Systemizing the Past: Papers in Near Eastern and Caucasian Archaeology Dedicated to Pavel S. Avetisyan on the Occasion of His 65th Birthday 1803273925, 9781803273921

Citation preview

Systemizing the Past Papers in Near Eastern and Caucasian Archaeology Dedicated to Pavel S. Avetisyan on the Occasion of His 65th Birthday

Edited by

Yervand H. Grekyan Arsen A. Bobokhyan

Systemizing the Past Papers in Near Eastern and Caucasian Archaeology Dedicated to Pavel S. Avetisyan on the Occasion of His 65th Birthday edited by

Yervand H. Grekyan and Arsen A. Bobokhyan

Archaeopress Archaeology

Archaeopress Publishing Ltd Summertown Pavilion 18-24 Middle Way Summertown Oxford OX2 7LG www.archaeopress.com

ISBN 978-1-80327-392-1 ISBN 978-1-80327-393-8 (e-Pdf) © the individual authors and Archaeopress 2023 Cover: Front cover illustration: Bird pendant of bronze, ca. 1400-1300 BCE, Late Bronze Age tomb no. 613, Karashamb Cemetery (in the background), Armenia.

All rights reserved. No part of this book may be reproduced, or transmitted, in any form or by any means, electronic, mechanical, photocopying or otherwise, without the prior written permission of the copyright owners. This book is available direct from Archaeopress or from our website www.archaeopress.com

Avetisyan Pavel

Contents Foreword....................................................................................................................................................................................... iii ‘Axe-Bull’: An Iron-Age Iconic Anagram................................................................................................................................1 Levon Abrahamian Armenian Standing Stones as an Object of Archaeological Study .................................................................................6 Hayk Avetisyan, Artak Gnuni, Levon Mkrtchyan and Arsen Bobokhyan Neolithization of Armenia: General Trends and Patterns of Development................................................................22 Ruben Badalyan and Armine Harutyunyan Groups of Three Deities in Middle and Neo-Assyrian Times..........................................................................................30 Felix Blocher Water Management in Ancient Armenia: Problems and Perspectives........................................................................42 Tork Dalalyan, Roman Hovsepyan, Levon Abrahamian, Arsen Bobokhyan and Boris Gasparyan The Archaeological Site of Garni, Armenia. Pre-Arsacid Archaeological Evidence and an Urartian Inscription of Argišti on a Vishap.........................................................................................................................................56 Roberto Dan, Arsen Bobokhyan, Onofrio Gasparro, Boris Gasparyan, Artur Petrosyan and Mirjo Salvini The Kurtan Belt..........................................................................................................................................................................79 Ruben Davtyan and Michael Herles Achaemenid Habitats in Beniamin II (Shirak, Armenia) from the End of the 6th Century BC to the End of the 4th Century BC................................................................................................................................................................88 † Stéphane Deschamps, François Fichet de Clairfontaine and Felix Ter-Martirossov New Findings on Urartian Rock-Cut Tomb in Mazgirt/Kaleköy Fortress ................................................................116 Serkan Erdoğan Dalarik-1: A New Lower Paleolithic Cave Site in the Republic of Armenia...............................................................127 Boris Gasparyan, Artur Petrosyan, Phil Glauberman, Ani Adigyozalyan, Hayk Haydosyan, Soseh Aghaian, Makoto Arimura, Ellery Frahm, Samvel Nahapetyan, Dmitri Arakelyan, Jennifer Sherriff, Teo Karampaglidis, Masha Krakovsky and Ariel Malinsky-Buller The Tušpa Mound Columned Hall........................................................................................................................................142 Bülent Genç and Erkan Konyar Climate Change and the Transition from the Early to the Middle Bronze Age in the Armenian Highland.....149 Yervand Grekyan A Prehistoric Aggregated Cell Structure at 2850 m asl on Mount Aragats, Armenia.............................................162 Pavol Hnila, Alessandra Gilibert and Arsen Bobokhyan Ceramic Technology at the Kura-Araxes I and II Site of Khizanaant Gora, Shida Kartli, Georgia......................172 Mark Iserlis and Raphael Greenberg Inscribed and Seal-Impressed Clay Finds from the Urartian Fortress of Çavuştepe..............................................189 Kenan Işık and Rıfat Kuvanç Iron Age Pottery from Metsamor. New Observations Based on Assemblage Discovered in 2019 Season.........197 Mateusz Iskra and Tigran Zakyan Urartian Priestesses, How Important They Were? Some Observations of the Iconographic Features.............209 Krzysztof Jakubiak Getahovit - 2 Cave in the Middle Ages................................................................................................................................220 Irena Kalantaryan and Astghik Babajanyan Shaft Hole Axes of Stone and Metal from the Checon Settlement of the Maikop-Novosvobodnaya Community ...............................................................................................................................................................................249 Sergey N. Korenevsky and Aleksandr I. Yudin i

Hatti and Išuwa: Anatolians in the Upper Euphrates Valley........................................................................................259 Aram Kosyan The Fortress of Aramus in the Early Iron Age..................................................................................................................265 Walter Kuntner, Sandra Heinsch and Hayk Avetisyan Woven Traces: Notes from the 2017 and 2018 Excavation Seasons at Masis Blur...................................................294 Kristine Martirosyan-Olshansky and Alan Farahani New Evidence from the Necropolis of Karashamb: Excavations of the Tomb no. 444...........................................308 Varduhi Melikyan and Artak Hakhverdyan Who were the Caucasian Owners of the Mitannian Cylinder Seals?..........................................................................325 Goderdzi Narimanishvil and Nino Shanshashvili Archaeological Prospection in the Ararat Valley – Drilling into the History of Ancient Artaxata, Armenia.337 Nikolaas Noorda, Achim Lichtenberger, Cornelius Meyer, Torben Schreiber and Mkrtich Zardaryan A Middle and Late Bronze Age Settlement in Armenia: The Aggregated Cells of Arteni......................................360 Bérengère Perello, Christine Chataigner, Olivier Barge, Irena Kalantaryan, Karen Azatyan, Roman Hovsepyan and Aurélien Creuzieux ‘Axe-Bull’ – Order of the Thunder God ..............................................................................................................................385 Armen Y. Petrosyan Overlooking the River Hrazdan Valley: The Fortified Site of Tghit in the Tsaghkunyats Mountains, Kotayk Region, Armenia ........................................................................................................................................................390 Artur Petrosyan, Roberto Dan, Priscilla Vitolo, Onofrio Gasparro and Boris Gasparyan From Landjik to Dvin: Armenian Evidence of Decapitation from Prehistory to the Mediaeval Era..................406 Daniel Thomas Potts Reconstructing the Lifeways of the Kura-Araxes ...........................................................................................................411 Mitchell S Rothman Middle Bronze Age Ceramics in Macro and Micro Perspectives .................................................................................424 Karen S. Rubinson Monitoring Heritage At Risk: Caucasus Heritage Watch and the Armenian Monuments of Nagorno-Karabakh..................................................................................................................................................................428 Adam T. Smith, Lori Khatchadourian and Ian Lindsay Tigran the Great and Mithradates Eupator: Two Parallel Kings of Kings?...............................................................440 Giusto Traina Modelling of Bronze and Iron Age Monuments at the Northwestern Slopes of Mount Aragats based on a Case Study of Lernakert......................................................................................................................................................446 Benik Vardanyan and Levon Mkrtchyan The Early Medieval Complex of Agarak..............................................................................................................................454 Nora Yengibaryan and Lilit Ter-Minasyan Women in Urartian Rituals....................................................................................................................................................485 Nora Yengibaryan

ii

Foreword Pavel Avetisyan, a modern leading Armenian archaeologist, a specialist with wide international recognition, was born in Tbilisi, Georgia, but spent his childhood and youth in Talin, a town in Aragatsotn Region of the Republic of Armenia. After graduating from Yerevan State University in 1980, the paths of his life was connected mostly with archaeology. As a researcher Pavel Avetisyan began his scholarly career at the ‘Erebuni’ Museum of History of Yerevan, continued his scientific research at the Chair of Archaeology and Ethnography and Scientific Research Laboratory of Archaeology at the Yerevan State University. Since 1993, he has been a member of the Institute of Archaeology and Ethnography, National Academy of Sciences of the Republic of Armenia, the main scientific research centre for Armenian archaeology. As a result of his extensive publication record, practical field experience, proven ability to develop local and external collaborations, as well as deliver research of international excellence he was appointed as Director of the Institute of Archaeology and Ethnography, a role he has been serviced since 2006. The scientific legacy of Pavel Avetisyan, represented by several books and more than a hundred articles, can conditionally be divided into two main categories. The first category is comprised of published field and laboratory reports covering the vast number of ancient cemeteries and settlements he excavated during the last 30 years. Pavel’s field projects were focused on the study of Bronze and Iron Age sites of Armenia (c. 3500-500 BC) and represent the main sphere of his scientific interests. The first excavations directed by Pavel were carried out in 1980s at the Mastara and Talin necropoleis located on the slopes of mount Aragats. In fact, these excavations formed an integral part of his scientific worldview. Another important stage was the excavations of Agarak, a Bronze and Iron Age multilayer settlement which was excavated under Pavel’s directorship in 2000s. The excavations of this rock-cut settlement and sanctuary significantly contributed the field of landscape archaeology in Armenia. Pavel’s contribution is essential to the works of the Armenian-American project ‘ArAGATS’, which began in the early 2000s. For the first time in the history of Armenian archaeology ‘ArAGATS’ realized regular intensive and extensive research projects, including the excavations of several important settlements and surrounding burial mounds in the Tsaghkahovit plain in northern Armenia. Among the numerous projects directed by Pavel, the excavations carried out at the Early Bronze Age high altitude settlement Tsaghkasar and the expansive excavations in the Middle and Late Bronze Age cemetery of Karashamb are of particular importance. These endeavours provide a fundamental basis for understanding the socio-political developments during the Bronze Age in Armenia. Pavel Avetisyan’s contribution to the study of Neolithic and Chalcolithic period sites (c. 6000-3500 BC), such as the agro-pastoral settlement of Aratashen in the Ararat Plain, has also been significant. Since 2004, Pavel has codirected the Armenian-French joint expedition at Godedzor, a semi-permanent settlement located in southern Armenia. Research at Godedzor has revealed phases dated to the transition from the end of the Chalcolithic and the beginning of the Early Bronze Ages and has provided additional evidence on the question of the local origin of the Kura-Araxes cultural phenomenon. At the same time, it has clearly demonstrated the relations and connections that existed during the end of the Chalcolithic phase between Armenia, the Iranian Plateau, and other regions of the Near East. Research conducted at these settlements has greatly advanced our understanding of the establishment and development of early agricultural societies in the Armenian Highland. The above-mentioned projects, particularly ones focused on the Bronze and Iron Age sites, formed the bases of Pavel’s theoretical framework. Problems concerning the chronology and periodization of Armenian archaeological cultures is especially worthy of a mention. Until the end of 1980s the Armenian archaeological periodization was guided by the famous work of Harutyun Martirosyan based on the Bronze and Iron Ages of Armenia and published in 1964. It was only in the 1990s that Armenian archaeologists, among them Pavel, began to re-examine both the available data and enrich the existing record with new researches bringing the investigation of periodization into a higher theoretical level. The results of this work were summarized in his Ph.D. and Habilitation theses entitled ‘Chronology and Periodization of the Middle Bronze Age of Armenia’ and ‘The Armenian Highland during the 249th Centuries BC: The Dynamics of Socio-Cultural Transformations, according to the Archaeological Data’. In contradiction to other scholars investigating the problems of chronology and periodization, Pavel formed his theories on the basis of a great number of radiocarbon data and by creating a corresponding theoretical and methodological background. For the first time, he introduced the problem under consideration in context of common developments of the corresponding periods, transformation of social environments and culture sequences, which naturally made his theories towards chronology and periodization more probable. Instead of basic evolutional

iii

theory he argues for the use of modern sociological methods, which infer not only logical sequences of cultural developments but also their coexistence and crossings in various niveaus of time and space. Pavel Avetisyan’s research contributions to the fundamental archaeological problems of Ancient Armenia gradually shifted the accent to meta-archaeological levels. So, for clarifying the position of the Armenian Highland in context of the Ancient World and the Ancient Near East, in particular, Pavel recurs to the ‘World-System’ theory and its main concepts (such as borderland, marginal zone, and frontier). His theoretical interests also touch problems concerning the formation and development of complex societies in the Armenian Highland, demonstrating the features typical to regional shifts within the common Near Eastern context. Pavel has also investigated and written about various topics dealing with ceramic typology, burial rites, palaeodemography, sacred landscape, and others. Pavel’s contributions are notable by the integration of present-day theoretical approaches, the application of scientific methodology, and the multidisciplinary nature of his research. And it is also in this regard that Pavel Avetisyan’s research stands out by its scientific value and rises Armenian archaeology to an international level. Hundreds of references of his publications both in Armenian and international scientific circles are testament to the undeniable value of his academic contributions. Pavel’s archaeological activites coincided with the recreation of the Armenian statehood and along with this the radical change of direction of scientific relations and worldviews. In this sense, Pavel Avetisyan’s name undoubtedly lies at the basis of new archaeological school of Armenia. We would like to thank all participants of this volume, our colleagues and friends from all over the world, as this publication would not have been possible without their valuable contributions. We express our sincere gratitude to David Davison from Archaeopress, who kindly met us in realising this publication. Finally, we would also thank Kristine Martirosyan-Olshansky for her kind help during the process of edition of this book. Yervand H. Grekyan Arsen A. Bobokhyan Yerevan July 2021

iv

‘Axe-Bull’: An Iron-Age Iconic Anagram Levon Abrahamian

Institute of Archaeology and Ethnography, NAS RA, Yerevan, Armenia Abstract: Obscure bronze artifacts found in the Iron Age burials in Armenia, pendants and decorative pins, are interpreted as iconic anagrams of a plot which has parallels to the Mediterranean ritual-mythological story of the bull-headed master of the labyrinth slayed by his half-brother Theseus, whose name is related to the Hurrian god of Thunder Tessub with ‘axe’ in his name and iconography. Possible origins of these paired prestigious adornments are traced back to the Middle Bronze Age vessel decorations and petroglyphs and reliefs depicting horn animals. Keywords: iconic anagram, ‘axe-bull’, labyrinth, prestigious adornments, oral narrative, memory transfer

In this article, I will discuss obscure artifacts found in the Iron-Age burials in different regions of Armenia, mostly from the site of Shirakavan.1 Those are bronze pendants consisted of three, sometimes of two circular strips, which form concentric circles, having in its center a protrusion of ambiguous configuration (Figures 1, 2, 3A, 3C). The concentric circles resemble labyrinth, while the protrusion resembles both a stylized bull’s head and an axe. At first, I spotted an axe following the interpretation by the archaeologists who published these objects from Shirakavan.2 Then Hamlet Petrosyan (personal communication) pointed me to a bull’s head resemblance of this protrusion. Combining these two interpretations, I propose to name it ‘axe-bull’. Hence the composition of the pendant could be called ‘The axebull in the labyrinth’.3 The composition could refer to the theme of the Minotaur, the monstrous bull-headed prisoner of the Cretan labyrinth, who was killed by his half-brother Theseus evidently using axe as the weapon of the slaughter. In any case the labyrinth derives its name from the double axe labrys, a word of unclear etymology, while Theseus bears in his name, according to the credible reconstruction by Armen Petrosyan,4 a root related to the Indo-European root *tek’s-, on the base of which the terms meaning ‘axe’ and actions involving the axe are derived, including the name of the Hurrian god Tessub and Urartian god Teiseba, the both depicted as axe-bearing gods of Thunder. Avetisyan et al. 2018: 30-31.  Torosyan et al. 2002: 106.  3 This was the title of my earlier short publication in Armenian (Abrahamian 2004). By not accepting this interpretation in general (Avetisyan et al. 2018: 41), the authors present it incorrectly: it turns out that I follow D․ Vardumyan’s assumption that the composition of the pendant is a labyrinth, while I just thank him on another occasion, for drawing my attention to the word bavigh (‘labyrinth’ in Armenian); it is not clear why I suggest seeing in the center of the composition ‘a stylized depiction of a man yoking a horned animal’, while no word is said about it; finally, the main ‘axe-bull’, the Minotaur-Theseus hypothesis is not even mentioned. 4 Petrosyan 2002: 251-253; 2012: 147-151.

Figure 1. Pendant from a burial in Shirakavan, diameter 12 cm (after Torosyan et al. 2002: Table LXX/8).

1 2

Systemizing the Past (Archaeopress 2023): 1–5

Figure 2. Pendant from a burial in Shirakavan, diameter 9.2 cm (after Torosyan et al. 2002: Table LXXI/2).

Systemizing the Past

Figure 3. The artifact from Nor Bayazet (height 22.5 cm) reconstructed by P. Avetisyan et al. (2018: 46–49). A) The model in its entirety; B) the upper part of the model as an incomplete pin; C) the lower part of the model as an incomplete pendant; D) final reconstruction of the pin (after Avetisyan et al. 2018: Figure 9).

Thus Theseus who bears ‘axe’ in his name, realizes his main heroic action, kills Minotaur in the labyrinth related to the double axe labrys. One may suppose that the composition of the pendant presents an anagrammatic image of the mentioned plot. Anagrams refer to words, composed by letters’ rearrangement, so that the original word which is hidden in this way in the text would be heard indirectly. This principle was especially used in sacred texts which had to hide a secret or a tabuated name.5 For example, the name of the Vedic goddess of speech Vāc is not present but is heard anagrammatically in Rigveda’s mandala X.125 dedicated to this goddess.6 The old Indian, especially Vedic poetic tradition in general preserves many features of the Indo-European mythopoetic tradition, including the anagrammatic principle of constructing sacred texts.7 It is noteworthy that the Armenian Song of Vahagn, which is considered to present one of the most ancient Indo-European poetic texts,8 seems to be constructed with anagrammatic principle as well.9

Figure 4. Small axe from a burial in Shirakavan, height 18.8 cm (after Torosyan et al. 2002: Table LX/14).

name of the bull-slayer has a hidden (etymological) axe-related meaning, which is deciphered by modern researchers, who add one more level in the anagrammatic interpretation of the original composition.

I suppose that in a similar way the artifacts under discussion present visual or iconic anagrams, composed by images hinting the original meaning of the anagrammatic image.10 Thus we have a protrusion repeating the form of small axes known from archaeological finds of the same period (Figure 4) and at the same time shaping the head of a bull, while the

Тhe expression of the anagrammatic image is conditioned by the pictorial context. Thus, ‘axe-bull’ protrusions decorating from outside the rhomb-shaped pins (another reminiscent of a labyrinth) (Figure 5), possibly, accentuating the ‘axe’ meaning, may reflect the victorious end of the narrative, when the bullslayer gets out of the labyrinth and is ‘glorified’, placed inside ‘radiant nimbuses’ (Figures 3A, 3B, 3D). Rhombshaped decorative pins are found in the burials of the same time, in some places together with the pendants.11

On the theory and analysis of anagrams see Ivanov 1998 (with the history and literature of the question) and Toporov 2004. 6 Toporov 1966; Yelizarenkova and Toporov 1979: 63-67. 7 Yelizarenkova and Toporov 1979.  8 Ivanov 1969.  9 Petrosyan 2019.  10 I am grateful to Vardan Hayrapetyan and Sergey Sokolovskiy for the useful discussion of this phenomenon. 5

11

2

Avetisyan et al. 2018: 30. 

Levon Abrahamian: Axe-Bull

Figure 6. Vessel from a Middle Bronze Age burial in Shirakavan, height 14.4 cm (after Torosyan et al. 2002: Table X/8).

their original meaning. There are many examples when the signs with symbolic meaning used to be perceived as pure adornments (for example, some bas-reliefs of medieval Armenian monasteries). There are also many examples when signs with semantic content are used centuries, even millennia later, in a different form but with the same semantic structure. Such is, for example, the three-parted mirror-symmetrical birth-giving scheme, the versions of which are found both on the Kura-Araxes vessels of the 4th millennium BC and in the composition of medieval khachkars.16 However, the last example refers to the universal archetypal schemes, while the labyrinthine anagrammatic image of the Iron-Age adornments refers to a concrete narrative, though also with a birth-giving, or rather, regenerating plot.

Figure 5. Decorative pin from a burial in Shirakavan, height 36 cm (after Torosyan et al. 2002: Table LXXV/17).

This last artifact (Figure 3A) acquired by Yervand Lalayan from Nor Bayazet (Gavar) in 1906, as it turned out recently, gained its unique two-part composition (which became the subject of many arbitrary astronomical interpretations12 later in the museum, when the two pieces of Lalayan’s acquiring (a pendant and a broken decorative pin) were joined together.13 Thus, this newly created mysterious object returns to its original, more understandable, but no less mysterious form (Figures 3C, 3D).14 If we consider that both components of this ‘deconstructed’ artifact, like many other pendants and decorative pins, are involved in the same ‘labyrinthine’ mythological field, then the pendant component reflects the plot in general, while the pin component ‘glorifies’ the victorious bull-slaying hero.

It is noteworthy that in the older images (on the vessels of the first half of the 2nd millennium BC) (Figures 6 and 7) there are often compositions that can be considered a stylized expression of the double-bitted axe (labrys) iconography of the labyrinth,17 the rhombshaped section located between two ‘labryes’ being comparable to the observed ‘labyrinthine’ decorative pins. The vessels with such schematic compositions were widely spread in the Middle Bronze Age in the same area, but were further greatly reduced.18 R. Torosyan, O. Khnkikyan and L. Petrosyan19 relate these vessels to some ethnic group that appeared in the area after the period, when the life in the early Bronze Age was interrupted for a while. We may suppose that this new pottery tradition with ‘labyrinthine’ theme was brought by a people with an appropriate narrative about the double-axe bearing hero/deity.

