roma, Roma, pp. 863-881. Vacca A. 2014, rd Orsi V. 2011, Crisi e Rigenerazione nella valle dell’Alto Chronology and Distribution of the 3 , in: Pizzimenti S., Romano Khabur (Siria). La produzione ceramica nel passaggio Millennium BC Flasks L. (eds.) dal Bronzo Antico al Bronzo Medio, Firenze. Šime ummiānka. Studi in onore di Paolo Matthiae in occasione del suo 75° compleanno offer- Postgate N., Moon J., 1984 Late Third Millennium ti dall’ultima generazione di allievi (Contributi e Pottery from Abu Essalabikh, «Sumer» 43, pp. 69-79. Materiali di Archeologia Orientale XVI), Roma, Reichel C. 2004/2005, Diyala Project, «Oriental pp. 251-286. Institute Annual Report», pp. 27-35. Vacca A., Mouamar G., D’Andrea M., Lumsden Riehl S., Pustovoytov K, Dornauer A., S. 2018, A Fresh Look at Hama in an Inter-regional Sallaberger W. 2013, Mid-to-Late Holocene Context. New Data from Phase J Materials in the Agricultural System Transformations in the Northern National Museum of Denmark, «Studia Eblaitica» Fertile Crescent: A review of the Archaeobotanical, 4, pp. 17-58. Geoarchaeological, and Philological Evidence, «Climates, Van Ess M. 1988, Keramik von der Akkad- bis zum Ende Landscapes, and Civilizations» 198, pp. 115-136. der Altbabylonischen Zeit aus den Planquadraten N Rova E. 2011, Ceramic, in: Lebeau M. (ed), pp. 49-128. XV und XVI und aus dem Sinkashid-Palast in Uruk- Sallaberger W. 2011, History and Philology, in: Warka, «Baghdader Mitteilungen» 19, pp. 321-441. Lebeau M. (ed.), pp. 327-342. Van Ess M. 1993, Kap. I.1.e Inventarisierte Keramik Sallaberger W., Schrakamp I., 2015 Part I: der Kampagne 39 (1989), in: Bock B., Boehmer Philological Data for a Historical Chronology of R.M., Boesseneck J., Uruk 39, «Baghdader Mesopotamia in the 3rd Millennium, in: Sallaberger Mitteilungen» 24, pp. 52–83.

West & East 56 Monografie, 4 The pottery repertoire at the end of the third millenniumBC as chronological marker…

Volk K. 2004, Beschrifte Objecte aus Tall Mozan/Urkeš, Woolley C.L. 1934, Ur Excavation II. The Royal «Mitteilungen der Deutschen Orient-Gesellschaft» Cemetery, Philadelphia-London. 136, pp. 87-101. Zettler R. 2008, Context and Text: Nippur Area Weiss H. 2012 (ed.) Seven Generation Since the Fall of TB Level IV and the «Archive» of Lama-Palil, in: Akkad (Studia Chaburiensia 3), Wiesbaden. Michalowski P., (ed.) On the Third Dynasty of Ur: Studies in honor of Marcel Sigrist, Boston, pp. 287-301.

West & East 57 Monografie, 4

Thinking data. Integrative big data approaches towards an ‘introspective’ digital archaeology in the ancient Mediterranean

FRANCESCA CHELAZZI*, SIMONE BONZANO**

* University of Glasgow, ** Independent Researcher

Abstract

Archaeological data is what economists call a ‘non-rivalrous’ good: it can be processed again and again with no diminishing of its value. The proliferation of urveyss and excavations, coupled with the large-scale adoption of digitalization in archaeology, exponentially increased the amount of data. Instead of keeping isolated ‘data silos’, one of the current challenges is the aggregation and correlation of archaeological data in the 3V’s perspective of ‘Big Data’: high volume, high velocity and high variety. Archaeologists make traditionally use of SQL RDBMS databases, although the rising importance of Big Data in Computer Science has recently brought to our attention a new typology of Database Management System: the NoSQL. This typology of database can much more effectively handle Big Data by preserving a ‘more human’ approach through dynamic queries and enhanced functions of data visualisation. In this perspective, NoSQL may prove to be a fundamental tool in moving from ‘data silos’ to a more complex strategy of data management. This paper explores the potential of a specific type of NoSQL Graph database (Neo4j) of handling archaeological ‘Big Data’, through the discussion of a specific case study in Bronze Age South-western Cyprus.

Keywords

Big Data, data integration, data management, SQL, NoSQL, Bronze Age, Cyprus,Eastern Mediterranean

West & East 59 Monografie, 4 Francesca Chelazzi, Simone Bonzano

1. Introduction that archaeology is going digital, as demonstrated by the proliferation of surveys and excavations over As archaeologists, we are well aware that many of time, and this process has exponentially increased our primary field methods are generally destructive the amount of data at our disposal. While the ‘com- and that our field results cannot be replicated by fu- putational turn’ appears to be nowadays a consoli- ture generations. Once we have dug a site, sampled dated practice,4 one of the current challenges is to a filling or emptied a grave, its research potential is replace the isolated data silos with advancements in substantially already decided; even when we survey data aggregation and correlation through the 3V’s a field, the picture right in front of us will never be perspective of ‘Big Data’: high volume, high velocity the same again. As archaeologists we know that it and high variety. is of primary importance to keep a comprehensive Since the well-known 2002 workshop at the record of our work and that our field notes, labels, University of Michigan, and its resulting volume pictures, drawings and maps will be probably much ‘Side-by-side Survey: Comparative Surveys in the of our winter duties, to develop our analysis and to Mediterranean World’,5 the debate about working plan the next stages of our fieldwork.1 in a comparative format has progressively gained In a regional and landscape perspective, we start- popularity in Mediterranean archaeology,6 so that ed to accumulate a significant amount of quantita- today an increasing number of scholars is aware tive data at least from the 1960s and the 1970s, when about the urgency of combining local datasets in or- regional studies progressively acquired an impor- der to produce wider landscape narratives.7 This is tant and autonomous position in the archaeologi- not only a theoretical issue but also a tangible ne- cal agenda, as the emergence of the so-called ‘New cessity: urban and rural changes are strongly trans- Archaeology’ and ‘New Geography’ determined the forming our territories and frequently the only leg- adoption and the development of new techniques acy at our disposal is a very heterogeneous corpus of for field and lab research. Until the 1980s, howev- old reports, survey projects, casual discoveries, res- er, the vast majority of archaeological data was cre- cue excavations and grey literature. ated and stored on paper, according to very patchy It is unquestionable that a remarkable number and heterogeneous organizational schemes. It’s only of surveys has been operating in the last decades, from the following ‘New Wave’ that computer with a large variety of theoretical backgrounds and technology spread and analytical tools improved.2 survey methodologies. Consequently, this implies From the ‘digital turn’ of the 1980s onwards, in that data comparability is often a problem that ar- fact, an ever-growing amount of data has been col- chaeologists have to address. The heterogeneity of lected, produced and processed using new digital regional datasets does not mean that archaeological techniques and analytical models, while advanced legacies are useless or incomparable; it rather means computer-based technologies and tools have been that researchers must be aware of how, when and largely developed inside and outside archaeology. why data has been collected, paying a particular attention to their own research methodologies, techniques and objectives.8 2. Thinking (big) data Before even beginning a second-generation anal- in Mediterranean archaeology ysis, a striking doubt rises at the centre of the methodological issue: can we rely on datasets collected in the past by someone else? And – if yes – how can we Archaeological data is what economists call a ‘non- rivalrous’ good: it can be processed again and again 4 Huggett 2016. with no diminishing of its value.3 There is no doubt 5 Alcock, Cherry (eds.) 2004, pp. 1-9. 6 See several papers in Smith (ed.) 2011 or the recent 1 Allison 2012. volume by Cadogan et Al. (eds.) 2012. 2 Cherry 1994. 7 Kintigh 2006. 3 Gattiglia 2015, p. 1. 8 Allison 2008; Kantner 2008.

West & East 60 Monografie, 4 Thinking data. Integrative big data approaches towards an ‘introspective’ digital archaeology in the ancient Mediterranean aggregate geographically and/or temporally circum- 3. Data storage in archaeology scribed datasets in a wider and cross-disciplinary ‘Big Data’ and ‘Linked’ perspective? Unfortunately, this ‘What is data?’ is the question every scholar, not point is frequently a deterrent and interesting soon- only archaeologists, tries to avoid for their whole to-be comparative studies cease to exist even before life. Philosophers do not agree about a unique de- birth. The ‘either go big or go home’ mantra,9 in fact, finition, because probably there is not a single me- largely conflicts with a perceived impression of data aning, and centuries of speculation provided grand fetishism, with a general lack of collective awareness theoretical scenarios but little practical help. Etymo- about Big and Linked Open Data, with a general- logically ‘data’ comes from the Latin datum, which ized resistance to data sharing and with a substantial means ‘given’; it could be so described as an ‘infor- deficiency of theoretical and methodological tools mation which is given’ from one to another, as the to synthesise and analyse information.10 verb ‘to give’ needs the existence of a receiving part- One of the preliminary problems, in fact, is the ner. If we accept the Latin etymology, the nature of incontrovertible difficulty in aggregating and cor- data lays in the existence of an external relationship, relating data of heterogeneous provenience. On top in the existence of a ‘giver’ and of a ‘receiver’. On of that, archaeological legacies are frequently formed the contrary, our traditional way to acknowledge by grey literature that makes datafication and com- data surprisingly focuses on the properties of the putational analysis even more difficult.11 The inclu- observed reality, which means that it focuses on in- sion of low quality data, for example, can support ternal factors that belong to the observed entity it- the reprocessing of legacy datasets to explore new re- self, rather than on the external factors that belong search topics and to reach a level of granularity and to data usage. The Cambridge Dictionary provides a detail that sampling strategies cannot provide.12 As very concise and stringent definition for data: «in- Newhard recently argued, «archaeology is a place formation collected for use», which reinforces the within the social sciences and Humanities where perception that usage constitutes the very nature of the nature of the work deals with Big Data».13 Giv- data.15 en this inherently heterogeneous nature of archaeological data, we must accept messiness as an inevitable characteristic of the archaeological inquiry and 3.1 Setting the question we need to collectively (re)think the data quality question, designing methodologies that can support Generally speaking, «a database is a collection of scholars in evaluating data metrics.14 information that is structured and recorded in a consistent manner […] to store and retrieve data records in the most efficient way possible».16 Bur- rough and McDonnel listed four ‘famous’ specific functions that a Database Management System (DBMS) should provide: quick access to data, facil- ity for inputting/editing/updating data, ability to 9 Wesson, Cottier 2014, p. 1. define rules to ensure data consistency and ability to 17 10 Boyd, Crawford 2012. protect data. This extreme flexibility and efficien- 11 Evans 2015. cy, coupled with the development of the Relational 12 Gattiglia 2015, p. 2; Van Eijnatten, Pieters, Model in the 1970s, made DBMS an essential tech- Verheul 2013, p. 59. nology in the hands of archaeologists. 13 Newhard 2013. For a critique to archaeological data as ‘Big Data’ see Huggett 2016. 14 Bevan et Al. 2013a, 2013b; Costas et Al. 2013; 15 Cambridge Dictionary 2017. Crema 2012; Crema, Bevan, Lake 2010; De Runz et Al. 16 2007; Habert, Huc 2010; Jaroslaw, Hildebrandt-Rad- Conolly, Lake 2006, p. 51. ke 2009; Lawrence, Bradbury, Dunford 2012. 17 Burrough, McDonnel 1998, p. 50.

West & East 61 Monografie, 4 Francesca Chelazzi, Simone Bonzano

While there is not a single data model of record- averse, Archaeology Data Service, Figshare, Open ing archaeological data, there is an implicit ‘internal’ Context, tDAR, Zenodo and UCL Discovery) and agenda that every archaeologist is asked to adhere national (e.g. Arachne in Germany, DANS in the to, in order to organize and manage collected data Netherlands, Mappa in Italy and SnD in Sweden) in the most efficient way. Completeness, correct- data repositories can support the sharing of existing ness, accuracy, consistency and structure are the five datasets among current and future scholars nowa- ‘magic words’ to fix in the mind when designing our days. Despite Linked Open Data still miss to break databases and conceptual models. Moreover, Ber- the resistance of Academic archaeology, data shar- ners-Lee provided us with the famous ‘5 stars’ goal ing appears to be one of the Gran Challenges of to produce fine datasets for the Linked Open Data twenty-first century archaeology, to create intercon- (LOD) realm and the Semantic Web.18 nected sets of information, hub of knowledge and Within the life-cycle of archaeological data (cre- opportunity to explore new trajectories in research.23 ating-processing-analysing-preserving-giving ac- The ‘don’t publish, push!’ model can certainly im- cess-reusing), it is significant that emphasis is tra- prove data quality through the integration of pri- ditionally – and maybe unconsciously – placed in vate editorial feedback and public version control; the phases of data-collection and database design. it can open new analytical and comparative research Yet the life-cycle contains further additional stag- avenues, encouraging a greater dynamism and new es, including data reuse, even if it still remains rare.19 collaborations among researchers and institutions.24 As the Archaeology Data Service correctly argued, Cyberinfrastructures such as Open Context «imagining data being reused by someone else move in this direction by providing access to pri- may cause you to approach the creation and de- mary data from multiple projects and making them sign of your data in a new light».20 A recent study re-analysable and comparable through the adoption developed by the ARIADNE project revealed of controlled vocabularies and ontologies, the rigor- that Academic archaeology is surprisingly lag- ous review of internal consistency and the provision ging behind Heritage Management in developing of accompanying metadata.25 The entire workflow a Linked Data approach: diverse organisational of these infrastructures, that aim to transform mul- settings in charge of data collection and manage- tiple sets of primary data into annotated products ment, project-oriented data management prac- ready for analysis, emphasises and optimises the idea tices and a general low level of open sharing of re- that data is a dynamic product, the very nature of search data are among the present unfavourable which is not fixed or static and the scientific capa- conditions that impede the uptake of the Linked bility of which is not limited to a defined number Open Data approach.21 Moreover, the results of of uses.26 Additionally, many project-specific data the AthenaPlus project’s 2013-2015 survey re- sharing facilities such as the Çatalhöyük database vealed that, despite a general awareness about can provide a fundamental contribution in support- LOD, there is a persisting lack of experience and ing second generation analysis, even if they are not engagement with LOD projects.22 readily scalable and comparable with external data- Yet, in recent years, a growing set of data shar- sets, unless they conform to the specific recording ing infrastructures has been developed. New in- systems.27 frastructures (e.g. ARIADNE, MedArchNet) have In particular, the increasing availability of com- been created in order to share data, theories and plex sets of primary data can positively support the knowledge; several international (e.g. JOAD Dat- dissemination of NoSQL-based analysis, providing

23 Huggett 2015b. 18 Berners-Lee 2009. 24 Kansa E.C., Kansa S.W., Arbuckle 2014. 19 Wallis, Rolando, Borgman 2013. 25 Kansa 2010. 20 Archaeology Data Service 2014. 26 See footnote 1. 21 Geser 2016, pp. 43-46. 27 Engel, Grossner 2015; Lukas, Engel, Mazzu- 22 AthenaPlus 2015, pp. 7-12. cato 2018.

West & East 62 Monografie, 4 Thinking data. Integrative big data approaches towards an ‘introspective’ digital archaeology in the ancient Mediterranean

a native data complexity and heterogeneity that can rows, where data needs to be preliminary structured capitalize the horizontal mechanisms, the flexibili- and grouped into schemata. By the logic of SQL, ty and the relationship-based structure of NoSQL. emphasis is placed on categories, that are general- As successfully demonstrated, published primary ized and predefined typologies that can aggregate data can improve data reuse and re-examination, fragmented data through a set of complex relation- increasing the overall quality of legacy datasets and ships, defined by the user according to the limits of supporting new interpretations and more detailed the machine. Archaeologically speaking, scholars thematic narratives.28 have a good grasp of SQL databases as they are tra- Hopefully, in the next decade LOD and data re- ditionally used in structuring organized collections use practices will have greater impact and diffusion, of archaeological data in a day-by-day perspective. going beyond Heritage Management and directly Data organization passes through the schematiza- entering the archaeological discipline. Rewarding tion and abstraction of the body of information mechanisms, credit systems, reducing barriers to (e.g. chronology, function, artefacts assemblage, public participation, new data modelling practices etc.) where data complexity is deconstructed and re- and a collective recognition of data reuse as a profes- constructed. sional goal will certainly minimise the current reluc- NoSQL DBMS are built upon a different ap- tance to share data.29 proach, which aims to save the complexity of data Berry and Huggett have recently advocated a without oversimplifying the process of knowledge ‘third wave’ in the use of digital technologies in the creation. They support the storage of unstructured Humanities, including Archaeology, arguing that data across multiple nodes, with no need to organize it’s time to collectively examine the result of this data into schemata and fixed tables.31 This horizon- ‘going-digital’ process and, in particular, to focus on tal scaling mechanism help to accelerate the whole the impact of the digital transformation on our pro- workflow and to handle ‘Big Data’, a vast range of cess of knowledge creation.30 The use (and reuse) of heterogeneous datasets that the SQL rigid and ver- archaeological data can not miss the chance to play tical structure can hardly manage. In contrast to a vital role in this ‘third wave’. RDBMS that generally adopt SQL language for querying and maintaining the database, NoSQL databases use many typologies of language to support 3.2 Comparing, replacing or integrating SQL different strategies and preferences in data collec- and NoSQL (R)DMBS? tion and processing.32 Generally speaking, NoSQL databases are It is necessary to take a step back and cut our minds progressively reaching the level of the famous SQL off from the traditional idea of ‘database’ (shaped relational databases such as Oracle and mySQL in by a long-term engagement with the SQL language) terms of their wide distribution. Fig. 1 displays the to acknowledge the difference between SQL and adoption of the main NoSQL databases according NoSQL. to the general families described by Yen.33 In our Commonly, Structured Query Language (SQL) opinion, Document Store databases and Graph Relational Database Management Systems (RD- databases are of great interest as they can provide BMS) are organised according to a certain over- an innovative contribution to the ‘introspective’ simplification of the process of data collection and discourse of Digital archaeology, in particular knowledge creation, which is reduced to a two-fold concerning our reflexive engagement with data and system based on standardized set of columns and the process of knowledge creation.34

31 For a general overview see Strauch 2009. 28 Atici et Al. 2013. 32 Several classifications can be found inCattell 2010; 29 Harley 2013; Kansa E.C., Kansa S.W. 2013; Pio- North 2009; Popescu 2010; Strauch 2009; Yen 2009. war, Vision 2013. 33 Yen 2009. 30 Berry 2011; Huggett 2015a. 34 Huggett 2015a.

West & East 63 Monografie, 4 Francesca Chelazzi, Simone Bonzano

Figure 1 Graph displaying the diffusion of the main NoSQL DBMS; for every main family (as described in Yen 2009) the commonest software has been included in this graph. Ranking values are updated at March 2018 (© DB-Engines.com)

Instead of splitting information into tables and zation, as requested by SQL databases, thus main- rows, Document Store databases such as MongoDB taining the overall complexity of legacy data. Simi- are able to store data belonging to different text larly, Graph databases such as Neo4J can provide a formats, thus supporting the maintenance of data significant contribution to archaeology, particular- complexity, the distribution of information across ly by improving the traditional techniques for data multiple servers and the movement/replication of visualization. Graph databases permit users to split entire objects. Entries can differ from one anoth- data into two different types of information, namely er in terms of their content/structure (in this case nodes (data itself) and their reciprocal relationships. the two elements fully coincide) as they are totally Nodes and relationships are defined by properties free from the limitative scheme based on tables and set by the user in order to ease data storage and anal- columns. This type of NoSQL database seems to ysis. The main benefit is represented by the relation- be particularly suitable for archaeological research ships themselves; while they operate as simple joins as it permits scholars to aggregate diversified legacy in SQL RDBMS, in NoSQL databases they possess datasets such as, for example, the results of surveys a defined set of properties that contain a meaningful carried out back in time, with variable methodolo- part of the information. gies, sampling strategies, theoretical backgrounds Certainly, we need to consistently advance in and historical questions. Diversified datasets can be digital technologies for archaeology in order to ful- correlated without requiring a secondary standardi- ly introduce NoSQL databases into our traditional

West & East 64 Monografie, 4 Thinking data. Integrative big data approaches towards an ‘introspective’ digital archaeology in the ancient Mediterranean

toolbox. One of the main concerns is the limited ca- What follows is our preliminary experiment pacity of NoSQL to integrate with Geographic In- in this direction, aiming to combine SQL (Post- formation Systems (GIS), that have been developed greSQL/PostGIS) and NoSQL (Neo4j) databases within the SQL relational structure.35 The synergy in a strictly GIS-based (QGIS) archaeological re- with the main GIS desktop software is very experi- search. The selected case study is a discrete test area mental and still in progress, such as in the case of the in Bronze Age Southwestern Cyprus. integration between MongoDB and QGIS or of the Neo4J Spatial library.36 Differently, SQL databases can be externally – but easily – related/joined to the 4. A case study from Bronze Age Cyprus main geospatial packages or they can even include a spatial extension into relational databases, such as Cyprus is the third largest island in the Mediterra- in the case of the widely used PostgreSQL/PostGIS nean Basin, as well as an extraordinary case study combination. given its ecological diversity and long-term trajecto- An additional and game-changing potential of ry in human history. Located in close proximity to NoSQL databases in archaeology is their usability the Levantine and Anatolian coastline, Cyprus was in predictive modelling, addressing large-scale set- an active crucible of ideas, technologies, goods and tlement dynamics through stochastic processes. In ideologies throughout Mediterranean Prehistory a comparative perspective, for example, NoSQL and Protohistory.39 DBMS (particularly Graph databases) are currently The Bronze Age history of Cyprus can be gen- used in intelligence-related studies (e.g. in advanced erally described as a complex and gradual route data mining on social networking37), but unfortu- from a relatively egalitarian, insular and village- nately scientific literature about their application based organization in the Early and Middle Bronze in history-related studies is still missing. As demon- Ages (2000–1750/1700 cal. BC) through to the strated by Durand, Belacel and LaPlante38, Graph hierarchical, urban and internationally-connected NoSQL DMBS such as Neo4j can be successfully society of the Late Bronze Age (1750/1700‑1300 employed in Learning Path Prediction, where the cal. BC). inductive nature of retro-simulative environments suggests that similar techniques can also support the investigation of historical processes, including pre- 4.1 Trends in Cypriote archaeology dictive modelling in archaeology. To conclude, currently we can not entirely re- When one googles ‘Cyprus crossroad’, approximately place SQL with NoSQL databases (assuming that 528,000 results come into view: from the 2011 exhi- this is our final goal) as we have a lot of ground to bition at the Smithsonian Museum entitled Cyprus cover before fully integrating NoSQL databases crossroad of civilizations to thousands of recent arti- with GIS and before elaborating an archaeology- cles in geopolitics that emphasize the role of Cyprus based body of theoretical and methodological liter- as a ‘crossroad of three continents’. From the archae- ature. We can certainly combine SQL and NoSQL ological perspective, Cyprus is paradoxically located databases in order to successfully manage an inin a position of academic isolation: it is neither part creasing amount of heterogeneous archaeological of Aegean nor of Near Eastern archaeology. Much (Big) legacy data and to effectively progress towards vaunted as a crossroad of cultures, Cyprus is tossed a proper data-driven approach. from one academic sector to another: Aegeanists and Orientalists consider Cyprus as one of the latest is- 35 Bennett 2015; McCarthy 2014. lands along the cross-Mediterranean trade route, de- 36 Altaweel 2016. See also Schutzberg 2011 about pending on whether it started, whether from the East the use of NoSQL for geospatial tasks. 37 Corbellini et Al. 2015. 39 Knapp 1986, 1993; Manning, Hulin 2008; Sher- 38 Durand, Belacel, LaPlante 2013. ratt S., Sherratt A. 1993.

West & East 65 Monografie, 4 Francesca Chelazzi, Simone Bonzano

or from the West. Therefore, ironically, the archae- The archaeological legacy from the Southwest of ology of Cyprus progressively became a discipline in Cyprus is the product of more than fifty years of re- its own right and a certain sense of insularity became search projects and a century of antiquarian inter- the backbone of several studies about Cypriote past.40 est. Under the supervision of the local Department ‘Insularity’ and ‘connectivity’ are recurring words in of Antiquities, six foreign survey projects record- Cypriote archaeology, frequently used in a constant ed and mapped the archaeological heritage of the tension between the island and what was located be- Southwest (fig. 2); namely these are: the Canadian yond its coastline.41 In this way, Cyprus started to Palaipaphos Survey Project (CPSP), the Kent State be delineated as a single monolithic block, different University Expedition at Episkopi Phaneromeni from what was positioned around/beyond/close to (KSU), the Kouris Valley Project (KVP), the Sotira it. Under some points of view, this dichotomy has Archaeological Project (SAP), the Sotira Kaminou- flattened our analytical perspective and the island’s dhia Survey (SKS) and the Western Cyprus Project complex ecosystem turned into a geographically ho- (WCP). Rescue and research-oriented excavations mogeneous mass, saturated by copper in every square provided a further impressive sequence of datasets, kilometre. Excluding some chronological stages or exploring the life at the Bronze Age communities specific themes, the regional scale of analysis has been of Sotira Kaminoudhia, Erimi Bamboula, Episko- omitted, at least, until the last decade.42 The repeated- pi Phaneromeni, Erimi Laonin tou Porakou and ly mentioned Early and Middle Bronze Age Cypriote Pitharka, Alassa Paliotaverna and Pano Mandilaris, regionalism, for example, has resulted in a prolifera- Prastio Mesorotsos, Kouklia Evreti and Asproyi, Maa tion of typologies of material production – especially Palaeokastro and Kissonerga Skalia. ceramics – while several topics related to the territory It is a matter of fact that regional and land- (e.g. settlement pattern, natural resources, econom- scape studies have a proud and successful tradition ic organization) still persist in a substantial island- in Cypriote archaeology. Thanks to the pioneer- wide perspective. Despite this fragmentation in our ing and multi-thematic approach by Gjerstad and knowledge, new interesting insights about the ‘di- the Swedish Cyprus Expedition, Cypriote archae- vergent trajectories’ in Cypriote Prehistory and Pro- ology was permeated by a wide-ranging theoretical tohistory have been recently provided by Webb and and methodological focus from the 1930s onwards, Frankel, tracing a new route in regional studies.43 with the important surveys directed by Dikaios and Catling.45 Systematic surveys in Cyprus underwent a successful period of flourishing from the 1970s 4.2 Addressing the problem of data integration onwards, thus producing a remarkable quantity of in the Southwest of Cyprus collected data that constituted a solid basis for several studies about the local settlement pattern.46 In There is currently no body of data allowing a com- 1974 the Turkish invasion of the north of the is- plete and regional-based comparative analysis about land suddenly interrupted the activities by the local Bronze Age Cyprus, and, excluding few important Survey Branch of the Department of Antiquities, exceptions, there is not a comprehensive diachronic and archaeological projects shifted from the north- study at a regional scale of analysis.44 ern part of the island towards the south coast and the southwestern region of Paphos.47 Nevertheless, 40 Thisis rather common for insular cultures; see Vo- giatzakis, Pungetti, Mannion (eds.) 2008. the new regional focus of the 1970s inspired for- 41 Held 1993; Knapp 2007, 2008. eign institutions to organize new projects on the 42 See Barlow, Bolger, Kling (eds.) 1991 for regionalism in pottery production and distribution, and Webb, 45 Gjerstad 1926, p. 1; Gjerstad, Lidros, West- Frankel 1999 concerning the passage from the Late Chalco- holm 1934, p. xiv; Catling 1962; Stanley Price 1979, lithic period to the Philia phase. p. 51. 43 Webb, Frankel 2013. 46 Very successful examples are Merrillees 1973 and 44 See Georgiou 2006 for a territorial-based compara- Frankel 1974. tive discussion about Bronze Age Cypriote regionalism. 47 Hadjissavvas 1977.

West & East 66 Monografie, 4 Thinking data. Integrative big data approaches towards an ‘introspective’ digital archaeology in the ancient Mediterranean

Figure 2 Spatial distribution of the six foreign survey projects in the Southwest of Cyprus and the location of the original research area and of the selected test area (the Ezousas river valley)

island, such as the Kent State University Expedi- prus Project (SCSP) that aimed to investigate «the tion at Episkopi Phaneromeni (KSU), the Sotira relationship between the production and distribu- Kaminoudhia Survey (SKS) and the Vasilikos Val- tion of agricultural and metallurgical resources» in a ley Project (VVP).48 In this period the debate on wide terra incognita.50 The diachronic approach and data collection methodologies developed in paral- the noticeable interconnections between environ- lel with the emergence of new interpretive frame- mental factors and socio-cultural interactions com- works. At least until the end of the 1980s, with pletely fit into the landscape paradigm characterizing the development of surveys as a stand-alone prac- the 1990s. This attempt to comprehend a particular tice, it was rather common to notice that data col- geographical area in a more detailed and comprehen- lection strategies were changing from year to year, sive way progressively produced a narrowing of sur- even within the same survey project. This process vey areas also in Cypriote archaeology.51 Examples is clearly exemplified by the Canadian Palaipaphos of this ‘high-resolution wave’ are two recent projects Survey Project (CPSP), where a site-oriented proj- carried out in the Southwest of the island: the sur- ect turned into a large-scale survey.49 vey in the surroundings of Prastio Mesorotsos and the In the 1990s, important projects operated Kouris Valley Survey Project (KVP).52 throughout the island, such as the Sydney Survey Cy- 50 Given, Knapp, Coleman (eds.) 2003, p. 1. 51 The ‘Mediterranean Myopia’ according to lantonB 48 Held 2003; Swiny 1979; Todd 1996, 2004. 2001, p. 68. 49 Sørensen, Rupp (eds.). 1993; Rupp, Kling 1983. 52 Jasink et Al. 2008; McCarthy et Al. 2010.

West & East 67 Monografie, 4 Francesca Chelazzi, Simone Bonzano

4.3 Methodology Neo4J has a very intuitive – for being a NoSQL database – structure, even if the entire process of Whilst Cypriote archaeology proudly experienced data entry needs to be carried out through com- a long history of successful landscape studies and mand-strings. Although this can appear to be an ob- survey projects, surprisingly there is a general lack stacle at first, command lines facilitate the control of web-platforms and repositories for sharing and over data structure. The database is organized accord- comparing datasets.53 This evident discrepancy ing to two types of features, where nodes and relation- points to the necessity to promote a new season of ships constitute the technique to add information as data management and to engage in a better under- each of them is accompanied by a rather infinite and standing of the theoretical and methodological im- fully editable set of properties.57 According to Neo4j plications of the overall ‘going-digital’ process. syntax, a node is the simplest form among the avail- This paper presents just a preliminary test about able sets of information: it can be a substantive, a per- the integration of SQL and NoSQL databases. Chela- son’s name or – in archaeology – the name of the site zzi recently re-processed a large amount of archaeo- or its identity document (ID). Additional sets of in- logical legacy data from southwestern Cyprus, setting formation can be placed both in the node’s label or an effective – but simple – method to manage differ- as properties. Nodes are thus correlated through re- ent datasets.54 Her methodology was shaped accord- lationships that do not operate in a standardized way ing to the characteristics of the Bronze Age Cypriote but rather as a dynamic component that can be ful- archaeological record, but it originated from the eval- ly customized by users and adapted to ad hoc needs.58 uation of wider theoretical issues that are not limit- Nodes, labels, properties and relationships create a ed to this specific spatial and temporal focus.55 The dynamic and potentially infinite network of relation- use of data quality estimation procedures supported ships. The process of data query can be structured on the characterization of the overall archaeological leg- this network, which – despite the standard language acy in terms of periods, regions, sites, typologies of – facilitates the creation of semantic queries. Thanks artefacts that are more or less reliable. This research to its elevated level of customization, this database al- largely made use of PostgreSQL/PostGIS where data lows users to visualize not only specific data or data was structured according to the traditional RDBMS collections, but also to create new relational data. The framework, based on tables using columns and rows. entire process is performed through a graph system With this paper we aim to go one step further providing data with its own visual evidence, which and to test the integration of NoSQL databases was what the column/row structure of SQL RDMBS in this specific case study. In particular, we tested obscured.59 Graph databases, including Neo4j, are the use of the commonest type of Graph database, traditionally classified as highly flexible but highly Neo4J, which can provide a remarkable contribu- complex; their own set of given features, however, tion to data visualization and is easy to use on web was considered as a crucial benefit in the performance browser, even when offline. To develop a coherent of this test analysis.60 and rigorous test analysis, we defined a subset of the Keeping in mind what we achieved with the original research area, that coincides with the Ezou- SQL-based representation of the archaeological re- sas river valley (fig. 2), where evidence of human oc- cord in the Ezousas river valley and aiming to cre- cupation is documented at least from the Aceramic ate the most consistent and coherent comparison Neolithic at Kritou Marottou Ais Yiorkis.56 between the two typologies of DBMS, we decided to run our comparative NoSQL test using the same 53 See Kydonakis, Chliaoutakis, Sarris 2013 for categories previously employed in the PostgreSQL/ the GIS-based application for the management of the archaeological heritage in Cyprus. 57 54 Methodology is broadly described in Chelazzi 2016, Baton, Van Bruggen 2017. pp. 44-93. 58 Hunger 2014. 55 See footnote 15. 59 Robinson, Webber, Eifrem 2015. 56 Simmons 2005. 60 Popescu 2010.

West & East 68 Monografie, 4 Thinking data. Integrative big data approaches towards an ‘introspective’ digital archaeology in the ancient Mediterranean

PostGIS database.61 This permitted us to accelerate At the same time, the Red Polished Punctured ware the whole process without limiting its performance. characterizes part of the ceramic production along Nodes included ID of sites, their function, their pe- the south-central coast. Already Åström attributed riods of occupation, their geographic location and the production of this ware to the South of the island the occurrence of the most common ceramic wares and more recently Herscher hypothesized its produc- of the Cypriote Bronze Age. This approach allowed tion at Episkopi Phaneromeni, renaming this ware as us to create a heterogeneous set of ‘nodes’ which the ‘Episkopi ware’.64 This ware has been broadly re- contained the basic set of information concerning corded not only along the Kouris valley but also at every archaeological site; in our opinion this set of several sites in the Southwest.65 data constitutes a solid base to create a consistent Once we structured the database, we run the test network of relationships. This last point, stage by through the use of the NoSQL main tool: the query. stage, resulted to be the main strength of the graph Our first test concerned the co-occurrence of both database. From a purely theoretical point of view, these ceramic wares in the Ezousas valley, aiming in fact, this system is virtually boundless and con- to map the convergence of ceramic traditions both sequently it allows the database to avoid any loss of from the coastal Southwest and South-centre of the semantic value for each single entry. The system is island. In the SQL environment, this test is based so data-oriented that it supports also the creation on a very simple query that every archaeologist is fa- of ‘weighted relationships’ to increase the network’s miliar with: we used the SELECT clause (which is information capacity. the most complex statement in SQL) for the ‘site number’, ‘Drab Polished Blue Core’ and ‘Red Pol- ished Punctured’ fields. Our main condition was 4.4 Discussion about data the WHERE clause which links the occurrence (YES clause) of the ‘Red Polished Punctured’ or A test of the real consistency/scalability of the (OR clause) the ‘Drab Polished Blue Core’. node/property structure was not among the main objectives of this preliminary study, thus we decided SELECT * WHERE “Red Polished Punctured” to bound our discrete test of relationships. In par- OR “Drab Polished Blue Core” = YES ticular, test query#1 focused on the ‘attributed-to’ relationship, which describes the hypothesized peri- The result of this query included five sites where both ods of occupation at each site, and on the ‘found-in’ the wares are documented, one site which provid- relationship, which indicates the collected material ed only the Drab Polished Blue Core and three sites assemblage. where only the Red Polished Punctured was collected. Given the selected set of variables, the test explored We ran the same test in the Neo4j environment. the local material assemblage and, in particular, two The software uses the Cypher language which needs distinctive Middle Bronze Age local ceramic vari- to perform a type of statement using both nodes ants of the island-wide Red Polished ware: the Drab and relationships. Polished Blue Core and the Red Polished Punctured wares.62 Generally speaking, the Drab Polished Blue match(b:pottery)-[r:FOUND_IN]->(a:site), Core belongs to a ceramic tradition of southwestern (b:pottery) -[x:ATTRIBUTED_TO]-> (c:period) Cyprus as it is documented by the extraordinary occur- where c.name=“MC I” XOR c.name=“MC II” rence of this ware in the area of Kissonerga Ammoud- XOR c.name=“MC III” XOR c.name=“LC IA” hia and Skalia, where it accounts for more than 70%.63 return a,b,c,r,x

61 Chelazzi 2016. 64 Åström 1972a, p. 95 Type VIII B; Herscher 1976, 62 Åström 1972, pp. 83-84; Carpenter 1981, pp. 61- 1991. 64; Herscher 1976, 1991, 2003; Swiny 1979, pp. 232-236. 65 Bombardieri et Al. 2012, pp. 96; Guldager Bil- 63 Crewe et Al. 2008; Graham 2012; Philip 1983. de 1993, pp. 6, 18-19, pl. 1.

West & East 69 Monografie, 4 Francesca Chelazzi, Simone Bonzano

Figure 3 Test query#1: the Neo4j graph

To perform this very simple query, Cypher was The outcome was, in fact, a nuclear graph, where used in its simplest performance potential as the a radial pattern based on chronological periods or- ‘search, find, elaborate and display’ query basical- ganized the information in a meaningful way. Fig. 3 ly matches the SQL statement. In short, we asked displays the created network in relation to the spe- the database to elaborate and visualize (match) all cific sites’ functions and IDs, the chronological pe- the possible relationships between the ceramic as- riods and the two selected ceramic wares. This com- semblage (b:pottery) and their corresponding lo- plex network was rather unexpected and it provided cation (a:site) in a time span (c:period) which a certainly more meaningful information when com- stretches from the Middle Bronze Age I (MC I) pared with the traditional SQL-based query. to the Late Bronze Age IA (LC IA) (c.name). Be- Beyond the impact of this small test in the ar- sides the specific Cypher clauses (match, where, chaeological discourse, the main difference among XOR, return), we requested the software to cre- the SQL and the NoSQL approach is self-evident. ate two new relationships (r:FOUND-IN) and Queries performed in Neo4j and Cypher do not (x:ATTRIBUTED‑TO). request the user to restrict the analysis to pre-iden-

West & East 70 Monografie, 4 Thinking data. Integrative big data approaches towards an ‘introspective’ digital archaeology in the ancient Mediterranean

Figure 4 Test query#2: the Neo4j graph

(for reasons of simplification and data readability, relationships related to place- name are visible only for site ID 017.1 by example)

tified typologies of results (which is the prelimi- function of each site (f:INTERPRETED_AS and nary condition sine qua non of the SQL language). d:function). The goal was to visualize if there was Instead of (pre)imagining the result in order to any meaningful association between pottery distri- build the query, we actually asked the software bution and site classification. itself to produce data: in short, we asked the database not only to answer to our initial question optional match (a:pottery)-[r:FOUND_IN]-> but also to contextualise its result, thus producing (b:site), (c:site)-[f:INTERPRETED_AS]-> new archaeological data. When compared with the (d:function) where a.class=“Drab Polished Blue ‘comfortable’ SQL inductive process, NoSQL que- Core” XOR a.class=“RP Punctured” return ries are much more complex but they provide an distinct a,r,b,d,f unexpected support for producing new data and new relationships. The outcome of this test query#2 was a nuclear To conclude our test, we asked Neo4j to re- graph, displayed in Fig. 4, where settlements and late the test query#1 results with the hypothesized cemeteries (shown as blue nodes) were almost

West & East 71 Monografie, 4 Francesca Chelazzi, Simone Bonzano

equally documented in terms of the occurrence tion of legacy data without the requirement of sec- of the Drab Polished Blue Core and Red Polished ondary processes of data standardization, such as Punctured wares, while farmsteads were remarka- those demanded by SQL databases. This comparably less ‘proximal’. Even during this second test, the tive test has been performed on the archaeologi- NoSQL query provided new data, as the software cal evidence from a discrete area of Southwestern did not require any pre-existing criteria. Cyprus, the Ezousas river valley, where a previous PostgreSQL/PostGIS-based study already provided interesting preliminary data. The replication of the 5. Conclusions SQL-based queries in a NoSQL environment (using Neo4j) permitted us not only to replicate the In the light of the recent claim for an ‘introspec- results but also to generate new networks of infor- tive’ and reflexive discourse about the use of digital mation without the need of over-imposing (pre)de- technology in the Humanities, including Archaeol- fined analytical conditions. ogy, this paper explores the semantic and functional The research potential of this integrative meth- transformation of archaeological data through new odology is particularly compelling, especially be- logics of data storage and mining. We developed a cause NoSQL Graph databases enhance visual func- comparative integration between the traditional tions and provide unexpected associations in an SQL RDBMS and the NoSQL DBMS. Despite the immediate and particularly explicit way. The main NoSQL/GIS integration is still experimental, the aim of this test study is to invite the archaeological underpinning logic of NoSQL databases facilitates community to experiment innovative strategies of not only the preservation of a more complex ap- data mining and to explore their theoretical impli- proach to archaeological data but also the aggrega- cations and methodological novelty.

West & East 72 Monografie, 4 Thinking data. Integrative big data approaches towards an ‘introspective’ digital archaeology in the ancient Mediterranean

BIBLIOGRAPHY

Alcock S.E., Cherry J.F. (eds.) 2004, Side-by- Cadogan G., Iacovou M., Kopaka K., Whitley J. Side Survey: Comparative Regional Studies in the (eds.) 2012, Parallel Lives. Ancient island societies in Mediterranean World, Oxford. Crete and Cyprus. Paper arising from the Conference Allison P. 2008, Dealing with Legacy data – an intro- in Nicosia organised by the British School at Athens, duction, «Internet Archaeology» 24, (https://doi. the University of Crete and the University of Cyprus, in org/10.11141/ia.24.8). November-December 2006, London. Åström P. 1972, The Middle Cypriot Bronze Age. The Carpenter J.R. 1981, Excavations at Phaneromeni: Swedish Cyprus Expedition Vol. IV, Part 1B, Lund. 1975-1978, in: Biers J.C., Sørensen L.W. (eds.), Studies in Cypriote Archaeology, Los Angeles, pp. 59-78. Atici L., Kansa S.W., Lev-Tov J., Kansa E.C. 2013, Other People’s Data: A Demonstration of the Catling H.W. 1962, Patterns of Settlement in Bronze Imperative of Publishing Primary Data, «Journal Age Cyprus, «Opuscola Atheniensia» 4, pp. 129-169. of Archaeological Methods and Theory» 20/4, Chelazzi F. 2016, Landscape Strategies in Bronze Age pp. 663-681. Southwestern Cyprus (2500-1100 B.C.), (Unpublished Barlow J.A., Bolger D.L., Kling B. (eds.) 1991, PhD dissertation, Department of Archaeology, Uni Cypriot Ceramics: Reading the Prehistoric Record, versity of Glasgow), Glasgow. Philadelphia. Cherry J.F. 1994, Regional Survey in the Aegean: the Baton J., Van Bruggen R. 2017, Learning Neo4j 3.x ‘New Wave’ (and after), in: Kardulias P.N. (ed.), – Second Edition, Birmingham. Beyond the Site: Regional Studies in the Aegean Area, Lanham, pp. 91-112. Berry D.M. 2011, The Philosophy of Software: Code and Mediation in the Digital Age, London. Conolly J., Lake M. 2006, Geographical Information Systems in Archaeology, Cambridge. Bevan A., Connolly J., Hennig C., Johnston A., Quercia A., Spencer L., Broom J. 2013b, Corbellini A., Mateos C., Godoy D., Zunino Measuring chronological uncertainty in intensive A., Schiaffino S. 2015, An architecture and plat- survey finds: a case study from Antikythera, Greece, form for developing distributed recommendation al- <> 55(2), pp. 312-328. gorithms on large-scale social networks, «Journal of Information Science» 41/5, pp. 686-704. Bevan A., Crema E.R., Li X., Palmisano A. 2013a, Intensities, interactions and uncertainties: some new Crema E.R. 2012, Modelling Temporal Uncertainty in approaches to archaeological distributions, in: Bevan Archaeological Analysis, «Journal of Archaeological A., Lake M. (eds.), Computational Approaches to Method and Theory» 19, pp. 440-461. Archaeological Spaces, Walnut Creek, pp. 27-52. Crema E.R., Bevan A., Lake M. 2010, A probabi- Blanton R.E. 2001, Mediterranean Myopia, listic framework for assessing spatio-temporal point «Antiquity» 75, pp. 627-629. patterns in the archaeological record, «Journal of Archaeological Science» 37/5, pp. 1118-1130. Bombardieri L., Scirè Calabrisotto C., Chelazzi F., Albertini E. 2012, Dating the con- Crewe L., Croft P., Graham L, McCarthy A. texts (or contextualising the datings?): new eviden- 2008, First preliminary report of excavations at ces from the southern cemetery of Erimi-Laonin tou Kissonerga-Skalia, 2007, «Report of the Department Porakou, «Cahier du Centre d’Études Chypriotes» of Antiquities, Cyprus», pp. 105-120. 41, pp. 87-108. De Runz C., Desjardin E., Piantoni F., Herbin M. Boyd D., Crawford K. 2012, Critical questions for Big 2007, Using fuzzy logic to manage uncertain multi- data: provocations for a cultural, technological and scho- modal data in archaeological GIS, in: Stein A. (ed.), th larly phenomenon, «Information, Communication Proceedings of the 5 International symposium on and Society» 15/5, pp. 662-679. Spatial Data Quality SDQ 2007, Modelling qualities in space and time, ITC, Enschede, The Netherlands, Burrough P.A., McDonnell R. 1998, Principles of 13-15 June 2007, Enschede. Geographical Information Systems, Oxford.

West & East 73 Monografie, 4 Francesca Chelazzi, Simone Bonzano

Durand, G., Belacel, N., LaPlante, F. 2013, Graph Held S.O. 2003, The Regional Archaeological Survey, in: theory based model for learning path recommendation, Swiny, Rapp, Herscher (eds.) 2003, pp. 467-474. «Information Sciences» 251, pp. 10-21. Herscher E. 1976, South Coast Ceramic Style at the Engel C., Grossner K. 2015, Representing the End of Middle Cypriote, «Report of the Department Archaeological Process at Çatalhöyük in a Living of Antiquities, Cyprus», pp. 11-19. Archive, in: Hodder, I., Marciniak, A. (eds.) Herscher E. 1991, Beyond Regionalism: Toward an Assembling Çatalhöyük, Leeds, pp. 13-24. Islandwide Middle Cypriote Sequence, in: Barlow, Evans T. 2015, A Reassessment of Archaeological Grey Bolger, Kling (eds.) 1991, pp. 45-50. Literature: semantics and Paradoxes, «Internet Herscher E. 2003, The ceramics, in: Swiny, Rapp, Archaeology» 40 (https://doi.org/10.11141/ia.40.6). Herscher (eds.) 2003, pp. 145-210. Frankel D. 1974, Middle Cypriot White Painted Huggett J. 2015a, A Manifesto for an Introspective Pottery: An Analytical Study of the Decoration, Digital Archaeology, «Open Archaeology» 1/1, Göteborg. pp. 86-95. Gattiglia G. 2015, Think big about data: Archaeology Huggett J. 2015b, Challenging digital archaeology, and the Big Data Challenge, «Archäologische «Open Archaeology» 1/1: pp. 79-85. Informationem» 38, pp. 1-12. Hunger M. 2014, Neo4j 2.0 – Eine Graphdatenbank Georgiou G. 2006, The topography of Human für alle, Frankfurt. Settlement in Cyprus in the Early and Middle Bronze Jaroslaw J., Hildebrandt-Radke I. 2009, Ages (Unpublished PhD dissertation, Department Using of Archaeology, University of Cyprus) (in Greek), multivariate statistics and fuzzy logic system to analyse Nicosia. settlement preferences in lowland areas of the temperate zone: an example from the Polish Lowlands, «Journal Given M., Knapp A.B., Coleman D. (eds.) 2003, The of Archaeological Science» 36/10, pp. 2096-2107. Sydney Cyprus survey project: social approaches to re- Jasink A.M., Bombardieri L., Fossataro D., gional archaeological survey, Los Angeles. Menozzi O. 2008, The Kouris river survey project: Gjerstad E. 1926, Studies on Prehistoric Cyprus. 2007 preliminary report, «Report of the Department Inaugural dissertation, Uppsala. of Antiquities Cyprus», pp. 159-182. Gjerstad E., Lidros J., Westholm A. 1934, The Kansa E.C. 2007, Publishing Primary Data on the Swedish Cyprus Expedition. Finds and results of the World Wide Web: Opencontext.org and an Open Excavations in Cyprus 1927-1931. Vol. I, Stockholm. Future for the Past, «Technical Briefs in Historical Graham L. 2012, The necropolis at Kissonerga- Archaeology» 2/1, pp. 1-11. Ammoudhia: new ceramic evidence from the Early Kansa E.C., Kansa S.W. 2013, We all know that 14 Middle Bronze Age in western Cyprus, in: Georgiou is a sheep. Data publication and professionalism in A. (ed.) Cyprus, An Island Culture: Society and Social archaeological communication, «Journal of Eastern Relations from the Bronze Age to the Venetian Period, Mediterranean Archaeology and Heritage Studies» Oxford, pp. 38-47. 1/1, pp. 88-97. Guldager Bilde P. 1993, Prehistoric Pottery, in Kansa E.C., Kansa S.W., Arbuckle B. 2014, Sørensen, Rupp (eds.) 1993, pp. 1-36. Publishing and Pushing: Mixing Models for Hadjissavvas S. 1977, The archaeological survey Communicating Research Data in Archaeology, of Paphos. A preliminary Report, «Report of the «International Journal of Digital Curation» 9/1, Department of Antiquities, Cyprus», pp. 222-231. pp. 57-70. Harbert B., Huc C. 2010, Building together digital Kantner J. 2008, The Archaeology of Regions: From archives for research in social sciences and humanities, Discrete Analytical Toolkit to Ubiquitous Spatial «Social Science Information» 49/3, pp. 415-443. Perspective, «Journal of Archaeological Research» Harley D. 2013, Scholarly communication: Cultural 16, pp. 37-81. contexts, evolving models, «Science» 342, pp. 80-82. Kintigh K. 2006, The Promise and Challenge of Held S.O. 1993, Insularity as a Modifier of Culture Archaeological Data Integration, «American Change: The Case of Prehistoric Cyprus, «Bulletin of Antiquity» 71/3, pp. 567-578. the American Schools of Oriental Research» 292, Knapp A.B. 1986, Production, Exchange, and Socio- pp. 25-33. Political Complexity on Bronze Age Cyprus, «Oxford Journal of Archaeology» 5/1, pp. 35-60.

West & East 74 Monografie, 4 Thinking data. Integrative big data approaches towards an ‘introspective’ digital archaeology in the ancient Mediterranean

Knapp A.B. 1993, Thalassocracies in Bronze Age Eastern stigations 1979-82 (Report of the Department of Mediterranean trade: Making and breaking a myth, Antiquities, Cyprus), pp. 9-55. «World Archaeology» 24/3, pp. 332-347. Piowar J., Vision T.J. 2013, Data reuse and the open Knapp A.B. 2007, Insularity and island identity in the data citation advantage, «PeerJ» 1, e175, (https:// prehistoric Mediterranean, in: Antoniadou S., doi.org/10.7717/peerj.175). Pace A. (eds.), Mediterranean Crossroads, Athens, Robinson I., Webber J., Eifrem E. 2015, Graph pp. 37-62. Databases, 2nd Edition. New Opportunities for Knapp A.B. 2008, Prehistoric & Protohistoric Cyprus. Connected Data, Sebastopol. Identity, Insularity, and Connectivity, Oxford. Rupp D.W., King R.H. 1983, Canadian Palaipaphos Kydonakis, A., Chliaoutakis, A., Sarris, A. Survey Project (CPSP), in: Keller D.R., Rupp D.W. 2013, GIS Based Application for the Management of (eds.), Archaeological Survey in the Mediterranean Monuments and Antiquities of Cyprus, in: Perakis, Area, Oxford, pp. 323-327. nd K., Moysiadis, A. (eds.), 32 EARSeL Symposium Sherratt S., Sherratt A. 1993, The growth of the st Proceedings “Advances in Geosciences”, 1 EARSeL Mediterranean economy in the early first millennium Workshop on Temporal Analysis of Satellite Images, BC, «World Archaeology» 24/3, pp. 361-378. Mykonos, pp. 141-151. Simmons A.H. 2005, Ais Yiorkis, An Upland Aceramic Lawrence D., Bradbury J., Dunford R. 2012, Neolithic Site in Western Cyprus. Progress Report of Chronology, Uncertainty and GIS: A Methodology the 2003 Excavations (Report of the Department of for Characterising and Understanding Landscapes Antiquities, Cyprus), pp. 23-30. of the Ancient Near East. Landscape Archaeology Smith E. (ed.) 2011, Conference (LAC 2012), «eTopoi. Journal for The Comparative Archaeology of s, Cambridge. Ancient Studies» Special Volume 3, pp. 353-359. Complex Societie Sørensen L.W., Rupp D.W. (eds.) 1993, Lukas D., Engel C., Mazzucato C. 2018, Towards a The Land of , Göteborg. Living Archive: Making Multi Layered Research Data the Paphian Aphrodite. Volume 2 and Knowledge Generation Transparent, «Journal of Stanley Price N.P. 1979, The Structure of Settlement Field Archaeology» 43/1, pp. 19-30. at Sotira in Cyprus, «Levant», pp. 46-83. Manning S., Hulin L. 2008, Maritime Commerce Swiny S. 1979, Southern Cyprus. 2000-1500 B.C and Geographies of Mobility in the Late Bronze Age (Unpublished PhD dissertation, Institute of of the Eastern Mediterranean: Problematization, in: Archaeology, University of London) London. Blake E., Knapp A.B. (eds.), The Archaeology of Swiny S., Rapp G., Herscher E. (eds.) 2003, Sotira Mediterranean Prehistory, Hoboken, pp. 270-302. Kaminoudhia. An Early Bronze Age site in Cyprus, McCarthy A., Blakeman B., Collard D., Croft Boston. P., Graham L., McCartney C., Stork L. 2010, Todd I.A. 1996, The Vasilikos Valley: its place in Cypriot The Prasteio-Mesorotsos Archaeological Expedition: and Near Eastern prehistory, in: Coleson J.E., Second Preliminary Report of the 2009 Excavations, Matthews V.H. (eds.), “Go the Land I will show «Report of the Department of Antiquities, Cyprus», you”. Studies in Honour of Dwight W. Young, Winona pp. 53-76. Lake, pp. 317-351. McCarthy C. 2014, Does NoSQL have a place in GIS? Todd I.A. 2004, Field survey in the Vasilikos Valley, in: – An open-source spatial database performance com- Iacovou M. (ed.), Archaeological Field Survey in parison with proven RDBMS (Unpublished MSc Cyprus. Past History, Future Potentials. Proceedings dissertation, School of GeoSciences, University of of a conference held by the Archaeological Research Edinburgh), Edinburgh. Unit of the University of Cyprus, 1-2 December, 2000, Merrilleess R.S. 1973, Sanctuary and Cemetery in London, pp. 43-54. Bronze Age Cyprus, in: Hennessy J.B., Frankel Van Eijnatten J., Pieters T., Verheul J. et Al. D., Merrillees, R.S. (eds.), The Cypriot Bronze Age. 2013, Big Data for Global History. The Transformative Some recent Australian contributions to the prehistory Promise of Digital Humanities, «Low Countries of Cyprus, Sydney-Birmingham, pp. 44-57. Historical Reviews» 128/4, pp. 55-77. Philip G. 1983, Kissonerga Skalia ceramics, in: Vogiatzakis I.N., Pungetti G., Mannion A.M. Peltenburg E., Croft P., Stewart J.D., (eds.) 2008, Mediterranean Island Landscapes. Woodhead J., Elliott C., Philip G., The Natural and Cultural Approaches, New York-London. Prehistory of West Cyprus: Ktima lowlands inve-

West & East 75 Monografie, 4 Francesca Chelazzi, Simone Bonzano

Wallis, J.C., Rolando, E., Borgman, C.L. 2013, If Webb J.M, Frankel D. 2013, Cultural Regionalism we share data, will anyone use them? Data sharing and and Divergent Social Trajectories in Early Bronze Age reuse in the long tail of science and technology, «PLoS Cyprus, «American Journal of Archaeology» 117/1, ONE» 8/7, e6773, (https://doi.org/10.1371/ pp. 59-81. journal.pone.0067332). Wesson C.B., Cottier J. 2014, Big Sites, Big Questions, Webb J.M, Frankel D. 1999, Characterizing the Philia Big Data, Big Problems: Scales of Investigation and facies. Material culture, chronology and the origin Changing Perceptions of Archaeological Practice in the of the Bronze Age in Cyprus, «American Journal of Southeastern United States, «Bulletin of the History Archaeology» 103/1, pp. 3-43. of Archaeology» 24/16, pp. 1-11.

Webliography

Allison L.A. 2012, Preserving Digital Archaeological Geser G. 2016, WP15 Study: Towards a Web of Data, «Library of Congress», November 19, 2012 Archaeological Linked Open Data, Octoer 6, 2016 (https://blogs.loc.gov/thesignal/2012/11/pre- (http://www.ariadne-infrastructure.eu/content/ serving-digital-archaeological-data/), last access: download/8392/49194/file/ARIADNE_archaeolo- January 2, 2019. gical_LOD_study_10-2016.pdf), last access: January Altaweel M. 2016, GIS and NoSQL Database, «GIS 2, 2019. Lounge», May 5, 2016 (https://www.gislounge. Huggett J. 2016, Biggish Data. Introspective Digital com/gis-nosql-databases/), last access: January 2, Archaeology, «Introspective Digital Archaeology», 2019. May 20, 2016 (https://introspectivedigitalarchae- Archaeology Data Service 2014, Data ology.wordpress.com/2016/05/20/biggish-data/), Management and Sharing Plans, August 2014 last access: January 2, 2019. (http://archaeologydataservice.ac.uk/advice/ Newhard J. 2013, Archaeology, Humanities, and DataManagementPlans.xhtml), last access: January Data Science, «The ArchaeoInformant», August 2, 2019. 1, 2013 (http://blogs.cofc.edu/thearchaeoinfor- AthenaPlus 2015, D4.2 Review on Linked Open Data mant/2013/08/01/archaeology-humanities-and-da- Sources, October 2013 (http://www.athenaplus.eu/ ta-science./), last access: January 2, 2019. getFile.php?id=190), last access: January 2, 2019. North K. 2009, Databases in the cloud, «Dr. Dobb’s», Bennett T. 2015, NoSQL for GIS Applications, September 2009 (http://www.drdobbs.com/databa- «Linkedin», September 9, 2015 (https://www.lin- se/218900502), last access: January 2, 2019. kedin.com/pulse/nosql-gis-applications-thomas- Popescu A. 2010, NoSQL at CodeMash – An Interesting bennett), last access: January 2, 2019. NoSQL categorization, February 18, 2010 (http:// Berners-Lee T. 2009, Linked Data, «W3C», June nosql.mypopescu.com/post/396337069/presenta- 18, 2009 (https://www.w3.org/DesignIssues/ tion-nosql-codemash-an-interestingnosql), last ac- LinkedData.html), last access: January 2, 2019. cess: January 2, 2019. Cambridge Dictionary 2017, Data (https://dic- Schutzberg A. 2011, NoSQL Databases: What tionary.cambridge.org/dictionary/english/data), last Geospatial Users Need to Know, «DM Insights on access: January 2, 2019. Location», February 17, 2011 (https://www.direc- tionsmag.com/article/2045), last access: January 2, Cattell R. 2010, High Performance Scalable Data 2019. Stores, February 22, 2010 (http://cattell.net/datastores/Datastores.pdf), last access: January 2, 2019. Strauch C. 2009, NoSQL Databases, June 10, 2009 (https://www.christof-strauch.de/nosqldbs.pdf), Costas R., Meijer I., Zahedi Z., Wouters P. 2013, last access: January 2, 2019. The Value of Research Data – Metrics for datasets from a cultural and technical point of view, «Knowledge Yen S. 2009, NoSQL is a horseless carriage, November Exchange», April 2013 (http://repository.jisc. 2009 (https://www.slideshare.net/northscale/ ac.uk/6205/1/Value_of_Research_Data.pdf), last nosqloakland-200911021), last access: January 2, access: January 2, 2019. 2019.

West & East 76 Monografie, 4 The Crisis Areas Archaeological Database (caad): a WebGIS for monitoring and safeguarding archaeological heritage

FRANCESCA CIOÈ*, MARZIA MERLONGHI MIANI**

* InfoCamere scpa, ** Indipendent Researcher

Abstract

The current political instability in the Near East is at the heart of complex issues of management and protection of local archaeological heritage. In the last century, many archaeological sites have been victim of conflicts, which caused huge damages. Although it constitutes a well-known problem inside the international community, a consistent and reliable report on damage is still missing. The Crisis Areas Archaeological Database (CAAD) aims to create an online open source database, a WebGIS platform, for collecting data relating archaeological heritage in Near Eastern crisis areas, monitoring their status in real time and documenting the extent of the damage suffered by them with photos, maps and, where possible, a comparison with existing archaeological documentation. The data collected will be accessible through a dynamic, searchable and interactive on-line maps, which, will allow access to several data such as name, geographical references, date of survey, presence of regular excavations/ restorations, type of damage, date of damage and eventual multimedia contents of the site. At present, a demonstration version of CAAD WebGIS regarding southern Levant is under construction. The goal of the project is the creation of a WebGIS available and updatable by all the scholars who, in their work, encounter damage to archaeological heritage in crisis areas and in all the Near East.

Keywords

Database, WebGIS, cultural heritage, conflicts, Near East, looting

West & East 77 Monografie, 4 Francesca Cioè, Marzia Merlonghi Miani

1. Introduction (historical, archaeological, artistic and architectural heritage). In war theatres (civil wars, wars, unconventional In this perspective, the CAAD project aims to conflicts, terrorist attacks), in addition to civilian develop a WebGIS in order to share data about casualties, cultural heritage is also subject to serious the conditions of the archaeological heritage in the damage. The loss of cultural property and archaeo- Near East. The data sharing will allow the monitor- logical heritage is a serious problem during conflicts ing of cultural heritage in crisis areas in real time.4 and post-conflict times. The situation is getting At present, the project is still in the early stag- worse, especially in Syria and Iraq (not to mention es: we developed a CAAD WebGIS demonstration Afghanistan or Yemen), but also in countries such version in order to test the feasibility and its real po- as Lebanon, Israel and Palestine. tential toward a full development in the future.5 In this climate of high political instability, cul- For this demonstration version we used a selec- tural heritage is damaged not only by acts of war – as tion of data collected by Marzia Merlonghi for her the ones provided by the Islamic State – but also by PhD project about damages to pre-classical sites in negligence and lack of supervision by the local au- Palestine and Israel.6 thorities and public institutions responsible for its Several challenges with different problems and protection. solutions will be addressed in the next years by ex- We are witnessing a never-ending drain of cul- perts of cultural heritage: from cultural genocide7 tural resources: in some cases, there are massive at- to illicit antiquities traffic.8 In this framework, it be- tacks against heritage, like in Palmyra or in Mosul, comes necessary to safeguard archaeological herit- in other cases, such as in Israel and Palestine, the de- age through a deep knowledge of the present condi- struction is slower but unrelenting.1 tions of the sites. Moreover, it is important to study There is a real risk that the Near East will lose it 2 all in a short span of time. Consequently, local peo- 4 Currently, there are other projects that aims to moni- ple will be facing not only the usual post-war chal- tor the state of conservation of archaeological sites in the Near lenges of social and economic nature, but also the East: the EAMENA database of the Oxford University (BEW- LEY, R. ET AL., 2016), for instance, uses almost only remote bewilderment caused by the loss of their own his- sensing and satellite images. This method is very useful during torical and cultural identity. wars since it is almost impossible go safely on the ground. A people without its history cannot exist3; as writ- On the other hand, CAAD, a was born to order data collected during survey activity. Using mainly data “from the ten in the revision of the Burra Charter approved by ground” involve a different method of observation and has dif- ICOMOS in 2013 “Places of cultural significance en- ferent purposes: it is helpful in post-war situations, where there rich people’s lives, often providing a deep and inspira- is a need to verify the situation together with local authorities and citizens. As far as we know, there are no other WebGIS tional sense of connection to community and land- programs developed mainly to order and collect data from scape, to the past and to lived experiences”. emergency survey. Current conditions of generalized and wide- 5 Unlike most of this kind of projects, CAAD is not spread crisis in the Near East require cultural herit- sponsored by a University or another institution. 6 age experts (from the Near East and from western Merlonghi 2015. The database we are developing is mainly dedicated to the southern Levant because of person- countries) to develop new methodological approach- al experience. The relative safeness of the region (if compared es for monitoring and safeguarding cultural heritage with, for example, Iraq and Syria) and, at the same time, a situation of never-ending conflict, contribute to make it an ideal laboratory for developing techniques of intervention in post- 1 The so called “second intifada” and the reoccupation of war situations (Ruggiero Maniscalco 2014, pp. 93-94). large parts of the West Bank in 2002 and 2003, the war against All the data we are using are original and the collection method Lebanon in 2006, the periodic military operations in the Gaza is explained in the next paragraph. Strip (Arrigoni 2009, passim) and the rockets of Hamas on 7 This term, even if it is widely disputed, indicates the de- southern Israel: all these episodes affect the conservation of her- liberate erasing of the tangible and intangible cultural proper- itage in southern Levant (Maniscalco 2005, p. 97). ties tied to a specific people Starrenburg( 2014; Akhawan 2 Nigro 2014, pp. 1-2. 2016). 3 Maniscalco, Mengozzi 2002, p. 79. 8 Yahya 2008, 39-55.

West & East 78 Monografie, 4 The Crisis Areas Archaeological Database (CAAD): a WebGIS for monitoring and safeguarding archaeological heritage

the anthropic9 causes of decay and to find possible areas”.12 This form allows the collection, during a solutions. preliminary visit to a monument, of the main infor- An on-line database with as much information mation needed in order to check the state of histori- as possible about sites (documenting the extent of cal buildings, monuments and archaeological sites damage with pictures, maps, plans, videos) is key in- (fig. 1). It was developed for the monuments and strument for understanding typology and causes of cultural properties in Bosnia and it formed the main damage and it will empower us to prevent them.10 tool for monitoring and safeguarding cultural prop- The definitive goal of the CAAD project is to erty at the end of the Balkan wars: it allows an easy, develop an online open-source application updated fast and accurate check of the situation in order to by all the scholars and local authorities who, during pinpoint the main actions needed. their work, encounter damage to archaeological This form, with some changes due to the differ- heritage in Near East: it will thus be possible to ac- ent operation set, was very helpful to check the sit- quire information on a growing number of sites in uation of the archaeological heritage in Palestine- conflict areas, monitoring their changes across time. Israel. In this case, the research started choosing a number of pre-classical archaeological sites:13 the first 2. Working method in post-war contexts: step was to study the history of the archaeological assessment of sites damage and data researches in these sites and to find all the availa- collection ble bibliographic and photographic material about them (a sample of 101 archaeological sites).14 The second step was to conduct field assessment In this regard, it is possible to define a working in order to check the transformations and the dam- method in emergency survey and intervention. This ages the sites had suffered in the last 50 years (from method is based on the experiences of Fabio Ma- 196715 to today):16 look at the way immoveable her- niscalco in Albania, Kosovo and Bosnia during the itage changes through time (confronting historical 1990s. Maniscalco, when he served as officer in the pictures and bibliographic material with what we Italian Army, developed one of the first methodolo- can see on the ground) can help to save, protect and gies for monitoring and protecting cultural heritage even restore monuments. An archive of more than in times of conflict.11 In the early years of the new 5000 digital pictures (originals, taken by the writer century he began working in Palestine but his pre- on field between 2011 and 2016) was set up. Each mature death left this work unfinished. site has a folder with pictures, joined with a digital Maniscalco’s pioneering surveys and emergency copy of the form completed during the survey. This interventions started a new phase of cultural her- database is constantly updated. itage protection during and after conflicts. In the surveys conducted during the peace-keeping operations in Albania and Bosnia Herzegovina, Ma- 12 Maniscalco 2007, p. 89. niscalco used a schedule developed by himself, the 13 Pre-classical sites are more fragile and exposed because “Form for the immovable cultural heritage in crisis of the materials used in these eras. Moreover, in historical Pal- estine, a main part of the ideological/political battles centre on 9 We excluded all the natural causes since the main goal of the pre-classical period, especially the early Iron Age. this project is to focus on the political crisis and war situations. 14 The sites were surveyed by Merlonghi during her PhD 10 The method for collecting this data will be explained in researches. The sample covers most of the territory of Historical the next paragraphs. Palestine (Israel and Palestine) from the Galilee to the Negev. 15 11 Sudiro, Rispoli 2015, pp. 40-44. Maniscalco also in- The 1967, the startingyear of the Israeli occupation in troduced the protection of cultural heritage in the Italian army Palestine (as stated by the UN Resolution 242) it is a symbolic and trained a team of soldiers during the peace-keeping mission date and marks a gap in the history of archaeology in southern “Alba”, in Albania in 1997, as an application of the 7th article of Levant. the Hague Convention for the Protection of Cultural Property 16 The fieldwork is very important in order to provide a in the Event of Armed Conflict, signed in the Hague in 1954 better understanding of the problems. Infact, the development (Sudiro, Rispoli 2015, pp. 64-68). of solutions could be different for every single situation.

West & East 79 Monografie, 4 Francesca Cioè, Marzia Merlonghi Miani

Figure 1 Form for immoveable cultural heritage in crisis area. (after Merlonghi M. 2016, p. 3)

West & East 80 Monografie, 4 The Crisis Areas Archaeological Database (CAAD): a WebGIS for monitoring and safeguarding archaeological heritage

Figure 2 Pie chart with percentage of the degree of damage in a sample of 101 archaeological sites (after Merlonghi M. 2016, p. 4)

For the sake of accuracy damage is evaluated ac- The second challenge was to manage all the data cording to a numeric scale from 0 to 5 where the collected and create the CAAD. major score indicates better conditions:17 From the survey emerged four kinds of damage: – 0: the site disappeared or is inaccessible for mili- military or war damages, damage from modern con- tary reasons; structions, damage due to illegal digging and general damage due to deterioration or missed restoration. – 1: the site is in very bad condition and there is a Often a place can suffer from two or more different risk to lose it; kinds of damage. – 2: the site is in bad condition (missing parts of Finally, some interviews with local archaeolo- buildings, mud brick structures melted, wide- gists and people living near sites (especially in the spread looting…); occupied West Bank) were collected in order to un- – 3: the site is in adequate condition but with derstand in greater detail the connection between some serious damage (the basic pattern of the population and archaeological heritage.19 The inter- site is still visible, some conservation damage is views were briefly reported in a section of the form: present or the overall environment has deterio- they allow to figure out what was the main cause of rated); damage. These interviews would also state the way – 4: the site is in good condition with some minor the presence of experts (both native and foreign) damage; can help to preserve the archaeological remains. – 5: the site is an archaeological park or a protected area in a very good state of preservation. 3. Four main categories of damage Using this scale, it is possible to extract statistical data. For example, in the total sample of 101 sites Obviously, the division of in four main catego- almost 12% have grade 0, 11 % have grade 1, 20% ries is instrumental to the analysis: the shades of have grade 2, 25 % grade 3, 17% grade 4 and 15% a damage to an archaeological monument could grade 5 (fig. 2).18 be innumerable. A rough division is needed in order to have a first impression of the issue.20 These 17 This is very useful for statistical computing. 18 Therefore, we can affirm that in Palestine and Israel, 19 the conservation of pre-classical sites is not so good. Generally This is useful in a second stage of work, when the ex- speaking, the causes are related to the situation of the occupa- perts going to start actions of rescue and preservation. tion and to the ideological and physical struggle between Jewish 20 In the form compiled during the survey is possible to and Arab population. find a more detailed description of every single situation.

West & East 81 Monografie, 4 Francesca Cioè, Marzia Merlonghi Miani

main kinds of damage are the most common dam- small territory.26 Other factors are, the military Is- age that occur not just in southern Levant but in raeli occupation of the West Bank and the con- every war theatre too. struction of illegal settlements (fig. 4).27

1. Military damages are those related directly with 3. Illegal diggings are widespread in all the a conflict situation, such as bombing, rockets, Near East and, usually, they are due, mainly, use of weapons on archaeological monuments, to the request of the illegal market. The illegal military installations on a cultural or archaeo- excavations are common where there is need logical site, occupied and closed areas. This kind for money and where there are negative feelings of damage are widely spread, e.g., in Syria, Iraq21 (or not feelings at all) for the past among local and Yemen. In addition, the damage caused by population.28 In a conflict area, the illegal traffic airstrikes and terrestrial attacks with mortars of ancient objects spreads whenever the national and artillery are very common. In Syria, Iraq and authorities are no longer able to control sites Afghanistan, we assisted to the intentional de- and antiquity shops. Illegal diggings are cause of struction of monuments using mines and vari- the destruction of archaeological stratification ous explosive devices. This is the worst kind of and damage to the underground structures. damage since a rocket or a mine can cancel in a Examples are Kamid el-Loz in Lebanon (Fisk few second an entire monument.22 In our sample 1989, 249-52), Nippur, Hatra and numberless of 101 sites, just the 10% suffered direct military places in Iraq and in Syria. These places were damage, especially sites in rural areas and near almost destroyed by looters (fig. 5). Remote the “separation wall” (fig. 3).23 sensing is very useful to investigate the presence of this kind of damage but only on-site 2. Modern constructions affect the conservation assessment allows to understand the frequency, of archaeological sites especially in places where the deep of the excavations and the damage to there is low attention to cultural heritage or the underground structures. where a central agency for building management is missing and there is a high rate of demographic 4. General deterioration: fires, vandalism and increase.24 Modern constructions are a common lack of conservation measures are the last kind problem in the countries rich in archaeological re- of damages investigated. Some factors influ- mains or in places where a sort of sensitivity about ence the politics of in situ conservation:29 a val- the past is missing for social and historical rea- ue-based approach could penalize some sites sons25. In the sample studied, 33% of the sites are just because they are not connected with the covered by modern constructions such as houses, predominant stakeholder groups.30 A correct streets and infrastructures: one of the main causes is the high rate of population increase in a very 26 Especially in seventies and eighties the management plans developed by the Israeli military authority, the Civil administration for Judaea and Samaria, disregarded the archaeo- 21 Fales 2005. logical heritage of the Palestinian hills, causing destruction of 22 Maniscalco 2007, pp. 67-96. ancient landscapes and sites (Piccirillo 2002, 271-73). 27 23 Maniscalco 2006, pp. 85-85. Iwais et Al. 2010, p. 103. Frequently Israeli colonial settlements are near or directly above biblical sites such as Shi- 24 Iwais et Al. 2010, . passim loh (Tell Seilun) or Hebron (Tell Rumeidah). 25 In the case of Palestine it is possible to observe a real de- 28 Sajey 2010, p. 62; Al-Houdalie 2010, p. 36; tachment from the pre-islamic heritage. The reasons are various Yahya 2008, p. 498. In the mind of many Palestinians, archae- but, basically, they are tied to the fact that archaeology in Pales- ological sites relate to confiscation of land by the Israeli army. tine had a colonial approach from its start in nineteenth cent. It is understandable that looting or destroying sites is a form of (the Arab population was only marginally involved in the re- defence against expropriation. searches). Even now, a decolonization of the archaeology is still 29 missing as underlined by Glock, Taha, Gori, and other scholars. Bandarin 2011, pp. 7-16. See Glock 1994, Taha 2010, Gori 2013. 30 Valentino, Misiani 2004, pp. 30-33.

West & East 82 Monografie, 4 The Crisis Areas Archaeological Database (CAAD): a WebGIS for monitoring and safeguarding archaeological heritage

Figure 3 Khirbet el-Makhruk, Jordan valley: firing position in concrete built directly on Iron Age remains (original picture taken by the author in 2012)

Figure 4 Hebron, Tell el-Rumeideh: a palace in the illegal settlement was built in 2004 directly on the ancient remains (original picture taken by the author in 2012)

West & East 83 Monografie, 4 Francesca Cioè, Marzia Merlonghi Miani

Figure 5 Tell Kheila, Hebron Governatorate: debris from an illegal excavation (original picture taken by the author in 2012)

Figure 6 Samaria, Omri’s Palace in 2014 (original picture taken by the author in 2014)

West & East 84 Monografie, 4 The Crisis Areas Archaeological Database (CAAD): a WebGIS for monitoring and safeguarding archaeological heritage

approach to the heritage should underline the 4. The CAAD WebGIS: universal importance of the historical proper- a quick overview ties: the concept of shared heritage is a modern and valuable approach to the interpretation of a A WebGIS is a geographic information systems cultural site. Shared heritage concept applied to (GIS) published on web. It is therefore an extension a post-conflict theatre allows avoiding possible of the web application born and developed to man- vandalism and reprisal against the cultural her- age digital cartography. A WebGIS project is distin- 31 itage of the enemy. The management of a site guished by a GIS project for the specific purpose of or a monument needs to involve all the commu- information and communication sharing with oth- nities living in the area, in order to get out of the er users. It is a really flexible platform, suitable for 32 ideological struggle. Moreover, community research and monitoring purposes. based management can be a tool of micro-eco- For our project we want to develop a platform nomic development in depressed areas: in many that is accessible to everyone, without restric- cases it was possible to develop very promising tions by proprietary software license. According 33 projects such as in the Samaria-Sebaste and to open-source policies, in order to design and de- 34 in Tell Balata. The concept of shared heritage velop CAAD we chose resources and tools that re- should point toward the intrinsic value of the spect this view. This choice is in accordance with the cultural property (fig. 6) as a living place, a place web services for making data available in conform- that bears significance by itself for all the com- ity with the so-called Open Geospatial Consortium 35 munities living in the area . standards.36 The CAAD development must take into ac- After identifying all the main damage, the last chal- count the following criteria: lenge was managing all the data collected. The better way to manage the complex situation, to study pat- 1: Low maintenance and development costs: one of terns for the various areas and to underline possible the issues in managing a WebGIS is to have a serv- similarities between distant places seemed to create er (hardware). Unfortunately, traditional servers a WebGIS platform. are too expensive and wasteful at level of energy Thanks to interactive cartography is possible to consumption. In addition, due to lack of funds, understand general and specific problems of conser- it is not possible to make use of cloud services, vation in a crisis area, managing various typologies such as cloud storage for storing large amounts of of data and identifying patterns in grade or in cat- data, making them available on Web. Similarly, is egories of damage. Since on line, these maps are a expensive to rent small servers in specialized da- precious source of information for scholars all over tacenter. The ideal solution for ensuring cheap the world: a tool to spread knowledge about cultur- support for the CAAD could be provided by al heritage in danger using a uniform recording sys- Raspberry Pi, a single-board, low-cost, but high- tem. This kind of tool is powered by WebGIS. performance computer first developed in the UK by the Raspberry Pi Foundation.

2: Easy-to-use accessibility: operators and users should be able to consult and operate on CAAD by means of not high-performance devices or in areas where, for various reasons, high-speed internet connectivity services are not availa- 31 Bandarin 2011, pp. 7-16. ble. An accessible WebGIS must be well coded 32 Valentino, Misiani 2004, pp. 26-27. 33 Benelli, Hamdan, Piccirillo 2007. 34 Van den dries, Van der Linde, 2012. 36 Castronova, Goodallb, Elag 2012. For more in- 35 Perring, Van der Linde 2009, pp. 197-200. formation: https://www.opengeospatial.org/standards.

West & East 85 Monografie, 4 Francesca Cioè, Marzia Merlonghi Miani

well, easy to navigate, and working in everyone’s 2) MapServer 7.0 GIS engine39: server for map pro- browser without the need for additional plugins duction. MapServer manages WMS (Web Map or special tools. Service)40, WFS (Web Feature Service)41 and WCS (Web Coverage Service)42 standards for 3: Simplicity and ease of use: we choose an intuitive supplying through web of raster and vector car- user-friendly interface. It is not overly complex but tography. is straightforward, providing quick access to com- 3) PostgreSQL: object-relational database (ORD�- mon features and commands for average users. BMS). It uses SQL language for data querying. 4) PostGIS: provides spatial objects for the Post-� In this regard, the final version of CAAD WebGIS greSQL database, allowing storage and query of will be developed on three access levels: information about location and mapping43. It’s able to handle both alphanumeric data and vec- a) Occasional User: public access. The access will tor elements in the same record format. be public, and it will be possible to query the da- 5) Pmapper: offers widespread functionality and tabase with a specific widget, display the alpha- multiple configurations in order to facilitate numeric results and the associated media files in the setup of a MapServer application based on an appropriate form. PHP/MapScript44. b) Registered User: private access. The user will be able to login only by personal credential to The MapFile defines the relationships between ob- download multimedia file. jects, points MapServer to the pace data are located, c) Administrator: private access. By personal cre- defines how things are to be drawn and, how to cre- dentials, the user be able to insert, delete and up- ate and use maps and their layers.45 date the data inside the WebGIS. The Template files are the common HTML pages provided with MapServer specific parameters and This feature is not present in demonstration ver- variables. They are what a user can watch on brows- sion, because it is not on line and it is working only er, so they are used to present maps and cartograph- on local server. ic objects. For coding the CAAD WebGIS we’re using, for The CGI is the real engine of the CAAD Web- the server side, PHP scripts. PHP is a server-side GIS. It is started up by the web server, it processes scripting language designed primarily for web devel- both the MapFile settings and the template files de- opment. Through the coding languages server-side fined by the user’s parameters and returns the pro- as PHP it’s possible to query the database data need- cessed outputs as maps, cartographic objects, varied for building dynamically web pages. ables values and query results shown in the template Indeed, for the client server, we’re using JavaS- files. Every CGI output is a temporary image or val- cript and HTML. For displaying map data in web ue updated at each CGI work session. browsers, with no server-side dependencies, we’re using OpenLayer, a JavaScript library. To facilitate the interaction between Client and Server side we use JQuery, a multi-browser JavaScript library designed to simplify the client-side scripting of HTML. quests via HTTP. 39 CAAD WebGIS is developed using different http://www.mapserver.org/ 40 tools: http://www.opengeospatial.org/standards/wms/ 41 http://www.opengeospatial.org/standards/wfs 1) Apache HTTP Server 2.437: web server38. 42 http://www.opengeospatial.org/standards/wcs 43 http://www.postgis.net/ 37 https://httpd.apache.org/ 44 http://www.pmapper.net/ 38 A web server is a computer system that processes re- 45 http://www.mapserver.org/mapfile/

West & East 86 Monografie, 4 The Crisis Areas Archaeological Database (CAAD): a WebGIS for monitoring and safeguarding archaeological heritage

Figure 7 Home Page of CAAD WebGIS

Figure 8 A sample query in CAAD WebGIS for looking at sites damage score

West & East 87 Monografie, 4 Francesca Cioè, Marzia Merlonghi Miani

The data entered CAAD WebGIS for each site 5. Conclusions currently are: – name; To summarise, CAAD WebGIS will allow to re- – date of survey; construct information on a growing number of archaeological sites in conflict areas of Near and Mid- – last excavation year; dle East and monitor their changes through time. – main damage; As cultural heritage experts, we should improve – other damages; the use of new technologies and web sharing instru- – score of the grade of damage; ments to strengthen our efforts in saving archaeological heritage in crisis areas. Keeping in mind that – link to a folder with images and multimedia files; technology is no more than a useful instrument, it – link to digitalized original schedules filled out is of the utmost importance to investigate the real during the survey. causes behind the destruction of cultural heritage in order to be ready to work, together with local popu- From the home page the user has access to MapS- lations, in post-conflict time. This would be the last erver template files, in order to query and open the step of the preservation work: the recovery of his- map (fig. 7). By clicking on each site is possible to torical memory of a community through the protec- read all the information about it. Changing the re- tion of its cultural heritage. Working in crisis and search parameters and the map layers, is also pos- post-conflict areas means especially to flank com- sible to see the archaeological sites as dots in gray- munities to restore dignity and to give the cultural scale, according to the level of damage severity, from property an active role in the reconstruction of soci- 0 to 5 (fig. 8). ety and economy.

West & East 88 Monografie, 4 The Crisis Areas Archaeological Database (CAAD): a WebGIS for monitoring and safeguarding archaeological heritage

BIBLIOGRAPHY

Al-Houdalieh N. 2010, Archaeological Heritage Glock A. 1994, Archaeology as Cultural Survival: the and Related Institutions in the Palestinian National Future of Palestinian Past, «Journal of Palestine Territories 16 Year After Signing the Oslo Accords, Studies» XIII/3, pp. 70-84. «Present Pasts» 2/1, pp. 31-53. Glock A. 1999a., Cultural Bias in Archaeology, in: Assi E. 2012, World Heritage Sites, Human Rights Kapitan T. (ed.), Archaeology, History and Culture and Cultural Heritage in Palestine, «International in Palestine and the Near East. Essays in Memory of Journal of Heritage Studies» 18/3, pp. 316-23. Albert E. Glock, Atlanta, pp. 324-42. Bandarin F. 2011, Heritage and Dialogue, in: AA.VV. Glock A. 1999b, Divided We Stand: the Problem of Cross Cultural City: Urban Context and Cultural Palestine in: Kapitan T. (ed.), Archaeology, History Diversity, Jerusalem, pp. 8-16. and Culture in Palestine and the Near East. Essays in Benelli C., Hamdan O., Piccirillo M. 2007, Memory of Albert E. Glock, Atlanta, pp. 343-65. Sebastya: storia, conservazione e comunità locale, Greenberg R., Keinan A. 2007, Thepresent Past of Jerusalem. the Israeli-Palestinian Conflict: Israeli Archaeology in Bewley R. et Al. 2016, Endangered Archaeology in West Bank and East Jerusalem since 1967, Tel Aviv. the Middle East and North Africa: Introducing the Greenberg R., Keinan A. 2009, Israeli Archaeological EAMENA Project, in: Campana S. et Al. (eds.) Activity in the West Bank, 1967 – 2007. A Sourcebook, CAA2015: Proceedings of the 43rd annual conference Tel Aviv. on computer Application and Quantitative Methods in GORI M. 2013, The Stones of Contention: The Role Archaeology, Siena, pp. 919-932. of Archaeological Heritage in Israeli–Palestinian Bouchenaki M. 2002, Mediterranée. A propos des sites Conflict, «Archaeologies» 9/1, pp. 213-229. du patrimione culturel dans les situation post-con- Hamdan O. 2005, Problematiche generali di conservazi- flictuelles, in: Maniscalco F. (ed.), La tutela del one e gestione del patrimonio culturale della Palestina, patrimonio culturale in caso di conflitto, Napoli, in: Maniscalco F. (ed.), Tutela conservazione e val- pp. 135-142. orizzazione del patrimonio culturale della Palestina, Boylan P. 2003, The 1954 Hague Convention on the Napoli, pp. 13- 24. Protection of Cultural Property and its Protocols, in: Huisman D.J. 2009, Degradation of Archaeological Schüpbach H. (ed.), Kulturgüterschutz betrifft uns Remains, The Hague. alle, Berna, pp. 31-49. Huisman D.J. 2012, Deep Impact: What Happens when Castronova A.M, Goodall J.L., Elag M., 2013, archaeological sites are built on?, «Conservation Models as web services using the Open Geospatial and Management of Archaeological Sites» 14/1-4, Consortium (OGC) Web Processing Service (WPS) pp. 60-71. standard, «Environmental Modelling and Software» 41, pp. 72-83. Ilan D., Dahari U., Avni G., 1989, Plundered! The rampant rape of Israel’s archaeological sites, «Biblical Cioè F., Merlonghi M. 2017, CAAD (Crisis Areas Archaeology Review» XV (March/April), pp. 38-42. Archaeological Database). Un WebGIS per la sal- Iwais M. 2011, vaguardia del patrimonio archeologico a rischio nel Conservation Policies in Palestine: a , «e-dialogos» 1, pp. 24-33. Vicino e Medio Oriente, «Archeomatica» 3, pp. 6-9. Critical Review Leanza U. 2002, Dweik A.S. 2006, The Israeli Separation Wall. II Protocollo aggiuntivo del 1999 Geographic Setting, Impacts and Policy Implications, alla Convenzione de L’Aja del 1954 sulla protezi- «Urbanistica PVS» 42/43, pp. 32-35. one dei beni culturali in caso di conflitto armato, in: Maniscalco F. (ed.) La tutela del patrimonio cul- Fales F.M., 2005, Saccheggio in Mesopotamia, Udine. turale in caso di conflitto, Napoli, pp. 25-40. Fisk R. 1991, The Biggest Supermarket in Lebanon: a Maniscalco F. (ed.) 2002, La tutela del patrimonio Journalist Investigates the Plundering of Lebanon’s culturale in caso di conflitto, Napoli. Cultural Heritage, «Berytus» XXXIX, pp. 243- 52.

West & East 89 Monografie, 4 Francesca Cioè, Marzia Merlonghi Miani

Maniscalco F. (ed.) 2005a, Tutela e conservazione del nella tutela e nel recupero dei Beni culturali nelle aree patrimonio culturale della Palestina, Napoli. a rischio attraverso le esperienze, gli studi e i progetti Maniscalco F. 2005b, La convenzione dell’Aja del di Fabio Maniscalco, «Restauro Archeologico» 1-2, 1954 e la crisi in Medio Oriente, in: Maniscalco F. pp. 92-95. (ed.), Tutela conservazione e valorizzazione del patri- Sahuri S. 2011, Safeguard Documentation in Palestine, monio culturale della Palestina, Napoli, pp. 25-40. «Forum» 18, pp. 87-98. Maniscalco F. 2005c, La legge n° 51/1929 e la tutela del Schipper F.T., Eichberger H. 2010, The Protection of patrimonio archeologico palestinese, in: Maniscalco the Cultural Property in the Event of Armed Conflict: F. (ed.), Tutela conservazione e valorizzazione del pa- the Cultural Property Officier as a Liason Between trimonio culturale della Palestina, Napoli, pp. 43-45. the Military and Civil Sector, «Present Pasts» 2/1, Maniscalco F. 2005d, The blue Shield project. Pratical pp. 169-176. experiences concerning the Protection of Palestinian Silberman N.A., Frierics E.S. 2001, Archaeology and Cultural Property, in: Maniscalco, F. (ed.), Tutela Society in the XXIst.Century, Jerusalem. conservazione e valorizzazione del patrimonio cultura- Silberman N.A., SMALL D. 1997, The archaeology of le della Palestina, Napoli, pp. 93-107. Israel. Structuring the Past Interpreting the Present, Maniscalco F. 2006, Il patrimonio culturale in medio Sheffield. Oriente fra Jihad, Intifada e “guerra al terrorismo”, Stone P. 2012, Human Rights and Cultural Property «www.webjournal.unior.it» Vol.2, pp. 77-99. Protection in Times of Conflict, «International Maniscalco F. 2007, Preventive Measures for the Journal of Heritage Studies» 18/3, pp. 271-284. Safeguarding of Cultural Heritage in the Event of Sudiro L., Rispoli G. 2015, Oro dentro. Un archeolo- Armed Conflict, «www.webjournal.unior.it. » Vol.3, go in trincea: Bosnia, Albania, Kosovo, Medio Oriente, pp. 67-96. Milano. Maniscalco F., Mengozzi G. 2002, I rischi “diretti” Taha H., Jadarat M., 2009, Inventories in Palestine, che minacciano i beni culturali in caso di conflitto ar- in: Braemer F. (ed.), Proceedings of the Workshop mato, in: Maniscalco F. (ed.), La tutela del patri- of Inventories, Euromed Heritage IV, Paris, monio culturale in caso di conflitto, Napoli, pp. 73-82. pp. 83-88. Mazar A. 2013, TheArchaeological Agenda in Israel: Valentino P.A., Misiani A. 2004, Gestione del patri- past sins and future atonement, in: Sullivan S., monio culturale e del territorio: la programmazione in- Mackay R.(eds.), Archaeological sites: conservation tegrata nei siti archeologici nell’area euro- mediterra- and management, Los Angeles, pp. 261-69. nea, Roma. Merlonghi M. 2015, Strati violati, siti negati. I danni Van den Dries M.H., Van der Linde S.J. 2012, antropici al patrimonio archeologico del Levante me- Collecting oral histories for the purpose of stimulating ridionale (Israele e Palestina) nel XXI sec: problemi, community involvement at Tell Balata, Palestine, in: proposte e soluzioni (PhD Thesis), available at http:// Schucker N. (ed.), Integrating Archaeology: Science dspace-uniud.cineca.it/handle/10990/612. – Wish – Reality. International Conference on the Nigro L. 2014, David e Golia: Filistei e Israeliti ad un Social Role, Possibilities and Perspectives of Classical tiro di sasso. Recenti scoperte nel dibattito sull’archeo- Studies, Frankfurt am Main, pp. 49-56. logia in Israele, «Quaderni di Vicino Oriente» VIII, Yahya A. 2008, Looting and Salvaging: how the Wall, pp. 1-17. illegal digging and antiquities trade are ravaging Perring D., Van Dan Der Linde S. 2009, The Palestinian cultural heritage, «Jerusalem Quarterly» Politics and Practice of Archaeology in Conflict, 33, pp. 39-55. «Conservation and Management of Archaeological Yahiya A. 2008b, Managing Heritage in a War Zone, Sites» 11/3-4, pp. 197-213. «Archaeologies» 4/3, pp. 495-505. Piccirillo M. 2002, Conservazione e distruzione in Yahiya A. 2010, Looting and Salvaging the Heritage of Terra Santa, in: Maniscalco F. (ed.), La tutela Palestine, «Present Pasts» 2/1, pp. 96-100. del patrimonio culturale in caso di conflitto, Napoli, pp. 271-276. Yahiya A. 2010b, The Palestinian-Israeli draft agreement on archaeological Heritage, «Present Pasts» Ruggiero Maniscalco M.R. 2014, Il Ruolo della 2/1, pp. 72-74. Formazione e della Cooperazione nella salvaguardia,

West & East 90 Monografie, 4 The Crisis Areas Archaeological Database (CAAD): a WebGIS for monitoring and safeguarding archaeological heritage

Webliography

MapServer.org, 2016. MapServer Website. Available at: Pmapper.net, 2016. P.Mapper Website. Available at: http:// http://www.mapserver.org/ [Accessed 15 November www.pmapper.net/ [Accessed 15 November 2016]. 2016]. Postgis.net, 2016. PostGis Website. Available at: http:// Opengeospatial.org, 2016. Open Geo Spatial Consortium www.postgis.net [Accessed 15 November 2016]. Website. Available at: http://www.opengeospatial. Apache.org, 2016. Apache HTTP Server Project org [Accessed 15 November 2016]. WebSite. Available at: https://httpd.apache.org/ [Accessed 15 November 2016].

West & East 91 Monografie, 4

Settling the land: settlements pattern and ceramics in the land behind Nineveh from the Middle Bronze Age to the establishment of the Middle Assyrian State*

COSTANZA COPPINI

University of Udine

Abstract

This paper deals with the second millennium BC ceramics, and a short excursus regarding the settlement pattern, in the region east of the Tigris and north of the Upper Zab, delimited to the north by the Dohuk plain and the Zagros foothills, to the west by the Mosul Lake and the Tigris river, and to the east and south by the Navkur Plain and the Bardarash region, the so-called Land behind Nineveh. This is the survey area of the Land of Nineveh Archaeological Project, led by Daniele Morandi Bonacossi and active since 2012. The focus of this paper is the Middle and Late Bronze Age, which surface ceramic assemblages will be illustrated and discussed along with problematics that arise from their analysis, especially concerning the definition of Mittanian and Middle Assyrian ceramic assemblages. The picture emerging from the assemblages combined with results from the analysis of the settlements pattern allow to assert the strategic importance of the Land behind Nineveh in the landscape of settlements from the Middle Bronze Age to the formation of the Middle Assyrian state.

Keywords

Pottery, Settlement Pattern, Iraqi Kurdistan, Northern Mesopotamia, Middle Bronze Age, Mittani, Middle Assyrian

* This article is an outcome of the PRIN 2015 project “Archaeological Landscapes of Ancient Iraq from Prehistory to the Islam- ic period: formation, transformation, protection, and management”, funded by the Italian Ministry of Education, University and Re- search (Grant no. 20154X49JT).

West & East 93 Monografie, 4 Costanza Coppini

Figure 1 The Kurdistan Region of Iraq (Kopanias, MacGinnis, Ur 2015, p. iv)

1. Introduction which is nowadays known as the Kurdistan Au- tonomous Region (fig. 1) and has become one of The northern part of Iraq, very mountainous and the new focuses of archaeological research since characterized by water-dominated plains, 1 has 2006. 2 The favourable political situation, despite been played a prominent role in the resurgence recent events involving attempts to damage this of archaeological field-research in this country. region and its inhabitants, has fostered a resume This paper focuses on the north-eastern part of of archaeological research in the whole Kurdistan the country, i.e the region that stretches from the region, with the involvement of local and interna- Diyala River to the south up to the Iraqi-Turkish tional institutions, as it is amply displayed in the border to the north and is delimited by the course proceedings of this conference. of the Tigris River to the west. This area is part of The northern sector of Iraqi Kurdistan is the the greater Kurdistan, namely the Iraqi part of it, area of research of four international projects (fig. 2), all collaborating in the frame of the Assyrian 1 Although the northern part of Iraq is much differen- Landscape Research Group (ALRG) and also work- tiated from the point of view of the climate and landscape, as ing together with the local Directorates of Antiq- it can be seen from the soil irrigation and the subsistence strategies (Mühl 2013, p. 3), it is characterized by reddish-brown soils with gypsum deposits (Altaweel 2008, p. 9) and a landscape that ranges from semi-desert to favorable to agriculture. 2 Kopanias, MacGinnis, Ur 2015.

West & East 94 Monografie, 4 Settling the land: settlements pattern and ceramics in the Land behind Nineveh from the Middle Bronze Age…

Figure 2 The four projects collaborating in the Assyrian Landscape Research Group (ALRG) (map courtesy of F. Simi)

uities, namely Dohuk and Erbil.3 Among them, the 2. The Land behind Nineveh Land of Nineveh Archaeological Project (LoNAP in the text, Morandi 2018a, fig. 4), is based in the The Land behind Nineveh (fig. 3) is located to the northernmost part of Iraqi Kurdistan. The project east of the great Assyrian capitals Nineveh and is being led since 2012 by Daniele Morandi Bona- Dur-Šarrukin, east of the Tigris and north of the cossi under the aegis of Udine University, with the Upper Zab, delimited to the north by the Dohuk support of the Italian Ministry for Foreign Affairs plain and the Zagros foothills, to the west by the and is located in the governorates of Ninawa and Mosul Lake and the Tigris river, and to the east and Dohuk. The area in which the project works is bor- south by the Navkur Plain and the Bardarash region dered by the Zagros piedmont to the north, the Eski to the east and to the south.5 It falls into the Assyri- Mosul lake to the west, thus encompassing the fer- an Triangle, 6 having its vertices in the three Assyri- tile plain located to the north-east of ancient Nin- eveh, that is the Land behind Nineveh.4 5 Morandi Bonacossi 2018a, p. 81. 6 Thisis a triangular area east of the Tigris and north of the Lesser Zab and southwest of the mountain barrier where 3 Coppini 2018, p. 65. Taurus and Zagros meet (Radner 2011, p. 321), thus consti- 4 For detailed information about the project, see Mo- tuting the core of the lands that were under the continuous rule randi Bonacossi, Iamoni 2015; Morandi Bonacossi of the Assyrian kings from the fourteenth to the seventh centu- 2013. ry BC (Radner 2006-08, pp. 45-48).

West & East 95 Monografie, 4 Costanza Coppini

Figure 3 The Land behind Nineveh, with the marked LoNAP survey area (after Morandi Bonacossi 2013, fig. 1)

an cities Assur, Arbela and Nineveh, which were al- ological investigations conducted in the area. In fact, ternatively capitals of the Assyrian empire – as for this region has never been archaeologically system- Nineveh and Assur – or regional capitals,7 as far as atically investigated before, 9 despite its relevance in we know about Arbela. 8 The great bulk of infor- the archaeological and historical processes of Upper mation about the Assyrian Triangle in general and Mesopotamia. Its vicinity to important urban cen- the Land behind Nineveh specifically concern the tres like Nineveh, which was a religious centre in the Neo-Assyrian period and come whether from the third and second millennium BC10 and then became Assyrian capitals or from the very scattered archae- the seat of an Assyrian capital in the first millennium BC, and to Dur-Šarrukin, makes the area a strategic 7 Radner 2011, p. 322. one to the purpose of agriculture and water manage- 8 Archaeological excavations in Arbela, modern Erbil, have started only recently (Nováček et Al 2008), and the . 9 information that we have up to now come mostly from its Morandi Bonacossi 2017a, p. 98; 2018a, p. 82. name engraved in one of the Neo-Assyrian reliefs from Nineveh 10 Westenholz 2004, p. 10; Morandi Bonacossi (Albenda 1980, p. 6). 2017b, p. 107; Reade 2005, p. 355; Ziegler 2004, p. 20.

West & East 96 Monografie, 4 Settling the land: settlements pattern and ceramics in the Land behind Nineveh from the Middle Bronze Age…

ment. Besides the exploitation of the land and its im- dom of Nurrugûm,19 that was independent until it portance in the first millennium BC, the less visible was conquered by Šamši-Adad and integrated in its records related to its life during the second millenni- Kingdom of Upper Mesopotamia.20 In the Middle um BC provide evidence of the occupation. Assyrian period the entire Assyrian Triangle was The entire Land behind Nineveh is characterized the core area of the Middle Assyrian Kingdom,21 by a very fertile soil (Brown Soils, Deep Phase)11 and and the Land behind Nineveh was strategic for the by an annual rainfall from 450 to 600 millimetres connection to the Assyrian enclave in the Upper Ti- per year: this makes the land reliable for agriculture.12 gris region,22 and, first of all, for the expansion of the The Nakvur Plain – the “plain of mud” in Kurman- state, which needed more cultivable and fertile land, ji13 – is the largest plain in the Land behind Nineveh, and water sources.23 being 30 km wide and triangular;14 it is crossed by The territory of the Land behind Nineveh con- the River al-Khazir, a tributary of the Greater Zab, stituted the economic solid background that was and by two smaller rivers, the Nardush and the necessary to the establishment and development of Gomel,15 and the Jebel Maqloub geographically political entities, whether small independent king- separates it from the plain of Nineveh.16 Given the doms or complex states. The occurrence in this vast presence of the above-mentioned water courses, the area, which still plays a primary role in the agricultur- Navkur Plain is very well supplied with ground wa- al production, of a huge number of sites allows the ter, thus turning this area into the ideal playground reconstruction of the occupation and the settlement for agricultural production, in fact the area is inten- pattern of this region. These sites have been identified sively cultivated. Therefore it is not surprising that during the survey carried out by LoNAP and restitut- the Navkur Plain hosts most of the sites detected in ed a consistent number of second millennium settle- the course of the survey conducted by LoNAP and ments, which have been dated through the examina- therefore was densely and continuously inhabited tion of the pottery retrieved. In this paper the ceramic during the Bronze and Iron Ages:17 the settlements assemblages from those sites and dated to the second pattern, which is going to be briefly sketched in the millennium BC, and a brief excursus on the relative following paragraphs, well illustrates the continuous settlements pattern are going to be illustrated. occupation of the Land behind Nineveh and in particular of the Navkur Plain.18 As it can be argued by textual source, it can be 3. Ceramics in the Second Millennium BC assumed that the availability of fertile soils and water in the Navkur plain played an important role The dataset here presented consists of the archaeo- throughout the second millennium BC. As we logical sites recovered during the five years survey know from the Mari state correspondence of the campaigns conducted by the LoNAP team, name- eighteenth century BC, the Land behind Nineveh, ly to the 2012 to 2016 survey campaigns. The sites or at least the Navkur Plain, was part of the King- have been recognized through two combined methodologies, that is the extensive24 and the intensive survey, which has been applied on a larger scale in a 11 Morandi Bonacossi 2018a, p. 84. delimited area in the territory around Gir-e Gomel.25 12 Morandi Bonacossi 2018a, p. 85. 19 13 Reade, Anderson 2013, p. 69. Morandi Bonacossi, Iamoni 2015, p. 24. 20 14 Morandi Bonacossi 2018a, p. 84. Charpin, Ziegler 2003, p. 79, p. 91-99, p. 101; Ziegler 2004. 15 Morandi Bonacossi 2018a, p. 87. 21 Radner 2006-08, p. 43-48. 16 Reade, Anderson 2013, p. 69. 22 Morandi Bonacossi 2018a, p. 88. 17 Morandi Bonacossi 2018a, p. 87. 23 Morandi Bonacossi 2018b, p. 49. 18 The settlement pattern will be fully and exhaustive- 24 ly discussed by D. Morandi Bonacossi in the final publication For more details about the survey methodology, see of the Land of Nineveh Archaeological Project (Morandi Morandi Bonacossi, Iamoni 2015, p. 13-14. Bonacossi in preparation). 25 For more details, see Simi in this volume.

West & East 97 Monografie, 4 Costanza Coppini

The ceramic material has been preliminari- the Navkur Plain30 on the eastern bank of the river ly processed during each survey campaign, as ex- of the same name. haustive illustrated in previous publications. 26 The The collected pottery dated to the MBA amounts results presented here refer to the work accom- to 3108 sherds. In the assemblage a distinction into plished during the study season 2017, with the four different wares is operated, following the sub- purpose of processing the sherds – 5449 in total – division usually employed when dealing with pot- with the support of the Ceramic Working Typol- tery from excavated sites: the assemblage is thus sub- ogy27 (CWT in the text). The aim of the work car- divided into Khabur Ware, Grey Ware, Common ried out during the study season was twofold: 1) to Ware, Cooking Ware (fig. 4). assess the occurrence of the ceramic types illustrat- Khabur Ware31 (1475 sherds) is the most rep- ed in the CWT; 2) to register the occurrence of resented in this period. From the survey material, survey-significant ceramic types that are not in- bowls with painted strokes on the rim (fig. 5: 1) and cluded in the CWT. 28 The results of the analysis, jars with horizontal painted bands (fig. 5: 2‑3) are as showed here, demonstrate the usefulness and the most occurring and easy to recognize. Chrono- grade of accuracy of the CWT, which is proved logically significant are the above-mentioned bowls, to be a very useful tool to process pottery from ar- with convex walls, which are diagnostic of MBA II chaeological surveys. Nevertheless, it needs to be Khabur Ware, and occur in high quantities in sites integrated with new types that are not attested in that bear MBA sequences. Valid comparisons are at- the corpus illustrated in it but are attested from tested at Tell Rijim32 in the Iraqi Jezirah, and Tell stratigraphic sequences retrieved in excavations in Leilan,33 Tell Brak,34 Tell Barri,35 Tell Chagar Bazar36 Northern Mesopotamia. Moreover, in the present in the Khabur triangle. Among closed shapes, very paper we propose a subdivision of the LBA – treat- recognizable MBA II types are large storage jars with ed in the CWT as one period – into the Mittanian ledge rim and with painted decoration on the rim / LBA I and Middle-Assyrian / LBA II. and on the upper part of the body (fig. 5: 2-3), and

30 Morandi Bonacossi 2018b, p. 60-61. 3.1. The Middle Bronze Age (MBA) 31 With the term Khabur Ware we intend the painted Ware that spread throughout Northern Mesopotamia from The Middle Bronze Age occupation is spread the beginning of the second millennium BC to the fifteenth century BC ca and characterized by a monochrome painting on through the whole area. Sites are scattered on the the clay surface. The painting can consist of horizontal bands, piedmont territory between the modern towns of strokes, geometric motifs, zoomorphic pattern, on the ceramic Dohuk, Ba’dreh, Sheikhan and Al-Qosh. There is a surface which is usually buff with chaff and calcite inclusions; the clay can be rich of sand or fine mineral inclusions if the ves- relatively high number of sites that are located along sel’s shape is a goblet or a beaker, or a small bowl. The first -oc water courses, as it is the case of the sites on the many currences of this ware are still debated: according to Schmidt wadis in the area south of the Jebel al-Qosh on the 2013 (p. 105) one of the oldest occurrences is documented in phase C7 at Tell Mozan, dated to the 2100-1950 BC (Early Je- Ba’dreh Plain. Most of the settlements are located zirah V, corresponding to Early Bronze Age V and to the histor- on the fertile and well-irrigated Navkur Plain and ical period Ur III/Isin-Larsa: see Schmidt 2013, p. 2, Tab. 1). were small-sized.29 The largest site is Gir-e Gomel, in On the other side Koliński (2014a, p. 30) proposes a later date for the earliest occurrence of Khabur Ware, i.e. 195-1800 BC (Old Jezirah I, corresponding to the Middle Bronze Age I and to the historical period defined as Old-Assyrian: seeNicol - 26 For a more detailed explanation, see Coppini 2018, le 2007: 183), thus contradicting the hypothesis of Schmidt p. 66; Gavagnin, Iamoni, Palermo 2016, p. 119. (Koliński 2014a, p. 31). 27 Ball, Tucker, Wilkinson 1989; Wilkinson, 32 Koliński 2000, Plate 38, C to E. Tucker 1995; Ur 2010, pp. 213-215. 33 Pulhan 2000, p. 445, fig. 37:4. 28 In order to do this, we considered the three criteria 34 mentioned by Ur 2010, p. 213: «an ideal survey type meets Oates D., Oates J., McDonald 1997, p. 181, three criteria: it occurs frequently, it is morphologically robust figs. 212-215, p. 187, figs. 292-298. and distinctive, and it is chronologically short-lived». 35 Baccelli, Manuelli 2008a, p. 199, Pl. 2.2 and 2.4. 29 Morandi Bonacossi 2018b, p. 60. 36 McMahon 2009, Pl. 13.1, Pl. 39.