The proposed interpretation is purely semasiological, we cannot make a definite conclusion about the perception of these two adornments by their wearers. The fact that both adornments were sent to the other world with the dead, who according to the available data were high-class equestrians,15 can only testify the fact that these adornments were of considerable importance, regardless of the meaning seen in them by their wearers and by those who did the burial. The adornments do not have to be prestigious because of

If we proceed from our juxtapositions, we can say that in the first half of the 2nd millennium BC there was an important plot in the territory of present-day

Avetisyan et al. 2018: 39-41. Avetisyan et al. 2018: 46-49. 14 It is not clear why, proving convincingly the artificial (if not forgery) nature of the New Bayazet (Gavar) model (Avetisyan et al. 2018: 46-49), the authors nevertheless offer a new astronomical version of the twoparted model disguised by themselves (Avetisyan et al. 2018: 44-45). 15 Avetisyan et al. 2018: 50-51.  12

Demirkhanyan 1982; Demirkhanyan and Abrahamian 1995. Cf. Khachatryan 1975: Figures 53-54, 57, 61-62. It was Pavel Avetisyan who drew my attention to this circumstance in 2004 during one of my first reports on the topic of labyrinths. 18 Avetisyan et al. 2018: 42.  19 Torosyan et al. 2002: 137. 16

13

17

3

Systemizing the Past the image (petroglyphs) and forming its sacrificial meaning (vishap stelae reliefs). Another riddle is the appearance of the paired (pendant and pin) elite adornment in Etiuni shortly before the Urartian conquest. It could reflect appearance of some elite equestrians in this country from Western areas, where, I hope, in the future some copies or versions of these adornments will be found. However, this move has no other evidences yet. Another possibility is supposed by Armen Petrosyan (personal communication): these adornments could be the sign of the warriors who have passed initiation rites.25 For this version, the sign would have exclusively local origin26 with vaguely imagined meaning reflecting the Middle Bronze Age symbolism, a combination of ancient images used by the ‘designer’ who was commissioned to create this sign marking the initiation passage. I propose that the sign could also reflect anagrammatically the labyrinthine narrative as well, which, probably, was circulating only orally at that time.

Figure 7. Vessel from a Middle Bronze Age burial in Harich, height 24 cm (after Khachatryan 1975: Figure 60).

Armenia related to the axe-labrys and the labyrinth, which in ancient Greece had a detailed parallel ritualmythological narrative about the monstrous Minotaur living in the Cnossos labyrinth in Crete and finding his death from his half-brother Theseus, whose name, weapon of the slaughter and place, where slaying takes place, were probably related to the Hurrian axe-bearing and also fratricidal Thunder god Tessub of Asia Minor.20 It is difficult to say what was the connection in content and chronology (if there was any) of the local plot anagrammatically depicted on the archeological finds and the well-known Mediterranean plot. We can only assert that centuries later, in the Iron Age, it suddenly appears in the form of the ‘axe-bull’ anagrammatic images in the same area, in the country of Etiuni, shortly before the conquest of that country by the Urartians․21 It is noteworthy that at that time the Urartians replaced the axe-bearing Hurrian god Tessub by Teiseba, also an axe-bearing god, who had probably already ‘forgotten’ the deeds of his early past, the echo of which was depicted in the ornaments of the conquered people.

Bibliography Abrahamian, L. 1995. Ritual-mythological sources of the labyrinth. In S. B. Harutyunyan and D. S. Vardumyan (eds), Republican Scientific Conference: 3-4․ Yerevan, Institute of Archaeology and Ethnography, NAS RA (in Armenian). Abrahamian, L. 2004. ‘Axe-bull’ in the labyrinth: towards clarification of one Broze- and Iron-Age composition. In S. Hayrapetyan (ed.), HistoricalCultural Legacy of Shirak, Proceedings of the 6th Republican Scientific Conference: 12–14. Gyumri, The Shirak Center for Armenian Studies, NAS RA (in Armenian). Avetisyan, P., Dan, R. and Petrosyan, A. 2018. Axes, labyrinths and astral symbols: bronze pendants and pins from the Armenian Highlands. Ancient West and East 17: 27–63. Demirkhanyan, A. 1982. Towards the problem of symbolism of the three-part compositions of Ancient Armenia. Patma-banasirakan handes (HistoricalPhilological Journal) 4: 154–162 (in Russian). Demirkhanyan, A. and Abrahamian, L. 1995. Crossstone as a version of three-part mirror symmetrical compositions. In G. Sh. Gyodakyan, A. V. Aghasyan, L. A. Barseghyan, K. E. Khudabashyan, L. H. Hakhverdyan, M. M. Hasratyan, M. M. Ghazaryan and V. H. Ghazaryan (eds), The Seventh Republican Scientific Conference on Art, Abstracts of Papers, Yerevan, 10-13 November 1995: 18–19. Yerevan, RA NAS Publishing House (in Armenian).

Could the labyrinthine pins of the Iron Age be the direct descendants of the rhomb-shaped images of the first half of the 2nd millennium BC? There are no data to assert such transition. The search among the petroglyphs for images resembling figures on the decorative pins, such as astral bodies and praying men22 seem to be out of the labyrinth context I am discussing here. While petroglyphs and vishap stelae reliefs, proposed as the origin of the pendants’ composition by Armen Petrosyan23 and P. Avetisyan, R. Dan, A. Petrosyan,24 could serve as iconic models for shaping Petrosyan 2002: 251-253; 2012: 147-151. P. Avetisyan, R. Dan and A. Petrosyan (2018: 50) recently suggested the precise date for these finds to be from the 9th/8th to the 7th/6th centuries BC, while R. Torosyan, O. Khnkikyan and L. Petrosyan (2002: 90-115) date them since the 11th century BC. 22 Avetisyan et al. 2018: 51. 23 Petrosyan 2015: 23-29. 24 Avetisyan et al. 2018: 50-51. 20 21

25 On the initiatory aspects of the Cretan labyrinth see Vidal-Naquet 1986: 112, of the labyrinth in general – Abrahamian 1995. 26 See Armen Petrosyan in this volume.

4

Levon Abrahamian: Axe-Bull Ivanov, V. V. 1969. Using for etymological studies onestem words in the ancient poetry in Indo-European languages. In O. N. Trubachev (Editor-in-Chief), Zh. Zh. Varbot, L. A. Gindin, G. A. Klimov, V. A. Merkulova and V. N. Toporov (eds), Etymology 1967: 40–56. Moscow, Nauka (in Russian). Ivanov, V. V. 1998. The theory of anagrams in the IndoEuropean verse. In V. V. Ivanov, Selected Works in Semiotics and History of Culture. Volume I: 617-627. Moscow, Languages of Russian Culture (in Russian). Khachatryan, T. S. 1975. Ancient Culture of Shirak (III-I millennia BC). Yerevan, Publishing House of the Yerevan University (in Russian). Petrosyan, A. 2002. The triad of Urartian principal gods and the problem of the origin of the ruling elite of the state. Patma-banasirakan handes (HistoricalPhilological Journal) 2: 256–266 (in Armenian). Petrosyan, A. 2012. The cities of Kumme, Kummanna and their god Teššub / Teišeba. In M. E. Huld, K. JonesBley and D. Miller (eds), Archaeology and Language. Indo-European Studies Presented to James P. Mallory (Journal of Indo-European Studies, Monograph 60): 141–155. Washington, DC, Institute for the Study of Man. Petrosyan, A. 2015. Thirty years passed: the vishap (dragon) stone monuments and the dragon-fighting myth. In A. Petrosyan and A. Bobokhyan (eds), The

Vishap Stone Monuments: 13–52․ Yerevan, Gitutyun (in Armenian). Petrosyan, A. 2019. On the anagrams of the ‘Song of Vahagn’. In V. L. Katvalyan, G. M. Mkhitaryan, L. S. Hovsepyan, G. G. Gevorgyan, A. E. Sargsyan and N. M. Simonyan (eds), Current Problems of the Historical Study of the Armenian Language: 167–176. Yerevan, Gitutyun (in Armenian). Toporov, V. 1966. On one example of the phonic symbolism (Rigveda X, 125). In R. Jakobson (ed.), Poetics / Poetyka / Poėtika, II: 75–77. The Hague, Paris and Warszawa: De Gruyer Mouton (in Russian). Toporov, V. N. 2004. Towards the study of anagrammatic structures (analyses). In V. N. Toporov, Studies in Etymology and Semantics. Volume I: 708–755. Moscow, Languages of Slavic Cultures (in Russian). Torosyan, R. M., Khnkikyan, O. S. and Petrosyan, L. A. 2002. Ancient Shirakavan (Results of the Excavations of 1977-1981). Yerevan, Gitutyun (in Armenian). Vidal-Naquet, P. 1986. The Black Hunter. Forms of Thought and Forms of Society in the Greek World. Baltimore and London, The Johns Hopkins University Press. Yelizarenkova, T. Ya. and Toporov, V. N. 1979. Old-Indian poetics and its Indo-European sources. In G. A. Zograf and V. G. Erman (eds), Literature and Culture of Ancient and Medieval India: 36–88. Moscow, Nauka (in Russian).

5

Armenian Standing Stones as an Object of Archaeological Study* Hayk Avetisyan, Artak Gnuni, Levon Mkrtchyan and Arsen Bobokhyan Yerevan State University and Institute of Archaeology and Ethnography, NAS RA, Yerevan, Armenia

Abstract: This study is dedicated to the problem of definition of standing stones and their types in Armenia. In the academic literature, these stones are considered as burial monuments (erected in memory of famous persons or events), boundary stones or road guides, altars, places of judgment or purification. The article focuses on criteria for the definition and typology of Armenian standing stones, according to their formal and contextual features. The emergence of the perception of ‘standing stones’ can be attributed to the early agricultural societies, although their clear existence is visible between the Bronze and Iron Ages. These monuments still play an important role in the world outlook of local communities. Keywords: Armenian Highland, Bronze and Iron Ages, standing stone, menhir, obelisk, stela, transitional form

Introduction In recent years, in the pages of academic literature, much attention has been paid to the research of megalithic monuments, particularly menhirs and menhir-type monuments, so-called standing stones. Moreover, the issue often goes beyond the academic sphere and becomes a subject of wider scientific and social discussions, dealing with interdisciplinary issues, even in the sphere of identity (for example, the Stonehenge in England and Zorats Karer in Armenia). In this sense, the study of standing stones and the clarification of their significance is anessential issue in Armenian archaeology. Stones and stone stelae have been worshiped in Armenia since the earliest times. The connection between the cults of stone and ancestors is mentioned by Agatangeghos in the 5th century AD.1 In mythology, fairy tales, and epics, a number of mentioned monuments can be identified with menhirs and their complexes (stone or column at the crossroads, a stone under which the master is buried,2 a stone on top of which a light emanates).3 There are also references that can be identified with natural or man-made standing stones, which are regarded as petrified giants (Getashen), with underground roots (Samshvilde),4 and tombs of saints (Great Abul).5 The alignments of menhirs (Zorats Karer, Nemrut, Kanaker) were analogued with petrified armies6 or herds of cattle (Minor Masis, Vaspurakan).7 Examining the Armenian

This work was supported by the Committee of Science, Republic of Armenia, under grants no. 21T-6A207 and no. 21AG-6A080. 1 Agatangeghos 1909: 172. 2 Gulakyan 1983: 82, 156, 304-305. 3 Gulakyan 1983: 156. 4 Djeyranov 1898: 64. 5 Rostomov 1898: 25-26. 6 Haxthausen 1857: 215-217; Lalayean 1898: 174-175; Srvandztyants 1874: 47-48. 7 Tsotsikean 1917: 413-414. *

Systemizing the Past (Archaeopress 2023): 6–21

myth about the petrified Queen Shamiram, Gh. Alishan considers it as a reference to a menhir.8 The scientific documentation of standing stones dates back to the 19th century. In particular the menhirs of Harzhis were known. It is mentioned that monuments of this type form a system that branches through almost the whole territory of ​​Vayots Dzor (Khachik, Amaghu, Keshishkend (Yeghegnadzor), Karmir Vank (Areni).9 The study of the megalithic complexes of Zorats Karer begins in the 19th century and continues into the 1920s by S. Lisitsyan,10 in 1970-1980 by O. Xnkikyan, and in the 2000s by A. Piliposyan. Rows of menhirs, cromlechs, their combinations are also found in Gegharkunik (Kuri Kharaba, Mrtbi Dzor),11 Vayots Dzor (Sultan Kelesi).12 Standing stones are also documented in the context of tombs. Several of them were documented by Ressler among the tombs of Gandzak region.13 A. Kalantar had the greatest contribution in this field. In 1923-1924 he investigated megalithic monuments in Syunik, Aragatsotn, Shirak and Lori and proposed a typology scheme for these monuments.14 In 1920-1930s T. Toramanyan mentions the menhirs located in the village of Shamiram, Aragatsotn region, and describes their structure.15 Here in 1969 a tomb was excavated by R. Torosyan, with a menhir placed on it.16 Since 1970s the systematic excavations of Shamiram under the direction of G. Areshian revealed new menhirs, which allowed to clarify the archaeological context of the previously known ones.17 Alishan 1910: 45. Bayern 1871: 308-309. 10 Lisitsyan 1938: 709-721. 11 Ivanovski 1911: 20; Lalayean 1908: 63. 12 Lalayean 1904: 246. 13 Ressler 1904: 44-46, 59. 14 Kalantar 1926: 210-211, 221-222. 15 Toramanyan 1942: 8. 16 Materials are not published and are kept in the HistoricalEthnographical Museum of Etchmiadzin, Inventory no. 2766-2771/40. 17 Areshian 1978: 503-504. 8 9

Hayk Avetisyan et al.: Armenian Standing Stones as an Object of Archaeological Study One of the earliest megalithic monuments was discovered in 1950-1960s during the excavations of the Early Bronze Age settlement of Shengavit, from the rectangular dwellign no. 1. It is a cylinder stone monument with a height of 230 cm.18 Sometime later, at the beginning of the 3rd millenium BC the Mokhrablur temple was built, on the top of which a menhir-type basalt monument stood.19 The question of the origin and cradle of the megalithic monuments still raises many hypotheses. From the 17th century onwards, it was common to attribute megalithic monuments, in general, and menhirs to the Classical period,20 ascribing them a local origin (O. Magnus, O. Worm, T. Corret).21 Since the discovery of megalithic structures in other regions since the 19th century, two main theories have been formed on their origin and distribution: multi-centered and singlecentered.22

Figure 1. Shamiram 13, preliminary elaboration of standing stone.

The views on the possible functions of menhirs, with all their diversity, can be divided into several groups: a) menhirs as burial monuments,23 b) erected in memory of a famous person or event,24 c) boundary stones or road guides,25 d) altars,26 and e) places of judgment or purification.27 Definition Criteria of Armenian Standing Stones According to elaboration According to the stone processing technique, standing stones can be classified into three groups:28 1) Stone processing, which was achieved by removing individual fragments of appropriate size and shape from the natural outcrops without any further modificaiton of the stone (e.g. Lusakn 2, Dzoraglukh, Shamiram 13, Figure 1); 2) Partially modification (it is documented in most standing stones), where in the surface of the stone was roughly processed to have a smoother appearance (Figure 2); c) Extensive modification, which is identified in Harzhis 21 and Gomk. While the Harzhis example is in a displaced state and its archaeological context cannot be reconstructed, the Gomk example is still in its original location, placed on top of a damaged barrow (Figure 3). There is a certain resemblance to the stone fragments unearthed in one of the tombs

Figure 2. Bazmaghbyur, rough elaboration of standing stone.

Sardaryan 2004: 295. Areshian 2005: 83-84. 20 Jones 1655: 5, 40-47, 65-66. 21 D’Auvergne-Corret 1797: 24; Magnus 1558: 22; Worm 1643: 4-5. 22 Bayern 1871: 314; De Morgan 1925: 239; Uvarova 1900: 197-198. 23 Lebeuf 2012: 41; Magnus 1658: 13; Worm 1643: 4. 24 Ferguson 1877: 52; Lisitsyan and Bayburdyan 1928: 20-21; Magnus 1658: 11. 25 Rowe 1830: 190. 26 Lebeuf 2012: 41; Lisitsyan and Bayburdyan 1928: 20. 27 Lebeuf 2012: 41. 28 The stages of stone processing are presented according to Berlin et al. 1979: 40. 18 19

excavated nearby Golovino, which probably belonged to the cromlech of the tomb.29 Among the standing stones, a group can be distinguished, in which the first and second versions of elaboration are combined. More often, they are slightly elaborated laterally (e.g. 29

7

Tumyan 1937: Figure 30.

Systemizing the Past

Figure 3. Gomk, fine elaboration of standing stone.

Figure 5. Harzhis 21, preliminary elaboration of the backside of the standing stone.

Shamiram 1, Garnarich 1, Harzhis 8 and 25, see Figure 4) or transversely (e.g. Bandivan, Attash 3, Harzhis 21, see Figure 5) which may indicate that the object was intended primarily for viewing from one side only.30 According to architectural traits Considering the standing stones as small architectural objects, it is necessary to mention a number of features, which refer to both the general structure and the structure of the body and the top. The largest group of standing stones has straight proportions (Figure 6). In some cases, the standing stones may be sloping (Agarak, Nerkin Sasnashen; Figure 7)31 or bent (e.g. Ambari Gyol, Harzhis 14, Rind 1; Figure 8). It should be noted that this feature is often found on vishapoids.32 The next component of the standing stones is the structure of the body. In the case of pre-elaboration, the body is mostly shapeless, while the more elaborate monuments have a shape of parallelepiped. Slab-shaped stones can be distinguished in this type of monuments (Harzhis 14 and 16, Krapashti Tner 2;33 Figure 9). The next group consists of pyramidal (e.g. Shamiram 8, Navur 2; Figure 10) or trapezoidal stones (Nerkin Sasnashen;34 Figure 11). The smallest group consists of stones with a cylinder or rounded body (Bazmaghbyur,

V. Stukeley notes that the frontal parts of the stone are better elaborated at Stonehenge (Stukeley 1740: 15). 31 Cf. in Denmark (Worm 1643: 62). 32 Bobokhyan et al. 2015: 299. For curved or inverted proportions of vishap stones, see Tumanyan 2015: 101. 33 Cf. Alto Alentejo, Portugal (Pope and Miranda 1999: 112), Kechili, Trache (Özbek 2006: 87). 34 Cf. Pyramidal megaliths at Devil’s Arrows, England (Aubery w.d ..). The body is slighly narrowing on some of menhirs of Stonehenge (Jones 1655: 60), but V. Stukeley notes that the slight narrowing to the top is not enough to qualify the form as a pyramidal (Stukeley 1740: 23). 30

Figure 4. Shamiram 1, preliminary elaboration of the side part of the standing stone.

8

Hayk Avetisyan et al.: Armenian Standing Stones as an Object of Archaeological Study

Figure 8. Ambari gyol, standing stone with an inverted body.

Figure 6. Lernakert 3, straight standing stone.

Figure 7. Agarak, standing stone with a sloping body.

Figure 9. Harzhis 16, slab-shaped standing stone.

9

Systemizing the Past

Figure 10. Navur 2, standing stone with a pyramidal body.

Figure 11. Nerkin Sasnashen, standing stone with a trapezoidal body.

Figure 12. Metsadzor, cylinder standing stone.

Metsadzor, Khnus, Shamiram 2; Figure 12).35 The round elaboration of body is typical of some vishap stones.36

pentagonal (Bandivan; Figure 18), or triangular (Shamiram 2, Harzhis 10; Figure 19).38

One of the most striking details of the architectural composition of standing stones is the structure of the upper part. In case of a variety of body designs, the elaboration of the top can often be a distinguishing feature. In general, the upper part can be straight (e.g. Khnus, Ghukasavan 2, Shamiram 7, 9-11, Ujan 1, Balak 4; Figure 13), sloping (e.g. Ghukasavan 1, Dzoraglukh 1 Shamiram 3, 6, 8, 12, 14; Figure 14), humped (e.g. Ghukasavan 3, Agarak, Aparan, Shamiram 4 Garni 3/2, Harzhis, 13; Figure 15), beak-shaped (Harzhis 17, 20, 24; Figure 16), curved (Yeghegnavan, Bazmaghbyur, Dashtadem 2, Dzoragyugh, Bnunis 1; Figure 17),37

By standing stones, in some cases, the lower part is narrowing or triangular (Agarak, Pemzashen, Ishkhanasar, Harzhis 13, 23, 24, Aylakh 1; Figure 20). Rare are those with pedestal or widening lower part, which are documented in Harichavank 3, Harzhis 5, 6 and Gomk (Figure 21).39

38 The group of standing stones with a hexagonal top is similar to the statues with hexagonal heads separated by H. Martirosyan (Martirosyan 1961: 82-84). Cf. Teishebaini, Lernapat, Tsaghkalanj (Martirosyan 1961: 83; Yesayan 1980: Table 41/1, 3-4), Argishtikhinili (Martirosyan 1974: 162). It should be noted that such a top is typical to the monuments of the 1st millennium BC. Hexagonal heads have also the anthropomorphic standing stones (Kornidzor, Lake Al). Perhaps this feature can be considered a sign of anthropomorphization of the standing stone. 39 Square and cylindrical pedestals are attested in a number of phallic and human shaped monuments (Lchashen, Karmir Blur, Dvin, Oshakan, Gunesh-Tepe), cf. Yesayan 1980: Table 53/6, 54/1-3, 6, 55/4, 8; Gnuni 2004: 124, on the phallic pipe-bowl found from the ‘worshiped ox’ tomb cf. Lalayean 1931: 151.

Similar columns are known from Denmark (Worm 1643: 64). It is possible that the columns found in Psekup (Western Caucasus) represent the same type of monuments (Kamenev 1867). 36 Bobokhyan et al. 2015: 299. 37 Menhirs with curved tops are documented in Scandinavia (Magnus 1558: 24). The upright or rounded upper part is typical of the Urartian pulusi type stelae (Avetisyan 2016: 112-113). 35

10

Hayk Avetisyan et al.: Armenian Standing Stones as an Object of Archaeological Study

Figure 13. Ghukasavan 2, standing stone with a straight top.

Figure 16. Harzhis 17, standing stone with abeak-shaped top.

Figure 14. Dzoraglukh 1, standing stone with a sloping top.

Figure 17. Dzoraglukh, standing stone with a curved top.

menhirs of Krapashti Tner 1, where the hole is located in the central part of the stone. The holes are mostly horizontal, rarely handle-shaped (Ghukasavan 2).40 The next sculptural element are the cupmarks and deep holes which are more rare on menhirs. Cupmarks and deep holes are present in the frontal part of Shamiram 1, Harzhis 10,41 Metsadzor, Ghukasavan 2 menhirs (Figure 12). A cupmark is present in the central part of Shamiram 10 menhir (Figure 23). It is noteworthy that the hole on the Soylan’s menhir (which is turned into a khachkar/crosstone) is located at the top, which was probably intended for certain pourings, while in Tsitsernavank and Harzhis 23 menhirs, the hole was on the floor, probably to strengthen the standing stone (Figure 24).42

Figure 15. Ghukasavan 3, humped standing stone.

According to sculptural peculiarities Holed stones are the most common (Yeghegnavan, Ghukasavan 1, 3, Agarak, Aparan, Artik 1, Artavan, Selim 2, Harzhis 7-9, 12, 15, 17, 24; Figure 22). The holes are usually made on in the top of the standing stone. An exception is Harzhis 15 (Figure 22) as well as the

Similar holes are documented in Bratsigovo, Bulgaria (Mishev 2016: 316-322); in Cornwall (Bottrell 1873: 31-32), in Stonehenge (Stukeley 1740: 61). 41 Avetisyan et al. 2019: 595. 42 Similar cupmarks are also attested in European monuments (Auden 1907: 7; Holmes 1907: 205-207; Škorpil 1905: 380, 383; Stukeley 1740: 63; Wilson 1888: 588). 40

11

Systemizing the Past

Figure 18. Bandivan, standing stone with a pentagonal top.

Figure 20. Harzhis 13, standing stone with trianglular lower part.

Figure 21. Harzhis 5, standing stone with a pedestal.

Figure 19. Harzhis 10, standing stone with triangular head.

The sculptures of standing stones can be divided into three groups, according to the technique of their preparation: engraved sculpture (in Ghurt Tapa with bent lines, ingraved rectangles; Figure 25), bas relief concentric circles (Harzhis 5, 23, 24; Figure 24), haut relief grooves (Lernarot; Figure 26), snake-shaped (Khnatsakh 1)43 or anthropomoprhic images (Harzhis 27).44

Figure 22. Harzhis 15, standing stone with a hole in the central part.

Avetisyan et al. 2015: 200. 44 Similar objects have been attested in Scandinavia and Spain (Gobo et al. 1995: 59; Magnus 1558: 23; Magnus 1658: 12-13). 43

12

Hayk Avetisyan et al.: Armenian Standing Stones as an Object of Archaeological Study

Figure 25. Ghurt tapa, standing stone with an engraved sculpture.

Figure 23. Shamiram 10, standing stone with a cupmark in the central part.

Figure 24. Harzhis 23, standing stone with a hole curved in the lower part.

Figure 26. Lernarot 2, grooved standing stone.

At the same time, in some cases, the standing stone gets anthropomoprhic (Al Lake45) or phallic (Krapashti Tner 1, Geghashen) countours (Figure 27). Anthropomorphic standing stones are also known (Kornidzor, Khoznavar).46

According to spatial principles and archaeological context

45 46

The standing stones could be placed in isolation, in rows, in a stone circle (cromlech), in parallel rows (alley) and grid rows,47 they can appear also in groups (menhir forests). Determining the location of the

Avetisyan et al. 2015: 66. Mkrtchyan 2015: 135.

47

13

Aubery 1862: 320; Bayens 1905: 8-10; Ferguson 1877: 51.

Systemizing the Past

Figure 27. Geghashen, standing stone with a phallic head (Drawing: V. Gayseryan).

isolated ones is quite difficult, as it can not be excluded that the monuments for some reason were moved in later periods. They are archaeologically documented in Mokhrablur and Shengavit, located in places of worship48 or erected isolated on ritual platforms. To this group should be added the idol placed on top of the tower of Ardar David.49

Figure 28),55 Novoseltsovo.56 The Berdik complex is worth mentioning, where the alignment of standing stones rise on separate tombs. In the central part of the necropolis there was a monumental structure with a cyclopean arrangement.57 Although surface material typical of the Late Bronze and Early Iron Ages has been found here, the contemporaneity of the stone alignments and tombs is questionable. In this regard, the so-called ‘ritual barrow’ no. 36 in Khanlar deserves a special attention. There were rows of boulders and stone circles on the top of the hill. According to Y. Gumel, there was also a large Bronze Age structure nearby.58

The next method of placing standing stones is the forest of menhirs, which also occur in cemeteries (Shamiram, Harzhis). Alignments of standing stones are attested both inside of the necropolis (Aygeshat, Norakert (megalithic towers are built here), Zorats Karer, Krapashti Tner50), outside of them (Khnatsakh 351) and in the context of cyclopean fortresses.52 In some cases, the alignment was subject to some logic. Thus, the tallest phallic menhir at the complex of Krapashti Tner (currently knocked down) was placed in the center of the row.53

The next method of placing standing stones is the closed one (cromlech).59 Cromlechs often surround tombs (Khnatsakh 5, Zorats Karer, Golovino), but are attested also on ritual platforms (Aylakh). The (Aubery 1862: 320). This hypothesis is confirmed by the example of Aruch 2, where a range of standing stones led to tomb N 5 (Areshian and Tumanyan 1991: 24-25). 55 Hmayakyan et al. 2010: 28-33. 56 Investigations by G. Sargsyan. 57 Avetisyan et al. 2015: 102. 58 Hummel 1940: 75, 77, 103-105. 59 This term has been clarifying since the 17th century. It is true that in this sense the term ‘stone circle’ was more applicable at that time. The word ‘cromlech’ was often used to describe the central altar of the circle (Borlase 1769: 193). Nowadays, the word ‘cromlech’ often refers to any stone rows around the tomb (see Bayburtyan and Lisitsyan 1928: 20).

Menhir alleys54 (parallel alignments) are documented in Ashotsk, Hartashen (in context of a necropolis; Areshian 2005: 83-84; Sardaryan 2004: 295. Investigations by L. Mkrtchyan. 50 Avetisyan et al. 2015: 216; Gnuni et al. 2017: 31; Torosyan 1971; Piliposyan and Avagyan 2016: 15. 51 Avetisyan et al. 2015: 216. 52 Hmayakyan et al. 2010: 29. 53 Cf. Beneteau-Douilard 2006: 570. 54 Menhirs placed in this way are called by J. Aubery procession roads 48 49

14

Hayk Avetisyan et al.: Armenian Standing Stones as an Object of Archaeological Study

Figure 28. Hartashen, alley of standing stones.

alignment approaching the total square of the area forms a grid (Lezk).60 As a separate type should be considered the combinations of chomlech and alignments (Kuri Kharaba).61 An Attempt of Typology of Armenian Standing Stones

Figure 29. Harzhis 8, menhir.

The standing stones can be classified into the following groups, which being similar to each other, anyway form several groups and differ both in structure, chronology and archaeological context.

Obelisks

All standing stones are often presented as menhir,62 with the main feature being the size of the stone (up to 3-4 m) and the elongated form.63 In addition to these features, the fact that it is roughly processed or unprocessed is often mentioned.64 Thus, being roughly elaborated is the main feature of menhirs. Almost all of the menhirs are straight, while the bodies are shapeless or in the form of a parallelepiped (Figure 29). The sculpting is limited to holes or cupmarks. The dating of the menhirs raises serious issues. The excavations of Shamiram menhir no. 165 and Aruch tomb no. 566 suggest that they were more widespread since the mid to the end of the 2nd millennium BC.

The next large group of standing stones are obelisks.67 The obelisk is usually defined as a square column that narrows upwards. However, the authors of this definition refer to the Egyptian monuments as an example.68 In general, this group is distinguished by fine processing.69 They are attested in Golovino,70 Gomk and Harzhis (Figure 30). The Gomk example is neatly cut, almost square in horizontal section, the lower part is slightly widened in form of a pedestal. It is noteworthy that in the looted chamber there are fragments of pottery typical to the Early Iron Age. The Harzhis obelisk generally resembling the example of Gomk, bears a sculpture of a concentric circle surrounded by cupmarks.71 This sculpture can specify the period of the existence of these monuments. This ornament is

Hmayakyan et al. 2010: 30. Ivanovski 1911: 20. This type is attested in early works on megalithic monuments (Aubery, w.d. 20; Rowe 1830: 194; Tsonev 2010: 51). 62 Brey and Tramp 1990: 155, 232. 63 Barkhudaryan 1935: 33; Martinov and Sher 1989: 9; Matyushin 1996: 36. 64 Lisitsyan and Bayburdyan 1928: 20. 65 Materials are kep in Historical-Ethnographical Museum of Etchmiadzin, inventory no. 2766-2771. 66 Areshian and Tumanyan 1991: 26.

O. Magnus mentions separately the pillars (obelisk) (Magnus 1658: 11). J. Auberry places images of well-elaborated stone steale in the necropolis at the site of Devil’s Arrows, along with rough stones (Auberry n.d. 38, Figure 2). Similar stelae (yak and booz) to be mentioned in the Bible (3 Kings, 7, 22-23). Although this type of monument is usually identified with menhirs, some researchers consider it a separate type (Cambry 1805: X; D ‘Anna and Pinet 2002: 580; Rowlands 1766: 48; Šcorpil 1905: 372-373). 68 Brey and Tramp 1990: 179. 69 Obelisks are mentioned in Northern Europe erected in memory of heroes, events or on tombs (Magnus 1658: 11-13). 70 Tumyan 1937: Figure 30. 71 Avetisyan et al. 2017: 5, Figure 5/15.

Menhirs

67

60 61

15

Systemizing the Past

Figure 30. Harzhis 6, obelisk.

often found on pottery and metal objects of the 7th-5th centuries BC72 and has parallels in Urartian and Iranian art.73 The mentioned obelisks of Gomk and Harzhis are similar to those known from the period of the Kingdom of Van in a number of features (e.g., fine processing, pedestal).74 Later, in the Sasanian period, cylinder columns were placed, e.g., on the altars of Bishapur.75 These data suggest that such monuments are typical for the beginning of the 1st millennium BC. The other type of obelisks have a shape of an upward-sloping parallelepiped (Hovhanavan). Stelae In context of the standing stones appear sculptured monuments (e.g. Christianized ones in Lernarot, Khnatsakh 1, Tegh, Areni; Figure 31).76 Characteristic features can be considered relatively wide proportions, and sometimes the cross section approaching a square. A further development of this type of stelae can be considered the sculptural stelae of later periods (Yervandashat,77 Hoghmik78). Other types Among the mentioned monuments, examples from Mokhrablur and Shengavit have a special place.79 In Cf. bronze earrings from Shvanidzor (Xnkikyan 2002: Plate XXXII/4-5), pottery from Shikahogh (Xnkikyan 2002: Plates LXXXIII/13, LXXXVI/28), from Keren (Gnuni 2014: 163-166). 73 Gnuni 2014: 159. 74 Avetisyan 2016: 112-114; Grekyan 2006: 178-179; Hmayakyan 1990: 69-70. 75 Ghirshman 1954: Table 41. 76 Monuments of this type are sometimes separated from the general context of megaliths and are attributed to the so-called paramegaliths (Tsonev 2010: 51). 77 Ter-Martirosov 2015: 37. 78 Akopyan 2003: 133. According to Z. Le Rouzic, similar monuments represent a transitional stage between the ‘border stone’ and menhir (Le Rouzic 1921: 13; 21; cf. Hakobyan and Gnuni 2007: 92-95). 79 According to H. Avetisyan, during S. Sardaryan’s expeditions, similar monuments have been documented also in other early agricultural settlements of the Ararat Valley.

Figure 31. Khnatsakh 1, stela.

72

contrast to the observed monuments, these two are placed in dwellings and sanctuaries. Their architectural features are not as clear as in the case of other menhirs. In addition, there are primitive phallic elements on the top of the Shengavit one.80

80

16

Areshian 2005: 83-84; Gnuni 2006: 245-246; Sardaryan 2004: 295.

Hayk Avetisyan et al.: Armenian Standing Stones as an Object of Archaeological Study

Figure 32. Artavan 2, menhir-obelisk.

Transitional forms

3. Obelisk-anthropomorphic sculpture

There are standing stones that combine various types:

The first monument (Harzhis 27), which was temporarily exibited in the History Museum of Armenia and has now been returned to its former place (near the House of Culture in Harzhis village), has a cylindrical body (dimensions: 225 × 37 × 35 cm). The human face is depicted on the upper part of the body, on a cylinder protrusion. The presence of a cylinder base to depict the face is not uncommon in Armenian monuments (Aygeshat, Erebuni, Karmir Blur).87 The eyes and the mouth are engraved, the forehead and the nose are depicted in relief.88 It resembles the stela from Aygeshat, which stands out with its elongated from (205 cm) and thickness (10 cm).89 Thus, it can be concluded that the example under discussion marks the anthropomorphization of menhirs.

1. Menhir-obelisk On a number of documented standing stones features typical of both menhirs and obelisks are present. The latter include relatively fine processing and/or a clear quadrilaterality in the section. Thus, the lateral and the front parts of the menhir at Artavan 2 are well worked and the corners are emphasized (Figure 32). Menhirs with square, sloping, and accentuated corners are attested in a number of places (menhir-khachkar at Dadivank monastery, Kaput Khach menhir-khachkar of Abovyan, menhir-khachkar inserted in the southern wall of Barevakhach chapel of Harzhis).81 Some examples of this type are known in settlements and necropolises (e.g. Shahumyan).

4. Menhir-phallic sculpture

2. Menhir-anthropomorphic sculpture

The idols from Harichavank 3,90 Geghashen,91 and Krapashti Tner 1, can be attributed to this group, which stand out with an accentuated phallic head, as well as their roughly processing.

This type of statue is sometimes considered one of the stages of anthropomorphization of the monument, which can be expressed by giving the object stylized human outlines while using natural stones reminiscent of human outlines.82 A similar example was found in the rocks surrounding the Lesser Al Lake (Lake 8).83 Although the monument is very stylized, some features (e.g., pentagonal upper part84) suggest that it is an anthropomorphic one.85 Schematic anthropomorphic monuments of the classical period from Hoghmik and Tsitsernakaberd can be included in the number of further examples of this type. 86

5. Vishapoid There is a group of isolated standing stones, that have elements of the standing stones with images of bull and fish (known as vishap stones) and menhirs. The vishapoid differs from the vishap stone by relatively small size, shape, and, most importantly, by the absence of images (Yeghnajur 1; Figure 33).92

Avetisyan et al. 2015: 65; Avetisyan et al. 2017: Figure 5/17. Beneteau-Dovilard 2002: 570. 83 Avetisyan et al. 2015: 66. 84 There is a pentagonal upper part also in the case of some menhirs, for example, on one example from Bandivan (Investigations by A. Bobokhyan). 85 At the same time, a certain genetic connection is observed between the menhirs and anthropomorphic statues: schematic profile representation, anthropomorphization of the menhir’s head (Beneteau-Douilard 2006: 570). 86 Hakobyan 2010: 23. 81 82

Yesayan 1980: 54-55, Tables 42/1, 43/3, 46/2. Avetisyan et al. 2015: 51, Figure 10. 89 Investigations by: H. Avetisyan, A. Bobokhyan, A. Gnuni, L. Mkrtchyan. Some parallels are observed in Arhz-Zelenchuk and Arghun region, where anthropomorphic stelae have been discovered (Kuznetsov 1977: 28). 90 Khachatryan 2003: 19-21. 91 Investigations by V. Gayseryan. 92 Cf. Bobokhyan et al. 2015: 298. 87 88

17

Systemizing the Past

Figure 33. Yeghnajur, vishapoid.

sculptural, spatial, contextual features), as a result of which their preliminary typology was suggested. According to this, the standing stones of Armenia can be classified into the following groups: menhirs, obelisks, stelae, other types, transitional forms (menhir-obelisk, menhir-anthropomorphic sculpture, obelisk-anthropomorphic sculpture, menhir-phallic sculpture, and vishapoid), and natural standing stones (for their distribution cf. Figure 34). These units differ in both structure and archaeological context and period of creation. The origins of the perception of standing stones can be attributed to the early agricultural societies, although their clear documentation is visible during the Bronze and Iron Ages. Many of these ancient monuments, transformed in their form and content, have survived to our times.

Natural standing stones Some natural objects, such as worshiped rocks, can be classified as standing stones.93 Among the similar natural monuments in Armenia are the worshiped rocks in Zangezur, Western Armenia, and Artsakh.94 On these rocks some natural and man-made details are present that bring them closer to the processed stones. Among these are holes (Tasik, Kyatuk, Badrkhan),95 phallic sculptures and carvings (Karchvan),96 and petroglyphs (Karahunj).97 The next group of unelaborated standing stones is represented by rocks on tombs (placed on the top of the tomb or in the chamber, cf. Aghvani, Karahunj).98 Conclusions Standing stone monuments have been the subject of special attention in Armenia since the earliest times. A number of monuments mentioned in Armenian mythology, fairy tales, and epics can even be identified with the menhirs and their complexes. Standing stones have long been the subject of research, however no generalization on the issue has been made yet. In this paper, for the first time, an attempt has been made to formulate the criteria for defining Armenian standing stones (according to the processing, architectural,

Bibliography Agatangeghos, 1909. History of Armenia by Agatangeghos (Comp. by G. Ter Mkrtchyan and S. Kanayants). Tbilisi, Martirosyants Printing House (in Armenian). Akopyan, A. 2003. Temple complex of Hoghmik. In L. Abrahamyan, R. Badalyan, L. Vardanyan, A. Kalantaryan and G. Kharatyan (eds), Archaeology, Ethnology and Folklore Studies of the Caucasus: 130–136. Etchmiadzin, Hazarashen (in Russian). Alishan, Gh. 1910. Ancient Faith or Pagan Religion of Armenia. Venice, S. Lazarus (in Armenian). Areshian, G. 1978. Investigations in Shamiram. In Archaeological Discoveries in 1977: 503–504. Moscow, Nauka (in Russian). Areshian, G. and Tumanyan, G. 1991. Tomb no. 5 of Aruch cemetery. In G. Tiratsyan (ed.), The Scientific Session

Cf. Shirakuni 1904: 183. Mirakhorean 1885: 34, 46, 119; Shirakuni 1904: 182-217; Lalayean 1897: 188-191; 1914: 26, 28-30. Cf. Mkrtchyan 2015: 132-133. 95 Lalayean 1897: 190-191. Cf. Mkrtchyan 2015: 132. 96 Lalayean 1897: 191. Cf. Mkrtchyan 2015: 132. 97 Mkrtchyan 2015: 132. 98 Mkrtchyan 2015: 132. The descriptive reference of Karahunj is composed by E. Ayvazyan. 93 94

18

Hayk Avetisyan et al.: Armenian Standing Stones as an Object of Archaeological Study

Figure 34. Distribution of standing stones of Armenia.