West & East 98 Monografie, 4 Settling the land: settlements pattern and ceramics in the Land behind Nineveh from the Middle Bronze Age…

Figure 4 Chart showing the distribution of the wares in the MBA ceramic assemblage from the LoNAP survey area

Figure 5 MBA ceramic assemblage from the LoNAP survey area (© LoNAP archives)

jars with long necks, decorated with painted bands with painted bands, and their occurrence is mostly (fig. 8: 4-5). These two types are very common in all related to domestic contexts, although there is evi- MBA assemblages from Northern Mesopotamia dence from Chagar Bazar that they were used also sites. Shouldered beakers appear to be a distinctive as grave good.37 feature in MBA ceramic assemblages from excavat- Burnished Grey Ware is a ceramic cluster pe- ed sites. Since they are characterized by a fine fab- culiar to the MBA assemblage: it is characterized ric – sandy with minute calcite inclusions – they are by a grey/greyish fabric with fine vegetal and min- not easy to collect, being vulnerable to breaks more eral inclusions. In the LoNAP material it amounts than the thicker jars and bowls. Despite this incon- to 85 specimens, which are all belonging to the venient, we were able to collect some shouldered open-shapes category. The most attested type is beakers specimens from LoNAP sites, precisely 43 37 sherds. They are characterized by a short (fig. 5: 6‑8) McMahon (2009, p. 117) stresses that these vessels may have been used in a pre-burial life. McMahon 2009, or a slightly longer neck and generally decorated p. 245 Pl. 1, p. 257, Pl. 7.

West & East 99 Monografie, 4 Costanza Coppini

the carinated bowl with ridges on the upper parts contexts, which pinpoint in both cases to a religious of the walls, with outside thickened rounded rim sphere of action and not only to a domestic one, their (fig. 5: -10).9 Comparisons from excavated sites can relevance is linked to their association to written re- be found at Tell Leilan,38 Tell Brak,39 Tell Barri,40 cords. Concerning Tell Leilan, we refer to the archive Tell Chagar Bazar41 in the Khabur triangle. found in the Qarni-Lim palace. 48 The texts from Tell Common Ware constitutes a matter of debate al-Rimah are dated by limu names, thus attributed to concerning survey material, since it is seen as bare- the Middle Bronze Age IIA. 49 The other paradigmat- ly recognizable from one period to the other as it ic chronological indicator is the shouldered beaker holds few characteristics that allow to recognize it.42 (fig. 5: 15), although not widely attested, but very im-� It is in fact beyond any doubt that painted wares portant. Specimens from the survey resemble those are more easily recognizable, but this consideration from the Gir-e Gomel graves, thus dated to the Mid- does not mean that it is impossible to isolate Com- dle Bronze Age II.50 Furthermore, comparisons are mon Ware and among it chronologically signifi- attested with the site of Kurd Qaburstan,51 located in cant types. Few Common Ware types are isolated the Erbil Plain, i.e. to the south of the Navkur Plain. as distinct types in the CWT, i.e. the Horizontal- They have been found in rooms associated to storage ly Grooved Jar Shoulder (fig. 5: 11), the External- vessels, therefore associated to food preparation and ly Grooved Bowl, the High Ring or Pedestal Base, consumption contexts.52 the Channel Base (fig. 5: 12), and the Concave Fine Bowl Base. 43 Among the collected material we were 3.2. The Mittanian period53 able to isolate new types, which seem to be chronologically significant on the base of comparisons with With the coming of the Mittanian period the set- specimens from excavated sites. One diagnostic type tlements number decreases, and this goes along is the so-called Old Babylonian bowl or rough-based with the settlements pattern in other regions of bowl,44 which consists of a bowl with ledge rim and Upper Mesopotamia. In the history of research, with convex profile (fig. 5: 13 ). Thi s typ e occurs in the low number of settlements attributed to the the MBA stratigraphic sequence recovered in the Mittanian period is seen in the difficulty of rec- excavation at Gir-e Gomel: from the same levels im- ognizing Mittanian pottery in surface ceramic as- portant structures undoubtedly dated to the MBA semblages.54 In general, Mittanian settlements are have been excavated, i.e. the barrel-vaulted graves.45 A smaller than the MBA settlements, or there is a further hint to their chronological attribution is pro- high number of smaller settlements compared to vided by comparisons found in other Northern-Mes- the MBA sites size.55 opotamian sites. Particularly relevant are the speci- In terms of collected and attributed sherds, they mens from Tell Leilan, Level 2 of the Lower Town are 823. Mittanian pottery from survey assemblag- Palace and Temple,46 from Tell al-Rimah, Level 3 es has long been underestimated, as already men- of the Old Babylonian temple. 47 Besides the finding tioned before concerning the number of Mittani-

38 Frane 1996, fig. 25.1-2. 48 Eidem 2011. 39 Oates D., Oates J., McDonald, p. 177, 49 Postgate, Oates D., Oates J. 1997, p. 23. figs. 170‑172. 50 Coppini 2018, p. 81, fig. 12. 40 Baccelli, Manuelli 2008b, fig. 2.3. 51 Schwartz et Al. 2017, fig. 25.4, 6. 41 McMahon 2009, p. 301, Pl. 29.13-16. 52 Schwartz et Al. 2017, p. 229. 42 Coppini 2018, p. 69. 53 We choose to use here the historical label for this pe- 43 Ur 2010, p. 218, Table B.1. riod as a matter of uniformity with previous works related to 44 Postgate, Oates D., Oates J. 1997, p. 65. LoNAP. For a more detailed explanation, see Coppini 2018, 45 Coppini 2018, p. 72. p. 67; Pfälzner 2007, p. 257. 46 Pulhan 2000, pp. 421-422. 54 Wilkinson, Tucker 1995, pp. 58-60; Ur 2010, 47 Postgate, Oates D., Oates J. 1997, p. 149 and p. 267; Algaze, Hammer, Parker 2012, p. 31. p. 151. 55 Morandi Bonacossi 2018b, p. 61.

West & East 100 Monografie, 4 Settling the land: settlements pattern and ceramics in the Land behind Nineveh from the Middle Bronze Age…

Figure 6 Chart showing the distribution of the wares in the Mittanian ceramic assemblage from the LoNAP survey area

Figure 7 Mittanian ceramic assemblage from the LoNAP survey area (© LoNAP archives)

an settlements. This is linked to the low amount that we name as “Mittanian” finds undeniable sim- of excavated Mittani sites, the low number of pub- ilarities with materials from excavated sites located lications and the tendency to see no difference, or in the Iraqi Jezirah and in the Khabur valley, thus at least few, between Mittanian and Middle-Assyr- confirming the high degree of differentiation with ian Common Wares, as if they are not distinguish- the pottery from the Middle-Assyrian period. The able from surface assemblages. 56 The baulk of sherds Mittanian period surface assemblage is divide into five groups: Nuzi Ware, Khabur Ware, Red-Edged 56 The problematic has been treated in Coppini 2018, bowls, Grey Ware, Common Ware (fig. 6). p. 70.

West & East 101 Monografie, 4 Costanza Coppini

Nuzi Ware has an extremely low amount – it is The same assemblages can be found in sites lo- attested by 3 sherds – nevertheless it is the most di- cated in Northern Mesopotamia. We find the agnostic ware for the Mittanian period (fig. 7: 1‑2).57 same kind of carinated bowls from sites in the Khabur Ware58 is represented by 38 sherds, among Khabur Basin, which was the core of the Mittani- which the chronologically significant types are the an Kingdom, as for example from Tell Brak Mittani straight-side goblet (fig. 7: 3-6), the squared ledge Palace61 and Tell Barri Phase E,62 and on the course rim jar (fig. 7: 7-9), and the grain measure. Red- of the Khabur river, further to the south, as for ex- Edged bowls (10 sherds) are included, together ample the Mittanian levels of Tell Bderi.63 Simi- with Nuzi Ware, in the sphere of the certain Mit- lar Mittani-period ceramic is attested from Kurd tanian pottery; 59 they occur in the form of shallow Qaburstan,64 where LBA contenxts have been exca- bowls, with outside or inside thickened rim, a buff vated and restituted a building.65 smoothed surface, and the peculiar red painting, which occurs as a band on the outer and on the inner wall of the vessel and can be polished (fig. 7: 10). 3.3. The Middle Assyrian period Grey Ware (1 specimen) is represented by bowls that show new shapes when compared to those of The Middle Assyrian period is characterized by a the MBA: from the LoNAP assemblage we have sharp growth in the number of settlements, 66 when one carinated bowl (fig. 7: 11), which follows the compared to the occupation registered for the Mit- shape of Common Ware specimens. tanian epoch and going then back to a similar situ- Common Ware is more articulated in the enu- ation already registered in the MBA. 67 This growth meration of diagnostic shapes. The carinated bowl does not find any reflection in the settlement pat- type is attested by the presence of 92 specimens tern in the Upper Tigris Valley, an area that was and shows the existence of two different types: the gradually incorporated in the Middle Assyrian simple and standard carinated bowls (fig. 7: 12-15) state; 68 in the neighbouring land west of the Tigris, is distinguished from the other carinated bowls, i.e. the Iraqi North Jezirah, where a decrease of set- which may present a higher variation of rims tlements is attested.69 In the LoNAP survey area, a (fig. 7: 16‑18). This separation is due not only to relatively high number of the settlements dated to chronological purposes, but also to considerations this period shows continuity with the Mittanian pe- leading to the identification of a site type or of a riod, thus witnessing the above-mentioned growth precise function of a site’s area. The cluster of large of settlements and, as proposed by Morandi Bona- storage jars is homogenized by the occurrence of cossi, a revitalisation of the settlements network.70 squared rims (fig. 7: 19‑21) on the majority of the The analysis of the Middle-Assyrian ceramic as- sherds (55 specimens). Another characteristic and semblage from the archaeological survey does not occurring type in the whole LoNAP region is the 60 Bonacossi et Al. 2018) and Tell Rijim (Koliński 2000) pie-crust postand (27 specimens, fig. 7: 22-23). corroborate. However, comparisons for survey specimens are attested with LBA / Mittani-dated specimens from Upper 57 For comparanda: Starr 1937, Pl. 79.BB1. Khabur basin sites, such as Tell Brak, Tell Barri, Tell Bderi. 58 We refer to the so-called Younger Khabur Ware. 61 Oates D., Oates J., McDonald 1997, p. 66. 59 They occur in the whole area that was controlled by 62 D’Agostino 2014, p. 249; p. 253. Coppini 2008, Mittani, as west as Tell Atchana/Alalakh (Horowitz 2015, p. 488, fig. 3. p. 167), in the core of the Kingdom, i.e. the Khabur basin 63 Pfälzner 1995. (D’Agostino 2014, p. 174, fig. 1), and in the Tigris Valley, i.e. 64 at Nemrik (Reiche 2014, p. 303, Pl. 2.3-4), just to cite selected Schwartz et Al. 2017, fig. 28. examples. 65 Schwartz et Al. 2017, p. 219. 60 Concerning pie-crust potstands, it is an actual mat- 66 Morandi Bonacossi 2018a, p. 87. ter of debate whether they should be assigned to the MBA or 67 Morandi Bonacossi 2018b, p. 61. to the first part of the LBA, namely to the Mittani period. Re- 68 Algaze, Hammer, Parker 2012, pp. 31-33. cent excavations in Iraqi Kurdistan have shown that they can 69 be found since the MBA II, as evidences from Kurd Qaburstan Wilkinson, Tucker 1995, pp. 59-60. (Schwartz et Al. 2017, p. 232), Gir-e Gomel (Morandi 70 Morandi Bonacossi 2018b, p. 61.

West & East 102 Monografie, 4 Settling the land: settlements pattern and ceramics in the Land behind Nineveh from the Middle Bronze Age…

Figure 8 Chart showing the distribution of the types in the Middle Assyrian ceramic assemblage from the LoNAP survey area

Figure 9 Middle Assyrian ceramic assemblage from the LoNAP survey area (© LoNAP archives)

rise problems, and this is due to a double factor. The text with cuneiform texts, the majority of which is first is reflected in the nature of the assemblage, i.e. dated through limu. The sherds dated to this period the standardization of shapes that affects this mate- amount to 1518, all of them have been classified as rial; the second factor is the relatively abundance of Common Ware. Most of the shapes has been recog- data from excavated sites that are moreover in con- nized with the support of the CWT types (fig. 8),

West & East 103 Monografie, 4 Costanza Coppini

the typical Middle-Assyrian ceramic assemblage as parisons from Tell Sabi Abyad,74 Tell Barri,75 Tell found in excavated sites is reflected in almost all Sheikh Hamad,76 Tell Mohammed Arab77 and As- of the Middle-Assyrian sites of the LoNAP survey sur78 show, the type can be reconducted to the time area. of decline of the Middle-Assyrian state. The biggest part of the material is constituted The remaining spectrum of shapes mirrors the by the sharp-carinated bowls (tot. 211 specimens), well-known Middle-Assyrian diagnostic types: which are differentiated into two different types. standard bottles (fig. 9: 11-13), large storage jars The first type – constituting the majority of the car- with squared rim (fig. 9: 14-15), and nipple bases inated bowls assemblage (141 specimens, sherds and (fig. 9: 16-18) belonging to goblets. complete vessels) – is defined as the so-called standard sharp-carinated bowls (fig. 9: 1-5), which is the mass-produced and widely spread shape of bowl 4. Conclusive remarks found in all sites included or under the influence of the Middle-Assyrian state. The second type re- The landscape of settlements emerging from the unites all carinated bowls (fig. 9: 6-7) that are recog�- MBA and LBA in the Land behind Nineveh pre- nized as Middle-Assyrians – on the base of the ware sents a varied distribution of inhabited areas, a con- of comparisons with pottery from excavated sites – tinuity in the intensity of occupation is registered but are not classifiable as standard sharp-carinated for the Navkur Plain, which is a constant in the bowls. This distinction is based on the study carried whole second millennium. The region is dominat- out by Pfälzner,71 in which he is able to differenti- ed by small size settlements, with the only excep- ate between official pottery and domestic pottery on tion of Gir-e Gomel. 79 The pre-eminence of this site the base of the finding context and of morphologi- during the MBA is attested not only by data from cal and formal features. It has to be underlined that the survey, but also by data from the excavation: we can not make assumptions about the context of although dug on a small surface, part of the settle- use, but the distinction between official and domes- ment at Gomel is proved to be an elite cemetery, tic pottery is used as a typological indication.72 The with baked-bricks hypogea resembling those found wide spectrum of standard sharp-carinated bowls throughout the whole Mesopotamia. 80 We can as- provides new hints and evidence of sub-types that sume that it must have been a prominent urban have a chronological relevance. The most significa- centre in the Navkur Plain, which was part of the tive has been recognized in the sharp-carinated bowl kingdom of Nurrugûm (see section 2), and there- with grooves under the rim and on the carination fore as an urban centre we can assume that it could (fig. 9: 8-10), defined as a fine-ware type in the Mid-� have controlled the small settlements dispersed in dle-Assyrian pottery.73 This type is usually character- the plain and the agricultural production gained ized by a well-smoothed surface and a very sharp an- from the fertile soils of the region, a dynamic that gle marking the carination and a straight wall above remains unchanged even after Nurrugûm was one the carination. Few sherds have been found among of the districts of the Kingdom of Upper Mesopo- the surface material, nevertheless it is well-known tamia. Contacts with the western part of Northern that this type is chronologically relevant. As com- 74 Duistermaat 2008, fig. IV.100w; Duistermaat 71 Pfälzner 1995, p. 161. 2008, fig. IV.115d. Both comparisons are from Level 4. 75 72 Nowadays it is still a matter of debate whether the D’Agostino 2006, fig. 1.13 (Area G, Shaft 200). distinction between an official and domestic use and produc- 76 Pfälzner 1995, Taf. 113.c (Mittelassyrische Stufe tion of the Middle-Assyrian pottery has to be accepted or not. IIa-c). According to Duistermaat (2008, p. 127-128), a distinction is 77 Pfälzner 1995, Taf. 192.d. not possible at Sheikh Hamad since evidences of ceramic pro- 78 duction have not been found and they are necessary in order to Beuger 2014, fig. 9 (Assur Deep Trench). asset different contexts or different recipients of the produced 79 Morandi Bonacossi 2018b, p. 60. goods. 80 Morandi Bonacossi et Al. 2018; Coppini 2018, 73 D’Agostino 2014, p. 241. p. 72.

West & East 104 Monografie, 4 Settling the land: settlements pattern and ceramics in the Land behind Nineveh from the Middle Bronze Age…

Mesopotamia are confirmed by the ceramic assem- problematic is more related to the phases of re-con- blage, which shows consistent similarities between quering the land by the Middle Assyrian power, dy- the Land behind Nineveh, the Iraqi Jezirah and the namic that affects the length of the Mittanian rule Khabur Valley. in this area and therefore the impact on settlements The decrease of settlements registered in the pattern and material culture. It is important to take Mittani period occurs in the whole Land behind into account that the Assyrian (re)-control of this Nineveh, and even the settled surface of Gir-e part of the Assyrian Triangle84 by the Assyrian state Gomel seems to shrink, 81 thus showing a loss of could have happened in different stages,85 of which pre-eminence by this site but without a replacement the final one shows up from survey data: we can by a new preeminent urban centre. It is beyond any presume that the (re)-conquest of the Land behind doubt that we face the opposite settlement pattern Nineveh was one of the first accomplished by the re- that characterizes the Khabur valley, where the Mit- surgent Assyrian State, first reconquest operated by tani period experiences an increasing of large sites. 82 Aššur-uballiṭ in the fourteenth century BC,86 about As Morandi Bonacossi asserts, the landscape in this which the preeminent places are mentioned, but we region is marked by a ruralisation, 83 nevertheless the can assume that the agricultural land was included sphere of influence in this rural landscape is clearly in the re-appropriation of these territories.87 The pertaining to the Mittanian political entity, as the presence of small centres commits to the evidence material culture reflects the same found in Mittani- that a network of small towns and villages composed an sites in the Khabur valley and in the Balikh val- the hinterland of Assyria, thus shaping a rural land- ley. Therefore, it is evident that the archaeological scape, reflecting the interest ofMāt Aššur in these presence of a homogeneous ceramic tradition that fertile grounds.88 The presence of Middle-Assyri- extends from the Balikh valley to the territories to an settlements and an uninterrupted exploitation the east of the Tigris valley, passing through and af- of the area, a higher exploitation indeed, witnesses fecting the Iraqi Jezirah, undoubtedly exists and is the further importance of the area as an agricultur- chronologically placed between the sixsteenth and al land. This can be attested also by written sources, the fourteenth century BC: it is then our duty to from which we know that the area was controlled decide if we want to call it LBA I ceramic tradition by the Assyrian administration through the dun- or Mittanian ceramic tradition. It is beyond any nu-institution.89 As for the pottery, we tentatively doubt that it occurs at the same time as the rule of estimate the presence of the so-called standard and the Mittanian kingdom in the above-mentioned re- not-standard ceramic production, thus committing gions, although it is not here the appropriate con- a majority of standard production, confirmed also text to suggest a connection of the recognized ce- by its presence in most of the sites. The occurrence ramic tradition with the rule of Mittani. The role of the so-called standard or official Middle-Assyrian of rural landscape for the Land behind Nineveh, in pottery induce to assert the existence of a capillary which the settlement pattern is dispersed, and the sites are small, fits with its position in the Mittanian 84 Following Jakob 2003, p. 5, the city of Aššur was the state and fulfils the need of strategic natural resourc- proper Assyrian enclave in the fifteenth century BC, maintain- es – water above all – agricultural facilities, and food ing an influence on a small portion of land but still under the threat of Mittanian expansion and will conquered by Šauštatar. production, all essential factors in order to maintain 85 Llop 2012, p. 594. a social wealth. 86 Jakob 2003, p. 56. If we are on the terrain of uncertainty for the 87 See Brown 2014, p. 94. Mittanian period, the subsequent epoch of the 88 Morandi Bonacossi 2018a, p. 62. He points out Middle Assyrian rule is relatively well-known. The also the importance of the Land behind Nineveh for the transit of crucial communication routes, which the newly ruling Assyr- ian dynasty is willing to control. 81 Morandi Bonacossi 2018b, p. 61. 89 Koliński 2015, p. 22: he refers to the only 82 dunnu Koliński 2014b, p. 208. known up to now in the area east of the Tigris, i.e. Tell Billa/ 83 Morandi Bonacossi 2018a, p. 87. Šibaniba.

West & East 105 Monografie, 4 Costanza Coppini

administrative control of the Land behind Nineveh been outside investigation areas, due also to politi- during the Middle-Assyrian period. We can assume cal issues (see Introduction), thus hindering a more that the Land behind Nineveh was a necessary space complete view of the settlements history and mate- where to practice extended agriculture since the rial culture of this region of Northern Mesopota- land around Assur was well-suited for crop cultiva- mia. Thanks to recent researches, it is evident that tion, but precipitation is insufficient.90 the Land behind Nineveh is a pivotal area in order To conclude, from this picture, although frag- to understand the settlements dynamics and the mentary and relying on survey material, it clearly productive processes that affected Northern Meso- emerges that the Land behind Nineveh has long potamia in the second millennium BC.

Acknowledgements

I would like to express my gratitude to the General Directorate of Antiquities of the Kurdistan Regional Govern- ment and its director Kaify Ali and former director Kak Abubakir Othman Zeineddin (Mala Awat), and to the Duhok Directorate of Antiquities and its director Dr Hassan Ahmad Qasim. A second but not less important thank is owed to the director of the LoNAP project, Daniele Morandi Bonacossi, and to the deputy director, Marco Iamoni (University of Udine), for the possibility of studying the here presented ceramic material. A huge thank to the staff of the project, particularly to my ceramic-specialist colleagues Katia Gavagnin and Marco Iamoni (University of Udine), and Rocco Palermo (University of Groningen), as well as to Alberto Savioli (University of Udine) for work on satellite imagery and maps and Francesca Simi for the fruitful discussions about the land around Gir-e Gomel.

90 Mühl 2015, p. 45.

West & East 106 Monografie, 4 Settling the land: settlements pattern and ceramics in the Land behind Nineveh from the Middle Bronze Age…

BIBLIOGRAPHY

Albenda P. 1980, An Unpublished Drawing of Louvre Rubio I., Martínez S. (eds.), Proceedings of the Ao 1194 in the British Museum, «Journal of Ancient 5th International Congress on the Archaeology of the Near Eastern Studies» 12, pp. 1-8. Ancient Near East. Vol. I (3-8 April 2006, Madrid), Algaze G., Hammer E., Parker B. 2012, The Tigris- Madrid, pp. 478-491. Euphrates Archaeological Reconnaissance Project: Coppini C. 2018, The Land of Nineveh Archaeological Final Report of the Cizre Dam and Cizre-Silopi Plain Project: Preliminary Results from the Analysis of Survey Areas, «Anatolica» 38, pp. 1-115. the Second Millennium Bc Pottery, in: Horejs B., Altaweel M. 2008, The Imperial Landscape of Ashur: Schwall C., Müller V., Luciani M., Ritter M., Settlement and Land Use in the Assyrian Heartland, Giudetti M., Salisbury R. B., Höflmayer F., th Heidelberg. Bürge T. (eds.), Proceedings of the 10 International Congress on the Archaeology of the Ancient Near East. Baccelli G., Manuelli F. 2008a, Middle Bronze Volume 2, Prehistoric and Historical Landscapes & , in: Córdoba Khabur Ware from Tell Barri/Kahat Settlement Patterns, Economy & Society, Excavation J. Ma., Molist M., Pérez C. Ma, Rubio th Reports & Summaries (25–29 April 2016, Vienna), I., Martínez S. (eds.), Proceedings of the 5 Vienna, pp. 65-82. International Congress on the Archaeology of the Ancient Near East. Vol. I (3-8 April 2006, Madrid), D’agostino A. 2006, La ceramica dal pozzo medioassi- Madrid, pp. 187-205. ro di Tell Barri (Siria): comunicazione preliminare dei dati, «Orient Express» 2006/1, pp. 15-26. Baccelli G., Manuelli F. 2008b, Osservazioni preliminari sulla ceramica della Media Età del Bronzo D’agostino A. 2014, The Tell Barri Sequence of Late di Tell Barri, in: Pierobon-Benoit R. (ed.), Tell Bronze Age Levels: Evolution Trends Within Late nd Barri: storia di un insediamento antico tra Oriente e 2 Millennium Ceramic Culture, in: Luciani M., Occidente, Napoli, pp. 234-244. Hausleiter A. (eds.), Recent Trends in the Study of Late Bronze Age Ceramics in Syro-Mesopotamia Ball W., Tucker D., Wilkinson T. J. 1989, The Tell and Neighbouring Regions (Proceedings of the Al-Hawa Project: Archaeological Investigations in the International Workshop in Berlin, 2-5 November North Jazira, «Iraq» 51, pp. 1-66. 2006), Rahden, pp. 211-236. Beuger C. 2014, 13. Pottery Traditions from the Duistermaat K. 2008, The Pots and Potters of Assyria. Mittanian to the Early Neo-Assyrian Period – Technology and organization of production, ceramic Evidence from Soundings in Ashur and Kar-Tukulti- sequence, and vessel function at Late Bronze Age Tell , in: Luciani M., Hausleiter A. (eds.), Ninurta Sabi Abyad, Syria, Leiden. Recent Trends in the Study of Late Bronze Age Ceramics in Syro-Mesopotamia and Neighbouring Eidem J. 2010, The Royal Archives from Tell Leilan. Old Regions (Proceedings of the International Workshop Babylonian Letters And Treaties From the Eastern in Berlin, 2-5 November 2006), Rahden/Westf., Lower Town Palace (PIHANS CXVII), Leiden. pp. 263-288. Frane J.E. 1996, The Tel1 Leilan Period I Habur Ware Brown B. 2014, Settlement Patterns of the Middle Assemblage, Ann Arbor. Assyrian State: Notes toward an Investigation of State Gavagnin K., Iamoni M., Palermo R. 2016, The Apparatuses, in: Bonatz D. (ed.), The Archaeology of Land of Nineveh Archaeological Project: The Ceramic Political Spaces: The Upper Mesopotamian Piedmont Repertoire from the Early Pottery Neolithic to the in the Second Millennium BCE (International Sasanian Period, «BASOR» 375, pp. 119–69. Conference, January 2010, Berlin), Berlin, pp. 85-106. Horowitz M. 2015, The Evolution of Plain Ware Charpin D., Ziegler N. 2003, Florilegium Marianum Ceramics at the Regional Capital of Alalakh in the 2nd V: Mari Et Le Proche-Orient À L’époque Amorrite. Millennium BC, in: Glatz C. (ed.), Plain Pottery Essai D’histoire Politique, Paris. Traditions of the Eastern Mediterranean and Near Coppini C. 2008, Mitannian Pottery from Tell Barri, East, London, pp. 153-182. in: Córdoba J. Ma., Molist M., Pérez C. Ma,

West & East 107 Monografie, 4 Costanza Coppini

Jakob S. 2003, Mittelassyrische Verwaltung und Morandi Bonacossi D. 2018a, Water for Nineveh. Sozialstruktur. Untersuchungen, Leiden. The Nineveh Irrigation System in the Regional Context Koliński R. 2000, Tell Rijim, Iraq: The Middle Bronze of the ‘Assyrian Triangle’: A First Geoarchaeological Age Layers, Oxford. Assessment, in: Kühne H. (ed.), Water for Assyria (April 2016, Vienna), pp. 77-116. Koliński R. 2014a, 20th Century Bc in the Khabur Triangle Region and the Advent of the Old Assyrian Morandi Bonacossi D. 2018b, The Creation of the Trade with Anatolia, in: Bonatz D. (ed.), Assyrian Heartland: New Data from the ‚Land Behind The Archaeology of Political Spaces: The Upper Nineveh‘, in: Stek T. D, Düring B. (eds.), The Mesopotamian Piedmont in the Second Millennium Archaeology of Imperial Landscapes. A Comparative BCE (International Conference, January 2010, Study of Empires in the Ancient near East and Berlin), Berlin, pp. 11-34. Mediterranean World, pp. 48-85. Koliński R. 2014b, Settled Space. Evidence for Changes Morandi Bonacossi D. (ed.) in preparation, Land of in Settlement Patterns of Northern Mesopotamia at Nineveh Archaeological Project 1. the Advent and at the Turn of the Mittani Era, in: Morandi Bonacossi D., Iamoni M. 2015, Landscape Cancik-Kirschbaum E., Brisch N., Eidem J. and Settlement in the Eastern Upper Iraqi Tigris and (eds.), Constituent, Confederate and Conquered Space. Navkur Plains: The Land of Nineveh Archaeological The Emergence of the Mittani State (International Project, Seasons 2012–2013, «Iraq» 77, pp. 9-39. Conference, summer 2009, Berlin), pp. 179-212. Morandi Bonacossi D., Qasim H.A., Coppini Koliński R. 2015, Making Mittani Assyrian, in: C., Gavagnin K., Girotto E., Iamoni M., Düring B. (ed.), Understanding Hegemonic Practices Tonghini C. 2018, The Italian-Kurdish Excavations of the Early Assyrian Empire. Essays dedicated to at Gir-e Gomel in the Kurdistan Region of Iraq. Frans Wiggermans (Consolidating Empires Project I), Preliminary Report on the 2017 and 2018 field sea- Leiden, pp. 9-32. sons, «Mesopotamia» 53, pp. 67-162. Kopanias K., Macginnis J., Ur J. 2015, Archaeological Mühl S. 2013, Siedlungsgeschichte im Mittleren Osttigris Projects in the Kurdistan Region in Iraq https://dash. gebiet. Von Neolithikum bis in die Neuassyrische Zeit, harvard.edu/handle/1/14022526. Wiesbaden. Llop J. 2011, The Creation of the Middle Assyrian Mühl S. 2015, Middle Assyrian Territorial Practices in the Provinces, «Journal of Ancient Oriental Studies» Region of Ashur, in: Düring B. (ed.), Understanding 131/4, pp. 591-603. Hegemonic Practices of the Early Assyrian Empire. McMahon A. 2009, Chagar Bazar Burial Practices, Essays dedicated to Frans Wiggermans (Consolidating in: McMahon A., Colantoni C., Frane J., Empires Project I), Leiden, pp. 45-58. Soltysiak, A. (eds.), Once There Was a Place: Nováček K., Chabr T., Filipský D., Janíček L., Settlement Archaeology at Chagar Bazar, 1999— Pavelka K., Šída P., Trefný M., Vařeka P. 2008, 2002, London, pp. 109-128. Research of the Arbil Citadel, Iraqi Kurdistan, First Morandi Bonacossi D. 2013, Il Paesaggio Archeologico Season, «Památky archeologické» 99, pp. 259-302. Nel Centro Dell’impero Assiro. Insediamento E Nicolle C. 2007, IX. La Jézireh au Bronze moyen Uso Del Territorio Nella ‘Terra Di Ninive’, «Atti et la céramique du Khabur, in: al-Maqdissi M., dell’Istituto Veneto di Scienze, Lettere ed Arti», Matoïan V., Nicolle C. (eds.), Céramique de pp. 181-223. l’Âge du Bronze en Syrie II. L’Euphrate et la région de Morandi Bonacossi 2017a, 18. Italian Research Jézireh, Beyrouth, pp. 179-231. in the Nineveh Region: Archaeological Investigation Pfälzner P. 1995, Mittanische und Mittelassyrische and Cultural Heritage Protection and Management, Keramik. Eine chronologische, funktionale und pro- in: Morandi Bonacossi D., Petit L. P. (eds.), duktionsokonomische Analyse, Berlin. Nineveh the Great City. Symbol of Beauty and Power Pfälzner P. 2007, X. The Late Bronze Age Ceramic (Leiden, Museum of Antiquities, Oct. 2017-March Traditions of the Syrian Jazirah, in: al-Maqdissi 2018), pp. 98-106. M., Matoïan V., Nicolle C. (eds.), Céramique de Morandi Bonacossi D. 2017b, 19. From Prehistory to l’Âge du Bronze en Syrie II. L’Euphrate et la région de the Arrival of the Neo-Assyrian Kings.”, in: Morandi Jézireh, Beyrouth, pp. 231-291. Bonacossi D., Petit L. P. (eds.), Nineveh the Great Postgate C., Oates D., Oates J. 1997, The City. Symbol of Beauty and Power (Leiden, Museum Excavations at Tell al-Rimah: the Pottery, London. of Antiquities, Oct. 2017-March 2018), pp. 107-08.

West & East 108 Monografie, 4 Settling the land: settlements pattern and ceramics in the Land behind Nineveh from the Middle Bronze Age…

Pulhan G. 2000, On the Eve of the Dark Age: Qami- Schmidt C. 2013, Die Keramik der Früh-Ğazīra V- Bis Lim’s Palace at Tell Leilan, Ann Arbor. Alt-Ğazīra Ii-Zeit, Wiesbaden. Radner K. 2006-08, S. V. Provinz. C. Assyrien, in: RlA, Schwartz G.M., Brinker C.D., Creekmore 11, pp. 42-68. A.T. III, Feldman M.H., Smith A., Weber J.A., Radner K. 2011, The Assur-Nineveh-Arbela Triangle. Excavations at Kurd Qaburstan, a Second Millennium Central Assyria in the Neo-Assyrian Period, in: Mühl B.C. Urban Site on the Erbil Plain, «Iraq» 79, S., Miglus P. (eds.), Between the Cultures: the pp. 213-255. Central Tigris Region from the 3rd to the 1st Millennium Simi F. in print, The Tell Gomel Archaeological Survey. BC, Heidelberg, pp. 321-464. Preliminary report of 2015-2016 campaigns, in: Reade J.E. 2005, The Ishtar Temple at Nineveh, «Iraq» Coppini C., Simi F. (eds.), Broadening Horizons 5 67, pp. 347-90. Proceedings, Vol. 3 (5-8 June 2017, Udine). Reade J.E., Anderson J.R. 2013, Gunduk, Khanes, Ur J. A. 2010, Tell Hamoukar. 1, Urbanism and Gaugamela, Gali Zardak – Notes on Navkur Cultural Landscapes in Northeastern Syria: The Tell and Nearby Rock-Cut Sculptures in Kurdistan, Hamoukar Survey, 1999 – 2001, Chicago. «Zeitschrift für Assyriologie» 103/1, pp. 69-123. Westenholz J. G. 2004, The Old Akkadian Presence in Reiche A. 2014, 14. Late Bronze Age Pottery from Nemrik Nineveh: Fact of Fiction, «Iraq» 66, pp. 7-18. (Northern Iraq), in: Luciani M., Hausleiter A. Wilkinson T. J., Tucker D. J. 1995, Settlement (eds.), Recent Trends in the Study of Late Bronze Age Development in the North Jazira, Iraq: A Study of the Ceramics in Syro-Mesopotamia and Neighbouring Archaeological Landscape, Warminster. Regions (Proceedings of the International Workshop Ziegler N. 2004, The Conquest of the Holy City of in Berlin, 2-5 November 2006), Rahden/Westf., Nineveh and the Kingdom of Nurrugûm by Samsî- pp. 289-332. Addu, «Iraq» 66, pp. 19-26.

West & East 109 Monografie, 4

‘Living Burials’: reopening actions in kurgans and tombs between Central Europe and the Eurasian Steppe

LORENZO CRESCIOLI

CAMNES - Center for Ancient Mediterranean and Near Eastern Studies

Abstract

In recent years increasing attention has been devoted to the re-opening and disturbance actions of several typologies of tombs, especially in monumental burial mounds, amply attested from Eurasia to Europe. It emerged that these funerary monuments were very dynamic areas and represented important places for the societies that built them. Traces of secondary rituals, intentional re-opening actions and other types of activities, and interferences have been often recorded in or around kurgan structures, but also in other typologies of graves. Their interpretations are still rather unclear, but it appears, however, that some of these rituals recur over a very wide geographical and chronological horizon (from the Bronze Age to the Medieval period), within a cross-cultural dimension. The paper presents some case studies, mainly related to the Eurasian Steppes (Scythian culture in particular) showing the typological multiplicity of the phenomenon and their possible interpretations, and stresses the need of a strict methodological approach to the excavation and analysis of these contexts, essential to correctly understand the evidence for re-opening actions and not mistake these with ‘simple’ robbing actions.

Keywords

Disturbed graves, post-funerary rituals, kurgan,Europe, Near-East, Eurasian Steppe

West & East 111 Monografie, 4 Lorenzo Crescioli

1. Introduction ideological approaches have often influenced different interpretations, as in the Soviet period when a The funerary rituals are, for archaeologists, one of strong pragmatism linked to Marxist ideology fa- the most interesting and studied topics in the re- vored for robberies, and only subsequently there construction of the past societies, most of all for the was an excess in the opposite direction.3 related social, symbolic and religious aspects. The In fact, in recent years, much more attention has burials do not represent static contexts, limited to been paid to this issue4, on one side demonstrating the burial moment, but are related to society and how the funerary tradition of many cultures could created by very dynamic social interactions that are include various types of ritual activities carried out conducted by the living community.1 Despite this, after the proper burial action, including different what happens once the tomb is closed remains a kinds of ritual voluntary reopening of tombs and much less studied subject. The research of this phe- graves. On the other side this highlights the complex nomenon in the Eurasian and Central Asian con- cultural and social dynamics related to post-mortem, texts is affected by multiple problems that have of- peri-mortem and successive manipulation of buri- ten a strong influence on the final interpretation. als, that sometimes are still part of the funeral pro- The tombs, and most of all the highly visible burial cess including secondary depositions. This research mounds of the steppes, are generally thought to be field still needs to progress further in the countries the target of robbers, which aim to rapidly obtain of former Soviet Union, mainly on theoretical ap- precious objects. This phenomenon usually started proaches and methodological perspectives, and it is immediately after the burial closure and the main growing the awareness of the need to integrate fig- consequence for scholars is the loss of rich objects ures of anthropologists and other various types of belonging to the funerary set. While most of the scientific analysis to get a comprehensive approach burials show strongly disturbed or damaged con- to the problem.5 texts, we will try to demonstrate that these can have The nature of these reopening activities, manip- several different interpretations. The second prob- ulations of human remains and associated objects lem is methodological and it is related to archae- is analyzed in this paper, going through few study- ological research dating to nineteenth and early cases in different cultural regions, showing the dif- twentieth century, often methodologically inappro- ferent level of research development and the po- priate or inaccurate, and ideologically influenced. tential of this area of study. The research focuses on The third is a terminological confusion in the defi- burial mounds mainly of Bronze and Iron Age, i.e. nition of these actions that inevitably led to difficul- Scythian period, in particular on a specific context ties on their interpretation and in the elaboration personally excavated in the Kazakhstan territories, of a general framework.2 Before any interpretation trying to define the nature of the recorded burial the term ‘disturbed burial’ or ‘manipulated burial’ reopening. The paper also briefly reports examples should be used because it has a neutral connotation. Since burial looting is a common phenome- 3 Yatsenko, Kilunovskaya 2016, p. 10. non in different ancient cultures and civilizations, 4 A great attention for this topic is generally growing, and for the reasons showed above, it has been used mainly in two different cultural research areas: the European and Scandinavian region where a network researching the phe- as the main and often the only explanation for the nomenon of reopening burials (http://reopenedgraves.eu/) disturbed layers inside burial mounds. Moreover, was founded, organizing conferences, symposia and publishing many studies. Their goal regards mainly the medieval central Europe and northern country as England and Viking Scandina- 1 See also the theory of Robert Hertz (1907) about ‘Col- via. The other cultural area is Russia where an important sym- lective representation of death’. posium (Chelyabinsk University, January 2015) about burial 2 For example the use of terms as ‘ritual robbing’ or ‘ritu- looting and ritual reopening in the Eurasian territories in the al looting’ in many cases is inappropriate because implies a vio- period between Bronze Age and Middle Ages was organized in lence that did not actually happen. Furthermore, this term was 2015, whose proceedings where published in 2016 and form a also used to indicate the real robbery and this could create inter- fundamental base for these studies in the region. pretative confusion. 5 Yatsenko, Kilunovskaya 2016, p. 8.

West & East 112 Monografie, 4 ‘Living Burials’: reopening actions in kurgans and tombs between Central Europe and the Eurasian Steppe

and study cases from different cultural areas, such as actions. Such a clear distinction is not always visible Near Eastern world, Central and Northern Europe, in the excavated contexts. where this field of study is more developed and more The identification of the moment in which these attention has been devoted to this issue, attempting actions took place is much debated and consequent- to highlight the need of a strict methodological ap- ly there are multiple interpretations. On the basis of proach to the excavation and analysis of this kind of human remains it is sometimes difficult to establish contexts in the Eurasian Steppes to fully understand exactly the time in which manipulation occurred. the complex social, ideological and funerary dynam- In most cases however, on the base of articulated ics characterizing these funerary contexts. bones, scholars suggested that the ‘manipulation’ was carried out chronologically rather close, 1-2 generations maximum, by people of the same wide 2. Burial interferences cultural community, who in some cases knew ex- and manipulations actly the structure and the positioning of the deceased inside the funerary chamber, so it is possible that this manipulation was still part of the funer- 2.1 The Eurasian Steppes and the study case ary or post-funerary rituals. This is the case of the of Kaspan necropolis, Southeastern Kazakhstan Sintashta culture where large kurgans have several different burials under one mound and the funer- The vast ‘world of the steppes’ is characterized by a ary ritual was constituted of two different phases: large presence of burial mounds, the so-called kur- a temporary burial related to the decomposition gan, dating from the fourth millennium BC to the phase, (a sort of ‘house of dead’), and later, a phase Middle Age. This typology is also attested in Central with the burial reopening and the transfer of the and northern Europe. It is constituted of a mound body to another place, along with other secondary that can vary in shape6, dimensions7 and architec- rituals through which the deceased got the new role 8 tural structure. They are related to the social com- of ‘ancestor’. plexity and often are characterized as family tombs Scholars from soviet and post-soviet cultures with several chambers, shafts and niches. Their usually labeled the interferences of the Bronze Age structure shows a great variety of solutions and very contexts as ritual robbing or in other cases as ‘neces- 9 often are strongly planned and elaborated structures sary robbing’ , assuming that in case of need, weap- related to the social, spiritual and ideological aspects ons could be recovered from the burials, but the ar- of the society that constructed them. chaeological evidences show a much larger number The majority of these contexts dating to the of possibilities. Bronze Age and Iron Age present traces of intru- Also, in the Tuva region (Southern Siberia) sions, in many cases almost contemporary to burial. there are evidences of body’s manipulation, attrib- Russian scholars have identified five main catego- utable to different motivations, not always easily in- ries of disturbed graves: burials without damage to terpretable, but without doubt ritual in nature. In the grave but with manipulation of the deceased or the Todz necropolis a female skeleton in situ shows animal remains; post-funerary damage to the burial the upper portion of the body completely removed. structures, without damage to the funerary set but It has been proposed that it could be part of rituals with manipulation of the bone remains; cenotaphs, to get in touch with the deceased of the community i.e. tombs without burial; classic robbery for remov- or, as in the case of Sintasha, to favor the passage in ing precious objects; robbery accompanied by ritual the afterlife (rites of passage and secondary burials). In the necropolis of Kun Sair, a mound contained 6 Mounds are usually rounded, but sometimes are also attested squared or oval shapes. 8 7 Mound diameters can range between a few meters up Koryakova, Epimakhov, 2007, p. 78; Zdanovich, to hundreds of meters, with heights also ranging from almost Zdanovich 2002. imperceptible up to more than 20 meters in height. 9 Kuzmina 1965.

West & East 113 Monografie, 4 Lorenzo Crescioli

Figure 1 Chinge Cemetery, Object 16 Sections and plan of Grave 3 showing disturbance and accurate covering. Aldy-Bel Culture (after Semenov, Kilunovskaya 2016, fig. 10, p. 29)

an individual (40-45 years old) that, because of bones were moved in another final place. This is -at some pathologies was probably lame, and to which tested for example in the necropolis of Albe-Belian the feet and other bones had been removed, perhaps culture where all the burials (including children’s to favor his passage to the netherworld, or because burials) were manipulated with the removal of some his ‘special status’, to prevent him from returning to bones, and then were carefully closed (fig. 1). the world of the living to disturb. The manipulation Equally, in the valley of the Eerbek river, 90% of of the bones is also attested in the necropolis of Ar- the burials are disturbed, except for those of the chil- gens, and in that of Dai-Bag, where in a mound a dren. In the necropolis of Beloe Ozero 3 four mon- skull had been voluntarily removed and placed be- umental mounds were manipulated before the con- tween the legs of the deceased.10 struction of the mound and the pits were sealed with The manipulation or destruction of the burial of- a layer of clay brought from the nearby lake.11 Three ten took place before the construction of the mound tumuli of monumental dimensions in the necropolis as part of a ritual, perhaps to help the passage of the of Kosh-Pei, not far from the famous necropolis of deceased to the netherworld. It is not clear if some Arzhan, present the same filling. They are character-

10 Semenov, Kilunovskaya 2016, pp. 18-25. 11 Semenov et Al. 2014.

West & East 114 Monografie, 4 ‘Living Burials’: reopening actions in kurgans and tombs between Central Europe and the Eurasian Steppe

Figure 2 Kosh Pei 1 cemetery. Burial mound 2, Uyuk culture. Plan of different levels of the wooden structure (after Semenov, Kilunovskaya 2016, fig. 12, p. 31)

ized by a complex structure formed by a large pit cov- es show a continuity of use and numerous paral- ered by a wooden structure and bark forming a sort of lel activities as they become the center of the spir- yurt/tent, and on the bottom there was the wooden itual and social life for the community, as a place of funeral chamber.12 The old dromos was used to access contact between the world of the living, the ances- the pit. Since there are no traces of re-opening, there- tors and the deities. A typical aspect of this region fore the various activities must be carried out before is the reuse of the central pits of the tumuli. First the completion of the mound (fig. 2).13 the primary burial remained closed for a certain pe- In the southern Transural region in the western riod of time with rituals carried out in adjacent ar- area of the steppes, numerous Bronze Age necropo- eas, then some rites of reopening and manipulation lis present different kind of manipulation or ‘viola- took place and new burials were added with many tion’. As in the Tuva region, the burials of children bones found in the filling of the graves. In Stepnoye are usually intact, while the large burials of adults VII cemetery, burial complex n. 4 is characterized are manipulated in 90% of cases. These complex- by two central pits: one completely manipulated with evident traces of intrusion, containing three 12 Semenov 1994; 2010. individuals (two of which embraced) and a very rich 13 Semenov, Kilunovskaya 2016, pp. 26-33.

West & East 115 Monografie, 4 Lorenzo Crescioli

Figure 3 Stepnoye VII cemetery, Pit 17, triple burial, plan showing the human remains and the several objects of the funerary set including fragmentary ceramic vessels (after Kupriyanova 2016, fig. 7, p. 91)

funerary set including fragments of a ceramic vessels part of the funerary use of this contexts character- whose other sherds are found inside the filling of the ized as a sort of ‘collective’ or family burials where adjacent pit. Therefore, the intrusion was contem- different bodies were added over time (fig. 3). porary and not aimed at the recovery of the precious It has also been emphasized how the manipula- objects that have been left in situ. Similarly in Kur- tions in this region often appear ‘excessive’: the ob- gan n. 1 the ceramic vessels were broken on purpose, jects are moved, the skeletons destroyed, and nu- and fragments of the same pots were found in the merous elements were left on the ground around the three central pits, while the metal objects were left funerary pit. These characteristics does not seem to in situ, showing a non-utilitarian character of this be linked to a utilitarian looting and it has been as- disturbance.14 Probably this ritual manipulation is sociated to a sort of fight against the dead, to prevent it from returning to disturb the world of the living. 14 Kupriyanova 2016, pp. 84-90. In addition, these disturbances show common fea-

West & East 116 Monografie, 4 ‘Living Burials’: reopening actions in kurgans and tombs between Central Europe and the Eurasian Steppe

Figure 4 Pazyryk kurgan n. 5, section of the mound and of the funerary chambers with traces of disturbance (after Rudenko 1970, fig. 6, p. 20)

tures in different cemeteries perhaps attributable to cially exposed to these types of activities, probably some kind of ‘manipulation rules’, for example the because of its easy identification in the landscape, its manipulation of only the upper body bones, while function as family tomb, and for the architectural the pelvis and legs were usually left untouched, so features of its structure which simplify the access to the deceased and his spirit could not return. The dis- internal sectors (dromos, vertical shafts, large cham- position of some items after the reopening seems to bers and wooden structures). The modern robbing emphasize it. The disturbance to the upper part of activities (eighteenth-twentieth centuries CE), hap- the bodies and other manipulations perhaps con- pened mainly under the push of Tsar Peter I the stituted preliminary operations for the subsequent Great, and actually led to the formation of the ‘Si- burials, as a sort of purification process, an interme- berian Collection’ today housed at the Hermitage diate phase before the new burial.15 Museum.17 These violent activities characterize From the Iron Age, the cultural Scythian hori- mainly southern Russia and Siberia and are easily zon, which is characterized by an enormous pres- recognizable, as compared to ancient looting which ence of mounds, spreads in the Eurasian continent. is much more difficult to locate. In some cases, a -ma Violation, looting and interferences of the Scythian nipulation almost contemporary to the burial took burial mounds are very attested and varies region- place, because the bodies were still articulated18 and ally, but to the north of the Black Sea for example it the scholars hypothesize (as for the bronze Age ex- stands between 70 and 100%.16 The burial mound amples) that the looters knew the exact construc- typology (most of all the royal kurgan) seems espe- tion of the mound (for example, in the Tuekta ne-

15 Kupriyanova 2016, pp. 90-95. 17 Rudenko 1962. 16 Mozolevskiy, Polin 2005. 18 Ochir-Goryaeva 2016, pp. 112-114.

West & East 117 Monografie, 4 Lorenzo Crescioli

Figure 5 Filippovka, Kurgan n. 13 Plan showing the structure, with dromos and two tunnels (after Yablonskiy, Meshcheryakov 2006, fig. 1)

cropolis on the Altai Mountains), and possibly even acter, but also simply to bring back the royal cou- participated in the construction of the mound and ple together, when the still alive member died and in the funeral ceremony. The looting dating to the joined his spouse.22 two different periods are distinguished by different In various regions such as the Black Sea, South- characteristics: a large vertical shaft for recent viola- ern Urals and Semirech'è, the access to the kurgan tions, and an oblique tunnel, usually from the north takes place through a tunnel from the edge of the base of the mound dug by contemporary people mound to the center of it, whose ritual character is who knew well the interior structure of the mound.19 demonstrated by the precise planning and correla- Since the 90s some scholars20 tried to study the tion with the elements of the entire structure. In the burial reopening based on ethnographic and his- necropolis of Filippovka in the Kurgan n. 13 two dif- torical comparisons or on the basis of the presence ferent tunnels were used to reenter the mound and of some utilitarian objects (burner for hemp, harp) were later filled with pure sand (fig. 5).23 perhaps used by individuals who entered the interi- Recently, it has been proposed that the Scythi- or of the tombs such as in Pazyryk kurgans (fig. 4).21 an burial mounds are characterized by a precise in- It has also been hypothesized that the dromos or ternal planigraphy24, where the dromos is generally other passages may have been used with ritual char- 22 Shulgin 2003. 19 Mozolevskiy, Polin 2005. 23 Ochir-Goryaeva 2016, pp. 115-116; Yablonsky, 20 Savinov 1996. Meshcheryakov 2006. 21 Ochir-Goryaeva 2016, p. 114; Rudenko 1970. 24 Ochir-Gorayeva 2011.

West & East 118 Monografie, 4 ‘Living Burials’: reopening actions in kurgans and tombs between Central Europe and the Eurasian Steppe

found along the southern side, i.e. a neutral zone on ing them. On the basis of this it has been hypoth- the border between the two parts of the mound: the esized that the objective of this violation was the western related to living people and the eastern to desire to recover only some objects with a symbolic the world of dead. Some scholars believe that they character, perhaps important for obtaining the sta- were built by thieves, but their position seems re- tus and legitimization of the government for the di- lated to ritual aspects, in addition to alleged loot- rect descendants. This would explain how only the ers would be agreed to dig the tunnel on the north main central burial suffered the disturbance.27 It has side, because it would be shorter since usually the been hypothesized that non-disturbed main central Scythian mounds had a steeper slope on the north- tombs can be linked to a symbolic passage of power ern side. However, it cannot be ruled out that these that has already taken place in life or to other cir- southern tunnels or dromos were subsequently used cumstances such as the lack of heirs or changes in by thieves as well. In the Besshatyr necropolis the the dynamics of control.28 wooden dromos built above ground level is gener- In general, the rite of visiting the tombs evolved ally found on the eastern side, but in some kurgan in some Scythian cultures, as in the Altai Moun- underground tunnels have also been found, which tains, from strong destruction of almost the entire from the southern side of the mound reached the room in the early Scythian periods, to less invasive, center of it, below the burial chamber. The accura- simpler and more symbolic rituals in later periods.29 cy of the realization and their dimensions suggest a These activities must be linked to ritual aspects, pos- ritual character, perhaps to communicate with the sibly part of the funerary process, ancestor worship, dead, as close as possible to the funerary chamber.25 or rituals linked to rite of passage. In the Ural Mountains Scythian cemeteries, typ- The burial ground of Kaspan, in south-eastern ical radial structure composed of wooden trunks Kazakhstan was recently excavated in a project of were uncovered above the funeral chamber of the collaboration between Ca’ Foscari University and mounds, which are only rarely preserved. They could Centro Studi e Ricerche Ligabue and contributed be accessed through specific vertical shaft and, given significantly to this discussion.30 The stratigraphic their accurate construction, were probably used for context of two excavated kurgans, dating to the sev- visits the dead and ceremonies of various kinds, as enth century BC and belonging to Scythian culture, attested in the mound of Bajkara.26 show a subsequent reopening of the burial, follow- In the large royal kurgan of Scythia proper, the ing the primary deposition or perhaps a period of entrance was through a central shaft that reached only partial and temporary closing of the burial pit, the burial chamber and not through inclined tun- followed by a complete removal of funerary goods nels. The shaft generally used the most recent en- and probably also of some bones. The presence of trance, since these great kurgans had several of them scattered bones in the pit filling, corresponding to because they were family tombs with multiple suc- some kind of activity and manipulation, demon- cessive burials. The manipulation and the removal strates the rituality of this action (fig. 7). of only some objects with symbolic meanings always Most of all the subsequent closure of the funer- concerned the largest and richest central chamber ary pit made with highly purified soil (coming from (fig. 6). In addition, the burials of the horses, which the superficial layers of the surrounding areas), were always on the western side of the mound and mixed with water and alternating with several lay- were richly adorned and also much more accessi- ers of stones, show the rituality of this activity. At ble, were never plundered. Thieves probably did not the end a small but deep pit reaching the bottom of know where these tombs were, but the member of the funerary chamber enclosing a dog burial was dug the communities knew about their existence, but 27 As attested in the Kurgan of Chertomlyk, Alekseev, probably they were not interested in manipulat- Murzin, Rolle 1991. 28 Ochir-Goryaeva 2016, p. 120. 25 Ochir-Gorayeva 2016, pp. 116-117. 29 Ochir-Goryaeva 2016, p. 126. 26 Parzinger et Al. 2003. 30 Beysenov et Al. 2015.

West & East 119 Monografie, 4 Lorenzo Crescioli

Figure 6 Royal burial mound of Chertomlyk, plan showing the catacomb with the deep shaft and the horse burials (after Ochir- Goryaeva 2016, fig. 3, p. 121)

West & East 120 Monografie, 4 ‘Living Burials’: reopening actions in kurgans and tombs between Central Europe and the Eurasian Steppe

Figure 7 Kaspan burial ground, filling of Kurgan n. 4 with stones and human bones mixed (courtesy of Centro studi e ricerche Ligabue)

and it seems somehow to conclude the whole cer- dead in the otherworld.31 Other less probable inter- emony (fig. 8), even if it is not clear how much time pretations are: ritual action of damnatio memoriae later it was realized. by a rival group or family which probably would ex- The interpretation of this re-opening operation pect a less attention to the enclosure process; or the and interference is rather difficult, but some hypoth- recovery of the ceremonial material used during the esis can still be made: the most plausible interpreta- funerary ritual, perhaps with a utilitarian purpose tion, due to the ritual character and the great care which does not fit with the almost complete remov- devoted to the enclosure process after the manipu- al of bones and other objects from the tomb. lation, is the presence of a secondary ritual, helping The kurgans belonging to the later Xiognu com- the deceased to conclude his journey to the after- munity of Iron Age of Mongolia have also been loot- life, finally getting the role of ancestor. The role of the dog can be read as guide and protector for the 31 Papanova 2004; Sinica 2006.

West & East 121 Monografie, 4 Lorenzo Crescioli

Figure 8 Kaspan burial ground, Kurgan n. 1, dog burial (courtesy of Centro Studi e Ricerche Ligabue)

Figure 9 Gol-Mod burial ground, Grave n. 1, schematic plan and section showing the disturbance traces (after Bendezu-Sarmiento, Grizeaud 2011, fig. 7, p. 44)

West & East 122 Monografie, 4 ‘Living Burials’: reopening actions in kurgans and tombs between Central Europe and the Eurasian Steppe

ed by ancient robbers, as well as by rival groups as part Figure 10 of ideologically based actions. There are also written Mentesh Tepe Kurgan, plan of the funerary chamber sources (Chinese chronicles) of first century BC re- (after Poulmarc’h, Pecqueur, Jalilov 2014, fig. 8, p. 241) porting this phenomenon.32 Some of these burials (tomb 64 and 94 of the Egiin-Gol necropolis in Mongolia, or the princely tomb of Gol Mod) (fig. 9) show plundered skeletons which are still in connection showing how this disturbance occurred soon after burial.33 In the Eurasian steppes it was a well-established practice for the raiders to destroy, raid and in some cases make new burials, in the old graves to prove their rights over new conquered lands and this practice could well be also applied to earlier periods.

2.2 The Caucasus region

In the Caucasus region there are no specific studies on the plundering and reopening actions of the burial mounds, but the latest research is showing how funerary contexts were very dynamic places within the community.34 In this region as well, since the Bronze Age, the kurgan typology is spreading, including examples of very large and monumental sizes, which can contain many individuals. Some of these ‘collective tombs’ are found in Azerbaijan, showing skeletons highly disturbed and other ones perfectly preserved, probably proving a periodic reopening of the structure to add members belonging to the same social group (about 39 individuals in Mentesh and 80 in Uzun Rama kurgans)35, where the bodies were shifted aside to make room for the new individuals (fig. 10). 36 Finally (probably after three-four generations), the closing ritual meant that the whole room was burned down, damaging or destroying bones and grave goods, perhaps not only as a religious act but also as an action against possible subsequent looting. In Mentesh kurgan no skeletons are complete and some bones (especially skulls) were taken out before closing the chamber, perhaps as memory or as ritualistic object.

32 Daffina 1999. 33 Bendezu-Sarmiento, Grizeaud, p. 43. 34 Palumbi 2011. 35 Poulmarc’h, Pecqueur, Jalilov 2014. 36 Akhundov 1999, 2001.

West & East 123 Monografie, 4 Lorenzo Crescioli

2.3 The Near Eastern world From some point of view also in some examples of so-called ‘collective burials’ or ‘family burials’ we Near Eastern cultures are an extraordinary case can find secondary burials. This tomb typology is thanks to the great number of investigated contexts characterized by structures (usually rock cut shaft and the presence of a huge corpus of written texts. graves) that allowed to add other deceased and to Here I will only briefly present some examples from re-enter several times for other purposes as ritual Syrian territories showing the complexity of the practices, which of course took also place in single phenomenon and the development of the studies. burials, but mainly from the outside.47 When new In this region different types of burial are recorded bodies were added to the tomb, if the space was lim- and the burial mound is a much rarer typology, ap- ited, the bones could be removed and moved to oth- parently limited to only some areas. In this cultural er places, keeping in the original tomb only some area of research generally all kinds of disturbed buri- of them, especially the skull or some long bones, as als or disarticulated bones were defined as second- shown in the Tell Barri48 and Mohammed Diyab ary burials, but probably the phenomenon is much hypogea.49 The entrance to the tombs of important more problematic and multifaceted.37 ancestors belonging to the elite could be linked to Many traces of manipulated and disturbed buri- the will to legitimate the role of descendants and to als are found in this northern Mesopotamian re- maintain the political order. Ebla texts refer to lat- gion38. The phenomenon of secondary burials is er rites (entry, gift, statues placed inside) but they widely attested for example at Tell Zeidan39, Tell are difficult to identify and mainly to distinguish Jerablus Tahtani40, Tell Oylum Höyük41, Umm between funerary or post funerary activities and el-Marra42, Tell Banat43 even if «careful recording their eventual chronological interval. In some cases and retrieval are required to determine the exis- it is possible, as in tomb T.302 from Tell Jerablus tence of secondary treatment».44 Some bones were Tahtani, that different kind of activities (inter- removed and transported to new locations, as clear- ments and placement of many objects) were sepa- ly shown in Tell Banat’s shaft and chamber tombs, rated by one or two centuries.50 where the bones of multiple bodies were moved Ancestor worship, in elite contexts, is anoth- from a primary deposition and grouped in large er explanation for manipulated contexts. At Umm heaps, mixed with animal bones and objects (main- el-Marra a mortuary complex in the center of the site ly ceramic vessels). It is not clear if they were simply was uncovered, where some of the numerous tombs moved among different places inside the chamber are related to other structures which could be used or if they were brought from outside (perhaps infor worship of ancestors, for preparation of bodies to side some ceramic vessels).45 Secondary rituals seem be buried, or maybe as part of a process transforming to be also related to the famous “White monument” the deceased into ancestors. These structures could of Tell Banat (fig. 11), a huge mound where human also to be related to the original funerary rituals, or and animal remains were definitely moved from the to both functions diachronically.51 In some tombs, tombs built in the surrounding area.46 objects were placed at higher levels in later periods and in three secondary burials, two of which appar- 37 Pfälzner 2016. ently contained inside a wooden container, so maybe 38 Felli 2015, 2016. transported from other (more or less) distant places. 39 Stein 2010, p. 107. Explanations can be various: need to wait for ritual- 40 Peltenburg et Al. 2015, pp. 44-45. ly proper time, desire to collect enough resources for 41 Engin and Helwing 2012, p. 97. 42 Schwartz 2012, p. 70. 47 43 Porter 2016. Peltenburg 2016, pp. 136-137. 48 44 Peltenburg 2016, p. 135. Valentini 2003, p. 281. 49 45 Porter 2016. Bachelot 1992. 50 46 Peltenburg 2006; McClellan 1998; McClel- Peltenburg 2016, p. 138. lan, Porter 1999. 51 Schwartz 2016, pp. 194-196.

West & East 124 Monografie, 4 ‘Living Burials’: reopening actions in kurgans and tombs between Central Europe and the Eurasian Steppe

Figure 11 View of White Monument of Tell Banat (after Peltenburg 2006, fig. 4, p. 224)

an elaborate ceremony, a sudden death far from its activities took place as shown from a large amount town, but also as part of precise sequence of rituals of bones inside ceramic vessels. During the final re- belonging to rite of passages.52 location the bones were moved to chamber two, a There are special contexts, like the royal Hypo- sort of ossuary.53 In the ante-chamber, the presence geum of Qatna, where the different activities took of ancestors statues, with several ceramic vessels and place together. Among about 20 bodies only one food offerings could be related to ancestor worship, was found in primary deposition, in anatomical con- recalling the ritual known in late (Amorite) writ- nection lying on a stone table. All the other bodies ten sources as ‘Kispum’, which seems taking place were found with some bones partially removed and in specific installations and places outsides palaces, collected after only partial decomposition on small but not in the tombs. 54 Its function was also to sus- heaps next to related bodies. In this intermediate tain the dead to prevent an interference with living phase different kinds of manipulation took place. world.55 When chamber two was full the bones were This could not be considered a secondary burial be- moved to a complete different tomb (number 7), in- cause the bones were almost in the same place of side different wooden boxes.56 primary deposition and only later will be moved to another location. This is a sort of pre-stage to final 53 Pfälzner 2016. 54 deposition, in which also procession and feasting Pfälzner 2006. 55 Peltenburg 2016, p. 143. 52 Schwartz 2016, pp. 197-199. 56 Pfälzner 2006, pp. 257-258.

West & East 125 Monografie, 4 Lorenzo Crescioli

Some written sources may show alternative in- Figure 12 terpretations for the reopening of a specific burial Austria, Cemetery of Brunn am Gebirge. type, (hypogeum) a rather systematic phenomenon Plan showing graves with re-opening traces (after Aspöck 2011, fig. 1, p. 301) during MBA, but also during LBA and generally attributed to simple looters, or to the re-use of burial in later periods (for example from Middle to Neo-Assyrian period). Indeed, some sources point out that when a king subjected a population, to break out its roots, apart from deporting the defeated, the tombs of the ancestors were violated, because through them the society ensured a bond and continuity over a given space. In this case violation is linked to political and ideological causes57; other sources point out that when a group or a family moved they carried the remains of their ancestors, because if the burial was destroyed they would not have a place of physical worship any more without benefiting of the care from the living people.58

2.4 Central Europe and Scandinavia

A completely different cultural, chronological and geographic framework is constituted by the funerary contexts of Merovingia (Central Europe) and Scandinavia, where, however, a major study on the phenomenon of the reopening of burials has recently been established.59 The reopening of burials characterizes the high- medieval period of a region ranging from Austria to England, especially in the seventh century when thousands of burials were violated in hundreds of necropolises (fig. 12). This action was already identified since the nineteenth century, but was studied more systematically only recently. These burials usually are not monumental, mainly single graves often in extra mural burial grounds. The number of disturbed burials varies greatly at regional level, it was a

57 Vernant 1982. 58 Bottéro 1980; Valentini 2003, p. 286. 59 The EAA Conference in Oslo 2011, was entitled ‘Past disturbances of graves: the reopening of graves for grave-robbery and other practices’; ‘Grave disturbance in Early medieval Europe International symposium 2017’, held at the department of Archaeology and Classical studies in Stockholm University.

West & East 126 Monografie, 4 ‘Living Burials’: reopening actions in kurgans and tombs between Central Europe and the Eurasian Steppe

widespread practice but causes and modalities must related to the consolidation of power and its struc- be examined locally.60 tures during this historical period. The second ex- In England (especially in the Kent region) in planation hypothesizes an action performed by the some of these cemeteries a clear selection in the re- same family or the descendants of the deceased to moval of bones and objects is attested. The skulls are retrieve heirlooms from the tombs.62 almost always damaged and the jaw is often missing The removal of bones shows other purposes or in secondary deposition, although this may de- (fig. 13). The skull is one of the most manipulated pend on the sarcophagus conservation status at the elements and it could be linked with two different time of reopening. Moreover, missing bones could interpretations, as a peri-mortem form of precau- indicate the areas from which the objects were re- tion against the possibility of revenge on the part moved and probably were not the primary target of of the deceased and his revenant, or as ‘memento’ the disturbance. In other cases, the removal of bones in the event that these groups moved to other re- could instead evidence a practice to prevent spir- gions. Some of these practices may even have been its from returning to the living world by being de- the cause of death of some deceased. stroyed of their ‘life functions’, through a re-killing In some cases, one or more methods are attested action. Sometimes bones belonging to other individ- simultaneously, to kill or to re-kill the newly dead uals appear together with the original deceased and body (fig. 14). These actions are related to what is were probably added later for uncertain reasons.61 usually defined ‘necrophobia’ (fear of the return of The precise selection of only a few types of ob- the dead), and to the increase in violence associat- jects is perhaps linked to the symbolic role they have ed with tightening of royal power.63 These practices assumed during the burial ritual, and to the different aimed at individuals who, due to events related to ways of holding or associating between individuals their life or death, were believed not to have the nec- or groups before burial. In particular, two categories essary serenity and the pre-conditions to complete of objects are missing: brooches and pins for female the journey. The manipulations are recorded both in burials and swords for men ones, which are highly the initial stages and in later phases, when the bodies symbolic and meaningful objects and were used in were already disarticulated, so the fear of a possible high medieval gift exchange and inheritance prac- return could last a long time. From the social per- tices. Other precious items (shields, knives, neck- spective it is considered an important action for the laces in semiprecious material such as amber) possi- community of the living rather than for the deceased bly used for gift exchange practices were instead left to whom it was applied, for example during crisis in the burial. In some cases the objects were taken episodes for the community. This practice moreover even in poor conservation conditions, so this ac- demonstrate the power of some individuals to con- tion need to be linked more to the deprivation of trol death, and it could be used as part of precise so- the deceased, than to the subsequent use of this ob- cial strategy to attack a deceased and most of all the jects by the living. Moreover, a precise selection of living people associated with him. Very similar prac- the graves (the larger and male burials) is recorded. tices are also well attested in the literature of ghost This disturbance could be explained with two main stories widespread in the Middle Age England.64 interpretations: an action performed by families or It is therefore assumed that the actions in ques- groups who wanted to harm rival communities, de- tion were carried out by the same individuals (or priving them of possession of symbolic objects (es- groups) who realized the burials and not by simple pecially weapons as knives, swords and belts) and thieves or rival communities. In this cultural con- attributes that carry with them the power, prestige text, the display of grave goods, as a competition ac- and value of the former possessor from which they were subtracted. This was probably part of conflicts 62 Klevnäs 2013, 2015, pp. 165-169; Aspöck 2011, p. 310. 60 Aspöck 2011; Klevnäs 2015, 2015a, pp. 177-178. 63 Reynold 2009. 61 Klevnäs 2015, 2015a. 64 Klevnäs 2015a, pp. 192-200.

West & East 127 Monografie, 4 Lorenzo Crescioli

Figure 13 England, Winnall II cemetery, decapitated burial 23 and burial 12 (after Aspöck 2011, fig. 12, p. 317)

tions between different groups, was a very impor- tional interpretation, intersections and overlaps of tant element within the funeral and it cannot be graves were perceived as completely accidental, but excluded that the community would later recapture currently the intentionality of the phenomenon is the materials used during the funerary ceremony.65 more emphasized. Probably it is related to differ- Similar activities, manipulations and interfer- ent kinds of interactions between dead and living ences in the burial ground are also widely record- people including the aim to transform existing so- ed and studied in the Scandinavian regions (Swe- cial identities, building new ones or as continuity den, Norway and Iceland).66 In the Scania region through kinship.69 (Southern Sweden) the practice of ‘overlying buri- As recorded for central Europe the burials were als’ is well attested: the burial re-opening is followed re-opened for different reasons: to recover objects, by the deposition of a new body, with the removal precious for economic or symbolic values; to per- of the older bones, which were put in the filling of form activities against deceased for the fear that the new graves.67 The overlying burials are also at- they could disturb the peace of the living (also re- tested in Denmark and in Sweden.68 In the tradi- ported in medieval Scandinavian literature); to so- cially attack the community or single individuals 65 Halsall 1995; Aspöck 2011. who have established contacts with them, especially 66 Klevnäs 2016. the chief or king, symbol of power. The control of 67 Skaarup 1989; Nagmer 1979. 68 Arcini, Jacobsen 2008; Nielsen 1991. 69 Thate 2007; Satalecki 2014.

West & East 128 Monografie, 4 ‘Living Burials’: reopening actions in kurgans and tombs between Central Europe and the Eurasian Steppe

Figure 14 England, Winnall II cemetery, grave 10 decapitate burial and missing legs (after Aspöck 2011, fig. 11, p. 316)

such important places for social identity and symbol fore being recovered and reused in important cer- of hierarchy, and power, is fundamental in the at- emonies or in the private sphere. In Iceland this ac- tempt to control society, from a political, but above tion probably needs to be related with the creation all, ideological perspective, as happened for exam- of identity of a ‘young group’ in the island (the Vi- ple, during the process of Christianization of Nor- kings), as the society had just formed and the com- way70 or for Norwegian royal mounds, manipulated petition for power with other rival groups could be probably with the purpose to attack those creating very strong. These symbols of power recovered from those social memory processes. the graves, showed the owner’s legitimacy, just as Iceland’s Viking burials show exactly the same the destruction of graves belonging to rival groups types of manipulation as in Scandinavia and central aimed to discredit the enemy’s importance (dam- Europe. The complete or partial removal of the fu- natio memoriae). The other explanations for these nerary set (especially weapons) could be carried out manipulations must be searched in the already ana- by thieves or for symbolic reasons, mainly related lyzed ‘fear of the return of the dead’, or in facilitate to the status that possession of such weapons con- the deceased passage into the Netherworld (rites of ferred. It is not to be excluded that weapons and passage), or in the use of some bones in rituals of other typologies of objects were thought to remain various kinds.71 in the burial only for a limited period of time, be-

70 Satalecki 2014. 71 Thorarinsdottir 2008; Eldjárn 2000.

West & East 129 Monografie, 4 Lorenzo Crescioli

3. Discussion and conclusions ate between a voluntary manipulation and causal movements due to the conservation of the con- Traditionally the history of burial ends with the text, details that are often ignored.74 deposition of the body and the funerary goods In general, the interest in these ‘bizarre’ burials and with the final closing of the tomb. Little at- defined in literature as ‘deviant burial’ or ‘divergent tention was paid to post-burial practices and such burial’, i.e. burials differing from the normal funeral re-openings were generally felt simply as a distur- rituals of a defined culture, period or region, arose bance of the dead, and not as something associ- in processual archeology and grew in the post-pro- ated with the ritual activities of the funerary ritual cessual theory. Today, a wider attention and knowl- or with the community that had buried their be- edge of this issue has shown that the so-called ‘devi- loved. The problem of antique and modern loot- ant burial’ are actually an integral part of the normal ing is real and causes serious damage. The ancient funerary practices of most communities. Indeed, populations themselves were aware of it so they there is not just one funerary model and different tried to use solutions against this problem.72 Re- concepts can coexist in the same cultural commu- cent studies show how, alongside the violent loot- nities, even without ethnic differences (for example ing, the tombs could be reopened, the funerary between elite and common people).75 set totally or partially removed and the bodies The study of this phenomenon is very important manipulated within ceremonies and activities of because it clearly helps in the understanding of cul- symbolic, spiritual or ritual nature, generally per- tural patterns, funerary rituals, social relationships formed by the same community, clan or family and dynamics, of the past civilizations. Of course, that buried the dead, during the funerary ritual or such different cultural contexts cannot be generi- at a later time. In the analysis of these problems, cally treated, and causes and interpretations must some fundamental aspects must be kept in mind be analyzed locally, and could be linked to histori- for a correct understanding, starting, for example, cal events and to cultural, social, and ideological as- from the whole ideological categories, the knowl- pects, even if Hertz has correctly highlighted the edge of the elaborate and varied funerary process, presence of some wide anthropological categories the concept of object possession and ownership, common to mankind.76 What needs to be highlight- on which the entire question is based.73 Did the ed is the importance of all these activities for the liv- objects belong to the deceased, to the family or to ing communities that performed them, more than the community? The answer to this question con- the people whom received it. notes a removal as violent or legitimate. Perhaps For example, in Scythian nomadic or semino- connected to this concept is also the precise selec- madic groups, the necropolis, and especially the tion of objects to be removed and their character- royal kurgans become the aggregation center of the istics of transferability that is recorded in all the community from religious, social and probably po- analyzed cultural contexts. The removal or ma- litical perspectives. Several different activities, in- nipulations of bones involve another problem, cluding memorial rituals, were carried out around i.e. the difficulty in establishing the exact moment and on top of the burial mounds as demonstrated in which this operation took place to distinguish from numerous peripheral structures. Some of the between peri-mortem, post-mortem or post-burial kurgan were also family tombs and therefore it treatment. The most important element in this should not be surprising if intense activities and rit- regard is the state of decomposition of the body, uals were carried out there, including reopening op- namely if human remains were still fleshed or al- eration in some specific occasions. In the case of Ear- ready skeletonized. It is also important to evalu- ly Middle Age Europe, some reopening actions may

72 In the kurgan for examples these expedients had been 74 adopted: funerary chambers not in axis with the mounds, very Klevnäs 2015a, p. 182. deep catacombs or pits, massive fillings of the funerary pits. 75 Aspöck 2008; Pfälzner 2016. 73 Klevnäs 2015b. 76 Hertz 1907.

West & East 130 Monografie, 4 ‘Living Burials’: reopening actions in kurgans and tombs between Central Europe and the Eurasian Steppe

perhaps also be linked to the spread of Christianity, These reopenings may be operations for the due to the will to move their loved ones and ances- recovery of funerary goods or bones from mem- tors to Christian cemeteries, so they can be buried bers of the same family or group. This may be in sacred places. In contrast, the process of Chris- part of popular or family-related beliefs, with the tianization in some cases imposed the violation of fear that the deceased might return to disturb the older contexts to affirm its superiority and its legiti- world of the living or claim their properties in life macy as the new belief system. The very strong bond (necrophobia). Some bones could have ritualistic between ancestors and living communities can also powers and were possibly used in benign or malev- be demonstrated by this desire to reopen the burials olent rituals in popular belief, or have a ‘memento’ that did not represent a ‘taboo’. However, it is very meaning for the family. The objects could have a difficult to identify the exact interpretation of each symbolic value and could be re-used in ritual cer- single case of reopening, especially on the basis of emonies, as talisman or heirlooms, or finally simply only archaeological data. A number of research ar- for utilitarian re-use. The weapons assumed special eas focuses on these issues only recently, so the gen- meaning as symbols of power so were one of the eral framework is still very generic, although some preferred manipulation targets. more detailed and comprehensive studies are start- Other reopening can be defined as intrusions, ing to appear. which are characterized by real violations for ideo- The main and most popular interpretations of logical reasons, for example attempts of damnatio these activities in different cultural areas can be memoriae from rival groups, clans or communities broadly summarized as follows. that can act at different levels, from the removal Besides simple plundering for personal and im- of single elements with highly symbolic value to mediate enrichment of the robber, these actions can the complete destruction of burials. The burial belong to extended funerary rituals, either for the ad- grounds assumed importance as places of power, dition of new burials in so called ‘familiar tombs’ (re- hierarchy and social identity, so their control al- location or removal of all or only few selected bones), lowed legitimacy to the ruler dynasty or to groups or for secondary rituals (secondary burials), integral aiming at it. For these reasons they were location part of the actual funerary ritual, with the movement of intense activities. of only some bones or of the whole body in a differ- A more in-depth analysis of these issues can re- ent and definitive place of burial (rite of passage). ally help to gain a better understanding of the ideo- Other manipulations could be post-funeral rites logical, social and cultural complexity of these com- linked to different activities such as the ancestors munities, their relationship with the world of the worship, mainly carried out in royal and elite tombs, dead and their conception of the afterlife, as well as and that could imply a re-entering inside the proper the social dynamics both of the family sphere, and tomb, or activities limited to the outdoor areas. the broader public dimension.

West & East 131 Monografie, 4 Lorenzo Crescioli

BIBLIOGRAPHY

Akhundov T. 1999, Drevejshie kurgan juzhnogo Kavkaza Daffina P. 1999, Histoire des Xiongnu, le premier empi- kultura podkurgannyzh sklepov, Baku (in Russian). re des steppes, Paris. Akhundov T. 2001, Severo-Zapadnyj Azerbajdzhan v Eldjárn K. 2000, Kuml og haugfé úr heiðnum sið á epokhu Eneolita i Bronzj (Akademija Nauk Respublik Íslandi, in: Friðriksson A. (ed.), 2. edition: Mál og Azerbajdzhan Institut Areologii i Ethnografii), Baku menning, Reykjavík. (in Russian). Engin A., Hewing B. 2012, The EBA-MBA Transition Alekseev A.Yu., Murzin V.Yu., Rolle R. 1991, in the Kilis Plain, in: Laneri N., Pfalzner P., Chertomlyk. Skifskii tsarskii kurgan IV v. do ne, Kiev Valentini S. (eds.), Les vivants et leurs morts: actes (in Russian). du colloque organisè par la College de France, Paris, les Arcini C., Jacobsson B. 2008, Vikingarna från 14-15 avril 2010 (Orbis Biblicus et Orientalis 257) Vannhög, Ale, «Historisk tidskrift för Skåneland» Fribourg/Göttingen, pp. 33-51. 1, pp. 1-14. Felli C. 2015, Dopo la Morte. Pratiche funerarie nella Aspöck E. 2008, What actually is a “Deviant Burial”?: valle del Medio Eufrate fra la fine del III e la prima comparing German-language and Anglophone rese- metà del II millennio a.C., Firenze. arch on “Deviant Burials”, in: Murphy E.M. (ed.), Felli C. 2016, How to cope with death: Mourning and fu- Deviant burial in the archaeological record, Oxford, nerary Practices in the Ancient Near East. Proceedings pp. 17-34. of the International Workshop, Firenze 5th-6th Aspöck E. 2011, Past “disturbances” of graves as a source: December 2013, Pisa. taphonomy and interpretation of reopenend early Halsall G. 1995, Settlement and Social Organization. medieval inhumation graves at Brunn Am Genirge The Merovingian Region of Metz, Cambridge. (Austria) and Winnall II (England), «Oxford Hertz R. 1960[1907]. A Contribution to the Study of Journal of Archaeology» 30/3, pp. 299-324. the Collective Representation of Death, in: Hertz R. Bachelot L. 1992, Une Tombe construite du deuxième (ed.), Death and the Right Hand, Glencoe. millénaire avo J.C., à Mohammed Diyab, in: Durand Klevnäs A. 2013, Whodunnit? Grave Robbery in Anglo- J.-M. (ed.), Recherche en Haute Mésopotamie. Tell Saxon England and the Merovingian Kingdoms, BAR Mohammed Diyab. Campagnes 1990 et 1991, International Series 2582, Oxford. Mémoires de NABU Vol. 2, Paris, pp. 31-38. Klevnäs A. 2015, Give and take: grave goods and gra- Bendezu-Sarmiento J., Grizeaud J.-J. 2011, Le pil- ve robbery in the early Middle Ages, in: Klevnäs A., lage des tombes en Asie centrale meridionale et steppi- Hedenstierna-Jonson C. (eds.), Own and be ow- que. Une affaire ancienne et pas seulement lucrative. ned. Archaeological approaches to the concept of posses- In: Compagnon G. (ed.), Halte au pillage, Paris, sion, Stockolm, pp. 157-188. pp. 31-48. Klevnäs A. 2015a, Overkill, reopening graves to maim Beysenov A.Z., Dzhumabekova G.S., Bazarbaeva the dead in Anglo-Saxon England, in: Gardela L., G.A., Barinova E., Crescioli L. 2015, Arkheolo Kajkowski K. (eds.), Limb, bones and reopened gra- gicheskiye issledovaniya mogil‘nika rannego zhelezno- ves in past societies, International Interdisciplinary go veka Kaspan-6 v Zhetysu. In: Sakskaya kul‘tura Meetings Motifs through the Ages Vol. 2, Bytow: Saryarki v kontekste izucheniya etnosotsiokul‘turnykh Muzeum Zachodniokaszubskie w Bytowie, protsessov Ctepnoy Yevrazii. Sbornik nauchnykh statey, pp. 177-213. posvyashchennyy pamyati arkheologa K. A. Akisheva, Almaty, pp. 59-70 (in Russian). Klevnäs A. 2015b, Introduction: the nature of belon- gings, in: Klevnäs a., Hedenstierna-Jonson C. Bottéro J. 1980, La mythologie de la mort en (eds.), Own and be owned. Archaeological approaches Mesopotamie anciennes, in: Alster B. (ed.), Death to the concept of possession, Stockolm, pp. 1-22. in Mesopotamia; Papers Read at the XXV/e Rencontre Klevnäs A. 2016, Assyriologique Internationale (Mesopotamia 8), Necromancy in grave reopening practi- Copenhagen, pp. 25-52. ces in Late Iron Age Scandinavia, EAA 2016, Vilnius

West & East 132 Monografie, 4 ‘Living Burials’: reopening actions in kurgans and tombs between Central Europe and the Eurasian Steppe

TH1-15 “Charting Otherworlds: Cemeteries as maya 2011 g., Rostov-na-Donu, Kagal’nik (Izd-vo cult sites”. YU NTS RAN), Rostov, pp. 179-192 (in Russian). Koryakova L.N., Epimakov A.V. 2007, The Urals Ochir-Goryaeva, M.A. 2016, Traces of post-funera- and the Western Siberia in the Bronze and Iron Age ry rituals in the Scythian burials of Eurasian steppes, (Cambridge World archaeology), Cambridge. in: Nosov E.N., Vishnyatsky L.B., Kashuba Kupriyanova E.V. 2016, Traces of post-burial activity M.T., Kilunovskaya M.E., Yatsenko S.A. in the Bronze Age cemeteries: robbery or ritual? (With (eds.), Ancient cemeteries and settlements: post-bu- particular reference to the sites near the Stepnoye villa- rial rites, symbolic interments, and grave plundering ge), in: Nosov E.N., Vishnyatsky L.B., Kashuba (Proceedings of IHMC RAS. Vol. 46), St. Petersburg, M.T., Kilunovskaya M.E., Yatsenko S.A. pp. 113-128 (in Russian). (eds.), Ancient cemeteries and settlements: post-bu- Palumbi G. 2011, The Arslantepe royal tomb and the rial rites, symbolic interments, and grave plundering “manipulation” of the kurgan ideology in Eastern (Proceedings of IHMC RAS. Vol. 46), St. Petersburg, Anatolia at the beginning of the Third millennium, pp. 84-95 (in Russian). in: Borgna E., Müller Celka S. (eds.), Ancestral Kuzmina E.E. 1965, Khronologiya nekotorikh kla- Landscapes. Burial mounds in the Copper and Bronze dov Semirech’ia (Novoe v Sovetskoy Arkheologii, Ages. Proceedings of the international Conference held Moskow), Nauka, pp. 106-110 (in Russian). in Udine (TMO 61, Maison de l’Orient et de la Méditerranée), Lyon, pp. 47-59. McClellan T. 1998, Tell Banat North: The White Monument, in: Lebeau M. (ed.), About Subartu: Papanova V.A. 2004, Pominal’no-pogrebal’nyye obrya- studies devoted to Upper Mesopotamia (Subartu 4/1), dy nekropolya Ol’vii, svyazannyye s kul’tom khtoniche- Turnhout, pp. 243-271. skikh bogov i geroyev, «NPFIZDU» 18, pp. 296‑299 (in Russian). McClellan T., Porter A. 1999, Survey of excavation at Tell Banat: funerary practices, in: del Olmo Lete Parzinger H., Zaybert V., Nagler A., Pleshakov G., Montero Fenollos J.-L (eds.), Archaeology of A. 2003, Der große Kurgan von Bajkara: Studien zu the Upper Syrian Euphrates, the Tishrin Dam area. einem skythischen Heiligtum, Mainz am Rein. Proceedings of the International Symposium held Peltenburg E. 2006, Enclosing the ancestors and the at Barcelona, January 28th-30th 1998, Barcelona, growth of socio-political complexity in Early Bronze pp. 107-116. Age Syria, in: Sepolti tra i vivi – Buried among the li- Mozolevskiy B.M., Polin S.V. 2005, Kurgany ving, evidenza ed intepretazione di contesti funerari Skifskogo Gerrosa IV v.do n.e. (Babina, Vodjana I in abitato. Atti del convegno internazionale (Scienze Soboleva Mogila), Kiev (in Russian). dell’antichità, Storia, Archeologia, Antropologia 14/1), Roma, pp. 215-247. Nagmer R.B. 1979, Gravfält från yngre järnålder – vi- kingatid samt boplats från gropkeramisk tid, bronsål- Peltenburg E. 2016, Mortuary Practices and embodied rd der och äldre järnålder: Stävie 4:1, Stävie sn, Skåne: identity in Northwestern Syria in the 3 Millennium arkeologisk undersökning 1973–1975, 1977, 1978. BC, in: Felli C. (ed.), How to cope with death: Mourning and funerary Practices in the Ancient Near Nielsen L.Ch. 1991, Hedenskab og kristendom. East. Proceedings of the International Workshop, , Religionsskiftet afsejlet i vikingetidens Grave Firenze 5th-6th December 2013, Pisa, pp. 133-156. in: Mortensen P., Rasmussen B. (eds.), Høvdingesamfund og Kongemagt, Århus, p. 245-268. Peltenburg E., Bolger D, Campbell S., Colledge, S., Cook G., Croft P., Deckers K., Eremin Nosov E.N., Vishnyatsky L.B., Kashuba M.T., K., Hamiltion D., Jackson A., Lunt D., Kilunovskaya M.E., Yatsenko S.A. (eds.) McCartney C., Northover P., Parras Z., phi- 2016, Ancient cemeteries and settlements: post-bu- lip G., Pickard C., Prag K., Ridout-Sharpe J., rial rites, symbolic interments, and grave plundering Sconzo P., Shortland A., Stevens C., Watt (Proceedings of IHMC RAS. Vol. 46), St. Petersburg M., Wilkinson T. 2015, Tell Jerablus Tahtani, Syria (in Russian). I Mortuary Practices at an Early Bronze Age fort of the Ochir-Goryaeva M.A. 2011, O planigrafii kurga- Euphrates River (Levant supplementary Series 17), nov Yuzhnogo Priural’ya pozdneskifskoy epokhi, Oxford. in: Yablonskiy L.T., Luk’yashko S.I. (eds.), Pfälzner P. 2006, Syria’s Royal Tombs uncovered, Pogrebal’nyy obryad rannikh kochevnikov Yevrazii: «Current World Archaeology»15, pp. 12-22. Materialy VY Mezhdunarodnoy nauchnoy konferent- sii «Problemy sarmatskoy arkheologii i istorii» 11‑15 Pfälzner P. 2016, Royal Corpses, Royal Ancestors and the Living: the transformation of the Dead in Ancient

West & East 133 Monografie, 4 Lorenzo Crescioli

Syria, in: Felli C. (ed.), How to cope with death: Semenov V.A. 2010, Kosh-Peyskiye kurgany v tuvinskoy Mourning and funerary Practices in the Ancient Near «Doline tsarey» (VI–V vv. do n. e.), in: Alëkshin East. Proceedings of the International Workshop, V.A. et Al. (eds.), Drevniye kul’tury Yevrazii. Matly Firenze 5th-6th December 2013, Pisa, pp. 241-270. mezhdu-nar. konf., posvyashchennoy 100-letiyu so Porter A. 2016, The materiality of mourning, in: Felli dnya rozhdeniya A. N. Bernshtama. SPb.: IIMK C. (ed.), How to cope with death: Mourning and fune- RAN, pp. 241-251 (in Russian). rary Practices in the Ancient Near East. Proceedings of Semenov V.A., Kilunovskaya M.E. 2016, Destruction, the International Workshop, Firenze 5th-6th December Plundering, Ritual, funeral rite in the context of an- 2013, Pisa, pp. 157-188. cient cultures (with particular reference to the archeolo- Poulmarc’h M., Pecqueur L., Jalilov B. 2014, An gical sites of Tuva), in: Nosov E.N., Vishnyatsky Overview of Kura-Araxes Funerary Practices in the L.B., Kashuba M.T., Kilunovskaya M.E., Southern Caucasus, «Paléorient» 40/2, pp. 233-248. Yatsenko S.A. (eds.), Ancient cemeteries and settlements: post-burial rites, symbolic interments, and gra- Reynolds A. 2009, , Anglo-Saxon Deviant burial custom ve plundering, (Proceedings of IHMC RAS. Vol. 46), Oxford. St. Petersburg, pp. 15-36 (in Russian). Rudenko S.I. 1962, Sibiriskaya Kollektsiia Petra I, Semenov V.A., Kilunovskaya M.Ye., Glukhov Archaeologia SSSR: Svod Archaeologicheskikh V.O., Zav’yalov V.A., Sadykov T.R., Smirnov Istochnikov, D-39, Akademiia Nauk SSSR, Moskva- N. YU. 2014, Predvaritel’nyye itogi issledova- Leningrad (in Russian). niy Tuvinskoy arkheologicheskoy ekspeditsii v 2013 Rudenko S.I. 1970, Frozen Tombs of Siberia: The g., «Byulleten’ IIMK RAN 4», pp. 327-355 (in Paziryk Burials of Iron Age Horsemen, Los Angeles. Russian). Satalecki F. 2014, Death as an architect of societies, Shulgin P.I. 2003, Mogil’nik skifskogo vremeni Lokot’- Burial and social identity during the Viking Age in 4a. Barnaul: Izd-vo Altayskogo un-ta (in Russian). (Master Thesis in Archaeology, South-western Scania Sinica B.C. 2006, O kul’te sobaki u skifskogo naseleniya Lund University), Lund. Severnogo Prichernomor’ya v VI–II vv. do n.e, in: Savinov D.G. 1996, Ob obryade pogrebeniy bol’shikh Mezhdunarodnyye otnosheniya v basseyne Chernogo pazyrykskikh kurganov, in: Alekseyev A.YU. et Al. morya v skifo-antichnoye vremya: sb. st. po materia- (eds.), Zhrechestvo i shamanizm v skifskuyu epokhu. lam XI mezhdunar. nauchn. konf. Rostov-na-Donu, SPb.: RGNF; IIMK RAN; Izd-vo GE, pp. 107-111 pp. 5860 (in Russian). (in Russian). Skaarup J. 1989, Doppeltgrave, «Skalk» 3/47, pp. 4-8. Schwartz G. 2012, Era of the Living Dead: Funerary Stein G. 2010, Tell Zeidan, The Oriental Institute, , in: praxis and symbol in third millennium BC Syria 2009-2010 Annual Report, Chicago, pp. 105-118. Pfalzner P., Niehr H., Pernicka E, Wissing A (eds.), (Re-)Constructing funerary rituals in the Thate E.S. 2007, Monuments and minds: monument re- ancient near East, Proceedings of the First internatio- use in Scandinavia in the second half of the first millen- nal Symposium of the Tübingen Post-graduate School nium AD, Lund. “Symbols of the Dead” in May 2009 (Qatna Studien Thorarinsdottir E. 2008, A grave revisited, On Supplementum 1), Wiesbaden, pp. 59-78. grave robbery in Viking Age Iceland (MA Thesis in Schwartz G. 2016, After interment/Outside the tombs: Archaeology), Reykjavík. Some mortuary particulars at Umm el-Marra, in: Valentini S. 2003, Le pratiche e l’ideologia funeraria a Felli C. (ed.), How to cope with death: Mourning Tell Barri/Kahat durante il Bronzo Medio, in relazio- and funerary Practices in the Ancient Near East. ne all’area Siro-Mesopotamica settentrionale, «Studi Proceedings of the International Workshop, Firenze Micenei ed Egeo-Anatolici» 45/2, pp. 273-305. th th 5 -6 December 2013, Pisa, pp. 189-216. Vernant P. 1982, Introduction, in: GnoIi G., Semenov V.A. 1994, Khronologiya kurganov znati mo- Vernant J.-P. (eds.), La Mort, les morts, dans les so- gil’nika Kosh-Pey I v Uyukskoy kotlovine v Tuve, in: ciétés anciennes, Cambridge, pp. 5-15. Masson V.M. et Al. (eds.), Elitnyye kurgany ste- Yablonskiy L.T., Meshcheryakov D.V. 2006, pey Yevrazii v skifo-sarmatskuyu epokhu (materialy Zagadka trinadtsatogo Filippovskogo kurgana. In: zasedaniy «kruglogo stola» 22–24 dekabrya 1994 Obydennova G.T., Savel’yev N.S. (eds.), g., Sankt-Peterburg). SPb.: Fond fundamental’nykh Yuzhnyy Ural i sopredel’nyye territorii v skifo-sar- issledovaniy RAN; IIMK RAN; Izdvo GE, pp. 184- matskoye vremya, Ufa: Gilem, pp. 38-45 (in Russian). 193 (in Russian).

West & East 134 Monografie, 4 ‘Living Burials’: reopening actions in kurgans and tombs between Central Europe and the Eurasian Steppe

Yatsenko S.A, Kilunovskaya M.E. 2016, The distur- Zdanovich, G.B., Zdanovich D.G. 2002, The “count- bed graves: Problems of their study, in: Nosov E.N., ry of Towns” of southern Trans-Urals and some aspects Vishnyatsky L.B., Kashuba M.T., Kilunovskaya of steppe assimilation in the Bronze Age, in: Boyle K., M.E., Yatsenko S.A., (eds.), Ancient cemeteries and set- Renfrew C., Levine M. (eds.) Ancient interactions: tlements: post-burial rites, symbolic interments, and grave east and west in Eurasia (MC Donald Institute for ar- plundering, (Proceedings of IHMC RAS. Vol. 46), St. chaeological research), Cambridge, pp. 249-263. Petersburg, pp. 7-14 (in Russian).

West & East 135 Monografie, 4

From the Caspian to the Aegean, from the Caucasus to the Carpathians. The circulation of the “long-sword” in the second millennium BC

VITTORIA DALL’ARMELLINA

Ca’ Foscari University of Venice

Abstract

Bronze swords with a thin and long blade which is generally over 60 cm and can reach up to 120 cm, often called “rapiers”, were found in second millennium BC elite graves over a wide geographical area from the Per- sian Talysh to the east, to the Aegean to the west. The paper will discuss the geographical and chronological distribution of these objects, which are strongly symbolic and can be connected to the ideals of a rising warrior aristocracy, with a special focus on identifying the routes and ways of diffusion of such ideals, and their possible antecedents, over the northern sections of the Near East and beyond these.

Keywords

Long-sword, rapier, Bronze Age, South Caucasus, Aegean, Anatolia, Talysh, Eastern Europe, aristocracy

West & East 137 Monografie, 4 Vittoria Dall’Armellina

1. Introduction the blade can vary considerably, and a midrib can be either present or absent. Especially in the ear- The so-called long-sword, an innovative type of lier periods, morphological variety is quite high. weapon that was also an emblem of the emerging The shape of the shoulder, the tang and the handle military aristocracy, spreads in the cultural land- can be very different from each other, probably as scape of the Near East and the Eastern Mediterra- a consequence of cultural trends, which tend to dif- nean of the second millennium BC. This is an ob- ferentiate the shape of the weapon on a regional ject with a strong symbolic value that accompanies basis. On the other hand, swords that come from the warriors even after their death. In fact, it is of- the same geographic area are often quite similar to ten part of the funerary apparatus in sumptuous each other.2 “royal tombs”, but it also appears in other prestigious elite contest from the Southern Caucasus to Greece. Sporadic specimens have also been found 3. South Caucasian, Anatolian outside the two limits of this “main diffusion and Aegean swords area”: in the Talysh region of Northwestern Iran and in Eastern Europe. If the spread of the “long The two most numerous corpora of long-swords sword” in the Southern Caucasus, in Anatolia and have been found in the Southern Caucasus and in the Aegean area seems to be fairly uniform and in the Aegean. We must however consider that it on the whole chronologically clear, the issue is is in Anatolia, in the late fourth millennium BC, more complex for these western and eastern exten- that the first sword prototypes were found, pre- sions as, on the one hand, the publications dealing cisely at Tulintepe, Arslantepe, and in Sivas re- with the Talysh swords are very dated and, on the gion.3 All these zones show connections with the other one, finds of this type of weapon in Eastern Caucasus since the Late Chalcolithic period, espe- Europe are very scarce and often result from fortui- cially with the South Caucasian Kura-Araxes, but tous discoveries. also with the North Caucasian Majkop/Novosvo- In the following, we will discuss the geographi- bodnaja culture.4 In fact a primitive sword model cal and chronological distribution of these objects, was also found in the Klady cemetery in the north- which are strongly symbolic and can be connected western Caucasus.5 Furthermore, the oldest long- to the ideals of a rising warrior aristocracy, with a swords, which are dated to the mid-third millen- special focus on identifying the routes and ways of nium BC were found in Anatolia, precisely in the diffusion of such ideals, 1 and their possible anteced- “Royal Cemetery” of Alacahöyük6 (fig. 1: A). There ents, over the northern sections of the Near East is however a second place, geographically very far and beyond these. away from Anatolia, where a possibly roughly contemporary long-sword was discovered: the necrop-

2. Typology 2 The typological question is rather complex and we have chosen not to treat it in this article. For the complete descrip- The long-sword is a metal weapon with a thin blade tion of the different types of long-sword see:Dall’Armelli - na 2017, pp. 146-153. with two cutting edges, often balanced by a midrib, 3 Yalçın, Yalçın 2009, p. 126 (Tülintepe); Di Noc- which can be more or less marked. In any case, the era et Al. 2004, pp. 66-67; Palumbi 2004, pp. 114-119; considerable length of the blade, which is mostly Frangipane 2007-2008, pp. 169-193 (Arslantepe); Zimmer- over 60 cm but can reach up to 120 cm, is its main mann et Al. 2011, pp. 1-7 (Sivas). 4 Palumbi 2011, pp. 47-59. distinctive feature. The shape of the upper part of 5 Anthony 2007, pp. 291-283; Rezepkin 2000, 1 To deepen other topics (for examples the typological pp. 62-67. classification, the contexts of discovery, the iconographic sourc- 6 Koşay 1944, pp. 118-120, pl. LXXXI, 26; Sandars es and of the written testimonies about the long-sword) see 1961, pp. 18-19, pl. 15,3-5; Gürsan-Salzmann 1992, pp. 73, Dall’Armellina 2017, pp. 143-182. 81-82, 91, 139-140, 142, 146.

West & East 138 Monografie, 4 From the Caspian to the Aegean, from the Caucasus to the Carpathians…

Figure 1 Examples of long swords from the Southern Caucasus, Anatolia and the Aegean area. A - Alacahöyuk (afterSandars 1961, pl. 15); B - Dzoraget (after Abramishvili 2001, Taf. 2); C - Lefkada (after Sandars 1961, pl. 17, 6); D - Malia (after Chapouthier 1938, fig. 4); E - Hattuša (after Taracha 2003, fig. 1); F - Ugarit (after Schaeffer 1955, p. 277)

olis of Nidhiri, on the island of Lefkada, where Central Anatolia,9 at Byblos in the Levant10 but also tomb cR7 contained a large set of weapons includ- in the Aegean area, with recoveries both in the is- ing a fragment of blade (fig. 1: C).7 We can thus suppose that Eastern Anatolia was the place where vili 2001, p. 4, Taf. 2,7), Mravaltskali (Picchelauri 1987, the “idea of sword” was born, and where later long- p. 37, Taf. XLIV; Abramishvili 2001, pp. 4-5, Taf. 2,2), Sa- duga (Abramishvili 2001, p. 5, Taf. 2,1), Maisan (Ares- swords developed. jan 1986, p. 494; Abramishvili 2001, pp. 5-6, Bobokyan Later on, starting from the late third/early sec- 2008, p. 60, Taf. 24) and Lori-Berd (Bobokhyan 2008, p. 60, ond millennium, the long-sword begins to spread Taf. 24). into various regions: in the Southern Caucasus with 9 Examples have been found at Diyarbakir (Güter- 8 bock 1965, 197-198, pl. XIII), Ganziantep (Summers 1991, the Trialeti culture (fig. 1: B) , in South-eastern and 184-186, figs. 6b, 8), Soli (Bittel 1940, pl. IV, fig. 6; San- dars 1961, p. 22, pl. 16,5) and Kültepe (Özgüc 1986, p. 75, 7 Dörpfeld 1927, p. 229; Sandars 1961, p. 26, pl. pl. 129,3). 17,6; Sandars 1963, p. 145. 10 Dunand 1937, pp. 148-149, figs. 138-139; Schaef- 8 Examples have been found at Samtavro (Abramish- fer 1948, pp. 60-61; Sandars 1961, p. 20, pl.16,6.

West & East 139 Monografie, 4 Vittoria Dall’Armellina

lands (fig. 1: D)11 and on the mainland.12 From the Attica,14 Eubea,15 in Crete16 and Rhodes17) and long seventeenth to the fifteenth century BC attestations swords make their appearance also in different sites in the Southern Caucasus and in Anatolia become in the Levantine area (fig. 1: F). On the contrary, less frequent, and the main propagation centre of they are no longer present in Anatolia and in the these weapons is the Peloponnese, especially Myce- Southern Caucasus. nae (fig. 2). It is important to understand which may have From the fourteenth century onwards, most finds been the main routes that allowed the long sword are located in the Aegean region (Peloponnese,13 to reach the Aegean area. We have three different

Figure 2 Examples of long swords from Mycene A- Circle A Tomb V cat. 730 (after Karo 1930, Taf. LXXXI); B- Circle B Tomb N (after Mylonas 1973, pl. 147 α)

14 At Athens (Immerwahr 1973, pp. 12-13, fig. 21; Driessen, Macdonald 1984, p. 69), Perati (Iakovides 1954, pp. 89-103, fig. 5; Sandars 1963, 152); Tangara (Or- 11 In particular at Malia (Charbonneaux 1925, landoi 1971, p. 18, fig. 15; Driessen, Macdonald 1980, pp. 1-18, figs. 4-11;Chapouthier 1936, pp. 15-30, pl. 1; p. 70) and Orchomenos (Driessen, Macdonald 1984, Sandars 1961, p. 17, pl. 17,1) and Arkalochori (Hazzidakis p. 69). 1912, pp. 35-47; Marinatos 1935, pp. 212-220; Sandars 15 Themeles 1972, p. 250, pl. 215a; Driessen and 1961, p. 17, pl. 17,3), Aegina (Kilian-Dirlmeier 1997, Macdonald 1984, p. 69. pp. 13-23, Abb. 5-7) and Lefkada Dörpfeld( 1927, p. 229; 16 Knossos (Evans 1906, pp. 33-37, 50, 62, figs. 66, 98 Sandars 1961, p. 26, pl. 17,5). figs. 38, 53, 66; Evans 1935, p. 849, fig. 832; Hutchinson 12 In the Circle B of Mycenae (Mylonas 1973), at Ar- 1956, pp. 68-73, figs. 2, 16; Hood 1956, pp. 81-99, figs. 3, 5, gos (Protonotariou-Deilaki 1990, p. 79; Gonzato pl. 14e; Hood, De Jong 1952, pp. 249, 265, fig. 15a, pl. 50; San- 2012, p. 123), Galatas (http://chronique.efa.gr/index.php/ dars 1963, pp. 144-147, pl. 21,1, pl. 23,12; Mervyn Popham, fiches/voir/1935/) and Thebes (Kilian-Dirlmeier 1993, Catling 1974, p. 226, figs. 16-17; Driessen, Macdonald p. 18, Taf. 5,31). 1984, p. 70), Kydonia (Andreadaki-Vlazaki 2010, p. 525) 13 At Dendra (Persson 1931, pp. 34-36, pl. XX; San- and Kato Symne (Lebessi 1976, p. 10; Driessen, Macdon- dars 1963, p. 147), Prosymna (Sandars 1963, p. 145); Nich- ald 1984, p. 71). oria (Wilkie 1987, p. 132, pl. XXXIIIa) and Epidaurus (Pa- 17 Maiuri 1926, p. 100, fig. 15; Jacopi 1930, p. 297, padimitriou 1948, pp. 101-102; Lambrinudakis 1981, figs. 42-43; Sandars 1963, p. 145; Driessen and Macdon- p. 62, fig. 13). ald 1984, p. 69.

West & East 140 Monografie, 4 From the Caspian to the Aegean, from the Caucasus to the Carpathians…

possible ways of communication. The first con- nates from Semacha (fig. 3: B). The scholar dated nects the Southern Caucasus to the area of the As- this item to the eighteenth/seventeenth centuries syrian merchant colonies in Central Anatolia and BC because it was found in association with some then, through this, reaches the coast of the Medi- black polished ware.20 The second sword comes terranean Sea between Cilicia and the Northern Le- from Astara (fig. 3: A), but we do not have any oth- vant and, from there, finds a sea outlet to the Aege- er information about the circumstances of finding an. This route may be reflected by some Anatolian or dating hypotheses.21 finds (Kültepe, Diyarbakir, Gaziantep, Soli) and The Talysh swords come from funerary contexts by an isolated in north Levant (Byblos), but also in which were archeologically investigated in the nine- the fact that most of the oldest Aegean swords were teenth century by J. De Morgan.22 Two of them found in the islands (Crete, Aegina, Lefkada and were found in the Veri (fig. 3: D-E) cemetery, one Amorgos). inside a dolmen tomb at Hovil (fig. 3: C), and the The second possible itinerary would pass to the last comes from a circular grave at Khodja-Daoud. south of the Pontus mountains, then would cross Unfortunately De Morgan provided insufficient in- the Bosphorus canal and then again reach Greece. formation about the chronology of these graves, did At last, we must consider also the situation of the not number them and often published materials be- site of Troy. No swords were found here, but the longing to their funerary equipment without speci- site indubitably returned materials that show strong fying which tomb they come from. connections, since the third millennium BC, with At a later date, C.F.A. Schaeffer inStratigra - the Near East, the Aegean area and with the South- phie Comparée et Chronologie de l’Asie Occidentale ern Caucasus.18 So we must take into account also a revised De Morgan’s work and dated some swords third route that, starting from the Southern Cauca- to his Talyche Recent I period (1550‑1450 BC) sus, passed through the area of Troy site and flowed and some other to his Talyche Recent II period from there into the Mediterranean. (1450‑1350 BC).23 Schaeffer’s chronological periodisation of the Talysh region is based on some Mi- tanni-style seals, found and published by De Mor- 4. Long-swords from Azerbaijan gan, which show strong similarities with specimens and the Talysh region from Palestine and Ugarit.24 More recently some scholars, such as Schachner25 and Abramishvili,26 Mention should now be made of some long-swords have however suggested that Schaffer’s chronology found in Azerbaijan and in the Talysh area of of the region should be revised and that, as a conse- Northwestern Iran. These finds attest a diffusion quence, these long swords might be contemporary of the long sword in the south-eastern direction, as with those found in Georgia and Armenia, which well. There are, however, some major problems con- are presently dated to the twentiet‑seventeenth cen- cerning the dates of these specimens. The literature turies. Although this can not be ruled out, it can not regarding long-swords discovered in Azerbaijan is be proved at the present state of knowledge, espe- fragmentary and approximate.19 Fortunately Gisela cially since in one of the Tombs of Veri, of which Burger in 1994 published two examples of swords a relatively good drawing was published, it appears coming from this country. The first is preserved at that daggers with a half-moon grip have been found the archaeological museum of Baku and it origi- 20 Burger 1994, p. 210. 21 Burger 1994, p. 211. 18 In particular, on the similarities of the various artefacts 22 De Morgan 1896, pp. 30-36, figs. 34, 38, 40, 56, found by Schliemann with contemporary examples from all 2005 p. 257, fig. 341. these areas, see Treister 1996, pp. 197-232; see also the table 23 Schaeffer 1958, pp. 415-486. of comparison of axe-hammers from Troy with examples from 24 the steppe in Kohl 2007, fig. 3.30. Castelluccia 2017, pp. 391-410. 25 19 Burger 1994, p. 210, Taf. VI, 2; Schachner 2001, Schachner 2005, pp. 179-180. p. 120, Taf. 36f – k; Areshian 2008, p. 59, Abb. 2,8. 26 Abramishvili 2010, note 61.

West & East 141 Monografie, 4 Vittoria Dall’Armellina

Figure 3 5. Bulgarian and Romanian long-swords Long swords from Azerbaijan and the Talysh area A - Astara (after Burger 1994, Taf. VI) B - Samacha (after Burger 1994, Taf. VI); In the second half of the second millennium, the C - Hovil (after De Morgan 1986, fig. 56,29) long-sword begins to spread also in other areas of D - E - Veri (after De Morgan 1896, figs. 56,1, 56,3) Europe, in particular in the Balkan and Carpathi- an regions. These European specimens, published in section IV of the series Prähistorische Bronzefunde,28 are probably closely related to the Greek ones, although unfortunately their dating is also very problematic. Two particularly significant examples should be mentioned in this regard. The first one is preserved at the Museum of Varna, in Bulgaria (fig. 4: A). Its archaeological context is unknown, but it was supposedly found in north-eastern Bulgaria.29 It is the fragment of a sword with a mid-rib and a blade that gradually nar- rows from the hilt plate to the tip. This particular profile shows a great similarity to the sword that was discovered in 1991 at the Lion Gate in the Hittite capital of Hattuša (fig. 1: E).30 The well-known Hit- tite sword has an inscription with the name of king Tudhaliya I, who lived in the end of the fifteenth century, and this fact provides a clear fixed point for the dating of the object. Taking into account the similarity of the two weapons, the sword preserved in Varna might be contemporaneous with that of Hattuša, although the possibility of a later dating should not be excluded. Another sword very similar to those of Varna and Hattuša was found in 1992 at Densuş in Roma- nia (fig. 4: B).31 The weapon, which is part of a private collection, was casually found along the banks of the river, thus its dating is unknown. We can observe, however, that the blade, in good preservation status and 73 cm long, has strong typological analo- gies with the specimens discussed above.32

along with the long sword.27 This material is clearly later than the Georgian and Armenian swords, 28 Kilian-Dirlmeier 1993; Bader 1991. and would imply a later date for the associated long 29 Athanassov, Kraus, Slavčev 2009, pp. 17-32. sword, although it cannot be excluded that a reuse 30 Ünal, Ertekin, Ediz 1991, pp. 46-52; Salvini, of the tomb, as frequently attested in the region, had Vagnetti 1994, pp. 215-235; Hansen 1994, pp. 213-215; not been noticed by De Morgan. Taracha 2003, pp. 367-376. 31 Gogâltan 1997, pp. 55-65. 27 De Morgan 1896, fig. 40. 32 Gogâltan 1997, pp. 55-65.

West & East 142 Monografie, 4 From the Caspian to the Aegean, from the Caucasus to the Carpathians…

Figure 4 Long swords from Bulgaria and Romania

A - Varna (https://www.researchgate. net/figure/Abb-1-Schwertklinge-aus- dem-Museum-in-Varna-Inv-Nr-I-3762_ fig1_228543207, 23/06/2018)

B - Densuş (after Gogâltan 1997, fig. 2)

West & East 143 Monografie, 4 Vittoria Dall’Armellina

6. Complete catalogue of long-swords from South Caucasus, Anatolia, Levant and Greece (mainland and islands)*33

SOUTH CAUCASUS Context of discovery Date

Samtavro kurgan No. 243 funerary 20th-18th c. Lilo kurgan No. 1 funerary 17th-15th c. Mravaltskali kurgan No.12 funerary 20th-18th c. Tetri Kvebi kurgan No. 1 funerary 17th-15th c. Saduga kurgan No. 2 funerary 20th-18th c. Dzoraget funerary unknown Ghatchaghan funerary (?) 20th-18th c. Maisan Valley funerary 20th-18th c. Lori-Berd unknown 20th-18th c. Vorotnaberd funerary unknown Angekhakot funerary unknown Dzora funerary unknown Nojemberjan unknown unknown Odzun unknown unknown Hovil unknown unknown Gachiani unknown unknown

ANATOLIA Context of discovery Date

Alacahöyük (3 pcs) funerary second half of 3rd mill. Gaziantep (2 pcs) hoard 20th-17th c. Çorum unknown unknown Soli Pompeiopolis hoard 20th-17th c. Kültepe (2 pcs) urban 20th-17th c. Bodrum unknown unknown

*33 Because of the different regional periodisations, which are difficult to convert into absolute dates, it was decided to divide the diffusion of the “long swords” into four main chronological periods: 1 – second half of the 3rd millennium 2 – 20th-18th centuries 3 – 17th-15th centuries 4 – 14th century or later.

West & East 144 Monografie, 4 From the Caspian to the Aegean, from the Caucasus to the Carpathians…

ANATOLIA Context of discovery Date

Bolu funerary unknown Nallihan (2 pcs) hoard unknown Diyarbakir unknown 20th-18th c. Hattuša sporadic find 17th-15th c. Kasatamonu sporadic find unknown

LEVANT Context of discovery Date

Byblos hoard 20th-18th c. Tell Es-Sa’Idiyeh funerary 14th c. or later Ugarit hoard 14th c. or later Ugarit urban 14th c. or later Ugarit ( 4 pcs) hoard 14th c. or later Gezer funerary 14th c. or later Alalakh urban 14th c. or later

CRETE Context of discovery Date

Malia (3 pcs) urban 20th-18th c. Arkalochori sacred 20th-18th c. Zakros (2 pcs) urban 17th-15th c. Knossos (13 pcs) funerary 14th c. or later Kato SymeViannou (3 pcs) sacred 17th-15th c. Kydonia funerary 14th c.

MYCENAE Context of discovery Date

Circle B tomb A ( 3 pcs) funerary 20th-18th c. Circle B tomb Γ (4 pcs) funerary 20th-18th c. Circle B tomb Δ funerary 20th-18th c. Circle B tomb Z funerary 20th-18th c.

West & East 145 Monografie, 4 Vittoria Dall’Armellina

MYCENAE Context of discovery Date

Circle B tomb I funerary 20th-18th c. Circle B tomb Λ funerary 20th-18th c. Circle B tomb N (2 pcs) funerary 20th-18th c. Circle A tomb IV (7 pcs) funerary 17th-15th c. Circle A tomb V (11 pcs) funerary 17th-15th c. Circle A tomb VI funerary 17th-15th c. Chamber tomb 81 funerary 17th-15th c. Chamber tomb 78 funerary 17th-15th c. Acropolis unknown unknown

CONTINENTAL GREECE Context of discovery Date

Kakovatos tholos B funerary 17th-15th c. Dendra tholos tomb (4 pcs) funerary 14th c. or later Dendra (?) unknown unknown Prosymna chamber tomb XXXVII funerary 14th c. or later Peramatos Ioannina funerary unknown Galaxidi unknown unknown Mount Olympus (?) unknown unknown Unknown unknown unknown Nichoria tholos tomb funerary 14th c. or later Perati tomb 12 funerary 14th c. or later Argo mound E tomb 5 funerary 20th-18th c. Agios Vasileios (batch of 16 pcs.) unknown 17th-15th c. Pylos funerary 17th-15th c. Galatas tomb 3 funerary 20th-18th c. Tangara funerary 14th c. or later Iolco tomb 6 funerary 17th-15th c. Orchomenos urban (?) 14th c. or later Athens Agora’s cemetery funerary 14th c. or later Thebes funerary 20th-18th c. Epidaurus sanctuary of Apollo (2 pcs) sacred 14th c. or later

West & East 146 Monografie, 4 From the Caspian to the Aegean, from the Caucasus to the Carpathians…

GREEK ISLANDS Context of discovery Date

Aegina King’s Grave funerary 20th-18th c. Rhodes Ialysos (3 pcs) funerary 14th c. or later Skopelos Staphilos Tomb funerary 17th-15th c. Levkas Nidhri tomb R7 funerary second half of 3rd Mill. Levkas Nidhri tomb 24, funerary 20th-18th c. Euboea funerary 14th c. or later Amorgos (4 pcs) unknown unknown

7. Long-swords from Arzebaijan and the Talysh region mentioned in the article

AZERBAIJAN Context of discovery Date

Semacha unknown 18th-17th c. Astara unknown unknown

TALYSH Context of discovery Date

Veri (2pcs) funerary 20th-17th c. (?) Hovil funerary 20th-17th c. (?) Khodja-Daoud funerary 20th-17th c. (?)

8. Long-swords from Bulgaria and Romania mentioned in the article

BULGARIA Context of discovery Date

North-eastern Bulgaria unknown 17th-15th c. (?) (preserved at the Mueseum of Varna)

ROMANIA Context of discovery Date

Densuş sporadic find 17th-15th c. (?)

West & East 147 Monografie, 4 Vittoria Dall’Armellina

9. Conclusions In conclusion, what merits to be evaluated with greater interest in the future is not only the inter- What emerges from the situation described above nationalisation, in the course of the Middle and esis that the first half of the second millennium is a pecially of the Late Bronze Age, of trade and the period characterised by a strong “cultural sprint”. wide-ranging distribution of specific products, but The Southern Caucasus and the Anatolian region also the sharing of common customs and ideolo- seem to be not only the places of incubation of the gies. The sumptuous elite graves in which many long long-sword, but also the starting point for the diffu- swords were found, along with other precious ob- sion of a series of symbols and traditions related to jects and weapons, are proof of a cultural koinè of it. The fact that some long swords have been found common values that result in a birth of a new aris- in Azerbaijan and in the Talysh region proves that tocratic class, which maintains, or exhibits its pow- there is not only a one-way diffusion vector, but a er through military exercise. The virtues that the series of multidirectional steps that lead to a broad- sword embodies thus show a common fil rouge from spectrum propagation of social values. The presence the Caspian to the Aegean, and from the Caucasus of long-swords in Eastern Europe seems to confirm to the Carpathians. this hypothesis even more

Figure 5 Distribution of the long sword from the third to the first half of the second millennium BC

West & East 148 Monografie, 4 From the Caspian to the Aegean, from the Caucasus to the Carpathians…

BIBLIOGRAPHY

Abramishvili M. 2001, Transcaucasian Rapiers and tional de thracologie (à participation internationale) the Problem of their Origin, in: Boehmer R.M., Băile Herculane, 14-19 septembre 1992), pp. 14‑19. Maran J. (eds.), Lux Orientis, Archäologie zwischen Castelluccia M. 2017, The Talesh Region in the Iron Asien und Europa, Stellerloh, pp. 1-8. Age and its Relations with Transcaucasia, in: Rova E., Abramishvili M. 2010, In search of the origins of me- Tonussi M. (eds.), At the Northern Frontier of Near tallurgy. An overview of South Caucasian evidence, in: Eastern Archaeology: Recent Research on Caucasia Hansen S., Hauptmann A., Motzenbäcker and Anatolia in the Bronze Age (Proceedings of I., Pernicka E. (eds.) Von Majkop bis Trialeti the Humboldt Kolleg Venice, 09-12/01/2013), Gewinnung und Verbreitung von Metallen und Turnhout, pp. 391-410. Obsidian in Kaukasien im 4.–2. Jt. v. Chr. Beiträge Charboneneaux J. 1925, Trois armes d’apparat du pa- des Internationalen Symposiums in Berlin vom 1.–3. lais de Mallia (Crète), «Monuments et Mémoires de Juni 2006, Bonn, pp. 167-178. la Fondation Piot» 28, pp. 1-18. Andreadaki-Vlazaki M. 2010, Khania (Kydonia), Chapouthier F. 1936, Deux épées d’apparat découver- in: Cline E.H. (ed.) The Oxford Handbook of the tes en 1936 au palais de Mallia, (Études Crétoises 5), Bronze Age Aegean, Oxford, pp. 518-528. Parigi. Aresjan G. E. 1986, Raskopki v Maisjane, in: Moscow Dall’Armellina V. 2017, Power of Symbols or Symbols B., Rybakov A. (eds.), Archeologiceskie otkrytija of Power? The “long sword” in the Near East and the 1984 goda, Moscow, pp. 427-428. Aegean in the second millennium BC, «Ancient Near Anthony D. W. 2007, The Horse, the Wheel and Eastern Studies» 54, pp. 143-182. Language, how Bronze-Age Riders from the Eurasian De Morgan J. 1896, La Mission scientifique en Perse. Steppes shaped the Modern Word, Princeton. Recherches archéologiques. Tome quatrième, Paris. Athanassov B., Kraus R., Slavčev, V. 2009, De Morgan J. 1905, Recherches au Talyche Persan Eine Bronzeschwerte Ägäisch-Anatolis typs aus dem en 1901, «Mémoires de la délégation en Perse, Museum von Varna, «Analele Banatului S.N.» Recherches Archéologiques» 8, pp. 251-342. XVII, pp. 17-32. Di Nocera, G., Huptmann A., Palmieri A. M. Bachhuber C. 2011, Negotiating Metal and the Metal 2004, Sviluppo della metallurgia al servizio dei capi. Form in the Royal Tomb of Alacahöyük in North- Le prime spade del mondo, in: Frangipane M. (ed.), Central Anatolia, in: Wilkinson T., Sherratt S., Alle origini del potere. Arslantepe, la collina dei leoni, Bennet J. (eds.), Interweaving Worlds: systemic in- Milano, pp. 66-67. teractions in Eurasia, 7th to 1st millennia BC, Oxford, pp. 158-174. Driessen J., Macdonald C. 1984, Some Military Aspects of the Aegean, «The Annual of the British Bader T. 1991, Die Schwerter in Rumänien School at Athens» 79, pp. 49-74. (Prähistoriche Bronzefunde IV/8), Wiesbaden. Dunand M. D. 1937, Fouilles De Byblos Tome Ier 1926- Bittel K. 1940, Der Depot von Soli-Pompeiopolis, 1927, Paris. «Zeitschrift für Assyriologie» 12, pp. 189-190. Dörpfeld W. 1927, Alt-Ithaka: ein Beitrag zur Homer- Bobokhyan A. 2008, Kommunikation und Austausch Frage; Studien und Ausgrabungen auf der Insel im Hochland zwischen Kaukasus und Taurus, ca. Leukas-Ithaka,München. 2500-1500 v. Chr. (BAR International Series1853), Oxford. Evans A. 1906, The Prehistoric Tomb at Knossos, London. Burger, G. 1994, Die transkaukasischen Rapiere Evans A. 1935, und die Möglichkeit ihrer Verbindung mit ande- The Palace of Minos at Knossos. Vol. IV , London. ren Rapierzentren im zirkumpontischen Gebiet, in: part II Roman P., Alexianu M. (eds.) Relations Thraco- Frangipane M. 2007-2008, The Arslantepe Royal Illyro-Helléniques (Actes du XIVe Symposium na- Tomb: New Funerary Customs and Political Changes

West & East 149 Monografie, 4 Vittoria Dall’Armellina

in the Upper Euphrates Valley at the Beginning of the Kohl P. 2007, The Making of Bronze Age Eurasia, Third Millennium B.C., «Scienze dell’Antichità» Cambridge. 14, pp. 169-193. Lambrinudakis V. 1981, Remains of the Mycenaean Gonzato F. 2011, Ostentazione di rango e manife- Period in the Sanctuary of Apollo Maleatas, in: Hägg stazione del potere agli albori della società micenea R., Marinatos N. (eds.) Sanctuaries and Cults in (Praehistorica Mediterranea) Milano. the Aegean Bronze Age (Proceedings of the First Gogâltan F. 1997, O rapiera de tip micenian de la International Symposium at the Swedish Institute Densuş, «Apulum» XXXIV, pp. 55-65. in Athens, 12–13 May 1980) Stockholm, pp. 59-65. Güterbock H. G. 1965, A Votive Sword with Old Lebessi A. 1976, A Sanctuary of Hermes and Aphrodite Assyrian Inscription, in: Güterbock H.G., in Crete, «Expedition» 13, pp. 2-12. Jacobsen T. (eds.), Studies in Honor of Benno Maiuri A. 1923, Jalisos, Scavi della Missione Landsberger on his Seventy-Fifth Birthday, April Archeologica Italiana a Rodi, «Annuario della R. 21, 1965 (Assyriological Studies 16), Chicago, Scuola Archeologica di Atene» VI, pp. 83-341. pp. 197-198. Marinatos S. 1935, Ἀνασκαφαὶ ἐν Κρήτη, Gürsan-Salzmann A. 1992, Alaca Höyük: A ΑΡΚΑΛΟΧΩΡΙ, «Praktikà» 1935, pp. 2012-2020. Reassessment of the Excavation and Sequence of Early Mervyn R., Popham E. A., Catling H. W. 1974, Bronze Age Settelment (Ph.D. Thesis, University of Sellopoulo Tombs 3 and 4, Two Late Minoan Graves Pennsylvania), Ann Arbor. near Knossos, «The Annual of the British School at Hansen O. 1994, A Mycenaean Sword from Boğazköy- Athens» 69, pp. 195-258. Hattusa Found in 1991, «The Annual of the British Müller K. 1909, Alt Pylos. II die Funde aus den School at Athens» 89, pp. 213-215. Kuppelgräbern von Kakovatos, «Athenische Hazzidakis J. 1912, An Early Minoan Sacred Cave at Mitteilunghen» XXXIV, pp. 269-283. Arkalokhóri, «The Annual of the British School at Müller-Karpe, A. 1994, Anatolische Bronzeschwerter Athens 19», pp. 35-47. und Südosteuropa, in: Dobiat C. (ed.) Festschrift für Hood, M. S. F., De Jong P. 1952, Late Minoan Otto-Herman Frey, Marburg, pp. 431-444. Warrior-Graves from Agios Ioannis and the New Mylonas G. M. 1973, Ο Ταφικός Κύκλος Β των Hospital Site at Knossos, «The Annual of the British Μυκηνών, Η εν Αθήναις Αρχαιολογική Εταιρεία, School at Athens» 47, pp. 243-277. Atene. Hutchinson, R. W. 1956, A Late Minoan Tomb Orlandoi A. K. 1971, ΤΑΝΑΓΡΑ, «Ergon» 1970, at Knosso, «The Annual of the British School at pp. 13-22. Athens» 51, pp. 68-73. Özgüç T. 1986, Kültepe-Kanish II. New Researches at Immerwahr S. A. 1973, Early Burials from the Agora the Trading Center of the Ancient Near East, Ankara. Cemeteries, Princeton. Palumbi G. 2004, La più antica tomba “reale”. Dati ar- Iakovides S. 1954, Ἀνασκαφὴ μυκηωαικῶν τάφον cheologici e costruzione delle ipotesi, in: Frangipane Περατῆς, «Praktika» 1954, pp. 89-103. M. (ed.), Alle origini del potere. Arslantepe, la collina Jacopi G. 1930, Nuovi scavi nella necropoli micenea di dei leoni, Milano, pp. 114-119. Jasso, «Annuario della R. Scuola Archeologica di Palumbi G. 2011, The Arslantepe Royal Tomb and Atene» XIII, pp. 253-345. the Manipulation of the Kurgan Ideology in Eastern Karo G. 1930, Die Schachtgräber von Mykenai, Anatolia at The Beginning of the Third Millennium, München. in: Borgna E., Müller Celka S. (ed.) Ancestral Kilian-Dirlmeier I. 1993, Die Schwerter in Landscapes: Burial Mounds in the Copper and Bronze Griechenland (ausserhalb der Peloponnes), Bulgarien Ages (Proceedings of the International Conference, und Albanien (Prähistorische Bronzefunde IV/12), Udine 2008), Lyon, pp. 47-59. Wiesbaden. Papadimitriou I. 1948, Ἀνασκαφὴ ἐν τὼι Ἀσκληπιείωι Kilian-Dirlmeier I. 1997, Das Mittelbronzezeitliche καὶ τῶι ἱερῶι τοῦ Ἀπολλ. Μαλεάτα ἐν Ἐπιδαύρωι, Schachtgrab von Ägina, Mainz. «Praktika» 1948, pp. 98-111. Koşay H. Z. 1944, Ausgrabungen von Alaca Höyük, Persson A. W. 1931, The Royal Tombs at Dendra near ein Vorbericht über die im Auftrage der Türkischen Midea, London. Geschichtskommission im Sommer 1936 durchgeführ- Philip G. 1995, Warrior Burials in the Ancient Near- ten Forschungen und Entdeckungen, Ankara. Eastern Bronze Age: The Evidence from Mesopotamia,

West & East 150 Monografie, 4 From the Caspian to the Aegean, from the Caucasus to the Carpathians…

Western Iran and Syria-Palestine, in Campbell S., am Beispiel einer Stele im Museum von Astara Green A. (eds.), The Archaeology of Death in the (Azerbaycan), «Archäologische Mitteilungen aus Ancient Near East, Oxford, pp. 140-154. Iran und Turan» 33, pp. 115–142. Platon N. 1949, Ὁ τάφος τοῦ Σταφύλοθ και ὁ Μινωικὸς Schachner A. 2005, Von Bronze zu Eisen: die ἀποικισμὸς τῆς Πεπαρήθοθ, «Kretika Chronika » III, Metallurgie des 2. und frühen 1. Jahrtausends v. Chr. pp. 534-573. im östlichen Transkaukasus, in: Yalçın Ü (ed.), Platon N. 1963, A New Minoan Palace, «Archaeology» Anatolian Metal III, Bochum, pp. 75-93. 16/4, 269-275. Summers G. D. 1991, Metalwork in Gaziantep Museum Pritchard J. B. 1980, The Cemetery at Tell Es-Sa’idiyeh, Said to Be a Hoard from the Region of Sakçagözü, Jordan, Philadelphia. «Anatolian Studies» 41, pp. 173-195. Picchelauri K. 1997, Waffen der Bronzezeit aus Ost- Taracha P. 2003, Is Tutḫaliya’s Sword Really Aegean?, Georgien, Espelkamp. in: Beckman G., Beal R., McMahon G. (eds.), Hittite Studies in Honor of Harry A. Hoffner Jr. Picchelauri K., Orthrnann W.1992, Der Kurgan on the Occasion of his 65th Birthday, Winona Lake, “Tetri Kwebi I”, «Georgica» 15, pp. 15-22. pp. 367-376. Protonotariou-Deilaki E. 1990, Burial Customs Treister M. 1996, I tesori troiani attribuzioni, cronolo- , in: and Funerary Rites in the Prehistoric Argolid gia, contesto storico, in: Danilova I., Treister M., Haegg R., Nordquist G.C. (eds.), Celebrations Pruss I. (eds.), Il Tesoro di Troia, Mosca-Firenze, of Death and Divinity in the Bronze Age Argolid, pp. 197-232. (Proceedings of the Sixth International Symposium at the Swedish Institute in Athens, 11-13 June, Ünal A., Ertekin A., Ediz I. 1991, The Hittite Sword 1988), Stockholm, pp. 69-83. from Bogazköy-Hattusa, found 1991 and its Akkadian Inscription, «Müze Museum» 4, pp. 50-52. Rezepkin A. D. 2000, Das frühbronzezeitliche Grä berfeld von Klady und die Majkop-Kultur in Wilkie N. C. 1987, Burial Customs at Nichoria: the Nordwestkaukasien, Rahden. MME Tholos, in: Laffineur R. (ed.) Thanatos. Les coutumes funéraires en Egée à l’âge du Bronze Salvini M., Vagnetti L. 1994, Una spada di tipo (Actes du colloque de Liège, 21-23 avril 1986), Liège, egeo da Bogazköy, «La parola del passato» XLIX, pp. 127-135. pp. 215-236. Wilkinson T. C. 2014, Tying the Threads of Eurasia: Sandars N. K. 1961, The First Aegean Swords and Their Trans-regional Routes and Material Flows in , «American Journal of Archaeology» 65/1, Ancestry Transcaucasia, Eastern Anatolia and Western Central pp. 17-29. Asia, c. 3000-1500BC, Leiden. Sandars N. K. 1963, , Later Aegean Bronze Swords Yaçin Ü., Yaçin H. 2009, Evidence for Early Use of Tin «American Journal of Archaeology» 67/2, at Tülintepe in Eastern Anatolia, «TÜBA-AR» 12, pp. 117-153. pp. 123-141. Schaeffer C. 1948, Stratigraphie Comparée et Zimmermann T., Dilek N., Önder T. K. 2011, Ein Chronologie de l’Asie Occidentale (III et II millénai- neues Schwert vom Typus Arslantepe – frühmetallzeit- , Oxford. res) liche Waffentechnologie zwischen Repräsentation und Schachner A. 2001, Zur Bildkunst des 2. Jahrtausends Ritual, «Praehistorische Zeitschrift» 86, pp. 1-7. v. Chr. Zwischen Kaspischem Meer und Van-See

West & East 151 Monografie, 4

The vaulted funeraryhypogea in Mesopotamia between the second and first millennium BC: localization and architectural features

MARGHERITA DALLAI

University of Florence

Abstract

Vaulted funerary hypogea are familial collective tombs containing multiple progressive burials. This funerary type, only sporadically attested in Mesopotamia in the third millennium BC, spread rapidly from the beginning of the second millennium BC and is attested throughout the first millennium BC across a vast geographical area: in the Middle Euphrates region, in Mesopotamia and in Khuzestan. Thesehypogea are characterized by specific architectural features and localiza- tions, which we will analyse in this work. In fact, they have a complex structure that required specific technical knowledge, especially for the realization of the vault roofing: corbelled vault, with radial arrangement or with tilted rows. Moreover, they were generally located under domestic dwellings or, to a rare extent, under palaces. Analysing the excavation reports which describe these tombs, it is clear that this relationship is a recurring element. This indicates the strong connection between the living and the dead, which is also suggested by the use of similar plans, both in graves and in houses.

Keywords

Vaulted funerary hypogea, Mesopotamia, Khuzestan, second millennium BC, first millennium BC, localization, architectural features

West & East 153 Monografie, 4 Margherita Dallai

1. Introduction tive tombs containing multiple progressive burials, since there is generally only one complete skeleton, In this paper, the data relating to the architecture while other disarticulated bones, belonging to other and localization of vaulted funerary hypogea exca- individuals, are located on either side of the burial vated in Mesopotamia, dating between the second chamber or even in another chamber. and first millennium BC, will be presented.1 This grave type was both geographically and chronologically widespread. Sporadically attested in Mesopo- 2. Materials and building techniques tamia during the third millennium BC (such as at of the vault the sites of Ur, Mari and Khafaja2), this funerary type spread rapidly from the beginning of the sec- The technology of the vault developed out of the ond millennium BC and is attested throughout swampy areas of Mesopotamia over 5,000 years ago. the first millennium BC across a broad geographi- The prototype was probably a structure made with cal area: from the Khabur basin (such as Tell Bar- bundles of reeds, which were planted in the ground, 3 4 ri) to Assyria (Assur and Nimrud) in Northern bent inward and tied together at the ends to form Mesopotamia, and from Central and Southern the roof. Although no ancient buildings of this type Mesopotamia (for example Ur, Khafaja, Uruk, Isin have been preserved, this technique is still used in and Larsa) to Khuzestan in south-west Iran (Susa, Southern Iraq, where large vaulted buildings with Tchoga Zanbil and Haft Tepe). Most of the anal- reed structures continue to be built.6 This type of ysed hypogea are intra muros funerary hypogea. They roofing has progressively spread from the Halaf pe- are tombs located underneath the surface level and riod up to the Iron Age and even into later periods, the specificationintra muros indicates all those hy- across an increasingly wider geographical area and pogea that were built under the floors of buildings. with different solutions from region to region and This specification is important because this kind of from period to period. In this study, we will there- grave has also been found in extra muros contexts, fore consider the various materials used in the con- as defined by Peltenburg, namely, «tombs in cem- struction of vaults, the different techniques of im- eteries and monumental or central settings»; these plementation and their evolution. are outside the scope of this study.5 Analysing excavation reports, it is possible to notice that the relationship between hypogea and dwellings is a perma- 2.1. Building materials nent feature and, from an anthropological point of view, this testifies to a closeness between the living Regarding the materials used for the building of and the dead. these hypogea, it is possible to subdivide them be- Architecturally speaking, these tombs, which tween constituents (stone and mudbrick) and bind- were either cut in the rock or built, were generally ers (mortar of clay, plaster and bitumen). characterized by vault roofing, which was realized Stone was mainly used in the Early Dynastic pe- through the use of different implementation tech- riod for the building of vaults. In these cases, only niques (corbelled, with radial arrangement, with blocks that were roughly shaped or worked into the tilted rows), which we will explore later. Finally, fu- form of a slab were used and generally implemented nerary hypogea are considered to be familial collec- without mortar.7 Mudbrick was the most common material used to realize the vault roofing. Initially, 1 Since this study has not yet been concluded, the results should not be considered to be completely definitive. the unbaked brick was used,but, from the second 2 Delougaz, Hill, Lloyd 1967, pp. 58-142. millennium BC onwards, it was replaced by the 3 Valentini 2000, 2001, pp. 82-85, 2003, pp. 273‑305, baked brick. The former was certainly easier to pro- 2016, pp. 217-240. 4 Haller 1954; Hockmann 2010. 6 Van Beek 1987, p. 82. 5 Peltenburg 2007/2008, p. 220. 7 Besenval 1984, p. 159; Valentini 2002, p. 8.

West & East 154 Monografie, 4 The vaulted funeraryhypogea in Mesopotamia between the second and first millennium BC: localization and architectural features

Figure 1 Different shapes of bricks (after Castel 1996, fig.16)

Figure 2 Bricks from Tell Barri (afterValentini 2003, p. 300)

duce, although the latter offered greater mechanical ral hydrocarbons present in ancient oil fields) were qualities.8 The unbaked brick was mainly used for only rarely used from the middle of the second mil- covering surface structures with a radial arrange- lennium BC.11 ment, while the baked brick is attested in almost all roofing with tilted rows and corbelled vaults in funerary hypogea. The bricks were mostly rectangu- 2.2. Techniques of implementation lar; only rarely did they have special shapes, such as those found at Susa9 and Tell Barri, in trapezoidal The techniques of implementation can be divided ashlar-shaped bricks and semi-curved ashlar-shaped into two categories: those that did not use temporary bricks, respectively (figs. 1-2). It is possible to argue supports (namely, a rib) for the masonry and those that, in general, the mudbrick was the most used that did. The first category includes corbelled vaults building material, especially the baked one, even and vaults realized with tilted rows (figs. 3-4). In the though its production involved high costs.10 As for first case, single rows were arranged, so that they pro- binders, mortar made with clay was the most com- truded in relation to each of the previous ones, until mon type because it was easy to find and prepare. the closing of the vault.12 In the second case, the line Plaster and bitumen (a colloidal mixture of natu- of the vault was indicated on the bottom wall with a thick layer of mortar; meanwhile, the first brick, rest- 8 Besenval 1984, pp. 21-22, 159. 9 Ghirshman 1965a, 1965b; Ghirshman, Steve 1966. 11 Besenval 1984, p. 23; Valentini 2002, p. 8. 10 Valentini 2002, p. 8. 12 Besenval 1984, pp. 41-42.

West & East 155 Monografie, 4 Margherita Dallai

Figure 3 Corbelled vault (after Valentini 2002, p. 8)

Figure 4 Figure 5 Tilted rows Radial arrangement (after Van Beek 1987, p. 85) (after Van Beek 1987, p. 84) ing on it, leaned on the short side tilted towards the ed as support for the vaulting masonry.15 The sec- bottom wall and, at the same time, towards the arch ond type was made using a wooden structure, whose span. A second brick was made to overlap the first, shape corresponded to the curve of the vault. This then a third above the second, which was inclined in- technique was more expensive than the previous wards and supported by the bottom wall. This pro- one, even though this rib could be reused for struc- cedure was repeated until the conclusion of the first tures of the same type.16 The use of a rib was essen- arch, with subsequent arches overlapping it.13 tial in vaults with a radial arrangement (fig. 5). This In the second category, the temporary support technique consisted of a progression of rows of was removed when the vault was closed, and the bricks, radially arranged in parallel to the axis of the structure was dry. For this reason, the main prob- vault. The first row of bricks was inclined towards lem associated with these techniques was the need the arch span (namely, the distance between the to build a support structure to sustain the entire piers) by the interposition of small stones or pottery surface of the vault before realizing the implemen- sherds in the extrados (namely, the external surface tation.14 Two types of rib can be distinguished. The of the vault). Starting with both springers at the first is defined as a “core rib”, which filled the space same time, the process was repeated, progressively that was to be covered by the vault using unbaked tilting row after row, until the arch was closed at the bricks or earth, such that the top of this filling act- top by bricks in a vertical position.17

15 Besenval 1984, p. 47. 13 Besenval 1984, pp. 42-45; Van Beek 1987, p. 85. 16 Besenval 1984, p. 47. 14 Besenval 1984, p. 47. 17 Besenval 1984, pp. 48-49; Van Beek 1987, p. 84.

West & East 156 Monografie, 4 The vaulted funeraryhypogea in Mesopotamia between the second and first millennium BC: localization and architectural features

2.3. The evolution of building techniques

As already mentioned, the model for the first open vaults may have been a structure formed by bundles of reeds tied together. Even today, this technique is used in Southern Iraq for the construction of the so-called mudhif (typical of the Nasiriyah Marshes), characterized by single-celled rooms with an arched tunnel shape and made with a series of arches, joined by bundles of reeds, which made the construction of barrel structures possible.18 Thus, it would seem that the first attempts at covering probably occurred in Southern Mesopotamia, maybe because of the lack of wood that was necessary for the construction of Figure 6 flat roofs. Unfortunately, this interpretation is com- Tholos at Tell Arpachiya, Halaf period (after Mallowan, Cruikschank Rose 1935, p. 30) plex, since the first evidence of this is documented in the north, not in open (barrel) vaults, but closed dome-shaped vaults made with the corbelled technique.19 In fact, this technique is initially attested in the Halaf period in the dome roofing of the tholoi, as found at the sites of Tell Arpachiyah and Yarim Tepe, before spreading to the south, where it is attested in monumental contexts, such as the tombs of the First Dynasty of Ur (figs. 6-7).20 Both at Tell Arpachiyah and at Yarim Tepe, the tombs were composed of a single circular chamber vaulted with blocks of pisé (raw clay) starting from the floor, while tomb 779 from Ur had four chambers Figure 7 made of limestone blocks covered by a vault starting Tombs PG779 at Ur, First Dynasty of Ur at one meter from the ground.21 In the third mil- (after Woolley 1934, pl. 24) lennium BC (Early Dynastic period), the corbelled technique was the main technique for covering underground vaulted structures, both with brick well as the covering of many surface structures, for masonry, such as at Khafajah, and with stone ma- the realization of the covering of two royal tombs 22 sonry, as at Mari and Ur (fig. 8). For example, at at Ur.24 A technique with radial rows and a succes- Mari, there were five tombs: one with a single room sive one with tilted rows (the latter is documented and four coupled together, all of which were cov- from the end of the third millennium) seem to have 23 ered with a corbelled vault starting from the floor. been developed in Southern Mesopotamia, start- From the beginning of the third millennium, a sec- ing from the prototypes of barrel vaults made from ond technique, involving radial rows, is attested, as bundles of reeds; this is evidenced by the examples 18 Valentini 2002, p. 8. found at Uruk, in some buildings in the area of Ean- 25 19 Valentini 2002, p. 9. na. Then, during the Middle Bronze Age, both 26 20 Besenval 1984, p. 160. techniques spread to the north and east. During 21 Mallowan, Cruikshank Rose 1935; Merpert, Munchaev, Bader 1976; Woolley 1934. 24 Besenval 1984, p. 160 22 Besenval 1984, p. 160. 25 Lenzen 1968 23 Jean Marie 1990. 26 Besenval 1984, p. 160

West & East 157 Monografie, 4 Margherita Dallai

Figure 8 Tombs 241, 242 and 300 at Mari, Early Dynastic period (after Jean Marie 1990, pl.40)

Figure 9 Tomb of Tepti Ahar at Haft Tepe, Middle Elamite period (after Negahban 1991, pl. 7)

the second millennium, these different implemen- estan, in the Early Elamite levels of Susa,30 the tilt- tation techniques coexisted. In the first part of the ed rows technique is attested, while the technique millennium, the corbelled vault is attested in Assyr- with radially arranged bricks is absent.31 Regarding ia, as well as in Southern Mesopotamia.27 During the latter half of the second millennium, the situa- the Isin-Larsa period, however, both this technique tion is more confusing. The Middle Assyrian tombs and the tiled rows are attested, especially at Uruk,28 at Assur seem to indicate a decrease in the impor- Isin29 and Ur. At all these sites, the vault arranged tance of the corbelled vault in favour of a vault with with tilted rows started on the side walls. In Khuz- radially arranged bricks.32 Here, six hypogea were

27 Besenval 1984, p. 161. 30 Ghirshman 1967, 1968a. 28 Lenzen 1961, 1962. 31 Besenval 1984, p. 161. 29 Spanos 1987, pp. 17-21. 32 Besenval 1984, p. 161.

West & East 158 Monografie, 4 The vaulted funeraryhypogea in Mesopotamia between the second and first millennium BC: localization and architectural features

covered with corbelled vaults which were generally altars; and a funerary area, containing the tombs, set on the floor, but most of the tombs were cov- which was far from the entrance (fig. 11). ered with a radial arrangement and set on side walls Despite this subdivision, it is not possible to in baked bricks or stone blocks, or directly on the identify specific features that architecturally distin- floor.33 Instead, the vaults in the Haft Tepe tombs, guish structures within a context compared to oth- dating back to the beginning of the Middle Elamite ers; instead, it is possible to isolate some features Period (fifteenth to fourteenth centuries BC), are that are common to the majority of the hypogea. covered with tilted rows, while, at Tchoga Zanbil, no structure involving this technique can be found, since the tombs were built with radial rows (fig. 9).34 Finally, this latter technique was still very much in use in Assyria during the first half of the first millennium BC, especially at Assur, where most of the tombs were made in this way, although a minority were made using tilted rows technique.35

3. Context and localization of the hypogea

The analysedhypogea can be divided into three groups, based on the context in which they are found. Most of them are located below private Figure 10 dwellings, typically situated in artisan and commer- Graph on the localization of the analysed hypogea cial neighbourhoods (such as Tell Mohammed Di- yab36 or Babylon37), to a lesser extent in dwellings near the palace. In some cases (at Uruk, Mari, Assur and Nimrud38), the tombs are below the palaces, especially under the private areas, but sometimes un- At all the analysed sites, access to the tomb is only der the administrative ones. Finally, at Haft Tepe39 possible from one of the rooms inside the building that houses it. In most cases, it is indicated by a kind and Tchoga Zanbil,40 in Khuzestan, such hypogea have been found underneath funerary buildings of baked brick sealing on the floor, as can be seen at (fig. 10). In both buildings, it is possible to identify Tchoga Zanbil, Ur and Tell Barri, while, in other three different areas: a public area, comprising the examples, it is highlighted with stone slabs, such as rooms closest to the entrance; a ritual area located at Assur (fig. 12). A unicum is instead represented in in the centre and characterized, in the first case, by some tombs within the North-West Palace of Nim- a courtyard with an altar in the centre and a pedes- rud, as indicated by bricks boxes, which were also 41 tal near the wall, and, in the second case, by fifteen used to leave offerings to the dead. The connection between the entrance of the tomb and the overhead room was generally realized thanks to a verti- 33 Pedde 2015. cal shaft, which could be excavated into the ground 34 Besenval 1984, p. 161. or built with baked bricks or stone, sometimes pro- 35 Besenval 1984, p. 161. 36 Bachelot 1992, pp. 31-34. vided with a few steps. Tchoga Zanbil has the only 37 Reuther 1926. example in which the vertical shaft was completely 38 Mallowan 1966; Oates J., Oates D. 2001. replaced by a staircase. Concerning the entrance to 39 Negahban 1991. 40 Ghirshman 1968b. 41 Mallowan 1966, p. 114.

West & East 159 Monografie, 4 Margherita Dallai

Figure 11 Schematic plan of the funerary building at Tchoga Zanbil and Haft Tepe, Middle Elamite period (after Mofidi-Nasrabadi 2012, p. 268)

Figure 12 Figure 13 Baked bricks sealing of the tomb III’s entrance Tomb 65 at Assur, Neo-Assyrian period at Tchoga Zanbil, Middle Elamite period (after Haller 1954, p. 161) (after Ghirshman 1959, p. 277)

the tomb, it can be argued that, in almost all cases, As for the plan of these tombs, it is possible to this was realized by an arch placed on one of the note that they generally had a rectangular or square- short sides of the chamber and closed by bricks or shaped funerary chamber; a trapezoidal shape was stones. Sometimes, however, access was possible much less common (fig. 13). In one case, at Chagar directly through the upper part of the vault, as ob- Bazar, the tomb had an apsed end with a semicircu- served in hypogea at Babylon. lar plan (figs. 14-15).

West & East 160 Monografie, 4 The vaulted funeraryhypogea in Mesopotamia between the second and first millennium BC: localization and architectural features

Figure 16 Tomb 70 at Assur, Neo-Assyrian period (after Haller 1954, p. 166) Figure 14 Tomb at Chagar Bazar, Old Babylonian period (after Mallowan 1937, fig. 8)

Figure 15 Figure 17 Graph on the plans of the one-chamber hypogea Tomb 36 at Assur, Middle Assyrian period (after Haller 1954, p. 113)

Most of the hypogea comprise only one room, such sometimes they are arranged at a right angle (such as those at Tchoga Zanbil and Haft Tepe. Some as at Tchoga Zanbil and Assur) (fig. 16), sometimes hypogea instead are equipped with two chambers: they are aligned with each other (as at Tell Bi’a and typically, the one closest to the entrance is regard- Assur) (fig. 17), sometimes they are arranged along ed as the antechamber, while the other is the funer- an axis (as at Nimrud) (fig. 18), and sometimes they ary chamber proper. Their arrangement is variable: are separated by a vertical shaft or the room is di-

West & East 161 Monografie, 4 Margherita Dallai

Figure 18 Tomb II at Nimrud, 8th century BC (after Damerji 1998, p. 54)

Figure 19 Tomb 24 at Assur, Middle Assyrian period (after Haller 1954, p. 106)

Figure 20 Graph on the plans of the two-chambers hypogea

West & East 162 Monografie, 4 The vaulted funeraryhypogea in Mesopotamia between the second and first millennium BC: localization and architectural features

vided into two parts by a partition (for example, at pogeum, unlike the others, has two entrances: the Assur) (figs. 19-20). Meanwhile, at Assur, there is main one is on the short side of the first longitudinal a tomb, probably dating back to the Middle Assyr- room, while the secondary one is on the short side of ian period, which comprises three rooms: two with the latitudinal room. At Assur, in some tombs dat- a longitudinal development on the same axis, and a ing back to the Neo-Assyrian period, the funerary third with a latitudinal development, which has, on chamber is divided into two or three areas by bricks the short side, an annex in the form of a large niche on the floor, whose function was probably to trans- with some grave goods (fig. 21). Moreover, this hy- form these areas into separate rooms (fig. 22).

Figure 21 Tomb 53 at Assur, Middle Assyrian period (after Haller 1954, p. 154)

Figure 22 Tomb 47 at Assur, Neo-Assyrian period (after Haller 1954, pl. 37b)

West & East 163 Monografie, 4 Margherita Dallai

Figure 23 Altar House at Ur, Isin-Larsa period (after Valentini 2016, p. 238)

Figure 24 Figure 25 Graph on the type of the rooms above the hypogea Graph on the relationship between hypogea and buildings

Unfortunately, the relation between the hypo- with a reception area and adjoining bathroom geum and the overhead space is not always under- zones. At Uruk, some burials have been found be- standable; sometimes, it is impossible to interpret low rooms in the administrative area (fig. 24).44 the function of the surface room because the struc- Moreover, I have tried to understand whether the tures in such examples have been almost completely tombs and the buildings that housed them were con- destroyed. Nevertheless, in general, we can say that temporaneous with each other. I have observed that most of the tombs were placed below courtyards hypogea are often planned at the moment of the re- (such as at Ur,42 Tell Mohammed Diyab and Tell alization of a building. This is the case with the struc- Barri). At Ur, many hypogea were located under- tures at Nimrud, Tell Barri and Babylon. On the neath the so-called domestic chapels, which were contrary, they were sometimes realized when the placed far from the building’s entrance (fig. 23). building was already in use, as confirmed by the Tell They were usually paved with burnt brick and, in Barri and Tell Bi’a examples. Finally, Tchoga Zanbil some cases, characterized by a low altar with a hearth has the only case of a tomb that was made before the and drains.43 In other cases, the tombs were under overhead building, while, at Tell Mohammed Diyab, private rooms, as in the case of the North-West Pal- the hypogeum was built later than the first dwelling ace at Nimrud, where there were several apartments phase, but earlier than the rebuilding phase (fig. 25).

42 Woolley, Mallowan 1976. 44 Lenzen 1961, pp. 21-23, 1962, pp. 28-29, 1963, 43 Valentini 2016, p. 13. pp. 35-36.

West & East 164 Monografie, 4 The vaulted funeraryhypogea in Mesopotamia between the second and first millennium BC: localization and architectural features

4. The chamber tombs at Ugarit and the large stone-built corbelled vaulted chamber tombs (exemplified by tombs 50, 101 and 20147), Subsequently, funerary hypogea excavated at Ugarit which are the ones we are specifically interested in have been analysed in order to compare the struc- (figs. 26-27).48 The realization system of these structures found on this site and those situated in Meso- tures is common, although there are many varia- potamia and Khuzestan.45 tions, and essentially consist of two parts: a verti- On this site, there are more than 200 chamber cal shaft or a hallway dromos( ) with stairs, giving tombs, distributed more or less across the entire ex- access to the funerary chamber. Generally, these cavated area. Based on Marchegay’s work, it is pos- graves are characterized by a square or rectangular sible to divide these tombs into two types: chamber plan and by an average size of two square meters, al- tombs with walls built of rubble stones and covered though they can exceed ten square meters.49 The fu- with slabs, which are the most common type (exam- nerary chamber usually comprised a series of niches ples of which are tombs 56 and 57, 103 and 106846), in the side walls, and sometimes even in the back,

Figure 26 Tomb 103 at Ugarit, Type 1, Late Bronze Age (after Teba 2015, p. 265)

Figure 27 Tomb 101 at Ugarit, Type 2, Late Bronze Age (after Teba 2015, p. 266)

47 On tomb 201 see also Margueron 1983. 45 Yon 1997. 48 Marchegay 1999, 2008, p. 99. 46 On tomb 1068 see also Salles 1987, pp. 158-173. 49 Marchegay 2008, p. 99.

West & East 165 Monografie, 4 Margherita Dallai

Figure 28 Tomb’s external entrance at Ugarit (after Teba 2015, p. 269)

which were probably used to place lamps or offer- 5. Concluding remarks ings.50 The extension of thedromos over time confirms the intent to make access to the tomb easier: at Based on the analysis of the different building mate- the beginning, this passage was very short, but it was rials, the techniques of implementation and the re- later changed into a longer passage and equipped lationship between funerary hypogea and the build- with a staircase.51 On account of the analysis of the ings located above them, as well as the comparison house-tomb complex, it is possible to argue that the of structures found at Ugarit, it is possible to make funerary structure was planned before the house some observations. was built. These tombs were placed near the access As for the materials employed in the realization and almost always involved independent access on of the vaults, the stone is certainly the least used the street to facilitate entry to people who most one. The brick, both baked and unbaked, is the most probably did not live in the house itself (fig. 28).52 common material, although the use of the former is In many dwellings, the tomb was located beneath limited because of the high cost associated with its two rooms: one was above the dromos, while the realization, that is due to the cooking process. other was above the funerary chamber; the former Regarding the evolution of the vault, on the ba- had a specific funerary role since the tomb was ac- sis of the data analysed to date, I can support the cessed from here, while the second most likely had a hypothesis suggested by Valentini about the exis- domestic function. Marchegay supposed that these tence of two distinct technological traditions.54 In spaces above the tomb could also be used for funer- fact, the dome-shaped roofing may have developed ary ceremonies, especially for the funerary feast in in Northern Mesopotamia, as well as a corbelled honour of the dead.53 vault technique, as documented in this area until later periods. Instead, the technique for radial rows and the barrel vault probably developed in South- ern Mesopotamia, inspired by the structures made with bundles of reeds. Without distinguishing between the different implementation techniques, 50 Teba 2015, p. 268. another hypothesis is supported by scholars such 51 Marchegay 2008, p. 102; Teba 2015, p. 267. as Besenval, Castel and Sauvage, who believe that 52 Marchegay 2008, p. 107. 53 Marchegay 2008, p. 107. 54 Valentini 2002, p. 9.

West & East 166 Monografie, 4 The vaulted funeraryhypogea in Mesopotamia between the second and first millennium BC: localization and architectural features

the vault technology belonged to a typical North- Finally, the most important difference concerns ern Mesopotamia tradition.55 Their thesis is also access to the tomb and its location inside the house. based on the fact that, in Northern Mesopotamia, In fact, while the hypogea at Ugarit were built in the vault is attested in private dwellings more fre- areas close to the building entrance and provided quently than in the south, suggesting that the tech- with direct access from outside, the tombs in Mes- nique was only mastered thanks to a long tradition opotamia and Khuzestan were often located as far of passing on relevant skills in this area.56 as possible from the entrance of the building, in ar- About the context in which the hypogea are eas that were difficult to access. Unfortunately, at found, I tried to analyse the data relating to their lo- present, we do not have enough elements to con- cation, access and plan in order to determine wheth- vincingly explain this difference. In the case of the er a chronological development could be outlined. Ugarit tombs, as Marchegay suggests, it can be as- Unfortunately, from the information collected, it is sumed that an independent entrance on the street not possible to identify any diachronic development could have facilitated access to people who were for any of the analysed elements. As for the planim- external to the house; for example, it was hypothe- etry of the tombs, it was not possible to hypothesize sized that this measure could have facilitated entry any rationale regarding the numerical variability of for relatives to the family tomb after the sale of the the rooms. house itself.58 Another hypothesis is related to the On the basis of comparison between the ex- observation that not all dwellings housed a hypoge- amined sites and the site of Ugarit, it was possible um, which means that more families possibly shared to highlight the existence of some similarities, but a hypogeum, but not the house above it. In this sense, also substantial differences between thehypogea on an independent entrance would have also made the these sites. At all these sites, most of the hypogea were tomb accessible to those who did not live in that found under private dwellings, although they were building.59 On the contrary, with regard to hypogea sometimes located under royal palaces. The plan of analysed in my study, the purpose of their location the funerary chamber was generally rectangular or seems to have been to hinder access to the tombs to quadrangular in shape, although its size was much external people, probably in order to keep the pub- larger (sometimes it exceeded ten square meters) in lic area of the house separate from the area that was the tombs of Ugarit than in those of Mesopotamia more connected to the afterworld. This concept is and Khuzestan. At all sites, the chamber had one or very clear in the case of the funerary buildings at more niches in the walls, although, in Ugarit, some Haft Tepe and Tchoga Zanbil, which have been of them connected the chamber with a space that previously described. was otherwise inaccessible; we do not know what its Finally, by analysing the relationship between function was, but Marchegay suggests that it could the hypogea and the overhead building, it is possible have served to host the ancestors.57 Thehypogea at to determine that, at Ugarit, the link between the Ugarit were all built with stone and covered by a tomb and the house was very strong because both corbelled vault, unlike those at other sites where dif- were planned and realized at the same time. Mean- ferent materials and techniques were used. General- while, at the other sites, despite their construction ly, the hypogea were characterized by a single room, mostly taking place contemporaneously, some hypo- while, in Mesopotamia and Khuzestan, there were gea were realized before or after the construction of sometimes two or, more rarely, three. the building that housed them.

55 Besenval 1984, pp. 160-161; Castel 1996, p. 275; Sauvage 1998, pp. 157-158. 56 Valentini 2002, p. 15. 58 Marchegay 2008, p. 107. 57 Marchegay 2008, p. 128. 59 Marchegay 2008, pp. 106-107.

West & East 167 Monografie, 4 Margherita Dallai

BIBLIOGRAPHY

Bachelot L. 1992, Une tombe construite du deuxième Lenzen H. 1963, Ausgrabungen in Uruk-Warka, Berlin. millénaire Av. J. C., à Mohammed Diyab, in: Durand Lenzen H. 1968, Ausgrabungen in Uruk-Warka, Berlin. J.-M. (ed.), Recherche en Haute Mésopotamie. Tell Mallowan M. 1966, , London. Mohammed Diyab. Campagnes 1990 et 1991 Nimrud and Its Remains (Mémoires de «NABU») Paris, pp. 31-34. Mallowan M., Cruikschank Rose J. 1935, , «Iraq 2, Besenval R. 1984, Technologie de la voûte dans l’Orient Excavation at Tall Arpachiyah 1933 » pp. 1-178. ancien 1-2. Des origines à l’époque sasanide, Paris. Marchegay S. 2008, Castel C. 1996, Un quartier de maison urbaines du Bronze Les pratiques funéraires à Ougarit , Moyen à Tell Mohammed Diyab (Djezireh Syrienne), in: au IIe millénaire. Bilan et perspectives des recherches in: Calvet Y., Yon M. (eds.), Veenhof K. R. (ed.), Houses and Households (40eRen- Ougarit au Bronze contre Assyriologique Internationale, Leiden 1993), moyen et au Bronze récent (Actes du colloque inter- Leiden/Istanbul, pp. 273-283. national, Lyon, novembre 2001), Lyon, pp. 97-118. Marchegay S. 1999, Ghirshman R. 1959, Tchoga Zanbil. Rapport Les tombes d’Ougarit. Architecture, , (Unpublished Préliminaire de la VIIe Campagne (1958-1959), localisation et relation avec l’habitat PhD Thesis, Université Lumière Lyon 2), Lyon. «Arts Asiatiques» 6/4, pp. 259-282. Margueron J.-C. 1983, Ghirshman R. 1965a, L’architecture élamite et ses tra- Quelques reflexions sur certai- , «Akkadica 32, ditions, «Iranica Antiqua» 5/2, pp. 93-102. nes pratiques funeraires d’Ugarit » pp. 5-31. Ghirshman R. 1965b, Suse du temps des Sukkalmah. Merpert N., Munchaev R., Bader N. 1976, Campagne de fouilles 1963-64, «Arts Asiatiques» The 11/2, pp. 3-22. Investigation of Soviet Expedition in Iraq, 1973, «Sumer XXXII», pp. 25-61. Ghirshman R. 1967, Suse, campagne de l’hiver 1965- Mofidi-Nasrabadi B. 2012, 1966. Rapport préliminaire, «Arts Asiatiques» 15, The Spatial Order in the pp. 3-28. Tomb Buildings of the Middle Elamite Period, in: Oates J., Oates D. 2001, Nimrud. An Assyrian Ghirshman R. 1968a, Suse au tournant du IIIe et du Imperial City Revealed, London. IIe millénaire avant notre ère, «Arts Asiatiques» 17, pp. 3-44. Negahban E.O. 1991, Excavation at Haft Tepe, Iran, Philadelphia. Ghirshman R. 1968b, Tchogha Zanbil (Dur Untash). Pedde F. 2015, Vol. II. Temenos, temples, palais, tombes, Paris. Gräber und Grüfte in Assur II. Die mittelassyrische Zeit, Wiesbaden. Ghirshman R., Steve M. 1966. Suse. Campagne Peltenburg E. 2007/2008, de l’hiver 1964-65. Rapport préliminaire, «Arts Enclosing the ancestors Asiatiques» 13, pp. 3-32. and the growth of socio-political complexity in Early Bronze Age Syria, in: Bartoloni G., Benedettini Delougaz P., Hill H., Lloyd S. 1967, Private houses M. (eds.), Sepolti tra i vivi. Buried among the Living. and graves in the Diyala region, Chicago. Evidenza e interpretazione di contesti funerari in abi- Haller A. 1954, Gräber und Grüfte in Assur, Berlin. tato (Atti del convegno internazionale, Roma, 26-29 Hockmann D. 2010, Gräber und Grüfte in Assur aprile 2006), Roma, pp. 215-247. I. Von der zweiten Hälfte des 3. bis zur Mitte des 2. Pfälzner P., Niehr H., Pernicka E., Wissing Jahrtausends v. Chr., Wiesbaden. A. (eds.), (Re-) Constructing funerary rituals in the Jean Marie M. 1990, Les tombeaux en pierres at Mari, Ancient Near East.Proceedings of the first internatio- «Mari. Annales de Recherches Interdisciplinaires» nal symposium of the Tübingen Post-Graduate School 6, pp. 303-336. “Symbols of the Dead” in May 2009, Wiesbaden, pp. 261-270. Lenzen H. 1961, Ausgrabungen in Uruk-Warka, Berlin. Reuther O. 1926, Die Innerstadt von Babylon (Merkes), Lenzen H. 1962, Ausgrabungen in Uruk-Warka, Berlin. Leipzig.

West & East 168 Monografie, 4 The vaulted funeraryhypogea in Mesopotamia between the second and first millennium BC: localization and architectural features

Salles J.-F. 1987, Deux nouvelles tombes de Ras Shamra, volta, in riferimento agli esempi di Tell Barri/Kahat in: Yon M. (ed.), Le Centre de la Ville, 38e‑44e cam- (Siria), «Bollettino degli ingegneri» 1/2, pp. 6-19. pagnes (1978-1984), Paris, pp. 157‑195. Valentini S. 2003, Le pratiche e l’ideologia funeraria a Sauvage M. 1998, La brique et sa mise en oeuvre en Tell Barri/Kahat durante il Bronzo Medio, in relazio- Mésopotamie. Des origines à l’époque achéménide, ne all’area siro-mesopotamica settentrionale, «Studi Paris. Micenei ed Egeo-Anatolici» 45/2, pp. 273-305. Spanos P. 1987, Nordost-Abschnitt I 1983 (7.‑8. Valentini S. 2016, Vaulted hypogea during the Middle Kampagne), in: Hrouda B. (ed.), Isin – Išān Bahrīyāṭ Bronze Age: a perfect example of the intra-muros III. Die Ergebnisse der Ausgrabungen 1983-1984, multiple tomb in Mesopotamia, in: Felli C. (ed.), München, pp. 17-21. How to Cope with Death: Mourning and Funerary Teba T. 2015, Practices in the Ancient Near East (Proceedings of the Keeping the Fathers Alive: The Conservation th th of Funeral Architecture in Ugarit, in: Amoêda International Workshop Firenze, 5 – 6 December R., Lira S., Pinheiro C. (eds.), REHAB 2015 2013), Firenze, pp. 217-240. (Proceedings of the 2nd International Conference Van Beek G. W. 1987, Archi e volte nell’antico Vicino on Preservation, Maintenance and Rehabilitation Oriente, «Le Scienze» 229, pp. 82-90. of Historical Buildings and Structures, Porto, 22-24 Woolley L. 1934, Ur excavations. Vol. II. The royal ce- July), Barcelona, pp. 263-278. metery: a report on the predynastic and Sargonid graves Valentini S. 2001, Alcune considerazioni sugli ipogei pa- excavated between 1926 and 1931, London. leobabilonesi rinvenuti a Tell Barri/Kahat, «Orient Woolley L., Mallowan M. 1976, Ur excavations, Express» 3, pp. 82-85. Vol. VII. The Old Babylonian Period, London. Valentini S. 2002, Gli ipogei funerari in Mesopotamia: Yon M. 1997, La cité d’Ougarit sur le tell de Ras Shamra, aspetti di tecnologia ed evoluzione della copertura a Paris.

West & East 169 Monografie, 4

Rotary querns from Tell Barri (Syria): chronology, use and function

MAURIZIO FASCITIELLO

Independent Researcher

Abstract

The rotary quern is one of the most advanced lever mill in use in the ancient world. Every mill of this type is composed of two discoidal querns, the upper one movable and the lower one stationary, usually made of basalt and functioning by friction. Over the last few decades, the research on this field has been especially focused on the western Mediterranean where the two querns are apparently always separated by a system of supports (rynd and spindle) to adjust the gap between them and produce fine or course flour. The site of Tell Barri, in the Khabur Valley (Syria), has revealed 21 rotary mills that suggest a new field of studies. They come from different contexts dating between the second and thirteenth c. AD. This paper analyses the different types and operational processes to argue for their use and determine a possible connection with the ones found in the Western Mediterranean.

Keywords

Mills, rotary mills, Tell Barri, querns, milling technology, grinding, grindstones

West & East 171 Monografie, 4 Maurizio Fascitiello

1. Introduction The rotary mill is composed of a stationary lower quern and a movable upper one, both circular, What kind of civilisation would the humans achieve usually made of basalt and functioning by friction without the introduction of the milling process? (fig. 1).4 Due to its importance in the human life, its in- The excavations at Tell Barri has produced 21 troduction was attributed to the gods: Pliny consid- rotary mills which date between the second and ered Ceres the responsible of teaching the agricul- thirteenth c. AD.5 ture and the grinding to the human beings, that is the reason why She was considered a goddess.1 The milling technology follows the evolution of the hu- 2. The Tenon and Mortise type man progression since the first domestication of the cereals: Hand mills, Assyrian mills, Olynthus The first type presented is the so called Tenon and querns and Rotary Mills are the main types of tools Mortise type.6 The upper movable quern has a cir- identified for the milling process.2 cular perforation on its centre (the mortise) while This paper is only focused on the rotary mills the lower stone distinguishes by a protrusion at the found in the site of Tell Barri, in the Khabur Valley centre of one of its faces (the tenon). The perfora- (in north-eastern Syria), to analyse the features of tion and the protrusion perfectly match in order to the tools and argue for their use.3 avoid the detachment of the tenon from the mortise during the rotation; this made the circular movement stable and increased the speediness of the system (fig. 2). Nine querns pertaining to this type have been found at Tell Barri: four upper (E.707, 3606, 4585, 6649) (fig. 3) and five lower (912, 2885, 6334, 6336, 6817) (fig. 4). They date between the end of the second c. AD and the beginning of the fifth c. AD.7 The mortise also functioned as a hopper to contain the grain that gradually slid on the grinding surface due to the movement: the upper quern E.707 is characterized by a raised circular collar set around the hopper in order to hold a larger quantity of Figure 1 grain in it. Fascitiello M. (manner of), Reconstruction The tenon and mortise mill could be easily man- of a rotary quern (after Amouric 1996, fig. 1a) aged by a single operator: once the mortise was filled with the grain, the rotation was applied by the use of

4 An exception is witnessed by the mills in limestone from the Aisne Valley, at Vendresse-Beaulne (north-eastern France): Naze et Al. 2011, pp. 269-283. 1 Nat.Hist. VII 191. 5 All the other types of milling tools found in the same 2 For the Hand mills see: Bombardieri 2010, site as mortars, hand mills, saddle querns, Assyrian and Olyn- pp. 33‑77. For the Assyrian Mills: Bombardieri 2005, thus querns have been studied and published by L. Bombar- pp. 157-161, 167-171; Bombardieri 2010, pp. 78-85. For the dieri: Bombardieri 2005, 2010. Olynthus querns: Frankel 2003, pp. 1-21; Bombardieri 6 2010, pp. 85-92. For the Rotary Mills: Buchsenschutz et For this mill type and the others of this paper, see the Al. 2011; Peacock D. 2013. catalogue below for details on the measures and features of the devises. 3 The excavation works, led between the years 1980- 7 2010, were directed by Prof. P. E. Pecorella (University of Flor- Thequerns 707 and 6817 were found in pits; 912 on ence) first and continued by Prof. R. Pierobon Benoit (Univer- the ground level; 3606 and 4585 in deposit levels; 6334, 6336 sity of Naples). For a summary of the site: Pecorella, Salvi- and 6649 were found on the floor but notin situ; 2285 was ni 1982; Pecorella 1998; Pierobon Benoit 2008. found in situ.

West & East 172 Monografie, 4 Rotary querns from Tell Barri (Syria): chronology, use and function

Figure 2 Fascitiello M., Intersection between an upper and a lower quern of the Tenon and Mortise type. Basalt. Max diam. 42cm. Late 2nd c. AD – early 3rd c. AD. Found at Tell Barri

Figure 3 Fascitiello M., Upper querns belonging to the Tenon and Mortise type. Basalt. Diam. between 36 and 44cm. Late 2nd c. AD – early 5th c. AD. Found at Tell Barri

West & East 173 Monografie, 4 Maurizio Fascitiello

Figure 4 Fascitiello M., Lower querns belonging to the Tenon and Mortise type. Basalt. Diam. between 35 and 44 cm. Late 2nd c. AD – early 5th c. AD. Found at Tell Barri

Figure 5 Fascitiello M., Back side of the upper quern E. 4585. Basalt. Max diam. 42 cm. Late 2nd c. AD. Found at Tell Barri

West & East 174 Monografie, 4 Rotary querns from Tell Barri (Syria): chronology, use and function

a handle. No trace of it was found during the exca- bricks platform so that the mill had more stability vations; they were probably made of wood or metal. during the movement. The hole where the handle was inserted is still visible Similar tools have been found in the Near East on the edge of every tool of this type: it has a circu- in contexts dating between the Chalcolithic Age lar shape and a double-cone section. On two of the and the seventh/eighth c. AD and beyond.8 Some upper querns (E.4585, 6649) the hole for the han- of the researchers on this field have doubted on the dle is located beyond the circular edge of the mill, function of these instruments as rotary mills for the on an extension that made the rotation easier to be Prehistoric contexts, suggesting their use as potter’s operated; on the grinding surface, the extension is wheels, instead.9 slightly raised, fitting with the edge of the stationary The instruments of this type from Tell Barri are quern, in order to give more stability and speed to surely rotary querns. In fact, the mills E.6334 and the rotation (fig. 5). The mortise E.6649 is featured E.6336 come from the same sector of a domestic by two holes set along the edge of the hopper. They complex which dates to the second half of the sec- were filled with two pieces of iron (one still visible) ond c. AD and surely used for grinding and storing that probably are the remains of a support to give food.10 Even the hole of the handle, visible on differ- more stability to the movement. ent tools suggests the same function, as the raised The grinding surface of the mortises is smooth circular collar set around the hopper of the E.707 to but the top is coarse and irregular, except for the hold a larger quantity of grain in it. mortise E.707 which has a regular upper surface em- Only future studies will clarify if these instru- bellished with a triangular decoration between the ments besides the ones from Tell Barri are rotary hole for the handle and the raised collar around the mills or potter’s wheels. hopper. The height of the moving upper quern ranges from 7,5 and 10 cm while the diameter is between 2.1. A possible variation of the Tenon 36 and 44 cm. The mortise can have a cylindrical sec- and Mortise quern tion (E.3606), a double-cone section (E.4585) or a truncated cone section (E.707); it is circular at the Different proposals have been made for the six up- base while it can be circular (E.707) or almost circular per querns E.438, 915, 1228, 1538, 6087, 6372 that (E.3606) at the top; the inner surface is smooth due to seem to have similarities with the ones just present- the friction with the tenon during the rotation. The ed (fig. 6).11 diameter of the hopper falls between 11 and 13 cm. Their shape brings to mind the upper querns of Regarding the stationary mill, the height is be- the Tenon and Mortise type but the size is different: tween 8 and 11,6 cm while the diameter between the height is between 3 and 6,7 cm, the mortise has 35 and 44 cm. It can have a truncated cone section (E.912, E.6336) or a double cone section (E.6334); 8 Catling 1972, p. 79, fig. 43;Moorey 1994, p. 146; the grinding surface is smooth due to the friction Bombardieri 2004, p. 97. with the mortise; the height of the tenon is between 9 For the supporters of the interpretation of these instruments as potter’s wheels see: Amiran 1956, pp. 46-49; Chil- 3 and 7,5 cm. de 1961, pp. 200-203; Trokay 1989, pp. 169-175; Bombar- A question on the functioning of this mill is nec- dieri 2004, pp. 96-99. For the supporters of the interpreta- essary: where was this tool located during the grind- tion as rotary hand mills see: Pruss 1994, pp. 205-206, fig. 63; Mecalister 1912, pp. 36-37, fig. 228;Loud 1948, p. 5, plate ing? On the floor, on a table or on a platform? 268:2. The lower surface of the stationary quern is al- 10 Palermo 2012, pp. 637-647. The food production is ways coarse and irregular. This is the reason why it clarified by the finding of a saddle quern in original position, couldn’t be located on a table or on the floor, oth- kitchen ware and two ovens for bread; the storing of food is suggested by the presence of 5 pithoi. erwise it wouldn’t have had enough stability to be 11 The querns 438 and 1228 were found in pits; 6087 in managed. Studying the features of the lower querns, deposit level; 915 was found on the floor but notin situ; 1538 they are supposed to be installed into a low mud- and 6372 come from an unknown context.

West & East 175 Monografie, 4 Maurizio Fascitiello

Figure 6 Fascitiello M., Upper querns belonging to a supposed variation of the Tenon and Mortise type. Basalt. Diam. between 22 and 50 cm. 11th/12th c. AD. Found at Tell Barri

a diameter between 3 and 5 cm while the diameter The dating of this type found at Tell Barri is is between 22 and 50 cm; in other words, they are about the eleventh/twelfth c. AD that can be con- thinner and wider. Some of the tools are featured by nected to the Islamic presence on the site. Similar a hole for a handle on the edge and furrows on the tools have been found at Gezev (600-1100 d.C.) grinding surface (fig. 7). and Ramat Hanadiv (dating between the Byzantine These instruments have been associated to a and the Islamic period).13 lower stationary quern with a slim tenon from an A few questions for future studies are neces- unknown context. The shape of the tenon matches sary: can this mill type be considered a variation of with the mortise of the upper mills and even though the tenon and mortise quern? If so, when did this the chronology of this tool is uncertain, their asso- change take place? ciation could be considered possible (fig. 8).12

12 The measures of this devise are the following: Diam. 13 Gezev: Macalister 1912, fig. 38b. Ramat Hanadiv: tenon at the base is 5 cm, at the top 2,9 cm (Max diam. 40 cm). Sidi 2000, p. 178, pl. 1, 7.

West & East 176 Monografie, 4 Rotary querns from Tell Barri (Syria): chronology, use and function

Figure 7 Figure 8 Fascitiello M., Back side of the upper quern E. 915. Fascitiello M., Lower quern with thin tenon Basalt. Max diam. 39 cm. Chronology unknown, from unknown context. Basalt. Max diam. 40. probably 11th/12th c. AD. Found at Tell Barri Found at Tell Barri

3. The device E. 1041 This mill type is the most common rotary hand mill in use in the Mediterranean Basin.16 Appar- A different mill type found at Tell Barri is repre- ently it was introduced in south-east Spain be- sented by the device E.1041 (fig. 9). tween the end of the sixth and the beginning of It has a deep elliptical hopper, a hole for the the fifth c. BC It seems this device first spread in handle on the edge and its diameter is 37 cm. The central and northern Spain (fourth c. BC) then, grinding surface is featured by two regular sockets quickly, from west to east.17 In southern France (3 × 2,5 cm; 1,8 cm deep) which housed the rynd. the oldest tools date back to the fourth-third c. This is part of the system of supports to sustain the BC; no hand-mills were found before the third c. weight of the upper stone and adjust the gap between BC in Sicily, while in northern Italy they appeared the upper and the lower mill to produce fine or course in the second-first c. BC18. It has been shown that flour. The rynd joins a vertical pivot (the spindle) -in 16 Recent studies have revealed local variations and dif- serted in a hole located at the centre of the lower mill ferent rotary hand mill types: Quesada Sanz et Al. 2014, (fig. 10).14 The device E.1041 comes from a deposit pp. 83-118. level excavated in a context of the sixth c. AD.15 17 Alonso 1996, pp. 187, 191. 18 For the devises found in southern France: Py 1992, 14 Amouric 1996, pp. 39-40; Robin, Boyer, 2011, p. 195; Alonso 1996, pp. 190-191; Reille 2000, pp. 263- pp. 351-358. 265, 269-270. For Sicily: Williams-Thorpe 1988, p. 260. 15 Pecorella 1993, pp. 21-23. For northern Italy: Donner, Marzoli 1994, p. 85.

West & East 177 Monografie, 4 Maurizio Fascitiello

Figure 9 Fascitiello M., Upper quern E. 1041. Basalt. Max diam. 37 cm. 6th c. AD. Found at Tell Barri

Figure 10 Fascitiello M. (manner of), Functioning of the mills with rynd and spindle (after Runnels 1990, fig. 1)

this technology arrived in Greece in the first c. BC second c. BC, followed the Roman armies in their but it was unknown in the Near East before the conquests eastwards. first c. AD.19 Since we know that the rotary mill Returning to the mill found in Tell Barri, how was a common feature of Roman army camps, it did it come from so far? seams that this device, at least starting from the Giving an answer to this question is very complicated at the moment. Only future researches could eventually give us a better idea of its possible spread 19 For the devises found in Greece: Runnels 1990, p. 153. Near East: Netzev 1991, pp. 232-235. in the Near East.

West & East 178 Monografie, 4 Rotary querns from Tell Barri (Syria): chronology, use and function

4. The quern type with “grain container”

The last mill type presented is probably the most controversial one: it is featured by a grain container carved into the lower grinding surface (fig. 11). It is possible to distinguish two variations. The upper moving querns E.2974 and E.5914 are characterized by the lack of hopper and the presence of a handle at the centre of the upper surface (fig. 12). This suggests that the upper mill had to be lifted to feed the devise: probably the grain had to be set at the centre of the stationary quern to be ground. The position of the handle implies two possibilities for the movement: either a “T” shaped handle for a semi-rotary motion or a “L” shaped handle for a rotary or alternative motion.20 Considering the features of this mill type, personally, a “T” shaped handle would probably be easier to be operated. On the other hand, the querns E.3602 and E.6226 (fig. 13) differ from the previous ones for the presence of a narrow hopper (diameter 3,7 and 6 cm) at the centre and the lack of any handle.

Figure 11 Figure 12 Fascitiello M., Back side of the upper quern with grain Fascitiello M., Two upper querns with grain container container E. 6226. Basalt. Max diam. 32,5 cm featured by a handle in their center. Max diam. 33 cm. 3rd – 5th c. AD. Found at Tell Barri Basalt. 2nd and 6th c. AD. Found at Tell Barri

20 For alternative motion it is meant an angle between 180° and 270°.

West & East 179 Monografie, 4 Maurizio Fascitiello

Figure 13 Fascitiello M., Two upper querns with grain container featured by a hopper in their center. Basalt. Max diam. 36,5 cm. 2nd and 6th c. AD. Found at Tell Barri

The presence of a hopper means that the operator did not need to lift the upper quern to feed the system and the lack of handle makes me think that the instrument described a semi-rotary motion applied by the operator’s hands placed on the opposite sides of the tool.21 All the querns of this type have similar proportions (the diameter is between 32 and 36,5 cm) and come from deposit levels dating between the second and sixth c. AD . At this point the question is: how was the grinding lower quern? From my research on the rotary hand mills from Tell Barri, I have noticed that the tool E.6428 (fig. 14) would perfectly match with the upper mills just presented.22 It was found in its original location (fig. 15), in a kitchen of a domestic complex which dates between the end of the second and the beginning of

21 Figure 14 No traces of grooves along the edge of these tools were Fascitiello M., Lower quern E. 6428; it might be the found to suppose the presence of any cords used to support a handle. stationary stone of the upper querns with grain container. 22 Basalt. Late 2nd – early 3rd c. AD. Fount at Tell Barri This association is suggested by the diameter and the inclination of the grinding surface, the same as the upper mills.

West & East 180 Monografie, 4 Rotary querns from Tell Barri (Syria): chronology, use and function

Figure 15 Fascitiello M., Lower quern E. 6428 in its original context. Basalt. Diam. 33 cm. Late 2nd – early 3rd c. AD. Found at Tell Barri

the third c. AD.23 This stationary mill is part of a nection between the upper mills with grain con- L-shaped platform made of four fired-bricks; this is tainer and the lower stationary mill E.6428 would a multifunctional working area used for preparing be more certain. Unfortunately, at this point of and cooking food (traces of a fire were found). the research, this association has been suggested by The place where the stationary quern E.6428 the general features of the tools but, unfortunately, was found, along two of the walls of the kitchen, none among the upper mills with grain container in the corner of the room, may suggest that the op- comes from the same archaeological context as the erator just moved the upper mill by following an stationary mill E.6428. Regarding the place where alternative motion24. If this assumption would be rotary querns were located during the grinding pro- confirmed by other findings, the hypothetical con- cess, the position of this mill on a low platform set on the floor shows that the operator had to work 23 This context is presented here for the first time. knelling. This finding does not exclude that other 24 The relation between the rotary or alternative motion mill types couldn’t be positioned on higher plat- and the mill’s location was approached for the first time by M. forms, supports or tables so that the operator could Py, studying the mills from Lattes: PY 1992, p. 225. The same work in a standing position. subject has been analyzed in Chartrain 2015, p. 450, fig. 13.

West & East 181 Monografie, 4 Maurizio Fascitiello

5. The catalogue of the rotary mills from one is flat and featured by furrows, in order to im- Tell Barri prove the grinding. The hopper has a double-cone section. Made of basalt. Partially preserved. Chro- nology unknown. E. 438 Measures: 39×19 cm; Ø 39 cm; H 6; Ø hopper Upper moving quern of a tenon and mortise mill. about 3 cm. The upper surface is irregular and coarse. The grinding lower surface is flat and smooth. The hopper has E. 1041 a cylindrical section and a circular shape. The hole Upper moving quern of a rotary mill. The upper sur- which held the handle has a truncated cone section face is regular. The hopper is oval-shaped. The hole and a circular shape; it is located on the edge. Made for the handle is preserved on its edge; it has a cir- of basalt. Intact. Chronology unknown. cular shape and a truncated cone section. The lower Measures: 35×25 cm; H. 4,6 cm; Ø 37 cm; Ø hop- grinding surface is flat and smooth; it is character- per 3 cm; Ø hole of the handle at the top 2 cm; Ø ised by two regular sockets (3×2,5 cm; 1,8 cm deep) hole of the handle at the bottom 3,5 cm. where a piece of iron (the rynd) was inserted to support the upper quern. The rynd is joined to a spindle E. 707 set in the lower quern: this equipment is essential Upper moving quern of a tenon and mortise mill. for the stability and adjustment of the rotary mill. Circular shape with a rectangular extension where Made of basalt. Intact. 6th c. AD. the hole for the handle is located; this hole has a Measures: Ø 37 cm; H 6,4 cm; hopper 11,5×15 cm; cylindrical section and is oval-shaped. Regular up- Ø hole for the handle at the top 3,5 cm, at the bot- per surface embellished by a triangular decoration tom 2,2 cm. between the hole for the handle and the raised collar around the hopper. The hopper has a truncated E. 1228 cone section and a circular shape. The grinding low- Upper moving quern of a tenon and mortise mill. er surface is flat and featured by furrows in order to The upper surface is somewhat irregular and coarse; improve the grinding. Made of basalt. Intact. Chro- the lower one is flat and smooth. The hole for the nology unknown. handle has a circular shape and a double cone sec- Measures: Ø 42 cm; H 7,5 cm; hole for the handle tion. Made of basalt. Partially preserved. Chronol- 4,5×3,5 cm; Ø hopper at the bottom 14 cm; Ø hop- ogy unknown. per at the top 11 cm. Measures: 16×14,5 cm; H 6 cm; Ø about 22 cm; Ø hole for the handle at the top 3,5 cm; Ø hole for the E. 912 handle at the base 2,8 cm. Lower stationary quern of a tenon and mortise mill. The grinding surface is regular and partially smooth. E. 1538 The tenon has a truncated cone section and it is Upper moving quern of a tenon and mortise mill. smooth due to the rotary motion. The lower sur- The upper surface is regular and partially smooth. face is course. Made of basalt. Intact. Chronology The hopper has a truncated cone section and a cir- unknown. cular shape. The grinding surface is flat and smooth. Measures: Ø 35 cm; H 6 cm; H tenon 7 cm; Ø ten- Made of basalt. Partially preserved. Chronology un- on at the top 9 cm; Ø tenon at the bottom 10 cm. known. Measures: 21×26 cm; H 6,7 cm; Ø 50 cm; Ø hop- E. 915 per 5 cm. Upper moving quern of a tenon and mortise mill. The upper surface is irregular and coarse; the lower

West & East 182 Monografie, 4 Rotary querns from Tell Barri (Syria): chronology, use and function

E. 2885 E. 4585 Lower stationary quern of a tenon and mortise mill. Upper moving quern of a tenon and mortise mill. The upper surface is regular and smooth; the lower The upper surface is course; the hopper has a dou- one is irregular and course. The tenon is thin and ble cone section. The grinding lower surface is flat has a truncated cone section. Made of basalt. Intact. and smooth. The hole for the handle has a double 3rd c. AD. cone section; it is located over the circular edge of Measures: Ø 38 cm; H 5,2 cm; Ø tenon at the top the mill, on an extension. On the grinding surface, 7,5 cm; Ø tenon at the base 11,5 cm; H tenon 2 cm. the extension is slightly raised, fitting with the edge of the stationary quern in order to give stability to E. 2974 the rotation. Made of basalt. Partially preserved. Late 2nd c. AD. Upper moving quern with grain container. The up- Measures: 41×30 cm; H 10 cm; Ø 42 cm; Ø hopper per surface is flat and somewhat smooth. At the cen- 13 cm; Ø hole for the handle at the top 4 cm, at the tre of the tool a fragmentary iron handle is still visible. bottom 3 cm. Lower grinding surface, regular and smooth, which surrounds an inner circular space used to contain E. 5914 grain during the grinding. Made of basalt. Broken in three pieces; reassembled. 2nd – 3rd c. AD. Upper moving quern with grain container. The up- Measures: Ø 33 cm; H 4,5 cm; Ø handle 3,5 cm; Ø per surface is regular and somewhat smooth. At inner circular space 19,5 cm; depth inner circular the centre of the tool a fragmentary iron handle is space 1,3 cm; width grinding surface 6,8 cm. still visible. The lower grinding surface is flat and smooth; it surrounds an inner circular space used to E. 3602 contain grain during the grinding. Made of basalt. Partially preserved. 5th – 6th c. AD. Upper moving quern with grain container. The up- Measures: 32×17,5 cm; Ø 32 cm; H 6 cm; Ø han- per surface is regular and coarse. The hopper has a dle 4 cm; Ø inner space 24 cm; depth inner circular double cone section, a circular shape and is not ex- space 1,6 cm; width grinding surface 7 cm. actly set at the centre of the tool. The lower surface is coarse and not well preserved. No handle. Made E. 6087 of basalt. Broken in four pieces; reassembled. Late 5th c. AD. Upper moving quern of a tenon and mortise mill. Measures: Ø 36,5 cm; H 7 cm; Ø hopper at the top The upper surface is regular and somewhat smooth. 6 cm, at the bottom 5,5×4,5 cm. The hopper has a truncated cone section. The grinding surface is flat and smooth. Made of basalt. Par- E. 3606 tially preserved. 12th – 13th c. AD. Measures: 18×29 cm; H 3,7 cm; Ø 38 cm; Ø han- Upper moving quern of a tenon and mortise mill. dle 6 cm. The upper surface is irregular and coarse. The hopper has a cylindrical section, a circular shape at the E. 6226 bottom, an almost circular shape at the top. The hole for the handle has a double cone section. The Upper moving quern with grain container. The up- lower surface is flat and smooth. Made of basalt. In- per surface is regular and slightly rounded. The hop- tact. Late 4th – early 5th c. AD. per has a circular shape and a double cone section. Measures: Ø 41,5 cm; H between 9 and 7 cm; Ø The lower surface, regular and smooth, surrounds hopper at the top 12,5×10 cm; Ø hopper at the bot- an inner circular space used to contain grain during tom 11 cm; Ø hole for the handle 5 cm. the grinding. Made of basalt. Intact. 3rd – 5th c. AD.

West & East 183 Monografie, 4 Maurizio Fascitiello

Measures: Ø 32,5 cm; H 5 cm; Ø hopper 3,7 cm; E. 6649 Ø inner circular space 16,5; depth of the inner circular space 1,6 cm; width grinding surface 8 cm. Upper moving quern of a tenon and mortise mill. The upper surface is somewhat irregular. The hop- E. 6334 per has a cylindrical section, circular shape at the bottom, almost circular at the top. Along the edge Lower stationary quern part of a tenon and mortise of the hopper there are two holes, filled in with two mill. The grinding upper surface is flat and smooth. pieces of iron (one still visible), used for holding the The tenon has a double cone section; it is smooth lower and the moving mill together, probably with a because of the friction with the hopper of the up- system of cords. The hole for the handle has a trun- per quern. The lower surface is coarse and rounded. cated cone section; it is located on an extension. On Made of basalt. Intact. Late 2nd – early 3rd c. AD. the grinding surface, the extension is slightly raised, Measures: Ø 44 cm; H 5 cm; Ø tenon at the top fitting with the edge of the stationary quern in order 8,3 cm; Ø tenon at the bottom 11 cm; H tenon to give stability to the rotation. The lower surface is 7,5 cm. flat and smooth. Made of basalt. Intact. Late 2nd – early 3rd c. AD. E. 6336 Measures: Ø 36 cm; length 45 cm; H 8 cm; hopper 11×9,5 cm; Ø hole for the handle 4 cm. Lower stationary quern part of a tenon and mortise mill. The grinding upper surface is flat and smooth. E. 6817 The tenon has a truncated cone section; it is smooth because of the friction with the hopper of the upper Lower stationary quern of a tenon and mortise mill. quern. The lower surface is coarse. Made of basalt. The grinding surface is regular but coarse. The tenon Intact. Late 2nd – early 3rd c. AD. is partially preserved. The lower surface is coarse and Measures: Ø 39 cm; H 11,6 cm; Ø tenon at the top rounded. Made of basalt. Partially preserved. Chro- 6,5 cm; Ø tenon at the bottom 11 cm; H tenon nology unknown. 6,5 cm. Measures: 40×20 cm; Ø 37 cm; H 8 cm; H tenon 2 cm. E. 6372 Upper moving quern of a tenon and mortise mill. The upper surface is regular and course. The hopper has a truncated cone section. The grinding lower 6. Conclusions surface is flat and smooth. Made of basalt. Partially preserved. Chronology unknown. The rotary quern is one of the most advanced le- Measures: Ø 50 cm; H 5,5 cm; Ø hopper 5 cm. ver mills in use in the ancient world and this paper gives a better idea of the sophistication and com- E. 6428 plexity of this device, as well as an idea on the variety of rotary hand mills in use and their different Lower stationary quern. The upper grinding surface mode of operations. is regular, smooth and slightly rounded. The lower It could be hoped that this work, with the dif- surface is regular, but coarse. Made of basalt. Intact. ferent rotary querns in use at Tell Barri, is the base Late 2nd – early 3rd c. AD. for future research on this field, especially in the Measures: Ø 33 cm; H 6,7 cm. Near East.

West & East 184 Monografie, 4 Rotary querns from Tell Barri (Syria): chronology, use and function

BIBLIOGRAPHY

Alonso N. M. 1996, Els molins rotatius: Origen I expan- Childe G. 1961, Rotary Motion, in: Singer C., Hall sio en la mediterranea occidental, «Revista d’Arqueo- A.R., Holmyard E.J., Williams T.I. (eds.), Storia logia de Ponent» 6, 183-198. della tecnologia, Vol. 1, Oxford, pp. 188-216. Alonso N. M. 2002, Le moulin rotatif manuel au nord- Donner M., Marzoli C. 1994, La macinazione: est de la Péninsule ibérique: une innovation technique Evoluzione delle tecniche e degli strumenti, in Donner dans le contexte domestique de la mouture des céréales, M. (ed.) Il grano e le macine, Catalogo dell’Esposizio- in Procopius H., Treuil R. (eds.), Moudre et bro- ne, Museo Provinciale di Castel Tirolo, 73-98. yer, II. L’interpretation functionnelle de l’outillage de Frankel R. 2003, The Olynthus Mill, Its Origin, and mouture et broyage dans la préhistoire et l’antiquité, Diffusion: Typology and Distribution, «American 2: Archéologie et Histoire: du Paléolithique au Moyen Journal of Archaeology» 107/ 1, pp. 1-21. Age, Paris, pp. 111-127. Jodry F. 2011, Le lexique, in: Buchsenschutz O., Amiran R. 1956, Millstones and the Potter’s Wheel, Jaccottery L., Jodry Fl., Blanchard J. L. «Eretz-Israel» 4, pp. 46-49 (Hebrew with English (eds.), Evolution typologique et technique des meules summary). du Néolithique à l’an mille, Actes des III rencontres ar- Amouric H. 1996, L’anille et les meules (Colloque cheologiques de l’Archéosite gaulois (Saint-Julien-sur- International Tecniques et économie antiques Garrone, 2-4 octobre 2009), Bordeaux, pp. 19-33. et médiévales: le tempes de l’innovation Aix-en- Jodry F., Jaccottey L., Robin B., Picavet P., Provence 21-23 Mai), pp. 39-47. Chaussat A. 2011, Typologie et foncionnement Bombardieri L. 2004, Il dispositivo a pivot per tournette. des manchons des moulins rotatifs manuels du- Sviluppo e diffusione tra la Mesopotamia settentrionale rant le deuxième age du Fer et le Haut-Empire, in: e l’area egea, «Orient-Express» 2004/4, pp. 96-101. Buchsenschutz O., Jaccottery L., Jodry Bombardieri L. 2005, Mulini assiri a scanalatura e mu- Fl., Blanchard J. L. (eds.), Evolution typologique lini a tramoggia. Lo sviluppo della tecnologia molito- et technique des meules du Néolithique à l’an mille, ria a Tell Barri e nella Mesopotamia Settentrionale Actes des III rencontres archeologiques de l’Archéo- nel corso del I millennio a.C., in: Pecorella P.E., site gaulois (Saint-Julien-sur-Garrone, 2-4 octobre Pierobon-Benoit R. (eds.), Tell Barri/Kahat, La 2009), Bordeaux, pp. 299-317. campagna del 2002, Relazione preliminare, Firenze, Loud G. 1948, Megiddo II, Chicago. pp. 156-179. Loud G. 1948, Megiddo II, Chicago. Bombardieri L. 2010, Pietre da Macina, Macine per Macalister R.A.S. 1912, The Excavation of Gezer, mulini, Definizione e sviluppo delle tecniche di macina- 1902-1905 and 1907-1909, Vol. 2, London. zione nell’area del Vicino Oriente e del Mediterraneo Moorey P.R.S. 1994, orientale antico, (British Archaeological Reports Ancient Mesopotamian Materials 2055), Oxford. and Industries, the Archaeological Evidence, Oxford. Buchsenschutz O., Jaccottery L., Jodry Fl., Moritz L.A. 1958, Grain Mills and Flour in Classical , Oxford. Blanchard J. L. (eds.) 2011, Evolution typologique Antiquity et technique des meules du Néolithique à l’an mille, Naze Y., Fronteau G., Robert B. 2011, L’atelier Actes des III rencontres archeologiques de l’Archéosite de meules rotatives en calcaire à cérithes de Vendresse- gaulois (Saint-Julien-sur-Garrone, 2-4 octobre 2009), Beaule (Aisne). Note à propos des outils de moutu- Bordeaux. re en calcaire Lutétien, in: Buchsenschutz O., Jaccottery L., Jodry Fl., Blanchard J. L. Catling H. W. 1972, An Early Byzantine Pottery (eds.), Factory at Dhiorios, «Levant» 4, pp. 1-81. Evolution typologique et technique des meules du Néolithique à l’an mille, Actes des III rencontres ar- Chartrain A. 2015, Il mulino, una macchina dell’an- cheologiques de l’Archéosite gaulois (Saint-Julien-sur- tichità, in Archetti G. (ed.), La civiltà del pane, Garrone, 2-4 octobre 2009), Bordeaux, pp. 269-283. Spoleto, pp. 433-468.

West & East 185 Monografie, 4 Maurizio Fascitiello

Netzev E. 1991: Masada III. The buildings: stratigraphy Robin B., Boyer F. 2011, La liaison des meules rotati- and architecture, Jerusalem. ves manuelles protohistorique et gallo-romaones: oeil et Palermo R. 2012, A Parthian Domestic Complex at anille, in: Buchsenschutz O., Jaccottery L., Tell Barri, in: Matthews R., Curtis J. (eds.), Jodry Fl., Blanchard J. L. (eds.), Evolution typo- Proceedings of the 7th International Congress on the logique et technique des meules du Néolithique à l’an Archaeology of the Ancient Near East, 12 April -16 mille, Actes des III rencontres archeologiques de l’Ar- April 2010, the British Museum and UCL, London chéosite gaulois (Saint-Julien-sur-Garrone, 2-4 octo- Volume 3: Fieldwork & Recent Research. Posters bre 2009), Bordeaux, pp. 351-358. Wiesbaden, pp. 637-647. Runnels C. 1990, Rotary querns in Greece, «Journal of Peacock D. 2013, The Stone of Life. Querns, mills Roman Archaeology» 3, pp. 147-154. and flour production in Europe up to c. AD 500, Sidi N. 2000, Roman and Byzantine Small Objects, in: Southampton. Hirschfeld Y. (ed.), Ramat Hanadiv Excavations, Pecorella P.E., Salvini M. 1982, Tell Barri/Kahat Jerusalem, pp. 177‑185. 1. Relazione Preliminare sulle campagne 1980-1981, Storck J., Teague W.D. 1952, A History of Milling, Firenze. Flour for Men’s Bread, University of Minnesota. Pecorella P.E. 1993, Tell Barri/Kahat, La campagna Thurmond, D.L. 2006, A Handbook of Food Processing del 1992. Relazione Preliminare, Firenze. in Classical Rome, Leiden-Boston. Picavet P., Fronteau G., Boyer F. 2011, Le meules Trokay M. 1989, Le deux documents complémentai- romaines de sept chefs-lieux de cité de Gaule Belgique res en basalte du Tell Kannas, Base de tournette ou occidentale, étude du matériel et synthèse bibliographi- meule?, in: Lebeau M. (ed.), Reflets des duex fleuves, que, «Revue du Nord» 393, pp. 167-226. volume de mélanges offerts à André Finet( Akkadica Pierobon Benoit R. 2008 (ed.), Tell Barri: storia di Supplementum VI), Leuven, pp. 169-175. un insediamento antico tra Oriente e Occidente, (La Watts M. 2011, A newly identified milling artefact from Parola del Passato, 63), Napoli. Roman Britain, in: Williams D., Peacock D. Pruss A. 1994, Kleinfunde aus Stein, in: Waalke (eds.), Bread for the people: the archaeology of mills and Meyer J.-W. (ed.), milling, Proceedings of a colloquium held in the British Die Kleinfunde von Tell Halawa th th A, pp. 205-229. School (Rome, 4 – 7 November 2009), Oxford, pp. 93-96. Py M. 1992: Meules d’époque protohistorique et romaine provenant de Lattes, «Lattara» 5, 183-232. Wefers S., 2011, Still using your saddle quern? A com- pilation of the oldest known rotary querns in western Quesada Sanz F., Kavanagh De Prado E., Lanz Europe, in Williams D., Peacock D. (eds.), Bread Dominguez M. 2014, Los molinos del yacimiento for the people: the archaeology of mills and milling, del Cerro de la Cruz (Almedinilla, Cordoba): classi- Proceedings of a colloquium held in the British School fication y analisis de los ejemplares de época ibérica y (Rome, 4th – 7th November 2009), Oxford, pp. 67-76. emiral, «SPAL» 23, pp. 83-118. Williams-Thorpe O. 1988, Provenancing and Archae Reille J.L. 2000, L‘apparition des meules rotatives en ology of Roman Millstones from the Mediterranean Languedoc oriental (IVe s. avant J.C.) d‘après l‘étude Area, «Journal of Archaeological Science» 15, du site de Lattes, «Gallia» 57, pp. 261-272. pp. 253-305.

West & East 186 Monografie, 4 Globalization of zoomorphic motifs during the Iron Age in Iran and its neighbours

SOHEILA HADIPOUR MORADI

University of Turin

Abstract

Several local polities and kingdoms in the north, northwest, and west of the modern territory of Iran were located on the periphery of the great kingdoms during the Iron Age (circa the first millennium BC). These governments were not only in contact with each other but also with the major empires, through war, domination or trading – and in some cases with their distant neighbours, too. These local Iron Age societies were large-scale producers of various types of metalwork; many of these artefacts were decorated with animals and fantastic creatures. Such objects have been found extensively throughout this area and beyond. This paper discusses the various zoomorphic motifs used on metalwork in Iran during the ninth to seventh centuries BC in detail and gives a general overview of metalwork produced by Iran’s neighbours.

Keywords

Zoomorphic, metalwork, Iron Age, Falak- ol- Aflak Museum, Iran, Mesopotamia

West & East 187 Monografie, 4 Soheila Hadipour Moradi

1. Introduction which flows in the southwestern corner of the country. Lake Orumie (in the far northwest) is the coun- The use of diverse precious metals in the production try’s largest water body. of artistic and religious zoomorphic artefacts was commonplace during the Iron Age in western Iran (1500–550 BC). The majority of such zoomorphic 3. The Iron Age in the Iranian Plateau artefacts were discovered in north, northwest and west Iran. The term “Iron Age” is employed to identify the Iron Age craftsmen in Iran used a wide variety cultural modifications introduced around the mid- of animal motifs on these objects. It seems that each second millennium BC;1 this period was followed part of the territory had its own special material, mo- by several changes with respect to the developed tifs, shape, and style; the zoomorphic objects discov- long-distance trade and early urban culture of the ered reflect the religious concepts of the people who Bronze Age. More specialized local traditions are created and used them for their ritual ceremonies. attested in the finds assemblages from local ceme- The style of the objects indicates their utiliza- teries.2 The Iron Age in Iran is generally dated from tion; they include a wide range of functional and 1500 to 550 BC: Iron Age I (1500–1000 BC), II ritual artefacts. Most of the objects discovered in (1000–800 BC) and III (800–550 BC).3 Archaeo- graves had zoomorphic motifs related to ritual cer- logical studies have shown that during the Iron Age emony scenes, such as a combination of real and metalwork production was a primary activity in the mythical creatures in battle, real or mythical crea- Iranian Plateau, especially in the western part of ture-plant compositions and real or mythical crea- Iran, central Zagros region or Luristan, north-west- ture-plant-human compositions, depicting hunting ern Iran, Azerbaijan, and northern Iran, Gilan, and or battle scenes. Mazandaran.4 The largest quantity of these finds Perhaps due to conflicts or other kinds of con- has come from the Iron Age levels at sites such as tact, such as commercial relations, Iran absorbed Hasanlu,5 Sangtarashan6 and Sagzabad7, as well as several zoomorphic styles from its western neigh- Marlik8 and several other localities. Most of these bours, and at the same time created its own local ar- metal objects were found in sanctuaries and graves; tefacts. This influence has been shown particularly they were deposited as votive offerings in burials with regard to metalwork. (fig. 2).9 Artefact types and shapes, metals and decorations differ from one area to another. However, they did have some similarities with the metalwork 2. The geography of Iran from neighbouring countries.

Geographically, Iran is located in West Asia; it is bounded to the north by Azerbaijan, Armenia, Turkmenistan, and the Caspian Sea, to the east by Pakistan and Afghanistan, to the south by the Per- sian Gulf and the Gulf of Oman, and to the west by 1 Muscarella 2006. 2 Turkey and Iraq. Danti 2013. 3 Iran is a rugged country of plateaus and moun- Muscarella 2006; Overlaet 2005. 4 tains, dominated by the Elburz Mountains in the Haerinck 1988; Pigott 1990; Moorey 1982. 5 Winter 1980, 1989. north, and the Zagros Mountains along its western 6 Oudbashi et Al. 2013. borders. The central and eastern portion is occupied 7 Mortazavi et Al. 2011. by the Iranian Plateau. 8 Negahban 1996. Dasht-e Kavir is a sandy (and salty) desert pla- 9 Fallahian 2013; Overlaet 2005, 2013; Musca- teau (fig. 1). The most significant river is the Karun, rella 2006; Danti 2013.

West & East 188 Monografie, 4 Globalization of zoomorphic motifs during the Iron Age in Iran and its neighbours

Figure 1 Map of Iran, https:// store.mapsofworld.com

Figure 2 Important archaeological sites where zoomorphic objects have been discovered in Iran

West & East 189 Monografie, 4 Soheila Hadipour Moradi

4. Variety of objects occur alone or in relation to other motifs. Most objects with such decorations were discovered in The Iron Age is associated with an increase in graves and sanctuaries. metal production and the creation of multi-functional and decorative objects. The use of precious a. Real or mythical creature-creature composi- metals to make artistic and religious artefacts was tions, in battle or ritual ceremony scenes. commonplace during this period in western Iran b. Creature-plant motifs, real or mythical. (1500–550 BC). Metalwork with a wide range of c. Human-creature compositions, real or mythical, zoomorphic motifs was prevalent in Iran during with hunting, battle or ritual ceremony scenes. the Iron Age. d. Real or mythical creature-plant-human compo- This metalwork influenced Median and later sitions, with hunting, battle or ritual ceremony Achaemenid production. Some objects had func- scenes. tional and ritual usage. The different decorations distinguish objects’ usage. Such objects may be: Objects with zoomorphic motifs were used especially during ritual ceremonies. Some were discovered a. Functional objects during systematic archaeological excavations and i. Armaments (sword, dagger, axe head, mace others purchased from the antiquities market. Most head, shield, adze, belt, quiver plate, whet- come from Luristan, in western Iran. The majority stone, handle). are bronze ritual objects with diverse zoomorphic ii. Ornaments (ring, armlet, bracelet, decora- motifs, such as rhytons, different kinds of plate, and tive plaque, pin, disc-headed pin and mirror). statues and statuettes of various animals. Most ar- iii. Horse gear (horse-harness, trappings and tefacts from Luristan were produced by casting. In horse-bit with decorative cheek-piece, breast- the north and northwest of Iran, gold was the metal plate). most frequently used, in ritual plates, buckets, rhytons and so on; the technique of soldering was used iv. Kitchen utensils. in these areas. It seems that special objects with distinctive motifs were created in each area of Iran, us- b. Ritual objects ing characteristic materials and techniques. i. Ritual vessels (spouted vessel, rython, beaker, bowl, decorative vessel, and plate). ii. Amulets and idols. 6. Iran and its neighbours; influences iii. Sculptures and small statues of various animals. and contacts iv. Horse gear (horse-harness, horse-bit with decorative cheek-pieces, breastplate). During the first millennium BC small states in the north, northwest, and west of Iran bordered on v. Armaments (sword, dagger, axe head, mace powerful empires such as Assyria, Urartu and Elam. head, shield, adze, belt, quiver plate, whet- These local governments had contacts with the em- stone handle). pires in their vicinity. vi. Ornaments (ring, armlet, bracelet, decora- The art produced in these small polities in Iran tive plaque, disc-headed pin). was influenced by the great empires, but this did not lead to the production of exact copies; they achieved new and native styles of art. Archaeo- 5. Variety of subjects logical finds indicate that, due to geographical circumstances, governments in the west, north, and Zoomorphic motifs are commonly found on Iron north-west of Iran had close relations with Mes- Age objects. Real or mythological creatures might opotamia and a part of Anatolia. The impact of

West & East 190 Monografie, 4 Globalization of zoomorphic motifs during the Iron Age in Iran and its neighbours

Figure 3 The decorative scheme of the Hasanlu bowl, based upon the 1974 drawing by M.T.M. de Schauensee. The highlighted area suggests the focus of the bowl’s decorative scheme (Winter 1989, p. 90)

these communications may be discerned on metal the Hasanlu bowl shows influences from ninth cen- objects with zoomorphic motifs. tury carved stone reliefs from Carchemish, for exam- Ritual ceremonies were important, and contact ple a woman with mirror seated on a lion, and gods with societies to the west led to different gods and standing on the back of a couchant lion (fig. 4: a and goddesses reaching Iran. They came on a variety 4: b).10 The same motifs appeared later in Urartian of vessels rather than rock reliefs and seals. Several art11 as well. A Babylonian cylinder seal in the Metro- vessels with deities were found in the west of Iran, politan Museum of Art shows a god standing on the which portray special animals, real or imaginary. back of an animal from whose mouth rain or water The Golden Bowl of Hasanlu is one of the sig- is pouring, as may also be seen on the Hasanlu bowl. nificant zoomorphic objects found in Iran. It could In addition, according to Lambert the three be considered a real and mythical creature-human charioteers above the scene might show Hurrian in- composition, with the hunting, battle and ritual fluence; there are also frequent references to the im- ceremony scenes. In fact the subjects on the bowl portance of storm, sun, and moon, gods in the Hur- are widespread in Mesopotamia. It seems that the rian pantheon.12 inhabitants of the Iran plateau and the Zagros The leader of the three men confronting the di- Mountains acquired such motifs through contact vine array holds out a goblet or beaker before the with their western neighbours; parallel concepts lead chariot, while each of the two individuals fol- have also been identified in the religions of the An- lowing him brings a ram, presumably for sacrifice. cient Near East. We focus here on some of its zoo- This scene is well known from the Hittite and Neo- morphic subjects in relation to neighbouring west- Hittite periods to the west and Elam to the south.13 ern cultures (fig. 3). The close resemblance of composition and motifs 10 Winter 1989, p. 91. between the Hasanlu bowl and items from other ar- 11 Van Loon 1966. eas needs further systematic study, but scholars have 12 Lambert 1978. debated the question. Muscarella has suggested that 13 Dyson 1989, p. 11.