Reality: 591–623. Yerevan, Institute of Archaeology and Ethnography Press (in Armenian). Bayern, F. 1871. On ancient structures in the Caucasus. In Collection of Information about the Caucasus 1: 306– 325 (in Russian). Baynes, E. 1905. The Megalithic Remains of Anglesey. London, Society. Barkhudaryan, S. 1935. Monuments of Material Culture of Soviet Armenia. Yerevan, Pethrat (in Armenian). Beneteau-Douilard, G. 2006. Anthropomorphic standing stones alignment in South Vandee. In Origin and Development of the Megalithic Monuments of Western Europe. International Conference, Bougon, 26-30 October 2002, vol. 2: 560–576. Bougon, Museé des Tumulus de Bougon. Berlin, Yu., Sychev, Yu. and Shalaev, I. 1979. Processing of Decorative Building Stone. Leningrad, Stroyizdat (in Russian). Bobokhyan, A., Gilbert, A. and Hnila, P. 2015. Archaeology of vishap stones. In A. Petrosyan and A. Bobokhyan (eds), The Vishap Stones: 299–396. Yerevan, Gitutyun (in Armenian). Bottrell, W. 1873. Traditions and Hearthside Stories of West Cornwall. London, Beare and Son. Bray, W. and Trump, D. 1990. Archaeological Dictionary. Moscow, Progress (in Russian).

dedicated to the Results of Archaeological Fieldwork in the Republic of Armenia in 1989-1990: Abstracts of Reports: 24–26. Yerevan, Academy of Sciences of the Armenian SSR (in Armenian). Areshian, G. 2005. Early Bronze settlements in the Ararat Plain and its vicinity. Archäologische Mitteilungen aus Iran und Turan 37: 71–89. Auberry J., w.d. Templa Druidum, P. 1, http://www. sarsen.org/2013/07/monumenta-britannica-byjohn-aubrey.html. Auberry, J. 1862. Wiltshire. The Topographical Collection. London, Longman & Co. Auden, A.1906. Historical and Scientific Survey of York and District. York and London, Sampson-Simpkin, Marshall & Co. Ltd. Avetisyan, H. 2016. Pulusi stelae discovered from Aramus. Hayagitut‘yan harc‘er (Issues of Armenology) 2: 109–115 (in Armenian). Avetisyan, H., Bobokhyan, A., Gnuni, A. and Sargsyan, G. 2015. Bronze and Iron Age Sacred Landscape of Syunik. Yerevan, Yerevan State University Press (in Armenian). Avetisyan, H., Gnuni, A., Mkrtchyan, L. and Bobokhyan, A. 2019. The cupmarks of vishap stones and their general parallels. In A. Bobokhyan, A. Gilibert and P. Hnila (eds), The Vishap on the Border of Fairy Tale and 19

Systemizing the Past Cambry, M. 1805. Monumens celtiques ou Recherches sur le culte des pierres. Paris, Chez. mad. Johanneau, Librarie. D’Auvergne-Corret, T. 1797. An V de la Republique Francaise. Origines gauloises, des plus anciens peoples de l’Europe. Paris, Chez. Quillau, Imprimeur Libraire. De Morgan, J. 1925. Prehistoric Humanity. Moscow, State Printing House (in Russian). Ferguson, J. 1872. Rude Stone Monuments. London, John Murrey. Girshman, R. 1954. Iran. London, A Pelican Book. Gnuni, A. 2004. Types and ritual functions of phallic monuments of the Armenian Highland. Banber Erevani Hamalsarani (Herald of Yerevan University) 2: 122–130 (in Armenian). Gnuni, A. 2006. Shrines located in Shengavit cultural settlements and their complexes. Patma-banasirakan handes (Historical-Philological Journal) 1: 236–253 (in Armenian). Gnuni, A. 2014. Keren at the crossroads of civilizations. Hayagitut‘yan Harc‘er (Issues of Armenology) 2: 149–167 (in Armenian). Gnuni, A., Tadevosyan, A. and Mkrtchyan, L. 2017. Shresh Blur. In A. Piliposyan (ed.), Metsamorian Readings, vol I: 25–38. Yerevan, Service for the Protection of Historical Environment and Cultural Museum-Reservations (in Armenian). Gobo, J. R., Castilio, A. D. and Guintana, J. C. L. 1995. Menhirs/Monolitus. Estucturas monoliticas en el esctor la cornisa lanta briea. In Actas de XXII congress nacional de Arqueología, Vigo 1993, vol. 1: 55–62. Vigo, Artes Gráficas Galicia. Grekyan, Ye. 2006. The will of Menua and the gods of Urartu. Aramazd: Armenian Journal of Near Eastern Studies 1: 150–195. Gulakyan, S. 1983. Index of Motives of Armenian Fairy Tales. Yerevan, Yerevan University Press (in Russian). Hakobyan, H. and Gnuni, A. 2007. Observations on the unity of the functional-symbolic functions of the road and street. Etchmiadzin 9: 89–97 (in Armenian). Hakobyan, H. 2010. The main results of excavations at Tsitsernakaberd in 1998, 2007. Hushardzan (Monument) 6: 21–44 (in Armenian). Haxthausen, A. 1857. Transcaucasian Territory. St. Petersburg, Printing House of the General Headquarter of Educational Institutions (in Russian). Hmayakyan, S. 1990. The State Religion of the Kingdom of Van. Yerevan, Academy of Sciences of the Armenian SSR (in Armenian). Hmayakyan, S., Hakobyan, H. and Melkonyan, M. 2010. Ashotsk, Hartashen and Navur menhirs. In A. Nersisyan (ed.), Collection of Scientific Articles: 25–33. Yerevan, Yerevan State University Press (in Armenian). Holmes, R. 1907. Ancient Britain and the Invasions of Julius Caesar. Oxford, Clarendon Press. Hummel, Ya. 1940. Archaeological Essays. Baku, AzFAN Press (in Russian).

Ivanovski A. 1911. On the Transcaucasia. Materials on the Archaeology of the Caucasus, vol. VI. Moscow, A. I. Mamontov’s Press (in Russian). Jeyranov, F. 1898. Chaikend village of Elizavetpol District of the same province. Collection of Materials on the Description of Localities and Tribes of the Caucasus 25: 59–100 (in Russian). Jones, I. 1655. The Most Notable Antiquity of Great Britain Vulgarly Called Stone-Heng. London, Printed by James Flesher. Kalantar, A. ​​1925. The Stone Age in Armenia. Nork 5-6: 208–232 (in Armenian). Kamenev, N. 1867. A Few Words about the Colonization of the Western Caucasus in General and the Psekupsk Regiment in Particular. Kuban Military Statements, Ekaterinodar, October 7, N 39 (in Russian). Khachatryan, N. 2009. Stone stelae from Harich Monastery. In A. Kalantaryan and R. Badalyan (eds), Ancient Culture of Armenia 3: 118–121. Yerevan, Mughni (in Armenian). Kuznetsov, V. 1977. In the Upper Reaches of Zelenchuk. Moscow, Isskustvo (in Russian). Lalayean, E. 1897. Varanda. Azgagrakan handes (Ethnographical Journal) 2: 3–244 (in Armenian). Lalayean, E. 1898. Zangezur. Azgagrakan handes (Ethnographical Journal) 4/2: 7–116 (in Armenian). Lalayean, E. 1904. Sharur-Daralageaz Province. Azgagrakan handes (Ethnographical Journal) 12: 235– 293 (in Armenian). Lalayean, E. 1908. Excavations in Nor Bayazet Province. Azgagrakan handes (Ethnographical Journal) 19/1: 59– 84 (in Armenian). Lalayean, E. 1914. Vaspurakan, Faith. Azgagrakan handes (Ethnographical Journal) 25: 21–60 (in Armenian). Lalayean, E. 1931. Excavations of Tombs in Soviet Armenia. Yerevan, Melkonyan Foundation Press (in Armenian). Lebeuf, A. 2012. What does the Bible tell us about Megaliths? Studia Religiologica 1: 37–56. Le Rouzic, Z. 1921. The Megalithic Monuments of Carnak and Locmariaquer. London, H.J. Hall and Company. Lisitsyan, S. and Bayburdyan, E. 1928. The List and Description of Antiquities. Tbilisi and Yerevan, Press of the Central Provincial Bureau(in Armenian). Lisitsyan, S. 1938. Koshun-Dash megalithic settlement in Sisian (Zangezur). Collection in Memory of acad. N. Ya. Marr: 709–772. Moscow and Leningrad, Press of the Academy of Sciences of the USSR (in Russian). Magnus, O. 1558. Historiae de gentibus septentrionalibus. Antverpiae, Apud Ioannem Bellerum. Magnus, O. 1658. A Compendious History of the Goths, Swedes and Vandals and other Northern Nations. London, Printed by J. Streater. Martirosyan, H. 1961. City of Teishebaini. Yerevan, Academy of Sciences of the Armenian SSR (in Russian). Martirosyan, H. 1974. Argishtikhinili. Yerevan, Academy of Sciences of the Armenian SSR (in Russian). 20

Hayk Avetisyan et al.: Armenian Standing Stones as an Object of Archaeological Study Martynov, A. and Sher, Ya. 1989. Methods of Archaeological Research. Moscow, Vysshaya shkola (in Russian). Matyushkin, G. 1996. Archaeological Dictionary. Moscow, Prosveshenie (in Russian). Mirakhorean, M. 1884. Descriptive Trip to the ArmenianPopulated Province of Eastern Turkey, vol. 1. Constantinople, Dimaxean Press (in Armenian). Mishev, G. 2016. The stone-faith and believs from Bratsigovo, megalithic monuments and cult practicies. In Proccedings of the II International Symposium: 316–322. Blagoevgrad, ‘Neofit Rilski’ University Press. Mkrtchyan, L. 2015. Monuments of the sacred landscape of Syunik in state archives. In H. Avetisyan, A. Gnuni, A. Bobokhyan and G. Sargsyan, Bronze and Iron Age Sacred Landscape of Syunik: 131–144. Yerevan, Yerevan State University Press (in Armenian). Özbek, O. 2006. Manhirs in the graveyards: fact or fiction? A reconsideration of erected stone monuments of Galipoli Peninsulia. In O. Özbek (ed.), Funeral Rites, Rituals and Ceremonies from Prehistory to Antiquity (Proccedings of the International Worshop ‘Troas and Neighbours’): 83–96. Istanbul, IFEA Press. Piliposyan, A. and Avagyan, A. 2016. Zorats Karer. Yerevan, RA Ministry of Culture Press (in Armenian). Pope, V. and Miranda, A. 1999. Geomorfology of megaliths: Neolithic landscapes of the Alto Alentejo, Portugal. Middle States Geographer 32: 110–124. Rostomov, I. 1898. Akhalkalaki District in archaeological terms. Collection of Materials on the Description of Localities and Tribes of the Caucasus 25: 1–133 (in Russian). Rowe, S. 1830. Antiquarian Investigations in the forest of Dartmoor, Devon. In Transactions of the Plymouth Institute: 180–212. Plymouth-London, Printed and published by Rowe. Ressler, E. 1904. Archaeological research by E. Resler in the Elizavetpol Province. News of the Imperial Archaeological Commission 12: 36–60 (in Russian). Sardaryan, S. 2004. Armenia, the Cradle of Civilization. Yerevan, Yerevan State University Press (in Armenian). Shirakuni, N. 1904. Zangezur District, Elizavetpol Province. Collection of Materials for the Description of Localities and Tribes of the Caucasus 34: 182–217 (in Russian).

Škorpil, K. V. 1905. Megalithic monuments, materials for Bulgarian antiquities. Aboba Plisaka. Bulletin of the Russian Archaeological Institute in Constantinople 10: 372–384 (in Bulgarian). Srvandztyants, G. 1874. Grots and Brots – Traditional Armenian Tails and Davit of Sasun or Mher’s Door. Constantinople, Tntesean Press (in Armenian). Stukeley, W. 1740. Stonehenge, a Temple Restor’d to the British Druids. London, Printed for W. Innys and R. Manby, at the West End of St. Paul’s. Ter-Martirosov, F. 2015. Ervandashat. In E. Minasyan, H. Petrosyan, H. Melonyan, Z. Hakobyan and L. Mikayelyan (eds), Genesis Forest: 33–54. Yerevan, Yerevan State University Press (in Russian). Toramanyan, T. 1942. Materials on the History of Armenian Architecture, vol. 1. Yerevan, ArmFAN (in Armenian). Torosyan, R. 1971. Diary no. 1, Necropolis of Aygeshat Village of Etchmiadzin Region, Collection of R. Torosyan of the Historical-Ethnographical Museum of Etchmiadzin (in Armenian). Tsonev, L. 2010. Megalites in Bulgaria. Sofia, Farago (in Russian). Tsotsikean, S. 1917. Caucasus-Ararat, vol. 1. Fresno (in Armenian). Tumanyan, G. 2015. The cultic-applied functions of vishap stones. In A. Petrosyan and A. Bobokhyan (eds), The Vishap Stones: 99–120. Yerevan, Gitutyun (in Armenian). Tumyan, H. 1937. Historical Monuments of Soviet Armenia, Dilijan Region. Yerevan, Yerevan State University Press(in Armenian). Uvarova, P. S. 1900. On the desirability of studying the Caucasian dolmens. Drevnosti (Antiquities) 16: 157– 164 (in Russian). Wilson, T. 1888. Megalithic monuments of Brittany. American Naturalist XXII/259: 574–589. Worm, O. 1643. Danicorum Monumentorum. Hafniae, Apud Ioachimum Moltkenam Bibliopolam ibidem primar. Xnkikyan, O. 2002. Syunik during Bronze and Iron Ages. Barrington, Mayreni Press. Yesayan, S. 1980. Sculpture of Ancient Armenia. Yerevan, Academy of Sciences of the Armenian SSR (in Russian).

21

Neolithization of Armenia: General Trends and Patterns of Development Ruben Badalyan and Armine Harutyunyan

Institute of Archaeology and Ethnography, NAS RA, Yerevan, Armenia Abstract: The oldest archaeological culture in the South Caucasus based on a production economy, with the first documented examples of housebuilding, ceramic production, and metalworking, is the Neolithic ‘Aratashen-Shulaveri-Shomutepe’ (AShSh) culture (c. 6000 – 5400 BC). This Neolithic complex demonstrates an already fully developed agricultural and pastoral economy based on the breeding of cattle and caprines and the cultivation of cereals. The formation process of this producing economy and, in general, the genesis of this culture has not yet been investigated. It is clear that the formation and development of the AShSh complex took place in the conditions of active cultural and economic contacts with synchronous cultures in the west and south, which were sources and conductors of not only some exotic, obviously prestigious, items and materials, but, possibly, a number of technological and cultural innovations. The article attempts to compare the main development trends and trace the general patterns of development of a number of Neolithic technologies in the Upper Tigris and Euphrates basins of the 8/7th millennium BC and the Ararat valley of the 6th millennium BC. Keywords: Neolithization, Ararat valley, Aknashen, Aratashen-Shulaveri-Shomutepe culture, ceramic production, obsidian, Halaf/Samarra pottery

In the sequence of archaeological cultures in Armenia, the Neolithic ‘Aratashen-Shulaveri-Shomu’ (hereafter AShSh) culture has formed only relatively recently and taken its place on the periodization-chronological scale.1 The significance of this complex, dating from 6000-5400 BC, is determined primarily by the fact that it represents the oldest culture known in the South Caucasus today, based on a production economy, with the first documented examples of housebuilding, ceramic production, and metalworking. This Neolithic complex demonstrates an already fully developed agricultural and pastoral economy based on the breeding of cattle and caprines and the cultivation of cereals (among which Triticum aestivum and Hordeum vulgare dominate). The formation process of this producing economy and, in general, the genesis of this culture has not yet been investigated. The interval of one and a half thousand years (about 7500-6000 BC) between the AShSh culture and rare early Neolithic sites scattered over a vast territory (Kmlo-2/Apnagyugh-8, etc.) is not supported by real material. Only recently there has been a tendency toward its partial filling as a result of excavations of the site of the first half of the 7th millennium BC Lernagog at the northwestern tip of the Ararat valley/ southwestern foot of Aragats.2 Clearly, under such conditions, both autochthonous and migration models of the Neolithization of the South Caucasus, discussed since the Soviet era, look equally probable. 1 2

Chataigner et al. 2014. Arimura et al. 2018.

Systemizing the Past (Archaeopress 2023): 22–29

Although all known AShSh settlements arose in uninhabited places and already in their lower layers there is a formed artifactual complex and an almost complete set of domesticated plant and animal species, there is still no sufficient reason to consider the AShSh horizon intrusive in the sequence of archaeocultures of the South Caucasus,3 as suggested, for example, by P. Kohl.4 In fact, no culture is known in the Near East that could be directly linked to the Neolithic culture that developed in the South Caucasus. Additionally, it is clear that the formation and development of the AShSh complex took place in the conditions of active cultural and economic contacts with synchronous cultures in the west and south, which were sources and conductors of not only some exotic, obviously prestigious, items and materials, but, possibly, a number of technological and cultural innovations.5 It should be noted that although the Neolithic settlements of the Kura basin have been studied on a much larger scale than those in the Araxes basin, these connections are more clearly manifested in the sites of the Ararat valley. These connections have long been documented by finds on the Neolithic sites of the South Caucasus, primarily Badalyan et al. 2007: 60. Kohl 2007: 68. 5 Cf. Özdoğan 2018: 15: ‘besides the primary components of the Neolithic package, the rest of the material evidence of the Caucasian Neolithic is incompatible with that of the Near East, indicating that the process of neolithisation was not due to an endemic migration but rather to the transfer of technologies.’ 3 4

Ruben Badalyan and Armine Harutyunyan: Neolithization of Armenia Detailed similarities are evident for samples of the obsidian industry (the single-platform prismatic cores, long standardized blades), several categories of bone tools (picks, arrows, scapulas, spoons), sharpeners/ straighteners, adze/axes. Because of this, Tilkitepe appeared as a potential source of Halaf ceramics found at the sites of the Araxes basin (in Aratashen, Aknashen, Masis Blur, Nakhichevan Kültepe I). However, Halaf ceramics Tilkitepe certainly differs from the findings of Halaf ceramics in the South Caucasus.16 It should also be noted that, along with the Halaf, Samarra-related ceramics were also found at the Neolithic sites of the Kura-Araxes interfluve (Aknashen,17 Hacı Elamxanlı18), which is absent from Tilkitepe.

in the Araxes basin, samples of painted Halaf pottery (Nakhichevan Kültepe I).6 Since then, the number of finds of Halaf (and Samarra) pottery in the Ararat valley has been continuously growing with each new site and each field season (Aratashen, Aknashen, Masis Blur).7 Aside from this visually detectable import, precision analyses have established the fact of infiltration of single obsidian samples from the deposits of the Van Lake basin on the settlements of the Ararat valley and the Araxes basin (Meidan Dağ – Aratashen, Aknashen, Masis Blur, Nakhichevan Kültepe I, Nemrut ‘South’ – Sardarapat).8 Clearly, their infiltration in non-marketable and, accordingly, economically unmotivated volumes in the region, which are abundantly provided with their own obsidian outlets, was not caused by a need for raw materials. Rather, these finds reflect non-economic contacts or, at least, contacts not directly related to the obsidian trade. The direct correlation between the finds of Halaf pottery and ‘imported’ obsidian suggests that the latter penetrated into the sites of the Araxes basin in the course of the influx of Halaf pottery.9

In recent years, as a result of intensive research at 6th millennium BC sites in the Araxes and Kura basins, not only has the number of objects of various categories significantly increased, reflecting the links between the South Caucasus and the North Mesopotamian world, but above all, the number of observed patterns of development of the main Neolithic technologies has significantly increased.

Single samples of painted ceramics have also been found in a number of settlements in the Kura basin, in Imiris Gora10 and Gargalartepesi;11 however, neither the stratigraphic position nor the cultural attribution of these finds has been established. According to Narimanov, these ceramic sherds are more similar to the pottery of the Mil steppe sites.12

The localization of possible sources of these innovations, it seems to us, shifts the focus of attention to the southwest, to the region of the basins of the upper Tigris and Euphrates. We consider it necessary to emphasize that the purpose of this article is not to study the mechanisms of the Neolithization of the South Caucasus and the formation of the AShSh complex, nor the identification of typological analogies. Rather, our priority is to draw attention to an area whose aggregate data mark, in our view, a source of potential influence.

To date, only at one, the earliest, Kura site – Hacı Elamxanlı Tepe  –  two samples of Halaf/Samarra pottery have been discovered.13 It is significant that obsidian from the sources of the basin of Lake Van is not represented at all at the synchronous sites in the Kura basin; even in the largest series of obsidian analyzed – 900 samples from the Göytepe – there is not a single Van or Nemrut sample.14 The only imported sample from a source near Dogubayazit (Tendürek?) was recorded by J. Blackman’s analysis in Khramis Didi Gora. Thus, there is a direct correlation between the absence/paucity of imported Samarra/Halaf pottery in the settlements of the Kura basin and the absence of Van obsidian.

The Aknashen Settlement: General Characteristics of the Artifactual Complex and Its South Caucasian Analogies The currently known sites of the culture under consideration form two clearly localized groups – two ‘oases’ confined to the alluvial river valleys of the Araxes and Kura and representing two main variants of the culture.19 The northern ‘oasis’ in the middle reaches of the Kura River (Kvemo Kartli and Ganja-Gazakh plains) includes the settlements of Shulaveris Gora, Imiris Gora, Gadachrili Gora, Dangreuli Gora, Arukhlo, Khramis Didi Gora, Hacı Elamxanlı Tepe, Shomutepe, Toyretepe, Gargalartepesi, Göytepe, and Mentesh Tepe . The southern ‘oasis,’ located inside Armenia, includes the settlements Aratashen, Aknashen, Masis Blur, and

The closest outpost of the Halaf culture to the sites of the Araxes basin is the settlement of Tilkitepe on the eastern shore of Lake Van.15 The materials of its lower, III, layer show the greatest number of analogies to the AShSh complex outside the Araxes and Kura basins. Munchaev 1975. Badalyan et al. 2010; Palumbi 2007. Badalyan 2002; Martirosyan-Olshansky 2018b. 9 Badalyan 2002; 2010. 10 Kiguradze 1976: Tables 35/12, 44/9. 11 Narimanov 1987: 121, Figure 27. 12 Narimanov 1987: 122. 13 Nishiaki et al. 2015a: Figure 3f, g. 14 Nishiaki et al. 2019b. 15 Korfmann 1982. 6 7 8

Palumbi 2012; personal observations of the authors at the Istanbul University museum in 2019. 17 Badalyan et al. 2010. 18 Nishiaki et al. 2015a: Figure 3f, g. 19 Chataigner et al. 2014. 16

23

Systemizing the Past Tsaghkunk, is localized in the middle reaches of the Araxes River, in the Ararat valley; the sites are confined to the left tributaries of the Araxes – the Sevjur (Metsamor), Kasakh, and Hrazdan rivers.

AShSh (approximately 6000/5950-5800 cal. BC), and is recorded more or less simultaneously on the sites of both the Ararat valley and the Kura basin.25 The above-lying (V-II/I) horizons of Aknashen are characterized by circular architecture; the buildings were erected using lumps and layers of mud/cob. The corresponding complex of material culture allows us to trace the process of emergence and development of ceramics, which looks as follows: in the upper level of horizon V (5780-5750 cal. BC) the first relatively few samples of ceramics with mineral inclusions26 of local production are found (with a predominance of Grittempered II pottery);27 in horizon IV (5750-5690 cal. BC), the amount of local ceramics with mineral inclusions increases sharply, the insignificant predominance of ceramics of the Grit-tempered II group in comparison with the Grit-tempered I group continues; in horizons III (5690-5600 cal. BC) and II (5600-5450/5400 cal. BC), the amount of ceramics doubles, the leading position is occupied by the Grit-tempered I group. In this horizon, ceramics with plant inclusions seem to appear for the first time. The upper horizon (I) is characterized by its complete predominance.