West & East 191 Monografie, 4 Soheila Hadipour Moradi

a b Figure 4 a – Relief from Carchemish in North Syria showing the sun and moon gods on a lion. (Archaeological Museum, Ankara, No. 10078. Photo courtesy of the British Museum, Department of Western Asiatic Antiquities) b – Relief from Carchemish in North Syria showing a goddess (or priestess) seated on a lion. (Archaeological Museum, Ankara, No. 141. Photo courtesy of the British Museum, Department of Western Asiatic Antiquities)

Figure 5 Figure 6 Assyrian bronze plate (Curtis, Reade (eds.) 1995) Silver beaker from Luristan, 102A Ancient Near East, Cleveland Museum of Art

West & East 192 Monografie, 4 Globalization of zoomorphic motifs during the Iron Age in Iran and its neighbours

Figure 7 Ziwiye gold belt fragments, Private Collection (Ghirshman 1950, pp. 194- 195, fig. 19)

The repetition of zoomorphic motifs in a rough- These subjects are also engraved on several Hittite ly similar style from piece to piece indicates that a seals.15 A silver beaker attributed to Luristan, kept common style was shared among the inhabitants. in the Cleveland Museum of Art, is decorated with Perhaps it was a fairly universal use of symbolism in scenes of animals and men in combat in two bands which animals were employed to represent the forc- (fig. 6). This scene is frequent in Iranian art, appeares of the universe. ing for example on a belt at Ziwiye Treasury (fig. 7); Another zoomorphic motif consists of a combi- this belt decoration resembles a hunting scene on a nation of different scenes which appeared on objects belt discovered in an Urartian tomb in Zakim near with different shapes, such as bowls, large plates, Kars. Its surface is divided into compartments with vessels, disc-headed pins, quivers, and beakers. lions, winged horses, bulls and kneeling archers in- Battles between humans and diverse animals and side brackets.16 Several objects with the same subjects ritual ceremonies are common scenes in Mesopota- have been discovered in west, north and northwest mia and also reached Iran; many examples may be Iran: a silver cup with hunting scene attributed to seen on Babylonian cylinder seals14 and Assyrian rock northwest Iran, said to be from Amlash (fig. 8),17 and reliefs. An Assyrian plate shows great similarity to plates discovered at a number of sites in Iran (fig. 5). 15 Ward 1899. 16 Piotrovskii 1959, p. 249, fig. 85. 14 Metropolitan Museum and British Museum collections. 17 Shepherd 1966, p. 40.

West & East 193 Monografie, 4 Soheila Hadipour Moradi

Figure 8 Silver cup with hunting scene, northwestern Iran (Shepherd 1966, p. 39)

a ritual vessel depicting a warrior with extended arms tion, there are animal-head finials among the Oxus who holds a leopard by the throat with each hand Treasure (fig. 9: d).23 discovered at Marlik, north Iran. Later, this subject Another bracelet with animal-head finials can be became a prevalent motif on objects and rock reliefs seen in Esarhaddon’s relief in the Southwest Palace during the Achaemenid Empire. at Nimrud (681-669 BC).24 A similar motif is pre- Zoomorphic pieces of jewellery are among the sent on the chariot of a king on a relief of Assurba- more common Iron Age finds from Iran. Brace- nipal in Nineveh, Room C (669-633 BC);25 and on lets were striking jewellery, and there exists a wide a relief of Assurnasirpal II (884-859 BC).26 range, such as a gold bracelet consisting of a circu- Rhytons – drinking cups or animal-shaped ves- lar tube of gold, jointed in the middle, with a finial sels – are another type of ritual Iron Age object. at each end in the form of a small but wonderfully More recent versions consist of the forepart of an expressive animal head. Bracelets with animal-head animal combined with a vessel. A Hittite rhyton finials made of bronze have been found in various with a fully antlered stag forepart is rendered beau- parts of the Near East including Luristan (fig. 9: a).18 tifully and naturally (fig. 10).27 There are several A bracelet with lion-head finials, most highly dec- bronze rhytons in the Falak- ol- Aflak Castle Mu- orated in the middle joint, was found at Ziwiye19 seum in Luristan which come from the Kalmakare (fig. 9: b). On a bracelet from Marlik there are two Treasure found in Luristan (fig. 11). The rhyton bands around the lions’ necks (fig. 9: c) and they foreparts are in the shape of various animals such have opened mouths.20 This style was used later dur- as lion, ibex, horse, bull, and sometimes a combina- ing Achaemenid period, too. A similar inlaid gold tion of animals (fig. 12). A kneeling bull holding a armlet, belonging to the Achaemenid Period, was spouted vessel is another similar object, which is at- discovered at Susa.21 A bracelet with animal-head tributed to the Proto-Elamite period (fig. 13). This finials perhaps came from Persepolis.22 In addi- silver vessel consists of a kneeling bull wearing hu-

18 Pudelko 1933- 1934, p. 85-88. 23 Ibid, Pls. Xviii, Xix, Xx. 19 Prada 1962, p. 13. 24 Barnet 1963. 20 Negahban 1965, p. 319. 25 Barnett 1970. 21 Dalton 1905, p. 4, fig. 1. 26 Porada 1945. 22 Ibid, p. 57, fig. 39. 27 Benzel et Al. 2010, p. 84.

West & East 194 Monografie, 4 Globalization of zoomorphic motifs during the Iron Age in Iran and its neighbours

a b

c d

Figure 9 a – Armlets from Luristan, National Museum, Tehran, Iran b – Armlet from Iran (http://www.iranicaonline.org/articles/art-in-iran-i-neolithic-to-median) c – Armlet from Marlik (Negahban 1965, p. 312) d – Armlet from the Oxus Treasure (Dalton 1905, pl. XX)

West & East 195 Monografie, 4 Soheila Hadipour Moradi

Figure 10 Stag vessel forepart (Benzel et Al. 2010, p. 85)

Figure 11 Rhytons, Kalmakare Treasure, Falak- ol- Aflak Castle Museum, Khorram Abad, Luristan, Iran

West & East 196 Monografie, 4 Globalization of zoomorphic motifs during the Iron Age in Iran and its neighbours

Figure 12 Rhytons, Kalmakare Treasury, Falak- ol- Aflak Castle Museum, Khorram Abad, Luristan, Iran

Figure 13 Kneeling bull holding a spouted vessel (Benzel et Al. 2010, p. 57)

man clothes, that holds up a vase with its hooves as Lions or lion-heads guarding or spewing forth a ritual offering.28 In Khorsabad a rock relief shows from their jaws the blade of a weapon have a long men drinking from vessels similar to the Kalmakare history in the Near East, beginning in the Early rhyton (fig. 11).29 During the Achaemenid Empire, Bronze Age and continuing until the Neo-Assyri- advanced styles of rhytons were made. an period in Mesopotamia. They are a regular fea- In the Near East the axe-head was a tool with a ture on the more elaborate bronze axe heads, flange- wide range of functional and ritual uses. From the hilted daggers and halberds. 30 Although they may Near East to the Far East, zoomorphic motifs are represent the same subjects, the zoomorphic motifs common decorations on ritual axe-heads from ar- and styles on objects in different regions were each chaeological contexts. Various scenes are portrayed created in their own way. Perhaps the inhabitants of on them, such as battles, ritual ceremonies or en- Iranian Plateau were in contact with their western counters between different creatures. neighbours and this cultural influence left its mark on their craft production. 28 Benzel et Al. 2010, p. 57. 29 Rawlinson 1870, Pl. Cxxxviii, fig. 1. 30 Moorey 1991, p. 7.

West & East 197 Monografie, 4 Soheila Hadipour Moradi

Figure 14

a,b – Axe-heads from Luristan, Falak-ol- Aflak Castle Museum, Khorram Abad, Luristan, Iran

c – Spiked axe-head from Luristan, d – Hurrian axe-head in the form of the forepart Reza Abbasi Museum, Tehran, Iran of a lion (Benzel et Al. 2010, p. 69)

e – Hittite ritual axe-head, Sivas Province (TUR). f – Shaft-hole axe head from Bactria-Margiana, Near East Museum, Pergamon Museum, Berlin Central Asia (Benzel et Al. 2010, p. 75)

West & East 198 Monografie, 4 Globalization of zoomorphic motifs during the Iron Age in Iran and its neighbours

Axe-heads in Luristan were different from those of Mesopotamian and Hittite metalwork (fig. 15: b, in the Iranian Plateau. Zoomorphic motifs are rep- c, f). Hittite figures such as bulls with stags, a stag, resented on almost all of these weapons (fig. 14: a, horse and man may be seen in the collection of the b, c). These axe-heads generally show western influ- Museum of Anatolian Civilizations in Ankara, Tur- ences; during the excavation at Hasanlu some axe- key (15: b, c). There are significant similarities -be heads were discovered that were partly similar to tween Hittite objects and those from Luristan; fu- others found in Luristan. ture studies would throw light on this. There is a large axe-head of the halberd type, Theinhabitants of Luristan produced a large with a crescent-shaped blade, in Falak-ol- Aflak number of zoomorphic objects of various kinds. Castle Museum in Khorram Abad, Luristan prov- Luristan developed a special vocabulary for some ince (fig. 14: b); the central portion is a lion mask, objects. They created pendants, pins, heraldic-ani- from which a fan-shaped bundle of leaves emanates. mal finials, idols – special zoomorphic pins which Along the upper edge of the blade, on either side of were found at Hasanlu too- and disc-head pins the lion mask, a row of saw-tooth lines is engraved. (fig. 15: d, e, g). A tube was added on top to attach the blade to As horses had a vital role in ancient times, horse the handle; this is adorned with a lion, lying with gear was also significant, and is commonly depict- its head towards the top of the weapon, and its tail ed in Assyrian rock reliefs. During the excavation at curled towards its back. Hasanlu some Assyrian cheek- pieces, harnesses and There is a Hurrian axe-head decorated with bits were found which are similar to those shown a snarling lion (fig. 14: d) which probably comes on Assyrian rock reliefs. 35 The cheek-pieces and har- from northeastern Syria and features Hurrian nesses from Luristan are the most famous Iron Age cuneiform,31 but is conceptually similar to the ones from Iran. Abundant zoomorphic horse gear Luristan axe-heads. The zoomorphic spike on the of various types is displayed in various museums in top of a spiked axe-head from Luristan, kept in Reza Iran and all over the world. Abbasi Museum, resembles a Hittite axe-head with Contacts with neighbours appear to have in- zoomorphic spikes (fig. 14: e). creased during the Iron Age. Even distant neigh- Perhaps the tradition of such axe-heads came bours, it seems, were interested in metal objects from farther afield than Mesopotamia. 32 A silver from Iran. Two important finds come from outside shaft-hole axe is a masterpiece of sculpture, partly in Luristan. One of these objects was excavated in the the round and partly in relief. Expertly cast in sil- temple of Hera on the Aegean island of Samos, and ver and gilded with gold foil, it portrays the struggle an openwork pendant of possible Luristan work- for survival of a bird-like creature with a boar and a manship was found at Fortetsa on Crete (figs. 18, dragon (fig. 14: f).33 16).36 How and why these two objects travelled A large amount of bronze and silver statues, idols west remain intriguing questions. Perhaps it shows and figurines with zoomorphic objects were discov- the interrelation of these two areas during the Iron ered at Ziwiye, Hasanlu, Marlik, and at several sites Age, but it is curious that so far no other site in Iran in Luristan, such as Surkh Dom Lori and Sangta- or the Near East has yielded a single specimen of a rashan. A variety of different animals – for instance Luristan bronze (figs. 17, 19).37 a stag at bay, crouching leopard,34 ibex and lion – and hunting are common motifs (fig. 15: a). The style, shape, and sometimes the production technique of zoomorphic objects in Luristan remind us

31 Benzel et Al. 2010, p. 68. 32 Yetts 1931. 35 De Schauensee 1989. 33 Benzel et Al. 2010, p. 74. 36 Muscarella 1977. 34 Negahban 1965, Pl. Liv- Lv. 37 Segall 1943, p. 72.

West & East 199 Monografie, 4 Soheila Hadipour Moradi

Figure 15

a – Bronze figure of crouching leopard, Marlik, Iran b – Hittite figures of bulls and stags, (Negahban 1965, Pl. LIV) Museum of Anatolian Civilisations in Ankara, Turkey

c -Hittite bronze horse and man figurine, d -Idol from Luristan, Falak-ol-Aflak Castle Museum, Museum of Anatolian Civilizations, Ankara, Turkey Khorram Abad, Luristan, Iran

e -Bronze finial form Luristan, Louvre Museum f -Bronze casting from throne, Nimrud (Lancaster 1952, p. 98) (Rawlinson 1870, Pl. LXXIV)

g -Heraldic animal finials from Luristan, Falak-ol- Aflak Castle Museum, Khorram Abad, Luristan, Iran

West & East 200 Monografie, 4 Globalization of zoomorphic motifs during the Iron Age in Iran and its neighbours

Figure 16 Figure 17 Bronze circlet from Crete, Luristan Harness ring from Luristan, (Seagall 1943, p. 73) Reza Abbasi Museum, Tehran, Iran

Figure 18 Figure 19 Bronze jug from Samos, Luristan Bronze jug Sangtarashan, (Seagall 1943, p. 73) Falak- ol- Aflak Castle Museum, Khorram Abad, Luristan, Iran

West & East 201 Monografie, 4 Soheila Hadipour Moradi

Conclusion deduced on the basis of widely accepted examples of creature-symbolism. Although they acquired many zoomorphic mo- The first millennium BC marked the apex of Irani- tifs from their western neighbours, at times these an metalwork. Numerous metal objects were used objects may indicate a more vigorous indigenous in ritual ceremonies, which featured animals, real style and less influence from western craftsmen and and imaginary. Through contacts with their west- imported objects from their western neighbours. ern neighbours, Iran was influenced by their reli- The discovery of zoomorphic objects from Luristan gions. Zoomorphic motifs reveal the many cases in Greece indicates existence of trade contacts and of conceptual similarity between the religions of cultural influence between different regions in the the Ancient Near East and those of Iranian poli- ancient world during the Iron Age. These items of ties. Metalwork exhibits a wide repertory of zoo- metalwork functioned as models for the objects of morphic motifs; the approximate meanings can be the Achaemenid period.

West & East 202 Monografie, 4 Globalization of zoomorphic motifs during the Iron Age in Iran and its neighbours

BIBLIOGRAPHY

Barnett R. D. 1970, The Assyrian Palace Reliefs, Investigation of Copper-Based Artifacts at Tepe Illustrated edition: Assyrian palace reliefs in the British Sagzabad in Qazvin Plain, Iran (1500-800 BC), Museum, London. «Iran Journal of Archaeological Studies» 1/2, Barnett R. D., Faulkner M. 1963, The Sculpture of pp. 49-59. Assur-Nasir-apli II (883-859 B.C.) Tiglath Pileser III Muscarella O.W. 1977, the Archaeological Evidence (745-727 B.C.) Esarhaddon (681-669 B.C.) from the for Relations between Greece and Iran in the First Central and South-west Palaces at Nimrud, London Millennium BC, «Ancient Near Eastern Society», and Bradford. 9/1, 2256. Benzel K., Graff S. B., Rakic Y., Watts E. W. Muscarella O.W. 2006, Iron Age, Encyclopaedia 2010, Art of the Ancient Near east, The Metropolitan Iranica Online, Fasc. 6, pp. 605-609. Museum of Art, New York. Negahban E.O. 1965, Notes on Some Objects from Curtis J.E., Reade J.E. (eds.) 1995, Art and empire: tre- Marlik, «Journal of Near Eastern Studies» 24/4, asures from Assyria in the British Museum, London. pp. 309-327. Dalton O.M. 1905, The Treasure of the Oxus, with Negahban E.O. 1996, Marlik: the Complete Excavation Other Objects from Ancient Persia and India, London. Report, Philadelphia. Danti M.D. 2013, The Late Bronze and Early Iron Age Oudbashi O., Emami S.M., Malekzadeh M., in northwestern Iran, In the Oxford handbook of an- Hassanpour A., Davami P. 2013, Archaeo cient Iran, Oxford, pp. 327–376. metallurgical Studies on the Bronze Vessels from De Schauensee M. 1989, Horse gear from Hasanlu, Sangtarashan, Luristan, W-Iran, «Iranica Antiqua» «Expedition» 31/2, pp. 37-52. 48, pp. 147-174. Dyson Jr. R. 1989, Rediscovering Hasanlu, «Expedi Overlaet B. 2005, The Chronology of the Iron Age in tion» 31, pp. 3- 11. the Pusht-i Kuh, Luristan, «Iranica Antiqua» 40, pp. 1–33. Fallahian Y. 2013, Investigation of Burial Patterns in Iron Age of Gilan, Iran, «Ancient Asia» 4, pp. 1–13. Pigott V.C 1990, Bronze: in Pre-Islamic Iran, Ency clopaedia Iranica Online, London and New York, Ghirshman R.H. 1950, Notes Iraniennes IV. Le Trésor pp. 457–471. de Sakkez, les origines de l’Art Mède et les bronzes du Luristan, «Artibus Asiae» 13/3, pp. 181-206. Piotrovskij B. 1959, Vanskoe Tsarstvo (Urartu), Moscow. Haerink E. 1988, The Iron Age in Guilan: proposal for a chronology, Bronze Working Centers of Western Asia, Porada E. 1962, Alt-Iran, Die Kunst in vorislamischer c. 1000-539 B.C., London, pp. 63–78. Zeit, „Kunst der Welt“, Baden-Baden. Lambert W. G. 1978, The Mesopotamian Background of Porada E. 1945, The Great King, King of Assyria: the Hurrian Pantheon, «Revue Hittite et asianique» Assyrian Reliefs in the Metropolitan Museum of Art, 36, pp. 129-134. New York. Lancaster C. 1952, Luristan bronzes: their style and Pudelko G. 1933, Altpersische Armbänder, «Archiv für symbolism, «Archaeology» 5/2, pp. 94-99. Orientforschung» 9, pp. 85-88. Moorey P.R.S. 1982, Archaeology and Pre-Achaemenid Rawlinson G. 1870, The seven great monarchies of the Metalworking in Iran: a fifteen year retrospective, ancient eastern world; or, the history, geography, and «Iran» 20/1, pp. 81–101. antiquities Of Chaldaea, Assyria Babylon, Media, Persia, Parthia, And Sassanian, or New Persian Moorey P.R.S. 1991, The Decorated Ironwork of the Empire, Vol. 1 of 7, Published October 4th2009, Early Iron Age Attributed to Luristan in Western Public Domain Books. Iran, «Iran» 29/1, pp. 1-12. Segall B. 1943, Greece and Luristan, «Bulletin of the Mortazavi M., Ahmad S.K., Mohammad A.G., Museum of Fine Arts» 41, pp. 72-76. Hassan T. 2011, Preliminary Metallurgical

West & East 203 Monografie, 4 Soheila Hadipour Moradi

Shepherd D.G. 1966, Four Early Silver Objects from Winter J.I. 1980, Hasanlu Special Studies, Volume Iran, «The Bulletin of the Cleveland Museum of I: A Decorated Breastplate from Hasanlu, Iran, Art» 53/2, pp. 38-50. Vol. 1, University of Pennsylvania Museum of Talái H., Aliyari A. 2009, Haftavan IV (Iron II) set- Archaeology. tlement cemetery: N-W Iran, Azerbaijan, «Iranica Winter J.I. 1989, the Hasanlu Gold Bowl, «Expedition» Antiqua» 44, pp. 89–112. 31/2, pp. 87- 106. Van Loon M.N. 1966, Urartian Art: its distinctive traits Yetts W.P. 1931, Chinese Contact with Luristan, «The in the light of new excavations, Nederlands historisch- Burlington Magazine for Connoisseurs» 59/341, archaeologisch instituut, Vol. 20, Leiden. pp. 76-81. Ward W.H. 1899, The Hittite Gods in Hittite Art, «American Journal of Archaeology» 3/1, pp. 1-39.

Webliography https://store.mapsofworld.com. https://www.britannica.com/place/Iran. http://www.iranicaonline.org/articles/art-in-iran-i-neolithic-to-median.

West & East 204 Monografie, 4 Defining the MB-LB transition in northern Mesopotamia: some archaeological considerations on the new data from the Erbil Plain and neighbouring regions

VALENTINA OSELINI

Sapienza University of Rome

Abstract

The transitional phase between the Middle and the Late Bronze Age in Northern Mesopotamia is difficult to define, both from a historical and an archaeological perspective. This paper seeks to address the main issues in the identification of diagnostic ceramic features for this phase, by comparing the evidence from the Erbil plain with the data gathered from surveys and excavations in the Middle and Upper Tigris region (i.e. Yorgan Tepe, Tell Rimah, Tell Billa) and in the Syrian Jazirah (i.e. Tell Brak, Tell Bderi and Tell Barri). Through the analysis of the ceramic assemblages coming from the Kurdistan Region of Iraq and their comparison with materials belonging to better known stratified sequences, I will focus on the diagnostic features characterizing the pottery repertoire during the Middle Bronze Age - Late Bronze Age transition.

Keywords

Transition, pottery, Northern Mesopotamia, Dark Age

West & East 205 Monografie, 4 Valentina Oselini

1. Introduction information about the small kingdom of Tunip- Teššup, probably located East of the Tigris and The transition between the Middle and the Late contemporary with the reign of Hattusili I, is pro- Bronze Age in Mesopotamia (hereafter MBA and vided by textual documents whose provenance is LBA) is traditionally dated to the middle of the dubious.6 Textual sources coming from other ar- second millennium BC, and it represents a prob- eas7 suggest that Mitanni should be considered lematic period for what concerns both the defini- as the most powerful polity in northern Mesopo- tion of an absolute chronology and the identifi- tamia before the rise to power of the Middle As- cation of its archaeological phases.1 The centuries syrian Kingdom. The peak of Mitanni’s influence between the Old Babylonian and the Middle Kas- was reached during the fifteenth and the first half site period are known as “Mesopotamian Dark of the fourteenth century BC, when a vast terri- Age”2 owing to the shortage of textual sources for tory extending from North-Western Syria to the this epoch. Moreover, the absence of well-docu- Eastern Tigris Region was under its control.8 The mented and exhaustive stratified contexts and ar- capital city of the Mitannian state was Washush- chaeological sequences for the northern Mesopo- kanni, initially identified with Tell Fekheriye,9 tamian area so far (especially in northern Iraq) while secondary capitals were located along the does not allow a comprehensive reconstruction of Khabur and Balikh rivers(fig. 1).10 Excavations at the historical background that could fill in the gaps Tell Brak (Area HH),11 Tell Fekheriye,12 Tell Barri of the documentation. There is consensus among (Area G, phases BF-BA)13 and Tell Bderi14 in the scholars in considering the time frame of the col- Syrian Jazirah allow us to recognize the main fea- lapse of the First Dynasty of Babylon, traditionally tures of the LBA material culture in the heartland set in 1595 BC3, and of the decades following this of the Mitannian state. event as a thorny issue.4 According to Novák, the first period that can be For what regards northern Mesopotamia, the properly defined Mitannian -both from an archaeo- scarcity of textual data concerning the phase be- logical and historical point of view- corresponds to tween the Old Assyrian and the Middle Assyrian the overlapping of the last phase of the MBA with periods, spanning from the end of eighteenth to the end of fifteenth century BC, has prevented a dle of the fifth century BC, the city of Assur appears as the capital of a territorial state (Veenhof, Eidem 2008, pp. 23-27). reliable reconstruction of the Assyrian history at 6 Salvini 1998, 1996. the middle of the second millennium BC.5 Some 7 Western Syria, Egypt, Hittite and Babylon Kingdoms, Cancik-Kirschbaum et Al. 2014, pp. 2-3. 1 Pruzsinsky 2009; Schwartz 2008, p. 450. 8 De Martino 2014, p. 61; Schwartz 2014, 2 Pruzsinsky 2009, p. 41. pp. 270‑271. 3 According to the Middle Chronology. 9 The identification of Washushkanni with Tell Fekheri- 4 Gasche et Al. 1998, pp. 6-7; Pruzsinsky 2009, ye was first suggested by von Oppenheim (von Oppenheim p. 17. 1931, p. 60). It has to be said, though, that this association has 5 The Assyrian King List is unclear for the end of the not been proved by archaeological excavations or documentary eighteenth century, when the so-called Dark Age begins. At the evidence so far (Bonatz 2015, p. 26; Salvini 1998, p. 114). time of Shamshi Adad, the city of Assur had a prominent role 10 Such as Ta’idu, seat of the sovereigns of Khanigal- as religious hub and trading post connected to the northern bat, possibly to be identified with Tell Hamidiya:Schwartz Mesopotamian area. We also know that Hammurabi defeated 2014, p. 269. many towns in the area of Subartu and Ekallatum during his 11 Oates D., Oates J., Mcdonald 1997. 33th year of reign and, despite the lack of a direct mention of 12 Bonatz 2014, 2015; Bonatz, Bartl 2012. Assur, it is likely that this city was included among the towns he 13 triumphed over. Textual data from Sippar, contemporary with According to Pecorella, Tell Barri was one of the main the reign of Samsu-Iluna (1750-1712 BC), describe Assur as an religious centres of the Mitannian state during the sixteenth important trade centre and Tell Leilan as the capital of the Old and fifteenth century BC: Pecorella 1997, p. 310. For the Assyrian kingdom. At a later stage, Shaushatar of Mitanni con- analysis on the MB-LB archaeological sequence and pottery quered the city of Assur which was then dominated or at least from Tell Barri, see Baccelli, Manuelli 2008; Coppini influenced by this political entity. Textual documents increase 2008; D’Agostino, Coppini 2014; D’Agostino 2014. at the beginning of the Middle Assyrian period and, at the mid- 14 Pfälzner 1990, 1995.

West & East 206 Monografie, 4 Defining the MB-LB transition in northern Mesopotamia…

Figure 1 The main centres of the Mitannian state in northern Mesopotamia and in northern Syria during the 15th-14th century BC. Map made by the author

the early LBA (ca. 1600-1522 BC). De Martino, in- 2. Mid-second millennium BC in the stead, argues that the Mitannian polity developed Syrian Jazirah and in the Middle and 15 later, by the end of the sixteenth century BC. Not- Upper Tigris Region: contexts and withstanding the archaeological excavations in the Syrian Jazirah, the main features of the material cul- ceramics ture characterizing the transition between the MBA The bulk of the archaeological material dating to and the LBA are still difficult to identify.16 the middle of the second millennium BC in North- New contributions to the understanding of ern Mesopotamia was brought to light in the Syrian the “Mesopotamian Dark Age” are provided by Jazirah; in particular, the most significant sequences the data from the recent archaeological activities known for the MBA-LBA transitional period were conducted in the Kurdistan Region of Iraq (i.e the excavated at Tell Brak and Tell Barri. The stratigra- Dohuk, Erbil and Sulaymaniyah provinces), where phy of Area HH at Brak shows a continuity of occu- extensive surveys, soundings and archaeological pation from the Old Babylonian period to the Mi- excavations have been carried out at several sites tannian period.17 The analysis of the materials from since 2012. the superimposed domestic buildings in Trench F, indeed, shows elements both of continuity and

17 Levels 7 and 6 of area HH (trenches A-D) are char- 15 Novák 2013, pp. 349-352; According to De Martino, acterized by residential buildings lying on the Old Babylonian a Hurrian polity existed in the Mitannian area even before the layers found in Trench D and predating the construction of time of the Hittite king Mursili I, although the Mitannian state the Mitannian palace (Oates, Oates, McDonald 1997, emerged in a phase posterior to his reign. De Martino 2004; pp. 18‑21, 35); the domestic building in area HH, trench F, can 2014, pp. 66-69. be dated to the Mitannian period and is set on the Old Babylo- 16 Novák 2007, pp. 389-390. nian level (McMahon, Oates 2007, p. 147).

West & East 207 Monografie, 4 Valentina Oselini

change between the two periods.18 Lingering ce- Rimah, traces of re-occupation of the site after the ramic types are: shouldered beakers often decorated abandonment and destruction of the MB buildings with painted stripes (Younger Khabur Ware), the were found in Level 2 of the Great Temple (Site A), so-called “grain measures”, Grey-burnished ware and in two occupation phases of a large residential build- large jars or vats incised with wavy lines.19 ing in Level 5 (Site C) and in the edifices brought to Excavations in Area G and Area P at Tell Barri light in Level 4 of Site D. Level 3 of the Great Tem- revealed a change in the layout and destination of ple area and the so-called “Kitchen floor” of Site this part of the town from residential to productive. C (Level 6a), in particular, are of great interest for Moreover, new ceramic types appeared in the Mi- what concerns the MB-LB transitional period.24 tannian phase; among them can be listed: Young- The pottery assemblage from these two sequenc- er Khabur Ware, red-edge bowls, carinated bowls, es allows us to delineate the ceramic horizon of the straight-sided footed beakers, Nuzi ware and Grey Middle Tigris Region around the middle of the sec- ware.20 D’Agostino and Coppini proposed a corre- ond millennium BC as consisting of undecorated, lation of the MB II with the late Old Babylonian/ incised and painted ware types. Old Jazirah III period, while the LB I should be as- Plain pottery from Yorgan Tepe/Nuzi includes sociated to the earliest phase of the Mitanni period storage jars, small jars, bottles, straight, concave side or Middle Jazirah I.21 shouldered beakers, flat or ring base Grey burnished Until the recent upswing of archaeological ac- ware bowls and other everyday shapes, less frequent- tivities in the Kurdistan Region of Iraq, the most ly found, such as plates, infant burial jars or drains.25 significant discoveries dating to the middle of the The sequence of Area C At Rimah, in particular, second millennium BC in Northern Iraq, and esis significant for the presence of the more typical pecially in the Middle Tigris Region, were made at shapes dating from the MBA to the LBA, such as Yorgan Tepe in the 1920s and at Tell Rimah in the beakers with wide rim, shallow rough-based bowls, 1960s.22 The archaeological evidence from Strata I Grey burnished ware, large unpainted jars and vats and II excavated at Nuzi, as Yorgan Tepe was called with incised wavy decorations.26 in the second millennium BC, consists of buildings, The production of painted pottery, Younger including a palace, a temple and houses, along with Khabur Ware and the so-called Nuzi Ware is gen- a large amount of ceramics and objects.23 At Tell erally associated to the beginning of the LBA. The monochrome painted Khabur Ware is considered 18 A domestic building with a collapsed vaulted roof and a street dating to the Mitannian period were brought to light as belonging to the Old Babylonian tradition and, in Trench F during the 2006 excavations, while an Old Baby- according to Oguchi, the later phase of the Khabur lonian multi-room house emerged directly below it in the 2007 Ware (Younger Khabur Ware), represented by the campaign. The materials coming from these buildings, including the pottery, are currently under study. McMahon, Oates painted pottery from Tell Brak and Tell Rimah, 2007, pp. 146-147. has to be dated to the period spanning from 1550 19 Oates, Oates, McDonald 1997, pp. 64-66. and 1400 BC.27 For Pfälzner, the Younger Khabur 20 D’Agostino, Coppini 2014, pp. 387-388. 24 21 D’Agostino, Coppini 2014, table 1. According Postgate, Oates D., Oates J. 1997, pp. 21-26, 37, to Pfälzner, thin-walled shouldered beakers painted in red or 43-44, 56. brown, Grey-burnished ware, White Paste inlay ware, carinated 25 Yorgan Tepe/Nuzi, Strata I and II. For a detailed de- bowls with vertical rim and accentuated, thickened carination scription of the Nuzi pottery see Cecchini 1965 and Starr are the most characteristic shapes of the Middle Jazirah Ia pe- 1939. riod, while Red-Edge bowls are typical of the Middle Jazirah Ib 26 These are typical of the MB-LB transition, recurring phase and Nuzi Ware is common to the entire Middle Jazirah I in Level 6a of Site C. Unpainted ware dating to the LBI peri- period (Pfälzner 2007, pp. 237-250). od was found in Level C5 and consists of pie-crust potstands, 22 For a reappraisal of the survey activities carried out in straight sided and footed beakers, large vessels with geometric the Mesopotamian area see Wilkinson, Ur, Hritz 2013, decorations on the rim and impressed grey ware with white in- while for a focus on the Iraqi territory, see Ur et Al. 2013, lays (Postgate, Oates, Oates 1997, pp. 55-56). p. 89. 27 Oguchi 1997, pp. 196-197, table 1. The archaeologi- 23 Starr 1939, pp. 42-179, 207-262, 304-322, 331-345, cal sequence at Tell Brak and Tell Rimah showed how features 387-494. specific of the Khabur Ware occur in MB/Old Babylonian and

West & East 208 Monografie, 4 Defining the MB-LB transition in northern Mesopotamia…

Figure 2 Relative Chronology of the Northern Mesopotamia area in term of ceramic phases. Grey is the period considered within the present analysis, corresponding to the so called “Dark Age” (until 1550 BC) and to the beginning of LBA (Mitanni period)

Ware, of which the thin-walled shouldered beaker sidered one of the most representative chronological painted in red or brown constitutes the most com- markers of the LBA, often associated to the presence mon shape, is typical of Middle Jazirah Ia-b contexts of Mitanni. It has been mainly related to luxury con- dating from 1550 to 1270 BC and it is contempo- texts such as religious buildings and the residences of rary to the Nuzi Ware production.28 the elite.29 Moreover, it was found in the lower oc- The peculiar white on blackNuzi Ware, attested cupation level of the Kassite palace of Dur Kurigalzu from the Amuq plain to north-eastern Iraq, is con- 29 Cecchini 1965, pp. 19-20. Although Nuzi Ware has LB/Mitannian contexts and illustrated that the distinction be- been found in domestic contexts as well, and even if it doesn’t tween “Older” and “Younger” Khabur Ware is not so strict: seem to be exclusive to the residences of the elite, it is gener- Postgate, Oates, Oates 1997, pp. 53-54. ally uncommon within private dwellings: Postgate, Oates, 28 Pfälzner 2007, pp. 243-244, 248. Oates 1997, p. 54.

West & East 209 Monografie, 4 Valentina Oselini

in association with other small precious objects and nected to the Northern Mesopotamian landscape, this assemblage represents the southernmost deposit settlement distribution and dynamics of interaction of this distinctive ceramic material.30 between sites in the mid-second millennium BC. In the Upper Tigris basin, excavations conduct- Moreover, the increase in field activities in this ex- ed at Tell Billa revealed that the site was occupied tremely important area shall provide a great oppor- around the middle of the second millennium BC. tunity to better define the characteristics of the sec- The earliest stages of this occupation date to the end ond millennium BC local ceramic horizon (fig. 3).34 of the seventeenth-beginning of the sixsteenth cen- The methodology used to distinguish the dif- tury BC (Stratum 4), while the latest are contempo- ferent phases of the second millennium BC in the rary with Stratum II at Nuzi (Stratum 3, 1600-1400 surveys carried out in the Kurdistan Region of Iraq BC). 31 Even if the two Strata were consecutive in generally derives from the archaeological periodiza- the archaeological sequence, the pottery assemblage tion based on the ceramic sequence established for from Stratum 4 lacks some of the shapes that char- the Syrian Jazirah. As proposed by Pfälzner, the MB acterize Stratum 3; this is the case for shouldered should correspond to the Old Jazirah I and II-III beakers decorated with horizontal bands on the periods (ca. 2000-1650 BC), while the LB should body and for Nuzi-style goblets and cups. be equated to the Middle Jazirah I-II periods (ca. Several sites were investigated in the framework 1550-1150 BC).35 Nevertheless, the terms “Mitan- of the Eski Mosul Dam Salvage Project between nian” or “Middle Assyrian” are frequently used in 1981 and 1984. Domestic buildings and graves char- substitution of the periodization of the LBA, in acterized by the presence of Khabur and Nuzi Ware accordance with the stratigraphic sequences high- were found in seven of them, indicating an occupa- lighted in two soundings at Assur and Kar Tukulti tion of this area in the mid-second millennium BC Ninurta36 and with the analysis and classification of and the existence of a common ceramic repertoire the ceramic materials found herein.37 shared with the neighbouring regions.32 At Tell Ri- The archaeological survey conducted in the hin- jim in particular, the early LBA sequence sees the terland of Nineveh (LoNAP)38 revealed that dur- re-use of the latest MBA buildings, testifying to the ing the MBA and the LBA, and especially in the continuity of occupation between the two phases.33 Middle Assyrian period (LB II), this area was quite densely occupied.39 The survey and the test trench

34 The first results of the recent archaeological activities 3. Cultural horizon: diagnostic pottery conducted in the Iraqi Kurdistan region have been published from the Trans-Tigridian region by K. Kopanias and J. MacGinnis in 2016, following the comprehensive article of 2015: Kopanias, MacGinnis, Ur (eds.) 2015; Kopanias, MacGinnis (eds.) 2016. The recent archaeological survey and excavation 35 See Pfälzner 2017 for the final third – early second projects in the Kurdistan Region of Iraq offer the millennium BC; and Pfälzner 2007 for the late second mil- chance to better understand relevant aspects con- lennium BC. 36 Beuger 2014. 30 Baqir 1945, fig. 15. 37 According to Coppini, the LoNAP team uses MBA I, 31 Two seasons of archaeological excavations were carried IIA and II B for the early second millennium BC pottery and out at Tell Billa in 1930-1933 by the same team that was exca- Mitannian or LBA I and Middle-Assyrian or LBA II for the vating at Tepe Gawra, under the direction of E.A. Speiser. Un- late secomd millennium pottery, taking into consideration the fortunately, the excavations did not cover the entire surface of nature of the historical events in the region (Coppini 2018, the tell: Stein 1984, pp. 13-16. p. 67). 32 Tell Fisna (Levels 4 and 3); Kirbet Hatara (Mixed Lay- 38 For a preliminary report on the MBA and LBA pottery ers in the Hatara Village and floors in the Hatara Cemetery); from the LoNAP survey see Coppini 2018. Tell Jikan (Nort-East and East side, Levels 4, 3 -Area A- and 39 Land of Nineveh Archaeological Project in the Dohuk Level 2 -Area B-); Karhal Sufla (three graves); Tell Rijim (Lev- and Nineveh provinces (LoNAP): Morandi Bonacossi, Ia- el 4); Wadi Khatkhum (Period 4): Killick, Black 1985, moni 2015, pp. 23-25, fig. 8. According to Coppini, the MBA pp. 227-239. See also Nemrik (Mitanni settlement: defence sites are clustered in the Navkur Plain, along the Zagros Pied- wall and four Houses, A, B, D, E), Reiche 2014. mont and the Gomel River, East of the Al-Khazir River and 33 Kolisńki 2000, p. XII. in the area around the site of Jerahiyeh. Early LBA sites in the

West & East 210 Monografie, 4 Defining the MB-LB transition in northern Mesopotamia…

Figure 3 Focus on the northern Iraq with the mention of the archaeological sites quoted in the text and of the surveyed areas in the Kurdistan Region of Iraq. Map made by the author (QGIS®)

showed that the largest settlement identified in the In the north-west part of the Kurdistan region Navkur Plain, namely Gir-e Gomel, has a well pre- (EHAS), pottery dated to the LBA was identified served archaeological sequence for the second mil- in ten sites; four among them were probably occu- lennium BC.40 The ceramics shapes used as chrono- pied in the LB I and LB II and about 6% of the total logical references for the beginning of the LBA in number of sherds retrieved in Bassetki can be dat- this area are “pie-crust” pot stands, Grey-burnished ed to the LBA as well. Diagnostic types of the LBA ware, red-edge bowls and Nuzi and Khabur Ware.41 period find parallels with similar shapes from Tell Diagnostic shapes indicating a LB II occupation of Sabi Abyad, Tell Sheikh Hamad, Tell Brak and Tell the sites are carinated bowls and plates, large stor- Barri.43 During the first two seasons of archaeologi- age jars with thickened, squared or rounded rim and cal excavations carried out at Bassetki in 2015 and nipple bases.42 2016, three stratigraphic phases that can be attributed to the sixteenth century BC were recognized in LoNAP area amount to 122 settlements and they are scarcer in Area A (Phases A10-A8). Following Pfälzner, these the eastern part of the plain, more regularly distributed in the Piedmont (Coppini 2018, p. 70). ni 2018, pp. 71-72. 40 Coppini 2018, pp. 72-73. 43 Eastern Ḫabur Archaeological Survey (EHAS) of the 41 Gavagnin, Iamoni, Palermo 2016, pp. 141-142; Dohuk province: Pfälzner, Sconzo 2016a, 2016b. For the Coppini 2018, pp. 70-71. LBA pottery report from the survey see Puljiz in Pfälzner, 42 Gavagnin, Iamoni, Palermo 2016, p. 142; Coppi- Sconzo 2016b, pp. 39-40.

West & East 211 Monografie, 4 Valentina Oselini

are generically dated to the MB III, although some Rouault. An earlier phase datable to the LBA I pe- pottery types diagnostic for the LB occur as well. riod was discovered in area A during the field activ- For this reason, Pfälzner himself proposes to adopt ities, thus demonstrating the existence of a strong the definition of MBA/LBA transitional or “Proto- continuity between the LBA I and the LBA II in Mitanni” phase for this assemblage.44 this relevant site.50 The 2012 survey conducted along the Greater Archaeological excavations are being carried out Zab and in the Bastora Cay (UZGAR45) demon- at Kurd Qaburstan by a team of the Johns Hopkins strated that second millennium BC settlements, University, directed by G. Schwartz.51 The occupa- identified thanks to the presence of band painted tional sequence singled out here covers most of the Younger Khabur Ware,46 are quite abundant in the second millennium BC: the lower town was mainly area and larger in comparison to the third millen- occupied during the MBA, while three levels of occu- nium BC ones. According to the available data, it pation dating to the LBA have been identified at the appears that during the early second millennium top of the mound, encompassing a timespan covering BC the occupation concentrated more on the east- the early LBA, all the LBA I, and the LBA II.52 ern bank of the Greater Zab, while during the LB According to Schwartz, the pottery from Kurd I the western bank of the river was settled as well.47 Qaburstan can be dated to the LBA I period thanks According to Kolinski, during the second millenni- to parallels with Nuzi, Tell Rimah, Tell Brak and um BC the eastern area of the Dohuk province was Tell Barri, even if the Nuzi Ware is very rarely found sparsely occupied.48 within this archaeological context.53 As shown by the results of EPAS Survey, which Starting from 2013, the Italian Archaeological is currently being carried out in the Erbil Plain un- Expedition in the Erbil Plain (MAIPE) has carried der the direction of J. Ur, a progressive increase in out an intensive survey and three campaigns of ar- the number of sites during the MB and LB periods chaeological excavation at Helawa, a site located in can be seen in this area. In this context, Ur identi- the south-western part of the Erbil Plain.54 The dis- fied Qasr Shemamok and Kurd Qaburstan as the tribution of second millennium potsherds, repre- two major settlements, respectively for the LB II senting the 25% of the total amount of the survey and the MB periods, since the beginning of his ac- collection, shows that they are spread on the en- tivities on the field.49 The situation of the second half of the LBA in 50 Masetti-Rouault 2017, p. 113. the Erbil Plain seems to be consistent with this pic- 51 Schwartz 2016; Schwartz et Al. 2017. ture as this period is well represented in the archae- 52 The early phase of the LBA has probably to be associated with a pre-Mitannian occupation and it is represented by ological remains of Qasr Shemamok, where baked a distinctive building characterized by a system of toilets and bricks bears cuneiform inscription about the pal- baths on the High Mound East and by drains and walls on the ace of Adad-Nirari I were identified in Area B by High Mound West. The discovery of a cylinder seal belonging to the Mitanni Common Style in Phase 2 confirms the chrono- the team directed by O. Rouault and M.G. Masetti- logical frame for the objects and the production area is dating to this phase. Few later graves and a kiln have been associated to 44 This interpretation was suggested by the presence of the later occupation of the top of the mound and dated to the flat conical bowls, red-edge bowls and knob-based beakers LBA II/Middle Assyrian period (Phase 1): Schwartz et Al. among numerous specimens of Late Khabur Ware: Pfälzner, 2017, pp. 217-223. Qasim 2017, p. 24. 53 Schwartz et Al. 2017, p. 235. 45 TheUpper Greater Zab Archaeological Reconnaissance 54 The works on the field in the Helawa and Aliawa area project in the Dohuk and Erbil provinces (UGZAR), http://ar- began in 2013 and are still in progress. An intensive survey was cheo.amu.edu.pl/ugzar/indexen.htm; Koliński 2018. carried out at Helawa and Aliawa in 2013 and 2015 and archae- 46 Koliński R. 2012. ological excavations under the direction of L. Peyronel (Uni- versity of Milan) have been conducted in Area B and Area D 47 Koliński R. 2014. at Helawa from 2016: Peyronel, Vacca, Zenoni 2016. The 48 Koliński R. 2015. preliminary results of the survey and the archaeological excava- 49 Erbil Plain Archaeological Survey (EPAS) directed by tions (2016-2018) at Tell Helawa will be published in the pro- J. Ur (Harvard University) in collaboration with the Director- ceedings of the 10th ICAANE conference in Munich, Peyro- ate of Antiquity of Erbil, Ur et Al. 2013, pp. 99-101. nel, Vacca forthcoming.

West & East 212 Monografie, 4 Defining the MB-LB transition in northern Mesopotamia…

Figure 4 Topographic plan of Helawa indicating the distribution of the MB-LB pottery types on the site (results of the 2013-2015 surveys). In general, the MB-LB pottery represents the 25% on the total of potsherds collected. MAIPE©

tire area, with higher concentrations on the north- evident in the case of painted and incised wares. eastern and on the southern slopes of the mound Moreover, both the repertoires present the same (fig. 4). Excavations carried out since 2016 in Area plain ware shapes: carinated bowls with verti- D, on the north-eastern lower mound, brought to cal rims, bowls with outward rims, triangular or light three different architectural phases character- squared in section and deep bowls and jars with ized by domestic and productive structures to be thick squared rims. dated to the middle of the second millennium BC The Helawa pottery assemblage is also character- (Phases 2-4). The ceramic assemblage is homogene- ized by the presence of Younger Khabur Ware and ous and representative of a single ceramic horizon. its typical shouldered beakers with a black-greenish Chronological and typological references for or red painted horizontal band and bowls with ex- the Helawa pottery assemblage can be found in Tell panded rim decorated with a zig-zag motif painted Barri, Tell Bderi, Tell Brak, Tell Rimah, Tell Billa on its upper part (fig. 6: 1-3, 8).55 The unpainted and in the pottery from the recently excavated High Mound at Kurd Qaburstan (figs. 5-6). 55 The red-stripe decoration on shouldered beakers and other shapes found at Tell Rimah and Brak is characteristic of Compelling typological similarities between the the Old Babylonian and Mitannian traditions (Postgate, pottery from Helawa and that from Qaburstan are Oates, Oates 1997, p. 53).

West & East 213 Monografie, 4 Valentina Oselini

Figure 5 Plain pottery, diagnostic open shapes of the beginning of the Late Bronze Age from different sites in northern Iraq and Syrian Jazirah (re-drawn by the author). Helawa materials come from the survey (TH.13.) and the excavations (TH.16.; TH.17.) in Area D

West & East 214 Monografie, 4 Defining the MB-LB transition in northern Mesopotamia…

Figure 6 Painted pottery and diagnostic closed shapes of the beginning of the Late Bronze Age from different sites in northern Iraq and Syrian Jazirah (re-drawn by the author). Helawa materials come from the survey (TH.13.) and the excavations (TH.16.; TH.17.) in Area D

West & East 215 Monografie, 4 Valentina Oselini open forms include bowls with vertical rim (fig. 5: 1), 4. Conclusion bowls with thickened rim (fig. 5: 5), slightly carinated bowls with everted rim (fig. 5: 8-9), carinated The Archaeological survey and excavations in the Grey ware bowls (fig. 5: 13) and deep bowls with ex- Kurdistan Region of Iraq show that this area was panded or squared rim (fig. 5: 15, 17). Among the settled at the beginning of the LBA, an historical closed forms, the most characteristic for this peri- phase that remains partly obscure in the textual as od are the vats and jars with squared rim (fig. 6: 11, well as archaeological documentation. Settlements 13) and the pithoi with thickened and squared rim dating to the beginning of the LBA seem to be fewer (fig. 6: 15). The archaeometrical evaluation of this than those dating to the MBA or to the second part material found that the pastes are usually coarse, of the LBA. organic and mineral tempered, with calcareous and Some sites are larger than others and prob- mica inclusions.56 The characteristic painted ware ably had an administrative role; these are Kurd found both at Helawa and Kurd Qaburstan seems Qaburstan and Qasr Shemamok in the Erbil Plain, to belong to the Younger Khabur Ware horizon, Bassetki in the Eastern Khabur and Gird-e Gomel datable to the beginning of the LBA. in the Navkur Plain. On the other side, smaller set- LBA markers of the Syrian Jazirah and north- tlements like Helawa were established in the coun- western Iraq ceramic horizon, i.e. the white on black tryside as domestic and productive sites with a painted Nuzi Ware, have not yet been found at shorter life cycle. More excavations both on the big- Helawa. Moreover, typical Middle Assyrian (LB II) ger sites and the smaller rural settlements in the Er- shapes, such as the wide carinated bowls, the inward bil Plain are needed in order to clarify the social and rim bowls and the nipple bases are completely absent economic dynamics connecting contemporary sites in this site. On the other side, very few specimens of during the MB-LB transitional period. Nuzi Ware were discovered at Kurd Qaburstan. Chronological and typological correlations be- Three hypotheses can be formulated to explain tween the ceramic material dating to the middle the absence of Nuzi Ware at Helawa and its scarcity and late second millennium BC from the Syrian at Kurd Qaburstan: Qaburstan was occupied longer Jazirah, the Middle and Upper Tigris Region and than Helawa; Nuzi Ware was not as typical and fre- the Kurdistan Region of Iraq show that a common quent in the Erbil Plain as in the Syrian Jazirah; the ceramic horizon was shared in Northern Mesopota- presence of Nuzi Ware at Kurd Qaburstan could mia. However, further and more in-depth analyses be related to the building with representative func- could reveal the existence of smaller ceramic regions tions located on High Mound East. characterized by similar pottery types and technol- The similarity of the ceramic materials from the ogies as is probably the case with the Erbil Plain. two sites as well as their proximity, as they are both Close parallels have been found, indeed, between located in the south-western part of the Erbil Plain the Kurd Qaburstan and Helawa repertoires, show- (ca. 7 km away from each other), suggest that they ing that the two sites were contemporary and shared were probably contemporary and somehow con- a distinctive local ceramic tradition. nected. The comparison between the data from Hel- The analysis of the archaeological and ceramic awa, Kurd Qaburstan and Qasr Shemamok, which sequences of the Kurdistan Region of Iraq could is situated slightly to the north of the first two sites, allow us to establish an updated and independent in the Erbil Plain, is crucial to define the chrono- regional periodization, which would shed light on logical and morphological markers of the material some aspects that remain uncertain due to the lack culture of the area in a regional perspective. of data for other previously investigated areas in Northern Mesopotamia.

56 The archaeometric analyses were conducted by dr. L. Medeghini of the Department of Earth Science (Sapienza, Uni- versity of Rome).

West & East 216 Monografie, 4 Defining the MB-LB transition in northern Mesopotamia…

Acknowledgments

I would like to express my gratitude to Luca Peyronel, director of the MAIPE project of the University of Milan, for giving me the possibility to study the second millennium BC ceramic materials from Helawa. I also wish to thank Agnese Vacca, deputy director of the project, for the useful discussion on the pottery sequence at the site. I am much obliged to both of them for the stimulating suggestions they always provide, enriching the perspectives of my ongoing research. Finally, I wish to extend warm thanks to Costanza Coppini for our frequent exchanges concerning the MBA-LBA ceramic materials in Northern Mesopotamia.

West & East 217 Monografie, 4 Valentina Oselini

BIBLIOGRAPHY

Armstrong J.A., Gasche H. 2014, Mesopotamian International Congress on the Archaeology of the Pottery. A Guide to the Babylonian Tradition in the Ancient Near East (Madrid, April 3-8 2006), Madrid, Second Millennium BC, in MHEM VI, Ghent. pp. 477-491. Baccelli G., Manuelli F. 2008, Middle Bronze Coppini C. 2018, The Land of Nineveh Archaeological Khabur Ware from Tell Barri/Kahat, in: Cordoba Project: Preliminary Results from the Analysis of the J., Molist M., Perez C., Rubio I., Martínez S. Second Millennium BC Pottery, in: Salisbury R. (ed.), (eds.), Proceedings of the 5th International Congress Prehistoric and Historical Landscapes & Settlement on the Archaeology of the Ancient Near East (Madrid, Patterns, Proceedings of the 10th International Congress April 3-8 2006), Madrid, pp. 187-206. on the Archaeology of the Ancient Near East (Vienna, Baqir T. 1945, Iraq Government Excavations at 25–29 April 2016), Wiesbaden, pp. 65-82. Ἁqar Quf, Second interim report 1943-44, «IRAQ D’Agostino A. 2008, Between Mitannian and Middle- Supplement», pp. 1-15. Assyrians: Changes and Links in Ceramic Culture at Beuger C. 2014, Pottery Tradition from the Mitannian Tell Barri and in Syrian Jazirah during the end of the nd to the Early Neo-Assyrian Period – Evidence from 2 millennium BC, in: Cordoba J., Molist M., Soundings in Ashur and Kar-Tukulti-Ninurta, in: Perez C., Rubio I., Martínez S. (eds.), Proceedings th Luciani M., Hausleiter A. (eds.), Recent Trends of the 5 International Congress on the Archaeology in the Study of Late Bronze Age Ceramics in Syro- of the Ancient Near East (Madrid, April 3-8 2006), Mesopotamia and Neighbouring Regions (Proceedings Madrid, pp. 525-547. of the International Workshop in Berlin, 3-5 D’Agostino A. 2014, The Tell Barri Sequence of Late November 2006), Berlin, pp. 263-288. Bronze Age Levels: Evolution Trends within Late nd Bonatz D. 2014, Tell Fekheriye in the Late Bronze 2 Millennium Ceramic Culture, in: Luciani M., Age: Archaeological Investigations into the Structures Hausleiter A. (eds.), Recent Trends in the Study of Political Governance in the Upper Mesopotamian of Late Bronze Age Ceramics in Syro-Mesopotamia Piedmont, in: Bonatz (ed.), The Archaeology of and Neighbouring Regions (Proceedings of the Political Spaces. The Upper Mesopotamian Piedmont International Workshop in Berlin, 3-5 November in the Second Millennium BC, Berlin, Boston, 2006), Berlin, pp. 235-261. pp. 61-84. D’Agostino A., Coppini C. 2014, Life at the Edge Bonatz D. 2015, Tell Fekheriye – An Introduction to the of the Settlement: the MBA-LBA Transition at the Current State of Archaeological Research, in: Hulínek Northern and Southern Slope of Tell Barri (NE Syria), D., Bonatz D., Kováč M. (eds), Archaeology on in: Bieliński P., Gawlikowski M., Koliński Three Continents 2006-2001, Bratislava, pp. 13-30. R., Ławecka D., Sołtysiak A., Wygnańska Z. (eds.), Proceedings of the 8th International Congress on Bonatz D., Bartl P. 2012, Tell Fekheriye 2010: th the Archaeology of the Ancient Near East (University Report on the 4 Excavation Campaign, «Chronique of Warsaw, 30 April – 4 May 2012), Wiesbaden, Archéologique en Syrie» 6, pp. 161-167. pp. 385-410. Cancik-Kirschbaum E., Brisch N., Eidem J. 2014, Gasche H., Armstrong J.A., Cole S.W., in: Cancik-Kirschbaum E., Brisch Introduction, Gurzadyan V.G. 1998, Dating the Fall of Babylon, N., Eidem J (eds.), Constituent, Confederate, and a Reappraisal of Second-Millennium Chronology, in Conquered Spaces. The Emergence of the Mitanni MHEM IV, Ghent. State, in Topoi: Berlin Studies of the Ancient World 17, Berlin, pp. 1-10. Gavagnin K., Iamoni M., Palermo R. 2016, The Land of Nineveh Archaeological Project: The Ceramic Cecchini S. 1965, , Rome. La Ceramica di Nuzi Repertoire from the Early Pottery Neolithic to the Coppini C. 2008, Mitanni Pottery from Tell Barri, Sasanian Period, «BASOR» 35, pp. 119-169. in: Cordoba J., Molist M., Perez C., Rubio Hunger H., Pruzsinszky R. (eds.) 2004, I., Martínez S. (eds.), th Proceedings of the 5 Mesopotamian dark age revisited: proceedings of an in-

West & East 218 Monografie, 4 Defining the MB-LB transition in northern Mesopotamia…

ternational conference of SCIEM 2000 (Vienna 8th-9th Tušratta, in: Hunger H., Pruzsinszky R. (eds.) November 2002), Vienna. 2004, pp. 35-42. Iamoni M. 2012, The Late MBA and LBA Pottery de Martino S. 2014, The Mitanni State: The Horizons at Qatna. Innovation and Conservation Formation of the Kingdom of Mitanni, in: Cancik- in the Ceramic Tradition of a Regional Capital and Kirschbaum E., Brisch N., Eidem J (eds.), the Implications for Second Millennium Syrian Constituent, Confederate, and Conquered Spaces. Chronology, in SAQ 2, Udine. The Emergence of the Mitanni State, in Topoi: Berlin Killick R., Black J.1985, Excavations in Iraq 1983-84, Studies of the Ancient World 17, Berlin, pp. 61-74. «IRAQ» 47, pp. 215-239. Masetti Rouault M.G. 2017, A Summary of the 5th Koliński, R. 2000, Tell Rijim, Iraq. The Middle Bronze Campaign of the French Archaeological Mission at Layers, «BAR International Series» 837. Qasr Shemamok (Kurdistan, Iraq), 21 September – 19 October, 2016, «As-sharq»1, April 2017, Koliński, R. 2012, Report on the field activities of The pp. 112-123. Upper Greater Zab Archaeological Reconnaissance Project in year 2012, http://archeo.amu.edu.pl/ug- Mc Mahon A., Oates J., 2007, Excavations at Tell zar/REPORT%20UGZAR%202012.pdf Brak 2006-2007, «IRAQ» 69, pp. 145-171. Koliński, R. 2014, Report on the field activities of The Morandi Bonacossi D., Iamoni M. 2015, Landscape Upper Greater Zab Archaeological Reconnaissance and Settlement in the Eastern Upper Iraqi Tigris and Project in year 2014, http://archeo.amu.edu.pl/ug- Navkur Plains: the Land of Nineveh Archaeological zar/UGZAR%20Report%202014cor.pdf Project, Seasons 2012–2013, «IRAQ» 77, pp. 9-39. Koliński, R. 2014, Settled Space. Evidence for Changes Novák M. 2007, Mittani Empire and the Question of in Settlement Patterns of Northern Mesopotamia at Absolute Chronology: Some Archaeological Conside the Advent and at the Turn of the Mitanni Era, in: rations, in Bietak M., Czerny E. (eds), The Synchro Cancik-Kirschbaum E., Brisch N., Eidem nisation of Civilisations in the Eastern Mediterranean J (eds.), Constituent, Confederate, and Conquered in the Second Millennium B.C. III. Proceedings of the nd th Spaces. The Emergence of the Mitanni State, in Topoi: SCIEM 2000 – 2 Euro Conference Vienna, 28 of st Berlin Studies of the Ancient World 17, Berlin, May – 1 of June 2003, Wien, pp. 389-402. pp. 179-212. Novák M. 2013, Upper Mesopotamia in the Mittani Koliński, R. 2015, Report on the field activities of The Period, in Orthmann W., Matthiae P., al- Upper Greater Zab Archaeological Reconnaissance Maqdissi M. (eds), Archéologie et Histoire de la Project in year 2015, https://www.academia. Syrie I. La Syrie de l’époque néolithique à l’âge du fer, edu/17604803/UGZAR_Report_2015 Wiesbaden, pp. 345-356. Koliński, R. 2018, An Archaeological Reconnaissance in Oates D., Oates J., McDonald H. 1997, Excavations the Greater Zab Area of the Iraqi Kurdistan (UGZAR) at Tell Brak, Vol. 1: The Mitanni and Old Babylonian 2012–2015, in: Salisbury R. (ed.), Prehistoric periods, London. and Historical Landscapes & Settlement Patterns, Oguchi H. 1997, A Reassessment of the Distribution of Proceedings of the 10th International Congress on the Khabur Ware: an Approach from an Aspect of its Main Archaeology of the Ancient Near East (Vienna, 25–29 Phase, «al Rāfidān» 18, pp. 195-224. April 2016), Wiesbaden, pp. 13-26. Oguchi H. 2006, The Date of the Beginning of Khabur Kopanias K., MacGinnis J. (eds.) 2016, The Ware Period 3: Evidence from the Palace of Qarni- Archaeology of the Kurdistan Region of Iraq and Lim at Tell Leilan, «al Rāfidān» 27, pp. 45-58 Adjacent Regions, Oxford. von Oppenheim M.F. 1931, Der Tell Halaf. Eine neue Kopanias K., MacGinnis J., Ur J. (eds.) 2015, Kultur im ältesten Mesopotamien, Leipzig. Archaeological Projects in the Kurdistan Region in Pecorella E. 1997, Gli Scavi della Missione Italiana Iraq, The Directorate of Antiquities of Kurdistan. a Tell Barri, in: Barich B., Stoppiello A.A., Maidman P. 1995, Nuzi: Portrait of an Ancient Missioni archeologiche italiane: la ricerca archeologica, Mesopotamian Provincial Town, in: Sasson J., antropologica, etnologica, Roma, pp. 309-314. Baines, J., Beckman, G., Rubinson K.S. (eds.), Peyronel L., Vacca A. forthcoming, Survey and Civilizations of the Ancient Near East (Volume II), Excavations at Helawa, Southwest Erbil Plain, New York, pp. 931-947. Kurdistan, Iraq, (Proceedings of the 11th Interna de Martino S. 2004, A Tentative Chronology of the tional Congress on the Archaeology of the Ancient Kingdom of Mitanni from its Rise to the Reign of Near East).

West & East 219 Monografie, 4 Valentina Oselini

Peyronel L., Vacca A., Zenoni G. 2016, Helawa: Salvini M. 1996, The Habiru Prism of King Tunip- a New Northern Ubaid/Late Chalcolithic Site in the Tessup of Tikunani, in Documenta Asiana 3, Roma. Erbil Plain, in Kopanias K., MacGinnis J. (eds.) Salvini M. 1998, The Earliest Evidences of the Hurrians 2016, pp. 309-321. Before the Formation of the Reign of Mitanni, in Pfälzner P. 1990, Tell Bderi. The Development of a Buccellati, G.; Buccellati, M.K. (eds) Urkesh and the Bronze Age Town, in Kerner S. (ed.), The Near East Hurrians, Studies in Honor of Lloyd Cotsen, BiMes in Antiquity. German contributions to the archaeo- 26, pp. 99-115. logy of Jordan, Palestine, Syria, Lebanon and Egypt, Schwartz G.M. 2008, Problems of Chronology: Amman, pp. 63-79. Mesopotamia, Anatolia, and the Syro-Levantine Pfälzner P. 1995, Mitannische und Mittelassyrische Region, in: Aruz J., Benzel K., Evans J.M. (eds.), Keramik. Eine Chronologische, Funktionale und Pro Beyond Babylon. Art, Trade and Diplomacy in the duktionsökonomische Analyse, (Berichte der Aus Second Millennium B.C., New York, pp. 450-452. grabung Tall Šēḫ Ḥamad / Dūr-Katlimmu 3), Berlin. Schwartz G.M. 2014, Reflections on the Mitanni Pfälzner, P. 2007, The Late Bronze Age Ceramic Emergence, in: Cancik-Kirschbaum E., Brisch Tradition on the Syrian Jazirah, in: al-Maqdissī N., Eidem J (eds.), Constituent, Confederate, and M., Matoïan V., Nicolle C. (eds.), Céramique de Conquered Spaces. The Emergence of the Mitanni l‘Âge du Bronze en Syrie, II, L‘Euphrate et la Region State, in Topoi: Berlin Studies of the Ancient World de Jézireh (Bibliothèque archéologique et historique 17, Berlin, pp. 265-277. 180), Beyrouth, pp. 231-291. Schwartz G.M. 2016, Kurd Qaburstan, A Second Pfälzner P. 2017, Ḫabur Ware and Social Continuity: Millennium BC Urban Site: First Results of the Johns The Chronology of the Early to Middle Bronze Age Hopkins Project, in Kopanias K., MacGinnis J. Transition in the Syrian Jezireh, in: Höflmayer F. (eds.) 2016, pp. 385-401. (ed.), The Late Third Millennium in the Ancient Near Schwartz G.M., Brinker C.D., Creekmore A.T., East. Chronology, C14, and Climate Change (Papers Feldman M.H., Smith A., Weber J.A. 2017, from the Oriental Institute Seminar, Held at the Excavations at Kurd Qaburstan, a Second Millen Oriental Institute of the University of Chicago 7–8 nium B.C. Urban Site on the Erbil Plain, «IRAQ» March 2014), in OIS 11, Chigaco, pp. 163-202. 79, pp. 213-255 Pfälzner P., Qasim H.A. 2017, The First and Second (https://doi.org/10.1017/irq.2017.2). Seasons of the German-Kurdish Excavations at Starr R.F.S. 1939, Nuzi. Report on the Excavations at Bassetki in 2015 and 2016, «ZOrA» 10, pp. 10-43. Yorgan Tepa near Kirkuk, Iraq, conducted by Harvard Pfälzner P., Sconzo P. 2016a, From the banks of University in conjunction with the American Schools the Upper Tigris River to the Zagros Highlands. The of Oriental Research and the University Museum of first season (2013) of the Tübingen Eastern Ḫabur Philadelphia 1927-1931, Cambridge. Archaeological Survey, in Kopanias K., MacGinnis Stein D. 1984, Khabur Ware and Nuzi Ware: their J. (eds.) 2016, pp. 323-331. Origin, Relationship, and Significance, in Assur 4, Los Pfälzner P., Sconzo P. 2016b, The EasternḪ abur Angeles. Archaeological Survey in Iraqi Kurdistan. A pre- Ur J., De Jong L., Giraud J., Osborne J.F., liminary Report on the 2014 Season, «ZOrA» 9, MacGinnis J. 2013, Ancient Cities and Landscapes pp. 10-69. in the Kurdistan Region of Iraq: the Erbil Plain Postgate C., Oates D., Oates J. 1997, The Archaeological Survey 2012 Season, «IRAQ» 75, Excavations at Tell Rimah: The Pottery, Warminster. pp. 89-117. Pruzsinsky R. 2009, Mesopotamian Chronology of Veenhof K.R., Eidem J. 2008, Mesopotamia: The the 2nd Millennium B.C. An Introduction to the Old Assyrian Period, in OBO 160/5, Fribourg Textual Evidence and Related Chronological Issues, – Göttingen. in Contributions to the Chronology of the Eastern Wilkinson T., Ur J., Hritz C. 2013, Settlement Mediterranean 22, Vienna. Archaeology of Mesopotamia, in: Wilkinson T., Reiche, A. 2014, Late Bronze Age Pottery from Nemrik Gibson McG., Widell M. (eds.), Models of (Northern Iraq), in: Luciani M., Hausleiter A. Mesopotamian Landscapes. How Small-Scale Processes (eds.), Recent Trends in the Study of Late Bronze Age Contributed to the Growth of Early Civilizations, in Ceramics in Syro-Mesopotamia and Neighbouring BAR IS 2555, Oxford, pp. 34-55. Regions (Proceedings of the International Workshop in Berlin, 3-5 November 2006), Berlin, pp. 289-332.

West & East 220 Monografie, 4 Matching Near Eastern seals and sealings: current issues and new perspectives

MELISSA RICETTI

MiBAC

Abstract

Seals and seal impressions known at present from the Ancient Near East amount to some tens of thousands. Despite this number only a few matches between extant seals and impressions have been identified. The reason may be ascribed to the mobility of seals and sealed objects in the past as well as in more recent times: clay bullae and tablets could be sent to distant places after being sealed. Actual seals have been often brought to light by -il legal excavations and sold on the antiquity market. Seals and sealings might therefore not be found in the same expected place. Moreover, the common practice of re-carving seals caused the loss of many original designs whose existence is now testified only by ancient impressions. Other reasons, however, may be considered. Museum and private collections have been published in several editions and online catalogues. Though a number of innovative methods for imaging and presenting seals and seal impressions have been developed, the cross comparison of seal designs for the discovery of new matches remains a long and laborious process. This paper will discuss some issues related to the matching of seals and their impressions in current studies by presenting a new cataloguing approach for the dataset offered by Kültepe/Kaniš.

Keywords

Seal, seal impression, glyptic art, matching, Kültepe/Kaniš

West & East 221 Monografie, 4 Melissa Ricetti

1. Introduction1 The seal’s or impression’s design, which is commonly called “composite seal”3, is the linear imma- Stamp and cylinder seals are small objects whose terial element on which scholars mostly base their surface is engraved with a design that appears in studies. relief when they are respectively stamped or rolled For decades glyptic studies have been the ideal over a plastic material. place for iconographic and stylistic debates. In the ancient Near East, impressions making The tendency to use one seal for many centuries was only one of the seal’s functions. Seals were not and to carry it across large regions initially made dif- only legal instruments, whose impressions marked ficult to assess date and provenience of certain sam- ownership, authority, acknowledgment and obli- ples, even when they were found in archaeological gation on documents and properties, but were also contexts. Objects where seal impressions appear can amulets, votive objects, and status symbols. provide valuable clues to the chronological and geo- Since their origins, glyptic studies have dealt graphical reference of the seal that made them. Seals with the duality constituted by seals and seal im- are frequently applied on inscribed objects. Once pressions. Seals are unique objects characterized texts are dated and located, same dates and places by some physical features, such as size, material, can be referred to seals impressed on them and to colour, texture, etc. Though some seals are cop- other examples in comparable style.4 ies made upon a shared model, each of them is the As long as research was improved, scholars were product of a distinct craftsmanship process. Im- able to trace with more accuracy the long evolu- pressions of the same seal can appear singly or mul- tion of glyptic art and to accordingly fit in seals on tiply on one object or on different ones. Since they the basis of material, shape, size, carving technique, are produced by the same seal, they substantially styles, represented scene, and single iconographic look the same. Nonetheless they are necessarily dif- motifs. First publications proposed classifications ferent because each of them is the archaeological of seals according to the time and place of origin. evidence of a different human action, i.e. the act of Rich catalogues offered large amounts of material impressing the seal on the object appointed to re- for iconographical and stylistic analyses.5 tain the impression. Each impression is thus differ- In the second half of the last century, a new ho- ent from the other and has some intrinsic features listic approach focusing on the importance of seals that identify it: position on the object impressed, and their impressions, not only for the History of legibility, distortion caused by the irregular rolling Art of the Ancient Near East, but also for the under- of the seal (when it is a cylinder), deterioration or standing of social, economic and ideological dynam- incompleteness are all aspects affecting the final ics of the past, was inaugurated.6 Scholars started appearance of the impression. joining archaeological, art historical and philologi- Seals and seal impressions are therefore comple- cal competences and explored seals not only as ob- mentary and find theirtrait d’union in the design jets d’art, but also as windows from which to observe they both display: one in negative – the intaglio of ancient life. Studies were devoted to manufacturing, the seal, which is directly perceivable from the seal’s diffusion and consumption in order to trace the dif- surface – and the other in positive – the relief of the ferent phases of the seal’s life, from the retrieval of impression, which is the mediated form of the seal’s the raw material to the carving of the intaglio and intaglio.2 from workshop production to their use by custom-

3 Tsouparopoulou 2014, p. 52. 1 I wish to thank the Broadening Horizons 5 Organizing 4 Committee for accepting my paper and my anonymous review- W. W. Hallo in Introduction to Buchanan 1981, ers for their helpful comments. p. xi. 5 2 For cylinder seals and cylinder seal impressions, a fur- Ward 1910; Frankfort 1939; Moorgat 1940; ther point to be considered is the difference for beholders in Porada 1948. perceiving the cyclical design of the intaglio and the linear 6 E.g. Cassin 1960; Gibson, Biggs (eds.) 1977; Po- freeze of the relief. rada 1993. On this theme see, in particular, Winter 2001.

West & East 222 Monografie, 4 Matching Near Eastern seals and sealings: current issues and new perspectives

ers. Iconographical and stylistic researches were pro- Re-carvings not only occurred at considerable gressively refined and new issues and links to ancient distances and long intervals from the original hab- world daily life, imagery and images perception were itat of the seal, but were also requested by current searched into carved designs. Prosopographic and owners when they advanced in rank or had to ac- sealing practices analyses allowed the exploration of knowledge new supervisors,10 as perhaps happened single identities, groups and communities.7 in cases ns. 4 and 15 (Table 1). Seal 1195 in Har- In the following pages I will present an approach vard Semitic Museum was initially claimed by E. for dealing with seal’s and impression’s designs that D. van Buren to have produced the impression brings together the materiality of seals and sealed on the bulla Tell Asmar 33.104.11 The publica- objects with the immateriality of the design. tion side by side of the seal and the bulla by Boeh- mer in 1965 showed, however, that designs differ in the inscription and in a few other details: the 2. Seal-impression matches: a survey one on the impression identifies the owner, Puzur- rum, as a scribe (dub-sar) while the one on the seal 12 Seals and seal impressions known at present from makes him a book-keeper (sag-su18). In light of the ancient Near East amount to some tens of thou- their great similarity, Hallo suggests that the seal sands. Despite their number only a few matches be- may represent “a modification of the seal impres- tween actual seals and impressions have been identi- sion, or vice versa, [the change was] made after the fied so far. In 2001 W. W. Hallo presented a survey owner’s assumption of a new role”, though he does on the alleged matches of seals and seal impressions not even exclude that the impression was made by known to him since that time: of about a dozen cas- another seal of the same owner.13 Case 15, similar- es he mentions and a few others added in the last ly, shows a seal impressed on a receipt mentioning years, only three seem to be genuine cases (grey Aššur-remanni as receiving party, who in this role highlighted in Table 1), while others remain mostly is expected to impress his seal on the tablet (LB hypothetical. 2532- Böhl collection).14 The impression, howev- Therefore, the inevitable question is: why only a er, has no inscription or caption allowing the iden- few matches have been detected so far? tification of the seal, which displays a stag eating Scholars suggest different explanations. Seals, from the “sacred tree” in a common Middle As- which were often made of prestigious stones, had syrian scheme. The same design has been detect- frequently their “life span” extended far beyond ed on a seal preserved at the Bibliothèque Natio- the place and time of their original user.8 Ancient nale and known as “seal of Aššur-remanni” thanks seals were sufficiently valuable for a new owner to to the inscription running horizontally across the wish acquiring one and changing the intaglio, so top of the cylinder: KIŠIB dA-šur-ARHUŠ-ni/A that re-carvings frequently and systematically hap- Mu-KAR-dA-šur, “seal of Aššur-remanni, son of pened. Ur III seal impressions on the envelopes Mušezib-Aššur”. Such a coincidence would tempt from Kültepe/Kaniš prove the appreciation of later generations for ancient seals. Alterations and addi- cious stones Ur III seals were usually made of, such as lapis lazuli), but also from their intrinsic qualities. Ancient seals were tions were generally produced on Ur III re-used seal relics of a glorious time, whose prestige as antique valuables surfaces, though a few seals have been re-used with- was increased also by the prominence of their ancient owners, out changes and occasionally preserve the ancient e.g. the seals of scribes Ur-lugal-banda (CS 421) and Lukalla 9 (kt m/k 63), whose names were significantly maintained in inscription together with the new one. the inscription by the new owners of the seals (Ricetti 2014, pp. 121‑128). 7 For Old Assyrian studies, to which the following proj- 10 Boehmer 1997; Mayr in Hallo 2001, pp. 250-251. ect is devoted, see in particular Larsen 1977; Teissier 1994; 11 Lassen 2012; Ricetti 2014. Van Buren 1946, p. 8, ad fig. 8. 12 8 Hallo, introduction to Buchanan 1981, pp. x, xiv. Lambert 1985, p. 92, n. 17. 13 9 The prestige Old Assyrian merchants attributed to Ur Hallo 2001, p. 246. III seals not only derived from their material value (i.e. the pre- 14 Hallo 1973, p. 183.

West & East 223 Monografie, 4 Melissa Ricetti

Table 1 – Alleged matches surveyed by Hallo in 2001 plus more recent additions1

Match supposed N. Seal Seal impression Type Authenticity /proposed by

1 Illustrated London News, Jan. Buchanan 1965, Stamp, Persian Buchanan: “this design so closely Disproved: the seal 28, p. 142, fig. 3; Buchanan p. 204, n. 4; pl. XVI, Gulf, Failaka resemble the impressions on the tab- and the seal that 1965, p. 204, n. 4; pl. XVI, D C (YBC 5447). Island. let that at first sight one might think made the impres- (Falaika V); Kjareum 1983, they had been made by that seal” sion are variants pp. 78-79, n. 176. (Buchanan 1965, p. 204). on a common theme (Hallo 2001, pp. 239-240, fig. 1-2). 2 Mackay 1938, p. 328, pl. XCIX, Mackay 1938, Stamp, Indus Buchanan (1965, p. 204), who, how- Doubtful. n. 661. p. 328, pl. XCI, Valley. ever, was doubtful in a margin note n. 20. to Mackay’s volume reported by Hal- lo (2001, p. 239, n. 4). 3 Esin, Harmankaya 1987, fig. 14; Esin 1985, p. 260, Stamp, ‘Ubaid Reichel 2003, p. 364, n. 29. Not confirmed. see also Esin 1989, p. 137. n. 14; see also Esin period, Değir- 1989, p. 137. mentepe. 4 Ward 1910, p. 363, fig. 1246; Van Buren 1946, Cylinder, Old Van Buren 1946, p. 8. Disproved (Lambert Boehmer 1965, pl. XLVIII, p. 8, ad fig. 8; Akkadian. 1985, p. 92, n. 17), fig. 566, n. 1308 (HSM 1195). Frankfort, Jacob- though the seal im- sen 1955, p. 49, pl. pression or the seal 61, n. 649; Boehmer might have been 1965, pl. XLIII, modified after the fig. 567 (Tell Asmar owner’s assumption 33.104). of a new role (Hallo 2001, p. 246). 5 Collon 1982, pl. XLV, n. 382. Delaporte 1920, 18 Cylinder, Ur III. Waetzoldt 1989, though the impres- Not confirmed, see (T. 155) (AO 4213). sion has not been published yet. Hallo 2001, p. 247.

6 Moorey, Gurney 1973, p. 72, pl. Yıldız, Waetzoldt Cylinder, Ur III. Mayr, as reported in Hallo 2001, Not confirmed. XXXIV, 4. 1994, p. 900. p. 250, affirms that one cannot entirely dismiss the possibility that the seal was the source of the impression mentioned. 7 Mayr 2005, n. 85. Owen 1991, p. 230. Cylinder, Ur III, D. I. Owen, as reported in the CDLI Not confirmed. from Umma. webpage mentioned in note 8. 8 Eisen 1940, pl. 13, n. 128; Speleers 1943, Cylinder, Ana- For Speleers 1943, pp. 169-170, n. Confirmed, but Özgüç 1986, p. 51, figs. 4-6 pp. 169-170, n. 660 tolian. 660: Buchanan (1965, p. 204, n. 4), impressions have (L55.49.50 – Moore Collection). and Matouš 1962, who, however, admits the possibility been made in recent Ka 662. of a modern impression. times (Leinwand For Matouš 1962, Ka 662: N. Özgüç 1984, pp. 119, 150; (1986, p. 50), who likewise suggests Özgüç 1986, pp. 50- that the impressions on the unin- 51, 53, n. 15). scribed tablet from Kültepe have been made in recent times (1986, p. 53, n. 15). 9 Porada 1948, pl. 51, n. 347; Hrozný 1952, pl. Cylinder, Old Özgüç (1986, p. 50), though she ad- Not confirmed. Özgüç 1986, 50, fig. 4.5. LXXIII, 50B. Babylonian mits that only an examination of the from Kültepe. pieces side by side could determine whether they are in fact identical.

West & East 224 Monografie, 4 Matching Near Eastern seals and sealings: current issues and new perspectives

Match supposed N. Seal Seal impression Type Authenticity /proposed by

10 Meijer 1989, p. 11, fig. 6. Meijer 1989, p. 11 Cylinder, Sy- Meijer (1989, p. 11), but the bulla is Not confirmed. (bulla). ro-Anatolian fragmentary and the match is at best (see Mazzoni probable, not certain (Hallo 2001, 1975) from p. 249). Hammam al Turkman on the Balikh 11 Parrot 1966, pp. 333-335; Po- Parrot 1959. Cylinder, Clas- Parrot 1966, pp. 333-335; Buchanan Confirmed, but the rada 1980, p. 34, fig. I-14. Pp. 169-185, pl. sic Syrian. 1970, p. 53; Porada 1980, p. 17. seal has its inscrip- XLVIII, 71-73, 81b. tion re-carved since it was acquired by another owner. 12 At first it was mistaken with: Reichel 2003 (As. Cylinder, Old Sollberger 1965, pp. 29-30, n. 26; Confirmed. Moortgat 1940, n. 254; Frayne 30: T. 643; As. 30: Babylonian. Franke 1977; Hallo 1983, p. 12, n. 1990, p. 518 at Berlin Museum. T. 650; As. 31: T. 10; Collon, Geller 1983, p. 205; Po- The correct reference is: Jacob- 256). rada 1983, p. 476, n. 1; Hallo 2001, sen 1940, p. 145, n. 12; Frank- pp. 245, 252; Reichel 2003. fort 1940: pp. 215-216; Frank- fort, Jacobsen 1955, p. 50, n. 709; Franke 1977, pp. 63-64; Frayne 1990, pp. 489-490 (As.30: 1000 – A.7468, Oriental Institute Chicago) 13 Seal in the collection of Mr. and Matthews 1992, Cylinder, Kas- Mayr in Hallo 2001, p. 251, n. 43. Not confirmed. Mrs. Jonathan Rosen (New p. 12. site. York); see Mayr in Hallo 2001, p. 251, n. 43. 14 Grayson 1987, pp. 280-281. Wiseman 1958, Cylinder, Kas- Wiseman (1958, pp. 19-22), but “this Not confirmed. pp. 19-22. site [possibility] must remain in the realm of hypothesis” (Hallo 2001, p. 252). 15 Delaporte 1910, p. 173, pl. Hallo 1973, p. 181 Cylinder, Mid- Hallo (1973, p. 184), who, however, Doubtful. XXI, n. 307; Frankfort 1939, (LB 2532). dle Assyrian. considers more likely that the pl. XXXI, j; Moortgat 1942, seal was not the one making the pp. 70‑73, fig. 46. impression, though owned by the same user. 16 Radner 2012, pp. 687-688, figs. Radner 2012, Stamp, Radner 2012, pp. 687-690. Confirmed. 1-2 (BM 2002-05-15, 1). pp. 688-690, figs. Neo-Assyrian. 3-5 (BM 84671; BM 84553; K. 348+Ki. 1904-10‑9, 246) For BM 84671 see also Reade 1987, p. 145, fig. 7; Herbordt 1992, pl. 32, 3 For K. 348+Ki. 1904-10-9, 246 see also Fales, Post- gate 1992, p. 94; Herbordt 1992, p. 252, Ninive 203, pl. 14, 13.

West & East 225 Monografie, 4 Melissa Ricetti

to regard this seal as the one that produced the im- though it might well have been recovered during pression on LB 2532, but it seems more likely that an illegal excavation or have been looted in antiq- they are in fact two distinct seals probably owned uity as well as in more recent times. by the same Aššur-remanni, who cancelled the in- An Old Syrian seal dated at the beginning of scription on clay or commissioned an uninscribed the second millennium BC and found in Italy in a seal similar to the one at Biblithèque Nationale.15 Roman grave of the third century AD is a further Case 11 is particularly emblematic both for the evidence of seals’ second life after a recovery in the place where the seal turned up and for the change past. The seal’s aesthetic quality and alleged amulet- of inscription and owner experienced by the seal. ic powers made an individual born more than ten The seal originally belonged to a high official of centuries after its production to wish to be buried king Zimri-Lim of Mari, Ana-Sin-taklaku, as the with it. The seal, now preserved at the Archaeologi- impressions of an inscribed seal on many jar seal- cal Museum in Florence, had lost its role as admin- ings at Mari testify: A-na-(il) Sin-ták-la-ku/mâr istrative tool and had become the exotic object to be Da-ri-ish-li-bur/warad Zi-im-ri-li-im, “Ana-Sin- shown off in a context belonging to a different era, taklaku, son of Dariš-libur, servant of Zimrilim”.16 thousands of kilometres far from its original place.21 A seal with identical carving (a king offering game? to a god with scimitar trampling on a little man lying on the ground; a suppliant goddess stand- 3. Obstacles to the search for matches ing behind the god followed by a semi-nude goddess in war-like attire, Ishtar?; three-lines inscrip- The task of matching has been hampered not only tion panel),17 but with the inscription of a certain by past and recent looting, which was certainly fa- Adad-šarrum is now at Louvre Museum after being cilitated by the seal’s small size, nor only by re-carv- bought in Teheran in 1966 (AO 21988, inscrip- ing activities, as has been pointed out in the previ- tion: “Adad-šarrum, son of Šamaiatum, servant of ous paragraph. Although known in thousands of the god Nin-š[ubur]”).18 From an accurate obser- samples, seals remain a quite extraordinary finding vation of the modern impression, it appears clear during excavations in comparison to the number of that the original cuneiform signs in the inscription objects bearing seal impressions. panel have been erased to make way to the signs of Seals were generally carved from hard stones, but the new inscription. The previous small size wedg- some of them were made of organic material such as es emerge in some traces between the larger signs bone, ivory and wood, which are rarely preserved in carved on their place.19 After being used by Ana- archaeological contexts. As prestigious and expen- Sin-taklaku, who probably resided in Upper Syr- sive artworks and representatives of the individual, ia and from there sent his shipments to Mari, the seals were rarely abandoned or lost. If a seal was lost fine Classic Syrian seal ended up by acquisition or or stolen, it was a matter of great concern. The own- inheritance to Adad-šarrum, who understandably er would warn his fellows about the invalidity of did not change the general setting of a fine and el- any impression that would have appeared after the egant seal’s design, but just modified the inscrip- event. Seals are more expected to be found in partic- tion with his name. The way the seal was then ular contexts, such as workshops and graves, which brought to Iran leaves room for most disparate are often targeted by illegal diggers, while are rather speculations. Parrot excludes the seal being stolen,20 fortuitous finds in other places. Owners always carried their seals along,22 even in their afterlife. Many 15 Hallo 1973, p. 184. seals were part of the funerary equipment of the 16 Parrot 1959, p. 254; Parrot 1966, pp. 333-334. 21 About this seal, see Felli 2013, pp. 329-331, n. 26. 17 Porada 1980, p. 30. 22 E.g. Before leaving for the imminent destruction of the 18 Parrot 1966, p. 334. city, the inhabitants from Kültepe/Kaniš apparently had been 19 Porada 1980, p. 17. able to take along their seals and other valuable possessions 20 Parrot 1966, pp. 334-335. (Veenhof 2008, pp. 56-57).

West & East 226 Monografie, 4 Matching Near Eastern seals and sealings: current issues and new perspectives

buried: most of the unprovenanced seals probably tunately only few archaeological remains have been come from robbed graves. identified for that period.23 Likewise, official docu- If excavations do not yield many actual seals, the ments and diplomatic gifts exchanged between dis- opposite happens to sealings, which not surprising- tant royal courts were put under their senders’ seal: ly constitute a substantial part of the archaeological it is the case of the first seal of Zimri-Lim of Mari24 remains. Though these artefacts have not been saved impressed on two envelope fragments found at Tell from illegal digging activities and looting either, the al Rimah,25 which probably enclosed a letter for phenomenon was partly balanced by their abundan- Hatnu-rapi, king of Qattara;26 and of the impres- cy within the archaeological layers. Their number sion of the seals of Aplaḫanda from Carchemish, basically depends on two reasons: statistic (one seal of Yaḫdun-Lim’s daughter Nagiḫanum from Mari, produce countless sealings) and profit (clay artefacts and their servants found at Acemhöyük, together have lower economic appeal than seals). In archaeo- with several impressions of Šamši-Adad’s seals.27 logical researches, accidentalness plays a role as well. Other problems arise from the possibility that Sealings were made of both wet and dried clay that some seals were not used for making impressions – in many cases were baked by conflagrations destroy- they could have been exclusively conceived as jew- ing ancient palaces and cities. Baked clay artefacts, ellery items – or used for decorative purposes with though subjected to crack or break, become strong painting on perishable material, such as textile or and durable to most environmental conditions and leather, as the many stamp seals from Çatal Höyük, are generally well preserved even after millennia. The no one found impressed on clay, would suggest.28 custom of sealing clay objects in daily life activities All these claims are certainly important and such as trade, administration and communication clearly affect the search for matches, but I would go produced masses of sealings that have been often further following Waetzoldt’s observations, which found in administrative offices, public and private offer a statistic explanation to the phenomenon. storerooms and archives or, once useless, discarded Waetzoldt asserts that, in spite of a very large num- in pits and dedicated rooms. ber of scribal seals (20.000 – 30.000 estimated), at As a result we have, on the one side, a large num- the time he wrote, not much more than a hundred ber of seals that often lack any archaeological infor- of them are published. Moreover, text editions often mation and, on the other, a larger number of seal- do not provide the drawings of the seals impressed.29 ings that have been mostly found in context. The He then claims the discrepancy between poten- balance between seals and sealings is further affected tially available and actually available data. Still un- by the above frequently stressed practice of re-sur- known matching seals and impressions might there- facing seals: making the previous design no more fore had been already retrieved from the soil and lay identifiable inevitably cuts down the number of ex- unnoticed among thousands of samples waiting for tant seals. publication. Moreover, the dismemberment of col- Further congenital obstacles to the matching lections and the silent acquisition of samples by un- process must be then considered. One is the dislo- known private collectors have caused some seals and cation of seals and their impressions, which means that the search for matches must expand the lim- 23 Veenhof 2008, pp. 38-41. its of the context, site, region, and area from where 24 Beyer 1985, p. 378. the sample under study comes, requiring a big effort 25 Dalley, Walker, Hawkins 1976, p. 250, n. 5, from the scholar. Within the wide network cover- pl. 107, seal 5. ing the ancient Near East, merchandise and docu- 26 Hallo 2001, pp. 248-249. ments could be sent far from the place where seal- 27 Özgüç 1980, pp. 61-71. The nature of the items ac- ings were applied. Many shipments and letters sent companied by these seal impressions is not clear. Bullae marked by royal seals could be attached to commercial merchandise ei- to merchants in Kaniš during the Old Assyrian pe- ther. riod bear impression of seals that should be seek for 28 Türkcan 2013, p. 245. in Assur, one thousand kilometres far, where unfor- 29 Waetzoldt 1989.

West & East 227 Monografie, 4 Melissa Ricetti

sealings to end up in collections that have very few merely give a clue to their appearance. Line draw- chances to be published.30 ings are rarely provided, given the good legibility of The prospect of finding matches thus largely most modern impressions. depends on the capacity of scholars of making the Seal impressions are usually published in black enormous glyptic dataset – sometimes still inacces- and white flatbed or camera pictures where carving sible – available to the public and of surveying col- details are generally more distinguishable than in lections in the most systematic way: an issue that is colour. Visual efficacy varies according to picture’s closely related to the way glyptic samples are collect- size and quality and is highly affected by the state ed, classified and reported within catalogues that of- of conservation of the sample. Pictures are seldom ten do not provide efficient search conditions, as accompanied by handmade line drawings that help will be better pointed out in the next paragraphs. the untrained eye to the comprehension of the seal Accordingly, the following project aims at mak- impression’s design. Some catalogues provide line ing the search for matches within the dataset offered drawings but no pictures of the impression, so that by Kültepe/Kaniš a less fortuitous and laborious we must necessarily rely on the personal evaluation task trough a new cataloguing approach. of the drawing’s author. Different interpretations are not unusual: the low legibility of the surface or the incompleteness of the impression seldom result- 4. The matching process ed in different drawing solutions. Seals’ designs are generally described by a short Ideally the matching is a very easy operation. By text that is very helpful for deciphering representa- browsing among pictures and line drawings of seals tions in presence of unfamiliar motifs and for clari- and seal impressions published in catalogues and ar- fying details that do not clearly emerge on pictures chaeological reports or, when possible, by browsing and drawings. among material in excavations and museums store- The recourse to such mediated forms (i.e. flatbed rooms, extant seals are identified as the ones that and camera pictures and line drawings) can some- produced at least one known impression. how reduce glyptic research perspectives. Modern The process may be described in three phases: impressions serve quite well art historical and epi- 1) Seal or impression design acquaintance; graphic analyses, but can be misleading for an in- depth study on the way craftsmen worked with 2) Searching; the material and on how seals were originally per- 3) Matching. ceived.31 Likewise, composite seal’s designs do not reflect the way impressions appear on tablets or other objects and do not represent any authentic use 4.1 Seal or impression design acquaintance of seals. Nonetheless any image support is always a great advantage over nothing at all. Seals and impressions are published in several printed and online catalogues and differ in the way they 31 Two examples offered by Wayne T. Pitard, direc- are catalogued and imagined. tor of Spurlock Museum, show how the seal’s carver adapted Seals are generally displayed with colour or black the natural morphological features of the stone to his decora- and white camera pictures of the present-day lab- tive purposes. In a Persian seal preserved at Spurlock Museum (1900.53.0052A) the artist inserted the representation with- oratory impression, occasionally accompanied by in the wedge created by the banded waves of the agate stone, shots of the seal from one or different angles, which which appear as the rays of the winged sun carved at the bottom. Similarly, on a Neo-Assyrian seal (1900.53.0097A), the 30 A few private collections, such as the Frida Hahn and sacred tree has been carved on a hematite red patch emerging the Rudolf Blanckertz Collections, disappeared: the first was on the white agate of the surface that highlights it as the focus of auctioned at Sotheby’s and tablets are now scattered in differ- the scene. These artistic expedients would go completely lost by ent countries, while for the second any trace has been lost. Old- looking only at the modern impression of the seal, rather than er collections have been frequently sources of later ones and oc- at the seal both as artwork in itself and tool for making impres- casionally names of their former owners were not revealed. sions (Pitard 2014).

West & East 228 Monografie, 4 Matching Near Eastern seals and sealings: current issues and new perspectives

Nowadays new imagining techniques adopt- largest number of terms of comparison. Published ing laser scanning and Reflectance Transformation collections are increasing in number, but a substan- Imaging (RTI) offer more satisfying solutions for tial part of them are still not or not sufficiently ac- the recording and publication of data, such as 3D cessible and often not well served by current photo models and digital rollouts of seals, which allow to documentation. Many museums offer free access to enjoy, together with designs, aspects of seals and their online databases. Despite excellent examples impressions that are usually neglected (i.e. colour, such as the British Museum and the Penn Museum, texture, grain, translucency of the stone, position of which give access to a rich image apparatus and al- the design on the seal surface and on the impressed low a good control over search criteria, museum on- object, etc.).32 line publication and inspection still appears as a long path to be accomplished.34 An important dataset is also offered by the project “Mesopotamian Seals” 4.2. Searching carried out within the Cuneiform Digital Library Initiative (CDLI), which, though still poor of im- Once the seal’s or impression’s design has been as- ages, catalogues more than 30.000 glyptic artefacts similated, we can advance to the second phase of the from some of the major world collections.35 matching process. Searching for the match means In the light of these figures, what can be done that designs have to be juxtaposed to thousands of for making the search faster and more effective? already known seals or impressions with the aim Though descriptions often rely on subjective inter- to find their exact counterpart. Considering the pretations, they offer an invaluable contribution, in number of samples to be compared with (tens of connection to museum databases which include one thousands if we sum glyptic artefacts of only major or more queryable description fields. When look- collections)33, the search is an extraordinary time ing for a specific feature, “description” parameters consuming task. Priority may be given to groups of can be entered and results filtered accordingly. El- samples somehow related to the design under exam ementary and fast, but is it really effective? Digital (i.e. those from the same certain or alleged site or re- catalogues tend to the simplicity and the unity of gion, though we have seen above that seals’ mobility language, but there is always a certain degree of ar- can be unpredictable) or to collections particularly bitrariness in adopting a certain vocabulary rather rich of samples in the same style. than another. Our enquiries might therefore not let Necessary but not sufficient condition is the pub- emerge data that in a differently set database would lication of the emerged material, in order to have the have been included in the results list.

32 I mention here a few initiatives that deserve attention. They all use different methods of reflectance imaging: Inscrip- 4.3. Matching tiFact developed by the West Semitic Research (USC) (Zuck- erman, Lundberg, Hunt 2014) and OCHRE at the Uni- versity of Chicago, as well as the Cultural Heritage Imaging Given the amount of different representations and (CHI), use the PTM/RTI (Polynomial texture mapping/Re- the consistency of certain patterns, it is rather un- flectance Transformation Imaging) approach based on the pi- likely that one would be able to recognize the same oneering work by Mark Malzbender (Hewlett-Packard Labs). Other projects, such as RTISAD/AHRC at the University of design in a seal and in a seal impression confiding Southampton and the GLYPCOL project in Brussels, as well as only on his or her visual memory. Databases al- the CDLI staff at the University of Oxford (Wagensonner 2015) use a RTI system re-edited by the Leuven team with the 34 British Museum Collection online can be browsed by introduction of the Portable Light Dome (Hameeuw 2014). visiting the following link: http://www.britishmuseum.org/re- 33 Some of the largest seals and sealings collections are search/collection_online/search.aspx. Penn Museum Collec- preserved at: the Louvre Museum (Paris), the British Museum tion is reachable from the Digital Penn Museum page https:// (London), the Penn Museum (Philadelphia), the Oriental In- www.penn.museum/collections/. In both cases research op- stitute (Chicago), the Royal Museums of Brussels, the Vorder- tions allow many different query solutions, but images are often asiatisches Museum (Berlin), the Pierpont Morgan Library still lacking. (New York) and the Ashmolean Museum (Oxford). 35 https://cdli.ucla.edu/projects/seals/seals.html.

West & East 229 Monografie, 4 Melissa Ricetti

low to quickly and systematically search among re- same design, but after a less superficial analysis it be- cords and are particularly suitable when looking for comes clear that they are in fact different. matches. Any database, however, is not perfect and The alleged match n. 9 in Table 1 shows anoth- a scrupulous preparation serves to contain any pos- er very common model: two mirrored groups, each sible weakness. Both language and query solutions composed of a bull-man fighting with a human- are tailored according to the nature of the samples headed bull, are at the sides of a vase supported by and to the personal perception of reality of the da- a high and narrow base ending in two curling arms. tabase developer. Moreover, each style and glyptic Between the figures of the first group there is a vase group counts hundreds of similar compositions and between the ones of the second group a ball- which differ for only small details: unlikely the de- staff. At the terminal of the composition, there is a sired record would stand out after the first research two-column inscription panel (dUTU / dA-a) over attempt. a small roaring lion. The same scheme can be seen Even if language ambiguities are limited and in at least two other seals besides Morgan Seal 347, queries are planned for providing as much as pos- WAM 42.0786 and BM 132533, which at first sight sible coherent answers, the first filtering inevitably might be mistaken for the parent seals of impression selects a number of “false positive” in addition to ICK 1, 50B.37 In comparison to the impression, the the desired results, whose entity depends on how seal preserved at British Museum has a different much detailed the request is. Their amount is any- inscription, an extra crescent and misses the small way much smaller, and therefore much more man- lion; weather the seal from Walters Art Museum ageable, than the total number of registered records. differs for the position of the human-headed bulls’ After a few filter attempts for refining research, re- front legs (on ICK impression, one leg is above the sults can be scrutinized and juxtaposed to the sam- other and on WAM seal, one leg is beside the other) ple whose counterpart (i.e. seal or impression) is and tail (on ICK impression, it falls behind the legs missing and with some luck the wanted record will and on WAM seal, it appears in between the legs). eventually come out. Designs seem also diverge in the rendering of the The compositional and iconographical consist- central object (on WAM seal, the terminal arms of ency of certain patterns inevitably enhances the the support slightly lean outwards) and of the small complexity of the search. The group formed by an lion, which, however, is only half preserved on ICK antelope on its hind-legs attacked by a rampant impression (fig. 2: a-c). lion and a griffin is a rather common sub-pattern in When looking for matches between pieces be- Old Babylonian style seals. It appears identical on a longing to distant collections, observations are pri- number of compositions where the other figures are marily made on available visual data. The capacity of also picked out from the same repertory and seem the scholar to determine similarities and differences very similar to each other (fig. 1).36 At first sight one therefore depends on the quality of pictures and de- may feel that compositions derive from exactly the signs. Impression ICK 1, 50B has been published as line drawing and no picture is available. Only an ex- 36 The group appears almost identically on impression kt amination of the clay impression side by side with 90/k 96D and seal BM 89124. See also seal BM 141643, where the seal would determine whether the above detect- the naked hero assumes the same posture as on impression kt ed differences are in fact real. 90/k 96D. On seal BM 89322 a kneeling man takes the place of the antelope within the group formed by the attacking lion and Line drawings are a primary concern in glyptic the griffin, while another sub-pattern formed by an antelope -at studies. Bad quality impressions drawings from old tacked by a rampant lion is juxtaposed to it. The group com- publications often arise doubts on design’s interpre- posed by the kneeling man attacked by the griffin is the same on impression kt 90/k 97G and on Morgan Seal 362, which tations, consequently affecting evaluations on pos- share with seal BM 22434 the iconography of the kneeling na- sible matches. Pictures are therefore highly recom- ked hero on a dais. BM 22434, in turn, has the group formed by mended to be included in publications. the hero subjugating the bull in common with kt 90/k 97G and the antelope attacked by the rampant lion is the same as in BM 89322. 37 Abbreviation for Hrozný 1952.

West & East 230 Monografie, 4 Matching Near Eastern seals and sealings: current issues and new perspectives

Figure 1 Cylinder seals in Old Babylonian style; a: kt 90/k 96D; b: BM 89124; c: BM 141643; d: BM 89322; e: kt 90/k 97G; f: Morgan Seal 362; g: BM 22434. Drawings by the author

c Figure 2 Cylinder seal impression (after HROZNÝ 1952, pl. LXXIII, 50B); b: Morgan Seal 347 (after PORADA 1948, pl. 51, n. 347); c: WAM 42.786 (© The Walters Art Museum, Baltimore)

West & East 231 Monografie, 4 Melissa Ricetti

5. An Iconographic Database for complex syntax of images composed of a multitude Kültepe/Kaniš glyptic corpus of elements which differ from each other in a number of significant details. The corpus therefore con- Database approach is indeed a great advantage for stitutes an ideal dataset for testing a glyptic icono- systematic recording and data handling, but only a graphic database, whose aims are mainly two: to careful strategy and strict methodology make a da- quickly retrace a recorded sample and to systemati- tabase still appropriate even when research purposes cally group records characterized by analogue icon- change. ographic features. Constructing a database is a complex process Similar goals are the basis of the praiseworthy where early stage choices heavily influence results. and ambitious project originated within the Cu- The first task is to select that part of reality that we neiform Digital Library Initiative (CDLI) and car- are to represent on the database and for what aims. ried out by Christina Tsouparopoulou which aims It is impractical to represent the whole reality and at the registration within a unique database of all no model can reflect it completely and accurately discovered seals and sealings. Iconographic descrip- for all purposes. Rather we should aim to hold only tions are provided, but through an intentionally a relatively small part, which will be a good basis for limited vocabulary in comparison to the variables future applications, and to produce the most appro- of each period and tradition. The database is there- priate model according to our scope. The material fore not completely appropriate for the matching selected for this project is the collection of seals and research, though it is an extraordinary attempt to sealings from Middle Bronze Age Kültepe/Kaniš, systematically classify glyptic iconography.38 More which offers one of the richest and diverse corpo- recently, the Institute for Near Eastern Archaeol- ra in Near Eastern glyptic art. More than one thou- ogy of the Ludwig-Maximilians-University Mu- sand and five hundred different sealings have been nich launched a project to develop a “Concept for already published and hundreds (if not thousands) the Digitization of Ancient Near Eastern Seals of them are still waiting for publication. Most of and Sealings (DigANES)”.39 The project has many them are preserved at Museum of Anatolian Civi- points in common with the one presented in this lizations in Ankara, while many others ended up paper (e.g. the creation of a semantic classifications in museum and private American, European and and a standardized vocabulary; the integration of Turkish collections before Turkish official excava- visual, material and textual data). Its goal, howev- tions started in 1948. In publications there is infor- er, is still more challenging since the dataset collects mation of roughly one hundred and twenty seals all corpora of Anciet Near Eastern seals and seal- recovered during excavations, though the actual ings. The work is still in progess, but once accom- number is certainly higher. In addition to them, plished, it will be of great benefit for the study of groups of so-called “Cappadocian” seals populate Ancient Near Eastern culture as a whole. A simi- many important seal collections all over the world lar approach, though on a different type of mate- and, though their origin is not certain, they prob- rial, was also developed within the “Corpus der ably come, if not from Kültepe, at least from one of minoischen und mykenischen Siegel (CMS)” pro- the other Assyrian centres placed on ancient trade ject, which aims at the systematic documentation routes between Mesopotamia, Syria and Anatolia. and publication of all known Aegean seals and seal- Thanks to their wide trading activity, Assyrian ings in a consistent manner. The project was estab- merchants could become familiar with and appre- lished and carried out in Marburg from 1958 to ciate styles and iconographic motifs typical of oth- 2011, when it was moved to its new home, the In- er regions and traditions. This is the reason why stitute of Classical Archaeology at the University of the sealing corpus from Kültepe includes samples in many different glyptic styles (i.e. Old Assyrian, Anatolian, Old Syrian, Ur III, Isin-Larsa, Old Bab- 38 Tsouparopoulou 2014. ylonian). Its iconographic repertory is formed by a 39 https://www.diganes.gwi.uni-muenchen.de/.