In the Ararat valley, the most complete, differentiated and long-lasting sequence of the Late Neolithic layers is attested in Aknashen. Its stratigraphic column with a total thickness of about 5 m, dated 5950-5400 BC, seems to reflect almost the entire chronological range of the culture and, moreover, includes a horizon (VII), the data from which makes it possible to raise the question of its formative stage.20 In this sequence, two complexes are distinguished, separated by a hiatus (horizon VI, flood episode) and differing in a number of important indicators. Considering all factors, we distinguish, first of all, the complex of the oldest horizon, VII. This horizon, dated 5950-5870 BC (according to the C14 dates currently available and which do not concern the deepest strata), overlies the virgin soil, is characterized by the coexistence of buildings in rectangular and circular planes erected using lumps and layers of mud/cob, the absence of local pottery with mineral and plant inclusions, a high content (compared to the overlying horizons) of imported painted Samarra/ Halaf and monochrome ceramic sherds and bladelets/ microblades, bullet cores and nuclei on pebbles, and microliths, as well as the predominance of barley. According to these parameters, this complex, on the one hand, differs from the overlying (V-II/I) horizons, which represent the AShSh in its fully formed form, on the other hand, it exhibits a significant degree of similarity with the synchronous (5950-5800 BC)21 Hacı Elamxanlı settlement in the Kura basin. We regard the complex represented by these two sites as the oldest, formative, stage of the AShSh.22 C14 dates indicate that settlements of this complex appeared and developed simultaneously in both the Kura and the Araxes River basins.23

The first steps in ceramic production observed according to the Aknashen data clearly reflect the general laws of the process of the appearance and development of ceramics in the South Caucasus in the first half of the 6th millennium BC. Similar trends were traced not only at other sites of the Ararat valley (Aratashen, Masis Blur), but also at most settlements in the Kura basin. Thus, a comparative analysis of data from the early Hacı Elamxanlı (c. 5950-5800 BC) and the late Göytepe (c. 5650-5450 BC) shows that ‘pottery was rare in the lowest levels but rapidly increased from the middle part of the sequence onward. The earlier pottery assemblages are also characterized by the almost exclusive use of mineral-tempered pottery, while the later ones showed more prevalent use of plant-tempered’ pottery.28 At the Hasansu settlement, the upper cultural layer contains relatively abundant fragments of ceramic vessels made of clay with inorganic inclusions and clay with a vegetal mixture. Below, fragments of ceramics are extremely rare and represent very small fragments of vessels

Structures of a rectangular plan, in addition to those excavated in the oldest, VII, Aknashen horizon, were also recorded at the Hasansu I settlement in the Kura basin. In the lower, above-subsoil horizon of the site, fragments of the construction remains of the rectangular-square plan were revealed. According to the author of the excavation, rectangular constructions precede round-oval ones.24 It should be emphasized that the 14C dates allow us to synchronize the lower horizons of both sites. Thus, it is clear that the rectangular plan is characteristic of the most ancient, formative, stage

25 Cf. Özdoğan 2018: 23: ‘however, the possibility of a short-lived earlier stage with rectangular buildings preceding the round-building phase should not be discarded.’ 26 A petrographic study of samples of clays and ceramics from Aratashen and Aknashen showed that the clays used for the production of ceramics contain a natural admixture of sandy lavas, crystals, and volcanic glass. In other words, the sandy ‘admixture’ could also be of natural origin, that is, be a part of the original material. In a number of cases, it is possible that a granulometrically larger sandy admixture from lacustrine-fluvial sediments was artificially added. The clay of the Grit-tempered II pottery was obviously sieved, cleaned of large impurities, after which chamotte was added, possible also manure/ organic matter (Harutyunyan 2011). 27 Harutyunyan 2014. 28 Nishiaki et al. 2015a: 283. See also Nishiaki et al. 2019a.

Badalyan and Harutyunyan 2014. Nishiaki et al. 2015a; 2015b. 22 Badalyan and Harutyunyan 2014. 23 Chataigner et al. 2014. 24 Museibli 2017. 20 21

24

Ruben Badalyan and Armine Harutyunyan: Neolithization of Armenia general vector at the present time can probably be expanded culturally and concretized in a geographic aspect.

(intrusion?). In the above-subsoil horizon, fragments of ceramic products were not found.29 An extremely small amount of ceramics, with mineral admixtures, is noted in the lower layers of Shulaveri (IX-III) and Imiris Gora (VII-VI).30 The Gadachrili Gora Neolithic complex (59205650 cal. BC)31 is dominated by ceramics with mineral inclusions; an extremely small amount of ceramics with organic additives appears only at the very end of Phase I and increases from 1% in the Neolithic to 19% in Late Neolithic/Early Chalcolithic.32

Isolated finds of Samarra-related ceramics in the settlements of the South Caucasus (Aknashen, Hacı Elamxanlı Tepe), despite their small number and, accordingly, random nature, nevertheless serve to a certain extent as an additional geographical marker. The northernmost sites with Samarra and/or Halaf pottery are located in the area of the upper basins of the Tigris and Euphrates; it is this area that seems to be the most likely source of imported pottery found in the South Caucasian settlements.

The obsidian industry at Aknashen, like the synchronous settlements of the Ararat valley (Aratashen, Masis Blur, Tsaghkunk) and AShSh in general, is morphologically and technologically characterized by a predominance of long standardized blades and bladelet blanks manufactured from prismatic cores and used in agricultural activities.

Such a probability is determined not so much by the external similarity of individual artifacts, which have a very wide distribution area, but by the totality of culturally significant features of material culture and the similarity of technology development trends. Despite the fact that, at first glance, the analogies given below between AShSh and the Neolithic of southeastern Turkey seem too general, their mutual compatibility gives the observation a certain significance.

These blanks are obtained through a combination of pressure techniques (lever and crutch), and indirect percussion. Use of the pressure technique (using a lever and/or a crutch) has been suggested also at sites in the Kura basin, such as Hacı Elamxanlı,33 Göytepe,34 Arukhlo,35 and Mentesh Tepe.36

A certain similarity between the Neolithic settlements of the Ararat valley, on the one hand, and the chronologically preceding sites of the region of the upper basins of the Tigris and Euphrates, on the other, can be traced in the field of housing construction, stone industry, and ceramic production, that is, in the most innovative areas. The extent to which this similarity is accidental, whether it is a consequence of contacts or a reflection of the general patterns of development of early agricultural societies, remains to be determined by further research.

Outside the South Caucasian ‘Oases’ - the Neolithic of the Upper Tigris and Euphrates Basins The aforementioned finds of imported ceramics in the settlements of the South Caucasus of the 6th millennium BC, being indisputable evidence of one or another form of contact of the AShSh carriers with Northern Mesopotamia, first of all reflect, as it seems, ‘the result of occasional and mediated interactions with the Halaf world’.37 Traditionally, among the evidence of contacts, but of a general nature, without a definite vector, there is also a round plan of buildings, characteristic of both the AShSh and the Halaf culture.38 This dominant

The northernmost settlements of the Halaf, in addition to the aforementioned Tilkitepe – Koruchutepe and Tülintepe – are located in the Upper Euphrates Valley. The northernmost settlement of the Samarra culture in the region is the settlement of Hakemi Use (61005950/5900 cal. BC) in the Upper Tigris basin.39

Museibli 2017. Kiguradze 1976. Cf. Batiuk et al. 2017: 196. 31 Hamon et al. 2016. 32 Batiuk et al. 2019. It should be noted that the data from Kültepe I of Nakhichevan clearly contradict the models observed for the appearance and the development of pottery in the South Caucasus. According to Marro et al. (2019: 92), the pottery, of which the absolute majority is plant-tempered, was already present in large quantity at the very beginning of the Neolithic sequence, dated to between 6200 and 6000 cal BC. The frequency was quite stable throughout the sequence, the fabrication technique had not changed, nor had the vessel shapes, except for the disappearance of carinated bodies in level 2. At the same time, to be noted is the significant morphological similarity of the Kültepe pottery with the Grit-tempered group of Aknashen and the rarity of decoration, inherent to both assemblages. 33 Kadowaki et al. 2016. 34 Nishiaki and Guliyev 2019. 35 Gatsov and Nedelcheva 2017. 36 Guiebeau et al. 2017. 37 Palumbi 2012. 38 For different views on the origins of round-bladed buildings and the connections between the architectural traditions of the ‘AratashenShulaveri-Shomutepe’ complex and the Halaf culture, see Areshyan 29 30

The episodic nature of the existence of rectangular buildings, recorded only at the formative stage of culture (Aknashen, horizon VII, Hasansu I) and later unknown in the AShSh area, suggests that they owe their appearance to an external impulse. It is for the aforementioned settlement of Hakemi Use that a rectangular plan of buildings is characteristic (only two buildings are found to have rounded outlines), erected in the same housebuilding technique as the buildings in the Ararat valley – lumps of mud and pisé without stone and Ghafadaryan 1996: 24-25; Baudouin et al. 2018; Javakhishvili 1973: 346-349; Munchaev and Merpert 1981: 193-196. 39 Tekin 2011.

25

Systemizing the Past foundations.40 At the settlement of Salat Cami Yani (6400-6200 BC) in the same Tigris Valley, rectangular pisé buildings are also attested.41

the Upper Tigris region, around the Bingöl and Nemrut Dağ volcanoes, which lies about 250 km southwest of the Araxes basin.

Of course, elementary morphological similarity is not enough to confirm contacts, let alone establish a genetic link between two geographically separated cultural areas. More valuable proof is the similarity in the evolution of the building techniques within these two areas, with the gradual disappearance of the cob technique to the benefit of the mud-brick in northern Mesopotamia at the outset of the sixth millennium BC matching with the appearance of the AShSh culture in the South Caucasus where both techniques are used concurrently at its beginning.42

Finally, judging by the aforementioned general parameters of the process of the emergence and development of ceramics in the South Caucasus in the first half of the 6th millennium BC,48 the locally started production followed the same development path that was documented in the first Near Eastern ceramic production centers. Both in individual examples of sites in the basins of the Tigris,49 Euphrates50 and the Khabur valley,51 and in the region as a whole, the same general trend in the development of ceramics technology is evident. According to Le Mière,52 three steps can be recognized in the process of gradual pottery development: in the first, the pottery is exclusively mineral-tempered; in the second, plant temper is introduced with early mineral-tempered pottery being still present; in the third, plant temper is used alone.

This similarity of architectural plans and trends in the development of construction techniques of the abovementioned sites of the Upper Tigris basin and the Ararat valley is accompanied by the frequency of bladelets in the obsidian industry of the Hakemi Use settlement;43 in Salat Cami Yani also ‘most of the obsidian artifacts consist of blades/bladelets;’ the presence of ‘some single-platform bullet cores’ is noted.44

Thus, for all the chronological difference, it is clear that there is a fundamental similarity, if not complete uniformity, in the process of the appearance and development of ceramics both in the settlements of the upper basins of the Tigris and Euphrates, and in the sites of the first half of the 6th millennium BC of the Ararat valley and the South Caucasus.

Clearly, in the field of obsidian industry between the regions under consideration, one can trace not only general similarities or particular analogies of a random nature. Technological innovation appears to be a more important aspect in this area. As noted above, the obsidian industry of Aknashen and AShShis generally characterized by a predominance of long standardized blades obtained through a combination of pressure techniques (lever and crutch), and indirect percussion. According to Chabot and Pelegrin,45 ‘pressure-related blade production can be usefully regarded as a marker of particular cultural traditions and of the diffusion of technical innovation.’

Conclusion More or less massive finds of identifiable artifacts and samples of obsidian raw materials are documentary evidence of the connection between the settlements of the South Caucasus and the outside world during the first half of the 6th millennium BC. Thanks to these findings, the direction of the most intense connections is determined. One of the directions — the western one — is documented by finds on all Neolithic sites of the Ararat valley (Aknashen, Aratashen;53 Masis Blur54), as

‘The production of large blades using a lever occurred as early as the second half of the eighth millennium cal BC at Çayönü Tepesi, likely between 7340 and 7080 cal BC. It was thus testified in the Balikh Valley a thousand years later, between 6100 and 6500 cal BC’.46 Although there is currently no direct evidence of the existence of this technology in the area under consideration at the very end of the 7th millennium BC, Chabot and Pelegrin47 believe that it is very possible that the origin of the South Caucasian obsidian pressure blade production techniques may well be found in the communities of

‘While pottery was very rare at the beginning, it began to increase in the second quarter of the 8th millennium cal BP: the initial mineral-tempered pottery soon gave way to plant-tempered pottery’ (Nishiaki et al. 2019a: 5). 49 Thus, in the above-mentioned settlement of Salat Cami Yani (64006200 BC) in the lower layer (Phase I) the density of pottery is evidently lower than in the upper two phases. The most predominant ware group is mineral-tempered. The above-laying layers (Phases 2-3) are dominated by vegetal-tempered coarse ware (Miyake 2011). 50 At Akarçay Tepe the total number of potsherds recovered from the earliest pottery-bearing layers (phase 1) are very low. Pottery first occurs in levels 11 and 10 and only from level 9 onward is a gradual increase in the number of sherds observed. Initially, in layers 11 to 7, the pottery was mineral-tempered exclusively. From layer 6 onwards plant-tempered sherds became dominant (Cruells 2017: 14). 51 The earliest pottery of Tell Seker al-Aheimar is exclusively mineraltempered. Plant temper will appear later. Gradually, plant-tempered ware became the most common ware, while mineral-tempered wares decreased and disappeared (Le Mière 2009). 52 Le Mière 2009: 75. 53 Chataigner and Gratuze 2014. 54 Marirosyan-Olshansky 2018a; 2018b. 48

Tekin 2011. Miyake 2011. 42 Baudouin et al. 2018. 43 Tekin 2011. 44 Miyake 2011. 45 Chabot and Pelegrin 2012. 46 Altınbilek-Algül et al. 2012. 47 Chabot and Pelegrin 2012. 40 41

26

Ruben Badalyan and Armine Harutyunyan: Neolithization of Armenia Badalyan, R. 2010. Obsidian of the Southern Caucasus: The use of raw materials in the Neolithic to Early Iron Ages. In S. Hansen, A. Hauptmann, I. Motzenbäcker and E. Pernicka (eds), Von Majkop bis Trialeti. Gewinnung und Verbreitung von Metallen und Obsidian in Kaukasien im 4. – 2. Jt. v. Chr.: 27–38. Bonn, Rudolf Habelt. Badalyan, R, Lombard, P., Avetisyan, P., Chataigner, C., Chabot, J., Vila, E., Hovsepyan, R, Willcox, G. and Pessin, H. 2007. New data on the late prehistory of the Southern Caucasus. The excavations at Aratashen (Armenia): preliminary report. In B. Lyonnet (ed.), Les Cultures du Caucase (VIe-IIIe millénaires avant notre ère). Leurs relations avec le Proche-Orient: 37–61. Paris, CNRS éditions, Editions Recherches sur les Civilisations. Badalyan, R., Harutyunyan, A., Chataigner, C., Le Mort, F., Chabot, J., Brochier, J., Balasescu, A., Radu, V. and Hovsepyan, R. 2010. The settlement of AknashenKhatunarkh, a Neolithic site in the Ararat plain (Armenia): Excavation results 2004–2009. Türkiye Bilimler Akademisi Arkeoloji Dergisi 13: 185–218. Badalyan R. and Harutyunyan A. 2014. Aknashen – the Late Neolithic settlement of the Ararat Valley: Main results and prospects for the research. In B, Gasparyan and M. Arimura (eds), Stone Age of Armenia: 161–176. Kanazawa, Center for Cultural Resource Studies, Kanazawa University. Batiuk, S., Jalabadze, M., Graham, A., Koridze, I. and Abu Jayyab, Kh., with contributions by Savulov C. 2017. The Gadachrili Gora Regional Archaeological Project: 2016 preliminary report. Anatolica 43: 173– 202. Batiuk, S. D., Jayyab, K. A., Jalabadze, M., Graham, A., Rhodes, S. and Losaberidze, L. 2019. The 2017–18 Gadachrili Gora Regional Archaeological Project (GRAPE): Second preliminary report. Anatolica 45: 43–76. Baudouin, E., Lyonnet, B. and Hamon, C. 2018. Architectural techniques and cultural relationships between the Caucasus and Mesopotamia at the beginning of the sixth millennium BC. In C. Douché and F. Pichon (eds), From the Caucasus to the Arabian Peninsula: Domestic spaces in the Neolithic: 49–84. Paris, Association Routes de l’Orient. Chabot, J. and Pelegrin, J. 2012. Two examples of pressure blade production with a lever: Recent research from the Southern Caucasus (Armenia) and Northern Mesopotamia (Syria, Iraq). In P. Desrosiers (ed.), The Emergence of Pressure Blade Making: From Origin to Modern Experimentation: 181–198. New York, Springer. Chataigner, C., Badalyan, R. and Arimura, M. 2014. The Neolithic of the Caucasus. Oxford Handbooks Online, Oxford University Press. DOI: 10.1093/ oxfordhb/9780199935413.013.13 Chataigner, C. and Gratuze, B. 2014. New data on the exploitation of obsidian in the Southern Caucasus

well as on a number of settlements in the Kura basin (Mentesh Tepe;55 Göytepe56) of a certain amount of obsidian from the sources of the Kars-Sarykamish area. However, the cultural context of the connection of the population of the South Caucasus with this area remains unclear due to the absence of any data on its Neolithic culture. Much more evidence in the form of finds of imported painted Samarra and/or Halaf ceramics and single samples of obsidian from deposits in the basin of Lake Van points to the southwest, to the region of the upper basins of the Tigris and Euphrates. Most of these findings were discovered in the settlements of the Ararat valley and the Araxes basin. Further to the north, the intensity of the connections determined by these materials, apparently, weakens and is recorded in a smaller volume at the sites of the Kura basin. At the 8,000-7,000 BC sites of the upper basins of the Tigris and Euphrates, those trends in the development of housing, obsidian industry, and ceramic production are observed that determined the nature of the AShSh culture of the South Caucasus in the first half of the 6th millennium BC. This observation is consonant with M. Özdogan, according to whom, ‘the process of neolitisation in the Caucasus must have occurred through the transfer of technologies and commodities’.57 Bibliography Altınbilek-Algül, C., Astruc, L., Binder, D. and Pelegrin, J. 2012. Pressure blade production with a lever in the Early and Late Neolithic of the Near East. In P. Desrosiers (ed.), The Emergence of Pressure Blade Making. From Origins to Modern Experimentation: 157– 179. New-York, Springer. Areshyan G. E. and Ghafadaryan, K. K. 1996. Architecture of Armenian Highland in the period of primitive society and the first state formations. In G. A. Tiratsyan (ed.), The History of Armenian Architecture, Vol. I: 14–86, Yerevan, ‘Gitutyun’ (in Armenian). Arimura, M., Petrosyan, A., Arakelyan, D., Nahapetyan, S. and Gasparyan, B. 2018. Preliminary report on the 2015 and 2017 field seasons at the Lernagog-1 site in Armenia. Aramazd: Armenian Journal of Near Eastern Studies 12/1: 1–18. Badalyan R. 2002. Obsidian of the Armenian Highland in the system of local and interregional contacts: New data. In A. Kalantaryan and S. Harutyunyan S. (eds), Culture of Ancient Armenia XII. Abstracts of the Reports of the Republican Scientific Session Dedicated to the 90th Anniversary of B. Arakelyan: 18–26. Yerevan, ‘Gitutyun’ (in Armenian). Lyonnet et al. 2012. Nishiaki et al. 2019b. 57 Özdogan 2018: 24. 55 56