West & East 232 Monografie, 4 Matching Near Eastern seals and sealings: current issues and new perspectives

Heidelberg.40 A few important issues for the cur- functional features that are distant from today per- rent project have been developed thanks to a careful ception. In their coding effort, scholars must reduce comparison with their methodologies. to a minimum the risk of symmetrisation with their The system chosen for classifying samples of own experience and knowledge, as well as avoid at- the sealing corpus from Kültepe bases on textual taching to images aprioristic values consolidated in description of the designs and on a flexible model past literature. which allows different and multiple combinations Textual coding concerns only the iconographic of research entries.41 Each seal’s or impression’s de- aspects of the representation, which are considered sign is segmented into its minimum significant ele- according to their relative spatial position within ments which are converted into text through fixed the scene. linguistic codes that univocally define them. Defini- Figures and objects are identified in terms of tions are ordered in sequence and formalized into genre, posture, gesture, and formal features (e.g. at- complete sentences forming overall descriptions tributes, hairstyle, garment, etc.). Each element is that can be filtered and queried according to the carefully described balancing the advantages of syn- presence or absence of specific codes. Language is thesis against a purely objective codification which specifically tailored to suit the corpus under study. would result in an excessive number of details. Same The elaboration of common codes results from a elements and attitudes must be described constant- critical research for the most appropriate form and ly using the same expressions without any ambiguity term and inevitably requires some progressive ad- or redundancy (fig. 3). justments. Linguistic choices depend on the per- Stylistically descriptions follow a few fixed rules. sonal evaluation and preparation of the scholar. In order to maintain textual integrity, terms appear The identification of single icons and sub-patterns declined and conjugated while, when needed, their (i.e. sets of icons which occur identically in differ- not-inflected form follows in brackets, so that the ent designs) requires an in-depth knowledge of the correspondence between data recorded and search glyptic lexicon and of its figurative organization, entries is bi-univocal. in particular, when attempting to understand the The iconographic search function has been meaning each element acquires according to its in- thought as an implementation of an already exist- teraction with the others. Elements have formal and ing relational database which links seal impressions to sealers and presents three interrelated sections: 40 https://www.uni-heidelberg.de/fakultaeten/philoso- the first devoted to sealed documents, the second to phie/zaw/cms/. sealers and the third to seal impressions.42 Sections 41 A few quantitative studies made on glyptic samples are all linked together, so that one can immediate- adopt a coding method for seal’s designs based on linguistic ly pass from the object where the seal impression principles. First developments date back to the 70s of the last century, where a broad sample of seals was measured by com- is displayed, to the individual who made it, and to puterized attributes in order to get better iconographic descrip- the impression itself. In most cases impressions are tion and classification Digard( et Al. 1975). Further studies not directly linkable to the user/owner of the parent appeared on the following decades addressing the analysis to the combination of different glyptic elements and to the composi- seal, since no label is provided next to the impres- tion as a whole, rather than dealing only with isolated parts of sion. It is therefore essential to recognize impres- the repertory (Kelly-Buccellati 1977; Matthews 1990). sions made by the same seal in order to identify its Scholars’ efforts were limited by insufficient coordination and little following within the field. Though promising, quantita- user/s through a cross comparison of sealers’ names tive data analyses are a difficult challenge. In the last years, the on different documents.43 mathematical and statistical approaches by Rova, Camiz and Di The seal impression’s search mask can be opened Ludovico (Camiz, Rova 2003; Di Ludovico, Camiz 2015) adopted a coding method for seal’s designs based on linguistic from the section reserved to seal impressions and of- principles, whose preliminary assumptions have some points in 42 common, though different purposes and general method, with Thedatabase has been created in Microsoft Access. the present project. At the moment a quantitative approach Pictures and line drawings are visible in the interface, though would be impracticable for the corpus under exam, because its stored separately as images. variety would exceed the method’s capabilities. 43 Ricetti 2014, pp. 71-74.

West & East 233 Monografie, 4 Melissa Ricetti

Figure 3 Microsoft Access database screenshot: the user interface of the seal impressions section

fers various query solutions: if we already know the The most interesting function of the search impression we can find the correspondent record by mask is the possibility to search for seal impressions entering its catalogue number or one of the refer- according to their designs. Search entries can be se- ences made in a secondary publication. Abbrevia- lected from a list of pre-set iconographic keywords tions and bibliographic standards are displayed in a and/or by entering textual codes in a number of free table accessible via hyperlink. The “type” and “style” search textual fields (fig. 4). fields allow filtering seals according to their shape Keywords are ordered into general categories, (i.e. cylinder, stamp, and substitute44) and carving such as “animals”, “deities”, “objects”, “astral sym- style, whose definition often oversimplifies real- bols”, etc., which collect the most recurrent el- ity for categorization benefits. If the seal is already ements of the composition. They are expressed identified or presents an inscription we can filter through textual codes and are constantly updated the correspondent data by entering part of the us- with the new elements emerging during the data- er’s name or part of the inscription. base populating process. Single entries can be selected through a combo box where both categories 44 Seal substitutes are objects that are not seals (or not and elements are alphabetically sorted. The num- properly used as seals), but left their impression on clay as a consequence of an intentional sealing action. Most common are ber of keywords that can be entered at once has rings, earrings, and textiles. been fixed to twelve, while the free search textual

West & East 234 Monografie, 4 Matching Near Eastern seals and sealings: current issues and new perspectives

Figure 4 Screenshot of the seal impressions search mask

fields have been limited to six, allowing different chive excavated at Kültepe/Kaniš (kt 90/k), which and multiple search combinations according to the display 214 impressions made by at least 191 differ- research requirements. ent seals45. Samples are still unpublished and textu- Unlike keywords, free search fields allow artic- al descriptions and drawings have been developed ulated queries, where not only single elements, but from the direct observation of the original ancient also their interaction can be used as a research pa- artifacts. rameter. Actions as well as figures often recur with- At this stage, the iconographic search mask has in glyptic compositions and, though they can con- been proved particularly helpful for the identifica- cern different characters and elements, they are tion of already known impressions and the paring expressed by identical patterns. The group of figures with partial or damaged impressions of the corpus, where a human or semi-human character subjugates allowing to quickly find out whether an impression an animal or a fantastic creature by forcing it upon a new object is already catalogued or not.46 side down and stepping with one feet on its head is a Before launching the database online, I plan to very common pattern within this corpus and can be enlarge the data set to the whole corpus of “Cappa- found by entering the correspondent action verb in docian” glyptic samples and to implement the mod- present tense (in this case, subjugate) in one of the el with a section devoted to extant seals from both free search fields. There are many other examples of Kültepe excavation and collections. Descriptions recurring actions depicted on seals: movement ac- will base on analyses and interpretative drawings tions, gestures, interaction with objects, with ani- made by different scholars. Any bias due to literary mals, etc. (Figs. 5-6). style will be eliminated during the recording process A tip box suggesting the correct entering of the and the iconographic content translated into the search criteria can be opened by clicking on the info database’s standard set of linguistic expression. symbol on the top right of the search mask. 45 The resultant query will provide only samples Ricetti forthcoming. 46 that meet all the fields entered. Results are shown Each impression must be registered under a unique identifier and identifiers must be linked to the same design, in a table which shows the information summary of when impressions are produced by the same seal. Circumstanc- the seal impression and its line drawing. es such as re-uses and transfers of seals, as well as the identifi- At the moment, the database collects data be- cation of the seal’s owner/user, can be hampered by the attribution of unconnected identifiers to impressions made by the longing to a lot of tablets yielded from a single ar- same seal.

West & East 235 Monografie, 4 Melissa Ricetti

Figure 5 Screenshot of the search mask results according to the free search of the verb “subjugate”

Figure 6 Screenshot of the search mask result according to the combination of the search by the keyword “male deity” and the free search of the verb “stand” and of the expression “on the back of a bull”

West & East 236 Monografie, 4 Matching Near Eastern seals and sealings: current issues and new perspectives

6. Conclusion the group under exam, as is the one here proposed. Once all critical points are solved and a common Going back to the question “why only a few registration procedure set up, all databases can be matches have been detected so far?”, to the above joined into a common network and eventually suggestions, I would add another explanation. Some merge into a single general model. matches might well be part of the evidence already Besides the benefits related to the identification emerged, but we are not able to identify them, either of parent seals and ancient impressions, such a data- because they are not accessible (a general mapping base approach would offer many advantages: of the existing seals and sealings collections is – It would offer the opportunity to get in just a unfortunately lacking) or because the search process click iconographic, textual and technological is not effective enough. data sorted according to the request made. A meticulous iconographic classification ex- – It would allow to quickly and univocally link tensively adopted, based on common standards impressions made by the same seal. and provided with both visual and textual infor- – It would provide some clues on the seal’s story mation may be a first solution for a rational man- life: the circumstance of its use, its primary owner, agement of the evidence at our disposal. A model its secondary user, etc. that would be suitable for all Near Eastern glyptic traditions can be seen as the final goal of a long The potential for research improvements is very adjustment process. Given the peculiarities of each high for both scholars who usually work on glyptic glyptic group and their abundance, I consider art and for those who usually do not and might more realistic a system of local databases, each one need information that would be otherwise difficult set on specific research parameters according to to find.

West & East 237 Monografie, 4 Melissa Ricetti

BIBLIOGRAPHY

Beyer D. 1985, Scellements de portes du palais de Mari, Overview, in: Giligny F., Djindjian F., Costa «MARI» 4, pp. 375-384. L., Moscati P., Robert S. (eds.), CAA2014 st Boehmer R. M. 1965, Die Entwicklung der Glyptik 21 Century Archeology. Concepts, Methods and nd während der Akkad-Zeit, Untersuchungen zur Tools. Proceeding of the 42 Annual Conference on Assyriologie und Vorderasiatischen Archäologie 4, Computer Applications and Quantitative Methods in Berlin. Archaeology, Oxford, pp. 29-34. Boehmer R. M. 1997, Einige ältere, in jüngeren Zeiten Digard F., Abellard C., Bourelly L., Deshayes wieder benutzte altorientalische Siegel, in: Pongratz- J., Cardin J.-C., Le Maitre J., Salomé N.-R. Leisten B., Kühne H., Xella P. (eds.), Ana šadî 1975, Répertoire analytique des cylindres orientaux Labnāni lū allik. Beiträge zu altorientalischen und publiés dans des sources bibliographiques éparses (sur mittelmeerischen Kulturen. Festschrift für Wolfgang ordinateur). Röllig, Neukirchener Verlag Neukirchen-Vluyn, Eisen G.A. 1940, Ancient Oriental Cylinder and Other pp. 23-41. Seals with a Description of the Collection of Mrs. Buchanan B. 1965, A Dated “Persian Gulf” Seal and Its William H. Moore, Oriental Institute Publications Implications, in: Güterbock H. G., Jacobsen Th. 47, Chicago. (eds.), Studies in Honor of Benno Landsberger on His Esin U. 1985, Some Small Finds from the Chalcolitic Seventy-fifth Birthday, April 21, 1965, Assyriological Occupation at Değirmentepe (Malatya) in Eastern Studies 16, Chicago, pp. 204-209, pl. XVI. Turkey, in: Liverani M., Palmieri A., Peroni R. Buchanan B. 1970, Cylinder Seal Impressions in the (eds.), Studi di Paletnologia in Onore di Salvatore M. Yale Babylonian Collection Illustrating a Revolution Puglisi, Roma, pp. 253-263. in Art c. 1700 B.C., «The Yale University Library Esin U. 1989, An Early Trading Center in Eastern Gazette» 45, pp. 53-65. Anatolia, in: Emre K., Mellink M., Hrouda B., Buchanan B. 1981, Early Near Eastern Seals in the Yale Özgüç N. (eds.), Anatolia and the Ancient Near Babylonian Collection, New Haven and London. East. Studies in Honor of Tahsin Özgüç, Ankara, pp. 135-142. Camiz S., Rova E. 2003, Quantitative Study of Images in Archaeology: I. Textual Coding, in: Schader M., Esin U., Harmankaya S. 1987, 1985 Değirmentepe Gaul W., Vichi M. (eds.), Between Data Science (Malatya Imamlı Köyü) Kurtarma Kazısı, in: VIII. and Applied Data Analysis, Berlin, pp. 624-632. Kazı Sonuçları Toplantısı I (Ankara, May 26-30 1986), pp. 95-137. Cassin E. 1960, Le Sceau: Un fait de civilisation dans la Mésopotamie ancienne, «Annales, Economies, Fales F.M., Postgate J. N. 1992, Imperial Admi Societes, Civilisations» 15, pp. 742-751. nistrative Records, Part I: Palace and Temple Administration, State Archives of Assyria 8, Helsinki. Collon D. 1982, Catalogue of the Western Asiatic Seals in the British Museum: Cylinder Seals II, London. Felli C. 2013, La glittica e altri oggetti di artigianato in pietra, in: D’Agostino A., Felli C., Valentini S. Collon D. 1987, , London. First Impressions (eds.), La Collezione orientale del Museo Archeologico Collon D., Geller M. J. 1983, Review of Buchanan Nazionale di Firenze, Volume II. Anatolia, 1981, «BO» 40, pp. 204-210. Mesopotamia, Iran, Collana di Studi Mesopotamici Dalley S., Walker C. B. F., Hawkins, J. D. 1976, The 5, Firenze, pp. 301-401. Old Babylonian Tablets from Tell al Rimah, London. Franke J. A. 1977, Presentation Seals of the Ur III/ Delaporte L. 1920, Catalogue des cylindres, cachets et Isin-Larsa Period, in: Gibson, Biggs (eds.) 1977, pierres gravées de style oriental du Musée du Louvre. I. pp. 61-66. Fouilles et Missions, Paris. Frankfort H. 1939, Cylinder Seals. A Documentary Di Ludovico A., Camiz S. 2015, Art History of the Essay on the Art and Religion of the Ancient Near East, Ancient Near East and Mathematical Models. An London.

West & East 238 Monografie, 4 Matching Near Eastern seals and sealings: current issues and new perspectives

Frankfort H. 1940, in: Frankfort H., Lloyd S., Kjaerum P. 1983, Failaka/Dilmun. The Second Jacobsen Th. (eds.) 1940, pp. 201-243. Millennium Settlements 1/1. The Stamp and Cylinder Frankfort H., Jacobsen Th. 1955, Stratified Seals, Aarhus. Cylinder Seals from the Diyala Region, Oriental Lambert W. G. 1985, The History of the muš-ḫuš in Institute Publications 72, Chicago. Ancient Mesopotamia, in: Borgeaud Ph., Christe Frankfort H., Lloyd S., Jacobsen Th. (eds.) 1940, Y., Urio I. (eds.), L’animal, l’homme, le dieu dans le The “Gimilsin” Temple and the Palace of the Rulers Proche-Orient ancien. Actes du Colloque de Cartigny, at Tell Asmar, Oriental Institute Publication 43, 1981, Leuven, pp. 87-94. Chicago. Larsen M. T. 1977, Seal Use in the Old Assyrian Period, Frayne D. 1990, Old Babylonian Period (2003-1595 in: Gibson, Biggs (eds.) 1977, pp. 89-105. BC), The Royal Inscriptions of Mesopotamia. Early Lassen A. W. 2012, Glyptic encounters: A stylistic and Periods 4, Toronto. prosopographical study of seals in the Old Assyrian pe- Gibson M., Biggs R. D. (eds.) 1977, Seals and Sealing riod – chronology, ownership and identity, Phd disser- in the Ancient Near East, Bibliotheca Mesopotamica tation, University of Copenhagen. 6, Malibu. Leinwand N. W. 1984, A Study of Anatolian Grayson A. K. 1987, Assyrian Rulers of the Third Weathergods of the Old Assyrian Period, Ph.D. Thesis, and Second Millennia BC (to 1115 BC), The Royal Bryn Mawr. Inscriptions of Mesopotamia: Assyrian Periods 1, Mackay, E. J. H. 1938, Further Excavations at Mohenjo- Toronto. Daro, Delhi. Hallo W. W. 1973, The Seals of Aššur-remanni, in: Beek Matouš, L. 1962, Inscriptions cunéiformes du Kultépé II, M.A., Kampman A.A., Nijland C., Ryckmans J. Prague. (eds.), Symbolae biblicae et Mesopotamicae Francisco Matthews D. M. 1990, Principles of Composition in Mario Theodoro de Liagre Böhl dedicatae, Leiden, Near Eastern Glyptic of the Later Second Millennium pp. 180-184. B.C., Freiburg. Hallo W. W. 1983, “As the Seal upon Thine Arm”: Matthews D. M. 1992, The Kassite Glyptic of Nippur, Glyptic Metaphors in the Biblical World, in: Orbis Biblicus et Orientalis 116, Göttingen, Gorelick L., Williams-Forte E. (eds.), Ancient Vandenhoek and Ruprecht. Seals and the Bible, Malibu, pp. 7-17, pl. xii. Mayr R. H. 2005, Seal Impressions on Tablets from Hallo W. W. 2001, Seals and Seal impressions, in: Umma. Unpubl. Ms. Hallo W. W., Winter I. J. (eds.), Seals and Seal Mazzoni S. 1975, Impressions. Proceedings of the XLVe Rencontre Tell Mardikh e una classe glittica siro- , «Annali dell’Istituto Assyriologique Internationale, Part II, Bethesda, anatolica del periodo di Larsa pp. 239-254. Orientale di Napoli» 35 (n.s. 25), pp. 21-43. Meijer D. J. W. 1989, Hameeuw H. 2014, Portable Light Dome System, from Hammam al Turkman on the , registration to online publication within the hour: sta- Balikh: First Results of the Fifth Campaign (1988) «Akkadica» 64-65, pp. 1-12. tus quaestionis Portable Light Dome project for cuneiform documents, online publication: https:// Moorey P. R. S., Gurney O.R. 1973, Ancient Near portablelightdome.files.wordpress.com/2014/12/ Eastern Seals at Charterhouse, «Iraq» 35, pp. 71-81, portable-light-dome-system-from-registration-to- pls. xxxiv-xxxv. online-publication-within-the-hour_1-1.pdf. Moortgat A. 1940, Vorderasiatische Rollsiegel: Ein Herbordt S. 1992, Neuassyrische Glyptik des 8.-7. Jh. Beitrag zur Geschichte der Steinschneidekunst, Berlin. V. Chr., State Archives of Assyria Studies 1, Helsinki. Moortgat A. 1942, Assyrische Glyptik des 13. Hrozný B. 1952, Inscriptions cunéiformes du Kultépé I, Jahrhunderts, «ZA» 47, pp. 50-88. Prague. Owen D. I. 1991, Neo-Sumerian Texts from American Jacobsen Th. 1940, Historical Data, in: Frankfort Collections, Materiali per il Vocabolario Neosumerico H., Lloyd S., Jacobsen Th. (eds.) 1940, 15, Rome. pp. 116-200. Özgüç N. 1980, Seal impressions from the palaces at Kelly-Buccellati M. 1977, Towards the Use of Acemhöyük, in: Porada E. (ed.), Ancient Art in Quantitative Analysis in Mesopotamian Sphragistics, Seals, Princeton, pp. 61-99. «Mesopotamia» 12, pp. 41-52.

West & East 239 Monografie, 4 Melissa Ricetti

Özgüç N. 1986, Seals of the Old Assyrian Colony Period Speleers L. 1943, Catalogue des intailles et emprein- and Some Observations on the Seal Impressions, in: tes orientales des Musées Royaux d’Art et d’Histoire, Canby J. V., Porada E., Ridgway B. S., Stech Brussels. T. (eds.), Ancient Anatolia: Aspects of Change and Teissier B. 1994, Sealing and Seals on Texts from Cultural Development: Essays in Honor of Machteld J. Kültepe Kārum Level 2, Publications de l’Institut Mellink, Madison, Wisconsin, pp. 48-53. Archéologique-Historique Néerlandais de Stamboul Parrot A. 1959, Mission Archéologique de Mari II/3. Le 70, Istanbul. Palais. Documents et Monuments, Paris. Tsouparopoulou C. 2014, Progress Report: An Online Parrot A. 1966, Une réapparition mystérieuse, «Syria» Database for the Documentation of Seals, Sealings and 43, pp. 333-335. Seal Impressions in the Ancient Near East, «Studia Pitard W. T. 2014, Circular Signatures: Getting a bet- Orientalia Electronica» 2, pp. 37-68. ter view of Mesopotamia’s smallest art form, «Biblical Türkcan A. U. 2013, Çatalhöyük stamp seals from Archaeology Review» 40/3, pp. 55-59, 70. 2000–2008, in: Hodder I. (ed.), Substantive Porada E. 1948, Corpus of Ancient Near Eastern Seals in Technologies at Çatalhöyük: Reports from the 2000– North American Collections I. The Pierpont Morgan 2008 Seasons, Çatalhöyük 9, British Institute at Library Collection, Bollingen Series 14, Washington. Ankara Monograph 48, Monumenta Archaeologica 31, London and Los Angeles, pp. 235–246. Porada E. 1980, Introduction, in: Porada E. (ed.), Ancient Art in Seals, Princeton, pp. 3-34. Van Buren, E. D.1946, The Dragon in Ancient Mesopotamia, «Or» 15, pp. 1-45, pls. i-viii. Porada E. 1983, Review of Brandes 1979, «JAOS» 103, pp. 476-478. Veenhof, K. R. 2008, The Old Assyrian Period, in: Wäfler M. (ed.), Mesopotamia. The Old Assyrian. Porada E. 1993, Why Cylinder Seals? Engraved Period, Annäherungen 5. Orbis Biblicus et Orientalis Cylindrical Seal Stones of the Ancient Near East, 160/5, Teil 1, Fribourg – Göttingen, pp. 13-264. Fourth to First Millennium B.C., «The Art Bulletin» 75, pp. 563-582. Yıldız F., Waetzoldt H. 1994, Die Umma-Texte aus den Archäologischen Museen zu Istanbul Nr. Radner K. 2012, The Seal of Tašmetum-šarrat, 601‑1600, Materiali per il Vocabolario Neosumerico Sennacherib’s Queen, and its Impressions, in: 16, Rome. Lanfranchi G. B., Morandi Bonacossi D., Pappi C., Ponchia S. (eds.), Leggo! Studies presen- Waetzoldt H. 1989, Original eines Siegels und dessen ted to Prof. Frederick Mario Fales on the Occasion of Abrollung, «NABU» 1989/3, 55 (79). his 65th Birthday, Leipziger Altorientalische Studien, Wagensonner K. 2015, On an alternative way of cap- 2, Wiesbaden, pp. 687-698. turing RTI images with the camera dome, Cuneiform Reade J. E. 1987, Was Sennacherib a Feminist?, in: Digital Library Notes 2015: 1, online publica- Durand J.-M. (ed.), La femme dans le Proche-Orient tion: https://cdli.ucla.edu/pubs/cdln/php/single. antique. Actes de la 33e Rencontre Assyriologique php?id=54. Internationale, Paris, pp. 139-145. Ward W. H. 1910, The Seal Cylinders of Western Asia, Reichel C. 2003, A Modern Crime and an Ancient Washington. Mystery: The Seal of Bilalama, in: Selz G. J. (ed.), Winter I. J. 2001, Introduction: Glyptic, History, and Festschrift für Burkhart Kienast zu seinem 70. Historiography, in: Hallo W. W., Winter I. J. Geburtstage dargebracht von Freunden, Schülern und (eds.), Seals and Seal Impressions. Proceedings of the Kollegen, Alter Orient und Altes Testament 274, XLVe Rencontre Assyriologique Internationale, Part Münster, pp. 355-389. II, Bethesda, pp. 1-14. Ricetti M. 2014, Testi e impronte di sigillo provenienti Wiseman D. J. 1958, The Vassal-Treaties of Esarhaddon, dal livello II del kārum Kaniš: uno studio comparato, «Iraq» 20, pp. 1-99, pls. 1-53. PhD dissertation, Sapienza – Università di Roma. Zuckerman B., Lundberg M., Hunt L. 2014, Ricetti M. forthcoming, Preliminary Results on the InscriptiFact: Internet Access to High-Resolution Seal Impressions from the 1990 Excavation Season at Images of Ancient Inscriptions and Artifacts from the Kültepe/Kaniš (kt 90/k). Near Eastern and Mediterranean Worlds, «Near Sollberger E. 1965, Three Ur-Dynasty Documents, Eastern Archaeology» 77/3, pp. 238-242. «JCS» 19, pp. 26-30.

West & East 240 Monografie, 4 Connecting cultures, dividing countries: a preliminary assessment of the Khabur Ware from Girnavaz Höyük, Southeastern Turkey

SERGIO GIUSEPPE RUSSO

University of Catania and British Institute at Ankara

Abstract

Girnavaz Höyük is a key site for better understanding the cultural and economic relationship between northern Mesopotamia and the Upper Tigris region between the third and the first millennia BC. The site is located in the liminal zone of the northern Jazirah at the start of the Tur ‘Abdin mountains, in a spot that most likely allowed it to play an important role in financial, cultural, and social exchange during the end of the third and the beginning of the second millennium BC when Assyrian merchants began engaging in long-distance trade with Anatolian settlements. During this period, the Khabur Ware (within the Upper Khabur river valley) and other types of Band Painted Wares (from other neighbouring regions) became clear archaeological correlates of this temporal phase. In this paper, I will focus on a preliminary analysis of the Khabur Ware from Girnavaz Höyük in order to define the similarities and differences between contemporary pottery assemblages discovered in the region as well as in other neighbouring areas. The aim will be to provide an initial definition of a relative inner chronology for this specific ceramic assemblage at the site that has never been previously published.

Keywords

Middle Bronze Age, Khabur Ware, Southeastern Anatolia, Upper Khabur Valley, Girnavaz Höyük, band painted ware, relative chronology

West & East 241 Monografie, 4 Sergio Giuseppe Russo

1. Introduction1 Furthermore, recent data from the Upper Khabur river valley also challenges such easy categorisations, The so-called Khabur Ware (hereafter KW) repre- as archaeological evidence reveals that the first ap- sents the most indicative repertoire of the first half pearance of the KW might be dated back to around of the second millennium BC, i.e. Middle Bronze 1950 BC.7 Earlier, P. Pfälzner proposed a controver- Age (ca. 2000-1600 BC, hereafter MBA), within sial hypothesis, also embraced by Schimdt and based Upper Mesopotamia. The first definition and classi- on the discovery of KW sherds within the ‘House fication of this ware was given by Sir Max Mallowan of Puššam’ at Tell Mozan (Phase C7 in inner chro- who, while excavating at Chagar Bazar in the Up- nology), which pushes the dates of the Early Khabur per Khabur river valley, noticed the high percentage Ware back to the very end of the third millennium of these painted vessels within the MBA levels.2 The (i.e., end of the Early Jezirah V period – 2100-2000 following excavations and the related discoveries led BC).8 This view has, however, been rejected by Kolin- scholars to review Mallowan’s work and examine in ski who stated that the sherds from this area do not depth the characteristic of KW, such as its chron- show resemblances with the earlier assemblage dis- ological span and distribution.3 Mallowan himself covered, for instance at Barri and Arbid, but belong distinguished between an earlier stage in which to the Classic Khabur horizon.9 Recently, thanks potters had used coarse buff ware and later periods to 14C analysis from short-lived plants deposited in marked by well-refined pinkish/creamy clay. How- Phase C7 accumulations, Pfälzner confirmed the ever, it was Hrouda who firstly subdivided the KW dates of these contexts to the twenty-first century BC into ‘older’ and ‘younger’, mostly on the basis of and proposed a new periodization of the KW.10 Thus, shape development and type occurrence.4 in his paper he hypothesizes that the first appearance In 1997, archaeologist H. Oguchi refined Mal- of this ware occurred during the last century of the lowan and Hrouda’s works by proposing a more de- third millennium BC. However, this proposal is far tailed periodization. Specifically, he argued for four from being generally accepted and so the debate on sub-periods: Period I (ca. 1900-1813 BC), defined the chronological issue of this pottery remains still also as ‘Early Khabur Ware’ period; Period II (ca. open. 1813-1700 BC) or the ‘Classic Khabur Ware’ peri- As concerns the distribution of the KW, since od; Period III (ca. 1700-1550 BC) or ‘Late Khabur the first publications of this painted pottery schol- Ware’ period; Period IV (ca. 1550-1400 BC) or ars have mainly adopted one criterion for determin- ‘Transitional Khabur-Mitannian’ pottery.5 Despite ing the diffusion of this assemblage, i.e. a qualita- this clear yet approximate chronological classifica- tive approach.11 These studies allowed recognizing tion several problems exist, especially in the first pe- several areas of distribution, in which the sites have riod: the transitional phase between the end of the been listed according to their location in the mod- third and the beginning of the second millennium has so far still produced scarce archaeological evi- 7 Kolinski 2014. According to him, clear examples dence, and is not supported by written sources, thus of Early Khabur Ware dated to the beginning of the twentieth century BC have been found at Tell Barri (stratum 34A) preventing an adequate calibration of the compared and Tell Arbid (Sector P). Moreover, dispersed sherds similar ceramic sequence.6 to those recovered from these sites can be identified at Tell Brak (Area TW), Tell Muhammad Diyab (chartier 5a) and Tell 1 This paper is based on my Master’s thesis titled Typo- Rimah (Area AS – Level A4). logical and Distributional Analysis of Khabur Ware and Band 8 Dohmann-Pfälzner, Pfälzner 2002, p. 154; Painted Wares in the Ancient Near East during the Middle Schimdt 2012, p. 173. Bronze Age: the Case of Girnavaz Höyük, Southeastern Turkey. 9 Kolinski 2014, p. 31. 2 Mallowan 1936; 1937, p. 102. 10 In his paper, the scholar proposes also a new period- 3 E.g., Hrouda 1957, 1989, 2001; Hamlin 1971; Ste- ization of the MBA: MBA IA, 2100-2000 BC; MBA IB, 2000- in 1984; Oguchi 1997, 1998, 2001, 2003, 2006. 1850 BC; MBA II, 1850-1650 BC (Pfälzner 2017, p. 176, 4 Mallowan 1937, p. 102; Hrouda 1957, pp. 38-40. tab. 7.2). 5 Oguchi 1997, pp. 195-196. 11 E.g., Hamlin 1971; Oguchi 1997; Faivre, Nicol- 6 Orsi 2010, p. 864. le 2007; Palmisano 2012.

West & East 242 Monografie, 4 Connecting cultures, dividing countries: a preliminary assessment of the Khabur Ware…

ern countries’ boundaries. In particular, Oguchi – cal dilemma on the appearance and diffusion of KW and successively the other scholars who based their and other band painted wares and on their connec- research on his studies – distinguished a main dis- tions is far from being solved. tribution zone, i.e. the Upper Khabur valley and The study of the unpublished KW sherds from northern Iraq, from a secondary area that included the site of Girnavaz Höyük, located in the Upper sites of southern and southeastern Turkey, western Khabur river valley, may provide further informa- Iran, and inner Syria.12 The spread of this ware over tion on those links. Due to its strategic position at such a large territory led to hypotheses that certain the crossroad between the Upper Khabur river val- relationships existed between the various regions, ley and the Upper Tigris region, the site represents perhaps linked to economic, cultural, and/or socio- an important centre for appreciating the connec- political factors.13 However, archaeometric investi- tions between northeastern Syria and southeastern gations on the MBA band painted vessels have dem- Turkey. Furthermore, a macroscopic and typologi- onstrated that in some peripheral regions, such as cal description of the vessels together with a com- the Upper Tigris valley in southeastern Turkey and parative analysis of the assemblages found in other the Ushnu-Solduz valley in northwestern Iran, the regions will be helpful in establishing an inner rela- KW is locally produced and thus it is better to label tive chronology of the MBA layers. it as Band Painted Ware.14 Indeed, the archaeologi-

12 Oguchi 1997. 2. The site and the history of research 13 Theorieson the diffusion of KW related to economic factors are mainly based on the establishment of the Old As- syrian long-distance trade system, being the Upper Khabur val- Girnavaz Höyük (UTM 37, 698198, 4108872) is ley both the main distribution zone of this painted pottery and located 4 km north of the modern city of Nusaybin, the region of utmost importance for the exchange of goods. C. in the district of Mardin in southeastern Turkey, Hamlin hypothesized that the presence of KW at Dinkha Tepe, in northwestern Iran, might be explained as a consequence of and 5 km north of the modern Syrian-Turkish bor- the role played by the site and the Ushnu-Solduz valley area der. The mound (ca. 300 m in diameter and 24 m within the commercial network (1971, pp. 306-307). H. Ogu- high, ca. 7 ha.) lies on the right bank of the Çağ Çağ chi, aware of the KW occurrence at Kültepe, in central Anato- lia, and in other Mesopotamian areas was much more careful in river, in a strategic position dominating the south- linking it to the Assyrian trade, stating that there were certain ern plain of the Upper Khabur valley and the north- contacts between the regions involved in the network (1997). ern mountainous lands of the Tur ‘Abdin (fig. 1). Instead, theories on the connection between KW and cultural factors suggest that this ware was a product of foreign The site’s proximity to both the Çağ Çağ and the tribal groups or populations, e.g. the Hatti, Hurrians or Am- Tur ‘Abdin might have led the inhabitants of the site orites, that slowly had been insinuated and adapted within the to diversify their economic subsistence, characterized Mesopotamian society (Speiser 1933, p. 273; Welker 1948, p. 218; Mallowan 1956, p. 218; Hrouda 1957, p. 44; by farming in the surrounding fertile lands and by Thomas 1967, p. 69; Faivre, Nicolle 2007, pp. 194-196). pastoral and hunting activities in the mountains.15 In At last, theories based on socio-political factors explain the addition, its position could grant control over the ac- occurrence of KW within a wide interregional context as the consequence of the establishment of the so-called ‘Kingdom of cess to Tur ‘Abdin, making Girnavaz a key-site with- Upper Mesopotamia’ of Shamshi-Adad I. The territory con- in the political landscape of the region. quered by the Amorite king, spanning from the Euphrates to The site was first recorded in scholarly litera- the Zagros, and the alliances with other rulers enlarged the influence sphere of Assyrians throughout Iraq, Syria, Turkey and ture by A.T. Olmstead who mentioned the city part of Iran. Either political or military control of these areas of Naşibina as a prominent settlement during the might have played an important role in the rising of a ‘court’ ninth century BC.16 After almost forty years, E. Lu- ware, represented by KW, highly valued and imitated in order to emulate Assyrian élites (this view has been suggested by K. cius and K. Sornig evoked again the site suggesting Kaercher in a paper presented at the Annual Meeting of the 15 American Schools of Oriental Research held in San Diego, No- E.g., a similar situation has been suggested for the vember 2014). site of Hirbemerdon Tepe, SE Turkey, which is close both to the Tigris river and the Tur ‘Abdin mountains (Laneri, 14 D’Agostino 2016, pp. 81-82. For archaeometric analyses, see Bedal et Al. 1995; D’Agostino 2012; Laneri Schwartz, Valentini 2012). et Al. 2015. 16 Olmstead 1918, p. 218, fn. 22.

West & East 243 Monografie, 4 Sergio Giuseppe Russo

Figure 1 Map showing the location of Girnavaz Höyük and of the main MBA sites of the Upper Tigris and Upper Khabur valleys

that it was located 1 km south of a village, Vesiki, solve this problem of identification, a survey on the which they thought to be the ancient Mitanni cap- mound was planned in 1980. Prof. Hayat Erkanal ital of Washukanni (Waššukanni).17 Later, Sornig and his team collected artefacts that allowed him to approached the issue on the localization of the Mi- hypothesize a human occupation spanning from the tanni centre in a new way, which this time he iden- Late Uruk period (i.e., end of the fourth millenni- tified as being Girnavaz.18 One year earlier, K. Kes- um BC) to the Neo-Assyrian Empire (tab. 1).20 sler published a study in which, according to some In 1982, the excavations began under the super- letters from Mari and Tell el-Amarna, he proposed vision of Prof. Erkanal with the permission of the that Girnavaz was in fact the Neo-Assyrian city of Mardin Museum. The archaeological investigation Nabula, called Nawala in the Old Babylonian pe- of the site was carried out for eight seasons (1982- riod.19 With this is mind, and with the attempt to 1984 and 1986-1991) but was abruptly stopped on 25.9.1991 when the team was struck by a terror- 17 Lucius, Sornig 1969, pp. 5, 10. 18 Sornig 1980, pp. 321-322. ARM 1, 90 and ARM 4, 52; Dossin 1950, p. 161; 1951, p. 7). 19 Kessler 1978-79. The Old Babylonian city of Nawala The Neo-Assyrian spelling Na-bu-la is instead attested is mentioned in one letter from Mari that is part of the royal cor- in two royal inscriptions: in the so-called ‘Broken Obelisk’ of respondence between the family members of Shamshi-Adad I. Ashur-bel-kala (1074-1053 BC), the king is recorded to have In ARM 4, 42, Ishme-Dagan gives information about his health fought the Arameans at Nabula (RIMA 2 A.0.89.7 iii, 10; to his brother, Yasmakh-Addu (Dossin 1951, p. 6). Moreover, Grayson 1991, p. 102). In his inscription, Shamshi-Adad V he apprises him of a famine that is affecting the town ofAmur - (824-811 BC) reminds a rebellion of 27 towns, among which sakim, which he is trying to eradicate with sending pears from Nabula, against his father Shalmaneser III (RIMA 3 A.0.103.1 Nawala. Probably, this famine was caused by the presence of i, 47; Grayson 1996, p. 183). Turukku, a tribe that was committing crimes in the region (see 20 Erkanal 1983.

West & East 244 Monografie, 4 Connecting cultures, dividing countries: a preliminary assessment of the Khabur Ware…

Table 1 Chronological phases identified at Girnavaz

West & East 245 Monografie, 4 Sergio Giuseppe Russo

Figure 2 Topographic map of the site with the excavated areas (modified from Uysal 1998, fig. 3)

West & East 246 Monografie, 4 Connecting cultures, dividing countries: a preliminary assessment of the Khabur Ware…

ist attack, which killed two archaeologists, Metin ber of ceramics considered to be diagnostic are 88 Akyurt and Bahattin Devam.21 (fig. 3).26 The work was conducted in three different sec- The KW from Girnavaz can be defined as a tors of the mound, respectively named ‘Central wheel-made monochrome painted ware with the Area’ (C), ‘Northern Terrace Area’ (N) and ‘North- colour of the fabric ranging from buff to brown-red- eastern Terrace Area’ (NE), covering an average of dish, including shades of beige, light yellow and light 0.18 ha, about 2.75% of the whole site (fig. 2).22 orange. The inner surface is often of the same colour Due to the interruption of the excavations, of the paste and in just a few cases a ‘sandwich’ sec- the layers dating back to the MBA period have not tion – in light grey – is found. The surface has a light been completely reached; a few remains have been slip or self-slip and it is sometimes smoothed; the brought to light in the N and NE areas. In the first tempers are quite depurated prevailing with chaff sector, which was marked by the presence of a prob- or sand in fine granulation and small inclusions of able official Neo-Assyrian architectural complex, chalk, lime and sometimes grits. The decorations the archaeologists opened a deep sounding (V- are mostly simple lines and parallel bands, some- Y/10-11). Here, scholars uncovered a garbage pit times combined with incised or impressed motifs. where pots and miniature vessels were discarded. Hatched triangles or cross-hatched motifs are also They labelled the vessels as Nuzi and Khabur wares attested, and they are generally located between the and related the pit to an adjacent wall built in the rim and the shoulder. More complex decorations early second millennium that was probably reused are extremely rare, usually consisting of triangles in the Hurrian-Mitanni period. Next to the wall, with dots or wavy lines (fig. 4). In one case (fig. 4: j), built with very large stones in an overlay technique, the combination with animal figures, e.g. a bird or a pebble floor was also recovered.23 In the NE sector, duck/goose, cannot be excluded. The most common above a cemetery of the third millennium BC, ar- colours for the painted decoration are brown-black- chaeologists unearthed remains of the MBA (layers ish, reddish-brown and orange. In terms of typolo- I and II) characterized by stone piles, pits, a proba- gies and shapes, the KW repertoire consists mainly ble tannur, mudbrick walls, and simple earthen bur- of bowls (17) and jars (35) (fig. 5). ials, and was marked by the presence of KW sherds.24 Bowls can be distinguished into either shallow Unfortunately, no complete structures and intact and deep: the former are characterized by a carination contexts have been recovered. and an internally thickened rim (fig. 8: 1).27 Painted bands are usually applied on the rim and occasionally on the inner surface. In one case, strokes are paint- 3. Description of the assemblage ed on the rim (fig. 9: 2).28 Deep bowls have straight (tabs. 2-3, figs. 6, 8-9) 26 Specimens like rims, bases, and decorated body sherds are here considered as diagnostic. First hand observation and A total of 224 band painted sherds have been re- macroscopic analyses were conducted. Thin sections micro- covered from sectors M 19-22 and O 22 (NE area), scopic analysis are planned for the future. Y-V/10-11, I 11 and B 11-12 (N area).25 The num- 27 Parallels can be found at Chagar Bazar (Mallowan 1936, fig. 17, n. 14; 1937, fig. 16, n. 18), Tell Barri (Baccelli, Manuelli 2008, pl. 2, n. 6), Tell Leilan (Weiss 1985, p. 13; Frayne 1995, fig. 41, n. 4), Tell al-Rimah (Postgate, Oates 21 Uysal 1998, p. 27. D., Oates J. 1997, p. 523), Hirbemerdon Tepe (Laneri 2016 (ed.), pl. 33, n. 76), İmikuşağı (Kozbe 1993, pl. 36, n. 3), and 22 Erkanal 1985. al-Andalus (Ristvet 2005, fig. A1.7, n. 6). 23 Erkanal 1991. 28 Parallels can be found at Tell Leilan (Frayne 1995, 24 Akyurt et Al. 1993. fig. 38, n. 1), in Tell al-Rimah (Postgate, Oates D., Oates 25 Unfortunately, no relevant information on the per- J. 1997, p. 520), Tell Mohammed Diyab (Faivre, Nicolle centages of other contemporaneous ceramic assemblages (i.e., 2007, pl. 11, n. 523), Tell Arbid (Bielinski 2005, fig. 10), Tell common wares, Plain Ware, Kitchen Ware) are here provid- Brak (Oates D., Oates J., McDonald H. 1997, fig. 190, ns. ed. A study with the aim of contextualizing the KW within the 210, 218), Dinkha Tepe (Hamlin 1974, fig 2, ns. 1a-b), and MBA pottery horizon is planned for the future. Tell Mozan (Pfälzner 2017, fig. 7.15, K2362).

West & East 247 Monografie, 4 Sergio Giuseppe Russo

Figure 3 Numbers and percentages of the KW diagnostic sherds collected from the different areas

Figure 4 Decorative patterns of the KW found at the site

Figure 5 Types and related painted decoration of the diagnostic sherds

West & East 248 Monografie, 4 Connecting cultures, dividing countries: a preliminary assessment of the Khabur Ware…

Figure 6 KW vessels exhibited at Mardin Museum (photo by Abdullah Bilen)

Figure 7 Preliminary chronological assessment of the KW found at the site

West & East 249 Monografie, 4 Sergio Giuseppe Russo

Figure 8 Selected sherds from the shallow (1-2) and deep (3-9) bowls and from the cups and beakers (10-13)

West & East 250 Monografie, 4 Connecting cultures, dividing countries: a preliminary assessment of the Khabur Ware…

Figure 9 Selected sherds from the barrels (1), the jars (2-7) and the basins (8)

West & East 251 Monografie, 4 Sergio Giuseppe Russo

or slightly curved profiles, externally and internally and beakers have painted bands in the upper part of thickened painted rims (fig. 8: 9).29 Carinated vari- the bodies, which present curved or carinated profiles ants are also present (fig. 8: 6-8).30 Of particular in- (fig. 8: 12‑13).34 terest are the small carinated bowls (or beakers) with Jars can be divided into two main categories, thin walls and painted bands or crosshatched trian- small and medium sized. The diameter of the small gles (fig. 8: 10-11).31 Incised and painted decorations jars varies between 9 and 18 cm. Jars with a high are found together in four examples (fig. 8: 3-4)32: or narrow neck and rounded body with a thick- one of the sherds presents a fishbone pattern and a ened rim belong to this group (fig. 9: 5-7).35 Paint- band on the rim whereas the other ones have incised ed bands and stripes are usually applied on the rim, lines on the rims or shoulders. Finally, a unicum is on the neck, and in some cases on the inner surface. represented by a double-rimmed bowl with a cross- One example, without a neck, has an ovoid body, hatched motif on the body (fig. 8: 5).33 A few cups painted bands and vertical strokes under the rim.36 The most complete vessel of this group is a small 29 Parallels can be found at Chagar Bazar (McMahon 2009, pl. 31, ns. 3, 5; pl. 37, n. 6), Tell Mozan (Pfälzner 2017, cal, in the unpainted variant, of the Late Chalcolithic period fig. 7.15, K2425), Tell Mohammed Diyab (Faivre, Nicolle in the Upper Khabur basin (e.g., in the Hamoukar area and 2007, pl. 10, n. 344), Tell Brak (Oates D., Oates J., Mc- in the Tell Leilan region; Wilkinson 2002, fig. 6, ns. 10-11; Donald H. 1997, fig. 195, n. 368), Tell al-Rimah (Post- Al Quntar, Qalidi, Ur 2011, fig. 7, ns. 1-2; Brustolon, gate, Oates D., Oates J. 1997, pl. 60, n. 564; pl. 79, n. 885), Rova 2007, fig. 3, n. 2). As concerns the painted decoration, Kültepe (Özgüç 1953, fig. 26), Hirbemerdon Tepe (Laneri cross-hatched motifs are found from the fifth millennium on- 2016 (ed.), pl. 64, n. 468), İmikuşağı (Kozbe 1993, pl. 21, wards. Interestingly, similar decorations to the one of bowl 8 n. 1), and Dinkha Tepe (Hamlin 1974, fig. 9, n. 35). can be found at Tell Brak (of unknown attribution, Oates 30 Parallels can be found at Tell Brak (Oates D., Oates D., Oates J., McDonald H. 2001, fig. 220; fig. 402, n. 286) J., McDonald H. 1997, fig. 192, n. 285; fig. 290), Chagar Ba- and in the late third to early second millennium period at zar (McMahon 2009, pl. 31, ns. 7-8; pl. 33, n. 5), Tell al-Rimah Tell al-Rimah (Postgate, Oates D., Oates J. 1997, pl. 19.a, (Postgate, Oates D., Oates J. 1997, pl. 59, n. 542; pl. 79, n. 7; pl. 60, n. 558). In addition, the same pattern is applied on n. 879), Dinkha Tepe (Hamlin 1974, fig. 2, n. 10b; fig. 10, jar 58, here attributed to the KW assemblage, which does not n. 29c), and Türbe Höyük (Baştürk 2009, pl. 2, a-c). Parallels show relevant differences in terms of fabric. However, it should to bowl 14 can be found at Chagar Bazar (McMahon 2009, be noted that a similar shape is found in northern Babylonia, pl. 31, n. 18), Tell Mohammed Diyab (Faivre, Nicolle 2007, at Tell ed-Der, which is dated from the end of the eighteenth pl. 10, n. 348), Tell Brak (Oates D., Oates J., McDonald century through the sixteenth (Armstrong, Gasche 2014: H. 1997, fig. 191, ns. 244, 248-249), Tell Leilan (Frayne 1995, p. 33; pl. 51, n. 7). Therefore, I suggest that this specimen at fig. 75, n. 4), Tell Mozan (Pfälzner 2017, fig. 7.5, K1136), Girnavaz may be dated to the MBAII. and Tell Barri (Baccelli, Manuelli 2008, pl. 4, n. 9). 34 Vessels 22-24 are most probably grain measures; parallels 31 Parallels can be found at Chagar Bazar (McMahon can be found at Chagar Bazar (McMahon 2009, pl. 47, n. 6), in 2009, pl. 37, n. 22; pl. 45), Tell Arbid (Kolinski 2014, fig. 4), Tell Brak (Oates D., Oates J., McDonald H. 1997, fig. 191, Tell Barri (Baccelli, Manuelli 2008, pl. 1, ns. 2-3; pl. 3, ns. ns. 256, 262, 265), and at Tell al-Rimah (Postgate, Oates D., 1-6), Tell Brak (Oates D., Oates J., McDonald H. 1997, Oates J. 1997, pl. 78, n. 871). Cups 25-28 have parallels at Cha- fig. 190, n. 206), Sagar (Ristvet 2005, fig. A1.8, n. 1), Tell al- gar Bazar (McMahon 2009, pl. 45, ns. 3-4, 15; pl. 46, ns. 17- Rimah (Postgate, Oates D., Oates J. 1997, pl. 79, n. 892), 18, 23), and Tell Brak (Oates et Al. 1997, fig. 191, ns. 245, Tell Mozan (Pfälzner 2017, fig. 7.15, K2427, K1141), and 249). Parallels to beaker 29 can be found at Tell Brak (Oates D., İmamoğlu (Kozbe 1993, pl. 100, n. 3). Oates J., McDonald H. 1997, fig. 197, n. 403), Tell Mozan 32 Bowls 5-7 are characterised by an atypical decoration for (Pfälzner 2017, fig. 7.21, K4359), and Tell al-Rimah (Post- KW. This, together with the uncertain and incomplete stratig- gate, Oates D., Oates J. 1997, pl. 76, n. 833) whereas paral- raphy of the excavated areas, makes their listing within the KW lels to cup 30 can be found at Chagar Bazar (Mallowan 1937, repertoire rather problematic. However, parallels to bowl 5 can fig. 24, ns. 12, 15), Tell Mohammed Diyab (Faivre, Nicolle be found at Tell Brak (Oates D., Oates J., McDonald H. 2007, pl. 10, n. 336), Tell Barri (Baccelli, Manuelli 2008, 1997, fig. 200, n. 455; fig. 214, n. 635; 2001, fig. 416, n. 467; pl. 1, n. 8), Tell Mozan (Pfälzner 2017, fig. 7.19, K4023), and fig. 428), Tell al-Rimah (Postgate, Oates D., Oates J. 1997, Tell al-Rimah (Postgate, Oates D., Oates J. 1997, pl. 59, pls. 64-65), Tell Mozan (Pfälzner 2017, fig. 7.5, K1134), Han- n. 536). sa (Ristvet 2005, fig. A1.7, n. 5), Tell Arbid (Kolinski 2014, 35 Parallels can be found in the Şırnak area (Kozbe 1993, fig. 5, n. 9), and Hirbemerdon Tepe (Laneri 2016 (ed.), pl. 44, pl. 111, n. 5; pl. 112, n. 1), at İmikuşağı (Kozbe 1993, pl. 4, n. 5; n. 470). Parallels to bowl 6 can be found at Tell Barri (Orsi 2011, pl. 17, n. 6), Tell Mozan (Pfälzner 2017, fig. 7.16, K2522; pl. 194, ns. 490-493) whereas parallels to bowl 7 can be found fig. 7.20), Tell al-Rimah (Postgate, Oates D., Oates J. at Tell Mozan (Orsi 2011, app. 4.IV, A7q1150-p121455) and 1997, pl. 90), and İmamoğlu (Kozbe 1993, pl. 99, n. 2). Tell Barri (Baccelli, Manuelli 2008, pl. 2). 36 Parallels can be found at Chagar Bazar (McMahon 33 Bowl 8 presents a channel or double rim that is typi- 2009, pl. 37, n. 2) and İmikuşağı (Kozbe 1993, pl. 32, n. 3).

West & East 252 Monografie, 4 Connecting cultures, dividing countries: a preliminary assessment of the Khabur Ware…

Figure 10 Comparisons of shallow carinated bowls: a, Tell Barri (afterBaccelli, Manuelli 2008, pl. 2, 6); b, Tell al-Rimah (after Postgate, Oates D., Oates J. 1997, pl. 59, 522); c, Tell Leilan (afterFrayne 1995, fig. 41, 4); d, Hirbemerdon Tepe (Laneri 2016 (ed), pl. 33, 76); e, İmikuşağı (afterKozbe 1993, pl. 36, 3); f, Dinkha Tepe (afterHamlin 1974, fig. 2, 1a); g, Mohammed Diyab (after Faivre, Nicolle 2007, pl. 11, 353); h, Tell Brak (after Oates D., Oates J., McDonald H. 1997, fig. 190, 210)

West & East 253 Monografie, 4 Sergio Giuseppe Russo

Figure 11 Comparisons of deep carinated bowls: a, Chagar Bazar (after McMahon 2009, pl. 33, 5); b, Türbe Höyük (after Baştürk 2009, lev. 2, b); c, Tell Brak (after Oates D., Oates J., McDonald H. 1997, fig. 192, 285); d, Dinkha Tepe (after Hamlin 1974, fig. 2, 10b); e, Chagar Bazar (after McMahon 2009, pl. 31, 7); f, Tell al-Rimah (after Postgate, Oates D., Oates J. 1997, pl. 79, 879); g, Mohammed Diyab (afterFaivre, Nicolle 2007, pl. 10, 348)

West & East 254 Monografie, 4 Connecting cultures, dividing countries: a preliminary assessment of the Khabur Ware…

Figure 12 Comparisons of small carinated bowls/beakers: a, Sagar (after Ristvet 2005, fig. A1.8, 1); b, Chagar Bazar (after McMahon 2009, pl. 45, 9); c, Tell Barri (after Baccelli, Manuelli 2008, pl. 3, 3); d, Tell Mozan (after Pfälzner 2017, fig. 7.15, K2427); e, Tell Mozan (after Pfälzner 2017, 7.5, K1141); f, Tell Arbid (after Kolinski 2014, fig. 4, 2)

West & East 255 Monografie, 4 Sergio Giuseppe Russo

Figure 13 Comparisons of medium jars with low necks: a, Hirbemerdon Tepe (afterLaneri 2016 (ed.), pl. 49, 323); b, Tell Barri (after Baccelli, Manuelli 2008, pl. 1, 9); c, Dinkha Tepe (afterHamlin 197, fig. 9, 32); d, Tell Mozan (afterPfälzner 2017, fig. 7.15, K2394); e, Chagar Bazar (afterMcMahon 2009, pl. 31, 23); f, Şırnak area (afterKozbe 1993, lev. 112, 4)

jar with a squared everted rim and rounded body tif. Medium‑sized jars have a diameter which is be- (fig. 9: 6).37 Painted bands are applied on the rim tween 18 and 35 cm and have a short or low neck. and on the upper and lower part of the body, while Most of the vessels have a hole-mouth rim, often the central area is occupied by a cross-hatching mo- internally and externally thickened, and a slightly curved profile (fig. 9: 2).38 A few carinated exam- 37 Parallels can be found at Tell Mohammed Diyab (Faivre, Nicolle 2007, pl. 12, n. 372) and İmikuşağı (Koz- 38 Parallels can be found at Chagar Bazar (McMahon be 1993, pl. 7, n. 5). 2009, pl. 31, n. 23; pl. 34, ns. 1, 5; pl. 35, n. 5; pl. 36, n. 2; pl. 37,

West & East 256 Monografie, 4 Connecting cultures, dividing countries: a preliminary assessment of the Khabur Ware…

ples are also attested (fig. 9: 1)39 as well as grooved In general, it seems that the site was frequent- rims (fig. 9: 3).40 As with the small ones, the medi- ed throughout the MBA, perhaps with a main oc- um‑sized jars have painted bands on the rim and cupation during MBA II (1800-1600 BC). The on the shoulder. Finally, two basins with painted site might have been inhabited in the earlier phase, bands, in one case combined with a rope decoration i.e. during the twentyfirt century BC, witnessed by (fig. 9: 8), are attested.41 the presence of bowls with combined incised/impressed and painted decorations, which are contemporary in other sites such as in Tell Mozan. Some 3.1 Proposal of inner relative chronology carinated vessels with painted rims and simpler patterns may belong to the MBA I (2000-1800 BC), The definition of an inner chronology for the site is whereas small-to-medium sized jars, cups/beakers rather difficult, considering that excavations were and bowls with a finer clay and more complex de- abruptly stopped and that, consequently, all the signs find parallels in the MBA II contexts (fig. 7). stratigraphic layers were not fully exposed. While some periods, such as the Early Bronze Age and the Neo-Assyrian period, do have related architectural re- 4. Conclusion mains that can help in dating the strata, for the MBA we can only take into account the pottery uncovered Girnavaz Höyük, located within the main distri- within mixed or uncertain contexts. Supplementary bution zone of KW, i.e. theUpper Khabur river information can be obtained from the comparison of valley, should be considered as the northernmost KW sherds with the contemporary repertoires of the edge of a homogeneous cultural area centred in sites of the surrounding regions (figs. 10-13). northeastern Syria, observed through the discovery of several MBA painted sherds. Furthermore, its location at a crossing point from/to southeastern Anatolia and the parallels to Anatolian band n. 6; pl. 50, n. 5; pl. 54, n. 1; pl. 55, ns. 5, 10-11, 23; pl. 56, ns. 8, painted wares support the hypothesis of the exist- 17; pl. 57, ns. 5, 12; pl. 58, ns. 8-9; pl. 59, n. 17; pl. 60, ns. 3, 14; ence at the site either of a trading outpost or, at pl. 63, n. 5), Tell Barri (Baccelli, Manuelli 2008, pl. 1, n. 5; least, a rest stop. In terms of settlement pattern, pl. 3, n. 7), Tell Mohammed Diyab (Faivre, Nicolle 2007, pl. 10, ns. 317, 320-321), Tell Leilan (Frayne 1995, fig. 60, the fact that no relevant second millennium archi- n. 6; fig. 65, n. 3; Ristvet 2005, fig. A1.8, n. 10), Tell Mozan tectures were found on the NE terrace could be ex- (Pfälzner 2017, fig. 7.15, K2394; fig. 7.16, K2616; fig. 7.21, plained by the presence of a related settlement in K4355), Tell al-Rimah (Postgate, Oates D., Oates J. 1997, pl. 85, n. 983; pl. 88), Tell Brak (Oates D., Oates J., other parts of the mound. The high percentage of McDonald H. 1997, fig. 192, n. 285; fig. 214, n. 646), Türbe KW uncovered from the northern sector suggests Höyük (Baştürk 2009, pl. 4, c), İmikuşağı (Kozbe 1993, pl. that it would perhaps be located on the N terrace, 32, n. 5; pl. 49, n. 1), in the Şırnak area (Kozbe 1993, pl. 112, n. 4; pl. 116, n. 2), and at Dinkha Tepe (Hamlin 1974, fig. 7, underlying the Neo- and Middle Assyrian levels. ns. 15a, 18; fig. 8, n. 22b; fig. 9, n. 33). Indeed, a residential sector might have also been 39 These vessels are most probably barrels; parallels can located either in the Central area – a flat portion be found at Chagar Bazar (McMahon 2009, pl. 17, n. 8; pl. 30, n. 22; pl. 39, n. 9), Tell Arbid (Kolinski 2014, fig. 5, of the mound where earlier layers were not inves- n. 7-9), Tell Leilan (Frayne 1995, fig. 73, n. 2; fig. 76, n. 3), tigated – or in the southern sector which was not Tell Mozan (Pfälzner 2017, fig. 7.15, K2401; fig. 7.21, excavated at all. Thus, the eastern area, due to its K4351), and in the Şırnak area (Kozbe 1993, pl. 113, ns.1-2; pl. 114, ns. 1, 3). steep slope, would not have been frquented in- 40 Parallels can be found at Chagar Bazar (McMahon tensely by humans; it probably represented the 2009, pl. 61, n. 2), Tell Barri (Baccelli, Manuelli 2008, shortest passage to the lower arable fields and to pl. 1, n. 9), Hirbemerdon Tepe (Laneri 2016 (ed.), pl. 49, ns. the river. If this hypothesis is reliable, the Neo-As- 323, 325-326), and Dinkha Tepe (Hamlin 1974, fig. 9, n. 32). syrian buildings would reflect the settlement pat- 41 Parallels can be found at Tell Barri (Orsi 2011, p. 876, n. 628) and Tell Mohammed Diyab (Faivre, Nicolle 2007, tern of the second millennium BC. pl. 11, n. 359).

West & East 257 Monografie, 4 Sergio Giuseppe Russo RØ :

slip slip slip slip slip slip slip slip, slip, slip, Surface smoothed smoothed smoothed smoothed smoothed treatments Decoration painted, 5YR 6/8 orange painted, 10YR 6/4 dull yellow orange incised; painted, 5Y 2/2 olive black incised; painted, 7.5GY 2/1 dark greenish gray painted, 5Y 2/2 olive black painted, 10YR 5/4 dull yellowish brown painted, 5Y 2/2 olive black painted, 5YR 4/4 dull reddish brown painted, 5YR 5/4 dull reddish brown painted, 5YR 4/4 dull reddish brown painted, 7.5GY 2/1 dark greenish gray painted, 10YR 6/4 dull yellow orange painted Fabric coarse-medium, limestone, grit coarse-medium, limestone, mica, grit coarse-medium, sand, chaff, limestone medium, sand, chaff, grit medium, sand, limestone, chaff coarse-medium, chaff, sand, limestone medium, sand, chaff, limestone medium, sand, grit medium-fine, grit, limestone fine, sand, grit medium-fine, sand, limestone medium, limestone, mica medium, sand, chaff, limestone Color (sec) 25YR 8/2 light gray 25YR 8/2 light gray 5Y 8/3 pale yellow 7.5Y 8/2 light gray 5Y 8/3 pale yellow 5Y 7/4 light yellow 5Y 8/3 pale yellow 5Y 7/4 light yellow 7.5YR 5/6 bright brown 7.5Y 8/3 pale yellow 10YR 6/3 dull yellow orange 25YR 8/2 light gray 5Y 8/3 pale yellow Color (in) 25YR 8/2 light gray 25YR 8/2 light gray 5Y 8/3 pale yellow 7.5Y 8/2 light gray 5Y 8/3 pale yellow 5Y 7/4 light yellow 5Y 8/3 pale yellow 5Y 7/4 light yellow 7.5YR 6/6 orange 7.5Y 8/3 pale yellow 7.5Y 8/3 pale yellow 25YR 8/2 light gray 5Y 8/3 pale yellow Class : general morphology: B: bowl; Ba: base; Be: beaker; Br: Barrel; BS: body sherd; C: cup; G: grain measure; M: miscellaneous; J: jar; Color (out) .: thickness in cm; Fabric : sequence of inclusions from highest to lowest density; Color with reference the Munsell Soil Colour Charts 2009) 5Y 8/2 light gray 5Y 8/2 light gray 5Y 8/3 pale yellow 7.5Y 8/3 pale yellow 5Y 8/3 pale yellow 5Y 7/4 light yellow 5Y 8/3 pale yellow 5Y 7/4 light yellow 7.5YR 6/6 orange 7.5Y 8/3 pale yellow 7.5Y 8/3 pale yellow 5Y 8/2 light gray 5Y 8/3 pale yellow 0,9 0,8 1,9 0,4 0,9 1,2 0,8 0,5 0,3 0,4 0,4 0,5 0,6 Th.

BØ 7 9 9 9 16 40 16 38 24 30 15 17 28 RØ B B B B B B B B B B B C C Class n/a Y' 10 Y' 11 Y' 11 Y' 10 Y' 11 Y' 11 V' 11 V' 11 Y' 10 O' 22 M' 22 M' 22 Sector N N N N N N N N N NE n/a NE NE Area 8.1 8.2 8.3 8.4 8.6 8.7 8.9 8.5 8.8 fig. 8.10 8.12 8.11 8.13 Area : N: Northern Terrace; NE: Northeastern ( rim diameter in cm; BØ : base Th T able 2 Catalogue of the selected sherds cited in text

West & East 258 Monografie, 4 Connecting cultures, dividing countries: a preliminary assessment of the Khabur Ware… RØ : slip slip slip Surface treatments slip, smoothed slip, smoothed slip, smoothed slip, smoothed Decoration applied; painted, 5YR 5/6 bright reddish brown painted, 5YR 5/4 dull reddish brown painted, 5YR 3/6 dark reddish brown painted, 5Y 2/2 olive black painted, 5Y 2/2 olive black grooved; painted, 5YR 5/3 dull reddish brown painted, 5YR 5/4 dull reddish brown painted, 5YR 4/4 dull reddish brown Fabric Color : with reference to the Munsell Soil Colour Charts 2009) medium, chaff, limestone, grit medium, limestone, grit medium, grit, chaff, limestone medium-fine, sand, chaff, limestone coarse-medium, sand, chaff, limestone coarse-medium, sand, chaff, limestone coarse-medium, limestone coarse-medium, sand, grit, limestone Color (sec) 25YR 7/4 light yellow 10YR 8/3 light yellow 5Y 8/4 pale yellow 5Y 8/3 pale yellow 5Y 8/3 pale yellow 5Y 7/4 light yellow 7.5Y 8/3 pale yellow 7.5Y 8/3 pale yellow Color (in) 25YR 8/4 pale yellow 7.5Y 8/3 pale yellow 5Y 8/4 pale yellow 5Y 8/3 pale yellow 5Y 8/3 pale yellow 5Y 7/4 light yellow 7.5Y 8/3 pale yellow 7.5Y 8/3 pale yellow Class : general morphology: B: bowl; Ba: base; Be: beaker; Br: Barrel; BS: body sherd; C: cup; G: grain measure; M: miscellaneous; J: jar; Color (out) .: thickness in cm; Fabric : sequence of inclusions from highest to lowest density; 25YR 8/4 pale yellow 5Y 8/3 pale yellow 7.5Y 8/3 pale yellow 5Y 8/4 pale yellow 5Y 8/3 pale yellow 5Y 7/4 light yellow 5Y 8/2 light gray 7.5Y 8/3 pale yellow Th 1,7 0,7 0,5 0,8 0,7 0,7 0,8 0,9 Th. 16 BØ 9 20 16 14 28 34 29 RØ J J J J J J J Ba Class I' 11 Y' 11 B' 12 Y' 11 Y' 11 Y' 10 Y' 10 O' 22 Sector N N N N N N N NE Area 9.8 9.4 9.5 9.6 9.7 9.1 9.3 9.2 fig. Area : N: Northern Terrace; NE: Northeastern ( rim diameter in cm; BØ : base T able 3 Catalogue of the selected sherds cited in text

West & East 259 Monografie, 4 Sergio Giuseppe Russo

Acknowledgments

I would like to thank Professor Hayat Erkanal for authorizing my research on the materials from Girnavaz Höyük and Dr. Ayşegül Aykurt for granting me access to the depository of Hacettepe University, Ankara where the materials are stored. I also thank the two anonymous reviewers whose comments helped improve and clarify this paper. Lastly, I would like also to dedicate this work to the two archaeologists, Metin Akyurt and Bahattin Devam, who lost their lives while working at the site.

West & East 260 Monografie, 4 Connecting cultures, dividing countries: a preliminary assessment of the Khabur Ware…

BIBLIOGRAPHY

Akyurt M., Devam B., Erkanal H., Ökse T.A. tra fine III e inizio II mill. a.C. (Studia Asiana 11). 1993, 1991 Girnavaz Kazıları, «Kazı Sonuçları Firenze. Toplantısı» 14/1, pp. 267-280. Dohmann-Pfälzner H., Pfälzner P. 2002, Al Quntar S., Qalidi L., Ur J. 2011, Protourbanism in Ausgrabungen Der Deutschen Orient-Gesellschaft the Late 5th Millennium BC: Survey and Excavations in Der Zentralen Oberstadt von Tall Mozan/Urkeš, at Khirbat al-Fakhar (Hamoukar), Northeast Syria, «Mitteilungen der Deutschen Orient Gesellschaft» «Paléorient» 37/2, pp. 151-175. 132, pp. 185-228. Armstrong J.A., Gasche H., 2014, Mesopotamian Dossin G. 1950, Correspondance de Šamši-Addu et de ses Pottery: A Guide to the Babylonian Tradition in the fils (Archives Royales de Mari 1), Paris. Second Millennium B.C. (Mesopotamian History Dossin G. 1951, Correspondance de Šamši-Addu et de ses and Environment). Ghent and Chicago. fils (Suite) (Archives Royales de Mari 4), Paris. Baccelli G, Manuelli F. 2008, Middle Bronze Khabur Erkanal H. 1983, 1982 Girnavaz Kazıları, «Kazı Ware from Tell Barri/Kahat, in: Córdoba J.M., Sonuçları Toplantısı» 5, pp. 131-135, 393-396. Molist M., Pérez M.C., Rubio I., Martínez S. Erkanal H. 1984, , «Kazı (eds.), Proceedings of the 5th International Congress 1983 Girnavaz Kazıları Sonuçları Toplantısı» 6, pp. 121-136. on the Archaeology of the Ancient Near East, Vol. I, Madrid, pp. 187-206. Erkanal H. 1985, 1984 Girnavaz Kazıları, «Kazı Sonuçları Toplantısı» 7, pp. 201-210. Baştürk E. 2009, Habur çanak çömleğinin Siirt- Türbe Höyük verileri ışığında değerlendirilmesi (1st Erkanal H. 1985, 1984 Girnavaz Kazıları, «Kazı International Symposium on the History and culture Sonuçları Toplantısı» 7, pp. 201-210. of Batman and its vicinity, Batman, 15-17 apr., 2009). Erkanal H. 1987, 1986 Girnavaz Kazıları, «Kazı Bedal L.-A, Fleming S., De Schauensee M., Sonuçları Toplantısı» 9/1, pp. 361-370. Hancock R. 1995, Second Millennium B.C. Erkanal H. 1988, 1987 Girnavaz Kazıları, «Kazı Pottery at Hasanlu Tepe and Dinkha Tepe: INAA Sonuçları Toplantısı» 10/1, pp. 273-286. and Petrographic Studies, «MRS Proceedings» 352, Erkanal H. 1990, 1988 Girnavaz Kazıları, «Kazı pp. 453-467. Sonuçları Toplantısı» 11/1, pp. 261-274. Bielinski P. 2005, Tell Arbid, the Ninth Season of Erkanal H. 1991, 1989 Girnavaz Kazıları, «Kazı , Syrian-Polish Excavations. Preliminary Report Sonuçları Toplantısı» 12/1, pp. 277-292. «Polish Archaeology in the Mediterranean» 16, pp. 475-489. Faivre X., Nicolle C. 2007, La Jézireh au Bronze moyen et la céramique du Khabur, in: Al-Maqdissi M., Brustolon A, Rova E. 2007, The Late Chalcolithic Matoian M., Nicolle C. (eds.), Céramique de Period in the Tell Leilan Region: A Report on the l’Âge du Bronze en Syrie, II, L’Euphrate et la région Ceramic Material of the 1995 Survey, «Kaskal» 4, de Jézireh, Beyrouth, pp. 179-229. pp. 1-42. Frayne J. 1995, The Tell Leilan Period I Khabur Ware D’Agostino A. 2012, Hibermerdon Tepe and The Upper Assemblage (Ph.D. dissertation, University of North Tigris Valley during the Early Second Millennium Carolina), Chapel Hill. BC: A First Assessment of the Local Pottery Horizon, in: Laneri N., Pfälzner P., Valentini S. (eds.), Grayson A.K. 1991, Assyrian Rulers of the Early Looking North, The Socioeconomic Dynamics of First Millennium BC (1114-859 BC) (The Royal Northern Mesopotamian and Anatolian Regions du- Inscriptions of Mesopotamia, Assyrian Periods 2), ring the Late Third and Early Second Millennium BC, Toronto. Wiesbaden, pp. 139-212. Grayson A.K. 1996, Assyrian Rulers of the Early D’Agostino A. 2016, L’Alto Tigri nelle età del Bronzo First Millennium BC II (858-745 BC), (The Royal Antico e Medio: Siti, sequenze e ceramiche rosso-brune Inscriptions of Mesopotamia, Assyrian Periods 3), Toronto.

West & East 261 Monografie, 4 Sergio Giuseppe Russo

Hamlin C. 1971, The Habur Ceramic Assemblage of Mallowan M.E.L. 1937, The excavations at Tell Northern Mesopotamia: an Analysis of its Distribution, Chagar Bazar and an Archaeological Survey of the (Ph.D. dissertation, University of Pennsylvania), Habur Region, Second Campaign, 1936, «Iraq» 4, Philadelphia. pp. 91-177. Hamlin C. 1974, The Early Second Millennium Ceramic Mallowan M.E.L. 1956, Twenty-five years of Mesopo Assemblage of Dinkha Tepe, «Iran» 12, pp. 125-153. tamian Discovery, London. Hrouda B. 1957, Die bemalte Keramik des zweiten McMahon A. 2009, Once There Was a Place: Settlement Jahrtausends in Nordmesopotamien und Nordsyrien, Archaeology at Chagar Bazar, 1999-2002, London. Berlin. Oates D., Oates J., McDonald H. 1997, Excavations Hrouda B. 1989, Die Habur-Ware in neuerer Sicht, in: at Tell Brak. Vol. 1: The Mitanni and Old Babylonian Emre K., Mellink M., Hrouda B., Özgüç N. periods, Cambridge. (eds.), Anatolia and the Ancient Near East. Studies in Oates D., Oates J., McDonald H. 2001, Excavations honor of Tahsin Özgüç, Ankara, pp. 205-214. at Tell Brak. Vol. 2: Nagar in the third millennium Hrouda B. 2001, About Habur-Ware, hopefully for the BC, Cambridge. last time, «Al-Rafidan» 22, pp. 89-92. Oguchi H. 1997, A Reassessment of the Distribution of Kessler K. 1978-1979, Nawala und Nabula, «Archiv Khabur-Ware: an approach from an aspect of its main für Orientforschung» 26, pp. 99-103. phase, «Al-Rafidan» 18, pp. 195-224. Kolinski R. 2014, 20th Century BC in the Khabur Oguchi H. 1998, Notes of Khabur Ware from sites out- Triangle Region and the Advent of the Old Assyrian side its main distribution zone, «Al-Rafidan» 19, Trade with Anatolia, in: Bonatz D. (ed.), pp. 119-133. The Archaeology of Political Spaces. The Upper Oguchi H. 2001, The Origins of Khabur Ware: A nd Mesopotamian Piedmont in the 2 Millennium BCE Tentative Note, «Al-Rafidan» 22, pp. 71-87. (TOPOI Berlin Studies of the Ancient World 12), th Berlin, pp. 11-34. Oguchi H. 2003, 20 century BC North Mesopotamia: an archaeological dilemma, «Al-Rafidan» 24, Kozbe G. 1993, Habur Çanak Çömleği ve Bu Çanak pp. 83-100. Çömleğin Yeni Buluntular Işığında Anadolu’daki Oguchi H. 2006, Dağılımı (Ph.D. dissertation, Ege Üniversitesi), Notes on Khabur Ware from the Izmir. Haditha Dam Area, in: Kepinski C., Lecomte O., Tenu A. (eds.), Studia Euphratica. Le mo- Laneri N. (ed.) 2016, Hirbemerdon Tepe Archaeological yen Euphrate iraquien révélé par les fouilles preven- Project 2003-2013 Final Report: Chronology and tives de Haditha (Travanx de la Maison 3), Paris, Material Culture, Bologna. pp. 203-216. Laneri N., Schwartz M., Ur J., Berthon R., Olmstead A.T. 1918, The Calculated Frightfulness of D’Agostino A., Hald M.M., Marsh A. Ashur Nasir Apal, «Journal of the American Oriental 2015, Ritual and Identity in Rural Mesopotamia: Society» 38, pp. 209-263. Hirbemerdon and the Upper Tigris River Valley Orsi V. 2010, During the Middle Bronze Age, «AJA» 119/4, The Passage from the Early Bronze to pp. 533-564. the Middle Bronze Age in Jezirah: A Parallel between Tell Mozan and Tell Barri Ceramic Sequences, Laneri N., Schwartz M., Valentini S. 2012, in: Matthiae P., Pinnock F., Nigro L. (eds.), Hirbemerdon Tepe: A Middle Bronze Age site Proceedings of the 6th International Congress on the in Mesopotamia, in: Laneri N., Pfälzner Archaeology of the Ancient Near East (5-10 may P., Valentini S. (eds.), Looking North, The 2008), Wiesbaden, pp. 863-881. Socioeconomic Dynamics of Northern Mesopotamia Orsi V. 2011, and Anatolian Regions during the Late Third and Early Crisi e Rigenerazione nella valle dell’Alto Second Millennium BC, Wiesbaden, pp. 137-147. Khabur (Siria): la produzione ceramica nel passaggio dal Bronzo Antico al Bronzo Medio, Firenze. Lucius E., Sornig K. 1969, Zum Lokalisierungsproblem Özgüç T. 1953, von Wassuganni und anderen mitannisch – hurri- Vorläufiger Bericht über die Grabungen schen Siedlungszentren, Istanbul. von 1950 in Kültepe Ausgeführt im Auftrage des Türk Tarih Kurumu, «Belleten» 17, pp. 109-118. Mallowan M.E.L. 1936, The excavations at Tall Palmisano A. 2012, Chagar Bazar and an Archaeological Survey of the Diachronic and Spatial Habur Region, 1934-5, «Iraq» 3, pp. 1-86. Distribution of Khabur Ware in the Early Second

West & East 262 Monografie, 4 Connecting cultures, dividing countries: a preliminary assessment of the Khabur Ware…

Millennium BC, «Journal of Open Archaeology Sornig K. 1980, Hurritische Spuren. Vielsprachigkeit Data» 1, p. e8 (doi: 10.5334/data.1334754978; im Alten Orient, der Tur Abdin und die Wassuganni- http://doi.org/10.5334/4f8d6ed49bd54). Frage, «Grazer Linguistische Studien» 11/12, Pfälzner P. 2017, Ḫabur Ware and Social Continuity: pp. 298-325. The Chronology of the Early to Middle Bronze Age Speiser E.A. 1933, The Pottery of Tell Billa, «The Transition in the Syrian Jezireh, in: Höflmayer F. Museum Journal» 23, pp. 249-283. (ed.), The Late Third Millennium in the Ancient Near Stein D.L. 1984, Khabur Ware and Nuzi Ware: their East: Chronology, C14, and Climate Change (Oriental origin, relationship, and significance, «Assur» 4, Institute Seminars 11), Chicago, pp. 163-204. pp. 3-65. Postgate C., Oates D., Oates J. 1997, The Thomas H.L. 1967, Near Eastern, Mediterranean and Excavations at Tell Al Rimah: The Pottery, Cambridge. European Chronology (Studies in Mediterranean Ristvet L. 2005, Settlement, Economy, and Society in Archaeology, vol. XVII/1). Lund. the Tell Leilan Region, Syria, 3000-1000 BC, Ph.D. Uysal B. 1991, Mardin Bölgesi M.Ö. 3. Bin Kolye ve dissertation, University of Cambridge. Bilezikleri (M.A. dissertation, Ankara Üniversitesi), Schimdt C. 2012, The Late Third and Early Second Ankara. Millennium BC Pottery Tradition in the Syrian Uysal B. 1998, Girnavaz Kazılarının Işığı Altında Jazirah and Beyond, in: Laneri N., Pfälzner Anadolu’da Ninive 5 Seramiği (Ph.D. dissertation, P., Valentini S. (eds.), Looking North, The Hacettepe Üniversitesi), Hacettepe. Socioeconomic Dynamics of Northern Mesopotamian Weiss H. 1985, , and Anatolian Regions during the Late Third and Early Tell Leilan on the Habur Plains of Syria «Biblical Archaeology» 48/1, pp. 5-34. Second Millennium BC, Wiesbaden, pp. 163-174. Welker M. 1948, Şenyurt Y.S. 1988, Girnavaz Kazıları Işığında Geç The Painted Pottery of the Near Assur Devri Düğme Dipli Vazoları, «Dil ve Tarih- East in the Second Millennium B.C. and Its Coğrafya Fakültesi Dergisi» 33/1-2, pp. 281-291. Chronological Background, «Transactions of the American Philosophical Society, New Series» 38/2, Şenyurt Y.S. 1994, M.Ö. 3. Binde Güneydoğu Anadolu pp. 185-245. Bölgesi’nde Metalik Seramik ve Bu Seramiğin Wilkinson T.J. 2002, Anadolu’daki Yayılmı, Ph.D dissertation, Ankara Physical and cultural landscapes Üniversitesi. of the Hamoukar area, (Syria), «Akkadica» 123/1, pp. 89-105. Şenyurt Y.S. 1995, Late Assyrian Painted Pottery Yalçıklı D. 1991, from Girnavaz, in Erkanal A. et Al. (eds.), In Güneydoğu Anadolu Bölgesinin M.Ö. (M.A. disser- Memoriam İ. Metin Akyurt ve Bahattin Devam III Bine Tarihlendirilen Tunç Silahları tation, Hacettepe Üniversitesi), Hacettepe. Anı Kitabı, Eski Yakındoğu Kültürleri Üzerine İncelemeler (Studies for Ancient Near Eastern Cultures), Istanbul, pp. 327-336.

West & East 263 Monografie, 4

(Hi)storytelling: the ancient Near East in western historical novels and archaeological writing

KATHRIN SCHMITT

Freie Universität Berlin

Abstract

Verbal aspects of historiography are barely considered in the German archaeological community. The highly descriptive archaeological writing is still widely seen as desirable because of its seemingly neutral and objective nature, especially compared with other forms of historiographical storytelling such as non-fiction, historical novels or documentaries. Even decades after the linguistic turn, its insights never fully entered the German archaeology of the Near East. Therefore, in this paper, I will address issues of narration and language in archaeological fact-production in comparison to historical novels, which are commonly accused of being ideological, euphemistic and sometimes escapist. This leads me to expose some hidden ideological elements in archaeological writing, which are mostly concealed within our structural way to approach past as well as present societies. While I focus on several ‘-isms’ such as colonialism, eurocentrism or androcentrism, the archaeology of the Ancient Near East serves only as a case study. Parts of the issues discussed here apply to varying degrees to the archaeologies of other times and regions as well.

Keywords

Historiography, historical novels, fact and fiction, postcolonial studies, ideologies in archaeology

West & East 265 Monografie, 4 Kathrin Schmitt

1. Introduction focuses almost exclusively on proving the fictional and thereby less valuable status of historical novels. As the March for Science movement has recently The resulting desideratum includes all questions on clearly demonstrated, claiming factuality in public a) the archaeological production of facts and factu- discourse becomes more and more difficult for reality, b) how archaeologists are writing history and searchers on an international level.1 As soon as one c) what biases their historiography. In this paper considers factuality and fictionality as a research my aim is to discuss especially the last two points. field in archaeology, one is almost automatically Therefore, I will first provide a brief definition of confronted with historical novels on one side and the terms fact and fiction and some methodologi- academic historiography on the other. The two gen- cal aspects of the linguistic analysis of our writing. res are traditionally the terminological pair repre- Afterwards I will focus on some analytical categories senting best the fact-fiction-divide in the historical influencing the factual and fictional status of both field.2 Thereby fiction is usually negatively connot- historical novels and archaeological writing and the ed and the quality of a historical novel is often (and ideologies hidden in them. outside of literary studies) measured by its historical accuracy.3 Furthermore novels are mostly accused of being full of ideologies, euphemisms and occasional- 2. Fact and fiction in historiography ly escapisms.4 On the other side, academic historiography and by this archaeological writing, especially While fictionality is usually defined in opposition in German Near Eastern archaeology, is still consid- to factuality, most studies lack a definition of factu- ered as more neutral and objective, as it is academic ality, since they implicitly equate it with truth and and thereby factual. Since the insights of the linguis- reality.7 For that reason I give here a very brief defi- tic turn never entered the German Archaeology on nition of the terms related to historiographical writ- the Near East, language is still widely seen as a neu- ing as I use them in this paper: tral tool of expression, which is one reason why I fo- Within the context of this article facts are un- cus mainly on the German archaeological writing in derstood as representations based on data and dis- this paper.5 Moreover, the dualistic concept of fact cursively negotiated narrations or verbalized forma- and fiction is rarely questioned in general.6 Research tions of possible historical worlds. In comparison, fictions are understood as imagined possibilities of 1 The March for Science is a global movement, which is historical worlds, which do not have to stick fully advocating for amplifying the role of science in policy, empow- to any given data or historical realities already dis- ering public engagement with science, fostering a diverse and 8 inclusive scientific community and building a global commu- cursively negotiated. This means that facts need to nity of science advocates. Further information can be found on bridge the gap between the written text and a veri- https://www.marchforscience.com/ (09/01/2019). fiable reality that is extra-textual and can be physi- 2 E.g. Paul, Faber (eds.) 2013; Schaeffer 2012; cally experienced. For example, a travel guide can be Storm 2010; Bernbeck 2005; Maier, Schweizer 1999; Nünning 1995. considered as factual if the directions given through 9 3 Ginz 2014, p. 23. The negative connotation is usually the text lead travellers to a historical monument. In based on the alleged deficiencies, particularly its lack of reality historiography, this gap is caused to a large extent and truth. by time difference. The metaphorical bridge into a 4 Aust 1994, p. 1. historical reality is therefore not fully traversable. 5 The term ‘linguistic turn’ goes back toRorty (ed.) 1968 and triggered a critical discourse on language in various archaeologists, while Anglo-American anthropology is much academic disciplines, especially in the humanities. The linguis- more engaging with the topic in general, cf. van Dyke, tic turn questioned the notion of language as a neutral tool of Bernbeck (eds.) 2015; Joyce, Preucel 2002; Clifford, communication: language is influencing every discourse in such Marcus (eds.) 1986. a way that, instead of simply mirroring or copying reality, it cre- 7 ates a ‘new reality’ of its own. Schaeffer 2012; 2010; Nünning 1995. 8 6 One of the rare exceptions is still White 1973, Cf. Felder (ed.) 2013. whose work is barely considered among German Near Eastern 9 Cf. Gardt 2013, p. 36 ff.

West & East 266 Monografie, 4 (Hi)storytelling: the ancient Near East in western historical novels and archaeological writing

However, for novels to be considered as historical the textual macro-structure and c) the textual mi- fiction, they need to reference at least partly to ex- cro-structure.14 tra-textual historical facts. Otherwise, they would The contextual frame includes a consideration be considered ‘counter-factual’ or fictional without of the authors and their background, the publishers the predicate ‘historical’. Thus, the demarcation line (scientific or belle lettres) and the cover’s design. In between fact and fiction in (Hi)storytelling is quite general it also includes the forms of communication fluid. Fact and fiction cannot be understood as two (written, spoken, pictorial). In this paper I focus separate phenomena, but as extremes of the same on written texts in printed books. Furthermore the sphere in which academic historiography as well as contextual frame includes a publication’s outreach historical novels and even the so called ‘non-fiction’ and an inquiry into the sphere of life or knowledge can be arranged or classified.10 Furthermore, the domain it is connected to. The knowledge domain constructivist character of factuality should be em- of the following examples is the history of the An- phasized, as factuality is related but not equal to the cient Near East, but the spheres of life to which they terms truth or reality, especially when one is dealing belong can be categorized briefly as ‘literature’ (his- with past worlds and societies and their present re- torical novels) and ‘science and humanities’ (archae- construction, as the bridge metaphor illustrates. To ological publications).15 The last meaningful factor avoid misunderstandings, this definition does not of the contextual frame is the anticipated reader, establish any proximity of factuality to the phenom- who is not to be confused with the actual reader. In- ena of falsehood and deception as they are closer but formation regarding these meaningful factors is not again not equal to the definition of fiction. always explicitly formulated in the books and requires either further research or prior knowledge to be analysed. Since knowledge is an abstract entity, 3. Methodological aspects its range can vary from person to person. Accord- ingly, it becomes obvious that the categorisation of a To analyse how facts and fictions are created in his- publication as factual or fictional generally depends toriographical writing, and more importantly what on an actual reader’s knowledge and the separation influences them and how they are biased, one can fo- is due to a relation rather than inherent in a text. cus on different ‘meaningful factors’.11 Such mean- The analysis of the textual macro-structure fo- ingful factors can be derived from the methods of cuses on the internal structure of the publication, two research fields, according to literary studies: lin- which is given not only by the format (e.g. layout, guistic discourse analysis and narratology.12 While outline), but also by a so called para-text.16 The para- in narratology the focus is more on the forms of tex- text describes all textual parts of a book, that are not tual representation, in linguistic discourse analysis the actual text itself: the designation of the text gen- the content itself as well as the form’s discursive ef- re in titles or subtitles (scientific report or study, his- fects are central.13 Furthermore, the already men- torical novel), the title and subtitle themselves, a po- tioned ‘meaningful factors’ can be grouped into tentially existing foreword as well as the epilogues or three main categories: a) the contextual frame, b) indexes.17 Especially the last three cannot be found

10 I prefer the German term Sachliteratur, as the English 14 The terminology used here is based on linguistic term puts the so labelled publications into a vague non-sta- discourse analysis as it allows to integrate the narratological tus. For the problems on this specific genre see also Hahne- aspects, see Gardt 2013. mann 2005. 15 The term ‘literature’ is used here as a pragmatic 11 Gardt 2013. category in order to differentiate it from other spheres of life 12 Felder, Gardt (eds.) 2015; Felder, Müller (eds.) and to avoid the term fiction. ‘Literature’ does not include any 2009; Nünning 1995. implications on the artistic quality of the novels. 16 13 Both, linguistic discourse analysis and narratology Genette 1989. support the analysis of different languages and while my research 17 The mentioned genre can be satiric and does not focuses mainly on writings in German, I will occasionally refer have to reflect the publication’s actual genre. Therefore, the to texts in other languages. designation of the genre is considered here as paratextual, while

West & East 267 Monografie, 4 Kathrin Schmitt

in every book, so that their presence or absence has and continuing with some broader examples from in itself analytical value. archaeological writing. The last category, the textual micro-structure denotes the actual text itself. Here the analysis includes the font, type of language or register (aca- 4. Case studies demic jargon or colloquial language) and terminology. Additionally, grammar and figures of speech The first case is the analysis of a book cover that is are studied, as well as the forms of argumentation a meaningful factor of a publication’s contextual and how they are used. frame. The historical novelDas zweite Buch is writ- The meaningful factors influence the categoriza- ten by classical archaeologist Michael Pfrommer tion of a publication as a whole, not one by one, as and was published by publisher Philipp von Zabern belonging more to a factual or fictional sphere. By in 2007.21 The cover is a montage of a watercolour analysing such meaningful factors, it is also possible painting from the Ishtar gate by Walter Andrae and to find hints of implicit bias or ideologies, such as a photograph taken by Birgit Betzelt, a photo jour- colonialism, eurocentrism, orientalism as well as an- nalist.22 The photo shows a desert scenery crossed by drocentrism and many other ‘-isms’.18 Most of them a person with a head scarf on a donkey. The paint- are historically and ideologically intertwined and ing of the Ishtar gate is emplaced between the sand thereby difficult to differentiate. However, they are dunes in the background.23 The purpose of the mon- all part of what I call here ‘hidden ideologies’: hid- tage is quite obvious: the picture makes the reader den because we are often unaware of them and be- aware of the story’s setting, which is taking place cause I consider them as structural parts not only of partly in the Near East (ancient Babylon and mod- our academic but also of our daily life. ‘Structural’ ern Iraq). The Near East is represented here by the is used here in a more extensive sense to include all stereotypes ‘desert’, ‘Ishtar gate’ and ‘person with aspects of life, such as political, social, institution- head scarf’. The donkey as a means of transporta- al and everyday structures, that are the context of tion underlines the novel’s historical character. Bet- our daily praxis, without reflecting or questioning zelt took the photo in Darfur, Sudan close to a refu- them at any given moment.19 I refer here in a broad- gee camp called Aboshok. The photo is showing an er sense to Grada Kilomba who maintains that the «internal displaced» woman.24 Due to the infor- question is not whether we, and hence our writing, mation given by the photographer and considering contain hidden ideologies, but rather how we can the political situation in Sudan one can potentially dismantle them.20 assume that the woman on this picture would not As historical novels are not subject to the aca- identify herself as ‘Arabian’ but as ‘African’.25 And demic conventions of neutrality and objectivity, these are already very generalizing and problematic hidden ideologies are more easily apparent than in academic historiography. To reveal some hidden 21 The second Book, literal translation. The title is a refer- ideologies, I will analyse some meaningful factors ence to Ezekiel’s prophecies, cf. Pfrommer 2007. Pfrommer, born in 1954 is currently adjunct professor at the University in a few case studies, starting with historical novels of Trier and wrote a series of historical novels taking place in the Ancient Near East and Ancient Egypt as well as some non- the text genre as a category in itself is part of the archi-text. Cf. fiction books. Genette 1989 as well as Nünning 1995, p. 157 ff. 22 Pfrommer 2007. 18 E.g. classism, imperialism, patriarchalism, racism, 23 For the original painting cf. Andrae 1992, p. 69, sexism. pl. 84. 19 While there is a discourse on Ideologies in Archaeology 24 Betzelt: Internal Displaced in Darfur. http://www. in the Anglo-American anthropology, the German archaeology birgitbetzelt.com/?id=6§ion=3&lang=en&fototask= is less concerned about it, Bernbeck, McGuire (eds.) 2011. bild&gid=168&fid=1509&folder=sudandarfur&forder=1 20 Kilomba focused in her work on racism, but (09/01/2019). her discourse can be partly transferred to other repressive 25 Betzelt: Internal Displaced in Darfur. http://www. systems, especially when these systems are interconnected, cf. birgitbetzelt.com/?id=6§ion=3&lang=en&fototask=text Kilomba 2008, p. 22. &gid=168&folder=sudandarfur (09/01/2019).

West & East 268 Monografie, 4 (Hi)storytelling: the ancient Near East in western historical novels and archaeological writing

labels themselves. The photo chosen represents con- common accusations concerning the historical nov- sequently less a stereotypical ‘Arabic woman’, but a els mentioned in the beginning. Additionally, the very Eurocentric worldview and thereby a western- novel addresses the author’s role in writing history ised imagination of an ‘Arabian’. Perhaps the use of through its main character, the historian and scribe stereotypes or general categories is not even so prob- Ethan. Ethan is constantly in conflict with himself lematic, as we are dependent on them, not only in while he is working on the ordered one and only true novels but in archaeology and other spheres of life report about the Israelite King David. On the one as well, to communicate about past and present side, he has to answer the government’s expecta- worlds.26 In this case it is rather the uncritical way tions, namely the wishes of David, Salomon’s son, of conceptualising and using stereotypes that cause which are clearly focused on the king’s legitimacy, a problem. The book’s cover becomes a paradigm of grandeur and moral impeccability. On the other Edward Said’s Orientalism and embodies how far side, Ethan’s sources portray a violent and – in reli- actual content and a respectful handling of sources gious terms – sinful man, whose legitimacy could be in a public discourse are sometimes considered less doubted. The scribe attempts to include the contra- important than the satisfaction of western expecta- dictions implicitly in the report but fails at the end tions and prejudices through the given fiction.27 and is therefore exiled. The novel’s theme and the The following case study is an example from the scribe’s role are confronting us on a meta-level with textual macro-structure as it concerns a novel’s top- questions of historiographical perspective: why and ic. In Stefan Heym’s The King David Report, his- for whom are we pursuing history in general and tory is used as a frame for a political statement for whom are we giving a voice in our histories?30 Fur- present conditions, rather than as a politicised way thermore, the novel is a constant reminder of times of retelling history in a novel.28 Heym used the his- during which archaeology and historiography were torical frame mainly for a socio-political criticism openly abused for the legitimation of a political re- directed at authoritarian states and their methods gime, be it the Nazi regime and later on the GDR of controlling the press, academics and thereby his- government, or broadly totalitarian Stalinism. tory. The author was always politically active and – In the next case study, I focus again on a factor of being Jewish – he had to flee Germany during the the textual macro-structure. I address a common lit- Nazi regime. Later he had to face suppression in the erary motif, exemplary in the novels Qatna and She US during the McCarthy era and then in the for- wrote on clay.31 In both novels the motif of the ‘in- mer GDR. Therefore, the novel’s topic reveals as dependent woman’ is used. A motif that figures a fe- much about the authoritarian regimes as it reveals male protagonist who is going to take her life in her about Heym’s own ideological stance in opposition own hands to make her way – not without strug- to them. Since history is used here as a frame for the gling – independently in a patriarchal society. Gen- political topic rather than the other way around, it is erally, she is still doing that with the help of men, not possible to unconditionally classify the book as mostly father figures, or is at least relying on their a historical novel.29 Because of its particular setting generosity. Relatively common is the final result, and structure, this text is at least partly beyond tra- where she ends up in a relationship with a man, op- ditional questions of fact and fiction and the most tional as expecting mother, and will therefore fulfil her duty as a ‘proper woman’ in society. In Qatna, 26 Cf. Lakoff 1987 (1990). the ‘independent woman’ is the city state’s princess 27 4 Said 1978 (2014 ). Kija. At first, she enjoys a childhood among the fa- 28 Heym 1972 (1992). The novel was written in English and translated by the author himself into German. This novel voured, as the king’s prioritised daughter. Later, her was already subject to several studies and literary critiques, personal odyssey brings Kija as Akhenaton’s second which all pointed out its politicised content: Ginz 2014; wife to Egypt, where she succeeds to free herself, Bernbeck 2005; Taberner 2000; Hutchinson 1986; Reich-Ranicki 1972. 30 29 Nevertheless it is – not only by me – considered a Cf. Ginz 2014. historical novel, e.g. Bernbeck 2005. 31 Courant 2009; Graetz 2013.

West & East 269 Monografie, 4 Kathrin Schmitt

finding peace and freedom in the relationship with the motif of the independent woman focuses less her childhood friend Talzu. In She wrote on clay, the on the role model of women in the Ancient Near ‘independent woman’ is named Iltani. She grows up East, but rather on a bourgeois European role model in a scribe’s house, being the preferred daughter too. of the eighteenth and nineteenth century and has Her father’s role model in mind, she wants to be- therefore probably little in common with women’s come a «nadītu» in the «gagû» of Sippar, which lives in the ancient Near East.36 Instead, the motif gives her the opportunity to become a scribe by her- «archaises» the idealized European role model via self.32 In the temple, she is challenged as woman, na- the historical representation by dating it back to an- ditu and scribe but after initial difficulties masters cient times and thus perpetuates the backwarded her tasks confidently. On her path, she is falling in concept until today.37 And this is a good occasion love with the king’s scribe Marduk-mušallim. To to turn towards archaeological writing and research, stay with him, she has to break with her family and as women as historical actors not only in the Near the temple community, but gives up everything for Eastern archaeology, but in archaeology in general marriage, where she is finally able to find her true were for decades totally silenced.38 When women destiny. are explicitly mentioned in the archaeological his- The motif of the independent woman refers to toriography, they are up to our days mostly pictured the female Entwicklungsroman and is internation- according to the same anachronistic model outlined ally well known from female novelists of the late in the historical novels as housewives and mothers.39 eighteenth and early nineteenth centuries, like Jane For this reason, the novel’s fiction is a fiction in -ar Austen or the Brontë sisters.33 At that time, women chaeology as well and this prefiguration is literally writers addressing contemporary women’s lives in blocking the view for establishing factual based his- society can be considered as practising acts of eman- torical narrations.40 cipation – even if these women published anony- The above-mentioned examples were mainly ad- mously or under (male) pseudonyms.34 Due to this, dressing the representation of history. My next ex- the motif can be considered to be anachronistic in ample focuses on the present representation of ar- historical novels about the Ancient Near East. In chaeologists as authors. The authors of a publication modern reception the motif of the independent are a meaningful factor of the contextual frame, as woman is reactionary, but as it can be easily trans- their own status or professionalism (e. g. trained ar- ferred into other patriarchal societies of any time chaeologist vs. novelist or journalist) on a certain and place, one can find it not only in historical nov- topic takes part in establishing a publication’s more els about the Ancient Near East, but also in comput- factual or fictional status and is therefore an impor- er games, comics or Disney movies.35 Consequently, (2017), Frozen (2013) or The little Mermaid (1989). 36 32 Cf. Graetz 2013. Naditu is the term for a legal po- On naditus cf. Yoffee 2005, pp. 116‑121. For other sition held by women, who were connected to certain temple examples cf. Bahrani 2001; Chavalas (ed.) 2012; Lion, precincts (gagû). Until today, their exact position in Babyloni- Michel (eds.) 2016. an society is not fully understood, but from what is known by 37 Cf. Röder 2013, p. 244. the texts, they were very independent and well-educated wom- 38 Cf. Yoffee 2005, p. 128; Chavalas (ed.) 2012, en, mostly from an elite background, see also Yoffee 2005, p. XII. pp. 116-121. 39 Röder 2013 is also pointing out, that the prefigura- 33 For the female Entwicklungsroman see Balmer 2011. tion of the bourgeois role model is concerning the interpreta- Especially Pride and Prejudice, Austen 1813 (2016), and tion of gender roles and family concepts in general. Sense and Sensibility, Austen 1811 (2003), as well as Jane 40 Cf. Röder 2013. Without any intention of belittling Eyre, Brontë 1847 (2014) can be read as well-known British past as well as current research made by and / or about wom- examples of the female Entwicklungsroman. en, it can be pointed out that this research field and connected 34 Lifelong Austen published under ‘by a Lady’ and the topics like an archaeology of gender are still on the fringes of Brontë sisters Charlotte, Emily and Anne under the male archaeological research while constantly facing a wide range of pseudonyms ‘Currer’, ‘Ellis’ and ‘Acton Bill’. inimical attitudes between ignorance and active refusal. Exem- 35 For computer games cf. Tomb Raider (1996) or Dis- plary shall be mentioned here Bolger 2012; Chavalas (ed.) honored 2 (2016). For comics cf. Wonder Woman (1941) or 2012 and with a focus on the situation in Germany Vogel Catwoman (1940). For Disney movies cf. Beauty and the Beast 2012.

West & East 270 Monografie, 4 (Hi)storytelling: the ancient Near East in western historical novels and archaeological writing

tant analytical category. For this analysis, scientif- er-relations established by language. The most com- ic and public CVs are the most useful source. The mon languages used in the western archaeological CVs of archaeologists are usually neutral and im- publications on the Ancient Near East are English, personal, especially on the webpages of the institu- French, German and Italian.42 Exemplary I mentions archaeologists are working at. The institutions tion here the Interdisziplinäre Studien zur Königs- have usually a template, which intends to establish gruft von Qatna (QS1), as the research at that site equality and leaves little space for individuality. The was contemporaneously carried out by a German- CVs include information about the current posi- Syrian, an Italian-Syrian as well as a Syrian team.43 tion, education, previous positions, experience in The publication includes an Arabic translation of field research, received grants, institutional respon- the preface, but all other 28 contributions are either sibilities and a list of publications. Information of written in German (19), English (6) or French (3). main research interests is optional. What the pub- Neither in the publication itself with a total of 606 lic CVs usually do not include is information about bibliographical references nor in the 19 page long the number of applications submitted and refused, project bibliography with 335 entries a publication times as part of the academic precariat, personal inin Arabic can be found.44 The example shows, that terests and background or political orientation.41 the European researchers publish to a large extend Consequently, every narration includes non-narrat- in their native language and refer mainly to publica- ed aspects and due to the narrated CVs conscious tions written in the European languages, but not to as well as unconscious ideologies and biases are well publications in the languages of the host countries. hidden. By listing all successes while concealing Reverse the Syrian researchers – in this specific case failed applications for jobs and grants the CVs are – publish mainly in French, English and German creating a picture of almost infallible researchers, but less in Arabic. This imbalance is on a large scale who are supposed to be the legitimate neutral and reflected in the German and broader Western pub- factual narrators of the Ancient Near Eastern his- lications, as most of the research published in Ar- tory. Undoubtedly, there are very good reasons for abic, Turkish or Persian is only rarely considered.45 this, and by pointing the problem out I do not want Furthermore, German students of the Near Eastern to call for the ‘transparent Archaeologist’ but sug- archaeology need to be able to read and speak – be- gest that it is worth to question the author’s (self-) sides German – at least English and most common- representation and its impact. ly French or Italian to pass their studies, but not Ar- Another example for hidden ideologies in west- abic, Persian or Turkish.46 In most jobs within the ern archaeological writing are the languages in use – not only in publications but also in the field doc- 42 I make no claims of completeness here and refer only to umentation. Focusing here on the archaeological the most visible languages in publications and in conferences. 43 publications, language is a meaningful factor on two Pfälzner, al-Maqdissi (eds.) 2011b. The quite recent publication represents the present state-of-the-art levels: The publication’s language itself is a meaning- in archaeological publications. The edited volume includes ful factor of the contextual frame and the languag- contributions from researchers of varying countries and es of the secondary sources, as they become visible disciplines. Furthermore, I choose this example, as the site is already addressed through the novel Qatna in this paper. through the bibliography, affect the textual macro- 44 Double mentions were excluded from that number. structure. The bibliography visualises a publication’s The Qatna-Bibliography covers all publications on the site be- intertextual connections within the archaeological tween 1924 and 2013: http://www.qatna-projekt.uni-tuebin- discourse and is therefore a good indicator for pow- gen.de/PDF/Bibliographie/Bibliographie.pdf (09/01/2019). 45 The situation is actually much better for Turkish in 41 Compare for example the faculty member’s CVs of the comparison to other languages. Institute for Near Eastern Archaeology at the Freie Universität 46 At the Freie Universität Berlin, English is required for Berlin: http://www.geschkult.fu-berlin.de/e/vaa/vaa/mitar- the Bachelor and the students can choose between Latin, French, beiter/index.html (09/01/2019). A rare exception is the post- Italian, Russian, Arabic, Persian or Turkish for the Master. Since modern alternative CV from Thomas Meier, currently at the the German Abitur requires already two foreign languages (most- University of Heidelberg: http://www.ufg-va.uni-hd.de/mitar- ly English and Latin or French), most students are already cov- beiter/meier_vita2.html (09/01/2019). ered and the need of learning another language is minor.

West & East 271 Monografie, 4 Kathrin Schmitt

field of Near Eastern archaeology the knowledge of logical Institute’s journal Zeitschrift für Orient- theses languages is welcomed but not required ei- Archäologie or – as an American example – the ex- ther. This state is by no means a normal condition hibition catalogue Pioneers to the past: American in all archaeological disciplines, since for the archae- archaeologists in the Middle East.50 All these terms ology of Eastern Europe or Russia the host country’s and therefore the publications that use them estab- languages are nearly always required. As in the case lish a geographical relation, with Europe as the cen- of the CVs, there may be reasons for the require- tral actor and the Near East as its peripheral region. ment of specific languages used in western archae- As soon as one adds the perspective of time as in the ology of the Near East. But the well-established re- notion ‘Ancient Near East’, occasionally comple- search languages are to a large extent the languages mented by ‘the cradle of civilization’ and / or refer- of the former colonialists and mandatory powers in ences to the Old Testament, ancient societies and the Near East.47 Simply by using these languages, cultures of the Near East are implicitly put in op- historically based imbalances of power are perpet- position to the present ones. At the same time, the uated on a structural level.48 Furthermore, they are terms are establishing a more or less subtle claim of providing very little access to research and publica- ownership of that history in the hands of today’s tions in the host countries. Simultaneously it goes western societies via a «translatio imperii», respec- without saying that colleagues from Near Eastern tively a linear construed genealogy.51 A version of countries have to learn what is commonly referred this can be read in the foreword of the mentioned to as research languages. Meanwhile, the vast major- catalogue Pioneers to the past: ity of western archaeologists never reach a fluency in Arabic and Persian or Turkish that enables them to Breasted immediately grasped the significance of this publish or give a talk. Thus the structures formerly historical moment and approached John D. Rocke- established by the colonialist systems are implicitly feller Jr. with an ambitious and detailed proposal for continued by academic conventions, in which pub- financial support to found the Oriental Institute as lications in the present Near Eastern languages ex- an interdisciplinary center for archaeological and philological research on the ancient Near East and its perience very little credit at an international level. role in the origins of Western civilization.52 Making one step further into archaeological writing, one can focus on the terminology and Thereby imbalances of power are reproduced again. grammar used in publications. Both are meaning- Consequently, I stop here writing about the ‘Near ful factors of the textual micro-structure. Very com- East’ and continue this paper by using ‘Western mon are still the terms Near or Middle East or the Asia’ instead. This may not be the best alternative, (ancient) Orient for the research area.49 Exemplary since the geographical term is less precise in the ex- I mention here the publication Ausgrabungen und act location and spatial extent, but the use of alter- Surveys im vorderen Orient, the German Archaeo- native terminologies creates awareness due to irrita- tion and sensitizes for the problems set out above. 47 Even if Germany was never a colonial power in the Near East, I would not exclude it, as I agree with March- A grammatical aspect of the textual micro-struc- and 1996, pp. 199‑208, who is pointing out the German qua- ture to be discussed here is the use of the generic si- (cultural) imperialist ambitions in the Near East and the role masculine. The generic masculine means the use of archaeology in it. The official non-colonialist status made it comparably easy for German archaeologists to avoid an engage- of the male grammatical gender in a generic sense ment with postcolonial studies for decades, especially, as the for individuals, whose biological sex is unspecified role of Germany as political actor in the Near East was not con- or unknown as well as for mixed groups of persons. sidered by Said 1978 (20144) either. 48 Of course this is not a unique problem for Near 50 Eastern archaeology, as one can find it in varying manifestations Eichmann (ed.) 2010; Emberling (ed.) 2010. nearly all over the world. Cf. Castro Varela, Dhawan 51 Liverani 2005, p. 224 describes the translatio imperii 2005 (20152), p. 34 ff. as «imperial power shifted from Assyria and Babylonia to Me- 49 Same for Naher / Mittlerer Osten, Alter Orient, le dia and Persia, then to the Greeks, finally to the Romans». Proche- / Moyen Orient or il vicino / medio oriente (antico). 52 Stein 2010, p. 7.

West & East 272 Monografie, 4 (Hi)storytelling: the ancient Near East in western historical novels and archaeological writing

While its impact in English is comparably minor persönliche Identität zu konservieren».57 Especially (but still existing), it is fatal to a language like Ger- through the singular male forms the text gives the man, which uses gendered articles as well as female- impression that only men could have had personal associated nouns that derive from the masculine grave goods, while the vast majority of the objects version and therefore seem subordinate to them.53 found in the royal tomb cannot be brought togeth- Similar problems occur in other languages such as er with a specific individual and moreover, the ones French or Italian too, but since my expertise lies in identified are of both sexes.58 The examples repre- the German language I restrict myself to this exam- sent the present state of language use in Germany ple. Critics of the use of the generic masculine are and show clearly, that past and present women as focusing mainly on the marginalisation of women historical actors and active archaeologists are textu- because many readers and writers associate them ally marginalised. less with what is said when using or reading the ge- Unlike common expectations the generic mas- neric masculine and women feel less likely included culine is not a continual and therefore unquestion- as well.54 Some examples can highlight this process able grammatical phenomenon, but was established of marginalisation. Due to the paper’s limitation in grammars with its full implications in 1850 in the I stick here to the already introduced QS1. In this United Kingdom and 1962 in Germany.59 It is no work, present archaeologists and past individuals surprise that this moment falls together in the UK are addressed, whereby the archaeological present with the ‘golden age’ of the female Entwicklungsro- as well as the historical past is affected. While for man, since bourgeois role models and generic mas- the acknowledgements the Binnen-I in «Mitarbeit- culine reflect an androcentric worldview in the end. erInnen» was used to address female and male staff This ideology is a persistent one and the efforts to members, readers are addressed in the singular and overcome it are rare, as the continuous use of the male form of «der Leser» only.55 Referring gener- generic masculine and the motif of the independ- ally to archaeologists and their tasks within the re- ent woman illustrate. Finally, the writer’s and read- search the male form is used only, as for the archae- er’s inability of including women in their thoughts ological «Bearbeiter».56 For historical individuals while using the generic masculine generates, in male forms in their generic function were used for terms of fact and fiction, academically written his- persons with unspecified or unknown sex, as it can torical and present worlds that are predominant- be seen in setting the theoretical frame for an inter- ly populated by men and are in this aspect simply pretation of the objects found in the royal tomb: fictional. Gender awareness in language is conse- «Dies sind solche Objekte, die kein spezifisches quently not only a feminist topic pushing for gen- Statussymbol darstellen, sondern die der Verstor- der equality, but rather an indicator for a factually bene zu Lebzeiten persönlich besaß und die ihm im oriented historiography. Tod mitgegeben werden, um seinen Besitz und seine

5. Conclusion 53 E.g. Archäologe (m) – Archäologin (f). In English e.g. craftswoman – craftsman, the male form is used generic. 54 A number of studies attest the marginalisation of The discussed cases clearly show that neither fic- women through the generic masculine in English as well as in tional narrative strategies nor (hidden) ideologies German, cf. Braun et Al. 1998; 2007; Doleschal 2002; are unique to fictional (hi)storytelling, respectively Gastil 1990; Hamilton 1988; Hamilton, Hunter, Stuart-Smith 1993; Miller, James 2009; Stahlberg, historical novels, as both can be found on all lev- Sczesny 2001. els of archaeological writing (contextual frame, tex- 55 Pfälzner, al-Maqdissi 2011a, pp. IX‑X, the French version of the Preface uses male forms only as well. 57 The German Binnen-I, so the capital letter I in a word’s female Pfälzner 2011, p. 49, male forms highlighted by the grammatical form is one possibility to indicate, that both bio- author. logical sex are meant. 58 Witzel 2011. 56 Pfälzner 2011, p. 39. 59 Doleschal 2002.

West & East 273 Monografie, 4 Kathrin Schmitt

tual macro-structure and textual micro-structure) toriography which is aware of its own ideological as well. Furthermore the two genres share common bias and is dealing with them in a responsible way pitfalls in the use of stereotyping, anachronisms should be attempted. Consequently, a critical un- and perspectivation, which include common preju- derstanding of the functions of language and lin- dices. From this point of view the mantra-like ad- guistic discourse is not only a matter of factuality herence to neutrality and objectivity especially in or fictionality, but it is also a useful method of re- the German archaeology of Western Asia becomes vealing hidden ideologies in archaeological writing. a fiction itself and is rather suitable to hide than to At the end, I want to return to the initial men- reveal ideologies and bias in factual historiography. tion of the March for Science movement. As a side Some of the pitfalls can be easily avoided by using effect, this brief comparative analysis of historical for example a gender-aware language instead of the novels and archaeological writing elucidates that generic masculine, so that traditionally perpetuat- popular media and academic discourse are influenced inequalities based on fiction are converted into ing each other. Therefore, the analysis enters the on- inclusive factually based historiographies. Others going public discourse on the production of facts in – like bourgeois role models, research languages or science, the humanities and public media and ad- terminology are firmly established in a westernised dresses archaeological responsibilities in it. The per- understanding of the world and academic struc- petuated inequalities and imbalances in the writing tures or conventions. They cannot just be avoid- of western Asian archaeology, revealed in the case ed by dropping them. Dismantling them requires studies of the author’s (self-) representation, re- a conscious, self-critical and creative approach as search languages, terminology and grammar, point well as an open and active discourse within the field out that archaeologists are not only in charge of re- of archaeology to focus on a transparent factually- search and archaeological fact production, but also based archaeological writing. Here the difference in the power over history. However, rather than between factuality and objectivity clearly emerg- claiming interpretative authority legitimised by ac- es and indicates, that the resulting historiography ademic neutrality and objectivity, an open minded will not become more ‘objective’, ‘neutral’ or ‘un- and self-reflexive approach to our research objec- biased’. In conclusion the unbiased writing is not tives may lead to a diverse and inclusive engagement possible and cannot be the goal either, but a his- in public discourse.

Acknowledgements

The research presented in this paper is part of my PhD, which is supported by the Gerda Henkel Stiftung. Also, I thank the Organizing Committee of the Broadening Horizons 5 Conference for all their efforts and Reinhard Bern- beck, Maria Bianca D’Anna, Nathalie Kallas, Arnica Keßeler and Chiara Schoch for their comments on the draft.

West & East 274 Monografie, 4 (Hi)storytelling: the ancient Near East in western historical novels and archaeological writing

BIBLIOGRAPHY

Andrae E.W. 1992, Bilder eines Ausgräbers: Die Clements R., Musker J. (directors) 1989, The little Orientbilder von Walter Andrae 1898 – 1919. Mermaid, United States. Sketches by an excavator, Berlin. Clifford J., Marcus G.E. (eds.) 1986, Writing cul- Aust H.1994, Der historische Roman, Stuttgart. ture: the poetics and politics of ethnography; a school Austen J. 1811 (2003), Sense and sensibility, London, of American Research Advanced Seminar, Berkeley, New York. Calif. Austen J. 1813 (2016), Pride and Prejudice, London, Condon B. (director) 2017, Beauty and the Beast, New York. United States. Bahrani Z. 2001, Women of Babylon: gender and repre- Courant M. 2009, Qatna: Historischer Roman aus der sentation in Mesopotamia, London. Zeit Echnatons, Ostfildern. Balmer S. 2011, Der weibliche Entwicklungsroman: in- Doleschal U. 2002, Das generische Maskulinum im dividuelle Lebensentwürfe im bürgerlichen Zeitalter. Deutschen. Ein historischer Spaziergang durch die Literatur – Kultur – Geschlecht. Studien zur deutsche Grammatikschreibung von der Renaissance Literatur- und Kulturgeschichte. Grosse Reihe, Köln, bis zur Postmoderne, «Linguistik Online» 11/2, Weimar, Wien. pp. 39-70. Bernbeck R. 2005, The Past as Fact and Fiction: From van Dyke R.M., Bernbeck R. (eds.) 2015, Subjects and Historical Novels to Novel Histories, in: Pollock S., narratives in archaeology, Boulder. Bernbeck R. (eds.), Archaeologies of the Middle East Eichmann R. (ed.) 2002, Ausgrabungen und Surveys im 4, Blackwell studies in global archaeology, Malden, Vorderen Orient 1 (Orient-Archäologie 5), Rahden/ Mass., pp. 97-121. Westfahlen. Bernbeck R., McGuire R.H. (eds.) 2011, Ideologies in Emberling G. (ed.) 2010, Pioneers to the past: American archaeology, Tucson, 2011. archaeologists in the Middle East, 1919-1920, Braun F., Gottburgsen A., Sczesny S., Stahlberg, Chicago. D. 1998, Können “Geophysiker” Frauen sein? Felder E. (ed.) 2013, Faktizitätsherstellung in Generische Personenbezeichnungen im Deutschen, Diskursen: Die Macht des Deklarativen. Sprache und «Zeitschrift für Germanistische Linguistik» 26/3, Wissen, Berlin. pp. 265-283. Felder E., Gardt A. (eds.) 2015, Handbuch Sprache Bolger D. (ed.) 2012, A companion to gender prehistory. und Wissen. Handbücher Sprachwissen Band 1, Hoboken, NJ. Berlin-Boston. Braun F., Oelkers S., Rogalski K., Bosak J., Sczesny Felder E., Müller M. (eds.) 2009, Wissen durch S. 2007, “Aus Gründen der Verständlichkeit...”: Sprache: Theorie, Praxis und Erkenntnisinteresse des Der Einfluss generisch maskuliner und alternativer Forschungsnetzwerkes Sprache und Wissen, Sprache Personenbezeichnungen auf die kognitive Verarbeitung und Wissen, Berlin, New York. von Texten, «Psychol. Rundsch.» 58/3, pp. 183-189. Finger B., Kane B. (creators) 1940, Catwoman. Brontë C. 1847 (2014), Jane Eyre, München. Gard T. (developer) 1991, Tomb Raider. Buck C., Lee J. (directors) 2013, Frozen, United States. Gardt A. 2013, Textanalyse als Basis der Diskursanalyse. Castro Varela M. do M., Dhawan N. 2005 (20152), Theorie und Methoden, in: Felder E. (ed.) 2013, Postkoloniale Theorie: eine kritische Einführung, pp. 29-55. Bielefeld. Gastil, J. 1990, Generic Pronouns and Sexist Language: Chavalas M.W. (ed.) 2012, Women in the ancient The Oxymoronic Character of Masculine Generics, Near East: a sourcebook, Routledge sourcebooks for the «Sex Roles» 23/ 11-12, pp. 629-643. ancient world, London. Genette G. 1989, Paratexte: das Buch vom Beiwerk des Buches, Frankfurt am Main, 1989.

West & East 275 Monografie, 4 Kathrin Schmitt

Ginz C.C. 2014, Die Ungefragten der Geschichte: Miller M., James L.E. 2009, Is the generic pro- eine Lektüre von Stefan Heyms «Der König David noun he still comprehended as excluding women?, Bericht» mit dem Diskurs um Gedächtnis und «The American Journal of Psychology» 122./4, Geschichte, Literatur – Medien – Religion 22, Berlin. pp. 483-496. Graetz S. 2013, She wrote on clay: a novel, Overland Nünning A.1995, Von historischer Fiktion zu histo- Park. riographischer Metafiktion, Literatur, Imagination, Hahnemann A. 2005, Aus der Ordnung der Fakten. Realität, Trier. Zur historischen Gattungspoetik des Sachbuchs, Paul I.U., Faber R. (eds.) 2013, Der historische Arbeitsblätter für die Sachbuchforschung 9, Berlin. Roman zwischen Kunst, Ideologie und Wissenschaft, Hamilton M.C. 1988, Using masculine generics: Does Würzburg. generic He increase male bias in the user’s imagery?, Pfälzner P. 2011, Das systemische und das archäolo- «Sex Roles» 19/ 11-12, pp. 785-799. gische Inventar der Königsgruft von Qatna und seine Hamilton M.C., Hunter B., Stuart-Smith S. Interpretationsmöglichkeiten, in: Pfälzner P., al- 1993, Jury Instructions Worded in the Masculine Maqdissi M. (eds.) 2011b, pp. 39-52. Generic: Can a Woman Claim Self-Defense When Pfälzner, P., al-Maqdissi, M. 2011a, Vorwort der “He” is Threatened?, in: Roman C., Juhasz S., Herausgeber, in: Pfälzner P., al-Maqdissi M. Miller C. (eds.), The Women and language debate: (eds.)2011b, pp. VII-X. a sourcebook, New Brunswick, pp. 340-347. Pfälzner P., al-Maqdissi M. (eds.) 2011b, Heym S. 1972 (1992), Der König-David-Bericht: Roman, Interdisziplinäre Studien zur Königsgruft in Qatna.̣ Frankfurt am Main. Qatnạ Studien 1, Wiesbaden. Hutchinson P. 1986, Problems of Socialist Pfrommer M. 2007, Das Zweite Buch. Ein historischer Historiography: The Example of Stefan Heym’s Roman, Mainz am Rhein. “The King David Report”, «The Modern Language Röder B. 2013, Urmenschliche Bürger – bürgerli- Review» 1/81, pp. 131-138. che Urmenschen. Zur Archaisierung des bürgerli- Joyce R.A., Preucel R.W. 2002, The languages of ar- chen Geschlechter- und Familienmodells über die chaeology: dialogue, narrative, and writing, Social ar- Urgeschichte, in: Grisard D., Jäger U., König T. chaeology, Oxford. (eds.), Verschieden sein: Nachdenken über Geschlecht Kilomba G. 2008, Plantation memories: episodes of und Differenz, Sulzbach, pp. 243-256. everyday racism, Münster. Rorty R. (ed.) 1968, The Linguistic turn: essays in philo- Lakoff G. 1987 (1990), Women, fire, and dangerous sophical method, Chicago. things: what categories reveal about the mind, Chicago. Said E.W. 1978 (20144), Orientalismus, Frankfurt am Lion B., Michel C. (eds.) 2016, The Role of Women Main. in Work and Society in the Ancient Near East, Schaeffer J.-M. 2010, Why fiction?, Lincoln. Boston-Berlin. Smith H. (director) 2016, Dishonored 2. Liverani M. 2005, Imperialism, in: Pollock S., Stahlberg D., Sczesny S. 2001, Effekte des generischen Bernbeck R. (eds.), Archaeologies of the Middle East Maskulinums und alternativer Sprachformen auf den 4, Blackwell studies in global archaeology, Malden, gedanklichen Einbezug von Frauen, «Psychologische Mass., pp. 223-243. Rundschau» 52/3, pp. 131-140. Maier V., Schweizer P. 1999, Dichtung und Stein G. J. 2010, Foreword, in: Emberling G. (ed.), Wahrheit: Zum Widerstreit von Wissenschaftlichkeit Pioneers to the past: American archaeologists in the und literarischer Fiktion im modernen prähistori- Middle East, 1919-1920, Chicago. schen Roman, in: Kümmel C., Müller-Scheessel Storm C. 2010, N., Schülke A. (eds.), Archäologie als Kunst: Die Imagination der Geschichte: Darstellung, Wirkung, Kommunikation, Tübingen, Authentizität, Historizität, Widerstand und pp. 147-160. Identität in chinesischen historischen Romanen, Veröffentlichungen des Ostasien-Instituts der Ruhr- Marchand S.L. 1996, Down from Olympus: archae- Universität, Wiesbaden. ology and philhellenism in Germany, 1750-1970, Princeton, N.J. Taberner S. 2000, The Writer’s Fascination With Power: Stefan Heym’s Der König-David-Bericht, Marston W.M., Marston E.H. (creators) 1941, «Neophilologus» 84, pp. 271-283. Wonderwoman.

West & East 276 Monografie, 4 (Hi)storytelling: the ancient Near East in western historical novels and archaeological writing

Vogel H. 2012. Das Konzept Gender in der Witzel C. 2011, Anthropologische Untersuchungen. Vorderasiatischen Archäologie, in: Günther-Saeed Vorläufige Ergebnisse der Untersuchung menschli- M., Hornung E. (eds.), Zwischenbestimmungen: cher Knochen und Zähne aus der Königsgruft seit der Identität und Geschlecht jenseits der Fixierbarkeit?. Bergung 2002, in: Pfälzner P., al-Maqdissi M. Würzburg, pp. 121‑137. (eds.) 2011b, pp. 367-382. White H.V. 1973, Metahistory: The historical imagina- Yoffee N. 2005, Myths of the archaic state: evolution of tion in nineteenth-century Europe, Baltimore. the earliest cities, states and civilizations, New York.

Sitography

Betzelt B. Internal Displaced in Darfur. http://www.birgitbetzelt.com/?id=6§ion=3&lang=de&gid=168&fotot ask=detail (09/01/2019). Freie Universität Berlin. Mitarbeiter/innen am Institut für Vorderasiatische Archäologie. http://www.geschkult.fu- berlin.de/e/vaa/vaa/mitarbeiter/index.html (09/01/2019). March for Science. https://www.marchforscience.com/ (09/01/2019). Meier T. Vita post-modern. http://www.ufg-va.uni-hd.de/mitarbeiter/meier_vita2.html (09/01/2019). Qatna-Projekt. Qatna-Bibliographie. http://www.qatna-projekt.uni-tuebingen.de/PDF/Bibliographie/Bibliographie. pdf (09/01/2019). Reich-Ranicki M. 1972, Eine nur leicht verkappte Polemik aus der DDR: König David alias Stalin, «Zeit», No. 33 (http://www.zeit.de/1972/33/koenig-david-alias-stalin/komplettansicht). Schaeffer J.-M. 2012, Fictional vs. Factual Narration. the living handbook of narratology, (http://www.lhn.uni-ham- burg.de/article/fictional-vs-factual-narration) (09/01/2019).

West & East 277 Monografie, 4

The Tell Gomel archaeological survey. Preliminary results of the 2015-2016 campaigns

FRANCESCA SIMI

University of Oxford

Abstract

This paper presents the methodology, goals and preliminary results of the Tell Gomel Archaeological Survey. The project has its origins in the wider ‘Land of inevehN Archaeological Project’ (LoNAP) which has been conducted since 2012 by the University of Udine in the northern Region of Iraqi Kurdistan. The area examined by the project is the heart of the Navkur Plain, an alluvial plain that covers the eastern hinterland of the Assyr- ian capital of Nineveh. The focus of settlement throughout the entire Navkur Plain was the ites of Tell Gomel, where a preliminary survey documented the existence of a settlement sequence ranging from the Chalcolithic to the Ottoman period. Due to its large size and position in the centre of the plain, Gomel must have played an important role in this region, presumably as its political and economic centre. The area around Gomel is also of great interest because of its position in the heart of the Navkur Plain, a trade route hub from the Late Chalco- lithic onwards, and the main focus of settlement for a much wider region. The project therefore aims to investigate the archaeological landscapes of this crucial and still unknown area.

Keywords

Iraqi Kurdistan, Upper Mesopotamia, Landscape Archaeology, Intensive Survey

West & East 279 Monografie, 4 Francesca Simi

1. Introduction large parts of the governorates of Ninawa and Duhok and aims at understanding the transformation of the This paper aims at illustrating the activities and the cultural and natural landscapes of a key region of preliminary results of an intensive survey carried out Northern Mesopotamia located in the hinterland of in the Navkur Plain during summer 2015 and 2016.1 the Assyrian capital of Nineveh. 2 The Tell Gomel Archaeological Survey is part of The LoNAP is itself one of four internation- the wider ‘Land of Nineveh Archaeological Project’ al archaeological projects that focus on the area of (LoNAP) which has been conducted since 2012 by north-eastern Iraqi Kurdistan, called the Assyrian the University of Udine under the direction of Prof. Landscape Research Group (ALRG, fig. 1). 3 Morandi Bonacossi in the northern part of the au- Due to the vast size of the region the research- tonomous region of Kurdistan, Iraq. The LoNAP in- ers decided to adopt a strategy based on a combi- vestigates an area of 3000 km2 in a region embracing nation of motor-vehicle and pedestrian field survey

Figure 1 Location of the various projects working in Northern Iraqi Kurdistan, namely, the Erbil Plain Archaeological Survey (directed by Jason Ur), the Upper Greater Zab Reconnaissance Project (directed by Rafał Koliński), and the Eastern H abur Archaeological Survey (directed by Peter Pfälzner)

2 1 This article is an outcome of the PRIN 2015 project For a preliminary overview see: Morandi Bonacossi “Archaeological Landscapes of Ancient Iraq from Prehistory 2012-2013, 2016; Morandi Bonacossi, Iamoni 2015. to the Islamic period: formation, transformation, protection, 3 For the preliminary reports of the other ALRG projects and management” funded by the Italian Ministry of Education, see: Pfälzner, Sconzo 2015, 2016; Pfälzner et Al. 2016; University and Research (Grant no. 20154X49JT). Koliński 2016a, 2018; Ur et Al. 2013.

West & East 280 Monografie, 4 The Tell Gomel archaeological survey. Preliminary results of the 2015-2016 campaigns

of sites previously identified as possible settlements atic and intensive survey system that follows a tradi-� through remote-sensing analysis. Therefore, until tion well established in the Mediterranean and Near the summer of 2015 no small-scale intensive survey Eastern landscape archaeology. It aims at testing the had been carried out in the Navkur Plain, although accuracy and the advantages of the system to an area a large number of sites were identified by the exten- already investigated with an extensive strategy. sive survey.4 The intensive field survey is crucial in order to -es The area examined by the Tell Gomel Archaeolog- tablish the presence of possible sites not recognized ical Survey (henceforth TGAS) is located in the the through the remote sensing analysis, or the exten- heart of the Navkur Plain (fig. 2), and extends south sive survey, and to explore ancient land-use, in par- of the Erbil-Duhok road, between the Jebel Maqloub ticular, with the field scatters recording. and the Bardarash region to the south and the Riv- Furthermore, we were interested in understand- er Al-Khazir to the east. The limits of the survey area ing whether the intensification of the survey might have been (arbitrarily) defined as a square centred on confirm, or perhaps deny, the diachronic patterns of Gomel, measuring 10 km on each side and covering settlement already obtained by the more extensive a total area of 100 km2. The TGAS applies a system- investigation.

Figure 2 The Tell Gomel Archaeological Survey location within the LoNAP licensed area

4 For an extensive discussion of the survey methodology and preliminary results see Morandi Bonacossi, Iamoni 2015.

West & East 281 Monografie, 4 Francesca Simi

Figure 3 The archaeological site of Gomel

2. Gomel and its region high productivity agriculture. It is therefore not surprising that the largest archaeological site in the en- The Navkur is a 30-kilometre wide and roughly tri- tire region is situated in this plain. angular plain in front of the Zagros foothills, it is This is the site of Gir-e Gomel (in local Kurd- crossed by the River Al-Khazir, a major tributary ish) or Tell Gomel (in Arabic),9 located on the River of the Greater Zab, and the minor Nardush and Gomel near the modern town of Kalakchi. Gomel Gomel rivers that join the Al-Khazir at the south- possibly corresponds with the Assyrian Gammagara ern end of the Navkur Plain. Numerous wadis and mentioned in the Jerwan B Inscription of Sennach- ephemeral water streams feed the main watercours- erib. 10 Furthermore, Sir Aurel Stein in his Limes es, however the plain is not only extremely rich in Report (1938-1939, published in 1985) proposed surface waters but also in ground waters thanks to that the battle of Gaugamela took place in the plain many springs, sustained by the aquifers located in surrounding Gomel, where in 331 BC Alexander the Zagros Range. 5 the Great defeated Darius III, thus definitively con- The plain it is also characterised by deep and quering the Achaemenid Empire. 11 fertile agricultural soils, mainly “Brown Soils The site features an elevated upper town (about (Deep Phase)” developed on alluvial sediments.6 40 m above the surrounding plain) that dominates Mean annual rainfall in this area is around 643 an extensive lower town (fig. 3). Due to its large size mm per year, 7 this means that today the region is and position in the centre of the plain, Gomel must located north of the so called “zone of uncertain- have played an important role in this region – pre- ty” and is part of the stable rain-fed zone of Upper sumably as its political and economic centre. Mesopotamia. 8 The preliminary survey of the Gomel archaeo- The abundancy of available water and the pres- logical site conducted by LoNAP in 2012 was able ence of deep fertile soils offer ideal conditions for to recognize the existence of a continuous settlement sequence ranging from the Chalcolithic to 5 For a first geoarchaeological assessment of the Navkur the Ottoman period (c. 5000 BC-early twentieth Plain see Morandi Bonacossi 2018. 6 Buring 1960. 9 7 The rainfall data are acquired from the climatic records Site n. 52 in Salman 1970, 280; 1976, Map 128. of the meteorological station of Bardarash and they are freely 10 Reade, Anderson 2013, 74; Fales, Del Fabbro accessible at www.Climate-Data.org. 2014. 8 Wilkinson 2004: 14; Wilkinson et Al. 2014. 11 See Stein, Kennedy and Gregory 1985.

West & East 282 Monografie, 4 The Tell Gomel archaeological survey. Preliminary results of the 2015-2016 campaigns

century AD). However, the site could have possi- identified sites using a strategy of complete cover- bly been occupied already in the Ubaid period since age throughout collection areas (ca 1 ha) selected Frankfort published an Ubaid stamp seal found at according to the site topography. 18 The corners and Gomel in the ’30s, now preserved in the Oriental the significant points on the boundaries of each unit Institute Museum of Chicago. 12 were visibly marked and their positions recorded via It is now widely recognized that no ancient GPS; only the diagnostic sherds were collected. site, especially an “urban” site, can be really un- Concerning the off-site,19 the methodology derstood without considering its hinterland, 13 for used consists of fieldwalking through long tran- this reason the TGAS was conceived as an inten- sects. These transects were covered by four sur- sive reconnaissance with a special interest in off- veyors positioned at a distance of 25 m from each site investigation. other. While walking along these predetermined transects, team members counted the number of pottery sherds visible on the surface with a tally 3. The fieldwork methodology counter and collected diagnostic artefacts.20 Eve- ry hundred metres the team stopped, and the data The fieldwork was preceded by the analysis and -in and ground observations were recorded in a stand- terpretation of the available cartographic material ard form. Due to the variable visibility conditions, and the satellite and aerial images. 14 The images that due mainly to different landuse, 10 × 10 m collec- until now have produced the most information are tion units were positioned along the transects at a the declassified CORONA satellite images, but the hundred metre intervals. These sample units were newly acquired U2 images, 15 which have a better selected according to the surface visibility and resolution and are slightly older than the CORO- ground conditions and were covered by only one NA, provided new interesting data. 16 This is espe- surveyor walking back and forth, trying to main- cially true for the site of Gomel, where the U2 im- tain the same light conditions. The placement of ages provided new insights regarding the shape and field scatter collection units was chosen with the dimension of the ancient settlement. aim of selecting homogeneous and high visibility In the field, the team used a handheld Global Po- conditions to ensure the compatibility of the data sitioning System (GPS) receiver17 to record all the collected, but as result of these requirements some information and observations. Back home, all these areas were left unrecorded. data were eventually included into a Geographic In- In nearly four months of intensive survey (2015- formation System (GIS) spatial database. 2016) we completed 58 transects and covered a The collection and survey strategy in Gomel was total area of 25 km2. The aim was to investigate a different from that adopted for the off-site survey, quarter of the entire TGAS area reaching an almost and for the investigation of new sites discovered and complete coverage (fig. 4). In only two campaigns a the re-reconnaissance of the sites already identified total of 1183 squares of 100 × 100 m were surveyed by the LoNAP team in the area. Regarding the lat- and data from 632 sample units was recorded. Un- ter, the methodology was quite common and wide- fortunately, several areas remained not investigated spread. The materials were collected from all the due to the thick straw layer covering the field sur-

12 Frankfort 1935, pp. 29-31, fig. 31 (A12466). 18 Many Near Eastern surveys have already employed 13 Wilkinson 2003: 100; Ur 2010: 1. these methods, see: Ball, Tucker and Wilkinson 1989; 14 The project used declassified U2 aerial photographs Lyonnet 1990; Wilkinson, Tucker 1995; Wilkinson and CORONA, Hexagon, Aster and OrbView-3 satellite im- 2000; Ur 2010. agery. 19 Similar sampling techniques have already been used in 15 I would like to thank Professor J. Ur who shared with several North Mesopotamian surveys: Wilkinson, Tucker me the U2 images in his possession. 1995; Ur, Karsgaard and Oates 2011; Ur 2010. 16 Hammer, Ur 2019. 20 Rims, bases, handles and decorated body sherds are 17 A Garmin Oregon 650 was used in the field. considered diagnostic.

West & East 283 Monografie, 4 Francesca Simi

Figure 4 The transects area within the TGAS project limits

Figure 5 Gomel in a CORONA image (A – Corona 1039, 28 Feb 1967) and in a U2 image (B – U2 1554, 29 Jan 1960)

West & East 284 Monografie, 4 The Tell Gomel archaeological survey. Preliminary results of the 2015-2016 campaigns

Figure 6 The intensive survey in Gomel

face as the result of harvesting activities. 21 This oc- collected data from a single square walking back and curred especially in the south-western sector. forth counting the number of fragments visible on Coming to the investigation of Gomel, one of the surface and collecting only the diagnostic ones. the goals of TGAS 2016 campaign was the intensive Overall, we examined a total of 496 squares (fig. 6). and full coverage survey of the site. The aim was also All the pottery collected during the two field- to cover a “halo”, probably evidence of the presence work campaigns was analysed, counting and divid- of anthrosols, 22 visible in the U2 images immediate- ing all the potsherds on the basis of chronological ly east of the previously estimated limits of the site and typological classification. 23 The typological clas- (fig. 5). The presence of the anthrosols could be re- sification was based upon the Working Ceramic Ty- lated to the existence of an extended “outer town”, pology (2013 version) developed from several sur- and it would consequently modify our interpreta- veys of North Mesopotamia, originally developed tion of the site’s dimensions. for the Tell al-Hawa survey (Ball, Tucker, and For the intensive survey a grid of 25×25m Wilkinson 1989), later updated by T. Wilkinson squares was positioned over the site. Each surveyor and D. J. Tucker (1995) and by J. Ur (2010), and

23 I am grateful and really indebted to the pottery special- 21 Both TGAS survey campaigns was conducted between ists of LoNAP team for their help in the preliminary processing mid-August and mid-October. of the TGAS pottery. In particular, I have to thank: C. Cop- 22 On anthrosols detection on satellite image see Savioli pini, K. Gavagnin, M. Iamoni, R. Palermo, C. Tonghini and (forthcoming), Koliński 2016b and Menze, Ur 2012. V. Vezzoli.

West & East 285 Monografie, 4 Francesca Simi

at present used and adapted by all the four ALRG (such as storage sheds, field buildings and other projects. similar non-permanent occupation features); – the third model relates the presence of field scatters on the surface of modern fields to natural 4. The Gomel intensive survey results transport and post-depositional activities. In this interpretation, the pottery fragments were re- The intensive survey of Gomel helped us to propose moved from their original location, a site, by rain, a more accurate estimated area of the site and it will wind and human activity, especially ploughing, help us to define with more precision the limits of ur- and thus spread over the areas between sites;26 ban expansion in the different phases of its history. – ephemeral pastoral or nomadic occupation is an- The results presented here are still preliminary, but other proposed explanation for the phenomenon; from the new data it is possible to suggest that the site in fact, campsite remains may have been broken of Gomel covered a total area of ca. 35 ha, substantial- up and then spread in the fields by ploughing.27 ly more than the 16 ha previously estimated. – in the last case scholars argue that organic rubbish (including small fragments of pottery) was dumped in the field as fertilizer. The so-called “ma- 5. The off-site survey results nuring hypothesis” was sustained in Near Eastern archaeology by Tony Wilkinson, who was the first In Near Eastern Archaeology, sites have sometimes who attempted to verify this theory in our field of been considered as synonymous with mounds, with studies. He linked the explosion of urbanization these tells simply floating in the landscape sur- in third- millennium north Mesopotamia with rounded by an archaeological void. However, there the development of intensive agriculture; in this exists a continuum of artefacts and other evidence model the practice of manuring was related to the distributed on the ground surface as a sort of carpet need to improve crop yields. According to this in- of archaeological data, and one of my goals in the terpretation, the presence of sherds in the fields is field was to recognize and interpret this evidence. explained by the dumping of the village rubbish The very small and battered pottery fragments that (including small fragments of pottery) rich in or- are dispersed throughout the landscape between ganic matter in the field as fertilizer.28 sites are called “field scatters”. They were recognized initially in Western Europe and then studied in as- Although all these factors may have contributed sociation with sites of different chronological peri- to the creation of field scatters, only the last mod- ods, such as for instance on the Islamic Iranian coast el could really explain the density and uniform dis- 24 or in Classical Greece. tribution of these fragments in off-site areas and it Below, I summarise the different explanations seems to be the most accepted at present. given for the formation of field scatters: Coming back to the results of TGAS survey, – the first, defined by Bintliff and Snodgrass as “a as it is possible to observe from the distribution of feature of archaeological folklore”, concerns a the fragments registered while walking through the mythical donkey fom whose back pots have fall- transects, there is a continuity in the dispersion of en, leaving these pottery fragments in the land- these sherds, not only around sites, but also in areas scape. 25 In other word, as the results of a combi- not close to any known archaeological site. nation of chance events;

– in the second hypothesis, off-site scatters repre- 26 See Alcock et Al. 1994 for a good review of the evi- sent the traces of non-habitation activity areas dences. 27 UR 2010, 74. 24 For a short outline of the field scatters see Wilkinson 28 Wilkinson 1982, 1988, 1989, 1994, 2003; Ur 2010; 2003, pp. 55-57. for some criticisms see Alcock et Al. 1994 and the com- 25 Bintliff, Snodgrass 1988, 507. ments in Wilkinson 1994.

West & East 286 Monografie, 4 The Tell Gomel archaeological survey. Preliminary results of the 2015-2016 campaigns

From the results obtained from the data record- Jazira region, hence new data sets are necessary to ed in the sample units, it is possible to observe that gain a proper inter-regional perspective. the quantity of the fragments in the TGAS area, although consistent, is much lower than those provided by other surveys in neighbouring regions. In 6. Site detection: intensive vs extensive fact, the mean in the TGAS-area is of 6 sherds per 100 m2 compared to an average of over 30 sherds per Regarding site detection, TGAS surveyed a total of 100 m2 recorded by other North-Mesopotamian 112 archaeological sites (fig. 7). The two field cam- surveys, such as the Tell Hamoukar Survey (average paigns led to the identification of 63 new archaeo- of 38 sherds per 100 m2) and the North Jazira Pro- logical sites, which may be added to the previous 49 ject (average of 20-39 sherds per 100 m2).29 Only a archaeological sites already identified by the LoN- few areas present more than 30 fragments and gen- AP team in the 2012 and 2013 seasons. Thus far, erally many areas yielded only up to 10 fragments in the TGAS area the site density per square kilo- and there are even areas where the sample units have metre is 1.12 sites/km2, positioning the Tell Gomel not recovered any ceramics. Archaeological Survey in first place with regard to This quantitative difference between the num- site discovery densities among other Mesopotamian ber of sherds recorded in the Gomel area and in the surveys. 33 North Jazira could be possibly put in relation with Almost a quarter of the archaeological sites dis- the environmental conditions. In fact, a higher an- covered was found trough the transect fieldwalking. nual rainfall can lead to the burial of the potsherds If we consider the site morphology of the new sites beneath the fields surface. 30 On the other hand, a discovered by TGAS, it is interesting to note that lower density of fragments could correspond to a most of the sites detected through the intensifica- lower necessity of systematic manuring. tion of the survey methodology are low-mounded The dating of these potsherds is difficult due to sites. This is interesting because it shows how the in- their poor state of preservation: they are often very tensification of the survey strategy could record not small and abraded. We collected ca. 800 diagnos- only flat sites but also small mounded sites unno- tic fragments and when diagnostic types were rec- ticed during the previous surveys. ognized among them, they appear to date mainly The preliminary results from the TGAS demon- to the mid-late third millennium or, to a lesser ex- strate how the intensification of the methodology tent, to the first half of the second millennium BC. applied in an already investigated area can change A similar trend was recorded in the North Jazira the interpretation of settlement distribution and Project31 and in the Tell Hamoukar Archaeologi- dating (fig. 8). This is especially true for the pre- cal Survey 32 where the majority of the field scatters and proto-historic periods. In fact, as shown in the date to the second half of third millennium. chart, the LoNAP team recorded, for instance only To conclude, the collected field scatters data 4 Early Pottery Neolithic and 1 Halaf sites in the seem to suggest that even in the Gomel area, al- region later studied by TGAS project, whereas the beit on a smaller scale, manuring was practiced in TGAS team found 17 Early Pottery Neolithic sites the second part of the third millennium and at the and 15 Halaf sites. This change in the data collected beginning of the second. However, off-site surface is due not only to the discovery of new sites, but also pottery densities are still not regularly recorded in to the recording of phases previously not observed Near Eastern field surveys and until now it is diffi- in sites already investigated by LoNAP. Other pe- cult to recognize a consistent pattern outside the riods in which the intensification of the survey had 29 Wilkinson, Tucker 1995, pp. 20-22; Ur 2010, a huge impact are the second part of the third mil- pp. 68-69. lennium, the Parthian and the Islamic periods. In 30 Bintliff, Snodgrass 1988. those cases, we detected dozens of small flat or low 31 Wilkinson, Tucker 1995, pp. 22-23. 32 Ur 2010, p. 73. 33 Ur et Al. 2013, p. 112, fig. 16.

West & East 287 Monografie, 4 Francesca Simi

Figure 7 Distribution map of the archaeological sites discovered in the TGAS area

Figure 8 The chart illustrates the results, in terms of identified sites per chronological period, from the LoNAP and TGAS surveys, and the two combined

West & East 288 Monografie, 4 The Tell Gomel archaeological survey. Preliminary results of the 2015-2016 campaigns

mounded one-period sites previously undetected by occurred only in the second half of the third mil- the extensive survey. Concerning the other periods, lennium BC. The following period of thriving oc- in the majority of the cases the survey intensifica- cupation was the Middle Bronze Age. Afterwards, tion seems simply to have increased the number of the Mitanni and the Middle Assyrian occupation sites for each epoch, almost without modifying the patterns laid the foundation for the later Neo‑As- general trend documented by the extensive survey. syrian period that represents the first-maximum expansion of human settlement in the region. After the collapse of the Assyrian Empire there is clear ev- 7. The settlement development idence of settlement reduction, until the Parthian period, which corresponds to the third-highest set- Regarding the study of settlement development, the tlement density in the TGAS research area. Finally, survey results reveal a complex and multi-layered the long period defined as Islamic is attested on 83 landscape (fig. 9). settlements, thus representing the absolute peak in The preliminary data shows a limited occupa- the occupation of the plain, but in this huge period tion of the area during prehistoric and protohis- are squeezed together all the sites dating to the Ear- toric times. Actually, the first strong development ly, Middle and Late Islamic periods. Further work of settlement took place in the Late Chalcolithic on the pottery materials is needed in order to obtain with also a distinctive grow of the total settled area. a better chronological sequence for this period. But the emergence of a dense rural landscape, with The data collected shows how the Navkur Plain is small-sized villages scattered throughout the plain, culturally strictly related to the Jazirah cultural con-

Figure 9 Preliminary chart of the sites discovered in the TGAS area during the 2015 and 2016 campaigns according to their chronology

West & East 289 Monografie, 4 Francesca Simi

text, especially the Khabur Triangle and the North test how the use of two different methodological Iraqi Jazirah. This strong relation is not only related approaches on the same territory could lead to two to the affinity in the material culture but also to the substantially different results in the understanding similar socio-political dynamics. In the Navkur Plain and interpretation of settlement development and it is possible to recognize, even if with some differ- land-use. ences, a settlement development close to the pattern From the methodological point of view, the re- recorded by other Upper Mesopotamian surveys.34 sult of the combination of three different strategies The Navkur Plain can be included among the to investigate the site of Gomel, the other archaeo- “Agricultural Plains” of the Northern Fertile Cres- logical sites in the TGAS area and the off-site was cent, as defined by Wilkinson et Al. (2014, 50), very effective. Such intensive approaches allowed and it forms as well a classic “Landscape of Tells” to: 1) obtain a high-resolution reconstruction of (Wilkinson 2003: 100) characterized by a dry- Gomel settlement’s development and limits; 2) de- farming staple economy. The peculiar settlement tect sites which would not have been found other- trajectory of the Navkur Plain is related to its fa- wise; 3) record ancient field scatters in the off-site vourable environmental conditions (high average areas. annual rainfall, deep soils and abundance of water One of the main accomplishments of TGAS resources) but also its distance and peripheric po- project was the demonstration how an extensive sition in relation to well-known communication methodology, strongly based on remote sensing routes and marginal lands (the steppe). analysis, could skew our understanding of settlement development and landuse. These results have profound implications not only in the study 8. Conclusions of pre- and proto-historic periods but also in the reconstruction of long-term population dynam- The TGAS project was the first project, in the ics, nucleation and dispersal phenomena and land field of Near Eastern Archaeology, designed to exploitation.

Acknowledgements

I want to express my gratitude for their support to the General Directorate of Antiquities of the Kurdistan Re- gional Government, in particular its former director Kak Abubakir Othman (Mala Awat), the new director Mr. Keify Ali and to the Duhok Directorate of Antiquities, in the person of Dr. Hassan Ahmed Qasim and all his team. Furthermore, I wish specially to thank Prof. Daniele Morandi Bonacossi and all the members of the LoNAP team for their constant and invaluable support. I am also grateful to Costanza Coppi- ni and Paola Sconzo for their comments and for the fruitful discussions regarding the results presented here.

34 North Jazira Project: Wilkinson, Tucker 1995; Tell Hamoukar Survey: Ur 2010; Tell Beydar Survey: Ur, Wilkinson 2008; Tell Brak Suburban Survey: Ur et Al. 2011.

West & East 290 Monografie, 4 The Tell Gomel archaeological survey. Preliminary results of the 2015-2016 campaigns

BIBLIOGRAPHY

Alcock S. E., Cherry J. F., Davis J. L. 1994, Intensive Koliński R. 2018, An Archaeological Reconnaissance in survey, agricultural practice and the classical landsca- the Greater Zab Area of the Iraqi Kurdistan (UGZAR) pe of Greece, in: Morris I. (ed.), Classical Greece: an- 2012–2015, in: Horejs et Al. (eds.), Proceedings of cient histories and modern archaeologies, Cambridge, the 10th International Congress on the Archaeology of pp. 137-170. the Ancient Near East, Wiesbaden, pp. 13-27. Algaze G., Hammer E., Parker B. 2012, The Tigris- Kopanias K., MacGinnis J., Ur J. A. (eds.) 2015, Euphrates Archaeological Reconnaissance Project. Archaeological Projects in the Kurdistan Region in Final Report of the Cizre Dam and Cizre-Silopi Plain Iraq, Directorate of Antiquities of the Kurdistan Survey Areas, «Anatolica» 38, pp. 1-115. Regional Government, Erbil. Ball W., Tucker D., Wilkinson T. J. 1989, The Tell Kopania K., MacGinnis J. (eds.) 2016, The Archaeology al-Hawa Project: Archaeological Investigations in the of the Kurdistan Region of Iraq and Adjacent Regions, North Jazira, «Iraq» 51, pp. 1-66. Oxford. Bintliff, J., Snodgrass, A., 1988, Off-site pottery di- Lyonnet B. (ed.) 2000, Prospection archéologique du stributions: A regional and inter-regional perspective, Haut-Khabur occidental (Syrie du N.E.), Vol. 1 «Current Anthropology» 29, pp. 506-513. (Bibliothèque archéologique et historique 155), Buringh P. 1960, Soils and Soil Conditions in Iraq, Beirut. Baghdad. Menze B., Ur J. A. 2012, Mapping patterns of long-term Eidem J., Warburton D. 1996, In the Land of Nagar: settlement in Northern Mesopotamia at a large scale, A Survey around Tell Brak, «Iraq» 58, pp. 51–64. «Proceedings of the National Academy of Sciences» 109, pp. 1-10. Fales F. M., Del Fabbro R. 2014, Back to Sennacherib’s Aqueduct at Jerwan: A Reassessment of the Textual Morandi Bonacossi D. 2012-2013, Il paesaggio ar- Evidence, «Iraq» 76, pp. 65-98. cheologico nel centro dell’impero assiro. Insediamento e uso del territorio nella ‘Terra di Ninive’, «Atti Frankfort H. 1935, Oriental Institute Discoveries in dell’Istituto Veneto di Scienze, Lettere ed Arti. Iraq, 1933/34: Fourth Preliminary Report of the Iraq Classe di scienze morali, lettere ed arti» 171, Expedition, (Oriental Institute Communications pp. 181-223. 19), Chicago. Morandi Bonacossi D. 2016, The Land of Nineveh Hammer E., Ur J. A. 2019, Near Eastern Landscapes Regional Project. Assyrian Settlement in the Nineveh , «Advances in and Declassified U2 Aerial Imagery Hinterland: A View from the Centre, in: MacGinnis Archaeological Practice» 7/2, pp. 1-20. J., Wicke D. (eds.), The Provincial Archaeology of the Herrmann J. T., Glissmann B., Sconzo P., Assyrian Empire, Cambridge, pp. 141-150. Pfälzner P. 2018, Unmanned Aerial Vehicle Morandi Bonacossi D. 2018, Water for Nineveh. (UAV) Survey with Commercial-Grade Instruments: The Nineveh Irrigation System in the Regional Context Ḫ A Case Study from the Eastern abur Archaeological of the ‘Assyrian Triangle’: A First Geoarchaeological , «Journal of Field Archaeology» 43/4, Survey, Iraq Assessment, in: Kühne H. (ed.), Water for Assyria pp. 269-283. (Studia Chaburensia 7), Wiesbaden, pp. 77-115. Koliński R. 2016a, Insights into the settlement histo- Morandi Bonacossi D., Iamoni I. 2015, Landscape ry of Iraqi Kurdistan from the upper Greater Zab ar- and Settlement in the Eastern Upper Iraqi Tigris and , in: Kopanias chaeological reconnaissance project Navkur Plain: The Land of Nineveh Archaeological K., MacGinnis J. (eds.), The Archaeology of the Project, Seasons 2012–2013, «Iraq» 77, pp. 9-39. Kurdistan Region of Iraq and Adjacent Regions, Oxford, pp. 163-172. Pfälzner P., Sconzo P. 2015, First Results of the Eastern Ḫabur Archaeological Survey in the Dohuk Koliński R. 2016b, The use of satellite imagery in an ar- Region of Iraqi Kurdistan: The Season of 2013, cheological survey in Iraqi Kurdistan, «Contributions «Zeitschrift für Orient-Archäologie» 8, pp. 4-32. in New World Archaeology» 9, pp. 113‑122.

West & East 291 Monografie, 4 Francesca Simi

Pfälzner P., Sconzo P. 2016, From the banks of Upper Tell Brak Suburban Survey, 2003-2006, «Iraq» 73, Tigris River to the Zagros Highlands. The Tübingen pp. 1-19. Eastern Ḫabur Archaeological Survey. Trial sea- Ur J. A., Wilkinson T. J. 2008, Settlement and Economic son 2013, in: Kopanias K., MacGinnis J. (eds.), Landscapes of Tell Beydar and its Hinterland, in: The Archaeology of the Kurdistan Region of Iraq and Lebeau M., Suleiman A. (eds.), Beydar Studies 1, Adjacent Regions, Oxford, pp. 323-331. (Subartu 21), Turnhout, pp. 305-327. Pfälzner P., Sconzo P., Beutelscheis R., Wilkinson T.J. 1982, The definition of ancient manured Edmonds A., Glissmann B., Herdt S., zones by means of extensive sherd-sampling techniques, Herrmann J. T., Heydari-Guran S., Köhler «Journal of Field Archaeology» 9, pp. 323-333. J., Müller-Wiener M., Puljiz I., Sharp M. Wilkinson T. J. 1988, 2016, The Eastern Ḫabur Archaeological Survey in The archaeological component Iraqi Kurdistan. A Preliminary Report on the 2014 of agricultural soils in the Middle East: the effects of , in: Groenman W., van Season, «Zeitschrift für Orient-Archäologie» 9, manuring in antiquity pp. 10-69. Waateringe, M. Robinson (eds.), Man-made soils: symposia of the association for environmen- Reade J. E., Anderson J.R. 2013, Gunduk, Khanes, tal archaeology 6, (BAR Int. Series 410), Oxford, Gaugamela, Gali Zardak: notes on Navkur and near- pp. 93-114. by rock-cut sculptures in Kurdistan, «Zeitschrift für Assyriologie» 103/1, pp. 68-122. Wilkinson T. J. 1989, Extensive sherd scatters and land use intensity: some recent results, «Journal of Field Salman I. 1970, Archaeological Sites in Iraq, Baghdad. Archaeology» 16, pp. 31-46. Salman I. 1976, Atlas of the Archaeological Sites in Iraq, Wilkinson T.J. 1994, The Structure and Dynamics Baghdad. of Dry-Farming States in Upper Mesopotamia [and Savioli A. forthcoming, Archaeological Sites or “False Comments and Reply], «Current Anthropology» Positives”? Tracking Anthrosols from CORONA 35/5, pp. 483‑520. Satellite Imagery in the Nineveh Hinterland. Wilkinson T. J. 2000, Archaeological Survey of the Tell Stein A., Kennedy D. L., Gregory S. 1985, Sir Aurel Beydar Region, Syria, 1997: A Preliminary Report, in: Stein’s Limes report: The full text of M.A. Stein’s un- Van Lerberghe K., Voet G. (eds.), Tell Beydar: published Limes report [his aerial and ground recon- Environmental and Technical Studies, Subartu 6, naissances in Iraq and Transjordan in 1938-39] edited Turnhout, pp. 1-37. and with a commentary and bibliography, (BAR Int. Wilkinson T. J. 2003, Archaeological Landscapes of the Series 272), Oxford. Near East, Tucson. Ur J. A. 2002, Surface Collection and Offsite Studies at Wilkinson T. J. 2004, On the margin of the Euphrates: Tell Hamoukar, 1999, «Iraq» 64, pp. 15-44. settlement and land use at Tell Es-Sweyhat and in the Ur J. A. 2003, CORONA Satellite Photography and upper Lake Assad area, Syria, Chicago. Ancient Road Networks: A Northern Mesopotamian Wilkinson T. J., Tucker D. J. 1995, Settlement Case Study, «Antiquity» 77, pp. 102–15. Development in the North Jazira, Iraq, Warminster. Ur J. A. 2010, Urbanism and Cultural Landscape in Wilkinson T. J., Graham P., Bradbury J., Dunford Northern Syria. The Tell Hamoukar Survey Project, R., Donoghue D., Galiatsatos N., Lawrence Chicago. D., Ricci A., Smith S. L. 2014, Contextualizing Ur J. A. 2017, Archaeological Renaissance in the Kurdistan Early Urbanization: Settlement Cores, Early States Region of Iraq, «Near Eastern Archaeology» 80/3, and Agro-pastoral Strategies in the Fertile Crescent pp. 176-187. During the Fourth and Third Millennia BC, «Journal Ur J. A., de Jong L., Giraud J., Osborne J. F., of World Prehistory» 27/1, pp. 43-109. MacGinnis J. 2013, Ancient Cities and Landscapes Wright H. T., Rupley E. S. A., Ur J. A., Oates in the Kurdistan Region of Iraq: The Erbil Plain J., Ganem E. 2006‑7, Preliminary Report on the Archaeological Survey 2012 Season, «Iraq» 75, 2002 and 2003 Seasons of the Tell Brak Sustaining pp. 89-118. Area Survey, «Les Annales Archéologiques Arabes Ur J. A., Karsgaard P., Oates J. 2011, The Spatial Syriennes» 49-50, pp. 7-21. Dimensions of Early Mesopotamian Urbanism: The

West & East 292 Monografie, 4 The human burials from Şinduxa (Iraqi Kurdistan). A bioarchaeological and archaeothanatological approach

ALESSANDRO CANCI*, HASAN AHMED QASIM**

* University of Udine, ** Directorate of Antiquities of Duhok

Abstract

In oriental archaeology the study of human skeletal remains sadly still today receives poor attention causing de- pletion to the archaeological and historical interpretations. Here we present the results of a study based on an integrated approach between archaeothanatology (among the topics taphonomy of human bone remains, spatial relationships between other elements of the tomb as the architectural structure or grave goods) and bioarchaeology (sex, age at death, ancestry, stature, palaeopathology) of the human bone remains coming from the excavations of Şinduxa necropolis nowadays embedded in the city center of Duhok, Iraqi Kurdistan. Briefly, four tombs were explored: one, placed in plain earth covered with stone slabs belonging to an adult woman and the other three consisted in stone funerary chambers above the ground were bone remains of 9 adult subjects of both sexes were lying. Interestingly in each of these structures a well preserved simple oval clay sarcophagus containing one or two skeletons was present. Regarding the tomb’s chronology C14 calibrated dating indicates a time frame from Neo-Assyrian to the Parthian epoch.

Keywords

Archaeothanatology, Bioarchaeology, Palaeopathology, Parthian; Burials

West & East 293 Monografie, 4 Alessandro Canci, Hasan Ahmed Qasim

1. Introduction Immediately the Direction of Antiquities of Duhok stopped the work and started a rescue excavation The site ofŞ induxa is located on the left bank of that continued for three more years (2010-2013 and the Rubar Duhok, on a terrace rising of about 15 m 2015), thanks to the financial support of the Kurdis- above the river level, at 518 m asl (fig. 1). Although tan Regional Government under the direction of Dr. today it is completely flattened, originally it had to Hasan Qasim, Director of Antiquities Department be a tell-formation. Nowadays it is embedded in the in Duhok. The rescue work brought to light no real city center of Duhok. architecture, rather a large array of new monumental The actual name is Şinduxa, the very same top�- statues and other objects of exceptional value dating onym - ALU Su-du-hi - was already mentioned in the from the Late Uruk period down to Hellenistic and ancient cuneiform texts from Mari as a city of the Parthian time. One of the most interesting discover- Kingdom of Mitanni. 1 ies was the presence of a large cemetery hosting over The site was discovered accidentally in 2010 100 graves of different type (fig. 2), that had been in during the construction work of the new Duhok cause for a long period of time. bleway. First to be noticed where some fragments of The cemetery was fully cleared and mapped a monumental stone statue of the Mitannian period. (fig. 3), but just four graves were fully excavated in

Figure 1 Localization from aerial photo of the site of Şinduxa, Duhok

1 Luckenbill 1912, 180; Heimpel 2003, 418.

West & East 294 Monografie, 4 The human burials from Şinduxa (Iraqi Kurdistan). A bioarchaeological and archaeothanatological approach

2015. These graves were located in the southern part belong to the Parthian epoch and testify a very com- of the site (square F3 in the map at fig. 3) and are plex burial practice probably connected to the Zo- object of the present contribution. If one of them roastrian religion widespread among the local pop- dates to the Neo-Assyrian Period, the other three ulation of the region.

Figure 2 A view of the necropolis. The semi-circular stone funerary chambers and simple burials covered by stone slabs are visible

Figure 3 Map of the archaeological site of Şinduxa. The square F3 indicates the localization of the necropolis

West & East 295 Monografie, 4 Alessandro Canci, Hasan Ahmed Qasim

2. Material and methods inside the tomb.4 More in detail this research had two main goals: the first one is devoted to recon- The integrated study of human skeletal remains struct the biological profile in terms of sex, age at at the aim to reconstruct the biological profile of death, stature and health conditions of the human the deceased sadly have received poor attention sample and the second one to furnish a reliable rein the past until the publication in the ’70 of the construction of the original manner of deposition Lawrence Angel’s research about human skele- of the deceased. tal remains from Mycene and Lerna.2 Until now On the whole this partial research allowed many others significant papers devoted to this top- us to anticipate a more extensive study suggest- ic have enriched the archaeology with stimulating ing the great interest of the necropolis of Şinduxa sources of information. In our opinion neverthe- concerning the funerary archaeology. In this study less, a milestone about the corpus of information four tombs, one single and three collectives, were devoted to support and help archaeological inter- excavated and studied and this report summarizes pretations and reconstructions results from the the preliminary results. application of the concept of Archaeotanatology The sex determination was done on the morpho- developed by Henry Duday and coauthors at the logical features of pelvic girdle, skull and general ro- ’80s of the last century (Duday et Al. 1990). Ar- bustness of the bone remains according to the indi- chaeotanatology consists in a synthesis about the cations reported in Buikstra and Ubelaker (1994) relationships existing between the deceased’s bone and about the estimation of age at death we consid- remains and their arrangement in the grave paying er the variations of the morphology of the auricular attention to the objects accompanying the dead surface of the ilium and the corresponding surface and their placement in respect to the human body; of the sacral bone. Furthermore, where possible, in this way the contribution of important infor- the morphological appearance of the fourth sternal mation that archeothanatology can furnish to the rib’s end according to Isçan et Al. (1984) and paleo- archaeology is obvious. pathological diagnosis was made according the diag- In this paper, we present as an example of the nostic criteria suggested by Ortner (2003). At last bioarchaeological and archaeothanathological ap- archaeothanatology interpretation of the burials proach, the results coming from the excavations of was made according to Duday (2002). four tombs found in an ancient cemetery localized in Duhok (Kurdistan Region). From sixteenth October to 16 November 2015 a project in order 3. Grave 1 to obtain bone samples to submit for radiocarbon dating has allowed to explore the tombs from the Grave 1 was a stone chamber burial where a skel- necropolis of Şinduxa. Furthermore, to obtain the eton was contained in a bath tube clay sarcophagus. bone samples a survey about the tombs selected The human bone remains belong to an anatomical- a preliminary bioarcheological research with the ly complete and well preserved skeleton of a young aim to describe the human sample sheltered in the adult woman, as suggested by a very marked preau- tombs was carried out. This study is the result of ricular sulcus on the pelvis, aged between 20 to 25 an integration between the more reliable and dif- years old with a stature estimated from 153 cm to fused methods to investigate human skeletal re- 156 cm (fig. 4 and 5). The skeletonat the discov- mains (physical anthropology and palaeopatholo- ery was lying in a supine position with the upper gy)3 and archeothanatology a new discipline based arms crossed and both hands placed on the chest; to investigate the relationships between the hu- the lower limbs appeared parallel and extended. man bones arrangement, goods or other structures The decomposition occurred in empty space but the marked straight appearance of the skeleton and 2 Angel 1971. 3 Byers 2002. 4 Duday 2004.

West & East 296 Monografie, 4 The human burials from Şinduxa (Iraqi Kurdistan). A bioarchaeological and archaeothanatological approach

Figure 4 Tomb 1. Oval clay sarcophagus containing the skeleton of a young adult woman

Figure 5 Plan of the tomb 1

West & East 297 Monografie, 4 Alessandro Canci, Hasan Ahmed Qasim

the verticalization of the clavicle seems to indicate close to the fingers of the right hand thus suggesting a wall-effect of compression thus suggesting that that the jewel was probably worn by the deceased the body was probably wrapped in a tissue, maybe (fig. 7). The other object, a needle in bronze meas- a shroud (fig. 6). During the excavation of the tomb uring about 5 cm, was discovered close to the right two small findings consisting in an iron ring and foot (fig. 8). The position and the nature of the ob- a very thin bronze needle came to light. The ring ject could indicate that the needle was used to sew was found at a higher level respect to the skeleton up the tissue wrapping the body.

Figure 6 – Tomb 1. The marked rectilinear disposition of the left upper and lower limbs of the skeleton suggests a wall effect probably resulting in compression due to a shroud wrapping the body

Figure 7 – Tomb 1. Figure 8 – Tomb 1. The ring accompanying the deceased that was lying The bronze needle accompanying the deceased that was lying inside of the sarcophagus inside of the sarcophagus

West & East 298 Monografie, 4 The human burials from Şinduxa (Iraqi Kurdistan). A bioarchaeological and archaeothanatological approach

Bone remains indicate as absolute dating an Very small fragments of long bones belonging to age of 1823±45BP with calibrated age (95,4% an unspecified number of individuals appeared close probability) ranging between 80 to 330 AD to the sarcophagus thus suggesting that it was a col- (fig. 9). lective tomb.

Figure 9 Absolute and calibrated age of the skeleton of the tomb 1

4. Grave 2

The chamber built in stone where the dead were placed, contained several unconnected human bones belonging to a minimum number of three adult subjects leaning on the stone wall (fig. 10, 11 and 12); a skeleton in poor anatomical connection, maybe an adult female, was lying on the soil on the lid of the sarcophagus used as the place of deposition (fig. 12); at last, an oval clay sarcophagus containing a complete skeleton of an adult male, about 25-35 years old, with a stature estimated in about 175 cm, was observable. This is an example of collective tomb where the differences of anatomical connection and better condition of preservation of the Figure 10 – Tomb 2. Collective burial where an oval clay sarcophagus containing a skeleton of an adult male deceased suggests that the sarcophagus was the en- (A) and fragmentary bone remains of three adult subjects vironment where the decay of the body took place. (B) placed outside of the sarcophagus are visible

West & East 299 Monografie, 4 Alessandro Canci, Hasan Ahmed Qasim

At the palaeopathological examination the stress maybe due to usual overload of heavy weight skeleton of the man contained in the sarcophagus on the back; a marked heavy dental wear on the showed signs of degeneration at the level of the mandibular teeth is observable as well (fig. 13). thoracic lumbar tract of the vertebral column con- Bone remains indicates as absolute dating an age sisting in the flattening of the vertebral bodies and of 1769±40 BP with calibrated age (95,4% proba- osteophytosis thus suggesting an incipient skeletal bility) ranging between 130 to 380 AD (fig. 14).

Figure 11 Plan of the tomb 2

Figure 12 – Tomb 2. Distribution of human fragmentary bone remains placed outside of the sarcophagus

West & East 300 Monografie, 4 The human burials from Şinduxa (Iraqi Kurdistan). A bioarchaeological and archaeothanatological approach

Figure 13 – Tomb 2. Palaeopathological evidences of the skeleton lying in the sarcophagus

Figure 14 Absolute and calibrated age of the skeleton lying in the sarcophagus

West & East 301 Monografie, 4 Alessandro Canci, Hasan Ahmed Qasim

5. Grave 3 the subject was seriously affected by a huge degree of dental wear probably because of a coriaceous diet The burial consisted in a simple ditch in the ground rich in contaminants likely consisting in remnants where a fragmentary and anatomically incomplete of the grindstone. At last, as a grave good a fragmen- skeleton was conserved (fig. 15 and 16). The skele- tary clay bowl placed at level of the skeletal face of ton belonged to a woman that was lying supine with the sublet was found. Bone remains indicate as ab- the upper limbs crossed on the chest and the low- solute dating an age of 2832±45 BP with calibrated er limbs extended. The age at death of the subject age (93,7% probability) ranging between 1130 BC was old, probably over 50 years old and the stature – 890 BC and with calibrated age (1,7% probabil- was estimated in about 154 cm. The dentition of ity) ranging between 880 to 850 BC (fig. 17).

Figure 15 – Tomb 3. Simple ditch on the ground where a fragmentary skeleton of a mature adult woman was lying. Fragments of a clay Figure 16 bowl (A) were placed at level of the head of the deceased Plan of the tomb 3

West & East 302 Monografie, 4 The human burials from Şinduxa (Iraqi Kurdistan). A bioarchaeological and archaeothanatological approach

Figure 17 Absolute and calibrated age of the skeleton of the tomb 3

6. Grave 4 The human sample occupies the funerary room in two spaces. The first environment was a clay sar- It is a collective burial were a whole of 5 subjects cophagus were the skeleton of an adult female be- were buried; the tomb was intercepted by an Islam- tween 25-35 years old of age was lying supine with ic tomb where a skeleton of an adult male lying on both legs extended and upper arms likely crossed on the right flank (fig. 18 and 19). the chest (subject B). During the excavation in pro-

Figure 18 – Tomb 4. Collective tomb consisting in a sarcophagus (A) containing a skeleton of an adult woman and fragmentary bone remains of lower limbs belonging to a probable adult male. Fragmentary bone remains without anatomical connection of three adult subjects (B) placed outside of the sarcophagus are observable. A tomb dated to the Islamic period intercepting the tomb 4 (C) is visible as well

West & East 303 Monografie, 4 Alessandro Canci, Hasan Ahmed Qasim

Figure 19 Plan of the tomb 4

Figure 20 – Tomb 4. Details of arrangement of the bone remains contained in the sarcophagus. The subject B was a young adult woman placed above the skeleton of the subject A a mature adult man (short arrows); both femurs of this skeleton where placed at the left side of the thoracic cage of the woman. These bones were probably wrapped in a bag or contained in a box of perishable material (long arrow)

West & East 304 Monografie, 4 The human burials from Şinduxa (Iraqi Kurdistan). A bioarchaeological and archaeothanatological approach

Figure 21 – Tomb 4. Details of the skeleton of the subject A was lying in the sarcophagus. The fall of the skull on the chest and the disjunction of the anatomical connection between cranium, mandible and vertebral column suggests the probable presence of a pillow in perishable material where the head was lied at the time of inhumation

Figure 22 – Tomb 4. The bronze needle accompanying the skeleton of the woman contained in the sarcophagus

gressive manner only the leg bones (tibia and fibula ical connection between the cranium from the man- of both sides) of a robust adult male appeared (sub- dible and the vertebral column (fig. 21). The loss of ject A). Besides a pair of femurs probably belonging anatomical connection could be the consequence of to the man were placed on top of the other on the the fall of the skull that was lying on a pillow in per- chest of the female skeleton. Probably this unusual ishable material. Besides, at the level of the left shoul- deposition is related with the aim to create the space der a thin hooked needle in bronze measuring about needed to shelter the female body in the sarcophagus 5,30 cm was found. As previously reported in the case (fig. 20). Observing the skeleton of the female is pos- of tomb 1 the needle could have been used to sew up sible appreciate the large disjunction of the anatom- a shroud in perishable tissue (fig. 22).

West & East 305 Monografie, 4 Alessandro Canci, Hasan Ahmed Qasim

The second place of deposition for the human As concern the palaeopathology, in grave 4 as bones is the space outside of the sarcophagus where well the oral health was very poor because of the skulls and unconnected postcranial bones corre- very heavy wear affecting the occlusal surfaces of the sponding to three individuals (probably two adult teeth of the female from the sarcophagus but above males and one subject of undetermined sex) were all in the case of a cranium found outside the sar- found. In this case as well, it is important to remem- cophagus were a complete ante mortem tooth loss ber that the tidy disposition of the bones without probably due to a serious masticatory stress affecting anatomical connection suggests a careful manipu- the mandible of a mature adult male was observable. lation with the aim to manage the space inside the Bone remains indicates as absolute dating age of tomb in order to contain in the best way the bones 1766±45 BP with calibrated age (95,4% probabili- of the dead (fig. 23). ty) ranging between 130 to 390 AD (fig. 24).

Figure 23 – Tomb 4. The tidy disposition of the bone remains outside of the sarcophagus suggest a packaging in a bag or in a box in perishable material or careful manipulation of the bones in order to manage in the best way the space in the tomb

Figure 24 Absolute and calibrated age of the woman’s skeleton from the sarcophagus of the tomb 4

West & East 306 Monografie, 4 The human burials from Şinduxa (Iraqi Kurdistan). A bioarchaeological and archaeothanatological approach

7. Concluding remarks The whole of the Parthian human skeletal remains previously described were sheltered in collec- In conclusion, as reported in table 1 in spite of the tive tombs consisting in a stone chamber of semicir- small number of burials explored the results indi- cular shape reopened over the course of time when a cate that dating allows us to affirm that the necropo- new deposition of a deceased member of the family lis cover a lifespan of about 700 years from the Neo- was necessary. At the interior of each room an oval Assyrian period to the Parthian epoch. clay sarcophagus, about 185-200 cm length, 40-45 In Şinduxa the most ancient tomb 3 dated to the cm wide and 25-30 cm depth containing the bones of Neo-Assyrian period and consists in a simple ditch one or more subjects, was present. The clay sarcoph- in the ground but it is in the Parthian-Ellenistic pe- agus is the place of decomposition of the body subse- riod that a real funerary building appears. A lot of quent to the deposition in the funerary chamber. matters concerning the funerary archaeology of Par- We suggest as reported by Olbrycht (2017) that thian epoch are still largely little known not only in this funerary practice fits well with Zoroastrianism’s Kurdistan region of Iraq, object of this paper, but in precets that impose to avoid any contamination of ancient Mesopotamia and Iran too. The recent re- the main elements (Earth, Fire and Water) because view about this topic written by Farjamirad (2015) after the death the body becomes impure. The sar- highlights the large number of different burial prac- cophagus is an environment where the complete de- tices among Parthian communities consisting in cay of the body after death takes places sparing the pithoi, shaft burials, sarcophagus of several mor- earth by an impure pollution due to putrefaction of phologies (oval, anthropoid, bathtub coffins) built the human tissues. At the end of this phase the bones in different materials (from simple clay to bronze), were assembled and placed outside of the sarcopha- monumental and rock cut chambers. gus in order to make available the environment of

Table 1

Tomb Structure Sarcophagus Sex Age at death Burial goods Cal. dates

1 Stone yes Female lying adult Iron ring 80-330 AD chamber in sarcophagus and bronze needle in the sarcophagus

2 Stone yes 1 male lyng in adult no 130-380 AD chamber sarcophagus, bone remains of 1 female (?), fragmentary and disconnected bones of two unsexed subjects outside of the sarcophagus

3 Simple ditch no female adult Clay bowl 880-850 BC

4 Stone yes 1 male and 1 female adult Bronze 130-390 AD chamber lying in the sarcophagus; needle in the disconnected bone sarcophagus remains of 2 males and 1 unsexed subject outside

West & East 307 Monografie, 4 Alessandro Canci, Hasan Ahmed Qasim

decay to a new deceased. Interestingly, we remark to adults of both sexes; no children bone remains that in the case of tomb 2 a skeleton outside of the were found. sarcophagus occupied by another deceased was lying At last, at the palaeopathological recognition on the sarcophagus lid placed on the ground prob- no signs of serious pathological lesions due to infec- ably to avoid in this case as well the contamination tions, traumas or bone evidences of poor nutrition of the tomb’s soil. At least in two cases (Tomb 1 and were observed but only few cases of osteoarthritis at Tomb 4) a bronze age needle probably used to sew vertebral column due to degenerative diseases and up the shroud wrapping the body was found. The ante mortem tooth loss were present thus suggesting human remains from the chamber burials belonged a fairly good health status.

Acknowledgments We thank Dr. Giuseppe Minunno for drawings of the tombs.

West & East 308 Monografie, 4 The human burials from Şinduxa (Iraqi Kurdistan). A bioarchaeological and archaeothanatological approach

BIBLIOGRAPHY

Angel J.A. 1971, The people of Lerna, Princeton. Heimpel, W. 2003. Letters of the King of Mari. A New Boucharlat R., Haerinck E. 2011, Tombs d’epoque Translation, with Historical Introduction, Notes, and Parthe, Leiden. Commentary, Philadelphia. Buikstra J.E.B., Ubelaker D.H. 1994, Standards for Isçan M.Y., Loth S.R., Wright R.K. 1984, data collection for human skeletal remains, Fayettville. Age estimation from the ribs by phase analysis: white males, «Journal of Forensic Sciences 29», Byers S.N. 2002, Forensic Anthropology, Boston. pp. 1094-1104. De Mecquenem R. 1931, Excavations at Susa (Persia), Luckenbill, D. D. 1912. Inscriptions of early Assyrian , «Antiquity» 5, pp. 330-343. 1930-1931 rulers, «The American Journal of Semitic Languages Duday H., Courtad P., Crubezy E., Sellier P., and Literatures» 28/3, pp. 153-203. Tillier A.M. 1990, L’anthropologie “de terrain”: Olbrycht M.J. 2017, Slipper coffins and funerary prac- , reconnaissance et interprétation des gestes funéraires tice in Parthia, «Anabasis» 8, pp. 301-313. «Bulletin et Mémoires de la Société d’Anthropolo- gie de Paris» 2, pp. 26-49. Ortner D. J. 2003, Pathological Conditions in Human Skeletal Remains, New York. Duday H. 2002, Lezioni di Archeotanatologia, Roma. Farjamirad M. 2015, Mortuary practice in ancient Iran from the Achemenid to the Sasanian period, (BAR International Series 2747), London.

West & East 309 Monografie, 4

Marine connections: the Gulf and the interactions between the Arabian Peninsula, Mesopotamia, the Iranian world and beyond

Adaptation and transformation of human setting from Middle Holocene to Early Bronze Age in south-eastern Arabian Peninsula

MAURIZIO CATTANI

University of Bologna

Abstract

This paper is a tribute to the forerunners scholars that devoted their research to the prehistory of Arabia. In particular, Serge Cleuziou and Maurizio Tosi, who staked on the investigation of the most extreme point of eastern Arabia and enlightened the way to carry out the research to many others (Tosi 1989; Cleuziou, Tosi 2007). Their main effort was made to introduce this region as a wide history of land, people and culture with several concepts and directions of the research of an outermost place among ancient civilizations of Near and Middle East. The aim of this paper is to illustrate the framework of the evolutionary path from the sixth to second millennium BC keeping on mind some keywords: adaptation to local (arid) environment; resource exploitation; social and economic options; interactions; trade and exchange; sharing technology.

Keywords

Oman, Middle Holocene, Bronze Age, settlement, graveyard

West & East 313 Monografie, 4 Maurizio Cattani

1. Introduction 2. The early phase

The geographical and geomorphological conditions The formative process during the Middle Holocene, of South-Eastern Arabia can be resumed in three between 6000 and 3000 BC can be illustrated by main elements: a central mountain chain raising up some of the excavated sites by the Italian Archae- to 3000 m; the desert inland extending to the west in ological Mission in Oman in almost forty years of the Rub’ al-Khali; the coast on the Gulf and the Indi- intensive explorations,4 like RH-5 and RH-6 at Ras an Ocean on the eastern side (fig. 1). The rocky ridge, al-Hamra – Qurum in the Capital Area of Muscat generally marking a boundary, is here a sort of cohe- and KHB-1 at Ra’s al-Khabba on Ja’alan coast. All sive force to link and unify different environmental of them located on the coast are better preserved features thanks to high availability of water. The re- than anywhere in the interior thanks to the dumps gion is included in the subtropical zone of arid cli- of collected shells that consolidated the archaeolog- mates with an average of rainfall well below the 250 ical deposit and preserved the prehistoric remains in mm conventionally marked as the limit of aridity1, much better conditions. They also resume the pos- but all around the mountains a widespread string of sibility to understand the complexities of a small aquifers and springs always represented the attraction community group that struggled to survive exploit- of an increasing population, especially during the pe- ing new resources. riod between the sixth to fourth millennium BC.2 The sites of RH-5 and RH-65 are two middens Another form of attraction was at the same made of shells and fishes located on the junction of time the seaside, as source of wealth with end- a tertiary calcareous terrace (Ra’s al-Hamra) with less resources (molluscs and fishes), well testified the sand beaches of Batinah, where the estuary of by several shell middens investigated in the last wadi Aday creates a wide mangrove swamp (Qurum decades. These corresponds probably to seasonal Nature Reserve). The sites allow to go back in Mid- campsites for small groups of food foragers mainly dle Holocene dating respectively between the end nourished by fishing, gathering and hunting with a of the fifth and the fourth millennium BC (RH‑5) few goats and cows aside in the Middle Holocene. and between the mid-sixth millennium BC and The similarities in funerary rituals and in mate- mid-fifth millennium BC (RH-6). The particular rial culture made of flint and stone tools or shell location allowed to exploit mangrove biomass, col- hooks extending from the Gulf in north-west to lecting shells like oyster (Saccostrea cucullata) and the seashore of Ja’alan show a continuous texture Terebralia palustris, as well as to for gathering wood. of identical form of adaptation and cultural mean- From the coast they exploited marine resources fish- ings that allow to propose a common path and a ing and collecting other sea shells. shared nature: presumably their force was in the The importance of these two sites, investigated mechanism of coordination, in which all peoples since early ’80s of last century with other campaigns have contributed and benefited from interactions in recent times, is marked by information on the with their neighbours. Archaeological record from dwellings of the settlement and especially on one Oman is strongly confirming the shaping of “so- of the most extended and ancient graveyard of the cial groups inextricably involved with other groups Arabian Peninsula.6 in weblike interconnections in which technologies In both sites the archaeological deposit up to 2 are diffused and modified by other groups caught m thick made of overlapping layers of shell and fish up in these same processes”.3 remains, contains dozens of huts and several structures (fireplaces, dump and waste pits, oven) used 1 Annual rainfall remains very low, ranging from 171 for cooking and processing fish and shells) with mm at Nizwa to 36 mm at Sohar. 2 Climate in the considered period varies from wetter 4 conditions until Fifth millennium BC to slightly more arid Frenez, Cattani 2019. during the Fourth millennium BC. 5 Biagi, Salvatori 1986; Biagi 1987. 3 Kohl 2008, p. 495. 6 Coppa et Al. 1985; Salvatori 2008.

West & East 314 Monografie, 4 Adaptation and transformation of human setting from Middle Holocene to Early Bronze Age in south-eastern Arabian Peninsula

Figure 1 Ecological zones and the wild animal populations of Oman (from Cleuziou, Tosi 2007: fig. 1.10)

Figure 2 Ra’s al-Hamra RH-5. View of the site with settlement and graveyard extensions (L.G. Marcucci)

West & East 315 Monografie, 4 Maurizio Cattani

Figure 3 RH-5 graveyard. Reconstruction of grave 411, a man has covered with layers of turtle bones, fish bones and shells (from Munoz 2008)

abundant objects of daily life. The material culture ogies, identify eating habits and rituals of death for describes the typical coastal prehistoric site with ancestors’ cult, and finally tentatively to reconstruct working of stone, shells and bones for manufactur- the society of these prehistoric fishermen.7 ing ornaments, as well as for items related to fish- The burials are mostly individual, placed in shal- ing and to hunting activities. From RH-6 a flint ar- low and oval pits lying on one side in a crouched po- rowhead and some stone adzes, small globular net sition with the arms bent and the hands placed in sinkers with a continuous incision or with pecked front of the head (fig. 3). In one part of the cemetery grooves around the waistline and the flat ones with multiple and secondary burials were found as well, notches at the end of the major axis, mark the early suggesting potential changes in funerary customs. fishing activities cutting stone tools, grooved stones, Among personal items found in the burials there grooved chlorite discs, shell scrapers (Veneridae), were personal ornaments such as necklaces, brace- multi-functional tools, an earring and rectangular lets, earrings, pins made from local stones and shells, beads in chlorite. and much rarely objects of common use such as shell The huts, ovals or circulars, were recognized by fishhooks, bone awls and net weights. In some cas- post-hole alignments or narrow trench foundation es, the burials contained a shell valve (Callista sp.), with few postholes inside. while two graves were covered with a whole cara- In both sites the excavations confirmed and en- pace of sea turtle (Chelonia mydas). larged the size of a graveyard which was in use be- Current analysis on the anthropological remains tween 3800 and 3300 BC at RH-5 and in the first are demonstrating a character of genetic isolation of quarter of the fourth millennium BC at RH-6. the local community marked especially by the ex- Thanks to almost five hundred graves of their -in tremely high incidence of spina bifida occulta, a mal- habitants the cemetery represents the most extraor- formation of one or more vertebrae that resulted in dinary archive of human records, where it is possible to analyse DNA and markers of diseases and pathol- 7 Coppa et Al. 1990.

West & East 316 Monografie, 4 Adaptation and transformation of human setting from Middle Holocene to Early Bronze Age in south-eastern Arabian Peninsula

painful scoliosis and deformity of feet. Life expec- Evaluating the economic activities based on the tancy scarcely reached 20 years, since almost 25% of exploitation of coastal resources (sea and lagoon) the individuals died between 20 and 25 years, while and on low level of hunting and pastoralism, it has less than 15% of the population reached 40 years. been suggested to recognize a nomadic nature of the The discovery of a shark tooth arrowhead still em- community, moving from the coast to the moun- bedded into the sacral vertebra of an adult male bur- tains, exploiting the wadis and mountains resources ied in the large collective grave clearly indicates the seasonally from spring to autumn and spending the evidence of a warfare event. winter between the coast and the lagoon.9 Among several sites discovered along the coast The exploitation of the sea has greatly influenced south of Ra’s al-Hamra, the site of Ra’s al-Khabba the maritime attitudes of the Neolithic communi- could be considered illustrative: KHB-1 after the ties: the successful adaptation of fishermen pushed early discovery by Paolo Biagi in 1987 was investi- to increase the confidence with the Ocean, pre- gated with several campaigns from 1999 to 2005 sumably finding the early boat for navigation and and published in several papers8 is one of the few opening to the great achievements of exploration fourth millennium BC specialized fishermen camp- confirmed by higher knowledge of astronomical sites along the eastern coast of Oman. The loca- characters as well as of marine currents and winds. tion on a narrow ridge of the coastal cliff next to a wide lagoon explains the attractiveness of resources obtained from the sea or the special environment 3. The fourth millennium BC where marsh water is frequently fed by fresh water of inner wadis. From this outline we can summarize our present The excavations recognized six phases of settle- knowledge about the settlement pattern of fourth ment occupation alternated with abandonment millennium BC: among the foraging societies of episodes marked by layers of inorganic loose sand. Oman, the role of domesticated plants and ani- The earliest phase belongs to the seventh millenni- mals remained marginal, even if we presume high- um BC and it is characterized by a moister climate er in the interior. Items of material culture show condition. Some post-holes related to huts, few fire- a remarkable level of integration connected to a places and a large pit filled by mussels. Other occu- unique social system, relied on social cohesion of pation phases belong to fourth millennium BC and nuclear families with its members sharing the same are characterized by several features and structures: living quarters. circular huts, shelter as wind screens, fireplaces. Few In few millennia, between the Early and the burials complete the archaeological evidence not Mid-Holocene, scattered groups of desert hunters too different from previous sites. Most of them are turning into food producers were able to extract located inside older abandoned structures, like a sec- an accumulative wealth of subsistence resources by ondary deposition (Grave 1) deposited in a shallow developing an interwoven texture of tribal allianc- oval pit belonging to phase V and a primary buri- es.10 Wealth accumulation, population density and al (Grave 2), unearthed in interphase II/III, located political complexity were developed along an orig- perfectly in the centre of a circular hut belonging to inal and unique evolutionary pathway based on a previous phase (phase II). kinship relations of equality that reversed the par- The material culture is mainly represented by adigm of hierarchy and specialization privileged by fishing tools with an advanced level of specialisation the settled farmers across the whole Middle East. in fishing techniques: stone net sinkers, shell fish- Compared to great civilizations of Near and Mid- hooks, shell lures, together with other tools used dle East, in South-Eastern Arabia the economical in the production of hooks (stone files) and in the foundations were built upon strategies of intensi- preparation of nets (bone needles). 9 Cavulli, Scaruffi 2013. 8 Cavulli, Scaruffi 2012. 10 Cleuziou 2002, 2005, 2009.

West & East 317 Monografie, 4 Maurizio Cattani

Figure 4 Map of Early Bronze Age 1 sites (3100-2700 BC). Graveyard of Hafit type cairns and settlements (modified from Bortolini, Munoz 2015: fig. 4)

fication and exchange based on kinship instead of ed by herding in the steppe-like desert regions of the kingship, tribal alliances instead of states, classes interior. Most of the few known sites in the interior and royalty. are therefore related to the presence of semi-mobile From the evidence of these coastal sites of the pastoral groups who reared sheep, goats, and cattle.11 fifth and fourth millennia BC we can recognize the So far we presume that the documentation about strategies to exploit resources necessary for their this period is simply partial and we should expect a subsistence and at the same time the radicated com- higher evidence. plexity of societal bonds evidenced from collective The suspect seems to be confirmed by a demo- and individual burials. We suspect that also in the graphic increase, connected to an extraordinary interior, unfortunately less visible in the archae- spread of collective monumental cairns that charac- ological record, the same system of communities terized the so called Hafit period from the end of related to building kinship were developed in the fourth through the early third millennium BC, de- course of the fourth millennium BC. These com- fined also as Early Bronze Age 1 (fig. 4). So far it is munities were supported by a parallel development debated if this development was due to internal or of exploitation of resources both domestic and wild external factors. species established along the piedmont aquifers on 11 e.g. Uerpmann, Uerpmann 2000; Uerpmann et both slopes of the al-Hajar Mountains and integrat- Al. 2012; Lemee et Al. 2013.

West & East 318 Monografie, 4 Adaptation and transformation of human setting from Middle Holocene to Early Bronze Age in south-eastern Arabian Peninsula

Since Mesopotamian Ubaid pottery do not go to mark the approaches and to delimit tribal terri- beyond the Gulf area there is no evidence of attest- torial boundaries.16 The main evidence comes from ed previous interactions with outside, most of the the very few settlements excavated like Hili 8 in the scholars suggest a local and internal development: UAE,17 Bat,18 Khashabah19 in the western piedmont higher availability of food generates more people of Omani Hajar and especially from the hamlet of and more people mean more work, and more work HD-620 at Ra’s al-Hadd on the easternmost corner means more food. It is the time of the Great Trans- of Arabian Peninsula. formation12 with the foundations upon which the The site of Hili, in the piedmont 150 km east of Magan Civilization could rise in the third millen- Abu Dhabi, was the early agricultural oasis recorded nium BCE, marked by major development in tech- since the excavations from ’70s. Even if the chronol- nology and building social relationships: copper ogy of the earliest phase is debated21 the cultivation exploitation, mud brick architecture, date palm cul- of date palm, various cereals, legumes and fruits,22 tivation and related setting up of oasis system, in- represents the evidence at the end of fourth to early crease of trade and exchange of technologies. On the third millennium BC of an already developed agri- other side we cannot exclude the increasing force culture. Possibly included in early phases are some from Mesopotamia or the Iranian country suggest- structures for water transformed afterwards in a ed by some of these new characters. What we can network of channels and ditches at Hili 8.23 assume is that at the end of fourth millennium BC At the site of Bat,24 in the area of Ibri, recent in- not only subsistence, but transformation to a social vestigations carried out by the American team un- structure based on tribal alliance and enforced by covered structures and archaeological deposit of new forms of production and exchange. Hafit period under the later towers of mid of third Except the archaeobotanic evidence of date palm millennium BC, confirming that the beginning of cultivation there are not many other remains to oases could be fitted earlier.25 A geomorphological demonstrate that the oasis system was already acti- research demonstrated the environment suitable for vated from the end of fourth to early third millenni- the development of agriculture along the terraces um BC. Most of the archaeological documentation aligned with wadi Sharsah since the fourth millen- of this period is related to graveyards and especially nium BC, but no evidence of large structures com- from this it appears more and more clear that a sig- parable with the later falaj (sloped drain exploiting nificant cultural transformation was establishing in groundwater up to the surface sometime at long dis- whole Oman. tance) were found so far. Even the hydraulic struc- Some authors13 explain the poor evidence of set- ture from Bahla, similar to a falaj dated to 3000 tlement remains with the ephemeral nature of occu- BC,26 remains a fable evidence for the chronological pation related to nomadic pastoralists living in tem- attribution still debated.27 porary perishable ‘campsites’. The numerous cairns The research carried out at Bat suggests an early spread all over the landscape could represent along phase influenced by regular floods managed artifi- this interpretation as markers of tribal territories related to pastoral groups, displaying their control 16 Cleuziou 2007. and ownership of grazing lands.14 17 Cleuziou 1989. I prefer a different interpretation linked to the 18 Cable 2012; Thornton, Cable, Possehl 2016. adoption of a full oasis-based model of agriculture 19 Schmidt, Döpper 2017. 20 from very early in the Hafit period,15 that includ- Cattani, Azzarà 2014. 21 ed sedentary settlements and tombs around them Potts 1990. 22 Cleuziou, Costantini 1980; Tengberg 2003. 23 Charbonnier 2014. 12 Cleuziou 2002. 24 Frifelt 1976. 13 al-Jahwari 2008, pp. 150-151, 2013, p. 163. 25 Thornton, Cable, Possehl 2016. 14 Deadman 2012; Al-Jahwari 2013b, p. 166. 26 Orchard, Orchard 2007. 15 Cleuziou 2002, pp. 200‑201. 27 Charbonnier 2014.

West & East 319 Monografie, 4 Maurizio Cattani

cially and utilized for irrigation practices, prior to blocks belong to different construction phases and 3100 BC, exploiting especially the water table.28 the stratigraphic evidence indicates several epi- Recent investigations carried out by the German sodes of transformations, but we can notice that expedition at Khashabah29 are indicating the pres- all of them belong to a single project of a closed ence of a large settlement dated to early third mil- and protected settlement (fig. 5). lennium BC, made of large structures comparable The first occupation is related to a series of huts with Hili 8 towers. Looking forward for a detailed made from perishable materials, as indicated by analysis of the discoveries it appears of extreme im- postholes, pits and fireplaces cut into the sand dune portance to extend the evidence towards the eastern (Phase 1). Immediately after these ephemeral struc- side of Oman and especially to find out an interme- tures, the stone wall and the mud-bricks buildings diate site of Hafit settlement between the western start to be constructed, following two main phases Hajar (Hili 8) and the coast of Arabian Sea, where characterised by a specific type of mud-brick. The Ra’s al-Hadd is a well-known evidence. older phase (Phase 2) displays structures made from The Early Bronze Age settlement of HD-6 ap- mud-bricks with a high component of sand, while a pears to be the most meaningful site to highlight second occupation, labelled as (Phase 3), is related historical and archaeological aspects of a very im- to real mud-bricks, with clay of better quality (Figs. portant step of ancient Ja’alan and, in extended 5-6). So far, fourteen buildings with different lay- way, of the development of Eastern Arabia civili- outs have been identified; eight of them were com- zations. Radiocarbon datings and archaeological pletely explored to the floor levels, while the others finds fit the chronology of the site between the end were excavated only partially. of fourth and the beginning of third millennium Analysis of the building plans provides evidence BC30 highlighting the significance of three main of different constructive patterns within the site. evidence corresponding to the Great transforma- The first one, related to Phase 2, consists of build- tion: mud brick architecture, date palm cultiva- ings with large and small rooms, some of them relat- tion, copper exploitation. ed to build a kind of platform similar to later tow- The settlement is located south of the modern ers in inner Oman. The later pattern is the tripartite village of Ra’s al-Hadd at the end of a low coast building, typical of Phase 3, that consists of a large where the terrace upraises, continuing southward elongated rectangular room with two or three abut- for several kilometres. The structures are placed on ting smaller rooms on each side. The whole com- the top of an ancient sand dune next to the entrance plex is usually delimited by a single row of stones of high tide, used probably as harbour, between the with several courses at the base of the outer face. seacoast and an inner lagoon. Each building seems devoted to the same domes- The main architectural compound can be rec- tic and craft activities. The elongated rooms are al- ognized by a large stone wall interpreted as the ways characterised by several fireplaces, sometime foundation of a mud brick wall that surrounded overlapping each other, demonstrating the lifetime an area of ca. 2500 m2 with irregular plan, resem- of several activities related presumably to food-pro- bling an oval shape with some rectangular outer cessing, associated with tools such as pestles, grind- structures that correspond to enlargements at lat- ers and querns. er phases. The interior is characterized by several The buildings are part of different compounds, architectural blocks including tripartite buildings including irregularly shaped open area interposed of mud bricks, frequently placed on the top of a between houses and the perimetral stone wall and foundation made of stones and mud, interposed that may be defined as yard. These spaces are clear- with yard with ovens and fireplaces. Some of these ly independent and mainly connected to their respective building associated: they were used to for 28 Desruelles et Al. 2016. 29 Schmidt, Döpper 2017a, 2017b. a range of domestic activities, including, fish pro- 30 Cattani, Cavulli 2004; Azzarà 2009; Cattani, cessing activities and accumulating local marine re- Azzarà in press. sources or dates imported from the interior locate.

West & East 320 Monografie, 4 Adaptation and transformation of human setting from Middle Holocene to Early Bronze Age in south-eastern Arabian Peninsula

Figure 5 Ra’s al.Hadd, HD-6 Settlement. General plan (graphic from the author)

Figure 6 Ra’s al.Hadd, HD-6 Settlement. View of building 1 (photo M. Cattani)

West & East 321 Monografie, 4 Maurizio Cattani

Figure 7 Ra’s al.Hadd, HD-6 Settlement. Oven next to building 6 (photo M. Cattani)

In the same area were installed fireplaces and a large erpearl, conus rings (fig. 8). Very few fragments of hearth oven (fig. 7). pottery indicate that in early third millennium BC The other rooms are either connected to manu- pottery was not produced yet, a feature not unusu- factures or meant for storing purposes. Most of the al for most of the contemporary settlements and for craft activities were not related to a group of spe- burials, where only imported vessels are documented. cialised people; the great uniformity of the evidence The Hafit period corresponds to the develop- within the buildings indicates that the households ment of exploitation of copper.32 We do not have shared similar tasks in both daily domestic activi- direct evidence of copper mines dated to the end ties and craft processing of local resources, like shells of fourth millennium BC, but again the findings at and stones.31 HD-6 are the best proof of metal processing. So far, The most significant finds from the excavation the metal objects from Ra’s al-Hadd HD-6 are the are copper objects, flint tools, beads from different best-known archaeological metal complex from the materials (stone, shell, enstatite) and several kinds of worked shells, such as discs and inlaid made of moth- 32 Hauptmann, Weisgerber, Bachmann 1988; 31 Azzarà 2009; Cattani, Cavulli 2004. Prange et Al. 1997.

West & East 322 Monografie, 4 Adaptation and transformation of human setting from Middle Holocene to Early Bronze Age in south-eastern Arabian Peninsula

Figure 8 Ra’s al.Hadd, HD-6 Settlement. Worked shell finds and necklaces (photo Joint Hadd Project)

Figure 9 Ra’s al.Hadd, H D-6 Settlement. Metal finds (photo Joint Hadd Project)

Hafit period.33 There is no evidence of casting activ- The productions of rings and pendants made ities and copper artefacts, mainly devoted to fishing from Conus and Pinctada margaritifera has been at- activities, were manufactured in situ only hammer- tested within all the excavated structures. The differing or cutting the imported ingots in small block- ent stages of the chaîne opératoire are documented lets. Metal objects were widely used for daily activ- by finished Conus rings and ornaments made from ities such as fishing and knitting, as indicated by mother of pearl, a series of broken or incomplete fish-hooks, pins, obtained through cold hammering rings and pendants, apexes and fragments in course of copper ingots (fig. 9). of processing and number of wastes and discarded body whorls. 33 Giardino 2017.

West & East 323 Monografie, 4 Maurizio Cattani

The main household activity was the manufac- connected to the processing of small fish and mol- ture of beads, based on a variety of raw materials luscs in every dwelling, and by small fireplaces lo- and the setting of complete necklaces and brace- cated in the courtyards, but the most significant lets. Most of the production derived from shells and evidence is represented by large ovens (up to 2 me- steatite softstone, but other stone materials such as tres in diameter), probably used for the processing quartz, pyrophyllite, crystal rock, alabaster and jas- of large marine species and large amounts of other per were used as well. The complete chaîne opéra- fishes to be smoked or dried to preserve and prob- toire of bead making is represented: together with ably to exchange and send to other communities finished objects, finds include blocks and rods of in the interior. According to ethnography, marine different raw material, flakes and wastes. All house- resources represented the main protein supply for holds had specific spaces related to bead manufac- all traditional communities in Oman, regardless of ture, as demonstrated by large amounts of waste their environment and level of social organisation.36 products and unfinished objects. In the exchange network involving inland and Large quantities of beads were produced from coastal communities, fishermen provided smoked synthetic enstatite, an artificial material obtained and dried fish while acquired dates and other farm- through the hardening of soapstone baked up to ing products from the oases in the interior. The large 1100 °C.34 In addition, large quantities of beads amount of charred stone dates found from layers since made from chlorite softstone, most likely not pro- earliest phases dating to the end of fourth millenni- duced in situ, as suggested by the lack of any trace of um BC are the best proof of the connection with oases processing, have been recorded within all the exca- developing in the interior of Oman at the same time vated contexts. with date palms and water management structures. The subsistence was based mainly on marine re- It is possible that HD-6 and contemporary sites sources: rich deposits of fish bones attest inshore responded to similar economic and social patterns, and pelagic species, with a large predominance of based on the interdependency of complementary Scombridae, such as Euthynnus affinis (kawaka- groups. wa), Thunnus tonggol (longtail tuna) and Thunnus At the light of the results of the excavation, the albacares (yellowfin tuna); taxa represented include settlement of HD-6 can be considered as the earli- also Carangidae (jacks and trevallies), Lethrinidae est evidence of the transformation in the social sys- (emperors) and Haemulidae (grunts), together with tem in Oman. Social complexity is reflected by the a smaller number of other species.35 Sharks, and archaeological record: a range of new technologies green turtles were also part of the exploited species, especially in pyrotechnology (metalworking, manu- both for the nutritional qualities of their flesh and facture of synthetic enstatite), architecture of mud- for the value of their by-products such as leather, bricks with standard units and processing of high fat and carapaces. The consumption of dolphins, at- quality, craft-specialisation committed not only for tested by huge quantities, is still doubtful but could local exploitation, but also for trading with the pro- confirm the requirement of fat (essential in prevent- duction of a surplus. All these activities and knowl- ing protein poisoning within the diet of non-agrar- edge background are shared by the entire population ian communities). Also molluscs and crustaceans, as unifying choice to be involved as a community at such as mytilidae in large quantity, urchins and a larger level who decided to build and live togeth- crabs, played a significant role in the diet. er. At the same time the arrangement of architectur- Processing of marine fauna was related to a seal units, each of them kept accurately independent, ries of fire-structures such as large earth ovens. Each indicates the presence of families with relationships household probably provided its own daily meal, as based probably on kinship. The same corporate trib- suggested by the presence of several indoor hearths al configuration seems to be confirmed by several groups of cairns spread on the terraces around the 34 Panei, Rinaldi, Tosi 2005. 35 Kerneur 2007; Tosi et Al. 2001. 36 El Mahi 2001.

West & East 324 Monografie, 4 Adaptation and transformation of human setting from Middle Holocene to Early Bronze Age in south-eastern Arabian Peninsula

Figure 10 Map of Early Bronze Age 2 sites (2600-2000 BC). Umm an-Nar tombs and towers (modified from Bortolini, Munoz 2015: fig. 5)

site, looking at and at the same time clearly visible naturally dispersed in a vast network of seasonal riv- from the settlement. According to the size of the er flows or in subterranean mountain aquifers. The site, an estimate of up to 200 inhabitants, seasonally oasis systems ensured the possibility to expand eco- even more, could represent the real transformation nomic activities in new sectors for production and comparing to previous demographic configuration exchange: from food production, including cultiva- of foragers expected for the Middle Holocene. tion of dates, cereals and dairies, to a whole range of The settlement of Ra’s al-Hadd is therefore pro- craft activities, related especially to copper mining viding the proof of strong changes that involved the and stone quarrying. The growth of these new sec- Eastern Arabia from the end of fourth millennium tors was necessarily related to the exchange, hence BC, confirming that this process was extended in to the demand for local resources by surrounding the whole region of piedmont. HD-6 is presum- and distant countries. ably a simple projection of the revolutionary out- We do not know if this process originated else- come of the Great Transformation for prehistoric where and was imported as a whole system in Oman, Oman was rising: the establishment of the oases as but what seems verified is the involvement of local centers of aggregation, true demographic complex- population, taking part more and more into a web es where people and plants could be concentrated of long-distance exchanges from the interior to the once the supply of water could be granted by irri- coast and from the coast to other countries by the gation works, exploiting great quantities of water end of the fourth millennium BC.

West & East 325 Monografie, 4 Maurizio Cattani

At the half of third millennium BC, the process Figure 11 seems to be completed and directed to a turning Ra’s al-Jins, RJ-2. View of room 1 of building 1 point in the organization of production and in the with bitumen storage and infant burial (photo M. Cattani) social system. From at least the mid of the millennium or something earlier in the piedmont of inner Oman the earliest settlements transform in huge oases with typical monumental buildings of stone towers (fig.10). Each of investigated area usually shows a group of towers (from 3 to 6) different in architectural details but all corresponding to a similar dimension, with a diameter between 20 and 24 m. About the meaning of these monuments, it appears clear that the symbolic role is one of the most winning interpretation. No specific ritual function, not supported by other evidence, not enough to imagine as defensive structure alone, neither to protect goods or to protect chief of the society. Besides these hypotheses it is necessary to stress their role in economic and social changes that contributed to enlarge production or define specialization, increase population, and especially to wide interactions. The key points again are stressing the availability of resources, the adaptation to a different social system and the capability to exploit material in order to make all the country particularly active: water to produce food and sustain people, copper to create wealth, exchange to create profit and prosperity. Corresponding to the Umm an-Nar period, that I prefer to call Early Bronze Age 2, the second half food and especially, since we know from Mesopo- of third millennium BC is clearly a period marked tamian sources, prestigious tissues. This outline cor- by a huge exploitation and trade of metal, according responds to the rise of early complex polities who Mesopotamian texts and archaeological evidence. To decided to show themselves with their impressive exploit metal, it was necessary to sustain people and towers37 and with monumental tombs magnificent- to produce enough food in a more arid climate it rely built with accurate stone faces. quired to invent constant water supply in arable lands. Not surprisingly the Eastern Arabian oases came At this moment the need to assure water for humans, into being as a product of social complexity, togeth- animals and soil became priority that pushed to create er with the cultural integration that unified in vast canalization, the earliest version of falaj. regions the different compartments of specific adap- In the oases in the interior a huge mass of food tive pathways, mentioned in the cuneiform sourc- was produced, the interaction with the exploitation es of Mesopotamia during the second half of the of other resources already established since centu- third millennium BC with the name of Magan as a ries increased and strengthened the tribal relation- country beyond the Inferior Sea, as a respected trade ships in controlling wealth and especially the way to partner of the Sumerian cities.38 carry out these operations. At a certain point, it corresponded to the explosion of long-distance trade, 37 Cattani et Al. 2017; Cocca et Al. 2019. importing precious stones, presumably different 38 Laursen, Steinkeller 2017.

West & East 326 Monografie, 4 Adaptation and transformation of human setting from Middle Holocene to Early Bronze Age in south-eastern Arabian Peninsula

4. Concluding remarks different material. These new tombs seem to have contained about 150 individuals buried with more The evolution of civilization in Oman was very dif- complex rituals: skeletons were removed, dismem- ferent, respect to other countries.39 While regions bered and burnt with many ornaments, periodically along the alluvial corridors irrigated by the Nile, the some bones were removed and disposed in external Euphrates-Tigris twin courses and the Indus flood- pits or underground chambers with offerings. The plain raised the complexity of their societies on hier- entrance too becomes monumental with specially archies and royalty in order to administrate the accu- worked stones placed above the ground level.40 mulated wealth from vast food reserves, the common The same cultural evidence are spread from the foundations of ancient Arabian society were built Gulf to the easternmost side toward the Ocean. on tribal kinship developing a unique different ci- Here, the excavations at Ra’s al-Jins41 and at Ra’s al- vilisation. Monumentality was representing these Hadd42 are attesting the intense relationship with bonds among individuals and families through the the other side of the Ocean. tower building and the turret cairns, from early The significance of exchange and interaction third millennium BC with Hafit type to the second between the Indus Civilization and Umm an-Nar half of third millennium BC with larger and more communities has often been discussed in the general articulated Umm an-Nar tombs. These burials, usu- context of local economic and sociotechnical devel- ally closer to the settlements, present many struc- opments in the Oman Peninsula. Interactions with tural innovations: larger diameter and lower height; Indus seafaring merchants are attested by numerous inner partitions, dividing the tomb in two or more fragments of Indus black slipped jars43 and other im- chambers; greater structural articulation and differ- ports of precious objects (ivory, carnelian and other ent access; more accurate external covering with stones).

39 Cleuziou, Tosi 2007, Magee 2014.

West & East 327 Monografie, 4 Maurizio Cattani

BIBLIOGRAPHY

al-Jahwari N.S. 2008, Settlement patterns, development Cavulli F., Scaruffi S. 2012, The Middle Holocene and cultural change in Northern Oman Peninsula: a settlement of KHB-1 (Ra’s al-Khabbah, Sultanate multi-tiered approach to the analysis of long-term set- of Oman): An overview, in: j. Giraud, G. Gernez tlement trends (PhD. Thesis Durham University). (Eds.), Aux marges de l’archéologie: ouvrage à la al-Jahwari N.S. 2013, The Early Bronze Age funerary mémoire de S. Cleuziou, Travaux de la Maison René- archaeological landscape of the western part of Ja’alan Ginouvès, Paris, pp. 405-429. Region: results of two seasons of investigation, «Arabian Cavulli F., Scaruffi S. 2013, Thoughts on nomadism Archaeology and Epigraphy» 24, pp. 151‑173. in Middle Holocene Oman, «Arabian Archaeology Azzarà V. 2009, Domestic architecture at the Early and Epigraphy» 24, pp. 15-27. Bronze Age sites HD-6 and RJ-2 (Ja’ ali’in), Sultanate Charbonnier J. 2014, Groundwater Management of Oman, «Proceedings of the Seminar for Arabian in Southeast Arabia from the Bronze Age to the Iron Studies» 39, pp. 1‑16. Age: a Critical Reassessment, «Water History» 7/1, Biagi P. 1987, The prehistoric fishermen settlements pp. 1-33. of RH5 and RH6 at Qurum, Sultanate of Oman, Cleuziou S. 1989, Excavations at Hili 8, a preliminary «Proceedings of the Seminar for Arabian Studies» report on the 4th to 7th campaigns, «Archaeology in the 17, pp. 15‑19. United Arab Emirates» 5, pp. 61‑87. Biagi P., Salvatori S. 1986, Gli scavi nell’insedia- Cleuziou S. 2002, Présence et mise en scène des morts à mento preistorico e nella necropoli di Ra’s al-Hamra 5 l’usage des vivants dans les communautés protohisto- -Muscat – Oman), 1980- 1985, «Rivista di archeolo- riques: l’exemple de la péninsule d’Oman à l’âge du gia» 10, pp. 5-23. bronze ancien, in: Molinos M., Zifferero A. (eds.), Cable C.M. 2012, A Multitude of Monuments: Finding I primi popoli d’Europa, Firenze, pp. 17‑31. and Defending access to Resources in Third Millennium Cleuziou S. 2005, Pourquoi si tard? Nous avons pris BC Oman (PhD Thesis, Michigan State University un autre chemin. L’Arabie des chasseurs‑cueilleurs de (unpublished)). l’Holocène au début de l’Age du Bronze, in Guilaine Cattani M., Cavulli F. 2004, La Missione J. (ed.), Aux marges des grands foyers du Néolithique. Archeologica Italiana in Oman, in: M.T. Guaitoli, Périphéries débitrices ou créatrices?, Paris, pp. 123‑148. N. Marchetti and D. Scagliarini (eds.), Cleuziou S. 2007, Evolution toward Complexity in Scoprire. Scavi del Dipartimento di Archeologia. a Coastal Desert Environment. The Early Bronze Catalogo della Mostra, Bologna, pp. 225-232. Age in the Ja’alan, Sultanate of Oman, in: T.A. Cattani M., Azzarà V. in press, The Early Bronze Age Kohler, S. Van Der Leeuw (eds.), The Model- site of Ra’s al-Hadd HD-6, Muscat. Based Archaeology of Socionatural Systems. Santa Fe, pp. 213-231. Cattani M., al-Lawati H., al-Bakri S., Tosi M., Cocca E., Armigliato A., Mantellini S., Cleuziou S. 2009, Extracting wealth from a land of Vinci G. 2017, The excavation at the Bronze Age to- starvation by creating social complexity: a dialogue wer of Al-Khutm (Bat- Sultanate of Oman): a preli- between archaeology and climate?, «Comptes Rendus minary evaluation of the monument (Papers from the Géoscience» 341, pp. 726-738. fiftieth meeting of the Seminar for Arabian Studies Cleuziou S., Costantini L. 1980, Premiers éléments held at the British Museum, London, 29 to 31 July sur l’agriculture protohistorique de l’Arabie Orientale, 2016), London, pp. 51-56. «Paléorient» 6, pp. 255-261. Cattani M., Kenoyer J.M., Frenez D., Law R.W., Cleuziou S., Tosi M. 2007. In the Shadow of the Méry S. 2019, New excavations at the Umm an- Ancestors. The Prehistoric Foundations of the Early Nar site Ras al-Hadd HD-1, Sultanate of Oman (se- Arabian Civilization in Oman, Muscat. asons 2016‑2018): insights on cultural interaction and Cocca E., Vinci G., Cattani M., Armigliato long-distance trade, «Proceedings of the Seminar for A., Di Michele A., Bianchi M., Gennuso I., Arabian Studies» 49, pp. 69-84.