27

Systemizing the Past (Armenia, Georgia) and eastern Turkey, Part 2: Obsidian procurement from the Upper Palaeolithic to the Late Bronze Age. Archaeometry 56/1: 48–69. Cruells W. 2017. Transitions and their chronologies in the Pottery Neolithic of the Near East. In W. Cruells, I. Mateiciucova and O. Nieuwenhuyse (eds), Painting Pots – Painting People. Late Neolithic Ceramics in Ancient Mesopotamia: 11–29. Oxford and Philadelphia, Oxbow Books. Gatsov, I. and Nedelcheva, P. 2017. Latest results on the obsidian assemblages from the Neolithic Period of Aruchlo I, Georgia. In B. Helwing, T. Aliyev., B. Lyonnet, F. Guliyev, S. Hansen and G. Mirtskhulava (eds), The Kura Projects. New Research on the Later Prehistory of the Southern Caucasus: 247–253. Berlin, Dietrich Reimer Verlag. Guilbeau, D., Astruc, L. and Samzun, A. 2017. Chipped stone industries from the Mil Plain (Kamiltepe) and the Middle Kura Valley (Mentesh Tepe), Azerbaijan. In B. Helwing, T. Aliyev., B. Lyonnet, F. Guliyev, S. Hansen and G. Mirtskhulava (eds), The Kura Projects. New Research on the Later Prehistory of the Southern Caucasus: 385–398. Berlin, Dietrich Reimer Verlag. Hamon, C., Jalabadze, M., Agapishvili, T., Baudouin, E., Koridze, I. and Messager, E. 2016. Gadachrili Gora: Architecture and organisation of a Neolithic settlement in the middle Kura Valley (6th millennium BC, Georgia). Quaternary International 395: 154–169. Harutyunyan, A. 2011. Some results of the research on the Neolithic-Chalcolithic pottery of the Ararat Valley. In G. Gambashidze (ed.), Archaeology, Ethnology and Folklore of the Caucasus: 51–55. Tbilisi, Gori and Batumi, Meridiani. Harutyunyan, A. 2014. On Neolithic pottery from the settlement of Aknashen in the Ararat Valley. In B. Gasparyan and M. Arimura (eds), Stone Age of Armenia: 191 –204. Kanazawa, Center for Cultural Resource Studies, Kanazawa University. Javakhishvili, A. I. 1973. Building and Architecture of the Settlements of the South Caucasus in the V – III Millennia BC). Tbilisi, Metsniereba. Kadowaki, S., Guliyev, F. and Nishiaki, Y. 2016. Chipped stone technology of the earliest agricultural village in the southern Caucasus: Hacı Elamxanlı Tepe. In O. Kaelin and H.-P. Mathys (eds), Proceedings of the 9th ICAANE, vol. 3: 709–722. Wiesbaden, Harrassowitz. Kiguradze, T. 1976. The Periodization of the Early Farming Culture of the Western Transcaucasia. Tbilisi, Metsniereba. Kohl, P. L. 2007. The Making of Bronze Age Eurasia. Cambridge, Cambridge University Press. Korfmann, M. 1982. Tilkitepe (Istanbuler Mitteilungen 26). Tübingen, Verlag Ernst Wasmuth. Le Mière, M. 2009. Early Neolithic pottery from the Near East: the question of temper and its implications. In L. Astruc, A. Gaulon and L. Salanova (eds), Méthodes

d’approche des premières productions céramiques: étude de cas dans les Balkans et au Levant. Table-ronde de la Maison de l’Archéologie et de l’Ethnologie (Nanterre, France) 28 février 2006: 73–80. Rahden/Westfalia, Leidorf. Lyonnet, B., Guliyev, F., Helwing, F., Aliyev, T., Hansen, H., and Mirtskhulava, G., 2012. Ancient Kura 2010– 2011: The first two seasons of joint field work in the southern Caucasus. Archäologische Mitteilungen aus Iran und Turan 44: 19–90. Marro, C., Bakhshaliyev, V., Berthon, R. and Thomalsky, J. 2019. New light on the Late Prehistory of the South Caucasus: Data from the recent excavation campaigns at Kültepe I in Nakhchivan, Azerbaijan (2012-2018). Paléorient 45/1: 81–113. Martirosyan-Olshansky, K. 2018a. Sourcing of obsidian artifacts from Masis Blur, Armenia. Aramazd: Armenian Journal of Near Eastern Studies 12/1: 19–34. Martirosyan-Olshansky, K. 2018b. Obsidian Economy in the Armenian Highlands During the Late Neolithic. A View from Masis Blur. UCLA. ProQuest ID: Olshansky_ucla_0031D_16582. Merritt ID: ark:/13030/m5fz26qq. Retrieved from https:// escholarship.org/uc/item/65b8v78q. Miyake Y. 2011. Salat Cami Yani. A Potery Neolithic site in the Tigris Valley. In M. Ӧzdogan, N. Başgelen and P. Kuniholm. The Neolithic in Turkey. New Excavations and New Research. Volume 1. The Tigris Basin: 129–149. Istanbul, Archaeology and Art Publications. Munchaev, R. M. 1975. Caucasus at the Dawn of the Bronze Age. Neolithic, Chalcolithic, Early Bronze. Moscow, Nauka. Munchaev, P. and Merpert, N. 1981. The Early Farming Settlements of Northern Mesopotamia. Moscow, Nauka. Museibli, N. A. 2017. The Neolithic period Hasansu settlement. In N. A. Museibli (ed.), Problems оf the Archaeology оf the Caucasus and Near East. NeolithicLate Bronze Age: 42–58. Baku, Afpoligraf. Narimanov, I. G. 1987. The Culture of the First FarmersBreeders in Azerbaijan – Chalcolithic Period VI-IV Millennia BC. Baku, Elm. Nishiaki, Y., Guliyev, F. and Kadowaki, S. 2015a. Chronological contexts of the earliest Pottery Neolithic in the South Caucasus: Radiocarbon dates for Göytepe and Hacı Elamxanlı Tepe, Azerbaijan. American Journal of Archaeology 119/3: 279–294. Nishiaki, Y., Guliyev, F., Kadowaki, S., Alakbarov, V., Miki, T., Salimbayov, Sh., Akashi, C. and Arai, S. 2015b. Investigating cultural and socioeconomic change at the beginning of the Pottery Neolithic in the Southern Caucasus: The 2013 Excavations at Hacı Elamxanlı Tepe, Azerbaijan. Bulletin of the American Schools of Oriental Research 374: 1–28. Nishiaki, Y. and Guliyev, F. 2019. Neolithic lithic industries of the southern Caucasus: Göytepe and Haci Elamxanli Tepe, west Azerbaijan (early 6th millennium cal. BC). In: L. Astruc, C. McCartney, F. 28

Ruben Badalyan and Armine Harutyunyan: Neolithization of Armenia Briois and V. Kassianidou (eds), Near Eastern Lithic Technologies on the Move. Interactions and Contexts in the Neolithic Traditions. 8th International Conference on PPN Chipped and Ground Stone Industries of the Near East, Nicosia, November 23rd – 27th 2016: 471–483. Nicosia, Astrom Edition. Nishiaki, Y., Zeynalov, A., Mansrov, M., Akashi, C., Arai, S., Shimogama, K. and Guliyev, F. 2019a. The Mesolithic-Neolithic interface in the Southern Caucasus: 2016–2017 excavations at Damjili Cave, West Azerbaijan. Archaeological Research in Asia 19: 100140. Nishiaki, Y., Maeda, O., Kannari, T., Nagai, M., Healey, E., Guliyev, F. and Campbell S. 2019b. Obsidian provenance analyses at Göytepe, Azerbaijan: implications for understanding Neolithic socioeconomies in the Southern Caucasus. Archaeometry 61/4: 765–782. Özdoğan, M. 2018. Some considerations on the emergence of Neolithic way of life in the Caucasus.

In A. Batmaz, G. Bedianasvhili, A. Michalewicz and A. Robinson (eds), Context and Connection – Essays on the Archaeology of the Ancient Near East in Honour of Antonio Sagona: 15–28. Leuven, Peeters. Palumbi, G. 2007. A Preliminary analysis on the Prehistoric Pottery from Aratashen (Armenia). In B. Lyonnet (ed.), Les Cultures du Caucase (VIeIIIe millénaires avant notre ère). Les relations avec le Proche-Orient: 63–76. Paris, CNRS Editions, Editions Recherches sur les Civilisations. Palumbi, G. 2012. The Chalcolithic of Eastern Anatolia. In G. McMahon and Sh. Steadman (eds), The Oxford Handbook of Ancient Anatolia (10,000-323 BCE). 10.1093/ oxfordhb/9780195376142.013.0009. Tekin, H. 2011. Hakemi Use. A newly discovered late Neolithic site in Southeastern Anatolia. In M. Ӧzdogan, N. Başgelen and P. Kuniholm (eds), The Neolithic in Turkey. New Excavations and New Research. Volume 1. The Tigris Basin: 151-172. Istanbul, Archaeology and Art Publications.

29

Groups of Three Deities in Middle and Neo-Assyrian Times Felix Blocher

Martin-Luther-Universität Halle-Wittenberg Abstract: Groups of three divine symbols and deities in Middle and Neo-Assyrian periods are investigated in two fields: art and royal inscriptions. Divine symbols as celestial bodies show combinations of three consisting of sun disk or winged sun disk, moon crescent, star or star disk, and Pleiades. This group also occurs on royal steles and slabs from private houses. Another group consists of three horned crowns, identifiable with Anu, Enlil, and Assur confirmed by royal inscriptions. The third subject from art is the winged sun disk showing three heads or busts above it in a triad-like way. Yet it still remains difficult to offer a solution for this conspicuous iconographic scheme. In the last part there is given a superficial overview over groups of three deities in the corpus of Middle and Neo-Assyrian royal inscriptions. The frequent group in art of three celestial bodies is not mirrored in the inscriptions, the Pleiades (Sebettu) lacking completely. Thus groupings of deities in art or inscriptions obviously follow different reasonings. Keywords: Assyria, Middle Assyrian period, Neo-Assyrian period, divine symbols, deities, sun, moon, star, Pleiades, winged sun, royal inscriptions

Introduction1 Groups of three deities, or triads with a more specific and significant expression, play a role in many religions worldwide.2 Concerning the Southern Caucasus we may recall the Urartian triad described by G. A. Melikishvili in 1953. Yet our focus in this contribution is Assyria during the Middle and Neo-Assyrian periods. First of all there is a strong tendency in grouping smaller and larger amounts of divine symbols, more rarely also anthropomorphic deities. Among the symbolic representations celestial phenomena are important, namely the following four: the multi-pointed star representing goddess Ištar or the planet Venus,3 the moon, in most cases as crescent, the sun in a variety of round disks or as winged disk, and finally the Pleiades.4 For all these four already in Middle Assyrian times (14th to 10th centuries BC) exist good examples. Middle Assyrian Period a) Celestial Phenomena All four phenomena mentioned above are grouped together on a cylinder seal (Figure 1).5 In most of the This article is devoted to Pavel Avetisyan in honouring his cooperation during the stay in Armenia in autumn 2011 and his great achievements in archaeology in Armenia. He was very helpful when Michael Herles and I were travelling through Armenia looking for a rewarding site. My thanks go to Yervand Grekyan for his support and patience during the production phase. Thanks go also to Michael Herles for many useful advices. 2 For a detailed superb overview and discussion of the scheme from ancient Egypt to Christianity see Griffiths 1996. Cf. also Balkan 1992; Glassner 1999; Jahangirfar 2018; Khvedilidze 1982; Parpola 2000: 202205. For other concepts of divinity in Assyria see e.g. Porter 2000. 3 I use the word ‘star’ in the descriptions despite the fact that Venus is a planet (cf. Hunger 2003-2005). 4 Cf. Blocher 2009; 2010. 5 Lambert 1979: no. 64. 1

Systemizing the Past (Archaeopress 2023): 30–41

Figure 1. Middle Assyrian cylinder seal showing star, crescent, sun disk, and Pleiades (Lambert 1979: Plate VII no. 64).

cases only three are shown, moon (crescent) and star (planet) forming a basic group, and the Pleiades or the sun may join the two (Figures 2-3).6 Neo-Assyrian Period a) Celestial Phenomena During the Neo-Assyrian period (10th to 7th centuries BC) this concept is continued. Even here glyptic examples exist that show all four phenomena together, e.g. impressions of a cylinder seal on clay tablets from Assur dating in the 8th century (Figure 4):7 Above a ‘mistress of animals’ crescent, star in peculiarly detailed way, Pleiades and winged disk are visible. The group of three, crescent – star – Pleiades, is very 6 7

Buchanan 1966: no. 568; Kantor 1958: Plate 72, no. XXII. Klengel-Brandt 2014: no. 205.

Felix Blocher: Groups of Three Deities in Middle and Neo-Assyrian Times

Figure 2. Middle Assyrian seal impression from Tell Fekheriye showing Pleiades, crescent, and star (Kantor 1958: Plate 72 no. XXII).

Figure 5. Neo-Assyrian cylinder seal showing Pleiades, star, and crescent (Delaporte 1923: Plate 88, 6, A. 680).

Figure 3. Middle Assyrian seal cylinder showing star, crescent, and sun disk (Buchanan 1966: Plate 38 no. 568). Figure 6. Neo-Assyrian cylinder seal showing crescent, Pleiades, and star (Collon 1987: 184 no. 880).

frequent above adoration or hunting scenes (Figures 5-6).8 The sun and the symbol of the sun god are now in most cases represented as the winged disk. Two cylinder seals show table scenes with winged disks and star and Pleiades or star and crescent (Figures 7-8).9 A hunting deity too may operate under the protection of star, winged disk, and crescent (Figure 9).10 We may draw different conclusions from these examples, firstly that the seal owner makes the choice of the celestial phenomena to be shown, secondly that

Figure 4. Neo-Assyrian seal impression from Assur showing crescent, star, Pleiades, and winged sun (Klengel-Brandt 2014: Taf. 42 no. 205).

Collon 1987: no. 880; Delaporte 1923: A. 680. Porada 1948: nos. 699-700. 10 Watanabe 1992: Taf. 71d. 8 9

31

Systemizing the Past

Figure 7. Neo-Assyrian cylinder seal showing winged sun disk, crescent, and star (Porada 1948: Plate CIII no. 699).

Figure 10. Upper part of a stele of Adad-nārārī III from Tell al-Rimah showing Pleiades, star, and crescent behind the king’s head and the winged sun disk in front of it (Orthmann 1975: Figure 212).

between the divine symbols represented and the deities mentioned in the inscription.11 On a famous stela of king Adad-nirari III. (810-783 BC) from Tell al-Rimah (Irak) the head of the king is dividing the upper part of the stela in two fields (Figure 10).12 In the left part (behind the king’s face) we see star, crescent and Pleiades, in the right part (in front of the king’s face) is the winged disk together with four other symbols.

Figure 8. Neo-Assyrian cylinder seal showing winged sun disk, Pleiades, and star (Porada 1948: Plate CIV no. 700).

We may assume that the decision of how to divide the upper part of the stele is not accidental. On the rock relief of Šikaft-i Gulgul near Ilam (Iran) (possibly early 6th century BC) we see behind the king’s face the same group as before – crescent, star, and Pleiades – whereas the winged disk and the horned crown are shown in front of the king (Figure 11).13 Again different is the situation on a relief from a private house in Assur probably from Sennacherib’s time (704681 BC) representing a god on a mixed being (Figure 12).14 Here the winged disk is situated behind the high headgear of the deity while the crescent, the star and the Pleiades are shown in front of it. Whether the piece is fragmentary and an adorant is to be restored remains an open question.

Figure 9. Neo-Assyrian cylinder seal showing star, winged sun disk, and crescent (Orthmann 1975: Figure 273 a).

the scene on the seal happens under various celestial, astronomical conditions. Divine symbols are also a prominent feature of the decoration on royal stelae. As on the seals they normally appear in the upper part of the scene, in this case close to the head of the king. Thus different sizes of groups of symbols result sometimes in consisting of three symbols. In some important cases we have correlations

Reade 1977: 39-41. Börker-Klähn 1982: no. 164. 13 Börker-Klähn 1982: no. 223. Despite the fact that most parts of the star disc are chipped off it is as such recognisable without doubts by the curved contour and the rest of a single point. Discussions about the dating are resumed by Börker-Klähn 1982: 215, no. 223. 14 Börker-Klähn 1982: no. 243. 11 12

32

Felix Blocher: Groups of Three Deities in Middle and Neo-Assyrian Times

Figure 11. Upper part of the Neo-Assyrian rock relief at Šikaft-i Gulgul showing Pleiades, star, and crescent inscribed in the full moon behind the king’s head and the winged sun disk in front of it (Grayson/Levine 1975: Plate X).

Figure 13. Relief from Assur showing an armed god and an adorant, above them winged sun disk, crescent, and star disk (Andrae 1938: Taf. 21c).

in court dress in front of him, standing on a low base. Above the scene the already mentioned sun, moon and star are hovering. Walter Andrae addressed the god as Ninurta.17 Let us add here a stele fragment of Adad-nirari III (810783 BC) from Tall Šayḫ Ḥamad/Dur-Katlimmu18 with yet another grouping of symbols in front of the king’s head: winged disk, star and lightning (Figure 14).19 An extraordinary group of three symbols is shown on a relief slab from a palace of Tiglath-pilesar III (744-727

Figure 12. Relief from Assur showing an armed god on a mixed being, behind him the winged sun-disk, in front of him crescent, Pleiades, and star (Andrae 1977: 79 Figure 57).

(Photo British Museum, restored state). The original photo suggests that the right upper corner has been preserved. J. Börker-Klähn’s statement that on stelae of courtiers not even sun, moon, and star were represented (1982: 223 under no. 242) is to be rejected with the example of Bel-Harran-beli-usur’s stela, (ibidem no. 232). The relief in Figure 13 may thus well be the work ordered by a courtier and displayed at Assur, also the base on which the courtier is standing being no problem. Following Peter Calmeyer (1973: 147, 149) a human figure standing on a base or plinth does not mean the representation of a statue but simply the use of this object by an adorant. 17 Andrae 1938: 50, note 1, 220 ad Taf. 21c. 18 Börker-Klähn 1982: no. 165. 19 Karen Radner was able to combine this part of the stela with another, much larger one in private ownership showing the king’s body and legs (Radner 2012, courtesy M. Herles), with a new edition of the inscription. Unfortunately, the part of the stele behind the king’s head is still missing.

One of the rare examples in the art of Assyria showing only a group of three celestial bodies – winged disk, crescent and star disk – is a small relief slab from Assur again (Figure 13).15 Also in this case we have to ask whether further parts are to be restored on the right side.16 The image shows a warrior god with an adorant Börker-Klähn 1982: no. 242. The upper right corner looks different on Andrae’s illustration compared with the image at Börker-Klähn 1982: Plates no. 242

15 16

33

Systemizing the Past

Figure 15. Relief from a palace of Tiglath-pileser III in Nimrūd/Kalḫu showing in front of the king crescent inscribed in the full moon, disk, and cross (Barnett/Falkner 1962: Plate VIII).

star, the planet Venus.22 There are many Assyrian examples showing the multi-pointed star against a disk as background (e.g., Figures 13, 14, 17, 18). The disk on the Tiglath-pilesar III slab shows either two further smaller disks or an imposed ring comparable to that on the cross. It is obvious that the cross is bigger than the two other celestial bodies. Is it also more important? The reason why there is no winged disk representing the sun or the sun-god may be that the campaign is in Babylonia where the sun is practically never represented by the winged disk.23 Let us now leave the consideration of groups of three celestial phenomena of bodies in the art of Neo-Assyrian times. In many cases we deal with ‘accidental triads’ since they often belong to larger groups of symbols. But the compositional grouping of symbols on tops of stelae behind or in front of the king’s head may well be intentional and may give hints on their importance in these particular cases.

Figure 14. Part of a stele of Adad-nārārī III from Tall Šayḫ Ḥamad showing winged sun disk, star disk, and lightning bundle (Börker-Klähn 1982: no. 165).

BC) in Nimrud/Kalḫu (Figure 15).20 The king is sitting on a throne on a small hillock, in front of him the turtan and other courtiers, behind the king an umbrella carrier. Above the scene crescent, a disk and a cross.

b) Three Horned Crowns

From other contexts we know that the cross is a symbol of the sun respectively of the sun-god;21 the crescent is self-explaining, so a problem is only the disk. It may stand for the star proper, morning star and evening

Contrary to the frequent groups of three celestial bodies in Assyria the combination of three horned crowns on seats or thrones (German Symbolsockel) is rare. The horned crown in ancient Near East is the ultimative hint in identifying deities. Yet the horned cap or crown in a symbolic context does not represent deities in their overallness but only the highest ones in the pantheon.

Barnett and Falkner 1962: Plate VIII. This slab was only known through a drawing by Austen H. Layard (Barnett and Falkner 1962: Plate VIII), but Polish archaeologists were able to re-find this and other slabs in their 1970s excavations (Mierzejewski and Sobolewski 1980: 157, Abb. 7). 21 Very important Calmeyer 1984; for an overview Herles 2006: 257259. Note that also the god Nabû has been connected with the cross (Pongratz-Leisten 1994: 99 concerning a statement of Simo Parpola [not accessible to me, but fully quoted by Pongratz-Leisten]). Beate Pongratz-Leisten in this context is also quoting Suzanne Herbordt (1992: 139 with note 55) who is clearly more cautious than Simo Parpola concerning the relation of Nabû and the cross. 20

Barnett and Falkner 1962: xvii call the symbols from left to right as ‘Sin, Šamaš and perhaps Ištar’ (ibidem 10 under Slab 6a). Neither Layard’s drawing nor the photo of the re-found slab show any structures or details on the disk. 23 Ehrenberg 2002: 72, is interpreting two of the symbols completely different: The cross is a symbol of Marduk and the disk stands for the sun(-god) (courtesy M. Herles). I keep following the opinion of Calmeyer 1984, because of the crosses integrated in sun representations, that the cross is a sun-symbol. 22

34

Felix Blocher: Groups of Three Deities in Middle and Neo-Assyrian Times

Figure 16. The group of divine symbols occurring uniformly (together with an image of king Sennacherib) in eleven niches in the garden valley of Khinnis/Bavian (Bachmann 1927: 21 Figure 15).

This is demonstrated by a few by-scripts on kudurrus of the Kassite era.24 In Assyrian art the horned crown means the god Aššur,25 the group of three horned caps consequently Anu – Enlil – Aššur (the sequence is of course arbitrary as long as an inscription on the monument does not give hints on it). Assyrian examples date from the reign of king Sennacherib (704-681 BC).

Figure 17a-b. King Sennacherib in front of divine symbols on two stelae from the town area of Nineveh. Note in the drawings the wrong stylus-like interpretation of the ramstaff beside the three horned crowns (Börker-Klähn 1982: no. 203-4).

Several identical rock reliefs close to the garden valley of Khinnis-Bavian show three horned crowns on seats side by side in the group of symbols (Figure 16).26 Immediately on the left side of this group is the head of the king. He is directed to the right in all rock niches. Twelve symbols are combined, the inscription mentions eleven of them, one name being illegible.27 Thus the three horned crowns are from left to right the representative objects for Aššur, Enlil, and Anu.28 Closely connected with Khinnis-Bavian are the representations on two steles from Nineveh’s town area (Figures 17-18).29 On the one stele (in Istanbul) the king is directed to the right side, on the other (in London) to the left, and the symbols are accordingly placed, i.e. the three horned crowns figure directly in front of Herles 2006: 212-213 for Anu; 219-220 for Enlil. Herles 2006: 213-214; Reade 1977: 38. 26 Bachmann 1927: 21, Abb. 15; Börker-Klähn 1982: nos. 189-199. 27 Bachmann 1927: 39, following Carl Frank. The lacuna concerns the double lion mace resp. the double mixed-being mace. Luckenbill 1927: § 331 restores for the lacking deity the name of Nusku but cf. Reade 1977: 39, note 9, who votes for Nergal. All other symbols are evident und they correspond to the current explanations. This is important for the still continuing discussion on Aššur and Šamaš (s. below) since here without any doubts the horned crown stands for Assur and the winged disk for Šamaš. 28 Luckenbill 1927: § 331. 29 Börker-Klähn 1982: nos. 203-204. 24 25

Figure 18. Upper part of the Sennacherib stele from Nineveh now in London. The ram-staff left of the three horned crowns is clearly discernible (Smith/Gadd 1938: Plate 34).

35

Systemizing the Past the king’s face. Unfortunately the inscription does not mention the deities represented by the symbols.30 This triad of horned crowns is in all these cases directly accompanied by a fourth symbol on a seat, the so-called ram-staff.31 This object is connected with the water-god Enki/Ea, and the inscription at Khinnis-Bavian is in accordance with the sequence of deities mentioning Ea at the fourth position.32 This position might be a memory of the fact that in the rare case of three horned crowns on Babylonian kudurrus the third one means Enki/Ea.33 On the other side the ram-staff on the above-mentioned Assyrian monuments shows clearly that none of the horned crowns is representing Enki/Ea, a fact which is by the way happily confirmed by the Khinnis-Bavian inscription.