West & East 328 Monografie, 4 Adaptation and transformation of human setting from Middle Holocene to Early Bronze Age in south-eastern Arabian Peninsula

Mohammed Al-Lawati H.E.H., Al-Bakri S., Laursen S., Steinkeller P. 2017, Babylonia, the Gulf Al-Maqbali S. 2019, Al-Khutm Project 2017/2018: Region and the Indus. Archaeological and Textual A Bronze Age Monumental Tower (Bat, Oman), Evidence for Contact in the Third and Early Second «Proceedings of the Seminar for Arabian Studies» Millennia BC (Mesopotamian Civilizations 20), 49, pp. 85-96. Winona Lake, Indiana. Coppa A., Macchiarelli R., Salvatori S., Santini Lemee M., Gernez G., Giraud J., Beuzen-Waller G. 1985, The Prehistoric Graveyard of Ra’s al-Hamra T., Fouache E. 2013, Jabal al-’Aluya: an inland (RH5): A short preliminary report on the 1981-83 Neolithic settlement of the late fifth millennium BC in excavations, «JOS» 8, pp. 97-102. the Adam area, Sultanate of Oman, «Proceedings of Coppa, A., Damadio S.M., Armelagos G. J., the Seminar for Arabian Studies» 43, pp. 197-212. Mancinelli D., Vargiu R. 1990, Paleobiology and Magee P. 2014. The Archaeology of Prehistoric Arabia: paleopathology: a preliminary study of the prehistoric fi- Adaptation and Social Formation from the Neolithic shing population of Ra’s al-Hamra 5 (Qurum, Sultanate to the Iron Age, Cambridge. of Oman, 3.700–3.200 BC), «Anthropologia Orchard J., Orchard, J. 2007, The third millennium Contemporanea» 13/4, pp. 329-336. BC oasis settlement of Oman and the first evidence of Deadman W.M. 2012, Defining the Early Bronze Age their irrigation by Aflaj from Bahla, in: Archaeology of landscape: a remote sensing-based analysis of Hafit the Arabian Peninsula through the Ages. Proceedings tomb distribution in Wadi Andant, Sultanate of of the International Symposium (7th-9th May 2006), Oman, «Arabian Archaeology and Epigraphy» 23, Muscat, pp. 143-173. pp. 26-34. Panei L., Rinaldi G., Tosi M. 2005, Investigations on Desruelles S., Fouache E., Eddargach W., ancient beads from the Sultanate of Oman (Ra’s al- Cammas C., Wattez J., Beuzen-Waller T., Hadd – Southern Oman), «ArchéoSciences, Revue Martin C., Tengberg M., Cable C., Thornton d’Archéométrie» 29, pp. 151-155. C., Murray C. 2016, Evidence for early irrigation Potts D.T. 1990, The Arabian Gulf in antiquity. Vol. I, at Bat (Wadi Sharsah, northwestern Oman) before From prehistory to the fall of the Achaemenian Empire, the advent of farming villages, «Quaternary Science Oxford. Reviews» 150, pp. 42-54 Prange M.K., Götze H.-J., Hauptmann A., El -Mahi A. T. 2001, The traditional pastoral groups of Weisgerber G. 1997, Is Oman the ancient Magan? Dhofar, Oman: A parallel for ancient cultural ecology, Analytical studies of copper from Oman, in: Young, «Proceedings of the Seminar for Arabian Studies» S. M. M., M. Pollard, P. Budd and R. A. Ixer 31, pp. 131‑43. (eds.), Metals in antiquity (BAR International Series Frenez M., Cattani M. (eds.) 2019, Sognatori. 40 792) pp. 187‑192. Anni di Ricerche Archeologiche Italiane in Oman. Schmidt C., Döpper S. 2017a, The development of com- Dreamers. 40 Years of Italian Archaeological Research plexity at 3rd millennium BC al-Khashbah, Sultanate in Oman, Bologna. of Oman: Results of the first two seasons 2015 and Frifelt K. 1976, Evidence of a third millennium BC town 2016, «Proceedings of the Seminar for Arabian in Oman, «Journal of Oman Studies» 2, pp. 57-74. Studies» 47, pp. 215‑226. Giardino C. 2017, Magan. The Land of Copper. Schmidt C., Döpper S. 2017b, Die Entstehung kom- Pehistoric metallurgy of Oman, Muscat. plexer Siedlungen im nördlichen Inner-Oman im 3. Hauptmann, A, G Weisgerber and H. G. Jahrtausend v. Chr.: Bericht über die Ausgrabungen Bachmann 1988, Early copper metallurgy in Oman, 2015 und 2016 in Al-Khashbah, «Mitteilungen der in: Maddin R. (ed.) The beginning of the use of metals Deutschen Orient-Gesellschaft» 149, pp. 121‑158. and alloy, Cambridge, pp. 34-51. Tengberg M. 2003, Archaeobotany in the Oman pen- Kerneur S. 2007, La pêche du Néolithique à l’âge insula and the role of Eastern Arabia in the spread du Bronze au Sultanat d’Oman. Synthèse des con- of African crops, in: Neumann K., Butler A., naissances actuelles (Mémoire de Master 2 d’Ar- Kahlheber S. (eds.), Food, Fuel and Fields: Progress chéologie, Université de Paris 1 Panthéon-Sorbonne in African Archaeobotany, Africa Praehistorica, 15, (unpublished)). Koln, pp. 229-237. Kohl P.L. 2008, Shared Social Fields: Evolutionary Thornton C.P., Cable C.M., Possehl G.L. (eds.) Convergence in Prehistory and Contemporary Practice, 2016, The Towers of Bat: Six Seasons of Excavations «American Anthropologist» 110/ 4, p. 495. (2007-2012), Philadelphia.

West & East 329 Monografie, 4 Maurizio Cattani

Tosi M. 1989, Protohistoric archaeology in Oman: the the Emirate of Sharjah (SE-Arabia) d Excavations first thirty years (1956–1985), in: Costa P. M. and 1995-1998, in: Mashkour M., Choyke A.M., M. Tosi (eds.) Oman Studies: papers on the archaeol- Buitenhuis H., Poplin F. (eds.), Archaeozoology ogy and history of Oman (Serie Orientale Roma 63), of the Near East IVB d Proceedings of the Fourth Rome, pp. 135‑161. International Symposium on the Archaeozoology of Tosi M., Cattani M., Curci A., Marcucci L.G., Southwestern Asia and Adjacent Areas, Groningen, Usai D. 2001, Missione archeologica nel Sultanato pp. 40-49. di Oman “Joint Hadd Project” Campagna di ricer- Uerpmann M., de Beauclair R., Handel M., ca 2000–2001, «Ocnus, Quaderni della Scuola di Kutterer A., Noack E., Uerpmann H.P. 2012, Specializzazione in Archeologia» 9/10, pp. 357-366. The neolithic site FAY-NE15 in the central region of Uerpmann M., Uerpmann H.P. 2000, Faunal re- the Emirate of Sharjah (UAE), «Proceedings of the mains of al-Buhais 18, an aceramic Neolithic site in Seminar for Arabian Studies» 42, pp. 385-400.

West & East 330 Monografie, 4 A transitional model between Umm An-Nar and Hafit cultures: the case study of Grave 4 of Al Arid*

MASSIMILIANO GHIRO*, SABATINO LAURENZA**, ENZO COCCA**

* University of Udine, ** Independent Researcher

Abstract

This paper presents the case study of a third millennium BC grave in Al-Arid, Ibri, Oman. The excavation and the analysis of the grave have been carried out during the archaeological rescue excavation project, directed by Dr. Sabatino Laurenza for the Ministry of Heritage and Culture of Oman, related to the works for the duplication of the road between Ibri and Yanqul, in Northern Oman. Controversial and interesting is the dating of the tomb: the structure is close to Umm an-Nar type (2600‑2000 BC), while the only element found inside it is a biconical small jar ascribable to Jemdet Nasr, Mesopotamian culture (3100-2900 BC). While the contacts between Jemdet Nasr culture and Hafit culture (3200-2800 BC) are already well known, our case study seems to represent a transitional moment between Hafit and Umm an-Nar cultures. Extremely interesting, at the same time, is the presence of skeletal remains only inside of the biconical small jar, while funeral chambers are completely empty.

Keywords

Al-Arid, Oman, grave, Hafit, Umm An-Nar, Jemdet Nasr, Bat-Type, Age of Bronze

* This paper is taken from bachelor’s thesis of Massimiliano Ghiro (2017).

West & East 331 Monografie, 4 Massimiliano Ghiro, Sabatino Laurenza, Enzo Cocca

1. Geographical and geomorphological results, Kondo proposes the hypothesis that those settings two terraces were formed during the humid period of the early Holocene (10.5–6 kya).2 The site of Al-Arid, in the north-western Omani region of Ad-Dhahirah, is located between the modern cities of Yanqul, approximately 20 km 2. Archaeological settings North, and Ibri, about 30 km South. Geographi- The area interested by the archaeological rescue ex- cally speaking, Al-Arid lies at the interface be- cavation has a rectangular shape with a width of 60 tween Al-Hajar Mountains, North East, and Rub m and a length of about 2 km (fig. 2). Arbitrarily ‘al-Khali desert, South West, area that plays a key subdivided in 7 sectors, the site is configured like a role for internal trade routes of Arabian Peninsu- typical example of settlement slope, with in the back- la.1 From a geomorphologic point of view the area ground, outside the investigated area, on the top of is basically an alluvial plan flooded and bounded by a hill, several visible 12 Hafit beehive graves. On the two main wadis: Wadi al-Khubayba and Wadi al- lowest part of the terrace A and widespread down to Kabir. The plain is located between the limestone the plain, there is a total of around 25 archaeological massif of Hawrat al-Arid, Jabal Wabah, and Jabal features: structures including probably the remain- Wahrah to the South. ing traces of platforms and Pre-Islamic towers, stone The area has an altitude comprised between 900 footings of dwellings, and other irrigation works, m and 400 m and presents two main alluvial depos- falaj traces of copper slag indicating the possible presence its that form terraces A and B (fig. 1). Both are more of smelting kilns, some scattered Islamic graves and than 1 or 2 m high above the present level of the an Islamic cemetery, a little bit outside the respect wadi’s watercourse. Terrace A is higher than terrace area of the alignment. B and antecedent to it. Based on the preliminary

Figure 1 Geomorphological and archaeological features in the Wadi al-Kabir basin (Kondo et Al. 2014, p. 228) Al-Arid=ARS01

1 Frifelt 1975, p. 383; Williams-Gregoricka 2013, p. 134. 2 Kondo et Al. 2014, p. 229.

West & East 332 Monografie, 4 A transitional model between Umm An-Nar and Hafit cultures: the case study of Grave 4 of Al Arid

Figure 2 Archaeological investigated area (Photo Archeological Rescue Excavation Team)

3. The Grave G4 graves were completely covered by debris and were not investigated. However, apparently, they appear The Grave G4 is part of a group of six graves (fig. 3) to be circular in shape, despite the collapse compro- that look similar for size, material and location. All mising the interpretation, and all the structures, like of them are standing on a natural terrace sloping the Grave G4, were built with irregular local lime- down towards the wadi. The Grave G4 is the closest stone stones, medium and large sized (20-60 cm), to the wadi bed (approximately 50 m from it) and it and light grey in color. is distanced about 50/60 m from the concentration The Grave G4 (fig. 4-5), excavated following of four graves on the South East side. Around 70 the stratigraphic method, has a circular shape of ap- m East there is also a sixth smaller grave. The other proximately 8 m diameter. Initially covered almost

Figure 3 The group of six tombs in Sector 1, Al Arid (Photo Archeological Rescue Excavation Team)

West & East 333 Monografie, 4 Massimiliano Ghiro, Sabatino Laurenza, Enzo Cocca

Figure 4 Plan of Grave G4 (Photo Archeological Rescue Excavation Team)

Figure 5 The Grave G4 (Photo Archeological Rescue Excavation Team)

entirely by the collapse, the grave appears well pre- (SU408), 2 m long and adjacent to the entrance. On served in the lowest rows. The removal of the layer the East side, between the Internal Wall (SU404) of collapsed stones started from the Northern side, and the Dividing Wall (SU408) there is a space oc- where it was possible to recognize an alignment cupied by an Inner Obstruction (SU407), seriously of 8 stones at the ground level. Following it, this damaged by the collapse. The entrance is blocked by alignment turned out to be the profile of the Ex- local unworked limestone stones. From the point of ternal Wall (SU402), which has a width of 80 cm. view of the construction techniques, the External From the top, the circular structure consists of two Wall (SU402) and the Internal Wall have a dou- concentric Walls (SU404, 402) and a Filling layer ble row of stones spaced out by a small filling (25 (SU403), with a central chamber divided in north cm) of small stones and sand without an ordered chamber and south chamber by a Dividing Wall arrangement. The Internal Wall measures 1,10 m

West & East 334 Monografie, 4 A transitional model between Umm An-Nar and Hafit cultures: the case study of Grave 4 of Al Arid

Figure 6 Figure 7 Grave 603, Bat (Böhme 2011, fig. 4) Section and plan of grave 603, Bat, Grid 1 m. (Böhme 2011, fig. 2)

and is enlarged slightly near the entrance up to a which presents peculiarities ascribed to the tombs width of 1,40 m, probably to ensure major stabili- both of the Hafit period and Umm an-Nar period. ty to the weakest part of the structure. The Filling In regards to the peculiarities Hafit we can between the External Wall and the Internal Wall, consider: wide around 70 cm, consists of medium stones and 1. use of local unworked stones;4 sand. The Dividing Wall is interconnected to the 2. absence of the plinth as basis of the grave; 5 Internal Wall on the West boundary and is orient- 3. single entrance at ground level;6 ed approximately 20° North of the entrance. Exter- 7 nally the Filling between the External Wall and the 4. blockage of the entrance with unshaped stones; Internal Wall leans both towards the Blockage of 5. absence of bones;8 the Entrance (SU410) and towards the Threshold 6. very poor grave goods;9 of the Entrance (SU412), suggesting that the Exter- nal Wall and the Filling were completed when the 4 Cleuziou-Tosi 2007, p. 112; Potts 2012a, p. 36, grave had already been closed. 2012b, p. 372; Williams-Gregoricka 2013, pp. 141, 146; Yule-Weisgerber 1998, p. 192. Although the Omani Bronze Age is divided into 5 Cleuziou-Tosi 2007, p. 125; Munoz-Ghazal- three main periods (Hafit, Umm an-Nar, Wadi Guy 2012, p. 453; Potts 2012b, p. 372. Suq), in keeping with the literature, there is a series 6 Cleuziou-Tosi 2007, p. 125; Potts 2012b, p. 372; of tombs that do not have all the canonical peculiar- Yule-Weisgerber 1998, p. 192. ities expected and cannot be classified as belonging, 7 Bortolini-Munoz 2015, p. 67; Cleuziou-Tosi 2007, p. 112; Potts 2012a, p. 36; Williams-Gregoricka unambiguously, to a specific period. It is therefore 2013, pp. 141, 143; Vogt 1985, p. 103. possible that Hafit and Umm an-Nar may be part of 8 “The cairns are often found empty or with few bones a single evolutionary process and not of two distinct and objects inside.” Cleuziou-Tosi 2007, p. 112; Frifelt periods3. This is the case of the Grave G4 of Al-Arid, 1970, p. 377; Williams-Gregoricka 2013, p. 147. 9 Cleuziou-Tosi 2007, pp. 112, 114; Williams- 3 Potts 2012b, p. 372. Gregoricka 2013, p. 147.

West & East 335 Monografie, 4 Massimiliano Ghiro, Sabatino Laurenza, Enzo Cocca

7. Jemdet Nasr jar;10 2. dimensions.17 Cleuziou and Tosi18 notice an in- 8. external podium, therefore the grave was con- crease in complexity and size of the tombs Umm ceived and constructed as a central body and an an-Nar over time, «from 7-8 metres at the be- external body,11 like grave 603 of Bat (fig. 6-7). ginning, the diameter reaches 10-12 metres by the end of the third millennium BC». On that last point, since the Hafit tombs were not collective in synchronic sense but diachronic, and Entrance facing West. About this, following not all Hafit tombs have the external podium, it is the study of Belmonte and González-García possible to advance the interpretation of the exter- (fig. 10A‑10B), the Hafit tombs have their single nal podium as an element that marks the end of the entrance localized on the third quadrant (West, use of the tomb. If this is true, can be the closure of South and South-West), while the Umm an-Nar the graves be connected with the extinction of the tombs have usually two gates with opposite direc- family branch, for which the tombs were intended? tions on the cardinal points, concentrated especially The Umm an-Nar elements are: on North, West and South.19 The almost totally ab- 1. tomb location on the alluvial terrace;12 sence of East gates is explained by the existence of Umm an-Nar tombs with a single West entrance. It 2. subdivision of the chamber by a dividing wall;13 is legitimate to suppose that these tombs with a sim- 3. chambers size: 2,60 m for the northern chamber, pler plan are the most ancient of the Umm an-Nar 14 2,30 m for the southern. period and also suggest continuity with the previous Hafit period. Finally, as transitional elements between the two periods we can consider: 1. construction techniques: double curtain of stones with filling, visible always in Hafit graves15 and in some Umm an-Nar graves16 (fig. 8-9);

10 Bortolini-Munoz 2015, p. 67; Cleuziou-To- si 2007, p. 114; Frifelt 1970, 1975a, p. 375, 1975b, Méry-Schneider 1996, p. 81; Potts 1986, p. 129, 2009, p. 32, 2012a, p. 42; Williams-Gregoricka 2013, pp. 141, 143-145, 147. 11 Böhme 2011, p. 26; Bortolini-Munoz 2015, p. 67; Potts 2012a, p. 36; Yule-Weisgerber 1998, p. 192. 12 Cleuziou-Tosi 2007, pp. 124-125; Condolu- ci-Degli Esposti 2015, p. 11; Frifelt 1975a, p. 374; Potts 2012b, p. 372; Yule-Weisgerber 1998, p. 194. Generally, the Hafit graves were positioned on the rocky ridges, crests and hills, or mountains. Bortolini-Munoz 2015, p. 67; Cleuziou-Tosi 2007, p. 107; Munoz-Ghazal- Figure 8 Guy 2012, pp. 452-453; Potts 2012a, p. 37, 2012b, p. 372; Grave 401, Bat Yule-Weisgerber 1998, pp. 191-192; Williams-Grego- (Böhme-Al Sabri 2011, fig. 26) ricka 2013, p. 146; Vogt 1985, pp. 70-71. 13 Bortolini-Munoz 2015, p. 71; Cleuziou-Tosi 2007, p. 125; Döpper 2017, p. 205; Munoz-Ghazal-Guy 2012, p. 453; Potts 2012a, p. 46, 2012b, p. 372; Vogt 1985, p. 110. 14 These measures can be considered related to the Umm an-Nar graves, whereas the internal diameters of Hafit graves are 2 m, or less. Döpper 2017, p. 205; Vogt 1985, p. 103. 17 Frifelt 1975a, p. 389; Yule-Weisgerber 1998, 15 Cleuziou-Tosi 2007, p. 112. pp. 192, 194. 16 Böhme-Al Sabri 2011, pp. 114-115; Döpper 2017, 18 Cleuziou-Tosi 2007, pp. 127-128. p. 191. 19 Belmonte-González García 2014, pp. 240-241.

West & East 336 Monografie, 4 A transitional model between Umm An-Nar and Hafit cultures: the case study of Grave 4 of Al Arid

Figure 9 Plan of grave 401, Bat (Böhme-Al Sabri 2011, fig. 4)

Figure 10 A-10 B Orientation diagrams of Hafit (a) and Umm en Nar (b) type tombs in ancient Magan (Belmonte González-García 2014, fig. 10)

West & East 337 Monografie, 4 Massimiliano Ghiro, Sabatino Laurenza, Enzo Cocca

4. Jemdet Nasr Jar collapse of the structure, not by anthropic violations. The vessel was filled with soil mixed with patinated, A fragment of small bronze pin and a small undec- blackened and rounded, bones fragments of an adult orated pottery jar were found near the entrance individual. The sex is not determinable. This kind of (fig. 11A-11B). The vessel has an everted lip and rim fragments has been linked to multiple reworking of with triangular section and a pointed profile. The bones inside funerary structures. All this evidence neck is trunk-conical, while the body is biconical substantiates the hypothesis that we are dealing with with a rounded shoulder. The bottom is concave. Its fragments that were already in secondary deposition: height measures 13.2 cm, while the rim diameter is indeed, no fragment can be associated with the sur- 6.8 cm, the maximum diameter is 15.2 cm, and the rounding ones. This means that the fragmentation base diameter is 6.7 cm. The fabric is very fine, with occurred outside of the jar before the fragments were microscopic mineral inclusions. The jar is very well collected. We cannot propose strong hypotheses on preserved. It was found in standing position, with a the origin of the fragment, but the lack of bones in single fragment of the neck with rim broken in situ. the grave and of sign of manumission, together with The standing position of the jar and its location near the position of the jar in the grave suggest that the the entrance and its optimal state of preservation place of first inhumation of these human remains is suggest its primary deposition, disturbed only by the another one than the Grave G4.

Figure 11 A, B Jar from Grave G4, Al Arid (Photo Archeological Rescue Excavation Team)

West & East 338 Monografie, 4 A transitional model between Umm An-Nar and Hafit cultures: the case study of Grave 4 of Al Arid

5. Discussion was dated to 2880–2670 BC (cal. 2s), and the charcoal was dated to 2600‑2480 BC (cal. 2s).30 As made evident by the description, the Grave G4 However, the most punctual comparison is giv- presents architectural elements related both to the en by the grave 602 (ex 1138) from Bat (fig. 13). It Hafit period and Umm an-Nar period and a Jemdet presents, in fact, an identical plan to Grave G4 in Nasr jar. For these reasons the research of compari- regards to the orientation of the entrance, the po- sons has been carried out keeping in mind the geo- sition of dividing wall and their alignment. More- graphical position, the architectural characteristics over, there is a rim fragment (fig. 12D) from this and their development and the jar. About the first grave that looks similar to G4’s jar. two elements several interesting parallels are from The jar can be compared with exemplars from Bat20, Jabal al-Emalah,21 Falaj al Qaba’il,22 Jaylah23, two Oman sites: according to the parallels, the ves- Jebel al-Buhais,24 Umm an-Nar,25 Al-Khubayb.26 sel can be attributed to the Jemdet Nasr period The research carried out during the Social, (3100-2900 BC). Spatial, and Bioarchaeological Histories of An- cient Oman (SoBO) project in north-western Oman is interesting. In particular, the necropolis of Al-Khubayb presents some sealed tombs des- ignated by Kimberly Williams and Lesley Grego- ricka as transitional between the Hafit and Umm an-Nar periods, which share several affinities with the Grave G4 of Al-Arid. These tombs (S007-001 and S007-003) have, in fact, the same geographic and geomorphological context, together with the presence of Jemdet Nasr jars inside the graves and other nearby tombs.27 There are also similar Figure 12 A Jar from Jebel Hafit, cairn 2, height 12.5 cm, rim diameter architectural features such as their general large 9 cm, base diameter 7.5 cm size for Hafit tombs, their carefully placed but (Potts, 1986, fig. 4:5 = Frifelt, 1970, fig. 13A) unworked stones and, in the case of tomb S007- Scale 1:3 001, the beginning of a multiple-chambered burial chamber.28 For the tomb S007-001 the bioap- atite sample from Individual C returned a date of 2861–2580 BC (cal. 2s) and the charcoal sample returned a similar date of 2910–2750 BC (cal. 2s),29 while for the tomb S007-003 the bone

20 Böhme 2011, 2012; Böhme-Al Sabri 2011; Fri- Figure 12 B felt 1975a. Jar from Jebel Hafit, cairn 22, height 10.5 cm, rim 21 Potts 2012b. diameter 8.5 cm, base diameter 6 cm 22 Potts 2012b. (Potts, 1986, fig. 1I = Frifelt, 1970, fig. 21C) 23 Yule-Weisgerber 1998. Scale 1:3 24 Jasim 2012a, 2012b. 25 Al Tikriti 2012. 26 Döpper 2017; Williams-Gregoricka 2013. 27 Williams-Gregoricka 2013, pp. 141-142. 30 “While some overlap between these samples is pres- 28 Döpper 2017, p. 205. ent, the more recent date from the wood charcoal may indicate tomb reuse for the interment of later occupants.” Wil- 29 Williams-Gregoricka 2013, p. 143. liams-Gregoricka 2013, p. 145.

West & East 339 Monografie, 4 Massimiliano Ghiro, Sabatino Laurenza, Enzo Cocca

6. Conclusions

The dating of the Tomb G4 of Al-Arid is problematic and interesting. From an architectural point of view, it is possible to contextualize the tomb in a horizon of transition between the Hafit (3200-2800 BC) and Umm an-Nar (2600-2000 BC) periods, Figure 12 C proposing a reasonable dating between 2800 BC Jar from Jebel Hafit, cairn 23, height 11 cm, rim diameter and the 2500 BC. 8 cm, base diameter 4 cm However, the presence, inside the sealed tomb, (Potts, 1986, Fig 1F = Frifelt, 1970, fig. 23A) of a jar referable to Mesopotamian culture Jemdet Scale 1:3 Nasr (3100-2900 BC) poses a chronological problem for the interpretation of the grave. In our opinion the asynchronism of this evidence can be explained by formulating two possible hypotheses. In the first hypothesis, the context could be earlier than the proposed date, and consequently the tomb would belong only to the Hafit horizon. Moreover, contacts between cultures Jem- det Nasr and Hafit are well documented in the archaeological record. Anyway, this cannot solve the problem of architectural elements typical of Umm Figure 12 D Rim from jar, Bat, grave 602 – ex 1138 an-Nar culture which characterize the tomb (tomb (Frifelt 1975b, fig. 12) location on the alluvial terrace, subdivision of the chamber by a dividing wall), unless one considers them to be anticipations of future architectural can- ons of Umm an-Nar tradition. While this is possible, it remains in our opinion highly unlikely. As for the second hypothesis, if we consider valid the proposed dating (2800-2500 BC), it can be assumed that the vessel comes from a more ancient Hafit-Jemdet Nasr context. This could have been clearer if it had been possible to date the bone fragments in the jar. On the basis of the available data, the second possibility seems to be more plausible.

Figure 13 Plan of grave 602 (ex 1138) (Frifelt 1975b, fig. 23)

West & East 340 Monografie, 4 A transitional model between Umm An-Nar and Hafit cultures: the case study of Grave 4 of Al Arid

BIBLIOGRAPHY

Al-Tikriti W. Y. 2012, Umm an-Nar, an ancient ca- Jasim S. 2012a, The Necropolis of Jebel al-Buhais – pital of Abu Dhabi: distribution of a culture and the Prehistoric Discoveries in the Emirate of Sharjah, current state of the site, in: Potts D.T., Hellyer P. United Arab Emirates, Sharjah, pp. 283-289. (eds.), Fifty years of Emirates Archaeology, Abu Dhabi, Jasim S. 2012b, Jebel al-Buhais 2008: A year in review, pp. 86-99. in: Potts D.T., Hellyer P. (eds.), Fifty years of Belmonte, J. A., González-García, A. C., 2014, Emirates Archaeology, Abu Dhabi, pp. 122-131. On the Orientation of Early Bronze Age Tombs in Kondo Y., Beuzen-Waller T., Miki, Noguchi A., Ancient Magan, «Mediterranean Archaeology and Desruelles S., Fouache E. 2014, Geoarchaeological Archaeometry» 14/ 3, pp. 233-246. Survey in the Wadi al-Kabir Basin, Wilayat Ibri, Böhme M. 2011, The Bat-Type. A Hafit period tomb con- Oman: A Preliminary Report, «Proceedings of the struction in Oman, «Arabian archaeology and epi- Seminar for Arabian Studies» 44, pp. 227-234. graphy» 22, pp. 23-31. Méry S., Schneider G. 1996, Mesopotamian Pottery Böhme M.2012, The restoration of Tomb 154 at Bat, Wares in Eastern Arabia from the th5 to the 2nd in: Potts D.T., Hellyer P. (eds.), Fifty years of Millennium BC: A Contribution of Archaeometry to Emirates Archaeology, Abu Dhabi, pp. 112-121. the Economic history, «Proceedings of the Seminar Böhme M., Al-Sabri B. A. 2011, Umm an-Nar burial for Arabian Studies» 26, pp. 79-96. 401 at Bat, Oman: architecture and finds, «Arabian Munoz O., Ghazal R. O., Guy H. 2012, Use of ossuary archaeology and epigraphy» 22, pp. 113-154. pits during the Umm an-Nar Period: New insights on Bortolini E., Munoz O. 2015, Life and Death in the complexity of burial practices from the site of Ra’s al- Prehistoric Oman: Insights from Late Neolithic and Jinz (RJ-1), Oman, in: Giraud J., Gernez G., Aux Early Bronze Age Funerary Practices (4th-3rd mill. marges de l’archéologie. Hommage à Serge Cleuziou BC), in: Proceedings of the International Symposium (Travaux de la Maison Archéologie & Ethnologie, “The Archaeological Heritage of Oman” (UNESCO, René-Ginouvès 16), pp. 451-467. Paris, September 7, 2012, Ministry of Heritage and Potts D.T. 1986, Eastern Arabia and the Oman pe- Culture, Sultanate of Oman), pp. 61-80. ninsula during the late fourth and early third millen- Cleuziou S., Tosi M. 2007, In the shadow of the ance- nium B.C., in: Finkbeiner U.,Rölling W. (eds), stors. The prehistoric foundations of the early Arabian Gamdat Nasr: Period or Regional Style? (Beihefte civilization in Oman, Muscat. zum Tübingen Atlas des Vorderen Orients, Reihe B 62), Wiesbaden, pp. 121-170. Condoluci C., Degli Esposti M. 2015, High places in Oman: the IMTO excavations of Bronze and Iron Potts D. T. 2009, The Archaeology and Early History Age remains on Jabal Salut (Quaderni di Arabia an- of the Persian Gulf, in: Potter L.G. (ed.), The Persian tica 3), Roma. Gulf in History, New York, pp. 27-56. Döpper S. 2017, Rescue excavation of a third-millen- Potts D. T. 2012a, In the land of the Emirates: the ar- nium BC tomb at Al-Khubayb, Sultanate of Oman, chaeology and history of UAE, London. «Arabian Archaeology and Epigraphy » 28/2, Potts D. T. 2012b, The Hafit-Umm an-Nar Transition: pp. 190-208. evidence from Falaj al-Qaba’il and Jabal al-Emalah, Frifelt K. 1970, Jemdet Nasr Graves in the Oman, in: Giraud J., Gernez G., Aux marges de l’ar- «Kuml», pp. 374-383. chéologie. Hommage à Serge Cleuziou (Travaux de la Maison Archéologie & Ethnologie, René-Ginouvès Frifelt K. 1975a, On Prehistoric Settlement and 16), pp. 371-377. Chronology of the Oman Peninsula, «East and West» 25, pp. 359-424. Potts D. T., Hellyer P. 2012, Fifty Years of Emirates Archaeology: Proceedings of the Second International Frifelt K. 1975b, A possible link between the Jemdet Conference on the Archaeology of the United Arab Journal Nasr and the Umm an-Nar grave of Oman, « Emirates, Abu Dhabi. of Oman Studies» 1, pp. 57-80.

West & East 341 Monografie, 4 Massimiliano Ghiro, Sabatino Laurenza, Enzo Cocca

Yule P., Weisgerber G. 1998, Prehistoric Tower Vogt B. 1985, Zur Chronologie und Entwicklung Tombs at Shir/Jaylah, Sultanate of Oman, «Beiträge der Gräber des späten 4.-2. Jtsd.v.Chr. auf der zur allgemeinen und vergleichenden Archäologie» Halbinsel Oman: Zusammenfassung. Analyse und 18, pp. 183–242. Würdigung publizierter wie auch unveröffentlichter Williams K.D., Gregoricka L.A. 2013, The Social, Grabungsergebnisse (unpublished dissertation, Spatial, and Bioarchaeological Histories of Ancient Georg-August-Universität, Göttingen), Göttingen. Oman project: The mortuary landscape of Dhank, «Arabian Archaeology and Epigraphy» 24, pp. 134‑150.

West & East 342 Monografie, 4 Graves, distribution and social memory: towards a new definition of funerary landscape in Oman

SABATINO LAURENZA, MARCO BIANCHI, ANTONIO DI MICHELE

Independent Researchers

Abstract

From 2014 Italian team, directed by Sabatino Laurenza, worked in the Sultanate of Oman for rescue archaeological excavations for roads construction. The excavations on the Batinah Expressway alignment in Sohar area touched three graveyards with different typologies of graves and with materials of different periods (from Bronze age to Sasanian and early Islamic). The results pushed us to apply a new approach to the “funerary landscape” study of the region. In this paper we present the “funerary landscape” as a specific type of archaeological landscape, focusing on a “mapscape” of the funerary sites and a detailed account of graves typologies and funerary finds (burialscape) and the relations between burials, disposal of the graves in the environment (funeraryscape) and the social memory of the group participating in the remembrance of the burial, through a series of standardized uses (i.e. graves organized in large groups over wide area, the repeated disposal of the dead in the same place, etc.). Those and other aspects let the area to become a place of remembrance of persons in a community’s social memory, reflecting the subscription of several communities to a similar set of guiding principles for creating and maintaining social memory.

Keywords

Oman, rescue archaeology, funerary landscape, burialscape, spatial analysis, graves disposal, social memory

West & East 343 Monografie, 4 Sabatino Laurenza, Marco Bianchi, Antonio Di Michele

1. Introduction memory of places leads to the fabrication of a social memory (fig. 1). This paper presents the results of the rescue exca- vations1 carried out between May 2014 and June 2015 in the Sultanate of Oman along the Pack- 2. Geographical background age 52 of the Batinah Expressway route, a multilane and historical studies of the region highway which will run across the Batinah region from Muscat to the UAE border. Our work started The Batinah region is a coastal gravel plain lying be- as a typical fieldwork technical rescue archaeologi- tween the Hajar mountains to the south-west and cal excavation but at the end, pushed by the huge the coast to the north-east. The area close to the sea archaeological record and information collected, is densely covered with modern cultivations and we developed a new theoretical and methodologi- settlements, instead the route of the Batinah Ex- cal approach to the final interpretation of funerary pressway runs more inland, where cultivable areas landscape in this region. exist only around the wadi beds. The topography is The basic concept of our approach is that characterized by low hills and gravel plains dissected acts and rituals can be more important in iterat- by wadiān valleys, that generally runs from south- ing memory than simply the production of mon- west to north-east. The majority of the excavated uments and objects. Therefore, from an archaeo- graves are located on the ridges or on the slopes of logical point of view, studies and analyses on the the hills overlooking the wadiān valleys. The first field must necessarily swing between the physical- archaeological investigation in the region began in ity of monuments and objects of funerary equip- the 70s, when a Danish team carried out a survey ment (material practices) and the representation locating several prehistoric sites, the most impor- and traces of so-called ritual practices and techno- tant of which was the second millennium graveyard logical know-how in the construction of the funer- of Wadi Suq.3 In the same years De Cardi surveyed ary monuments (immaterial practices) that place the regions of Zahirah, Dakhiliyyah and Sharqiyah, the subjects in a new time/space understanding. touching just the south-eastern of the Batinah4 and Thus, the material and immaterial practices act as it began the first surveys focused on the exploitation cultural indicators, as the containers of a past that of the copper ores of the region,5 with the discovery contribute to the creation of a “cultural memory”. of some mines used in ancient times.6 The explora- Moreover, communities with a well-defined cul- tion on the copper mining continued later with the tural memory tend to shape the landscape accord- accurate fieldworks of the Deutsches Bergbau-Mu- ing to certain characteristics designed to create a seum.7 Anyway, the most important work for the memory of the places to be preserved in the future. Batinah region remains the survey on the hinter- In this way, cultural memory combined with the land of Sohar aimed to the understanding of the agricultural system during the Islamic period.8 During 1 Theexcavation operations were carried out by an Ital- the 90s a survey conducted by Yule and Weisgerber9 ian Team composed by all professionals freelance archaeolo- identified some other sites, touching marginally the gists, anthropologists and surveyors directed on the field by Dr. Sabatino Laurenza, under the scientific direction of the De- Batinah region. partment of Excavations and Archaeological Studies of Minis- 3 try of Heritage and Culture, sponsored and supported by the Frifelt 1975. Ministry of Transport and Communication of the Sultanate of 4 de Cardi et Al. 1976; de Cardi et Al. 1977; Doe Oman. 1977. 2 The Batinah Express highway is a 265 km length road 5 Goettler et Al. 1976. (41 km is the length of Package 5); the project was divided in 6 Hastings et Al. 1975. 6 packages with different consultants and contractors for each 7 Weisgerber 1978; Hauptmann 1985. package. The archaeological works involved different teams of 8 different nationalities, under the Supervision of the Director of Costa, Wilkinson 1987. Dept. of Excavations and Archaeological Studies, Dr. Sultan Al 9 Yule, Kervran 1993; Yule 2001; Yule, Weisger- Bakri. ber 1998.

West & East 344 Monografie, 4 Graves, distribution and social memory: towards a new definition of funerary landscape in Oman

Figure 1 First original sketch of the diagram of “fabrication of ancient landscape” process

In recent years a team of the Sultan Qaboos Wadi al Jizzi region.12 The works for the Batinah University investigated the southern end of the Expressway Package 5 where our team worked was Batinah,10 while a German-Omani team worked in long 41 km and it begins roughly 18 km inland from Wadi Bani ‘Awf,11 and a Dutch team surveyed the the coast after Wadi Haibi in thewilayat of Sohar and ends in the wilayat of Liwa. Here we identified and excavated three graveyards, Wadi al Arad, Falaj 10 al Jahwari, El Mahi 2007; al Belushi, El Mahi as Souq and Liwa (fig. 2). 2008. 11 Häser 2000, 2003; Ribbeck, Gangler 2001. 12 Düring, Olijdam 2015.

West & East 345 Monografie, 4 Sabatino Laurenza, Marco Bianchi, Antonio Di Michele

Figure 2 The three graveyards along the Batinah Expressway Package 5 alignment

3. The graveyards of about 168 graves with only one type of grave, so- called ‘tower-shaped grave’. The structure has an oval or sub-rectangular plan, and a trapezoidal outline 3.1. Wadi al Arad with the sides slightly tapered and the roof flat. The grave can be divided in three distinct architectural The graveyard of Wadi al Arad is located southwest features: the external wall, made with big large stones of Sohar, on two low hills to the north and south tightly stacked, a filling wall composed of small stones of the homonymous wadi, that is part of the basin and sand, and the burial’s chamber wall made with of Wadi al Jizi. The archaeological evidence consists elongated and flatted face stones. This type of grave

West & East 346 Monografie, 4 Graves, distribution and social memory: towards a new definition of funerary landscape in Oman

Figure 3a Type of graves from Wadi Al Arad graveyard

Figure 3b The lower limbs of the buried corpse arranged in a “rhomboid shape”

has an average height of 1.5 m, an external diameter lower limbs arranged in a “rhomboid shape”, with the of 4 × 3 m, and a burial’s chamber of 2.2 × 0.7 m. and knees placed below the first row of the stones of the it was disposed along the slopes of the hill on both the burial’s chamber wall, clue to the originality of that sides of the wadi, isolated or in cluster of two, three position fig. 3b). Moreover, a constant symmetry ob�- or more graves (fig. 3a). The anthropological remains served in the position of the bones suggest us the in- are very poor due to the arid climate and the action of tentionality of the deposition. the looters. We identified 128 individuals, determin- Another peculiar feature of Wadi al Arad grave- ing the sex of 109 skeletons (39 men and 70 women) yard is that at least 50 individual of both sexes lack and estimating the age of 124 individuals (101 adults, of the upper part of the body, with the complete re- 10 young and 13 children). The data of the burial moval of the bones from the grave or with the trans- practice showed us 41 graves with a single body, 27 lation of certain bones of the upper part of the body with two bodies, 14 with three, 2 with four, and 1 in another side of the chamber. Probably it was due with seven. The corpse mostly laid out in supine po- to the action of the looters or to any unknown fu- sition with the legs stretched and 14 of them had the nerary rituals.

West & East 347 Monografie, 4 Sabatino Laurenza, Marco Bianchi, Antonio Di Michele

Figure 4 Relevant materials from the three graveyards, testifying the reuse of the graves in later periods

In Wadi al Arad graveyard only 61 excavated pared with some finds from Qasr-i Abu Nasr14 or, for graves have yielded at least one object of the funer- a later dating, with the one recovered in the fortress ary assemblage, the most common finds founded of Ashdod Yam:15 Moreover, we found four Sasanian are small beads of different materials (mostly carne- stamp seals of different materials from three different lian and shells) found in 45 tombs alone or togeth- graves (fig. 4: f). In one grave there is a seal of hema- er with other objects. Other common categories of tite with a crouching rabbit engraved, and another objects are the bronze rings found in 25 graves and seal of garnet stone with a monogram incised. From several fragmented glass vessels found in 13 graves, other two graves it comes a bronze seal with the draw which could give us some clue for the dating of the of a humped bull, and a basalt seal with a human fig- burials. A fragmentary dark fine glass vessel found in ure engraved. All the seals could be dated, for shape the grave with an elongated shape (fig. 4: b) has some and style, to the Sasanian period,16 more precisely to comparison in the Syro-Palestinian area during the the late fifth to half sixth centuries.17 late third to the fourth centuries13 or maybe later. Another miniature jar of fine dark glass can be com- 14 Whitcomb 1985, fig. 58d. 15 Ouahnoua 2014, p. 105, fig. 28.18. 16 Gyselen 1993; Osten 1931. 13 Hayes 1975, p. 207, n. 321. 17 Kutterer et Al. 2015, pp. 43-54.

West & East 348 Monografie, 4 Graves, distribution and social memory: towards a new definition of funerary landscape in Oman

3.2. Falaj as Souq and a max height of 1.5 m. The entrance was recognized in all the graves except for one sample, 14 The graveyard of Falaj as Souq is located northwest oriented to S, 6 to E, 3 to SE, and 1 to NE, W and of Sohar, on top of a hillock high 135 m above the SW. It’s quite astonishing that those two types of sea level, within the Wadi Souq basin, at the fall of graves are built in the same place but have a pret- Wadi al Jizi catchment basin. Moreover, the grave- ty clear different orientation, probably due to the yard is situated on one of the most important natu- fact that they are from different chronological pe- ral routes connecting the UAE with Oman through riods. Although few of them are isolated, 21 are ag- the Hajar al Gharbi mountain range. Our team has gregated in seven clusters made up from two to six surveyed about 250 archaeological features – most- graves. However, in four cases we have the occur- ly funerary cairns – and excavated 59 graves and 9 rence of both the types of graves; two groups have stone structures that were directly interested by the three ‘Horseshoe-shaped’ graves, oriented N-S, destruction for the highway project. The graves can and one ‘Dome-shaped’ grave, oriented E-W; one be grouped in two distinct typologies due their ar- group has two ‘Horseshoe-shaped’ graves, orient- chitectural features: the ‘Dome-shaped’ graves and ed NW-SE and one ‘Dome-shaped’ grave, orient- the ‘Horseshoe-shaped’ graves (fig. 5). The ‘Dome- ed E-W. Finally, there is an isolated group with shaped’ grave shows a circular plan and it was built an ‘Horseshoe-shaped’ grave, oriented N-S with a directly onto the surface, without foundation, with ‘Dome-shaped’ grave, oriented SW/NE. Archae- local stones of different sizes. Viewed by the out- ological investigation has find out clearly that the side, the structure is a mound of stones irregularly ‘Horseshoe-shaped’ graves were constructed above stacked together and the only detectable feature is the ‘Dome-shaped’, suggesting a clear chronological the wall of the burial’s chamber made with elongat- succession between them, indicating also in some ed stones put in three/four courses. cases that stones from ‘Dome-shaped’ graves were This type of grave has an average size of 5.3 × 4.8 recovered to build up the later ‘Horseshoe-shape’ m of external diameter and a burial chamber of graves.The anthropological data for the necropo- 2.7 × 1.1 m, with a height, where preserved, reach- lis of Falaj as Souq are very poor as the human re- ing 2.5 m. In some of the graves it was possible to mains were almost entirely compromised due to the detect the entrance to the burial’s chamber, which acidity of the soil and the action of the looters. Of in 14 graves was oriented to SW, in 10 graves to W the 49 subjects identified, we were able to determi- and just in 2 graves looking to S, practically in most- nate the sex of just six individuals, all female. Also, ly of the graves the entrance was directed toward the we have estimated the age of eleven individuals: 6 Wadi Souq. Those type of graves are always isolat- adults, 2 young, 1 child, 2 infants. The data are poor ed and never aggregated in a cluster, despite of the also for the burial practice with the identifications ‘Horseshoe-shaped’ graves, as we will see further on. of a double burial just in four graves, two ‘Horse- The ‘Horseshoe-shaped’ graves have an apse plan shoe-shaped’ and two ‘Dome-shaped’. The excava- with a structure characterized by an external wall, a tions of this graveyard yielded comparatively few filling wall, and a chamber wall and it was built di- finds: just 33 of the graves had at least one object. rectly onto the surface, without a foundation or a The most abundant finds were the small beads found platform. Those three features were made with local in 16 graves and made of different materials (carne- unworked stones of the same size, and just for the lian, shells, agate and rock crystal). Bronze objects chamber wall elongated and flatted face stones were were found in 14 tombs, in two ‘Horseshoe-shaped’ selected. When preserved, the top is covered with graves we found 10 small bronze rings, while, in the flat stones and gravel, a lintel made with a large flat ‘Dome-shaped’ graves we found 10 pins, 1 earring, stone also surmounted the entrance. The ‘Horse- 1 hook, 1 ring and 1 nail. Fragmentary glass vessels shoe-shaped’ graves are smaller than the ‘Dome- were found in just 6 graves; a light green fine glass shaped’, with an average of the external diameter flask with pentagonal globular body, straight neck of 4 × 2.75 m, a funerary chamber of 2.3 × 0.7 m, and everted rim with rounded lip could have some

West & East 349 Monografie, 4 Sabatino Laurenza, Marco Bianchi, Antonio Di Michele

Figure 5a Dome-shaped types of graves from Falaj as Souq graveyard

Figure 5b Horseshoe-shaped types of graves from Falaj as Souq graveyard

comparisons with the material from the Syro-Pales- Finally, there is a light blue thin glass with outplayed tinian area of the fourth centuries (fig. 4: a). 18 Two rim and rounded lips, probably belonging to a small flat rims of two blue glassunguentaria with long cyl- pear-shaped bottle (fig. 4: c). Similar bottles occur inder and flared neck have some comparison from in the sites of Veh Ardashir and Nineveh during the the site of Abu Skhair during the second half of the Early and the Middle Sasanian period (third-fifth third century and beginning of the fourth century.19 cent. AD). 20

18 Hayes 1975, p. 207, n. 320; Whitehouse 1997, p. 179, n. 315. p. 220, fig. 20.19.3. 19 Negro Ponzi 1972, pp. 215-237; Simpson 2014, 20 Simpson 2005, p. 147, fig. 1.11.

West & East 350 Monografie, 4 Graves, distribution and social memory: towards a new definition of funerary landscape in Oman

3.3. Liwa this grave shows also a kind of monumental entrance flanked by an angular jamb and closed with The graveyard of Liwa is located west of the homon- some big stones irregularly packed (fig. 6). ymous town, along the Wadi Bani Umar al Gharbi In Liwa graveyard the anthropological data are basin, and it is concentrated on two hills high 100 m very poor, identifying only 83 individuals (65 adults, above the sea level. The valley between the two hills 1 young and 4 infants), with 12 males and 18 fe- is occupied by a huge dump. The northern hill is a males between the adults. For the burial practice we wide plateau with 25 excavated graves; 19 are on the have 29 graves with a single burial, 13 with two, 4 top and 6 along the southern slope. The southern with three, 2 with four and 1 with six individuals. hill has been modelled in a pyramidal shape by the The ‘Dome-shaped’ grave 644 presents three adults, works for the near dump and here the Italian team one male, one female and one of indeterminate sex, has excavated totally 29 graves, 9 of them placed on and one infant; instead the grave 650 of the same ty- the top, and the others along the western slopes. pology has only one adult female. The dominant typology defined ‘Tumuli-shaped From the 54 graves of Liwa, only 28 has at least graves’ has an oval or circular plan made up with big one funerary object; the most represented funer- irregular stones stacked together like the ‘Dome- ary good are beads of different materials (found in shaped’ graves of Falaj as Souq. Otherwise, this type shows a wide range of adaptation to the surface on which grave was build up. As we can see, the graves built on the slope of the hill follow the inclination of the ground, with a tumulus of elliptical shape with a narrowest funerary chamber. For this reason, the measures of the external perimeter of those graves range from 6 × 5 to 2.5 × 1.35 m and the burial chamber from 2.2 × 0.8 m to 1.5 × 0.43 m. More- over, this type has a more irregular plan than the ‘Dome-shaped’ graves of Falaj as Souq, and that’s why we have preferred to distinct the two types. In this type of grave we did not find an entrance, but the chamber’s orientation is often E-W (21 graves) and the remaining graves were oriented N-S (7), SW-NE (6) and SE-NW (5). The ‘Tumuli-shaped graves’ are found both isolated and in clusters of two, three or five tombs. In Liwa we have also two examples of another type of grave, the graves 644 and 650, and they seem very similar at the ‘Dome- shaped’ graves found in Falaj as Souq. Both have a circular plan, but the grave 644 has a diameter of 7.17 × 6.95 m and the chamber of 4 × 1.69 m, instead the 650 has the diameter of 6.8 × 6.32 and the chamber of 4 × 1.45 m, as we can see they are bigger than those of Falaj as Souq. Another shared characteristic is that they were built near the edge of the northern hill, suggesting a high visibility on the surrounding landscape. The grave 644 is flanked on its east, west and south side by three ‘Tumuli-shaped Figure 6 graves’ that are stratigraphically later. Furthermore, Tumuli-shaped type of graves from Liwa grave yard

West & East 351 Monografie, 4 Sabatino Laurenza, Marco Bianchi, Antonio Di Michele

20 graves), some bronze rings in 5 graves and some sula, as from Wadi al-Qawr,29 90 km north to Sohar, fragmented glass vessels found in 7 graves. The frag- Jebel Buhais30 and Samad,31 all dated for the materials mentary rim with a folded-in lip and a conical neck to the Wadi Suq Period. If this is the case, also this can be dated to the Late Sasanian period with some type should be reused for the presence of glass vessels influence by Syrian productions.21 Another frag- in some graves, as pointed out earlier. The ‘Tumuli’ mented bottle of dark glass can be compared with of Liwa are more difficult to understand due to the some finds from Nippur.22 In another grave, we high variety of its forms; however, some comparison found a flagon jug characterized by a spiral thread of can be traced in Jebel Buhais,32 where a grave has the plain glass wound loosely round shoulder and low- same structure and it is dated to the Wadi Suq peri- er part of neck, dated to the Early Sasanian period 23 od. Another quite similar grave dated to the same pe- (fig. 4: d). Finally, in the ‘Dome-shaped’ tomb G644 riod was recently found from the anglo-omani exca- archaeologists recovered a Sasanian coin of Hormizd vations of the Packages 3-4 in Batinah.33 But there are II (303 – 309 AD) in which is represented: D/ a bust also several similar graves excavated in Samad dated right, wearing an eagle crown with korymbos; R/ fire to the Iron Age II-III (1000 – 250 BC) for the mate- altar with ribbon and bust left in flames flanked by rials found.34 ‘Tumuli’ type shows clearly a reuse also two attendants both with korymbos (fig. 4: e). for the Sasanian glasses found in some graves. The ‘Tower-shaped graves’ of Wadi al Arad share parallels with a number of structures recorded across the 3.4. Chronological Remarks Oman peninsula. The closest matches are the graves excavated by Frifelt in Wadi Jizzi, in northern al- With this few elements it is very difficult to suggest Batinah, dated to the Late Iron Age.35 Another com- a chronological sequence for the three graveyards; parison can be made with some graves excavated by however, some suggestions can be made considering De Cardi in Kalba, in the coastal Sharjah, also dated the architecture, measurements, orientation, building to the Iron Age.36 German teams have called this type techniques and some findings. The ‘Dome-shaped’ ‘hut-graves’ (Kastengräber) and found them in differ- graves typology is very common in the Batinah24 and ent regions of Oman dating to the Iron Age II-III37, across the Oman, where is dated to the Hafit period; or also reused in the Late Iron Age.38 in our cases, those from Falaj as Souq, with a simpler Considering those data, the graves of Wadi structure than the two from Liwa, are similar to graves al Arad can be dated to the Iron Age, but thanks find in Kalba,25 HD-1026 and Jebel Buhais.27 Further- to the comparison of the glass vessels and for the more, due to the Sasanian coin from Liwa and the stamp seals, it is possible determinate a reuse also glasses fragmentary vessels found there and in Falaj for this graveyard at least until the Sasanian period as Souq, it is strongly possible to suggest a reuse of the (Tab. 1 and 2). graves also thousands of years later. Besides, as seen before, the ‘Dome-shaped’ graves are stratigraphically fig. 18. earlier than the ‘Horseshoe-shaped’ tombs of Falaj as 29 Phillips 1997, p. 209. Souq and the ‘Tumuli-shaped graves’ of Liwa. For the 30 Uerpman et Al. 2006, pp. 25-27. ‘Horseshoe-shaped’ graves there are some few exam- 31 Yule, Kazenwadel 1993, p. 254. 28 ples in Batinah and many across the Oman penin- 32 Uerpman et Al. 2006, pp. 27-29. 33 Saunders 2016, pp. 126, fig. 279. 21 Lamm 1931, p. 361. 34 Yule 2001, p. 237, p. 245, p. 252. 22 Meyer 1996, p. 248, n. 20. 35 Frifelt 1975, p. 373, fig. 62. 23 Hayes 1975, p. 213, n. 436. 36 De Cardi 1975, p. 22. 24 Saunders 2016, type 1, 9, fig. 8. 37 Weisgerber 1981, pp. 224-225; Yule, Weisger- 25 Eddisford, Phillips 2009, p. 101, fig. 3. ber 1988, p. 14; Yule, Kazenwadel 1993, p. 254; Yule et 26 Salvatori 2001. Al. 1994, pp. 396-398. 27 Jasim 2012, p. 173, fig. 143, 157, figs. 190-191. 38 Yule 2001, pp. 396-397; Yule, Weisgerber 1988, 28 Frifelt 1975, p. 392; Saunders 2016, type 6, 14, p. 210.

West & East 352 Monografie, 4 Graves, distribution and social memory: towards a new definition of funerary landscape in Oman

Table 1 Measurements comparison between the different types of graves

Table 2 Hypothesis of chronological timetable of the three graveyards in the Batinah Region

4. The “fabrication of the ancient evidence coming out from excavations to be con- landscape”: from the recognizing nected with immaterial practices was the building of the material and immaterial practices techniques of the graves, considering and focusing on all a set of details linked to the building tech- At the end of the fieldworks we had in our hands niques of the funerary monuments. From the data almost 300 excavated graves, with a lot of informa- of the excavation of grave 21 of Falaj as Souq (made tion regarding the physicality of monuments and in a more detailed way, dismantling the rows of objects of funerary equipment (material practices) stones one by one) we arrived to elaborate the build- and the necessity to individuate the representation ing technique of those monuments understanding and traces of so-called ritual practices and techno- that behind such a physical and hard work there is logical know-how (intended as a specialized crafts- the hand of skilled workers able firstly to draw on manship transmitted by “master” to “apprentice”) the ground the base plan of the funerary monument in the construction of the funerary monuments and to choose accurately the stones for size, mate- (immaterial practices) that place the subjects in a rial and shape. The master was helped by at least 4 new time/space understanding: trajectories that skilled workers (number determined by the need to fuse the past, the present and the future. The first build up the at the same time the walls of the burial

West & East 353 Monografie, 4 Sabatino Laurenza, Marco Bianchi, Antonio Di Michele

chamber and the outer one, in order to guarantee wadi valleys, a kind of topographical markers along its stability) during the building of the monument the natural routes of the country, their distribution with the help of several workers involved in the car- along the main natural routes, represented by the riage and the preparation of the stones. ancient wadi beds, hypothesis confirmed by an ar- Another important aspect regarding the imma- ticle of 1925 of an English traveler Sir Percy Cox terial practices was the evidence of the reuse of such published on Geographical Journal39 and also the monuments. The practice of reuse regards in reality fact that all the graveyards were placed at an almost almost all the funerary monuments in Oman but in similar distance of 12/13 Km between the coast line the specific case of Falaj as Souq we have seen that and the main ore mines placed under the mountains the reuse regards not only the single monument but of Hajar, the mine of Arja for the graveyard of Wadi the place where funerary monuments were placed. Al Arad, Lasail for the graveyard of Falaj as Souq From the excavations at Falaj as Souq we show that and the mine of Fizh for the graveyard of Liwa, as that ‘Horseshoe-shaped’ graves (surely of a later pe- seen from the study of R. G. Coleman and E. H. riod) were built up on largest ‘Dome-shaped’ graves Bailey for U.S. Geological Survey of 1974.40 (of an earlier period), in some case reusing the same stones of the earlier graves for their building and removing also the burial bones in a closer exter- 5. Conclusions nal round pit grave. The reuse of such monuments was clear to us also observing the disposition of the Our work allows us to identify that process for bodies inside some graves besides the discovery of what communities with a well-defined cultural materials of later periods, as seen before in Wadi memory tend to shape the landscape according to al Arad graveyard. At this point, we arrived to de- certain characteristics, designed to create a memo- fine the material and immaterial practices, a set of ry of the places to be preserved in the future. So, in cultural indicators considered as containers of that such a way the cultural memory combined with the past which contributes to the production of cultur- memory of places leads to the construction of a so- al memory of a community. But this cultural mem- cial memory: I bury my dead in certain monuments ory is displaced in a precise and well-defined space, and with certain rituals that involve certain goods as we saw looking at the disposition of the graves in (cultural memory); I do all this in a given landscape, the three graveyards. In fact, the hugest graves of Fa- whose characteristics become for my group guiding laj as Souq and of Liwa were displaced all on the top principles to create and maintain memory and so- of the hills, besides the graves of Wadi al Arad and cial memory (fig. 7). the smallest of Liwa were placed all along the slopes So, our role as archaeologists must be simultane- above the wadi valley, with their entrance directed ously to be able to explore the moments and ten- towards the wadi beds. This allow us to think that sions between the memory of material and imma- communities with a well-defined cultural memory terial work. Therefore, we need to stimulate our tend to shape the landscape according certain char- method of doing research and digging, expanding acteristics designed to create a memory of the places our ability to read archaeological contexts, being to be preserved in the future. In our case clear indi- able to perceive and identify the so-called immate- cators of it were the spatial distribution of the fu- rial practices from the material data that the excava- nerary monuments, well visible from down of the tion gives us back.

39 Cox 1925. 40 Coleman, Bailey 1974.

West & East 354 Monografie, 4 Graves, distribution and social memory: towards a new definition of funerary landscape in Oman

Figure 7 The fabrication of ancient landscape chart

West & East 355 Monografie, 4 Sabatino Laurenza, Marco Bianchi, Antonio Di Michele

BIBLIOGRAPHY

Antonaras A. C. 2010, Early Christian and Byzantine Häser J. 2000, Formation and transformation processes glass vessels: forms and uses, in: Daim F., Drauschke of oasis settlements in the Sultanate of Oman: prelimi- J. (eds.), Byzanz-das Römerreich im Mittelalter, 1: nary report on a new field project, «Proceedings of the Welt der Ideen. Welt der Dinge, Mainz, pp. 383-430. Seminar for Arabian Studies» 30, pp. 115-118. al Belushi M. A., El Mahi A. T. 2008, The second Häser J. 2003, Archaeological results of the 1999 and season of 2006 excavations at the site of Bawsher, 2000 survey campaigns in Wadi Bani Awf and the re- Sultanate of Oman: the role of location and natural gion of al-Hamra (Central Oman), «Proceedings of resources in the inception settlement, «Adumatu» 18, the Seminar for Arabian Studies» 33, pp. 21-30. pp. 47-80. Hastings A., Humphries J. H., Meadow R. H. Cleziou S., Tosi M. 2007, In the Shadow of the 1975, Oman in the third millennium BCE, «Journal Ancestors, Muscat. of Oman Studies» 1, pp. 9-55. Coleman R. G., Bailey E. H. 1981, Mineral depo- Hauptmann A. 1985, 5000 Jahre Kupfer in Oman. sits and geology of northern Oman as of 1974 (U.S. 1. Die Entwicklung der Kupfermetallurgievom 3. Geological Survey Report 81), Muscat, pp. 39- 54. Jahrtausend bis zur Neuzeit, Bochum. Costa P. M., Wilkinson T. J. 1987, The hinterland of Hayes J. 1975, Roman and Pre-Roman glass in the Royal Sohar: archaeological surveys and excavations within Ontario Museum, Toronto. the region of an Omani seafaring city, «Journal of al Jahwari N. S., El Mahi A. 2007, The geography of Oman Studies» 9, pp. 1-238. location and topos-culture. Results of excavations at Cox P. Sir 1925, Some excursions in Oman, Bausher Site, Sultanate of Oman, «Adumatu» 15, «Geographical Journal» 66, pp. 195-227. pp. 7-34. De Cardi B., Collier S., Doe D. B. 1976, Excavations Jasim S. A. 2012, The necropolis of Jebel al-Buhais. and survey in Oman, 1974-1975, «Journal of Oman Prehistoric discoveries in the Emirate of Sharjah, Studies» 2, pp. 101-187. United Arab Emirates, Sharjah, pp. 43-54. De Cardi B., Doe D. B., Roskams S. P. 1977, Kutterer A., Jasim S. A., Yousif E. 2015, Buried Excavations and survey in the Sharqiyah, Oman, far from home: Sasanian graves at Jebel al-Emei- 1976, «Journal of Oman Studies» 3, pp. 17-33. lah (Sharjah, UAE), «Arabian Archaeology and Doe D.B. 1977, Gazetteer of sites in Oman, 1976, Epigraphy» 26, pp. 43-54. «Journal of Oman Studies» 3, pp. 35-38. Lamm C. J. 1931, Les verres trouvés à Suse, «Syria» 12, Düring B., Olijdam E. 2015, Revisiting the Suhar pp. 358-367. hinterland: the Wadi al-Jizi Archaeological Project, Meyer C. 1996, Sasanian and Islamic glass from Nippur, «Proceedings of the Seminar for Arabian Studies» Iraq (Annales du 13e congrès de l’Association 45, pp. 93-106. Internationale pour l’Histoire du Verre), Lochem, Eddisford D., Phillips C. S. 2009, Kalba in the pp. 247-255. third millennium (Emirate of Sharjah, UAE), Negro Ponzi M. 1972, Glassware from Abu Skhair, «Proceedings of the Seminar for Arabian Studies» «Mesopotamia» 7, pp. 293-384. 39, pp. 111-124. Ouahnoua B. 2014, The Glass Vessels from Ashdod Yam Frifelt K. 1975, On prehistoric settlement and chro- Citadel, in: Raphael S. K. (ed.), Azdud (Ashdod nology of the Oman peninsula, «East & West» 25, Yam). An Early Islamic Fortress on the Mediterranean pp. 359-383. Coast, Oxford, pp. 64-111. Goettler G., Firth N., Huston C. C. 1976, A pre- Osten H.H. von der 1931, The Ancient Seals of the Near liminary discussion of ancient mining in the Sultanate East in the Metropolitan Museum, Old and Middle of Oman, «Journal of Oman Studies» 2, pp. 43-55. Persian Seals, «Art Bulletin» 13, pp. 221-241. Gyselen R. 1993, Catalogue de sceaux, camées et bulles sassanides. 1. Collection générale, Paris.

West & East 356 Monografie, 4 Graves, distribution and social memory: towards a new definition of funerary landscape in Oman

Phillips C. S. 1997, The pattern of settlement in the Whitcomb D. S. 1985, Before the Roses and Wadi al-Qawr, «Proceedings of the Seminar for Nightingales. Excavations at Qasr-i Abu Nasr, Old Arabian Studies» 27, pp. 205-218. Shiraz, New York. Ribbeck E., Gangler A. 2001, Oasis Settlement in Whitehouse D. 1997, Roman Glass in The Corning Oman. Pilot Study 1999-2000, Stuttgart. Museum of Glass Volume Three, New York. Salvatori S. 2001, Excavations at the funerary struc- Yule P., Weisgerber G. 1988, Samad ash-Shan: pre- tures HD-10-3.1, 3.2, 4.1, 4.2 and 2.1 at Ra’s al-Hadd liminary report 1988 (Unpublished report submitted (Sultanate of Oman), «Rassegna di Archeologia» 25, to the Ministry of Heritage and Culture), Bochum. pp. 67-77. Yule P., Kazenwadel B. 1993, Toward a chronology Saunders B., 2016, Archaeological Rescue Excavations of the Late Iron Age in the Sultanate of Oman, on Packages 3 and 4 of the Batinah Expressway, in: Finkbeiner U., Boucharlat R. (eds.), Sultanate of Oman, Oxford. Materialien zur Archaologie der Seleukiden und Simpson S. J. 2005, Sasanian glass from Nineveh, Partherzeit im südlichen Babylonien und im (Annales du 16e congrès de l’Association Internatio Gulfgebiet, Tübingen, pp. 251-277. nale pour l’Histoire du Verre), Lochem, pp. 146-151. Yule P., Kervran M. 1993, More than Samad in Simpson S. J. 2014, Sasanian glass: an overview, in: Oman: Iron Age pottery from Suhar and Khor Keller D., Price A. J., Jackson C. M. (eds.), Rori, «Arabian Archaeology and Epigraphy» 4, Neighbours and Successors of Rome: Traditions of Glass pp. 69-106. Production and Use in Europe and the Middle East in Yule P., Weisgerber G., Kunter M., Bemmann the Later 1st Millennium AD, Oxford, pp. 200-231. M. 1994, Wadi Suq burial structures in Oman, Uerpmann H-P., Uerpmann M., Jasim S. A. 2006, «Nubica» 3, pp. 379-415. Evidence of ancient mining sites in Oman: a prelimi- Yule P., Weisgerber G. 1998, Prehistoric tower tombs nary report, «Journal of Oman Studies» 4, pp. 15-28. at Shir/Jaylah, Sultanate of Oman, «Beiträge zur Weisgerber G. 1978, Evidence of ancient mining sites Allgemeinen und Vergleichenden Archäologie» 18, in Oman: a preliminary report, «Journal of Oman pp. 183-241. Studies» 4, pp. 15-28. Yule P. 2001, Die Gräberfelder in Samad al-Shan Weisgerber G. 1981, Mehr als kupfer in Oman: ergeb-� (Sultanat Oman): Materialien zu einer Kultur- nisse der expedition 1981, «Der Anschnitt» 5, geschichte, Rahden. pp. 174-263.

West & East 357 Monografie, 4

Decorated shell discs from Sumhuram, Oman

SILVIA LISCHI

University of Pisa

Abstract

During the excavations of Sumhuram (1997-2018) a large number of findings were brought to light. An interesting and enigmatic group of objects, composed of 18 big shells that have particular decorative motifs, are particularly interesting for understanding the interconnection between Arabia Felix and the Near East. In fact, the objects and the iconography had no comparisons in south-western Arabia, but there are close comparisons from Iran and some similar objects from the Gulf area and the north of Oman. In order to understand if these objects were imported or belonged to a local production, they were analytically studied. Considering the few comparisons at the current state of research, it is difficult to establish their origin with certainly. It is possible, however, to say that close relations with Iran are present. This emphasizes the uniqueness of the city of Sum- huram also regarding its artistic production, which turns out to be a reinterpretation of models from different cultures and places.

Keywords

Pre-Islamic Arabia, decorated shell disc, Dhofar, Sumhuram, Indian Ocean trade, pomegranate, South Arabian Kingdoms

West & East 359 Monografie, 4 Silvia Lischi

1. Introduction the variety of their decorations. Some discs have already been published in the two volumes edited by In this paper I will present a group of 18 objects A. Avanzini5 while the discs that refer to the seasons that we found in the city of Sumhuram (300 BC – between 2004 and 2018 are discussed here for the AD 500). The city located in the area of Khor Rori first time. The only exception is the poster by Lischi (fig. 1) is the most important pre-Islamic settlement which summarizes the disc assemblage up to 2015.6 in the Dhofar region, southern Oman. The city was The study of these objects still is marginal within ar- founded in a geographically strategic place for the chaeological studies, such as beads and other orna- activity of the most important Indian Ocean trade mental materials.7 routes of the time. The lagoon in front of the city is The miniaturist characters of the decorations al- a perfect natural port, while behind it lies the Nejd low us to think that they were part of a corpus of region, where the best quality of frankincense was, luxury objects.8 The study of this type of materials, and still is, produced.1 neglected until most recently, is becoming increas- Since 1997 the Italian Mission to Oman, di- ingly important as it allows us to identify and un- rected by Prof. Alessandra Avanzini2 of the Univer- derstand a cross-section of the society of that pe- sity of Pisa in collaboration with the Office of the riod otherwise difficult to identify. Moreover, the Adviser to His Majesty the Sultan for Cultural Af- uniqueness of the materials and the techniques re- fairs, have been working in Sumhuram. During the quire highly skilled artisans who could operate in archaeological excavations a large number of find- well-defined centers, and with the necessary pre- ings were unearthed.3 These were indispensable for cautions identifiable or work in groups of itinerant understanding the activities of the pre-Islamic cities workers with very precise origins. The study of these along the coast of Arabia. Among these, of particu- characteristics allows us to identify the social role of lar interest are a group composed of eighteen orna- these objects with the aim of understanding the an- mental objects made from big shells that have par- cient economic models.9 ticular decorative motifs. What is extremely interesting is the lack of sim- This class of objects was almost unknown and in ilar findings in south-western Arabia, such as the the few known cases these artefacts were variously chronological incongruity that is delineated be- defined as buttons, medallions, pendants, discs or tween the materials found in Sumhuram and those bosses. 4 Attribution of the part of the shell used is present in the Emirates, in Saudi Arabia and in the still uncertain since the handiwork is so accurate north of Oman.10 It is probably difficult to obtain that there remains little evidence of the characteris- precise chronological indications from these arte- tics of the raw material. facts as they often have prolonged use over time, but This study aims to present an analysis of this it is possible to highlight typical patterns of a period group of objects, which represent a significant or characteristic trends of an area. group in terms of quantity and are interesting for

1 Albright 1982; Avanzini, Sedov 2005; Avanzi- ni 2008, 2011, 2014; Pavan, Sedov 2008. 2 My heartfelt thanks to Professor Alessandra Avanzi- 5 Lombardi 2002, pp. 168-169, cat. 121; Lombardi ni, director of the Archaeological Mission at Sumhuram, Ital- et Al. 2008, p. 403. ian Mission to Oman, University of Pisa, for making this study 6 Lischi 2015. possible and allowing to me to study these unpublished materi- 7 als that are presented here. Lischi 2018, p. 2; Francis 2007, p. 251. 8 3 An overview of the materials unhearted in Sumhur- Caputo, Genchi 2015. am can be found in the two volumes edited by A. Avanzini in 9 Liverani 2005. 2002 and 2008. 10 Al-Hajri 2006, Pl. 3.11.c; Weeks et Al. 2017, 4 Genchi 2013; Jasim 2012; Degli Esposti 2015; p. 56, fig. 24; Caputo, Genchi 2015; Hartnell, Barker Caputo, Genchi 2015; Lombardi 2002; Lombardi et 1999; Jasim 2012, p. 185, fig. 220; Yule, Kervran 1993, Al. 2008; Weeks et Al. 2017. p. 81, fig. 3:13.

West & East 360 Monografie, 4 Decorated shell discs from Sumhuram, Oman

Figure 1 The map of the area with Sumhuram and other sites involved in the Indian Ocean trades during the Classical period (map. S. Lischi)

2. Materials, techniques and typology 2.2. Production techniques

The shell working tradition developed, in south east- 2.1. Materials ern Arabia and surrounding, since the Neolithic period. This long history of local shell manufacture, This class of objects have a circular shape and is with the availability of the raw materials, makes it produced from a piece of marine gastropod shell. likely that, at least in part, the discs were produced As mentioned before these objects were finely pro- in this area. The shell discs that are found in different cessed, so during fabrication have removed the diag- areas of the ancient Near East, are similar to our but nostic elements necessary to identifying the species are differentiate by the presence of a single central of shell used.11 Recently, the shape and the size of perforation.13 This feature is uncommon in our as- the artefacts, have allow to find two species of shell semblage, which normally show a multiple perfora- which may have been used as raw material to pro- tions drilled on the obverse surface of the objects. Be- duce this objects: Lambis truncate sebae and Pleu- low are presented the drilling techniques. In the end, roploca trapezium.12 Given the variable morphology we can say that the product processing was probably of the archaeological artefacts discussed, it is possi- divided into 3 or 4 phases: the cutting of the shell to ble that have been used of different variety of shells obtain the desired shape, smoothing of the surfaces, found naturally on the shores. perforating to allow for the passage of the wire, engraving of the decorative motif (when present).

11 Moorey 1994, p. 130. 12 Weeks 2017 et Al. 2017, p. 56, fig. 24. 13 Bradl 1984.

West & East 361 Monografie, 4 Silvia Lischi

2.2.1. Drilling techniques position to form an acute angle. If the hole is almost perfectly circular this would suggest the use of a drill The shell discs of Sumhuram have on their rears in a vertical position. The drill points have a medi- from 3 to 4 holes made with multiple perforations um size of 0.3 cm. of one or more drills with probably metal point. It In the assemblage of Sumhuram it is possible to is possible that the points were different in size. The divide the objects with preserved holes in 4 different drills were used with coincident drilling directions typologies based on the number and on the position to create a central hollow space and two or more of the holes. The first type is composed of one circu- channels converging towards the centre. lar central hole that forms a hollow space connected Analysing in depth the technique of drilling with the other three holes arranged in a triangular we note that some holes show an oval enlargement shape. The second is composed of three holes in a which presupposes the use of a drill in an oblique line, the central one is circular and connected with

Table 1 Scheme of the drilling techniques used in the Sumhuram’s assemblage (ph. S. Lischi)

DRILLING TECHNIQUES

TYPE DEFINITION N. OF HOLES SCHEME

H.1 triangularly arranged and connected together 4

H.2 linearly arranged 3

H.3 rectangularly arranged and connected two by two 4

H.4 four-leaf arranged and connected together 4

West & East 362 Monografie, 4 Decorated shell discs from Sumhuram, Oman

the other two. In types three and four, both are tif, the geometric motif, the figurative motif and the composed of four oblique holes. In the third type plain discs. The second can be divided into two sub- the holes are connected two by two, in the fourth types: simple and complex. type the holes are connected together from a hollow The decorations, in some examples, show an space in the middle. A brief overview of these drill- incredible ability used by the crafters. The dimen- ing systems is present in the table 1. sions of the surface to be decorated are normally The use of a wire or string to tie the discs to the very small (the diameter is about 10 cm) and this in- supports are testified by the wear and tear on the volves the use of a “miniaturist” technique. For this external parts of the holes, caused by the action of reason, the tools probably had very small tips. Below something (wire or string) that had been rubbed on is a summary of the typology already mentioned. the surface of the shell.

Type 1 – pomegranate motif 2.2.2. Decoration techniques The seven discs that are included in this type, as the The artefacts show two worked faces. The ob- name suggests, present a complex carved decoration verse is smoothed and shows a reworked natural re- consisting of 6 or less concentric stripes, with differ- cess of the shell (fig. 2). The reverse is often inter- ent and continuous decorative motifs. The promi- spersed with engraving decorations. nent element is the pomegranate-flower motif. An This decoration can be identified by a typology important data is that, in general, the pomegran- composed of four main types: the pomegranate mo- ate is not present in the South Arabian iconogra-

Figure 2 Some examples of shell discs with a well visible reworked natural recess below and some traces of black substances inside the engravings. Clockwise direction: Sh778, Sh574, Sh327, Sh778, Sh376, Sh497 (ph. S. Lischi)

West & East 363 Monografie, 4 Silvia Lischi

Figure 3 Some examples of shell discs with pomegranate decorative motif. The two in the upper part show the pomegranate motif finely worked while the other two show a sketchy pomegranate motif. Clockwise direction: Sh327, Sh569, Sh183, Sh1 (drawings S. Martelli, V. Labianca)

phy, the first and unique representation known is Another important observation is the quality of from Zafar, but it is later (sixth century AD) and the pomegranate decoration in the Sumhuram as- represents an external artistic tradition.14 Further- semblage. There are, in fact, two different groups, more, it is important to remember that the pome- the first with pomegranate finely worked, composed granate seems a non-local motif also in southeastern by four pieces; and the second with pomegranate ex- Arabia. Even if the concentration of this decorative tremely sketchy, composed by three pieces. It is diffi- element in the Persian Gulf region suggests a local cult to propose a clear stratigraphic differentiation production of these objects, the greater familiarity for the city, given the many changes that took place of the Near East with the cultivation of pomegran- over time, it seems clear that the artifacts with fine- ate should not be forgotten.15 ly produced pomegranates are generally attributable to the most ancient period, while the most sketched 14 Yule et Al.2007, p. 535, pl. 37.3. ones have been found in more layers recent. 15 In the Near East there are archaeological and textual In the other stripes there are geometric or styl- evidence that suggests the begin of cultivation of the pomegranate at least to the Bronze Age (Tengberg 2012). ized decorative motifs consisting of triangles, saw-

West & East 364 Monografie, 4 Decorated shell discs from Sumhuram, Oman

tooth motifs, lines, squares and zigzag or stylized racal or Arabian leopard?). It is not clear if there is snakes (fig. 3). a humanoid figure highly stylized near the fracture. But there is a stylized sun which is engraved around the apex of the shell. Among the main figures there Type 2 – geometric decorative motif are small floral symbols (fig. 5).

In Type 2 there can be seen a carved decoration consisting of concentric and parallel stripes with some Type 4 – plain discs geometric signs inside (eight discs). Only one object has geometric motifs disposed in a completely Last but not least, a disc reveals its upper surface different position. In this case the decoration con- without decoration. On the back the holes are pre- sists in simple geometric symbols like triangles and sent. The disc could have been decorated with paint- squares divided by vertical lines in 8 segments, dis- ing, erased over the time, or the disc could be an un- posed in a radial position. As you can see from the finished piece (fig. 6). picture the difference is well visible, for this reason we divided this type in to two subtypes: complex and simple decoration (fig. 4). An important observation regards the engravings that frequently were filled with a dark substance, perhaps bitumen, with the probable intention of bring- Type 3 – figurative decorative motif ing out the decoration from the white surface of the shell. The material, the decorations and the presence This type is composed of one piece with figurative of the holes suggest a decorative use, probably as a decoration. The disk shows a figurative representa- boss of cloth, brooch or belt buckle. The reassumed tion of birds, perhaps birds of prey, and felines (ca- typology is present in the table 2.

Figure 4 In the picture there are visible some shell discs belonging to type 2. The first three objects belong to the group with complex decorative motif, while the last one belongs to the simple decorative motif. From left to right: Sh496, Sh174, Sh376, Sh564 (drawings S. Martelli)

West & East 365 Monografie, 4 Silvia Lischi

Figure 5 The shell disc belongs to type 3, Sh574 (photo S. Lischi, drawing S. Martelli)

Figure 6 The shell disc belongs to type 4, Sh515 (photo S. Lischi)

Table 2 Scheme of the typology of the shell discs discovered in Sumhuram

DECORATIVE TYPOLOGY TYPE DEFINITION SUBTYPE Type 1 Pomegranate decorative motif Simple Type 2 Geometric decorative motif Complex Type 3 Figurative decorative motif Type 4 Plain (probably painted)

West & East 366 Monografie, 4 Decorated shell discs from Sumhuram, Oman

3. Distribution The discovery context, in the above cases, is funerary. The graves where these materials are found At the moment of writing there is a lack of attesta- are dated from the Bronze Age (Wadi Suq peri- tions in the South Arabian Kingdoms of these deco- od), about second millennium BC, to the Iron Age, rative disks. Several comparisons are present in south- about first millennium BC. This information seems eastern Arabia (north of Oman and UAE) and also to suggest a prolonged use of these types of objects. in Saudi Arabia. These comparisons were found in The finest examples of Sumhuram show a series of Wadi Qwar and Fashgha 1, Shimal, Salut, Dibba, concentric stripes with different and continuous Tell Abraq, Jebel al-Buhais, Saruq al-Hadid and Tay- decoration motifs (triangles, lines, squares, saw- ma.16 One such comparison in the Iranian area is an tooth), and in some cases with more complex figura- object collected by The Metropolitan Museum of tive motifs such as pomegranate or stylized snakes. Art of New York attributed to the geographical area The use of pomegranate, fruit associated in antiqui- of Iran during the Iron Age, but unfortunately from ty to the funerary and fertility beliefs (rebirth) and a private collection.17 This seems to support the hy- the stylized snakes normally related to the chtonic pothesis that these kinds of objects come from the sphere are of particular interest and justify their use Mesopotamic area or were under Iranian influence. in funerary contexts.18

Figure 7 Distribution plan of the decorated discs inside Sumhuram (IMTO archive, S. Lischi)

16 Phillips 1987; Vogt, Kastner, Vogt 1987; Hartnell, Barker 1999; Al-Hajri 2006; Jasim 2012; Ca- puto, Genchi 2015; Degli Esposti 2015; Weeks et Al. 2017. 17 It is possible to find the comparison at the following link: https://www.metmuseum.org/art/collection/search/324300. 18 Caputo, Genchi 2015; Schneider 1945.

West & East 367 Monografie, 4 Silvia Lischi

The discovery in a city context, like Sumhuram, Some questions remain open. The use of the pome- underlines the intrinsic value and emphasized the granate decorative element is curious and almost uniqueness of the artistic production of the city due absent in South Arabian iconography. However, it to its role as a trading centre and cultural crossroads. is extremely common in Iran, where this product Looking at the distribution of these objects in- was purchased. This could suggest that these objects side the city of Sumhuram we note that most of were produced in Iran or in the Gulf area and ex- these (eight pieces) were found in Area F. This area ported to Sumhuram. In contrast with this observa- is mainly dedicated to the cult although residen- tion there is the presence of a shell disc with unfin- tial buildings were also found. We observed an in- ished perforations. This would seem to indicate that teresting concentration of ornamental disks in the the production happened, at least in part, inside the square area, called A20, in front of the intra-muros city or that some unfinished objects were traded. temple. Seven pieces come from Area A. This area In any case the detailed decoration and iconogra- is generally considered to be a residential area and phy gives these objects a great intrinsic value that there are small production structures, probably for appears to justify the long life of these discs. Also family use. There is a particular concentration of the trading and use in different contexts are com- two disks with the pomegranate decorative motif prehensible. Furthermore, the discovery of discs in a pit inside room A123. This room is included in with different qualities of pomegranate decoration a building complex where there are evident traces can suggest that there was a reinterpretation of an of structures connected with production activities. ideal model, which as it moved further away from In Area H we found only two disks with different the initial concept, resulted in making the decora- kinds of decoration and in Area G we found only tive system ever more abstract. This may also indi- one shell disc. cate a change in the production centres of these ob- A high frequency of findings of these objects in jects, which then reinterpret the model according the streets of the city or in the open areas has been to local styles. Taking a look at the objects found noted: 7 in Area F and 2 in Area A. Corresponding in the sites mentioned in the article we realize that approximately to 60% of the total. This observation the discs found in the most ancient contexts have probably confirms the hypothesis that these objects an inlay or strongly figurative decoration while over were used for the decoration of clothes. It is possible time the inlay is lost and the figures become more that they have since been worn away, dismembered and more geometric. and lost during the normal life of the city (fig. 7). Hopefully in the future we will be able to collect more information about these discs, to understand the places of production and origin of such sophis- 4. Conclusions ticated objects and have an ever more complete idea of the chronology and movement of materials during In conclusion these objects seem to be unique and the Bronze Age and the Iron Age along the Arabic with very few comparisons in the Middle East. coast; forever the crossroads between east and west.

Acknowledgements

I am particularly grateful to Dr Alexia Pavan for her invaluable comments, suggestions, and discussions on the subject.

West & East 368 Monografie, 4 Decorated shell discs from Sumhuram, Oman

BIBLIOGRAPHY

Albright F.P. 1982, The American archaeological ex- Jasim S.A. 2012, The necropolis of Jebel al- Buhais: pedition in Dhofar, Oman, 1952–1953, (American Prehistoric discoveries in the Emirate of Sharjah, Foundation for the Study of Man, 6), Washington United Arab Emirates, Sharjah. DC. Kastner J.M., Vogt B. 1987, The ring-chambered tomb Al-Hajri M.Y. 2006, Brief preliminary report on the SH99, in: Vogt B., Frank-Vogt U. (eds.), Shimal excavations at the industrial site in Tayma, «Atlal» 1985/1986: Excavation of the German Archaeological 19, pp. 21–26. Mission on Ras al-Khaimah UAE. A Preliminary Avanzini A. (ed.) 2002, Khor Rori. Report 1 (Arabia Report (Berliner Beitrage zum Vorderen Orient 8), Antica 1), Pisa. Berlin, pp. 49-65. Avanzini A. (ed.) 2008, Khor Rori. Report 2. A port in Lischi S. 2015, An archaeological investigation on de- Arabia between Rome and the Indian Ocean (3rd C. corated shell bosses from Sumhuram – Sultanate of BC–5th C. AD) (Arabia Antica 5), Rome. Oman (poster presentation at Rencontres Sabéennes 19, 18-20 jun. 2015), Pisa. Avanzini A. (ed.) 2011, Along the aroma and spice routes. The harbour of Sumhuram, its territory and the trade Lischi S. 2018, Macroscopic analysis of the bead assem- between the Mediterranean, Arabia and India, Pisa. blage from the South Arabian port of Sumhuram, Oman (seasons 2000–2013), «Arabian Archaeology Avanzini A. 2014, Un port d’Arabie entre Rome et l’In- Epigraphy» 29/1, pp. 65-92. de, «Comptes Rendus de l’Académie des Inscriptions et Belle Lettres» 1, pp. 483–505. Liverani M. 2005, The near east: the Bronze Age in the ancient economy: evidence and models, in: Manning Avanzini A., Sedov A.V. 2005, The stratigraphy of J.G. and Morris I. (eds.), Social Science History, Sumhuram: new evidence, «Proceedings of the Stanford, pp. 47-58. Seminar for Arabian Studies» 35, pp. 11–17. Lombardi A. 2002, Small finds from 1998 campaign: -ca Bradl B. 1984, , The engraved tridacna-shell discs talogue, in: Avanzini A. (ed.), Khor Rori Report 1 «Anatolian Studies» 34, pp. 15-41. (Arabia Antica, 1), Pisa, pp. 71–76. Caputo F.P., Genchi F. 2015, Seashell discs from the Lombardi A., Buffa V., Pavan A. 2008, Small fin- Early Iron Age collective tombs of Daba, Sultanate of ds, in: Avanzini A. (ed.), Khor Rori. Report 2. A (poster presentation at Seminar for Arabian Oman port in Arabia between Rome and the Indian Ocean Studies, The British Museum, 24-27 jul. 2015), (3rd C. BC–5th C. AD) (Arabia Antica 5), Rome, London. pp. 317-413. Degli Esposti M. 2015, , in: Salut and its neighbours Pavan A., Sedov A.V. 2008, The temple extra-mu- Avanzini A. (ed.), In the heart of Oman, the castle of ros, in: Avanzini A (ed.), Khor Rori. Report 2. A , Rome, pp. 77-86. Salut port in Arabia between Rome and the Indian Ocean Francis P. Jr. 2007, Personal adornments, in: (3rd C. BC–5th C. AD) (Arabia Antica 5), Rome, Sidebotham S., Wendrich W. (eds.), Berenike pp. 261–266. 1999/2000. Report on the excavations at Berenike, in- Phillips C. 1987, Wadi Al Qawr, Fashgha 1: The exca- cluding excavations in Wadi Kalalat and Siket, and vation of a prehistoric burial structure in Ras al-Khai- , Los Angeles, the survey of the Mons Smaragdus region mah, UAE (Department of Archaeology, University pp. 251–257. of Edinburgh Project Paper No.7), Edinburgh. Genchi F. 2013, Explorations at Daba, Oman. Schneider H. 1945, On the pomegranate, «The Preliminary Report Season 2013 (Unpublished Metropolitan Museum of Art Bullettin» 4/4, Internal Report of the Ministry of Heritage and pp. 117-120. Culture, Sultanate of Oman). Weeks L., Cable C., Franke K., Newton C., Hartnell T., Barker D. 1999, Decorated discs from Karacic S., Roberts J., Stepanov I., David- Sharm, «Arabian Archaeology and Epigraphy» 10, Cuny H., Prince D., Bukhash R.M., Radwan pp. 205–212.

West & East 369 Monografie, 4 Silvia Lischi

M.B., Zein H. 2017, Recent archaeological rese- Yule P., Franke K., Meyer C., Nebe G.W., Robin arch at Saruq al-Hadid, Dubai, UAE, «Arabian C., Witzel C. 2007, Zafar, Capital of Himyar, Ibb Archaeology and Epigraphy» 28, pp. 31-60. Province, Yemen: First Preliminary Report: Summer Yule P., Kervran M. 1993, More than Samad in Oman: 1998 and Autumn 2000, Second Preliminary Report: Iron Age Pottery from Suḥar and Khor Rorī, «Arabian 2002, Third Preliminary Report: 2003, Fourth Archaeology and Epigraphy» 4/2, pp. 69-106. Preliminary Report: 2004, «Archäologische Berichte aus dem Yemen» 11, pp. 479-547.

West & East 370 Monografie, 4 Christianity in the Gulf: vestiges of the East Syrian presence in late antiquity

MARA NICOSIA

University of Naples “L’Orientale”

Abstract

The Arabian-Persian Gulf area has been fertile soil for different civilisations through the centuries, and bears the traces of numerous settlements of many different historical periods. Specifically, the area known in the past as Bēt Qaṭrāyē shows the vestiges of the Christian Syrian presence. The coastal region stretching from Kuwait to the United Arab Emirates, including also some of the Persian islands, displays a strange unevenness between the literary sources and the archaeological data relating to the history of the Christian communities that lived there between the fourth and the ninth centuries. The main texts naming the area, specifically its upper clergy and its most important religious circumstances, are the acts of the councils, but there are also other mentions scattered in different sixth and seventh centuries authors’ texts. Following the path of the ups and downs of the Church of the East, this paper aims at reconstructing a hypothetical history of the dioceses and the settlements in the area that witnessed the Christian presence. Anyway, this history cannot be understood only by relying upon written sources, and needs then to be integrated with the archaeological reports produced during the last decades. The result of this integration is the picture of a highly lively community, strong and self-reliant, whose history was too often forgotten.

Keywords

Bēt Qaṭrāyē, Syriac, Church of the East, Syriac Christianity, Arabic-Persian Gulf

West & East 371 Monografie, 4 Mara Nicosia

1. Preliminary remarks so-called Jacobite or West Syrian Church, this was rather settled in other parts of the Arabian Peninsu- The history of Syriac Christians living in the area la, as testified, for instance, by the massive massacre known as Bēt Qaṭrāyē has been somewhat under- of Christians belonging to this Church that took studied by Syriac scholars, leaving thus many ques- place in Naǧrān in 523 CE.3 tions unanswered. However, if we take a closer look at the data we have about the centuries from the fourth to the ninth, we see a very fascinating and 2. Sources about BQ lively reality, made up of highly educated writers, distinguished theologians and, most of all, a very The sources mentioning BQ or testifying a Chris- strong religious class capable of releasing itself from tian presence in the area can be divided into writ- the Persian metropolitan authority. ten and archaeological sources: they are quite var- The name Bēt Qaṭrāyē (BQ) refers to a huge iegated and also hardly ever consistent one with the area spanning from the coasts of the Kuwait, other. They shall be treated in the following two to Saudi Arabia, Bahrain, Qatar and the Unit- paragraphs. ed Arab Emirates.1 This area has been affected by important movements of people, goods, cultures and cultural patterns. Specifically, it was inhabit- 2.1 Written sources about BQ ed for several centuries by some Eastern Christian communities.2 The textual evidences come both from Islamic and The Church that is to be found in the BQ region Christian sources. Reference to Christianity in is the Church of the East, also labelled Nestorian, the Arabian peninsula are found in the Qur’ān, East Syrian, Chaldean Church, Assyrian Church in the Sīrat Rasul Allah by Ibn Isḥāq (d. 767 CE), (of the East), etc.: the BQ Church was suffragan of in the Ta’riḫ al-umam wa-l-muluk by al-Ṭabari the Persian metropolitan see of Rev Ardashir (that (d. 922 CE) and in some genealogies, though all was under the authority of the patriarchate of Se- those sources often pose problems of historicity.4 leucia-Ctesiphon). Concerning the presence of the Regardless, certain sources written by Muslim authors, such as the Murūj al-ḏahab wa maʽādin al- 1 According to modern excavations, it must have by al-Masʽūdī (d. 958) and the ʽ stretched to a small portion of the coasts of Iran. See Briquel jawāhir Mu jam Chatonnet, Debié 2017, p. 122: ‘Cette dénomination real-buldān by Yāqūt al-Hamawī (d. 1228 CE), rep- couvre un espace plus large que le Qatar actuel qui en a hérité resent useful tools to reconstruct, for instance, the son nom, et s’applique à pratiquement toute la côte orientale toponomastics.5 On the Christian side, there are de la péninsule arabique et aux iles qui la bordent’. But later, p. 124: ‘Les traces archéologiques retrouvées de cette implan- acts of synods, chronicles or letters: these sources, tation chrétienne dans le Golfe se sont multipliées depuis un demi-siècle. Près de la rive orientale, le site insulaire de Kharg 3 (Iran) a été fouillé dans les années 1959-1960 avant l’installa- Gajda 2009, pp. 97-102; Robin 2010, pp. 69-79. tion d’un terminal pétrolier’. 4 Naturally, this statement does not rest on the personal 2 Briefly put, the two main councils of Late Antiqui- judgment of the present writer, but on the fact that the period ty, namely the Council of Ephesus in 431 CE and the Coun- under examination coincides with the early centuries of Islam, cil of Chalcedon in 451 CE, triggered the great Christological and the texts belonging to that moment were usually arranged schisms within Eastern Christianity. The so-called East Syrian to legitimize Islam and the Prophet Muḥammad, often belong- Church rejects the Christological positions of Ephesus, which ing to the anti-Christian disputes. Being Sacred Texts or texts were mainly the definition of Mary asTheotόkos , the repudi- about the figure of the Prophet Muḥammad, they often display ation of Nestorianism and the acceptance of miaphysitism mythical accounts and cannot be considered entirely historical- (which denotes the belief in the single and indivisible nature ly reliable. It is possible to say the same about genealogies, since of Christ, in which live together both his humanity and his it is not unusual to have invented characters or people that ap- divinity). On the other hand, the West Syrian Church rejects parently lived for more than one hundred years in order to fill the positions of Chalcedon, which were mainly the negation of the gaps between two or more characters. monophysitism (which denotes the belief in the one and only 5 For accounts on the Muslim sources and their prob- divine nature of Christ) and the superiority of the Patriarchate lems see Bin Seray 1996, pp. 324-325, Mourad 2014, of Constantinople over those of Antioch and Alexandria. pp. 40-44, 53-55 and Al-Murikhi 2014, pp. 250-253.

West & East 372 Monografie, 4 Christianity in the Gulf: vestiges of the East Syrian presence in late antiquity

primarily composed for practical reasons, present in the first lines is called, among the others, patri- anyway certain limitations, lacking in accurate in- arch of ‘the isles’: this name was used to indicate, formation about, for instance, the exact location of together with ‘the black islands’, the archipelago of the towns they mention. Bahrain before the seventh century, when the desig- The text that places a diocese in BQ earlier in nation Bēt Qaṭrāyē came into usage.12 Bishops com- time is the much discussed Chronicle of Arbela, ing from BQ are mentioned in many other synods mentioning the existence of a bishopric already at until the one that took place in 676, which will be the times of the bishop of Arbela Ḫairān (217-250 discussed later.13 CE),6 but, considering the scholarly querelle about What we learn, particularly from the first text, its authenticity, it is more prudent not to rely upon is that the reality of the BQ area was lively and tur- it. It should be noticed, though, that Christian mer- bulent already: there was at least one bishopric and chants might have lived in the region from the first some dissidents, people that apparently used to be century on, working on the flourishing routes con- bishops or members of the clergy but must have de- necting the Gulf and the Indian Ocean.7 viated from the directions of their Christian author- The first reliable mention of a diocese in the ities, being thus removed from their roles. Anyway, area of BQ is in the acts of the synod of Mar Isaac these dissidents still performed their functions, like in 410.8 Interestingly, these acts contain an invec- ordering new priests, and must have had a conspicu- tive against a certain Baṭai of Mašmahīg, whose re- ous number of followers, according to the words of moval from his functions as bishop is confirmed the acts of the synod. and repeated. For the time being, it will only be Among the most important sources describing stated that Mašmahīg is one of the most import- the history of the area there are the letters writ- ant towns in the BQ area. About Baṭai, we learn ten by the catholicos Išoʽyahb III, addressed to the that he has been accused of unspecified miscon- clergy, bishops, monks and people of BQ.14 As a ducts and, as Chabot translates, he « était censu- matter of fact, the religious controversies already ré et repoussé, est de nouveau censuré, excommu- described in the synod of Mar Isaac were not fated nié, déposé et reprouvé par tout le synode, lui et to end, and the whole area apparently released it- quiconque oserait recevoir de lui l’imposition des self from the control of the metropolitan. It seems mains».9 Furthermore, a certain Daniel, previous- thus that the ecclesiastical authorities of BQ were ly ordered by Baṭai, is said to have attended the sy- powerful enough to set an important rebellion, to nod: about him Chabot translates «qu’il soit lui- the extent that Išoʽyahb III (catholicos from 649 même censuré et excommunié par ce synode; qu’il to 659) felt the need to address them a set of letters sorte du pays où il a jeté le trouble et qu’on ne l’y with undisguised impetuousness and a certain lack rencontre plus ; qu’il s’en aille à sa place, censuré, of Christian charity, in order to reproach their re- déposé et réprouvé».10 Among the signatures con�- bellion and gain their loyalty back. These letters cluding these acts, there is the one of a certain Elias describe a community that acts independently, has bishop of Mašmahīg, that apparently agreed upon excommunicated bishops that order new priests, the two excommunications. 12 It should be remarked, though, that in a footnote The second mention is found in the acts of the Chabot (Ibid., p. 276) explains that the diacritic indicating the 11 following synod of Mar Yahbalaha I in 420, who plural form ‘the isles’ might as well be a scribal mistake, imply- ing that ‘the isle’ would point to the Jazira region. Regardless, 6 Kawerau 1985, p. 31 (text); Ramelli 2002, p. 43 in the previous synod the indication of ‘the isles’ unmistakably (transl.). designates the islands of the Gulf (Ibid., p. 34.20 text and p. 273 trans.). 7 Briquel Chatonnet, Debié 2017, pp. 122-129. 13 Beaucamp, Robin 1983, p. 181; Bin Seray 1996, 8 Chabot 1902, pp. 17-36 (text) and pp. 253-275 pp. 320-321; Bin Seray 1997, pp. 212-217. See also Briquel (trans.). Chatonnet 2010, pp. 179-180. 9 Ibid., p. 34.26-27 (text) and p. 273 (trans.). 14 For brief accounts on Išoʽyahb III’s life see Hoyland 10 Ibid., p. 35.9-13 (text) and p. 273 (trans.). 1997, pp. 174-182; Healey 2000, pp. 227-229; Healey 11 Ibid., pp. 37-42 (text) and pp. 276-284 (trans.). 2009, pp. 3-7 and Wilmshurst 2011, pp. 104-106.

West & East 373 Monografie, 4 Mara Nicosia

and, most of all, relies on the civil (Muslim) au- however, priests and deacons, fulfill the ecclesiastical thorities to solve the controversies.15 In order to ministry with all that is legally just without the com- give a hint of the concern displayed by Išoʽyahb, munion of those who were called your bishops until some brief quotations from his letters are provided a priestly chair is established for you by the Church here, in the remarkable translation recently offered of God according to the precept and ordinance of the spiritual laws. And you, faithful, in whose hands by Mario Kozah.16 is placed ancillary authority of the islands and the Letter XVII, addressed to the bishops of BQ: dwellers of the desert, I mean Dayrīn (dyryn) and Mašmahīg (mšmhyg) and Talon (tlwn) and Ḥaṭṭa You have hindered yourselves from coming to the (ḥṭʾ) and Hagar (hgr), strive at this time more than Church of God and have set off with thoughtless ha- ever to guard your faith and in legal establishment of ste towards the leader of your rebellion and you have the priesthood by which you have been consecrated, with unintelligent madness written and sealed the more so than for worldly affairs. […]19 rebellion against God. Furthermore, you have also brought the book of your rebellion to the door of the temporal rulers, and have already done enough to cut Scholars have been able to reconstruct the modern off all hope of a life in the Church. God’s Church […] location of the places mentioned by the patriarch has now […] removed you from the honor and the as follows: Dayrin has been recognized as a village rank through which you have become renowned. […] on Tarut island in Saudi Arabia, Mašmahīg as al- Therefore, let us know your situation as soon as pos- Muharraq in Bahrain, Ḥaṭṭa as al-Qaṭīf, in Saudi sible, not just with simple letters, as (you have done) Arabia and Hagar as al-Hufūf in Saudi Arabia as thus far, but with a true demonstration of action and well. Concerning Talon, the toponym resembles the Lord through his grace will help you to do what is the old name of Bahrain itself, as reflected in Greek 17 appropriate for your salvation, and be strong. Tylos, and thus might refer to the Bahrain archipelago generally speaking, even though it would be Letter XIX, addressed to the people of BQ: the only toponym referring to an entire region instead of a single town.20 In addition to that, con- The tale of evils which have occurred in your region, cerning the location of this place, in the final can- my beloved brothers, presses us to write to you a se- ons written for Jacob, bishop of Daray (probably cond time, since I am very troubled by the fear of Dayrin), by Išoʽyahb I, it is said that the island of other evils which are set to happen in your region. For from the beginning, when I heard that your bi- Daray was next to Talūn (the toponym spelled as 21 shops contrived wickedly, wrote and sealed a docu- t’lwn). Moreover, in the Passion of the martyr ment of apostasy from the Church of God, I reali- Yazdbozid, it is said that a monk, identified by Fiey zed that they were prepared to fall into every evil and with Moses II, partner of Abraham the Stranger, would cast down the people of God whose bishops was brought as a captive on an island called Talwn they are called.18 around the sixth century.22 Another interesting source for the history of Important details about the names of the towns in- BQ is the History of Mar Yawnan:23 it is a hagi- cluded in the region are provided by the Letter XVI- ographical text about the life of Mar Yawnan, the II, addressed again to the people of BQ. We read:

19 Ibid. pp. 51-53 (trans.) and 74-75 (text). Since, however, your bishops-in-name were not sat- 20 See Trimingham 1979, p. 281; Beaucamp, Rob- isfied with impiety against the Church of God, they in 1983, pp. 171-179; Healey 2000, pp. 233-235; Car- presented a public display of their rebellion […].You, ter 2008, pp. 100-103; Carter 2013, pp. 313-314 and Kozah 2017, p. 63. For a discussion on Tylos see Potts 1990, pp. 125‑149. 15 Kozah 2015, p. 44. 21 Chabot 1902, p. 165.27-28 (text) and p. 424 (transl.). 16 Kozah 2015, pp. 45-88. 22 Fiey 1966, pp. 135-136 and C. and F. Jullien in Steve 17 Ibid. pp. 45-47 (trans.) and 69-70 (text). 2003, p. 166. 18 Ibid. p. 57 (trans.) and 78 (text). 23 Payne 2011; Brock 2015, pp. 1-41.

West & East 374 Monografie, 4 Christianity in the Gulf: vestiges of the East Syrian presence in late antiquity

renowned founder of the famous monastery near mentioned in the same non-connoted way in the Piroz-Shabur/Anbar, on the Euphrates, who is Chronicle of Michael the Syrian during the events likely to have lived in the fourth century. This His- of the years 824/825, even though later, during the tory is however to be dated probably to the late sev- events of the 834/835, the people of Bahrain ap- enth or early eighth century. This text is the only pear to be Christians.28 extant example of hagiographic Syriac literature The last important mention of the dioceses concerning the Gulf area: the supposed author, of BQ is to be found in the synod held by Mar Zadoy, is indeed abbot of the monastery of Mar Giwargis I in 676 in the town of Dayrin.29 Mar Thomas on the ‘black island’ and writes with the Giwargis was Išoʽyahb III’s successor, and was ap- purpose of linking the monastic life of the region parently the one who tamed the rebellion and re- to that of the great founder fathers. The narration stored order. According to this last text, Dayrin offers thus a reliable proof of the monastic life in had a bishop at that time, Išoʽyahb, who signed BQ and provides us with a hint of the need that the acts of this synod, but there appears to be also the monastic communities had to validate them- a metropolitan bishop of BQ attending the syn- selves. od, named Thomas, whose signature was written Another mention of the region that has to do right after that of Giwargis. Did BQ become a with monasteries is to be found in the Chronicle of metropolitan see in the meantime? Fiey suggests Seert, chapter 62: it tells the story of the building that this title might have been ‘usurped’ by Thom- of some monasteries by ʽAbdišoʽ (who lived during as without Giwargis’s validation,30 but it might the catholicosate of Mar Tomarsa in 363-371 CE), as well be the case that Giwargis, for the sake of one of which was on an island of Bahrain.24 peace, accepted BQ as a metropolitan see and ad- Moreover, Qatar is mentioned in the Chron- mitted its exponents in the list of participants to icon anonymum de ultimis regibus Persarum,25 an win back their loyalty.31 anonymous chronicle written probably by an From this moment on, we will not find a single East Syrian monk around 660 CE, known also as written account regarding churches in this area. Chronicle of Khuzistan.26 The ‘islands of Qatar’ are Fiey states that he was able to find another men- mentioned after the narration of the last years of tion of Christians (generally speaking) in Bah- the emperor Heraclius I and of his death in 641. rain in 835 and that their last mention of these is The islands are mentioned along with other towns under catholicos Yūwānīs III (893-899), when a in the Arabian Peninsula, among which we find rebel šayḫ takes power and behaves well towards also Hagar and Ḥaṭṭa. The people living in these the Christians.32 This writer was however unable towns appear to be monotheistic, but it is not pos- to track these events in the work indicated by sible, in the opinion of the present writer, to decide Fiey, which is Mari Ibn Sulayman’s Aḫbār faṭār- whether they should be accounted as Christians or kat kursī al-mašriq min kitāb al-maǧdal (translat- Muslims, according to the information provided ed by Gismondi as De patriarchis nestorianorum by this text. In the same chronicle there is also a commentaria).33 mention of a certain Maʽnē, native of the island of Dayrin, who was interpreter for the Laḫmid king al-Numan III (579-601). Moreover, there is a mention of BQ and another one of Qatar, even pp. 18, 22 and 30 (Latin). 28 though in none of those cases there is a specific ref- Chabot 1899, III, pp. 61-62 and 85. 29 erence to a religious affiliation.27 Bahrain is also Chabot 1902, pp. 215-226 (text) and pp. 480-490 (trans.). 24 Scher 1910, 310-312. 30 Fiey 1969, p. 211. 25 Guidi 1903, p. 38.26 (Syriac text) and p. 31 (Latin 31 Beaucamp, Robin 1983, p. 184 and Kozah 2014, translation). p. 7. 26 Debié 2015, pp. 611-613. 32 Fiey 1969, p. 211. 27 Guidi 1903, pp. 20.9, 25.27 and 37.3 (Syriac) and 33 Gismondi 1899.