Figure 19. The so-called Sennacherib seal from Nineveh showing the sun god in the winged sun disk and two further figures in or behind the wings (Collon 1987: 174 no. 812).

c) Winged Disk with Three Figures, Busts or Heads We know a group of representations of the winged sun disk that show beside the central bust or head in the ring or in front of the disk two additional busts or heads on the stretched wings.34 Is this a sun triad? Before discussing former interpretations we may have a look at some of the objects. The so-called Sennacherib seal from Nineveh displays an adorant in front of a life tree behind which a stele with a royal representation is visible (Figure 19).35 Above the life tree a winged sun disk with a semi-figure directed to the right is shown coming from the central ring or disk. Above the wings we see two smaller figures each directed to the central one.

Figure 20. Neo-Assyrian cylinder seal showing three heads above the winged sun disk (Collon 2001: Plate XXXIX no. 211).

On another cylinder seal the winged sun disk is supported by two scorpion-men (Figure 20).36 Above the sun symbol we see two smaller and one bigger head to which belongs also a left arm. The bigger head wears a horned crown with a spherical end. The heads of the two smaller figures also end spherically and are thus divine headgears too. The scene is framed by an

adorant and a deity or genius. A second main motif is a lord of animals. A seal impression from Assur shows the winged sun disk with three heads above a life tree, which is framed by two male figures and a crouching sphinx. A crescent is visible too (Figure 21).37

30 Luckenbill 1927: §§ 473-476; RINAP 3/1: no. 38 (speaking of three stelae). 31 The drawings at Börker-Klähn 1982: nos. 203-204 in both cases show small staffs with a pointed end, and accordingly Börker-Klähn 1982: 209, no. 203 calls the object a stylus. But styli look different and appear in most cases in pairs. A closer look at the photographs (quotations at Börker-Klähn 1982: sub nos. 203-204) reveals that the objects are doubtless ram-staffs with their bended heads reaching into the frame of the stela. So it is obvious that Enki/Ea’s symbol is placed directly beside the three horned crowns. 32 Cf. Herles 2006: 218-219. 33 Woods 2004: 65, Anm. 211 identifies the three horned crowns on a Kudurru from the reign of Nebuchadnezzar I as the symbols of Anu, Enlil, and Sîn (instead of Ea) but without giving an explanation for this idea. 34 Collon 2001: 80-81. 35 Collon 2001: no. 173. 36 Collon 2001: no. 211.

The motif is also part of the design of stamp seals despite a reduced space for images. Stamp seal impressions on a clay tablet from Assur show three heads over a winged sun disk supported by two bull-men (Figure 22).38 Klengel-Brandt 2014: no. 195. Hrouda 1991: 101 Abb. 8. Compare the stamp impression from Assur Klengel-Brandt 2014: no. 82, and the cylinder seal with the same motif that was impressed in the way of a stamp seal, ibidem no. Ist 76. The central figure shows a star on her headgear. 37 38

36

Felix Blocher: Groups of Three Deities in Middle and Neo-Assyrian Times

Figure 21. Neo-Assyrian seal impression on a clay tablet from Assur showing the winged sun-disk with three heads (Klengel-Brandt 2014: Taf. 40 no. 195).

Figure 22. Neo-Assyrian seal impressions on a clay tablet from Assur showing the winged sun disk with three schematic heads (Hrouda 1991: 101 Figure 8).

A Lamaštu amulet in the Mosul museum features in the upper row five different divine symbols (Figure 23).39 At the left probably a lightning fork/bundle, next a star disk or radiant wreath with a figure inside directed to the right.40 In the middle position we see the winged sun disk with a figure in the central ring and ending in a bird tail. On or in front of the wings of the sun disk we see two further figures. The fourth symbol is the horned crown. At the right end of the upper row there is a moon representation with the crescent inscribed in the full moon. In the crescent is standing a halffigure directed to the right. Thus we have the very rare example of the three main celestial bodies sun, moon, and star, all with their ‘manning’. To come back to the winged disk with figures, we may assume that the central figure stands for the sun god, but who are the smaller figures on the wings? P. Calmeyer gave an overview on earlier interpretations and an own suggestion concerning the smaller figures: They should be defined as the servant deities Kīttu and Mīšāru, ‘Truth’ and ‘Justice’, who are walking at the side of the king while the whole group means the king in his role as ‘sun of justice’.41 S. Parpola, while identifying

Figure 23. Neo-Assyrian Lamaštu amulet showing among others the three divine symbols star, sun, and moon, each equipped with a small figure (Amiet 1977: Figure 574).

the winged sun disk with Aššur, sees in this group Enlil/ Marduk (in the center), Ninurta/Nabû (the figure whom the central god looks at), and Mullissu/Ištar (the third figure in the back of the central person).42 The volutes on top of the disk sometimes visible may count as the two figures on the wings. Thus the whole concept is not only a triad but a real ‘trinity-in-unity’ and as such a component of the Assyrian concept of god (Aššur).43

Amiet 1977: Figure 574. For Ištar in the radiant wreath cf. Podella 1994: 31-34, 126-132. 41 Calmeyer 1984: 146, note 73. Cf. also Herbordt 1992: 98-100; broadly Podella 1994: 26-31, 132-154. Ursula Moortgat-Correns kept identifying the figure in the winged sun disk with Aššur and the figures on the wings with Sîn and Ištar (1996: 159 with note 13). She argues that the smaller figures are characterised with star and crescent but this is not to be verified on the image she cites (Frankfort 1939: Plate 33e, here Figure 20). Actually there is a small star in the upper right of the seal but it has nothing common with the figure on the wing, and a crescent is lacking completely. She does not discuss the ideas promoted by Calmeyer 1984. Also Parpola 2000: 170-172 with Figure 1 identifies the god in the winged sun disk with Aššur. The article ‘Sonnengott’ in the Reallexikon presents a more recent state of knowledge (Krebernik 2009-2011, Kurmangaliev 2009-2011). For further figures who could be combined with the men on the wings cf. Krebernik, ibidem: 601 § 3. Kurmangaliev describes the two small figures as ‚seine zwei Gehilfen, vielleicht …kīttu und mīšāru‘ (ibidem: 619 § 3; see also Krebernik 2006-2008: 354 § 2). 39 40

I would like to propose another solution for this group: Burchard Brentjes (1994) suggested that the Neo-Assyrian scene with the life tree and two kings means father and grandfather of the acting king who was sitting in front of the life tree. Above the tree the winged sun disk with the sun god is shown. Accordingly 42 43

37

Parpola 2000: 203-204 with Figure 6. Parpola 2000: 204.

Systemizing the Past I would like to suggest that the two small figures could represent the two royal ancestors in a changing status.

In Tukultī-Ninurta’s I (1233-1197 BC) inscriptions are mentioned twice Aššur, Enlil and Šamaš (RIMA 1, A.0.78.5 lines 48-9; 23 lines 56-7, 124, 135-6) as well as once Aššur, Adad and Ištar (RIMA 1, A.0.78.5 lines 111120).

The group of celestial symbols on the Mosul Lamaštu amulet is important in combining all of them with ‘manning’ whereas normally on other Lamaštu amulets only the basic celestial bodies are shown.44 At the same moment this is good evidence that the two small figures on the winged sun disk are subordinate to the central person and do not represent high deities like Ištar or Sîn since these are present on the same monument.

The same group of Aššur, Enlil and Šamaš figures in an inscription of Aššur-nādin-apli (1196-1194 BC) (RIMA 1, A.0.79.1 lines 5-6). With Aššur-rēša-iši I (1132-1115 BC) Anu, Enlil and Ea are mentioned twice (RIMA 1, A.0.86.1 line 2; 2 line 1).

The triadic character of the figures in the winged sun disk is most important for the phenomen of triads or even trinities in the history of religions, especially when we compare it with later evidence in Christianity.45 Nevertheless the character of the Neo-Assyrian group of three figures in the winged disk still needs further research work.

In Tiglath-pileser‘s I (1114-1076 BC) inscriptions Aššur, Anu and Enlil (RIMA 2, A.0.87.11 line 9’) as well as Aššur, Šamaš and Adad (RIMA 2, A.0.87.15 lines 1-2) are mentioned. Aššur-bēl-kala (1073-1056 BC) speaks of Aššur, Anu and Adad (RIMA 2, A.0.89.6 line 6’).

Groups of Three Deities in Royal Inscriptions from the Middle and Neo-Assyrian Periods

b) Neo-Assyrian Period In the reign of Ashurnasirpal II (883-859 BC) groups of Aššur, Šamaš and Adad (RIMA 2, A.0.101.1 line 104) and Anu, Enlil and Ea (RIMA 2, A.0.101.40 line 10) play a role.

Among the many facets of written tradition in the ancient Near East the royal inscriptions are an attractive group of comparison, although a couple of objects investigated above do not stem from a court context, e.g. seals.46 In each context of an inscription we would have to ask for the purpose of it, for its eventual connection with a building or a votive object in order to understand the grouping of mentioned deities etc. Compare this to the groupings on royal stelae. The purpose of the following compilation is to look whether there were groups of three deities or not.

Once with Shalmaneser III (858-824 BC) Aššur, Adad and the Assyrian Ištar are mentioned together (RIMA 3, A.0.102.43 line 10). In an inscription of Šamšī-Adad V (823-811 BC) they are Aššur, Šamaš and Adad (RIMA 3, A.0.103.1 III lines 64-5). From Tiglath-pileser‘s III reign (744-727 BC) groups of Bēl (Marduk), Nabû and Nergal (RINAP 1, no. 24 lines 6-7) and Aššur, Šamaš and Marduk (RINAP 1, no. 47 Obv. line 3) are transmitted.

a) Middle Assyrian Period Starting with Aššur-uballit I (1353-1318 BC), Aššur, Adad and Bēl-šarri (RIMA 1, A.0.73.1 line 27) as well as Aššur, Adad and Ištar-Kudnittu (RIMA 1, A.0.73.4 lines 11-2) are mentioned.

Under king Sargon II (721-705 BC) things are slightly changing. There are 17 mentionings of the group of Aššur, Nabû and Marduk, always in this sequence (Fuchs 1994: 47247). Other groups, but mentioned only once, are Anu, Enlil and Ea-Ninšiku (ibid.) (cf. above p. 36) concerning the three horned crowns), Aššur, Ningal and Adad (ibid.), Aššur, Šamaš and Adad (ibid.), Aššur, Šamaš and Marduk (ibid.) as well as Sîn, Šamaš and Adad (Fuchs 1994: 474).

Adad-nārārī I (1295-1264 BC) counts in three cases Aššur, Ištar and Adad (RIMA 1, A.0.76.4 lines 48-51; 14 lines 29-36; 15 lines 39-45), but once Aššur, Anu and Adad (RIMA 1, A.0.76.17 line 20). Shalmaneser I (1263-1234 BC) speaks of Enlil, Aššur and Ištar (RIMA 1, A.0.77.1 lines 1-2) and of Aššur, Ištar and Adad (RIMA 1, A.0.77.6 lines 24-28).

King Esarhaddon (680-669 BC) speaks in four cases of Šamaš, Adad, and Marduk as a group (RINAP 4, no. 104 col. III lines 9-10; 105 col. III lines 40-1; 114 col. III lines 16-18; 116 rev. line 9). In two cases it is Marduk, Sarpanītu and Nabû (RINAP 4, no. 105 col. V lines 24-

E.g. Heeßel 2002: 199 no. 20, 209 no. 30. Cf. Courth 1988: 9-13 for the Old Testament. Parpola 2000: 171-172 with Figure 1, 204-205. 46 This is a short overview over royal inscriptions using the RIMA and RINAP volumes without any claim of completeness. I was not able to check other literary, magical, theological etc. texts from these two periods. Cf. the similar list given by Parpola 2000: 168-169 note 7 in another context, concerning Aššur. 44 45

Since Andreas Fuchs (1994) mentions the groups of deities in his index I simply quote the pages of his catalogue where the references to Sargon’s inscriptions are registered.

47

38

Felix Blocher: Groups of Three Deities in Middle and Neo-Assyrian Times 5, 36-7), and in one case Nabû, Tašmētu and Nanaya (RINAP 4, no. 113 line 4).

stelae or slabs the fact that the head of the king or an anthropomorphic deity is dividing the upper field of the monuments in two parts may produce intentional groups of three divine symbols.

Under Ashurbanipal (668-631 BC) we have a rich documentation mainly in the different prism inscriptions. We can note the grouping of Aššur, Bēl (Marduk), and Nabû (RINAP 5/1, Ashurbanipal 3 col. III line 10, col. VI line 40; 4 col. I line 58, col VI line 41; 6 col. II line 64’; 7 col. II line 35’; 8 col. IX line 2’’’; 11 col. I line 81; 12 col. V line 3, col. VI line 8’ and line 20’; 17 col. I’ line 2’). In one case it is Aššur, Bēl (Marduk), and Sîn (RINAP 5/1, Ashurbanipal 4 col. III line 4).

The three horned crowns are expectedly identified with Anu, Enlil, and Aššur as is demonstrated by monuments and inscriptions from Sennacherib’s times. An important but still enigmatic group or triad proper is the winged sun disk with three human figures. The suggestion is made here that they represent sun god, father and grandfather of the king.

Another combination is Aššur, Sîn, and Šamaš (RINAP 5/1, Ashurbanipal 3, col. III line 31; 4 col. III line 20; 5 col. V line 5; 6 col. IV line 10’’).

It is difficult to see a systematic preponderance of certain three-fold combinations of deities in the royal inscriptions. In almost every case Aššur is part of a group of three deities. In comparison with art astonishingly we never find the group of Sîn, Šamaš, and Ištar mentioned together in royal inscriptions, the Pleiades (Sebettu) lacking completely despite their frequent occurrence in art. Thus we can at least differentiate between a perspective in art, be it popular or courtly, and one in royal inscriptions. Images and divine symbols seem to have another effect on grouping as the deities proper when enumerated in inscriptions.

There is also the group of Aššur, Sîn, and Ištar (RINAP 5/1, Ashurbanipal 3, col. V, line 76; 7 col. VI line 15). More specific groups are Aššur, Mullissu, and Ištar from Arbela (RINAP 5/1, Ashurbanipal 23, lines 116, 119, and 160-1) as well as Aššur, Adad, and Ištar (RINAP 5/1, Ashurbanipal 61 rev. 22’). But there are also combinations of three deities without Aššur. So there are Sîn, Šamaš, and Adad (RINAP 5/1, Ashurbanipal 2, col. I line 6; 13, col. I line 3) or Šamaš, Adad, and Ištar (RINAP 5/1, Ashurbanipal 11, col. I line 6).

Bibliography Amiet, P. 1977. Die Kunst des Alten Orient. Freiburg i. B., Herder. Andrae, W. 1938. Das wiedererstandene Assur. Leipzig, Hinrichs. Andrae, W. 1977. Das wiedererstandene Assur, 2., durchgesehene und erweiterte Auflage, hg. von Barthel Hrouda. München, C. H. Beck. Bachmann, W. 1927. Felsreliefs in Assyrien: Bawian, Maltai und Gundük (Wissenschaftliche Veröffentlichungen der Deutschen Orient-Gesellschaft 52). Leipzig, Hinrichs. Balkan, K. 1992. The conception of Trinity in the tablets of Kültepe. In H. Otten, E. Akurgal, H. Ertem and A. Süel (eds), Hittite and Other Anatolian and Near Eastern Studies in Honour of Sedat Alp: 14–44. Ankara, Türk Tarih Kurumu Basımevi. Barnett, R. D. and Falkner, M. 1962. The Sculptures of Aššur-nasir-apli (883-859 BC), Tiglath-Pileser III (745-727 BC), Esarhaddon (681-669 BC) from the Central and SouthWest Palaces at Nimrud. London, The Trustees of the British Museum. Blocher, F. 2009. Sonne und Sonnengottheiten im Alten Vorderasien. In A. Bärnreuther (ed.), Die Sonne, Brennpunkt der Kulturen der Welt, mit Vorworten von Klaus Töpfer und Hermann Scheer: 40–53. München, Edition Minerva. Blocher, F. 2010. Gestirns- und Himmelsdarstellungen im alten Vorderasien von den Anfängen bis zur Mitte des 2. Jahrtausends v. Chr. In H. Meller and F. Bertemes (eds), Der Griff nach den Sternen:

And finally there are groups with a local focus, as Nabû, Tašmētu, and Nanaya (RINAP 5/1, Ashurbanipal 2, col. I line 8) or Great Anu, Šarrat Dēr, and Marbīti (RINAP 5/1, Ashurbanipal 23, line 73). Note also Ninagal, Kusibanda, and Ninkurra (RINAP 5/1, Ashurbanipal 8, col. I, lines 13’-14’). In the Babylonian inscriptions we count twice Aššur, Šamaš and Marduk (RIMB 2, B.6.32.2 lines 7-8; B.6.32.14 line 4-5) and once Aššur, Enlil and Ninurta (RIMB 2, B.6.32.15 line 10). Ashurbanipal’s brother Šamaš-šuma-ukīn (667-648 BC) calls Enlil, Šamaš and Marduk (RIMB 2, B.6.33.1 line 4) a group. Conclusions Groups of three deities are frequent in art and royal inscriptions during the Middle and Neo-Assyrian periods. In seals and seal impressions the celestial phenomena sun, moon, star, and Pleiades are the best documented group of divine symbols from which a group of three is choosen. The basic combination crescent and star is enlarged by sun or Pleiades. On royal and private 39

Systemizing the Past Wie Europas Eliten zu Macht und Reichtum kamen, Internationales Symposium in Halle (Saale) 16.-21. Februar 2005 (Tagungen des Landesmuseums für Vorgeschichte Halle (Saale) 5/II, Halle (Saale)): 973– 987. Halle (Saale), Landesamt für Denkmalpflege und Archäologie Sachsen-Anhalt. Börker-Klähn, J. 1982. Altvorderasiatische Bildstelen und vergleichbare Felsreliefs (Baghdader Forschungen 4). Mainz, Philipp von Zabern. Brentjes, B. 1994. Selbstverherrlichung oder Legitimitätsanspruch? Gedanken zu dem Thronrelief von Nimrud-Kalaḫ. Altorientalische Forschungen 21: 50–64. Buchanan, B. 1966. Catalogue of Ancient Near Eastern Seals in the Ashmolean Museum, Volume 1, Cylinder Seals. Oxford, Clarendon Press. Calmeyer, P. 1973. Zur Genese altiranischer Motive. Archäologische Mitteilungen aus Iran NF 6: 132–152. Calmeyer, P. 1984. „Das Zeichen der Herrschaft … ohne Šamaš wird es nicht gegeben“. Archäologische Mitteilungen aus Iran NF 17: 135–153. Collon, D. 1987. First Impressions, Cylinder Seals in the Ancient Near East. London, British Museum Publications. Collon, D. 2001. Neo-Assyrian and Neo-Babylonian Periods (Catalogue of the Western Asiatic Seals in the British Museum, Cylinder Seals V). London, British Museum. Courth, F. 1988. Trinität in der Schrift und Patristik (Handbuch der Dogmengeschichte Bd. II, Fasz. 1a). Freiburg, Basel und Wien, Herder. Delaporte, L. 1923. Musée du Louvre, Catalogue des cylindres orientaux, cachets et pierres gravées de style oriental, II: Acquisitions. Paris, Hachette. Ehrenberg, E. 2002. The Kassite Cross revisited. In C. Wunsch (ed.), Mining the Archives: Festschrift for Christopher Walker on the Occasion of his 60th birthday 4 October 2002 (Babylonische Archive 1): 65–74. Dresden, ISLET. Frankfort, H. 1939. Cylinder Seals. A Documentary Essay on the Art and Religion of the Ancient Near East. London, Gregg. Fuchs, A. 1994. Die Inschriften Sargons II. aus Khorsabad. Göttingen, Cuvillier. Glassner, J.-J. 1999. Triades archaïques dans les pantheons sumériens. In B. Böck, E. CancikKirschbaum and Th. Richter (eds), Munuscula Mesopotamica, Festschrift für Johannes Renger (Alter Orient und Altes Testament 267): 161–167. Münster, Ugarit-Verlag. Grayson, A. K. and Levine, L. D. 1975. The Assyrian relief from Shikaft-i Gulgul. Iranica Antiqua 11: 29–38. Griffiths, J. G. 1996. Triads and Trinity. Cardiff, University of Wales Press. Herbordt, S. 1992. Neuassyrische Glyptik des 8.-7. Jh. v. Chr. (State Archives of Assyria Studies 1). Helsinki, NeoAssyrian Text Corpus Project.