West & East 375 Monografie, 4 Mara Nicosia

2.2 Archaeological data news concerning the proceeding of the works.38 The site of al-Quṣūr displays two churches: the first Those who would like to see the vestiges of the towns one (called A1) was discovered already in 1989, of Dayrin, Mašmahīg, Ḥaṭṭa and Hagar would be while the second one (A2) had to wait until the very disappointed to learn that not a stone has been 2008 to be found. Another space has been exca- unearthed from the places where these towns used vated in 2009 near those churches, described as an to be, even though, in 1914, alleged ‘old founda- “ecclesiastical” space probably used for residential tions’ have been noticed near Mašmahīg.34 What purposes: the whole area displays more than 140 is even more incredible is that interesting sites structures, that have been interpreted has monks’ have been excavated in other places all around the cells, with a kitchen and a stocking area.39 The two area, places that are never mentioned in the writ- churches show different architectures: A1 is bigger ten sources. The sites that are going to be described (35 × 19 m), built in mud bricks, finely plastered briefly here are those of al-Quṣūr on the island of and has a big central nave and apse, with aisles Failaka in Kuwait, Sir Bani Yas in the United Arab on both sides preceded by a narthex, while A2 is Emirates, Jubail in Saudi Arabia and Kharg in the smaller (22 × 7 m) and with a single nave.40 This Iranian province of Bushehr. These sites are to be last building has to be considered an unicum in the dated between the seventh and the ninth centuries. Gulf area, considering that the other churches ex- Minor Christian vestiges have been excavated in cavated are similar to A1.41 A tomb, dating back to Akkaz (Kuwait), Thaj and Jebel Berri (Saudi Ara- 280 CE, has been found in the south wall of the bia) as well, but they will not be described in this central nave of A1: its date must not lead to the paper. conclusion that the Christian site belongs to the The site of al-Quṣūr, which is one of the big- third century, but testifies instead that the area gest on the island (2,80 × 1,60 km),35 has been sur- was occupied before the arrival of the Christians. veyed for the first time by an Italian team in 1975- According to the discoveries of the MAFKF, the 1976, followed by a French mission from 1988 to first installation of Christians in the site belongs to 199036 – when the Gulf War forced the mission the end of the Sasanian period, and lasted until the to interrupt its works –, and then again in 2007 ninth century.42 The hypothesis of the MAFKF is and from 2012 on.37 At the present time, the Mis- that the whole area used to be a monastic complex. sion Archéologique Franco-Koweïtienne de Faïla- The second site is Sir Bani Yas, in the United ka (MAFKF), launched in 2011, is responsible for Arab Emirates. This site displays an imposing mo- the excavations: the mission constantly updates nastic complex (90 × 70 m)43 that was found in its website with articles, information, pictures and 1992 by the Abu Dhabi Islands Archaeological Survey (ADIAS) mission, and has been excavated 34 Potts 1990, p. 124. Among the alleged remains of from 1993 to 1996.44 In 1995 the archaeologists the ancient Eastern Christian towns, there are also the three limestone funerary stelae from Tarut, displaying figures recall- found a finely plastered church (16 × 11 m), very ing those of the Christian stelae of Rasm al-Qanafez, in Syria. similar to A1 in al-Quṣūr, with a big central nave These three issues, kept at the Bahrain National Museum, have and apse, aisles on the sides and a narthex.45 Be- no record of acquisition with their original provenance, making it very difficult to account them as proper evidences. See Potts 1990, pp. 212-213. Besides, Baumer mentions a mon- 38 http://mafkf.hypotheses.org/. astery on the island of Tarut (where Dayrin was supposed to 39 Bonnéric 2016, p. 108. be placed), but since he gives no further reference on the top- 40 Salles, Callot 2013, pp. 242-259. ic there is no way to understand which structure he refers to, 41 since he is the only scholar mentioning it. See Baumer 2006, Bonnéric 2016, p. 108. p. 138. For a discussion on archaeological evidences in Qatar 42 Ibid. p. 111. see Al-Thani 2014. 43 Elders 2001, p. 48. 35 Bonnéric 2016, p. 107. 44 See King 1997 for the preliminary report of the 1990s 36 See Bernard et Al. 1991 for the preliminary report excavations. of the 1989 mission. 45 For the analysis of the phases of construction of the 37 Salles, Callot 2013, p. 237. church see Elders 2001, pp. 49-53.

West & East 376 Monografie, 4 Christianity in the Gulf: vestiges of the East Syrian presence in late antiquity

sides, the archaeologists found six courtyard hous- The last site is Kharg, in the Iranian province es in the surroundings of the monastery.46 The of Bushehr, that was excavated for the first time dating of the site has been defined by Carter on in 1959 by the Mission Archéologique Française. the basis of ceramic and radiocarbon evidences, This site displays a stunning monastic complex giving a mid-seventh to mid-eighth occupational measuring 123 × 88 m: it has a church with a big time span, even though the stucco work seems to central nave (5 × 15 m), aisles and narthex, a re- indicate rather a eighth/ninth century date.47 The fectory, a library and a sizable number of cells,53 a site of Sir Bani Yas thus appears to be contempo- structure that resembles those of the monasteries rary to the one in al-Quṣūr, even though the latter excavated in Egypt.54 Not very far from the mon- might have been occupied for a little longer.48 It astery, the archeologists found a housing area that is worth adding that another church with related they recognized as married priests’ residences, structures was excavated not far from Sir Bani Yas, since Eastern Christianity allows the clergy (but in Marawah, during the 2000 spring campaign: the not the monks) to get married.55 Someone else two sites are similar one to another.49 instead considers this area as rather inhabited by The third site is the one in Jubail: it displays a people linked to the monastery, for instance, by remarkable church that has been found by acci- working relationships, or by hosts of the monas- dent in 1986. This site has a complicated history tery.56 It has been estimated that the whole com- and unfortunately, as soon as some crosses on the plex could have hosted about 100 people, and dis- wall were discovered, it has been surrounded by played ingenious hydraulic and agricultural works fences restricting the access to it.50 The Saudi De- in order to provide its population with everything partment of Antiquities undertook excavations needed.57 The site displays also a certain number of in 1987, but apparently the report has never been Christian tombs, connoted by crosses engraved on published: the only study we have seems to be them and by some Syriac inscriptions.58 the unofficial one provided by Langfeldt in 1994, which states that the church appears similar to the ones in al-Quṣūr and Sir Bani Yas under the point 3. The authors of BQ of view of the stuccos, but different in the architecture, having a single square room instead of central The BQ region gave birth to several well-known nave with aisles. Due to these reasons, Carter sug- authors, who lived during the sixth-seventh cen- gested that the Church might belong to the Jaco- turies. Some of them were known everywhere in bite confession, even though there is no concrete the Syriac-speaking community as being refined evidence to confirm this hypothesis.51 Apparently, theologians, learned biblical exegetes and religious three Christian gravestones have been found in the poets. They are explicitly said to be born and edu- vicinity of the site, but the circumstances under cated in BQ, where most of them spent their en- which they have been discovered prevent us from tire life. The most important writer of this group is considering them as proper evidence.52 Isaac of Nineveh, who left BQ at a certain moment

46 Carter 2008, p. 73. 47 Ibid., pp. 89-92. See also Kennet 2007, pp. 89-94. covered by some inches of sand, and called the Department of Antiquities in order to leave a record of the discovery. 48 Carter 2013, p. 321. 53 Steve 2003, pp. 85-113. 49 Elders 2001, pp. 53-56. 54 Ibid., p. 86. 50 Langfeldt 1994, pp. 32-34. 55 Bowman 1974-5, pp. 58-59. On Persian monasticism 51 Carter 2008, p. 98. in the Gulf see the detailed account given by C. and F. Jullien in 52 Langfeldt 1994, pp. 42-44 writes that he was con- Steve 2003, pp. 155-183. tacted by three people that have found these gravestones some 56 Steve 2003, pp. 150-151. 200/250 metres north of the church and consequently decided 57 to bring them home, believing that the Saudis would destroy Ghirshman 1960, p. 23; Bowman 1974-5, them. Subsequently, however, they decided all together to re- pp. 52‑54; Steve 2003, pp. 150-152. locate the gravestones in one of the rooms south of the church, 58 Ghirshman 1969, p. 6; Steve 2003, pp. 45-59.

West & East 377 Monografie, 4 Mara Nicosia

during the struggle with the Persian metropolitan Leaving alone the city of Edessa – whose Arama- see: he was bishop of the city of Nineveh only for ic dialect turned to become what we call Classical five months, after which he resigned and went liv- Syriac –, all the towns belonging to the community ing as an eremite in the Khuzistan region.59 Other have their own mother tongues, not necessarily be- important names are those of Dadishoʽ Qaṭraya, longing to the Aramaic group. This happens because Aḥob Qaṭraya and Abraham bar Lipah.60 Among the boundaries of the Syriac world span from Tur- the authors of BQ there were also a certain num- key, to Mesopotamia, Persia and the Arabian Pen- ber of writers all named Gabriel, whose identities insula, and also as far as India and China, meaning and works have been studied by Brock.61 The ex- that the Christians of Syriac tradition speak differ- istence of this intellectual class implies the present languages.64 Of course, BQ is no exception: it is ence of a high-level pole of education and learn- worth remarking that, among the languages used in ing in BQ, able to provide refined knowledge on the Gulf area, there were different Arabic varieties, matters such as biblical exegesis and religious stud- Aramaic, Hasaitic65 and Persian. ies. As a matter of fact, Brock’s idea of the exist- Given that the religious language of BQ was ence of monasteries acting as higher education Classical Syriac, concerning the language these schools, like the famous one in Nisibis, appears Christians spoke, scholars came across a set of very well-grounded.62 Moreover, these monaster- glosses in the alleged “language of Bēt Qaṭrāyē” in ies are likely to have served as caravanserais, since a manuscript, the (olim) Diyarbakır 22, contain- there are evidences for an involvement of Chris- ing an East Syriac ninth century commentary to tians in the shipping and commercial activities in the Bible. These glosses have been investigated first the Gulf, with specific regards to trade in pearls.63 by Schall66 and later by Contini67, and even though Thus, we should not imagine these places as closed the two scholars analyzed different groups of words, structures whose access was restricted, but as gath- they drew the same conclusions: the glosses seem to ering places open to the population, where there pertain to a language identifiable as a south-eastern was an exchange of goods, knowledge and proba- variety of Aramaic. Admittedly, Syriac is also a dia- bly medical cares. lect of Aramaic, as explained above, but it is actually a central Aramaic dialect,68 whereas this spoken variety was southern Mesopotamian, as Mandaic. 4. The language of BQ What is even more interesting is that, according to the study of these glosses, some of the words display There is one last puzzling enigma concerning the Persian etymologies, as well as Arabic and common language spoken by the people of BQ. As a mat- Semitic ones, while others so far defied any attempt ter of fact, the identity of the Syriac communities is at identifying an etymology.69 grounded both on religious and linguistic bases, but the language used as the written variety and acting as a vehicular language is not actually a spoken variety.

64 See Contini 2004 on this topic. 65 59 On Isaac of Nineveh see, for instance, Brock 1999- For an account on Hasaitic in East Arabian context see 2000a, pp. 88-89; Brock 1999-2000b; Kozah 2014, pp. 7-13, Al-Jallad 2018, p. 32-34. Chialà 2014, Kessel 2015 and Hansbury 2015. 66 Schall 1989. 60 For accounts on Syriac writers from BQ see Brock 67 Contini 2003. 1999-2000a, 2009; Kozah 2014, 2015, 2017. 68 For an updated and recent discussion on Syriac as a 61 Brock 1999-2000a, pp. 89-92. dialect of Late Aramaic, and on its position in this group, see 62 Ibid., pp. 95-96. Butts 2019. 69 63 Carter 2008, p. 105. Probably also the pearl divers See Schall 1989 and Contini 2003 for all the de- mentioned in the XIX canon of the synod of Išoʽyahb were tails on these words. Moreover, a fair amount of so far unex- Christians from BQ, see Chabot 1902, p. 198.16 (text) and plained vocabulary of presumably external origin has been p. 448 (trans.). pointed out by Holes 2002, pp. 274-279 in Bahraini Arabic.

West & East 378 Monografie, 4 Christianity in the Gulf: vestiges of the East Syrian presence in late antiquity

5. Attempted conclusions Church.74 The disappearance might be due then to a conviction of these dioceses to some kind of dam- Considering all that has been described so far about natio memoriae. the BQ region, it is clear that the area shows a cer- None of these hypotheses, at least taken in isola- tain number of incongruities: first, given that the tion, in fully conclusive. The (entirely speculative) Christian presence is attested from the fourth to idea of a relocation of the communities from the the late ninth century, we are faced with written seventh century on, from the places mentioned by sources only up to the seventh century and after that the written sources to the places where the archae- only with archaeological data;70 second, nothing ological sites are, cannot explain the lack of textual has been unearthed from the towns mentioned in mentions of the towns and of the reasons for the the texts, while quite large archaeological sites have movement itself. Anyway, even though this was been excavated in places that the written sources not the case for the seventh and eighth century, the never mention; third, texts written by local authors spread of Islam might be responsible for the ulti- belong only to the late sixth – early seventh century. mate decay of Christianity in the Arabian Penin- These problems are to be considered unsolved, sula, at least from the ninth century on.75 even though it is possible to make some reflections. Concerning the authors of BQ, given that the The main question concerns the incongruity sixth and seventh centuries were the most active between written sources and archeological data. ones under the cultural point of view for the East- The first hypothesis is that before the seventh cen- ern Syriac presence in BQ, we cannot exclude that tury there were no proper Christian buildings in there existed other texts belonging to different peri- the area, built in bricks or stones, used as churches: ods that got lost, even though it seems unlikely that this implies that the congregations rather gathered these are not mentioned elsewhere. inside the houses of their members and that those One last remark can still be made: there is a houses were probably made of perishable material.71 chance that the monasteries that have been found Even though we have parallels elsewhere, it must by the archaeologists were not connected, for some be underlined that this community was strong and reason, to dioceses settled in BQ region, justifying well settled and was thus likely to have a physical the disappearance of the local dioceses from the place to give expression to its faith. Most likely then sources. Besides, if Payne’s remark about the mon- the vestiges of the ancient towns of BQ lie under the asteries of Kharg and Sir Bani Yas sharing a strik- modern constructions and palm gardens.72 ing number of features with the one described in the To explain the disappearance of BQ from writ- History of Mar Yawnan is correct, we would have ten records, Carter suggested that the bishops of found at least one of the missing pieces of the puz- BQ stopped attending the synods or ceased to be zle.76 This theory would be in perfect harmony with part of the East Syrian community:73 he pointed out the idea previously put forward: since the monaster- the possibility that Giwargis’s synod was a complete ies were placed in a land far from the important reli- failure and did not solve the controversies, leav- gious centres, the need of linking their own history ing BQ completely disconnected from the Mother with the more relevant ones appears justified.

70 Admittedly, as Morony 1982, p. 40 pointed out, there were many other bishoprics in the Mesopotamian and Persian provinces that disappeared from the sources already af- 74 Carter 2013, p. 327. ter the synod of 605, a fact that he explains as the reflex of the 75 As soon as the Muslim authorities started to gain the conflict between the East and West Syrian Churches and even- civil power, they imposed a poll tax on all the non-Muslims liv- tually of ecological changes. See Morony 1982 for a detailed ing under their control. Many people then decided to convert account on the changing in administrative geography of Iraq in for economical reasons: for instance, in Beth Mazūnaye, many Sasanian and Early Muslim periods. non-Muslims converted in order to avoid giving up their pos- 71 Carter 2008, pp. 103-104. sessions. See, among others, Carter 2008, p. 100, Al-Mu- 72 Carter 2013, p. 314. rikhi 2014, pp. 249-250 and Kozah 2015, pp. 44. 73 Carter 2008, pp. 103-104. 76 Payne 2011.

West & East 379 Monografie, 4 Mara Nicosia

Ultimately, it is very likely that the ruins of Day- In conclusion, further works would certainly be rin, Mašmahīg, Ḥaṭṭa and Hagar lie under the mod- necessary. Unfortunately, even though the history ern settlements, and that after the synod of Giwargis of BQ has recently been object of renovated inter- the importance of BQ was extremely reduced, caus- est, it appears still doomed to its destiny of being left ing the disappearance of these towns from the offi- aside: for example, whereas Hoyland’s able synthesis cial records. It is reasonable to believe that the only takes into account the events of BQ,77 no mention important Christian centers remaining in the area of them is made in Fisher’s comprehensive book, were the monasteries. since the only Christians taken into account are the miaphysites.78

Acknowledgements

The author wishes to thank Vittorio Berti, Emiliano Fiori, Luca D’Anna and Ahmad Al-Jallad for their remarks.

77 Hoyland 2001, pp. 27-32. 78 Fisher 2015.

West & East 380 Monografie, 4 Christianity in the Gulf: vestiges of the East Syrian presence in late antiquity

BIBLIOGRAPHY

Al-Jallad A. 2018, What is Ancient North Arabian?, Brock S.P. 2009, Gabriel of Qatar’s Commentary on the in: Pat-El N., Birnstiel D. (eds.), Re-engaging Liturgy, «Hugoye» 6/2, pp. 197-248. comparative Semitic and Arabic studies, Wiesbaden, Brock S.P. 2015, The History of Mar Yawnan, in: pp. 1-44. Kozah et al. (eds.), pp. 1-42. Al-Murikhi S.S. 2014, The influence of Christianity Butts A.M. 2019, The Classical Syriac language, in: among the Arab tribes in the Gulf area during the sixth King D. (ed.), The Syriac world, London-New York, and seventh centuries AD, in: Kozah et al. (eds.), pp. 222-242. pp. 249-267. Carter R.A. 2008, Christianity in the Gulf during the Al-Thani H. 2014, An archaeological survey of Beth first centuries of Islam, «Arabian Archaeology and Qatraye, in: Kozah et al. (eds.), pp. 23-35. Epigraphy» 19, pp. 71-108. Baumer C. 2006, The Church of the East. An illustrated Carter R.A. 2013, Christianity in the Gulf after the co- history of Assyrian Christianity, London. ming of Islam: redating the Churches and Monasteries Beaucamp J., Robin C. 1983, L’évêché Nestorien de of Bet Qatraye, in: Robin C.J., Schiettecatte J. Mâšmâhîg dans l’archipel d’al-Baḥrayn, in: Potts (eds.), Les préludes de l’Islam. Ruptures et continuités D.T. (ed.), Dilmun, new studies in the archaeology and dans les civilisations du Proche-Orient, de l’Afrique early history of Bahrain, Berlin, pp. 171-196. orientale, de l’Arabie et de l’Inde à la veille de l’Islam, Bernard V., Callot O., Salles J.F. 1991, L’église Paris, pp. 311-330. d’al-Qousour Failaka, État de Koweit – Rapport Chabot J.-B. 1899, Chronique de Michel le Syrien, préliminaire sur une première campagne de fouilles, Tome III, Paris. 1989, «Arabian Archaeology and Epigraphy» 2, Chabot J.-B. 1902, Synodicon Orientale ou Recueil de pp. 145-181. synodes nestoriens, Paris. Bin Seray H. 1996, Christianity in East Arabia, Chialà S. 2014, Two discourses on the “Fifth part” of Isaac «ARAM Periodical» 8, pp. 315-332. the Syrian’s writings: Prolegomena for Apokatastasis, Bin Seray H. 1997, The Arabian Gulf in Syriac sources, in: Kozah et al. (eds.), pp. 123-131. «New Arabian Studies» 4, pp. 205-232. Contini R. 2003, La lingua del Bēt Qaṭrāyē, in: Lentin Bonnéric J. 2016, Les établissement chrétiens du golfe J., Lonnet A. (eds.), Mélanges David Cohen: Études Arabo-Persique à la lumière du site archéologique d’al- sur le langage, les langues, les dialectes, les littératures, Quṣūr et des découvertes de la MAFKF, «Parole de offertes par ses élèves, ses collègues, ses amis; présentées à l’Orient» 42, pp. 103-123. l’occasion de son quatre-vingtième anniversaire, Paris, Bowman J. 1974-5, The Christian monastery on the pp. 173-181. island of Kharg, «Australian Journal of Biblical Contini R. 2004, L’orizzonte linguistico della Chiesa di Archaeology» 2/3, pp. 49-64. Persia: considerazioni preliminari, in: Vergani E., Briquel Chatonnet F. 2010, L’expansion du chri- Chialà S. (eds.), Storia, cristologia e tradizioni della stianisme en Arabie: l’apport des sources syriaques, Chiesa Siro-orientale, Milan, pp. 47-77. «Semitica et Classica» 3/1, pp. 177-187. Debié M. 2015, L’écriture de l’histoire en syriaque. Briquel Chatonnet F., Debié M. 2017, Le mon- Transmissions interculturelles et constructions identi- de syriaque. Sur les routes d’un christianisme ignoré, taires entre hellénisme et islam, Paris. Paris. Elders J. 2001, The lost churches of the Arabian Gulf: Brock S.P. 1999-2000a, Syriac writers from Beth recent discoveries on the islands of Sir Bani Yas Qaṭraye, «ARAM Periodical» 11-12, pp. 85-96. and Marawah, Abu Dhabi Emirate, United Arab Emirates, «Proceedings of the Seminar for Arabian Brock S.P. 1999-2000b, From Qatar to Tokyo, by way Studies» 31, pp. 47-57. of Mar Saba: the translations of Isaac of Beth Qaṭraye (Isaac the Syrian), «ARAM Periodical» 11-12, Fiey J.M. 1966, Notes d’hagiographie syriaque, pp. 475-484. «L’Orient Syrien» 11/2, pp. 133-154.

West & East 381 Monografie, 4 Mara Nicosia

Fiey J.M. 1969, Diocèse syriens orientaux du Golfe King G.R.D. 1997, A Nestorian monastic settlement on Persique, in: Graffin F. (ed.), Mémorial Mgr the island of Ṣīr Banī Yās, Abu Dhabi: a prelimina- Gabriel Khouri-Sarkis (1898–1968), fondateur et ry report, «Bulletin of the School of Oriental and directeur de L’Orient Syrien, 1956–1967, Louvain, African studies» 60, pp. 221-235. pp. 177-219. Kozah M. 2014, Introduction, in: Kozah et al. (eds.), Fisher G. (ed.) 2015, Arabs and empires before Islam, pp. 1-22. Oxford. Kozah M. 2015, Ishoʿyahb III of Adiabene’s Letters Gajda I. 2009. Le royaume de Ḥimyar à l’époque mo- to the Qataris, in: Kozah M., Abu-Husayn A., nothéiste. L’histoire de l’Arabie du sud ancienne de la Al-Murikhi S.S., Al-Thani H. (eds.) 2015, An fin du IVe siècle de l’ère chrétienne jusqu’à l’avènement Anthology of Syriac Writers from Qatar in the Seventh de l’islam, Paris. Century, Piscataway NJ, pp. 43-88. Ghirshman R. 1960, The island of Kharg, Tehran. Kozah M. 2017, The Syriac writers of Beth Qaṭraye in Ghirshman R. 1969, Les fouilles récent de Kharg, en the seventh century: an overview of the current state of Iran. La «route de la soie» et le pétrole (Séance publi- studies, in: Teule H., Keser-Kayaalp E., Akalin que annuelle des Cinq Académies, lundi 27 octobre K., Dorum N., Toprak M.S. (eds.), Syriac in its 1969), Paris. multi-cultural context, Leuven-Paris-Walpole, MA, pp. 59-80. Gismondi H. (ed.), 1896, Maris Amri et Slibae de patriarchis nestorianorum commentaria, Roma. Kozah M., Abu-Husayn A., Al-Murikhi S.S., Al-Thani H. (eds.) 2014, The Syriac writers of Qatar Guidi I. 1903, (CSCO Syr. III, 4), Chronica minora in the seventh century, Piscataway NJ, pp. 1-22. Leuven. Langfeldt J.A. 1994, Recent discovered early Christian Hansbury M. 2015 , Isaac the Syrian: the Third Part, monuments in Northeastern Arabia, «Arabian in: Kozah M., Abu-Husayn A., Al-Murikhi Archaeology and Epigraphy» 5, pp. 32-60. S.S., Al-Thani H. (eds.) 2015, An Anthology of Syriac Writers from Qatar in the Seventh Century, Morony M.G. 1982, Continuity and change in the ad- Piscataway NJ, pp. 281-440. ministrative geography of Late Sasanian and Early th Islamic al-‘Irāq, «Journal of Persian studies» 20, Healey J.F. 2000, The Christians of Qatar in the 7 cen- pp. 1-49. tury AD, in: Netton I.R. (ed.), Studies in honour of Clifford Edmund Bosworth, Volume ,I Leiden- Mourad S.A. 2014, Christianity in Arabia: An th th Boston-Köln, pp. 222-237. Overview (4 -9 centuries CE), in: Kozah et al. (eds.), pp. 37-60. Healey J.F. 2009, The patriarch Yšoyabh and the Christians of Qatar in the first Islamic century, in: Payne R. 2011, Monks, dinars and date palms: hagio- Hunter E.C.D. (ed.), The Christian heritage of Iraq, graphical production and the expansion of monastic in- Piscataway NJ, pp. 1-9. stitutions in the early Islamic Persian Gulf, «Arabian archaeology and epigraphy» 22, pp. 97-111. Holes C. 2002, Non-Arabic Semitic elements in the Arabic dialects of Eastern Arabia, in: Arnold W., Potts D.T. 1990, The Arabian Gulf in Antiquity, vo- Bobzin H. (eds.), Spricht doch mit deinen Knechten lume II, From Alexander the Great to the coming of aramäisch, wir verstehen es!, Wiesbaden, pp. 269-280. Islam, Oxford. Hoyland R.G. 1997, Seeing Islam as others saw it. Ramelli I. 2002, Il Chronicon di Arbela: presentazione, A survey and evaluation of Christian, Jewish and traduzione e note essenziali, Madrid. Zoroastrian writings on early Islam, Princeton NJ. Robin C.J. 2010, Nagrān vers l’époque du massacre: notes Hoyland R.G. 2001, Arabia and the Arabs, from the sur l’histoire politique, économique et institutionnelle Bronze Age to the coming of Islam, London. et sur l’introduction du christianisme, in: Beaucamp J., Briquel Chatonnet F., Robin C.J. (eds.), Kawerau P. 1985, (CSCO Syr. Die Chronik von Arbela Juifs et chrétiens en Arabie aux Ve et Vie siècles: regards 199-200), Leuven. croisés sur les sources, Paris, pp. 42-106. Kennett D. 2007, The decline of eastern Arabia in Salles J.F., Callot O. 2013, Les églises anti- , «Arabian Archaeology and the Sasanian period ques de Koweit et du golfe Persique, in: Briquel Epigraphy» 18, pp. 86-122. Chatonnet F.(ed.), Les églises en monde syriaque, Kessel G. 2015, Isaac of Niniveh’s Chapters on Paris, pp. 237-268. Knowledge, in: Kozah et al (eds.)., pp. 253-280.

West & East 382 Monografie, 4 Christianity in the Gulf: vestiges of the East Syrian presence in late antiquity

Schall A. 1989, Der nestorianische Bibelexeget Īšō’dāḏ Steve M.-J. 2003, L’île de Khārg, une page de l’hi- von Merw (9. Jh. N. Chr.) in seiner Bedeutung für die stoire du Golfe persique et du monachisme oriental, orientalische Philologie, in: Macuch M., Müller- Neuchâtel. Kessler C., Fragner B.G. (eds.), Studia semitica Trimingham J.S. 1979, Christianity among the Arabs in necnon iranica: Rudolpho Macuch septuagenario ab Pre-Islamic times, London. amicis et discipulis dedicata, Wiesbaden, pp. 271-282. Wilmshurst D. 2011, The martyred Church: a history Scher A. 1910, Histoire nestorienne (Chronique de of the Church of the East, London. Séert), première partie II, Paris.

West & East 383 Monografie, 4

The Ubaid in the Gulf: compositional analysis of ceramic material (sixth–fifth millennium BC)

ELEANOR LUCY PRESTON

UCL Qatar and UCL Institute of Archaeology

Abstract

Maritime trade in the Persian Gulf has been a pivotal feature of world trading systems from the Bronze Age to the present. Understanding its origins in prehistory is essential for any broader interpretation of early social and economic developments between the Indian Ocean and the Mediterranean. The key period in this respect is the Mid-Late sixth millennium BC, when material culture of Mesopotamian origin is first documented in the Gulf. The material remains are in the form of potsherds from the southern Mesopotamian Ubaid tradition, which have been found as far as the Straits of Hormuz, 900 km from Iraq. Chemical analysis, conducted on the Ubaid-ware found in Arabia, determined that these potsherds originated from southern Mesopotamia and were not local imitations. These Ubaid-ware sherds, alongside a coarse-ware, believed to be local, are the earliest ceramic material found in Arabia. However, after the Ubaid period came to an end by the end of the fifth millennium BC, ceramic artefacts are not seen again in Arabia for nearly a millennium. Therefore this early appearance of ceramics in the Gulf region takes on a central role in approaching the wider questions of cross- cultural interaction. However, questions concerning provenance remain. This paper details ICP-AES analysis conducted on both the Ubaid-ware and the local ware with the aim of furthering our understanding of the exchange relationship between Mesopotamia and Arabia and putting the Ubaid within the context of the Ara- bian Neolithic nomadic population.

Keywords

Mesopotamia, Arabia, Iran, Persian Gulf, Ubaid, Neolithic, Ceramics, ICP-AES

West & East 385 Monografie, 4 Eleanor Lucy Preston

1. Introduction 40 years since the NAA study, our understanding of this material remains inadequate as little new In the 1960s prehistoric ceramics were found by research has been conducted and for many years Grace Burkholder on a coastal site in Saudi Ara- study of the Ubaid in Arabia was marginalised, but bia and were identified by Geoffrey Bibby as Ubaid in more recent years, research into the Neolithic in (fig. 1).1 the region has progressed with re-examination of Given that Ubaid pots originate in Iraq, this sites and remains.3 The work discussed in this pa- discovery was surprising. Since this initial discov- per is part of this reassessment. ery more than 60 sites have now been identified on the Gulf coast as having Ubaid potsherds and constitute evidence of connectivity in the sixth- 2. The Ubaid in Iraq fifth millennium BC from Kuwait to the Straits of Hormuz (Map 1). Neutron activation analysis The term ‘Ubaid’ is used for a wide ‘interaction (NAA) on the sherds revealed that they were im- zone’ in west Asia,4 found from west Syria to the ports from southern Iraq and not local imitation,2 Caucasus, through Iraq into west Iran and the however how and why they were transported to Persian Gulf (Map 2). the Gulf could only be speculated on. Now, near- The Ubaid in Mesopotamia begins in southern ly 50 years after the first potsherd was found and Iraq in the seventh millennium BC (Ubaid 0 and

Figure 1 – Ubaid potsherds from Dosariyah (Saudi Arabia) (Frifelt 1989, p. 406)

3 Beech et Al. 2016; Drechsler 2012; Carter, 1 Burkholder 1972. Crawford 2010. 2 Oates et Al. 1977. 4 Carter, Philip 2010; Henrickson, Thuesen 1989.

West & East 386 Monografie, 4 The Ubaid in the Gulf: compositional analysis of ceramic material (sixth–fifth millennium BC)

Map 1 Ubaid and Ubaid-related sites around the Gulf coast (Carter, 2018 In Press)

Map 2 The Ubaid-horizon in west Asia (Adapted from Carter, Philip 2010, p. X)

West & East 387 Monografie, 4 Eleanor Lucy Preston

Figure 2 Ubaid pots from Eridu (Iraq) (Oates, 1960 Plate V)

Ubaid 1), but with influences from central and and design motifs; e.g. bands of paint, zigzags, chev- northern Iraq, most notably the Samarran culture.5 rons and reverse decoration and vessel shape domi- By Ubaid 3 (approx. fifth millennium BC) Ubaid- nated by open forms such as bowls, often with cari- style material culture was adopted by local groups nation (fig. 2).8 throughout Iraq and as far as west Syria and west This ceramic ware has been divided into six Iran.6 Evidence of connectivity across the region chronological subgroups based on continuity and can be seen in the movement of materials (e.g. ce- change in vessel shape and decoration, predomi- ramics and obsidian) and ideas (iconography and nantly observed at Eridu where the material has technology).7 The archaeology of Iraq in the sixth been found in well-defined stratigraphy.9 The and fifth millennium BC is one of change: when the changes from one style to the next were gradual foundations of the city states of the third millenni- and overlapped and, as with the other material re- um BC were established and when social stratifica- mains, because this buff-ware is found over a wide tion began to develop. The people of Mesopotamia region it displays differences in dominant vessel were settled in agricultural communities practicing shape, as well as variation in fabric and paint col- rain-fed farming (northern Iraq) and irrigated farm- our.10 Ubaid 0 is so far only found in southern Iraq ing (southern Iraq). while Ubaid 1-4 are more widespread and Ubaid 5 Ubaid is used for a ceramic type, usually de- (Terminal Ubaid) is a transitional type linking to scribed as a black-on-buff ware, with distinct shapes the next period.11

5 Oates 1960, pp. 42–44. 8 Oates 1960, 1982, 1987; Masry 1997, pp. 82–84. 6 Henrickson, Thuesen 1989; Carter, Philip 9 2010. Oates 1960. 10 7 Carter, Philip 2010; Henrickson, Thuesen Ibid.; Oates 1980, 1987; Masry 1997, pp. 82–84. 1989. 11 Oates 1960.

West & East 388 Monografie, 4 The Ubaid in the Gulf: compositional analysis of ceramic material (sixth–fifth millennium BC)

However, there is debate among archaeologists cent of that found in southern Iraq are referred to concerning the definition and use of the Ubaid as Ubaid-related.16 term both as a chronological label and as a cultural identifier. In this study Ubaid is used in reference to what is argued by many to be the core and origin of 4. The Ubaid in Arabia the Ubaid: southern Iraq. This definition includes the pots but also tripartite architecture, irrigated The archaeology of Arabia’s Gulf coast in the sixth farming, sedentism, other ceramic material (tools and fifth millennium BC is in contrast to that of and figurines) and personal adornments found on southern Iraq. This period is named the Arabian sites across southern Mesopotamia from the Mid- Middle Neolithic and while work continues far less seventh millennium BC until the end of the fifth is known about it than that of the neighbouring re- millennium BC.12 gions for the same time period. The Arabian Mid- dle Neolithic is characterised by mobile populations engaging in hunting, gathering, fishing and herd- 3. The Ubaid in Iran ing17 who had a visible lithic tool industry which is known as the Arabian bifacial tradition with a re- The archaeology of Iran in the seventh-fifth millen- gional type known as Qatar A, C and D.18 There nium BC is multifaceted and its chronology com- is evidence of extensive exploitation of marine replicated with many local labels used to define both sources including gathering molluscs from the in- material culture and time periods. These labels of- tertidal zone, fishing (line and net) and hunting ten overlap.13 Ubaid material did not penetrate far sea mammals.19 Domesticated species of cattle, goat into Iran but the relationship between Mesopota- and/or sheep are thought to have arrived in the re- mia, the Susiana plain and the Iranian Gulf Coast gion seventh–fifth millennium BC.20 The degree of must be considered when discussing the Ubaid: its mobility is debated but generally it is believed the extent, its influence and mechanism of connectiv- people were not highly mobile, moving seasonally ity. The term Ubaid is infrequently used in Iranian around the landscape.21 There is also evidence that archaeology, instead using local terminologies (e.g. some members of the population did live in perma- Middle Susiana and Bakun)14 in order to avoid as- nent settlements, primarily on the coast, and they sumptions of a core and periphery, but that can ob- may have done so in order to exploit particular nat- scure evidence of connectivity. The Iranian Ubaid ural resources while the remainder of their social/ is complex and while there is contact with southern cultural group moved seasonally to hunt, access Iraq there are also influences from within Iran, most other resources, graze their herds, and to follow the notably the Zagros Highlands and the Inland Fars.15 more clement weather. 22 There is as yet no evidence While there is a different material culture in Iran of agriculture in Arabia for this period. Carbonised than Mesopotamia in this period, there are some date stones have been found 23 suggesting there may shared characteristics including subsistence econo- have been some management of plant resources but my based on farming and herding. The ceramics are the species are not domesticated and farming (irri- a buff-ware with dark painted decoration and can be included in the Ubaid-horizon, however links have also been made with Bakun ware and material from 16 Ibid. the Susiana Plain and therefore ceramics reminis- 17 Potts 1990, pp. 28–58. 18 de Cardi 1978. 19 12 Carter, Philip 2010; Henrickson, Thuesen Masry 1997; Carter, Crawford 2010. 1989. 20 Drechsler 2012. 13 Ehrich 1992. 21 Uerpmann et Al. 2008. 14 Carter et Al. 2006, pp. 16–17; Ehrich 1992. 22 Lézine et Al. 2010. 15 Carter et Al. 2006. 23 Gros-Balthazard et Al. 2017.

West & East 389 Monografie, 4 Eleanor Lucy Preston

Figure 3 Ubaid potsherds from Kuwait, the UAE and Saudi Arabia (Beech 2013, p. 36; Carter, Crawford 2010, p. 343, Plate 15; Drechsler 2012)

Figure 4 Example of Arabian red coarse-ware (Carter, Crawford 2010, p. 343, Plate 14)

gation or rain-fed) is unknown.24 There is evidence to surface scatters, some of which cover a wide area. of access to other resources most notably bitumen, They share a material culture, including lithic tech- obsidian, haematite and copper.25 nology, faunal remains and personal adornment On the Arabian coast and the islands of the Gulf items, which coincides with the overall Arabian more than 60 sites have now been classified to the Middle Neolithic (sisxth-fourth millennium BC).27 Ubaid tradition due to the presence of Ubaid-ware The Arabian potsherds are predominately from potsherds (fig. 3); little to no other material has Ubaid 2, Ubaid 2/3 and Ubaid 3. As of yet no Ubaid been conclusively shown to be southern Iraq mate- 0 or Ubaid 1 has been found and only a limited rial culture.26 The sites range from small mound sites amount of Ubaid 4 and 5. These potsherds, alongside radio-carbon dating, place these sites within the 24 Drechsler 2012. 25 Connan et Al. 2005; Drechsler et Al. 2013; Drechsler 2012. 27 Potts 1990, pp. 28‑58; Carter, Crawford 2010; 26 Carter 2018, 2002. Bielinski 2015; Masry 1997; Drechsler 2012.

West & East 390 Monografie, 4 The Ubaid in the Gulf: compositional analysis of ceramic material (sixth–fifth millennium BC)

Late sixth to Mid-fifth millennium BC.28 The Ubaid not originate from southern Iraq.34 The chemical material are not local imitations but foreign intru- analysis, along with the fact that the Arabian red sion. Joan Oates et al conducted a NAA on a sample coarse-ware has only been found on Arabia’s Gulf and demonstrated that the Ubaid potsherds found coast, has led to the conclusion that it is local to in the Gulf originated from southern Iraq, notably Arabia and was manufactured there. This propos- the site of Ur.29 Roaf and Galbraith reassessed the al is generally accepted by archaeologists, although results and argued that the exact origin site of the no clay source has as yet been identified as a poten- sherds could not be determined due to the homog- tial source nor has any other evidence of manufac- enous nature of southern Iraq’s alluvial soil, but did ture (e.g. kilns or wasters). It is also not known if agree with Oates et al’s general conclusion concern- this ware was produced at one site and then trading southern Iraq as the place of origin.30 ed outwards along the Gulf coast (perhaps like the A red coarse-ware has also been found in the Ubaid pots) or if the Arabian pots were made at Gulf region (fig. 4) and it is of note for two reasons: multiple sites along the coast, where there was con- 1, it has so far only been found on sites that also have tact with Mesopotamians, and was only transport- Ubaid-ware sherds and 2, when the Ubaid intrusion ed short distances. It is also not clear who is making ends so too does the appearance of this coarse-ware these pots: the Arabians or the Mesopotamians. The and Arabia returns to an aceramic culture for near- Mesopotamians clearly had the technology and may ly a millennium and Mesopotamian material is not have passed this knowledge on. However, if this was found again until late into the Late Chalcolithic the case it may be argued that other, non-Ubaid- (fourth millennium BC).31 related Arabian Middle Neolithic sites would also The Arabian red coarse-ware is in stark con- have had ceramics, but this is not the case. Therefore trast to the Ubaid-ware. Where the Ubaid is fine while the Arabian-ware is part of the local material and carefully made the coarse-ware is thick walled, culture it is intrinsically linked with the foreign in- roughly made and the fabric rich in inclusions. The trusion of material from southern Iraq and must be fabric is usually a red/orange colour and frequently considered as a separate category or risk missing cru- has a blackened core. The shape typology is also dif- cial information related both to the macro and mi- ferent from the other ware; while the Ubaid pots cro scales of connectivity. are dominated by open vessel forms, the Arabian Other important finds associated with the pots appear to be more suited for cooking.32 Deco- Ubaid-ware in Arabia are evidence related to sea- ration of the coarse-ware is rare and when it does oc- faring vessels. These finds are predominantly from cur it is not necessarily painted but rather impressed H3 (As-Sabiyah) in Kuwait (see below). At H3 bi- and appliqué techniques are used instead. Some of tumen pieces with reed impressions on one side and the decorations appear to emulate Ubaid design but barnacles encrusted on the other have been inter- too few examples exist to conduct statistical analysis preted as caulking for seaworthy boats (fig. 5). or typological comparisons.33 These finds, alongside the fact that nearly all the Previous research on the Arabian coarse-ware Ubaid-related sites are coastal and include off-shore includes chemical analysis which showed it is not islands, have led to the conclusion that the Ubaid chemically similar to the Ubaid-ware and, despite pots were moved around the Gulf, at least in part, only ever being found on Ubaid-related sites, did via the sea.35 Other evidence for seafaring can be seen in the representations of boats in both pictorial (fig. 6) 28 Frifelt 1989; Hermansen 1993; Carter, Craw- ford 2010. and model form found at H3 (fig. 7), and also 36 29 Oates et Al. 1977. Dosariyah. 30 Roaf, Galbraith 1994. 31 Hermansen 1993. 34 Oates et Al. 1977. 32 Smogorzewska 2013, 2016; Carter 2010. 35 Carter, Crawford 2010. 33 Smogorzewska 2013, 2016. 36 Carter 2012, 2018.

West & East 391 Monografie, 4 Eleanor Lucy Preston

Figure 5 Bitumen remains found at H3 (Kuwait) interpreted as caulking for seafaring vessels (Carter, Crawford 2010, p. 351, Plate 28)

Figure 6 Figure 7 Reworked potsherd with boat painting Boat model from H3 (Kuwait) from H3 (Kuwait) (Carter, Crawford 2010, (Carter, Crawford 2010, p. 350, Plate 24) p. 351, Plate 27)

West & East 392 Monografie, 4 The Ubaid in the Gulf: compositional analysis of ceramic material (sixth–fifth millennium BC)

5. Importance period as it appears to be them utilising the pots.40 As with the macro and meso scales, the Ubaid ma- In the first millennium BC a trade network from terial in the Gulf may be used to identify connec- China to Europe, encompassing India, north Africa tivity this time via the movement of the potsherds and Arabia was in existence. This network did not between smaller geographical areas. spring up from nowhere. Over generations contacts However, a better knowledge of this material has were made and spread outwards till they encoun- implications beyond connectivity and trade. For the tered other expanding networks, creating a vast communities of Arabia with access to this foreign, web.37 The foundations of this network lie in deep exotic traded item it may have represented associa- prehistory and to understand the establishment of tion with ideas of prestige, linked to display and sta- this network, it is important to trace back in time tus. Social organisation of the people of the Arabian to these foundations. Ubaid material in the Gulf is coast in the Arabian Middle Neolithic is unknown important because it constitutes some of the earliest as evidence is scarce. If the Ubaid pots were traded evidence for far-reaching connectivity38 and there- for their own sake (which seems likely given their fore represents the foundations of the later network open forms and decoration)41 then they may be con- that connected the known world. sidered as evidence of some social stratification. The In addition to implications of prehistoric con- reasons for feasting, prestige and display are many nectivity on the macroscale the research can influ- but all are concerned with social structure, group ence interpretation of archaeological remains on identity and group inter and intra-relations.42 the meso and micro scales also. Mesoscale: the presence of Ubaid pots from southern Iraq in the Gulf provides information on 6. Current models connectivity within the region and is some of the earliest evidence of contact in the region that later The models proposed to explain how the Ubaid facilitated the connections between Mesopotamia, pots arrived in the Gulf can be placed into three cat- Dilmun (Bahrain) Magan (Oman) and Meluhha egories: mobility, ‘luggage’ and trade. (India). Ceramics and other remains (obsidian, bitumen and copper) help to reconstruct these interactions using chemical analysis and petrography.39 6.1. Mobility models The Ubaid’s association with the coast and the evidence of seafaring vessels also opens up the possibil- In the 1970s Abdullah Hassan Masry conducted ex- ity for archaeologists to understand the cultural re- tensive field research in the north-eastern regions of lationships with the sea and how this facilitated or Saudi Arabia. His work, Prehistory in Northeastern restricted connectivity. Arabia: The problem of interregional interaction 43 Microscale: the Arabian Middle Neolithic is not (1974, updated and reprinted 1997), presented prominent in the archaeological record and while the Arabian Neolithic in a new light and proposed research continues, knowledge of this period is not that the Ubaid material from Iraq was arriving by very great. Therefore a better understanding of the direct movement of people across the land in large Ubaid in the Gulf will contribute directly to the in- cyclical migrations that covered the Gulf coast from terpretation of the Arabian Middle Neolithic as the Iraq to the Straits of Hormuz. His model uses the reason for the presence of the Ubaid material must theory of interaction spheres to explain that Ara- have its roots in the cultural groups of Arabia in this bia’s nomadic population movements brought 40 Carter 2010. 37 Bentley 1993. 41 Ibid.; Carter, Crawford 2010. 38 Carter 2018, 2010. 42 Dietler, Hayden 2001; Bliege Bird, Smith 39 Weeks 2004; Connan et Al. 2005; Van de Velde 2005. 2015. 43 Masry 1997.

West & East 393 Monografie, 4 Eleanor Lucy Preston

them from southern Iraq, where they encountered settling at midway points in order to facilitate con- ceramics and domesticated herding animals, to the tact. While the subsistence economy at both sites is shores of the sea in the central Gulf region. This one of herding, fishing, hunting and gathering like means that the Ubaid potsherds were being traded/ that of other Arabian sites (with no evidence of ex- exchanged directly and then carried over great dis- tensive crop cultivation) there are remains that do tances by land. fit Mesopotamian material culture (e.g. jewellery Masry’s proposal that the Ubaid material arrived and body adornment items) and both sites also have in Arabia due to population movement via the land permanent, stone architecture which is unusual in can partly be discounted due to newer evidence of the Arabian Neolithic.47 Of particular interest is the seafaring, however when considering the mecha- building remains at Bahra 1 which are not rounded, nisms involved in Gulf connectivity during this pe- cellular structures, but square/rectangular buildings riod, interaction spheres, migration and nomadism reminiscent of Ubaid architecture. To call them tri- must all be considered and therefore Masry’s mod- partite at this stage may be premature, however an el is still one of importance when considering the Ubaid influence cannot be ruled out.48 Ubaid-phenomenon. Both Masry and Piesinger’s models share a view C.M. Piesinger’s proposal44 has many similari- of direct contact and exchange and despite their ties with Masry but for one crucial difference: the limitations, cannot be discounted with current lev- contact made between the populations of southern els of knowledge and both offer ideas that remain Mesopotamia and Arabia was via the sea and not important when considering the Ubaid material in the land. This proposal was based on the distribu- the Gulf. tion of the Ubaid potsherds along the Gulf coast. Since Piesinger wrote her thesis more evidence has been found confirming the presence of seafar- 6.2. ‘Luggage’ model ing vessels during this period.45 Piesinger also offers an additional note on her model of the movement After her chemical analysis of the Ubaid sherds,49 of material within the Gulf and that is concerning Joan Oates (and later Potts) proposed that the the navigation of coastal waters, which she believes Ubaid pots were being brought and used by Mes- must have been developed in the Gulf by those liv- opotamians and then discarded as they broke. She ing there, making them the primary agents of the proposed that the reason ceramics are not found movement of material. Also people in the Gulf at elsewhere in Arabia or after the Ubaid period for this time, were more likely to have come from the 1000 years is that it was not the Arabian popula- Arabian coast and not the settled populations living tion making the Arabian red coarse-ware, but the in the alluvial marshes of Southern Iraq.46 visiting Mesopotamians who were making cook- Both these mobility models share the same flaw: ing pots as they needed them. They did not teach or the lack of other Mesopotamian material remains in transfer the technology to the locals. Oates argues the Gulf. If, as part of their migration, groups were that the chemical data proves a one-sided relation- moving out of southern Iraq more Ubaid material ship between the Mesopotamians and Arabians, but culture would be expected to be found in the Gulf the evidence is not adequate to prove this. More re- and not just on coastal sites. However, this theory cent work at Bahra 1 and H3 (Kuwait)50 in fact sug- of a mobile group with direct contact with south- gests the opposite is true: there was a relationship ern Iraq cannot be completely discounted in the up- between the two groups, one of interaction and ex- per Gulf. H3 and Bahra 1 may be evidence of move- change (goods and ideas) and that the Mesopotami- ment of both Arabian and Mesopotamian peoples, 47 Carter, Crawford 2010; Bielinski 2015. 44 Piesinger 1983. 48 Bielinski 2015. 45 Carter 2006. 49 Oates et Al. 1977. 46 Ibid.; Piesinger 1983. 50 Carter, Crawford 2010; Bielinski 2015.

West & East 394 Monografie, 4 The Ubaid in the Gulf: compositional analysis of ceramic material (sixth–fifth millennium BC)

Figure 8 Figure 9 Plaster pieces from Dalma Island (UAE) with painted A small, closed form Ubaid pot found on MR11 decoration appearing to emulate Ubaid-ware motifs (UAE) (Beech 2013, p. 40) (Beech 2013, p. 39)

ans did not just arrive on the Arabian shores, carry their own use, it would explain the establishment of out their activities, discard their rubbish and leave. temporary, transitory settlements. However, if this At the time of publication (1977) Oates also argued was the case we would expect to find other Ubaid that there was no demand for a pottery trade as material culture at the Ubaid-related sites beyond most pottery in Mesopotamia was locally made (al- the ceramics and that is not the case. There are also though she concedes the possibility of itinerant pot- sites with complex stratigraphy (e.g. Dosariyah)53 ters) and that if there had been a demand we would and others with permanent stone architecture (e.g. expect to find imitations of Ubaid-ware. In the cen- H3)54 that suggests a more complex use and occupa- tral/lower Gulf (e.g. Dalma, UAE) such imitations tion/reoccupation of the sites. have now been identified in the form of white plas- Oates’ ‘luggage’ model (as with Masry’s) has be- ter vessels with black painted decoration (fig. 8).51 Is come outdated by more recent archaeological evi- this then evidence for a demand in Ubaid-ware? dence, but also, like Masry, there are important fea- While Oates’ suggestion that there was no possi- tures of the archaeological record that this model bility of a trade relationship has become outdated,52 highlights which must be considered. her model has a number of benefits in that it addresses issues surrounding the Ubaid material that other models do not. Firstly, it does explain the return to 6.3. Trade models an aceramic culture after contact with Ubaid Meso- potamia ends as ceramic technology was not passed Although it is not a new model, proposals centring on. Secondly, it does address the issue of the major- on trade in the Gulf have become the most promi- ity of sites being coastal. Finally, few of the Ubaid- nent and, like Masry and Pisesinger, this model sug- related sites have evidence of previous or post oc- gests direct contact. Initially it was considered pos- cupation to the Ubaid period. If the material was sible that it was the contents of the pots that were being brought by the Mesopotamians as luggage for being traded and the pots themselves were almost

51 Beech 2013. 53 Drechsler 2012. 52 Oates et Al. 1977, pp. 232–234. 54 Carter, Crawford 2010.

West & East 395 Monografie, 4 Eleanor Lucy Preston

incidental. However, studies of the pots’ function- in the lower Gulf is down the line, why stop at the al profiles and shapes indicate that the majority Straits of Hormuz? There is evidence of contact be- of the Ubaid-ware pots were open forms, none of tween the coastal people of both the lower Gulf and which are particularly large in size (fig. 9).55 While the Gulf of Oman,59 but trading pots is not part of this does not exclude the possibility of their use for their exchange network. This feature, coupled with containing luxury goods that may only be traded in the open shapes of the pots and the location of the small quantities, based on their typology, it is more sites near to the coast may mean a complex interac- likely that the pots were traded for their own sake as tion system is being undertaken between the direct tools for feasting, display and prestige. contact of the Mesopotamians and the Arabians: it Carter56 goes further with his model suggesting is not just the pots that are valuable, but the people that in the upper Gulf (Kuwait to Qatar) contact trading them and their relationship to the sea are was made directly by the populations of Arabia with also important and thus a highly ritualised mecha- those in Mesopotamia. The presence of mixed mate- nism of contact manifests. rial culture at two sites in Kuwait (H3 and Bahra 1) indicates a complex relationship that is not necessarily dominated by the Mesopotamians and could mean 7. Research questions and hypotheses that the boats themselves were run out of these sites (and not southern Iraq) and H3 has substantial evi- This study’s research questions are centred on ori- dence of sea worthy vessels.57 This model places agen- gins and connectivity and they are: cy with both cultural groups and the Kuwait sites may 1. Where did the ceramic material originate? represent an interfacing zone of interaction spheres. 2. Can production centres be identified for both However, Carter argues that the situation in the low- Ubaid-ware and the local coarse-ware? er Gulf is different as there has so far been far fewer potsherds and Ubaid-related sites found between 3. Did it come from multiple locations or just one? Qatar and the Straits of Hormuz. Instead of direct 4. How did both wares spread within the Gulf and trade with the Mesopotamians for their ceramics, are they distributed differently? Carter’s model proposes down-the-line trade as the mechanism for the movement of ceramics into this It is clear from previous work on the Ubaid-related region, which may also explain the presence of imita- sites in the Gulf region that there is no clear answer tion plaster vessels because, while there was demand, to the reason for the pots’ presence nor the mecha- the supply was less consistent than in the north. nism of their arrival. Based on previous discussion The trade model currently utilises more of the on the material I believe that a more complex situ- archaeological data than the other models, while ation existed than has so far been proposed. Cart- also incorporating Masry’s interaction spheres and er’s trading model goes the furthest in this com- Piesinger’s mobility via the sea. However, there is plexity, however there is still more data that must still much to consider when interpreting the pres- be addressed. I believe that the lack of Ubaid sherds ence of the Ubaid in the Gulf. The trade model uses found beyond the Straits of Hormuz is significant feasting and prestige as an explanation for the de- and the fact that most of the Ubaid-related sites are mand in ceramics, but signalling theory (including coastal is also crucial. These features make contact feasting and prestige) is complex and can be accom- with the sea/water important and therefore it is di- plished in different ways to achieve different goals rect contact with people from the sea that moves (some of which are contradictory).58 If the trade the material around all areas of the Gulf. While this could mean Arabians in the lower Gulf, it could also be Mesopotamians who were stopped by the Straits 55 Carter 2010; Smogorzewska 2013. of Hormuz and the more dangerous waters of the 56 Carter 2006; Carter, Crawford 2010. 57 Carter 2006. 58 Bliege Bird, Smith 2005; Dietler, Hayden 2001. 59 Charpentier, Méry 2008.

West & East 396 Monografie, 4 The Ubaid in the Gulf: compositional analysis of ceramic material (sixth–fifth millennium BC)

Omani Gulf: the sea itself was the barrier, not the and the Iran material by the Ashmolean Muse- land (the Qatar peninsula). um (Oxford) as part of the Williamson collection However, interaction spheres and down-the- housed at the time of sampling at Durham Univer- line trade were still likely present. Large sites with sity. 180 samples were taken from across the six se- repeat occupation (e.g. Dosariyah) may have been lected sites (see below) and sherds chosen to reflect where the people carrying the pots for trade landed the character of the assemblage in relation to quan- and took part in exchange with waiting locals. From tities of each type (Table 1). However in the case of these large sites the pots may then have been traded the Arabian red coarse-ware there was limited avail- to smaller, local camp sites both via land and sea, e.g. ability so all available sherds were sampled. an area of approximately 30-40km around Dosari- The methodology used to study this material yah contains at least 10 surface scatter sites with ma- was the bulk chemical analysis technique of induc- terial that matches the typology of Dosariyah.60 tively coupled plasma atomic emission spectrom- Therefore I propose that the material was etry (ICP-AES). ICP-AES records data for Al, Fe, moved around all parts of the Gulf via the sea di- Mg, Ca, Na, K, Ti, P, Mn as oxides and Ba, Ce, Co, rectly by the Mesopotamians. Additionally to this, Cr, Cu, La, Li, Ni, Pb, Sc, Sr, V, Y, Zn, Zr, in ppm. I propose there was some movement via the land It has become standard practice to run a statistical among the populations of Arabia, but only on a technique on such data that reduces the number of limited, local scale. variables without losing a significant amount of in- This study will use two techniques: bulk chemical formation especially concerning variability. Princi- analysis and petrography. This paper outlines the pro- pal component analysis (PCA) and cluster analysis gress made with the chemical analysis and future pub- are such techniques and this study will utilise IBM lications will address finds in the petrographic work. SPSS statistics 24 in order to carry out the statistical analysis.61 There is one main issue which is already known 8. Methodology and sampling due to the previous chemical analysis conducted on this material and that is that the south of Iraq, where Access was granted to the Iraq material by the Brit- Ubaid-ware is believed to originate, was a marshland ish Museum from their stores, the Kuwait materi- with alluvial soils deposited from both the Tigris al by Dr. Robert Carter who oversaw their excava- and Euphrates rivers and as such the soils/clays in tion, the Saudi Arabian material by Dr. Joan Oates southern Iraq are homogenised and it is thought dif- while at the McDonald Institute (Cambridge) ficult to differentiate. This may mean, as with Oates

Table 1 Sample data including site and ware information

60 Masry 1997, pp. 84–87. 61 Baxter 2015.

West & East 397 Monografie, 4 Eleanor Lucy Preston

et al’s NAA study, that it is impossible to identify The Kuwait site sampled, H3 is one of only a production centres due to the clays in southern Iraq handful of Arabian Chalcolithic sites with perma- being too similar to each other due to the deposi- nent, stone structures (fig. 10). It is of particular im- tion of the two rivers. That being said this research portance as it is at H3 that the bulk of the evidence is using a different technique and if production cen- for maritime activity is found. Of note is the nearby tres cannot be identified the new chemical data may site Bahra 1 which is also Ubaid-related and also has be sensitive enough to provide insight into different stone structures but unlike H3 they are rectangular aspects of the movement of the material. in character and described by the excavators as possibly being tripartite in design, a feature of southern Iraq Ubaid settlements.65 9. The sites Two sites were sampled from Saudi Arabia, Do- sariyah and Abu Khamis. Dosariyah has recently Material was sampled from six sites around the Gulf undergone re-excavation which showed the site to Coast from Iraq, Kuwait, Saudi Arabia, and Iran consist of complex stratigraphy, but little evidence (Map 3). of structures was found and only a limited number The Iraqi sites Eridu and Tell al-Ubaid were of permanent installations and therefore is charac- selected for two reasons; firstly, previous chemi- terised as a campsite with occupation and activity cal analysis conducted on the ceramic material zones moving over time. However, despite its size from the Gulf suggested that the Ubaid potsherds and complexity, Dosariyah’s material culture is in came from southern Iraq.62 The results from the keeping with the broader Arabian Middle Neolith- NAA led to the proposal that Ur was the most ic.66 Therefore Dosariyah, along with Abu Khamis, likely site of origin for the Ubaid-ware, how- are good representations for Ubaid-related sites in ever these results were further analysed and the Saudi Arabia.67 conclusions drawn were that while it did appear Finally, the site H200 on the Bushehr peninsula, that the Ubaid-related finds were from south- Iran, was also sampled as Chalcolithic Iran had con- ern Iraq, it was impossible to be more specific as tact with southern Iraq via Susiana and historically to the origin point.63 Since this second publica- there has also been contact across the Gulf to the tion no further chemical analysis on the mate- opposite shore where Dosariyah and Abu Khamis rial has been published. Both Eridu and Tell al- are located. The inclusion of this material is in or- Ubaid are in the area identified by both papers der to investigate the relationship between the Ira- as the region most likely to be the source of the nian coast and southern Iraq further as well as in- potsherds. And secondly, both sites have potential vestigate the potential of contact with Arabia in as production centres due to a surface survey iden- this early time period. H200 itself was first identi- tifying a number of possible kiln sites, although fied in 1969/1970 and the ceramic material initially as of yet excavation of such features has not oc- classed as Ubaid, however beyond this little else is curred.64 known about either the site or the potsherds.68

65 Bielinski 2015; Carter, Crawford 2010. 66 Drechsler 2012. 62 Oates et Al. 1977. 67 Masry 1997, pp. 78–92. 63 Roaf, Galbraith 1994. 68 Williamson, Prickett 1970; Carter et Al. 64 Moore 2002. 2006.

West & East 398 Monografie, 4 The Ubaid in the Gulf: compositional analysis of ceramic material (sixth–fifth millennium BC)

Map 3 The Ubaid in the Gulf with sampled sites highlighted (Adapted from Carter 2018, in Press)

Figure 10 Stone architecture at H3 (Kuwait) (Carter 2018, in press)

West & East 399 Monografie, 4 Eleanor Lucy Preston

10. Results and discussion at least two fabrics and possibly two different clay sources were used for the Arabian ware. Of particu- Preliminary outputs from PCA are promising with lar interest are samples 11 and 12 from H3 which are regard to identifying groups within the material considerably closer to the Ubaid and H200 results that share a similar chemical signature. Figures 11 than the other Arabian sherds. This too will require and 12 depict the same data but coloured to reflect further investigation into its significance. ware type and site. Finally is the Ubaid-ware. This material is very In figure 11 there are three groups with only tightly packed, but there does appear to be a divide a small amount of overlap between two of them. within the group cutting it in half. On the lower- Firstly it is clear that the Arabian coarse-ware has left is H3 and Dosariyah and on the upper-right is a different chemical signature than all the Ubaid both the Iraq sites and Abu Khamis. It is too soon and Ubaid-related (H200) material. There is no to speculate on what this may indicate, but it may overlap but it does appear these sherds have con- mean that differentiating between southern Iraq siderable variability especially when compared to clays is possible to some extent. all the Ubaid material (Ubaid medium, Ubaid fine, Unfortunately the results for the cluster analysis Ubaid coarse). All the Ubaid potsherds are fairly on all the samples is not as promising. While there tightly grouped suggesting a very similar chemical are numerous ways to conduct cluster analysis71 the signature which is expected considering the nature results when applied to all the samples for this proof the fabric (fine) and the possible sources of clay ject often produced a ‘chain’ result (fig. 13) which is in southern Iraq. difficult to interpret or draw conclusions from. The H200 results are interesting for a number However, if the ceramic categories defined by of reasons. When H200 was first discovered it was macroscopic investigation are used to create sub- identified as Ubaid69 but later examination of the groups and cluster analysis is then applied some pat- ceramic material cast this into doubt and linked the terns do appear. Of note are the Arabian red coarse- material more closely with Highland Fars and rede- ware results (fig. 14). The dendrogram shows at least fined it as Ubaid-related.70 This study’s data shows two subgroups within the Arabian coarse-ware fab- that chemically H200 pots are different from south- ric, which is in agreement with the PCA results. ern Iraq Ubaid material. However, there is a partial overlap with the Ubaid group. This could be for a number of reasons: 1, it may simply be part of the 11. Conclusions and future work group’s internal variability which appears to be fairly large or 2, some of the potsherds collected from This research project is still in its early stages, how- H200 are in fact from southern Iraq in origin and ever some conclusions can already be drawn from represent a foreign intrusion in the collection. This the results. Firstly, the PCA results support the pre- feature of the data will require further investigation, vious analysis conclusions that the Ubaid-related using both statistical and petrographic analysis, in material in the Gulf is chemically similar to that in order to determine its nature. southern Iraq and therefore it is probable that the Figure 12 is the same results as figure 11 but col- Gulf Ubaid-ware originated from southern Iraq. oured for site rather than fabric type. This shows Secondly, it also supports the previous analysis of some other noteworthy patterns. the red coarse-ware as not being from southern Iraq. While the variability of the Arabian ware is great, Despite only ever being found in association with there does appear to be a divide within it between Ubaid-ware, their chemical signatures are very dif- the material from H3 and that from Dosariyah and ferent and were not made from the same clays. No Abu Khamis. While it is not conclusive, this suggests clay source has been identified in Arabia as a possible source, nor have any kilns been located, but 69 Williamson, Prickett 1970. 70 Carter et Al. 2006. 71 Baxter 2015.

West & East 400 Monografie, 4 The Ubaid in the Gulf: compositional analysis of ceramic material (sixth–fifth millennium BC)

Figure 11 Figure 13 PCA results Cluster analysis for all variables applied to all with fabric ware types ware data

Figure 12 PCA results for all variables with site data

Figure 14 Cluster analysis on Arabian red coarse-ware only

West & East 401 Monografie, 4 Eleanor Lucy Preston

in light of the chemical analysis and that this ware two samples that do not group with the other sam- type is only found in Arabia it is argued to be of lo- ples and the reason for this may be clearer in petro- cal manufacture and this study’s results in no way graphic analysis. Also, the PCA and cluster analysis contradict this proposal. of the coarse-ware suggests there were at least two Finally, the H200 potsherds, when first exam- clays and maybe more than one recipe used in the ined were identified as Ubaid, but a later macro- production of this local ware. Identifying these dif- scopic study cast doubt on this classification, linking ferences may be clearer in the petrographic analysis the material more closely to inland Iran rather than and this type of study may explain in what manner southern Iraq.72 The PCA analysis of the chemical they are different (e.g. use of temper). data supports this later analysis as the chemical sig- These early results illustrate the potential of this nature is not similar to the Ubaid-ware from south- data-set to further understanding of the Ubaid in ern Iraq and therefore, while there may be similari- the Gulf which in turn may affect interpretation of ties between the decoration on the potsherds, it is both the wider Arabian Middle Neolithic and also more likely to be of local manufacture and it is more the Ubaid in Iraq itself. Understanding the mecha- accurate to call it Ubaid-related rather than Ubaid. nisms of connectivity for this early period has reper- Future work will involve continued analysis cussions on interpretations of networks beyond the of the chemical data (using both PCA and cluster region of the time period. To understand how the analysis), but also petrographic analysis as it has be- great trading networks of later antiquity developed come common practice to complement this type of it is necessary to recognise their foundations and the chemical study with petrography. This work may Ubaid in the Gulf is a key step in identifying these shed light on some of the PCA results, particular- foundations and ascertaining how they grew into a ly in relation to the Arabian coarse-ware. There are vast network that connected half the world.

72 Carter et Al. 2006.

West & East 402 Monografie, 4 The Ubaid in the Gulf: compositional analysis of ceramic material (sixth–fifth millennium BC)

BIBLIOGRAPHY

Baxter M.J. 2015, Exploritory multivaraiate analysis in Carter R., Challis K., Priestman S.M.N., archaeology, New York. Tofighian H. 2006, The Bushehr Hinterland: Beech M.J. 2013, In The Land of The Ichthyophagi Results of the first season of the Iranian-British ar- – Prehistoric Occupation of the Coast and Islands chaeological survey of Bushehr Province, November- of the Southern Arabian Gulf : A Regional review, December 2004, «Iran» 44, pp. 63‑103. «Adumatu» 27, pp. 31‑48. Carter R., Crawford H. 2010, Maritime interac- Beech M.J., Strutt K., Blue L., Al-Kaabi A.K., tions in the Arabian Neolithic: evidence from H3, as- Omar W.A., Al-Haj El-Faki A.A., Lingareddy Sabiyah, Leiden & Boston. A., Martin J. 2016, Ubaid-related sites if the sou- Carter R., Philip G. (eds.) 2010, Beyond the Ubaid. thern Gulf revisited: the Abu Dhabi coastal heritage Transformation and integration in the late prehistoric initiative, «Proceedings of the Seminar for Arabian societies of the Middle East, Chicago. Studies» 46, pp. 9‑24. Charpentier V., Méry S. 2008, A Neolithic settlement Bentley J.H. 1993, Old World encounters: cross-cultural near the Strait of Hormuz : Akab Island, United Arab contacts and exchanges in pre-modern times, Oxford & Emirates, «Proceedings of the Seminar for Arabian New York. Studies» 38, pp. 117‑136. Bielinski P. 2015, Bahra 1 Excavations in 2013. Connan J., Carter R., Crawford H., Tobey M., Preliminary Report on the Fifth Season of Kuwaiti- Charrié-Duhaut A., Jarvie D., Albrecht P., Polish Archaeological Explorations, Kuwait City and Norman K. 2005, A comparative geochemical stu- Warsaw. dy of bituminous boat remains from H3, As-Sabiyah Bliege Bird R., Smith E.A. 2005, Signaling theory, (Kuwait), and RJ-2, Ra’s al-Jinz (Oman), «Arabian strategic interaction, and symbolic capital, «Current Archaeology and Epigraphy» 16, pp. 21‑66, doi: Anthropology» 46, pp. 221‑248. 10.1111/j.1600-0471.2005.00041. Burkholder G. 1972, Ubaid Sites and Pottery in Saudi de Cardi B. 1978, Qatar archaeological report : excava- Arabia, «Archaeology» 25, pp. 264‑269. tions 1973, Oxford. Carter R. 2002, The Neolithic origins of seafaring in Dietler M., Hayden B. 2001, Feasts : archaeological the Arabian Gulf, «Archaeology International» 6, and ethnographic perspectives on food, politics, and po- pp. 44‑47, doi: 10.5334/ai.v6i0.129. wer, London & Washington DC. Carter R. 2006, Boat remains and maritime trade in Drechsler P. 2012, Archaeological excavations at the Persian Gulf during the sixth and fifth millen- Dosariyah: Final reports 2010-2012. nia BC, «Antiquity» 80, pp. 52‑63, doi: 10.1017/ Drechsler P., Berthold C., al-Naimi F.A., S0003598X0009325X. Eichmann R. 2013, Ceremonial objects or household Carter R. 2010, The Social and Environmental items? Non-destructive μ-XRD2 and μ-XRF studies on Context of Neolithic Seafaring in the Persian Gulf, in: three Neolithic hematite axes from Qatar, «Arabian Anderson A,. Barrett J.H., Boyle K.V. (eds.), Archaeology and Epigraphy» 24, pp. 119‑124, doi: The Global Origins and Development of Seafaring, 10.1111/aae.12029. Cambridge, pp. 191‑202. Ehrich R.W. 1992, Chronologies in Old World archaeo- Carter R. 2012, Watercraft, in: Potts D.T. (ed.), A logy Volume 1 and 2, Chicago and London. Companion to the Archaeology of the Ancient Near Frifelt K. 1989, Ubaid in the Gulf area, in: East, Oxford, pp. 347‑372. Henrickson E., Thusesen I. (eds.), Upon this Carter R. 2018, Globalising Interactions in the Arabian foundation: the ’Ubaid reconsidered: proceedings from Neolithic and the ‘Ubaid, in: Boivin N., Frachetti the ’Ubaid Symposium, Elsinore 1988, Copenhagen, M.D. (eds.) Globalisation and the people without hi- pp. 405‑418. story, Cambridge, in press. Gros-Balthazard M., Galimberti M., Kousa thanas A., Newton C., Ivorra S., Paradis

West & East 403 Monografie, 4 Eleanor Lucy Preston

L., Vigouroux Y., Carter R., Tengberg M., Oates J., Davidson T.E., Kamilli D., McKerrell Battesti V., Santoni S., Falquet L., Pintaud H. 1977, Seafaring merchants of Ur?, «Antiquity» 51, J.C., Terral J.F., Wegmann D. 2017, The pp. 221‑234, doi: 10.1017/S0003598X00071738. Discovery of Wild Date Palms in Oman Reveals a Piesinger C.M. 1983, Legacy of Dilmun: the Roots of Complex Domestication History Involving Centers in Ancient Maritime Trade in Eastern Coastal Arabia in the Middle East and Africa, «Current Biology» 27, the 4th/3rd millenniums BC., Michigan. pp. 1‑8, doi: 10.1016/j.cub.2017.06.045. Potts D.T. 1990, The Arabian Gulf in antiquity: Henrickson E., Thusesen I. (eds.) 1989, Upon this Volume 1, Oxford. foundation: the ’Ubaid reconsidered : proceedings from Roaf M., Galbraith J. 1994, the ’Ubaid Symposium, Elsinore 1988, Copenhagen. Pottery and P-Values – Seafaring Merchants of Ur – Reexamined, Hermansen B.D. 1993, Ubaid and ED pottery from «Antiquity» 68, pp. 770‑782, doi: 10.1017/ five sites at Ain as Sayh, Saudi Arabia, «Arabian S0003598X00047463. Archaeology and Epigraphy» 4, pp. 126‑144, doi: 10.1111/j.1600-0471.1993.tb00046.x. Smogorzewska A. 2013, Pottery from Bahra 1 (Kuwait). New evidence for the presence of Ubaid Lézine A.M., Robert C., Cleuziou S., Inizan culture in the Gulf, «Polish Archaeology in the M.L., Braemer F., Saliège J.F., Sylvestre Mediterranean» 22, pp. 555‑568. F., Tiercelin J.J., Crassard R., Méry .S , Charpentier V., Steimer-Herbet T. 2010, Smogorzewska A. 2016, Local and imported potte- Climate change and human occupation in the Southern ry in the Neolithic Gulf: A new perspective from the , «Polish Archaeology in the Arabian lowlands during the last deglaciation and site Bahra 1 (Kuwait) Mediterranean» 25, pp. 595‑617. the Holocene, «Global and Planetary Change» 72, pp. 412‑428, doi: 10.1016/j.gloplacha.2010.01.016. Uerpmann M., Uerpmann H., Jasim S.A., Haendel M. 2008, Masry A.H. 1997, Prehistory in Northeastern Arabia: Neolithic use of space and enviroment at , in: The problem of interregional interaction, London & Jebel al-Buhais (Emirate of Sharjah, UAE) New York. Hellyer P., Ziolkowski M. (eds.), Emirates Heritage Volume Two: Proceedings on the 2nd Annual Moore A.M.T. 2002, Pottery kiln sites at al ’Ubaid and Symposium on Recent Archaeological Discoveries in the Eridu, «Iraq» 64, pp. 69‑77. Emirates, Al Ain, pp. 44‑53. Oates J. 1960, Ur and Eridu, the Prehistory, «Iraq» 22, Van de Velde T. 2015, Digging into the Ubaid pe- pp. 32‑50. riod bitumen from Dosariyah, «Proceedings of the Oates J. 1982, Ubaid Mesopotamia reconsidered, in: Seminar for Arabian Studies» 45, pp. 405‑416. Young T.C., Smith P.E., Mortensen P. (eds.), Weeks L. 2004, Early metallurgy of the Persian Gulf, The Hilly Flanks and beyond. Essays on the prehistory Boston. of Southwestern Asia, Chicago, pp. 251‑282. Williamson A., Prickett M. 1970, Survey of the Oates J. 1987, ’Ubaid chronology, in: Aurenche O., Persian Gulf Coast (Unpublished Ashmolean Evin J., Hours P. (eds.), Chronologies in the Near East Museum Archive), Oxford. (BAR Internatianl series 379), Oxford, pp. 473‑482.

West & East 404 Monografie, 4 Rescue archaeology in the sultanate of Oman: methods and solution strategies

CRISTIANO PUTZOLU, CARLO BAIONE, ENZO COCCA, SABATINO LAURENZA

Independent Researchers

Abstract

This paper aims at presenting the results of the topographic fieldwork of a team of professional archaeologists invited by the Ministry of Heritage and Culture of the Sultanate of Oman to excavate and survey three graveyards in the area of Sohar (Falaji as Souq, Wadi al Arad and Liwa) in 2014 and 2015. The construction of the Batinah Express Highway would have led to the destruction of hundreds of burial mounds, therefore the team developed a quick and accurate surveying strategy to document them properly: after a first “test” campaign using monoscopic photogrammetry, the team opted for 3D SfM photogrammetry using a completely open source workflow. This workflow required two surveyors on the field and in the IT lab to ensure the archaeologists updated ortho- photos and to update the 2D and 3D vector plans. To manage the huge mass of data coming from the field the team opted for QGIS and the plugin PyArchInit. The mix of surveying methodology and managing system developed on site allowed the team to document the numerous Stratigraphic Units produced during the excavation of hundreds of graves, and also proved to be very helpful as hermeneutic tool as shown in the case of the excavation of Grave 21.

Keywords

Digital Archaeology, Open Source Photogrammetry, Rescue Archaeology, Oman Archaeology, Open Source, PyArchInit, QGIS

West & East 405 Monografie, 4 Cristiano Putzolu, Carlo Baione, Enzo Cocca, Sabatino Laurenza

1. Introduction – Measures designed to contribute to the cost of restoring, repairing and renovating registered From 2014 to 2015 a team led by Sabatino Laurenza monuments deemed to be of special historical, was invited by the Ministry of Heritage and Culture artistic and scientific importance. of the Sultanate of Oman to excavate a great number of graves located in three different graveyards Considering that Oman witnessed a huge infra- whose integrity was threatened by the construction structural development work, especially in the last of the Batinah Express Highway.1 years, very scarce appear to be the rules about rescue No specific laws in the Sultanate of Oman de- archaeology. mand rescue archaeology, but nevertheless a Na- The Batinah Express Highway is a 265 km road tional Heritage Protection Law preserves national which connects the Muscat Expressway up to the heritage by protecting cultural properties and up- border of the United Arab Emirates. The highway grading public awareness on national heritage.2 This will be located inland around 20 km from the sea, law is applied to all kind of monuments and antiq- and during the construction works it has been di- uities on the Oman territory, as well as to “chattels vided in 6 Packages, each one managed by differ- of cultural properties”, which also include fossils, rare ent consultants and contractors. Among the other archetypes of fauna and flora, fragments of artistic international teams operating in the construction ruins, ancient coins, engraving and marks, manu- site of the Batinah Expressway Project, our team scripts and books, documents and print matter of was charged with the investigations of the Pack- special historic, artistic, scientific and literary value, age 5 in the Batinah plain, which included Sohar as well as traditional style furniture items, painted and Liwa inland areas. Package 5 covers a stretch earthenware, musical instruments, jewelry, precious of 41 km from the kilometer 180 to the kilome- stones and weapons. ter 221 mark of the Expressway; it begins roughly The cited Decree requires: 18 km inland from the coast after Wadi Haibi in the Wylayat of Sohar and ends in the Wylayat of – Nationwide overall inventory which is to be Liwa (fig. 1). concluded and updated; Our team was invited by the Ministry of Her- – Measures designed to prohibit all kinds of activi- itage and Culture, the costs of the project was fi- ties or actions that may modify, alter or in a way nanced by the Ministry of Transports and Com- tamper with the property, be it a monument or a munications with BOTEC (Bosphorus Technical mobile cultural property; Consulting Corporation) acting as Consultant and – Measures designed to restrict or prohibit the all the needs for our team was arranged on-field by export of mobile cultural properties, as well as Ferrovial Agroman and Federici Stirling Batco LLC their purchase and sale; acting as Constructors. This is an important point to be stressed out,

1 The authors of this paper wish to thank the Ministry of because it was clear from the beginning that our Transport and Communication and the Ministry of Heritage goal was the complete excavation of all the graves and Culture of Sultanate of Oman, in particular Adj. Director that were going to be destroyed by the construction General for Archaeology and Museums, Dr. Sultan Al Bakri, who followed all the fieldwork operations giving us the support works, and any other result would have been unsuc- needed. Also we want to thank the Contractors staff of the Fer- cessful. All the excavated sites were located with- rovial and Federici Stirling & Batco and the Consultant staff of in the projected roadway, which was a corridor of BOTEC Ace for their precious direct support on the field. Spe- cial thanks go to our colleague and friend Claudia Tomaselli for about 75 meters, but in hilly areas the corridor was the graphics for this paper and to Vittorio Lauro for the excel- wider as the hills should have been blasted away in lent work done with the 3D rendering of Grave 21. Finally we order to create a correct slope. The expressway was would like to thank two anonymous reviewers for their useful feedback. fenced off with a 120 meters wide corridor, and the 2 Enacted by the Royal Decree No: 6/80, dated 10 Feb- burials within this zone would not be damaged by ruary 1980. the construction of the road. Outside those fenc-

West & East 406 Monografie, 4 Rescue archaeology in the sultanate of Oman: methods and solution strategies

Figure 1 The Batinah Express highway project Package 5

Figure 2 The different areas of operation

es there was the 200 meters wide Right Of Way investigated the Graveyards of Falaji as Souq, Wadi (ROW), where the burials were not harmed as well. al Arad and Liwa (fig. 2).3 All those burials were fenced with warning tape and with apposite signage to remain untouched. (S. L.) From June 2014 to June 2015 different teams of 3 First rescue campaign, which lasted from 30th May to Italian professional archaeologists, led by Sabatino 5th July 2014, was focused on the graveyard of Falaji as Souq, Laurenza but never composed by the same people, near Falaji Al Qabail interchange of the Batinah Expressway

West & East 407 Monografie, 4 Cristiano Putzolu, Carlo Baione, Enzo Cocca, Sabatino Laurenza

2. Research questions At the end of the first campaign, which was successfully concluded as scheduled, the survey team Given the extreme environmental conditions and had many issues to be solved in terms of speeding the construction site needs and timings, the survey the field data acquisition time and the quality of team that operated alongside the archaeologists in data produced on and off site. the first campaign opted for the monoscopic pho- The issues that needed to be solved in order to togrammetry as the methodological approach to improve the survey quality of the second campaign document the structures. This digitization meth- can be summed up as follows: od provides great accuracy and saves time for field – Too much time on the field for data acquisition; data acquisition, but works best when the surveyed – Too much time for the post-processing of digital features lie on the same plane. If the features to be surveys including maps drawing; surveyed are not on the same level, as it was for the – ETS 3D data acquisition was too long for a sat- funerary structures investigated in Oman, it is nec- isfactory data-set. In fact, recording less coordi- essary to divide them in different scenes and survey nates meant less data, thus less information; each one separately. To overcome the challenges of surveying such – Bad accuracy in the final maps due to the diffi- big structures and to give enough time to the ar- cult simplification of three dimensional volumes chaeology team to excavate them stratigraphically, in different planes. the monoscopic photogrammetry seemed the best option at the time. This choice unfortunately ended Before the end of the campaign a preliminary analy- expanding the time needed for the field data acqui- sis of the satellite imagery of the Wadi al Arad area sition, for the images processing and for the drawing showed hundreds of graves to be surveyed; the area of the plans.4 was soon covered with a fieldwalk survey to esti- Another negative aspect of this digitization meth- mate the exact number and disposition of the graves 5 od is its bidimensionality, which complicated the 3D inside the construction corridor. The big number data acquisition on field, which is crucial to properly of tombs inside the buffer not only required a big- read and interpret the original shape and volume of ger team but also a complete reorganization of the the excavated graves. The acquisition of the heights workflow and a change of the strategy and of the of every surface recorded with the monoscopic pho- method used on field and in the lab. togrammetry was made with the Electronic Total Station (ETS): this solution added more time on the (C. P., C. B.) field for the survey team and less time for the archaeologists to investigate the structures. 3. Methodology

Package 5. The team was composed of seven archaeologists and A mix between 3D Structure from Motion (SfM) three Omani representatives and we excavated in Falajii as Souq graveyard a total of 26 graves plus 6 circular stone structures, photogrammetry and ETS seemed the best option surveying (only in the area inside the buffering and fencing to fix the issues noted in the first archaeological zone) around 149 graves. During the second rescue campaign, campaign. 3D digital photogrammetry was already which lasted from 30th September to 23rd December 2014, the team was composed of fifteen archaeologists and three Oma- used by the topography team in their professional ni representatives who excavated 34 graves in Falaji as Souq daily life.6 The switch to SfM photogrammetry was graveyard and 95 graves in Wadi al Arad graveyard. The team quickly assimilated by the team and improved the that run third campaign, from 12th April to 18th June 2015, was composed of sixteen archaeologists plus three Omani representatives and investigated 74 graves in Wadi al Arad graveyard 5 Which is the area investigated for the second and third and 54 graves in Liwa graveyard. campaigns. 4 The investigated tombs are generally composed by the 6 At the time of the second campaign also composed by collapse of the dome (made of medium/big sized basalt stones) Enzo Cocca and Cristiano Putzolu. Bigliardi et Al. 2013; and placed at a different height, the funerary bed/chamber. Putzolu, Vicenzutto 2013; Cocca 2014.

West & East 408 Monografie, 4 Rescue archaeology in the sultanate of Oman: methods and solution strategies

field data acquisition also enhancing the quality of ed on the needs of the software selected to process the work. the images and to create the 3D point cloud: for the 3D digital photogrammetry is a low cost method Oman rescue archaeology campaign the team opted also known as close-range photogrammetry, which for Python Photogrammetry Toolbox (PPT) GUI, belongs to the wider group of the non-invasive, pas- an open source software for rendering a 3D point sive digitization techniques. It is an optics digitizing cloud from photos of an object at different angles.11 system that consists in the identification of points In order to supply the software with the cor- of reference in digital images shot from several an- rect sequence of pictures for the point cloud cal- gles, and in the determination of the relative posi- culation, the tombs were photographed either cir- tion of those identified points in a three dimension- cling around them clockwise or counter-clockwise, al space using triangulation.7 changing shooting angle and pole height every loop; The choice of this digitization method proved to only certain contexts required to photograph the be helpful to accurately record the tombs, further- tomb zigzagging on top of it, or even mixing two more fitting the environmental and archaeologi- different patterns in order to supply the software cal context the team of professionals was operating with more information to properly compute the in. The only equipment needed on site besides the 3D model. ETS – that was integrated to photogrammetry in Given the dimensions of the tombs, which were this survey – consisted of a compact digital camera often clustered in two, three or more, the cam- with a remote control, a telescopic pole with a camera was mounted on a telescopic pole to widen the era mount, a smartphone as camera remote control shooting angle and therefore framing most of the and a bag of paper Ground Control Points (GCP).8 structure in each picture. To speed up the image This approach to the graphic documentation of acquisition step each surveyor connected the cam- the features on site ensured: era to a smartphone via WiFi to control the shoot- – Fast data acquisition on site; ing angle, hence setting a proper distance from the tomb to be surveyed and also allowing a better qual- – Fast post-processing of the 3D surveys and ity control during the acquisition. prompt plans drawing; Each standard sized tomb required a minimum – Accurate and satisfactory 3D data acquisition of 60-80 pictures with at least a 75% overlap be- provided by the 3D point clouds; tween each one; a higher number of photos was re- – Great accuracy in the georeferenced photomod- quired for bigger tombs or for important details, el and in the orthorectified image. for the funerary chamber or also for tombs clusters. The maximum time required to shoot a stan- 3D digital photogrammetry and ETS provided the dard sized tomb with this setup was 10 minutes survey team with a quick and effective solution to circa including the camera setup, and a few min- document all the structures that were going to be utes with the ETS to record each of the 4 GCP destroyed by the construction of the Batinah Ex- placed around the tomb. 9 pressway. To improve the quality of the work and the pro- The acquisition of the image sets for the 3D cessing speed both on field and in the lab, for the last model creation followed different patterns, each campaigns the topography team was upgraded to a 10 one used to survey a different context situation. team of two surveyors.12 A bigger team allowed to The choice of following a regular pattern depend- work on different things at the same time, speeding up the survey on the field, the 3D model creation 7 Howland et Al. 2014. and the post processing of the survey data (fig. 3). 8 Wróżyński et Al. 2017. 9 Alby et Al. 2013; Alby 2015. 11 Moulon, Bezzi 2012; Belmonte et Al. 2017; see 10 Thesurveyed tombs were very often in groups of three infra. or more or else located on the edge of a cliff, thus complicating 12 Cristiano Putzolu and Giacomo Fontana in 2014 and the image acquisition. Cristiano Putzolu and Carlo Baione in 2015.

West & East 409 Monografie, 4 Cristiano Putzolu, Carlo Baione, Enzo Cocca, Sabatino Laurenza

Figure 3 Some moments of the survey methodology workflow: a) images acquisition, b) ETS survey of GCP, c) point cloud creation, d) checking drawings on plotted orthophotoplans, e) vectorized plan

The advantage of using 3D digital photogram- Before the beginning of the second archaeological metry was clear during the field data acquisition campaign the survey team opted for a complete open and became more obvious during the image pro- source workflow: from the images processing to the cessing phase: the survey team noticed an instant production of plans and sections. Most Free/Libre speeding of the preliminary surveys that improved and Open Source Softwares (FLOSS) are released un- the quality of 3D and 2D data, but most impor- der the General Public License, and the source code tantly provided the team of archaeologists with is openly distributed and shared. The team chose to more time to investigate the tomb without losing use FLOSS to benefit from decreased software costs, any relevant information.13 increased security and stability, and furthermore to have more control over the hardware. 13 The first step of the survey, before the excavation of the tomb, was the topographical survey of the structure and PPT GUI is a free/libre, open-source and its collapses. cross-platform photogrammetry software orig-

West & East 410 Monografie, 4 Rescue archaeology in the sultanate of Oman: methods and solution strategies

inally developed by the Arc-Team in September try of the photomodel, to georeference it and to pro- 2010. The software is composed of Python scripts duce an orthorectified image of the scene.16 The 3D that automate the 3D reconstruction process point cloud generated by PPT GUI was in fact pro- from a set of pictures. The processes performed cessed in Meshlab to create both geometries and tex- in PPT GUI can be summed up into two parts: ture, then it was georeferenced with the GCP coordi- camera pose estimation/calibration and dense nates.17 The georeferenced model and the point cloud point cloud computation. Open-source software imported in a GIS environment were then used to are employed to perform these intensive compu- extract cross sections, orthorectified photos and 3D tational steps: Bundler for the calibration (using information to the highest level of accuracy.18 SIFT algorithm) and CMVS/PMVS for the dense To create the digital georeferenced copy of a sin- cloud reconstruction. gle tomb, to vectorialize it and then add it in the db All in all PPT GUI is a user friendly applica- it took from 2 to 4 hours total; more time requested tion that produces 3D digital copies of virtual- a cluster of tombs, both on the field and in the lab, ly any scene or object, and it provides a low-cost, taking from 4 to 12 hours depending on the amount portable and effective solution to record archae- of pictures. ological structures with just a camera and a total During the excavation campaigns both teams of station, which are usual tools on any archeologi- topographers and archaeologists worked together cal excavation. This digitization method, especially to document the graveyards while the construction if executed in a controlled environment, provides site proceeded, and the workflow adopted demon- similar results to those produced with a laser scan- strated to fit the needs. The topographic work on ner, which can be difficult to handle and is not as site was organized in three macro chunks, which left cheap as a compact camera.14 the team the chance to change and adjust strategy in The PPT GUI rendering process could last case of need: from a minimum of 3 hours of image processing – Pre-excavation survey, to record the state of art for a 30-45 images set to 12 hours for a grave cluster condition of the tomb; with around or more than 120 pictures, depend- – During the excavation survey, to record the fu- ing on the hardware used and also on the align- nerary structure and the uppermost layer of ment settings chosen.15 The first step of the pro- abandonment inside the burial chamber; cess tends to employ more CPU than the CMVS/ – End of the excavation survey, to document PMVS process, which instead relies more on the how the tomb is left after the stratigraphic -ex RAM: during the images processing the user is free cavation. to work on other tasks and to run PPT GUI in the background with no need of waiting for the dense cloud creation. 16 The team opted for Meshlab (v 1.3.3), a FLOSS Falkingham 2012, 2013. 17 Once imported both the sparse cloud and the dense that provides systems for processing and editing three cloud in the software, the dense cloud can be processed with dimensional triangular meshes, to create the geome- the Surface Reconstruction: Poisson algorithm. The parameters needed are Octree Depth (suggested range is between 5 and 10, 14 Vicenzutto et Al. 2019; for an updated review of the higher is the parameter the better is the precision in the the potential of SfM photogrammetry in archaeology and an reconstruction but also the higher is the processing time) and accurate bibliography, see Willis et Al. 2016; for a compar- Solver Divide (default value is 8, it reduces the memory usage) ison between Laser Scanning and sfm photogrammetry and an and our choices were respectively 14 and 12. The mesh thus ob- accurate bibliography, see Skarlatos, Kiparissi 2012 and tained is cleaned and then textured using Parametrization + Wilkinson et Al. 2016. texturing from registered rasters algorithm using a Texture size of 15 The first step of PPT GUI pipeline is Run Bundler: the 8192. Finally the model is georeferenced using Reference scene parameters needed are Feature Extractor and desidered Photo and assigning to the different GCP visible in the textured mesh Width and our choices respectively siftvlfeat and 2400 (resolu- the corresponding coordinates. tions higher than 2400 pixels ended with PPT GUI crashing); 18 As further shown in this paper, this project opted for the second step is running either CMVS/PMVS or PMVS with- QGIS and the PyArchInit plugin to manage the alphanumeric out CMVS and our choice is the latter. data.

West & East 411 Monografie, 4 Cristiano Putzolu, Carlo Baione, Enzo Cocca, Sabatino Laurenza

As previously stated the images processing results in a 3D georeferenced point cloud that can be imported into the GIS as z-point feature or it can either be used in any other vector based software for 3D reconstruction. The 3D information thus produced allows archaeologists to hypothesize a reconstruction of the volumes of the funerary structures. The accuracy and fidelity of the final vectorial plan is ensured by the orthorectified image, which, compared to the several photomodels of different portion of the structure proved to be a very important improvement in terms of work quality.19

(C. B.)

4. Field strategy

The field survey workflow has been revised and improved several times throughout the archaeological campaigns with the final goal of gathering as much data as possible in the best way, and at the same time giving the team of archaeologists more time to investigate the contexts. The process development led the topography team to adopt an open-source solution that worked well on the Omani geographical and historical context. To better understand the operating mode it will be useful to describe, as an example, the documentation process of an average tomb. After a preliminary cleaning of the structure, necessary to understand and highlight the physical limits and the structural elements to be reproduced Figure 4 – Data acquisition in 3D, the topographer shoots the set of pictures and then records the GCP placed on the edges of the scene (fig. 4). After the field survey the images are processed with PPT GUI at the standard resolution and then As soon as the zenithal image has been pro- the mesh is created, textured and georeferenced duced it is printed and delivered to the archaeol- with Meshlab (fig. 5) as previously described in this ogists working on site to be used as a graphic base paper. Both the ETS data and the orthorectified im- on which to draw the limits of the different strati- ages are uploaded in the GIS project managed by graphic units (SU). Each SU is then vectorized and PyArchInit.20 loaded into PyArchInit together with its recording sheet (fig. 6).21

19 The workflow used during the first campaign, which gave us many distortion problems. 21 As well as the pottery, osteological and other recording 20 See infra. sheets.

West & East 412 Monografie, 4 Rescue archaeology in the sultanate of Oman: methods and solution strategies

Figure 5 Post processing: from the point cloud to the Figure 6 orthorectified 3D model: Plans and sections drawing: a) point cloud, a) orthorectified photomodel, b) 3d mesh, b) vector plan, c) textured 3d mesh, oblique POV, c) georeferenced 3d point cloud, d) textured 3d mesh, zenithal POV d) selected 3d points along the cross section

West & East 413 Monografie, 4 Cristiano Putzolu, Carlo Baione, Enzo Cocca, Sabatino Laurenza

The accuracy of the 3D renders produced with updated using the thesaurus, through which the user Meshlab allowed the topography team to use the can customize and standardize the specific vocabu- textured and geo-referenced models of the burial lary that best fits a given archaeological context. The chambers with the bone remains as a virtual envi- SU form is linked to the geometric table by means of ronment: the precision of the 3D scenes was used by three external keys that make the data univocal. the anthropology team to collect the metric infor- The vector file (PyUnitaStratigraphic) is im- mation they needed after the work on site in order ported into QGIS and is used to draw the topolog- to speed it up and let the archaeologists team and ical features. As SU form and geometric tables are the construction site to go on faster. The 3D models linked, the user can search not only alphanumer- proved to be helpful for the anthropologists to hy- ic data but also geometric data and quickly create pothesize the taphonomy and to get other informa- phase and period maps. tion from the bones. The implemented functions are: – Export of the forms into pdf; (C. P.) – Error check on stratigraphic relationships; – Export of the matrix media management; – Export of maps for SU and periods (currently a 5. PyArchInit trial tool).

PyArchInit is an open source project written in Py- The GIS time controller board allows the user to thon and directly integrated in QGIS designed to build phase and period plans for either absolute and manage archaeological fieldwork data. It was orig- relative chronology, thus allowing a faster export inally created by Luca Mandolesi in 2005, it was process than by constructing phase plans through then implemented and further developed by Enzo queries. Cocca for his Ph.D. thesis in 2010.22 In 2015 the As mentioned above, the core of PyArchInit re- project expanded to become a web community sides in its geodatabase, which can either be used called UnaQuantum. with a db SQLite or PostgreSQL db. To manage the The geodatabase containing the Batinah Express- data from the archaeological excavation in Oman the way emergency excavation data has been centralized survey team opted for a PostgreSQL db which would on a single server which can be remotely accessed by have given the possibility to centralize it remotely, so all the project members. This allowed the teams to that all teams could enter the data on a single db. All upgrade the geodatabase with their data and to view the excavation data could then be managed connect- and access in real time all the excavation data. ing the single db to the centralized PostgreSQL db PyArchInit proved to have a high capability of through a Transmission Control Protocol (TCP) adaptation to the specific needs of the project in connection. This furthermore allowed the institu- which is utilized, allowing the customization of the tions involved in the Cultural Heritage management existing applications and the development of new to have a direct and real-time supervision of the ar- integrated modules. PyArchInit currently contains chaeological fieldwork in progress. 20 forms that help the archaeologist to digitize and In the frame of the Batinah Expressway proj- manage all archaeological data in QGIS. ect the team proposed a management information system to facilitate the preservation and the valori- How it works: zation of the Cultural Heritage of the Sultanate of The alphanumeric data can be entered into the Oman. PyArchInit database via graphical interfaces: tabs The information management system here pre- provide the user with a tool to erase, search and in- sented is divided into two modules: put data, while some fields with a combobox can be – WebGIS; – GIS. 22 Mandolesi, Cocca 2013.

West & East 414 Monografie, 4 Rescue archaeology in the sultanate of Oman: methods and solution strategies

The WebGIS module is the management tool that implemented and developed, such as the 3D digi- allows 4 levels of secure access via login: tal reproduction of special finds for documentation 23 – The mapping and thematic view of the Oman and analysis purposes. territory; PPT GUI is just one of the many open source solutions for 3D digital photogrammetry, and com- – The positioning of archaeological sites (extra site pared to the more powerful – yet very expensive and level); not entirely customizable – proprietary software – The positioning of each site excavated (intra such as RealityCapture or Photoscan is undoubt- site level) with relative performance of work art edly slower but not less precise. In fact, even if the status; first part of the 3D modeling process in PPT GUI – Search and query features; is computationally slow, the 3D dense cloud pro- – Print pdf in various scales with custom templates; duced by the software is accurate and can be man- – Download data in shapefile format. aged by any program for processing and editing 3D models such as Meshlab. The PyArchInit GIS module is the tool that the ar- After all, the methodology chosen to record the chaeological operator can use to enter alphanumeric excavation processes proved to be effective and to and geometric data of the archaeological sites. This run well along with the stratigraphic method used tool offers an information standard with two pur- by the archaeology team also helping the anthropol- poses: on one hand, it will give the Ministry of Cul- ogy team in their work. tural Heritage a chance to monitor in real-time the None of the tombs investigated in the cam- archaeological work in progress and, on the other paigns was found intact at the beginning of the in- hand, to standardize the data recorded on site. The vestigations, thus complicating the understanding QGIS excavation projects will be made available the original shape of the structures and the distinc- daily and upgraded directly into the server. tion between natural collapses and looting holes. Grave 21 in the graveyard of Falaji as Souq is a dome (E. C.) shaped tumulus tomb with a heap of collapsed stones covering the funerary recess and thus making it scarcely perceptible. This tomb can be used as exemplification of the hermeneutical value of this 6. Conclusions methodology, since it has been digitally recorded in its three dimensionality in many different instants The Omani extreme environmental conditions re- of the excavation and let the archaeologists hypoth- quired accuracy and rapidity on the field on behalf esizing the construction technique and phases. of the teams of professional archaeologists. The In order to record all the stones from the out- photogrammetric workflow presented in this paper ermost collapse to the internal first course of seemed to fit the needs of the project and the hard- stones the team opted for a dedicated survey strat- ware available on field, allowing to investigate and egy which involved a 3D survey for each level re- document all of the tombs destroyed by the con- moved. The archaeologists recognized seven dif- struction of the Batinah Express Highway. ferent construction steps, all of them recorded in Although during the fieldwork was not possible 3D (fig. 7). to apply this method to survey special finds, recent Thanks to the accuracy of the 3D digital pho- papers have shown the ductility of SfM photogram- togrammetric acquisition the team was able to hy- metry for either micro-contexts and for individual pothesize all the construction phases of the tomb objects documentation. In fact, more than just the (fig. 8), and also its original aspect (fig. 9). many improvements made by the team from the first to the last emergency excavation campaign in 23 See for example Molloy, Milić 2018 or also Me- Oman, there are many aspects that could be further gale, Baione 2017 for the 3D digitization of artefacts.

West & East 415 Monografie, 4 Cristiano Putzolu, Carlo Baione, Enzo Cocca, Sabatino Laurenza

Digital photogrammetry and ETS together construction phases, and to transform the work- proved to be useful to read Stratigraphic Units and flow from an accurate time-saving technique to a to interpret the archaeological remains. Freezing precise interpreting tool. different frames of the same scene in a digital environment as done for the Grave 21 in Falaji as Souq helped the team to analyze the masonry and the (C. P., C. B., E. C., S. L.)

Figure. 7 Grave 21 “reverse building”

West & East 416 Monografie, 4 Rescue archaeology in the sultanate of Oman: methods and solution strategies

Figure 8 Hypothetical building and use of the grave (drawings by Claudia Tomaselli) Figure 9 Hypothetical aspect of the Grave 21 (3D model by Vittorio Lauro)

West & East 417 Monografie, 4 Cristiano Putzolu, Carlo Baione, Enzo Cocca, Sabatino Laurenza

BIBLIOGRAPHY

Bigliardi G., Cappelli S., Cocca E. 2013, Il sito di Remote Sensing and Spatial Information Adulis (Eritrea): raccolta e gestione dei dati tramite Sciences» XL-5/W2, pp. 19-24, doi: 10.5194/ software opensource, «Archeologia e Calcolatori» 4, isprsarchives-XL-5-W2-19-2013. pp. 222-227. Alby E. 2015, Point cloud vs drawing on archaeologi- Cocca E. 2014, Kojtepa 2013: The Use of 3D for the cal site, «ISPRS – International Archives of the Drawings of Excavation: A Methodological Approach, Photogrammetry, Remote Sensing and Spatial «Newsletter di Archeologia CISA» 5, pp. 1-20. Information Sciences» XL-5/W7, pp. 7-11, Howland M.D., Kuester F., Levy T.E. 2014, doi:10.5194/isprsarchives-XL-5-W7-7-2015. Photogrammetry in the field: documenting recor- Belmonte A. A., Biong M. M. P., Macatulad E. ding and presenting archaeology, «Mediterranean G. 2017, Dem generation from close-range photo- Archaeology and Archaeometry» 14/4, pp. 101-108. grammetry using extended python photogramme- Mandolesi L., Cocca E. 2013, Pyarchinit: gli svi- try toolbox, «ISPRS – International Archives of luppi dopo ArcheoFoss 2009, in «Archeologia e the Photogrammetry, Remote Sensing and Spatial Calcolatori» 4, pp. 128-138. Information Sciences» XLII-4/W5, pp. 11-19, doi: 10.5194/isprs-archives-XLII-4-W5-11-2017. Moulon P., Bezzi A. 2012, Python Photogrammetry Toolbox: A free solution for Three-Dimensional Falkingham P. 2012, Acquisition of high resolution th- Documentation, in: Cantone F. (ed.), ree-dimensional models using free, open-source, photo- ARCHEOFOSS. Open Source, Free Software e grammetric software, «Palaeontologia Electronica» Open Format nei processi di ricerca archeologica. 15/I, doi: 10.26879/264. Atti del VI Workshop (Napoli, 9-10 giugno 2011), Falkingham P. 2013, Low cost 3D scanning using pp. 153-170. off-the-shelf video gaming peripherals, «Journal Putzolu C., Vicenzutto D. 2013, Il rilievo delle su- of Paleontological Techniques» 11, pp. 1-9, doi: perfici tramite fotogrammetria 3D: dal microscavo dei http://jpaleontologicaltechniques.org/pasta3/ complessi tombali agli scavi in open area, «Archeologia JPT%20N11/Pdf/JPT_n011_Jun.pdf. e Calcolatori» 24, pp. 355-370. Megale C., Baione C. 2017, The Virtual Etruscan Vicenzutto D., Putzolu C., Tasca G. 2019, Tre Museum of Populonia Gasparri Collection: enhanc- dimensioni per un deposito archeologico: fotogramme- ing the visitor’s experience, «Archeostorie. Journal of tria 3D, potenzialità informative e fruizione. Il caso Public Archaeology» 1, doi: 10.23821/2017_3d. studio della Gradiscje di Codroipo, in: Borgna E., Molloy B., Milić, M. 2018, Wonderful Things? A Càssola Guida P., Corazza S. (eds.), Preistoria Consideration of 3D Modelling of Objects in Material e Protostoria del Caput Adriae, Atti della XLIX Culture Research, «Open Archaeology» 4/1, Riunione Scientifica dell’Istituto Italiano di Preistoria pp. 97‑113, doi: 10.1515/opar-2018-0006. e Protostoria (Udine-Pordenone, 8-12 ottobre 2014), Skarlatos D., Kiparissi S. 2012, Comparison (Studi di Preistoria e Protostoria 5), Firenze, of Laser Scanning, Photogrammetry and Sfm- pp. 697-702. mvs Pipeline Applied in Structures and Artificial Willis M.D., Koenig C.W., Black S.L., Castañeda Surfaces, «ISPRS – International Archives of the A.M. 2016, Archeological 3D mapping: the struc- Photogrammetry, Remote Sensing and Spatial ture from motion revolution, «Journal of Texas Information Sciences» I-3, pp. 299-304, doi: Archaeology and History» 3, pp. 1-36. 10.5194/isprsannals-I-3-299-2012. Webliography Wilkinson M.W., Jones R.R., Woods C.E., Alby E., Elter R., Ripoche C., Quere N. 2013, Gilment S.R., McCaffrey K.J.W., Kokkalas S., Close range photogrammetry applied to the documen- Long J.J. 2016, A comparison of terrestrial laser scan- tation of an archaeological site in Gaza Strip, «ISPRS ning and structure-from motion photogrammetry as – International Archives of the Photogrammetry, methods for digital outcrop acquisition, «Geosphere» 12/6, pp. 1865–1880, doi: 10.1130/GES01342.1.

West & East 418 Monografie, 4 Rescue archaeology in the sultanate of Oman: methods and solution strategies

Wróżyński R., Pyszny K., Sojka M., Przybyła C., try with a smartphone and a compact camera, «Open Murat-Błażejewska S. 2017, Ground volume as- Geosciences» 9/1, pp. 281-294, doi: 10.1515/ sessment using ‘Structure from Motion’ photogramme- geo-2017-0023.

West & East 419 Monografie, 4 Finito di stampare nel mese di aprile 2020 EUT Edizioni Università di Trieste