Herles, M. 2006. Götterdarstellungen Mesopotamiens in der 2. Hälfte des 2. Jahrtausends v. Chr.: das anthropomorphe Bild im Verhältnis zum Symbol (Alter Orient und Altes Testament 329). Münster, Ugarit-Verlag. Heeßel, N. P. 2002. Pazuzu. Archäologische und philologische Studien zu einem altorientalischen Dämon (Ancient Magic and Divination IV). Leiden, Boston and Köln: Brill. Hrouda, B. 1991. Vorläufiger Bericht über die neuen Ausgrabungen in Assur, Frühjahr 1990. Mitteilungen der Deutschen Orient-Gesellschaft 123: 95–109. Hunger, H. 2003-2005. Planeten. Reallexikon der Assyriologie und Vorderasiatischen Archäologie 10: 589– 591. Jahangirfar, M. 2018. The Elamite Triads: Reflections on the possible continuities in Iranian tradition. Iranica Antiqua 53: 105–124. Kantor, H. J. 1958. The Glyptic. In C. W. McEwan, L. S. Braidwood, H. Frankfort, H. G. Güterbock, R. C. Haines, H. J. Kantor and C. H. Kraeling, Soundings at Tell Fakhariyah (Oriental Institute Publications 79): 69–85. Chicago, University Press. Khvedilidze, M. 1982. Babylonian and Egyptian Triads. In H. Klengel (ed.), Gesellschaft und Kultur im alten Vorderasien: 137–141. Berlin, Akademie-Verlag. Klengel-Brandt, E. 2014. Die neuassyrische Glyptik aus Assur, mit Beiträgen von Sabine Böhme und Othmar Keel (Wissenschaftliche Veröffentlichungen der Deutschen Orient-Gesellschaft 140). Wiesbaden, Harrassowitz Verlag. Krebernik, M. 2006-2008. Richtergott(heiten). Reallexikon der Assyriologie und Vorderasiatischen Archäologie 11: 354–361. Krebernik, M. 2009-2011. Sonnengott. A. I. In Mesopotamien. Philologisch. Reallexikon der Assyriologie und Vorderasiatischen Archäologie 12: 599– 611. Kurmangaliev, A. 2009-2011. Sonnengott. B. I. In Mesopotamien. Archäologisch. Reallexikon der Assyriologie und Vorderasiatischen Archäologie 12: 616– 620. Lambert, W. G. 1979. Ancient Near Eastern seals in the Gulbenkian Museum of Oriental Art. Iraq 41: 1–45. Luckenbill, D. D. 1927. Ancient Records from Assyria and Babylonia, Vol. II: Historical Records from Assyria, from Sargon to the End. New York, Greenwood (=Reprint 1968). Melikišvili, G. A. 1965. Die Götterpaartrias an der Spitze des urartäischen Pantheons. Orientalia NS 34: 141– 145. Mierzejewski, A. and Sobolewski, R. 1980. Polish excavations at Nimrud/Kalhu. Sumer 36: 151–162. Moortgat-Correns, U. 1996. Das Grab des Nabonid. Studi Micenei ed Egeo-Anatolici 38: 153–177. Orthmann, W. 1975. Der Alte Orient, mit Beiträgen von Pierre Amiet et al. (Propyläen Kunstgeschichte 14). Berlin, Propyläen Verlag. 40

Felix Blocher: Groups of Three Deities in Middle and Neo-Assyrian Times Parpola, S. 2000. Monotheism in Assyria. In B. N. Porter (ed.), One God or Many? Concepts of Divinity in the Ancient World (Transactions of the Casco Bay Assyriological Institute 1): 165-209. Casco Bay, Casco Bay Assyriological Institute. Podella, T. 1994. Das Lichtkleid JHWHs: Untersuchungen zur Gestalthaftigkeit Gottes im Alten Testament und seiner altorientalischen Umwelt (Forschungen zum Alten Testament 15). Tübingen, Mohr. Pongratz-Leisten, B. 1994. Ina šulmi īrub. Die kulttopographische und ideologische Programmatik der akītu-Prozession in Babylonien und Assyrien im 1. Jahrtausend v. Chr. (Baghdader Forschungen 16). Mainz, Philipp von Zabern. Porada, E. 1948. The Pierpont Morgan Library Collection, Corpus of Ancient Near Eastern Seals in American Collections (The Bollingen Series 14). Washington, Pantheon Books. Porter, B. N. 2000. The anxiety of multiplicity: Concepts of divinity as one or many in ancient Assyria. In B. N. Porter (ed.), One God or Many? Concepts of Divinity in the Ancient World (Transactions of the Casco Bay Assyriological Institute 1): 211–271. Casco Bay, Casco Bay Assyriological Institute. Radner, K. 2012. The stele of Adad-nērārī III and Nergal-ēreš from Dūr-Katlimmu (Tell Šaiḫ Ḥamad). Altorientalische Forschungen 39: 265–277. Reade, J. 1977. Shikaft-i Gulgul: Its date and symbolism. Iranica Antiqua 12: 33–44. Smith, S. and Gadd, C. 1938. Assyrian Sculptures in the British Museum from Shalmaneser III to Sennacherib. London, The British Museum. Watanabe, K. 1992. Nabû-usalla, Statthalter Sargons II. in Tam(a)nūna. Baghdader Mitteilungen 23: 357–369. Woods, C. E. 2004. The Sun-God tablet of Nabû-aplaiddina revisited. Journal of Cuneiform Studies 56: 23– 103.

Abbreviations RIMA 1. Grayson, A. K. 1987. Assyrian Rulers of the Third and Second Millennia BC (to 1115 BC) (The Royal Inscriptions of Mesopotamia, Assyrian Periods Vol. 1). Toronto, Buffalo and London, University of Toronto Press. RIMA 2. Grayson, A. K. 1991. Assyrian Rulers of the Early First Millennium BC, I (1114-859 BC) (The Royal Inscriptions of Mesopotamia, Assyrian Periods Vol. 2) Toronto, Buffalo and London, University of Toronto Press. RIMA 3. Grayson, A. K. 1996. Assyrian Rulers of the Early First Millennium BC, II (858-745 BC) (The Royal Inscriptions of Mesopotamia, Assyrian Periods Vol. 3). Toronto, Buffalo and London, University of Toronto Press. RIMB 2. Frame, G. 1995. Rulers of Babylonia, from the Second Dynasty of Isin to the End of Assyrian Domination (1157-612 BC) (The Royal Inscriptions of Mesopotamia, Babylonian Periods Vol. 2). Toronto, Buffalo and London, University of Toronto Press. RINAP 1. Tadmor, H. and Amada, S. with the editorial assistance of Novotny, J. 2011. The Royal Inscriptions of Tiglath-pileser III (744-727 BC), and Shalmaneser V (726-722 BC), Kings of Assyria (The Royal Inscriptions of the Neo-Assyrian Period Vol. 1). Winona Lake, Indiana, Eisenbrauns. RINAP 3/1. Grayson, A. K. and Novotny, J. 2012. The Royal Inscriptions of Sennacherib King of Assyria (704-681 BC), Part 1 (The Royal Inscriptions of the Neo-Assyrian Period Vol. 3/1). Winona Lake, Indiana, Eisenbrauns. RINAP 4. Leichty, E., with a contribution of Grant Frame 2011. The Royal Inscriptions of Esarhaddon, King of Assyria (680-669 BC) (The Royal Inscriptions of the Neo-Assyrian Period Vol. 4). Winona Lake, Indiana, Eisenbrauns. RINAP 5/1. Novotny, J. and Jeffers, J. 2018. The Royal Inscriptions of Ashurbanipal (668-631 BC), Aššur-etelilāni (630-627 BC), and Sîn-šarra-iškun (626-612 BC), Kings of Assyria, Part 1 (The Royal Inscriptions of the Neo-Assyrian Period Vol. 5/1). University Park, Pennsylvania, Eisenbrauns.

41

Water Management in Ancient Armenia: Problems and Perspectives Tork Dalalyan, Roman Hovsepyan, Levon Abrahamian, Arsen Bobokhyan and Boris Gasparyan Institute of Archaeology and Ethnography, NAS RA, Yerevan, Armenia

Abstract: Water resources have gained greater importance all around the world in the context of growing volumes of water use and current environmental problems becoming a key factor in economic and political developments. Taking into account the current geopolitical and economic processes in the region, the study of all spheres of water use in the Armenian Highland, including the cultural aspects, becomes a vital question. Such an undertaking aims at bringing together the data of archaeology, ethnography, philology and other disciplines related to the field of water use, by involving the tools of comprehensive interdisciplinary research. It will allow revealing the past experience of water management in the region, the material and non-material cultural heritage regarding the exploitation of water resources, as well as the problems regarding the present and future of the use of water in Armenia. Keywords: Armenian Highland, water management, irrigation systems, agriculture, Neolithic to Middle Ages, interdisciplinary approaches

Introduction Although experts have paid certain attention to the early use of water resources in the Armenian Highland, the problem has never been considered systematically. Our investigation is the first attempt, which aims at highlighting the past experience of water use, the material and non-material cultural heritage related to water resource management, the issues related to the current challenges in the use of water resources of the region. We are coming out from the principle that the same relief and similar climatic conditions dictate similar mechanisms of water use since ancient times to the present day, patterns that remain almost unchanged over the centuries, while only the technical capabilities may change.1 Armenian Historian Ghazar Parpetsi in the 5th century AD, for the first time in Armenian reality, figuratively describes the natural irrigation system of the Ararat valley in central Armenia.2 Later the problem was always under the attention of the Armenian State and the Armenian Church. In modern times, the study of water economy and artificial irrigation systems became especially active in 1920s, due to economic and ideological reasons.3 Not accidentally, the period of large water-construction projects (cf. Ayghr Lake, Shirak and Sardarapat canals) coincides with the main time of investigation of water use in ancient Armenia. In connection with the construction of Ayghr Lake in 1925-1927, a well-known state-man and historian This work was supported by the Committee of Science, Republic of Armenia, under grant number 21AG-6A080. 2 Ghazar Parpetsi I, 21. 3 Aghakhanyan 1957: 233; Gasparyan 1968: 190; Shirmazan and Darbinyan 1961: 49. 1

Systemizing the Past (Archaeopress 2023): 42–55

Figure 1. A. Kalantar on the mount Aragats (State Museum of Armenia, A. Kalantar Collection, N 8561).

Ashot Hovhannisyan refers to early archaeological constructions of the country as predecessors of modern irrigation systems.4 In scientific research of early water economy and water-related cult of ancient Armenia the works of A. Kalantar,5 G. Shirmazan6 (Figures 1-4), as well as G. Aghakhanyan,7 R. Yepremyan8 are especially noteworthy. Recent research deals with the ancient ways of water resource management, the problems of creation and exploitation of agricultural-horticultural landscapes, ecological niches related to the generation of water resources, the problems of the impact of climate change on the livelihoods,9 communication Cf. Shirmazan 1961: 146. Gasparyan 1968: 190; Kalantar 1937. Shirmazan 1961; 1962; Shirmazan and Darbinyan 1961. 7 Aghakhanyan 1960. 8 Yepremyan 1992. 9 Arimura et al. 2014; Avetisyan et al. 2005; Barnard et al. 2010; 4 5 6

Tork Dalalyan et al.: Water Management in Ancient Armenia

Figure 2. Scetch of water supply system of the mount Aragats drawn by A. Kalantar (Family Archive of A. Kalantar, courtesy of Vahagn Gurzadyan).

It goes without saying that water use played a crucial role in the creation of early complex societies. In this regard, important is the well-known theory of K.-A. Wittfogel on the ‘irrigation’ state, according to which the state emerged as a result of the organization of irrigated agriculture, which presupposed collective work, general consolidation of forces and the emergence of a government administration.15 This theory can be useful also in the Armenian context, where water use, of course in its own ways, played an essential role in the formation of early organized society. In this context, it is noteworthy that both in ancient and modern times, the Ararat Valley, as a core region of historical Armenia, and the problem of its irrigation became a key issue in the shaping of the Armenian political landscape.16

systems, natural resources and water cult according to archaeological data,10 linguistic and cultural layers and epic traditions concerning the origins of waterrelated deities and water worship, agricultural terms and hydronyms,11 corresponding rituals and beliefs,12 places of pilgrimage associated with springs and healing waters, ‘water’-centered modern architectural concepts.13 All specialists associated the prehistoric monumental statues of the Armenian mountains called vishap (dragon stone) with the ancient cult of water and water use, considering them protectors of irrigation constructions.14 Gasparyan 2010; Hobosyan et al. 2020; Hovhannisyan et al. 2017; Hovsepyan 2015; Ollivier et al. 2009; 2010; Petrosyan and Gasparyan 2012; 2014; Sherrif et al. 2019; Wilkinson et al. 2012. 10 Bobokhyan et al. 2018; Hnila et al. 2019; Hovsepyan 2021. 11 Dalalyan 2004; 2006; 2008, 2011. 12 Hovsepyan 2017; 2018; Mkrtchyan 2006; 2010. 13 Abrahamyan 2011; 2012; 2019. 14 Abeghyan 1941; Atrpet 1929; Barseghyan 1967; Ghapantsyan 1952; Gilibert et al. 2012; Hobosyan et al. 2021: 50-62, 72-99; Kalantar 1933;

1937; Karapetyan 1979; Khachatryan 1971; Khanzadyan 1969; Marr and Smirnov 1931; Petrosyan and Bobokhyan 2015; Piotrovskij 1939. 15 Wittfogel 1957. 16 Аghakhanyan 1957: 234.

43

Systemizing the Past

Figure 3. G. Shirmazan by the Urartian inscription of Zvartnots (National Archive of Armenia, collection 810, list 1, unit 10, p. 16).

Figure 4. Page from the work of G. Shirmazan ‘Irrigation in Armenia from the ancient times till now’ (National Archive of Armenia, collection 810, list 1, unit 1, p. 4).

unique hydrographic node, from where the largest rivers of the Near East originate: Euphrates, Tigris, Araxes, Kura and Chorokh.17 Every year about 90 km3 of water flows from the Armenian Highland to the World Ocean through rivers. The total length of the Armenian Highland river network is 180 thousand km. One of the hydrographic features of the Armenian Highland is the abundance of lakes of different origins (tectonic, tectonic-volcanic, volcanic, glacial). The depressions of the three major lakes of the region Sevan, Van, and Urmia are of tectonic origin, and both Sevan and Van are also results of lava block. Among the tectonic volcanic lakes, Tsovak Hyusisoy (Tseli) is also noticeable. There are lakes of glacial origin, placed in cirques or moraines such as Kari Lake on Aragats, Akna Lake on the Geghama mountain range, Sev Lake on Ishkhanasar, numerous small lakes on the Byurakn plateau. Because of the landslides and collapses also blocked lakes were originated (Parz, Gosh). The crater lakes are widespread, the largest of which is Nemrut (9 km2), followed by Azhdahak and Armaghan on the summits of anthropogenic volcanoes (cf. Figures 5-22).

Landscape and Water Management Since prehistoric times, when people lived in shelters of natural origin, the existence of at least one source of water in the neighborhood was a necessary condition. There has never been a man-occupied area in the world that is waterless, but there have been and still are favorable and unfavorable zones. Thus, in the Near East, unlike the mountainous regions of the ‘Fertile Crescent’, the steppe zone of Syria-Mesopotamia nowadays is characterized by unfavorable water resources. Mountainous areas are not only rich in atmospheric precipitation during almost all seasons, but also, due to the presence of several climatic zones, create favorable environment, which provides ‘winter storage’ of water resources in the form of snow and ice. In the spring, these frozen storages of water thaw and form flooding ‘water arteries’, tributaries and rivers, and provide the annual water supply until the next ‘storage period’. The presence of large pools of water and the evaporation of water, especially during the hot seasons of the year, are additional impetus for the smooth operation of the natural atmospheric cycle. The Armenian mountains (and especially Ararat, Aragats, Geghama, Byurakn massifs, etc.) create a

17

44

Cf. Gabrielyan 2000; Hakobyan 1981; Hewsen 2001.

Tork Dalalyan et al.: Water Management in Ancient Armenia

Figure 5. Mount Ararat from Igdir (Photo by A. Bobokhyan).

Figure 8. Mount Ararat from Sakhurak (Geghama mountains) with vishap N 4 (Photo by A. Bobokhyan).

Figure 6. Slopes of mount Ararat from direction of Bayazet (Photo by A. Bobokhyan).

Figure 9. Mount Aragats from the peak Tirinkatar (Photo by A. Bobokhyan).

Figure 7. Snow accumulations on mount Ararat on c. 3000 m a.s.l. (Photo by A. Bobokhyan).

Figure 10. Kare Lake on mount Aragats (Photo by A. Bobokhyan).

45

Systemizing the Past

Figure 11. Gegharda Lake on Geghama mountains (Drone Photo by A. Mkrtchyan).

Figure 13. Crater pond on the mount Azhdahak, Geghama mountains (Photo by T. Dalalyan).

Figure 12. Lcharot pond on Geghama mountains (Photo by A. Bobokhyan).

Figure 14. A pond on the slopes of mount Amul Sar, Zengezur mountains (Photo by T. Dalalyan).

Data of Material Culture

eight thousand years ago. Possibly, other settlements of the above-mentioned tradition, such as the ones in the Ararat valley, also had similar irrigation systems.20 The presence of ancient irrigation networks, canals and dams near the Early Bronze Age settlements of Armenia (e.g. Mokhrablur) indicates the development of agriculture / horticulture.21

Prehistoric and early historic period Back in the early stages of the settlement at the Armenian Highland, in the Paleolithic period, local inhabitants organized their activities on the shores of lakes. Besides, traces of springs have been found near many stations and sites.18

Bronze and Iron Age irrigation systems based on the distribution of snowmelt water are known in the Aragats and Geghama mountains (Figures 2, 23). A number of rock-carved drawings have been preserved, which are interpreted as maps of the hydrographic system (Figure 24).22 Some rock-engraved images in Balahovit (Kotayk) and Partizak (Talin) villages presumably resemble

The origins of artificial irrigation in the Armenian Highland trace back to the early agricultural societies. An irrigation system has been discovered in the area of Gadachrili Gora Neolithic settlement in Georgia, which is the oldest in the region.19 This system was constructed and used by the Aratashen-ShulaveriShomutepe culture community, which existed about 18 19

Dolukhanov 1979: 70; Hovsepyan 2015; Torosyan 1976: 107-108. Harutyunyan et al. 2005: 19; Jalalbekyan 1974; Tumanyan 2008: 1011. 22 Khanzadyan 1978: 114; Vardumyan 2004: 142, 144. 20 21

Colonge et al. 2013; Gasparyan 2010; Montoya et al. 2013. Ollivier et al. 2018.

46

Tork Dalalyan et al.: Water Management in Ancient Armenia

Figure 15. Big Al Lake, Borderline between Armenia and Artsakh (Photo by A. Bobokhyan).

Figure 18. Van Lake from Artamet (Photo by A. Bobokhyan).

Figure 16. Sev Lake, Syunik mountains (Photo by A. Bobokhyan).

Figure 19. Urmia Lake from Aghavno fortress (Photo by T. Dalalyan).

Figure 17. Sevan Lake from Artanish, with the view on Geghama mountains (Photo by A. Bobokhyan).

Figure 20. A peninsula in Urmia Lake (Photo by T. Dalalyan).

47

Systemizing the Past Kingdoms of Van and Greater Armenia Constructions of irrigation systems reached their peak during the Kingdom of Van (9th-6th centuries BC), when permanently flowing waters of rivers have been used instead of the snowmelt waters of mountains. In addition, previously unknown engineering solutions were implemented (rock processing, tunnels), as well as ordinary water streams and pipes,26 possibly also wells.27 Some of these canals and reservoirs are still operating (for example, the ones in Sardarapat, Echmiadzin, Dalma, Kyanqi Aru, Mamri, Ashtarak, etc.). The canal built by the King Menua is particularly noteworthy, the construction of which is attributed by Movses Khorenatsi to the mythological queen Shamiram (Figure 26). That canal, starting from Hayots-Dzor, flows 75 km. There are 14 cuneiform inscriptions along its entire length, which tell about the construction of the canal. In addition, the Sardarapat inscription of Argishti I mentions 4 irrigation canals simultaneously beginning from the river Araxes, after construction of which vineyards were planted.28 One of the masterpieces of water construction works of the Kingdom of Van is the reservoir Keshish-gol (‘Lake of Rusa’) built by Rusa I, 30 km away from the city of Van, in the vicinity of the Ton village. After the Urartian period until the 7th century AD no any monument related to waterworks has been found yet.29

Figure 21. Tsovak Hyusisoy Lake (Photo by T. Dalalyan).

Middle Ages and Modern Period Archaeological data indicate that water supply and drainage systems reached a high level of development in the urban processes of medieval Armenia. Information on irrigation systems, disputes over their purchase and sale have been preserved in the works of Ghazar Parpetsi, Pavstos Buzand, Hovhannes Draskhanakertsi, Tovma Artsruni, Stephanos Orbelyan, Hovhannes Erznkatsi, Mkhitar Gosh, Zakaria Kanakertsi, Abraham Kretatsi and in the epigraphic inscriptions (for example, in Aruch). From the point of view of maintaining the national irrigation network, the centers of the Armenian Church, such as Echmiadzin, the Tatev monastery and the S. Karapet monastery have played an important role.30 Khachkars (cross-stones) were erected and chapels were built near the water sources. A number of canals were built especially in the High Middle Ages (Ashtarak, Yeghvard, Gndevaz, Akner),

Figure 22. Araxes river from Parspatunik (Photo by T. Dalalyan).

water-supplying systems.23 In the context of vishaps, depictions of irrigation networks are present also amongst the petroglyphs of the Geghama mountains.24 As mentioned above, vishaps were placed in the nodes of these water distribution systems, to be related, among others, also to the cult of water.25

Javakhketi, Trialeti and Tayk mountains. Nowadays, about 150 examples of these monuments are known, about 100 of which are in the territory of the Republic of Armenia (Bobokhyan et al. 2018; Hnila et al. 2019; Petrosyan and Bobokhyan 2015). 26 Belli 1997; Grekyan 2002; Hovhannisyan 1996; Meshchaninov 1931; Piotrovskij 1959: 138-140. 27 Bdoyan 1972: 322. 28 Harutyunyan 1964: 118. 29 Shirmazan 1962: 43. 30 Bdoyan 1972: 322-324; cf. also Babayan and Zhamkochyan 2015: 3438; Harutyunyan 1971: 48-58; Hasratyan 1974.

Asatryan 2004: 24; Grigoryan 1990: 99. Vardumyan 2004: 142, 144. 25 Shirmazan 1962: 50-51, 106-107. Main areas of distribution of vishaps in the Armenian Highland are Aragats, Geghama, Vardenis, 23 24

48

Tork Dalalyan et al.: Water Management in Ancient Armenia

Figure 23. Map of ancient water supply system of the Geghama mountains, according to A. Kalantar (Reconstruction by B. Gasparyan).

49

Systemizing the Past

Figure 24. Prehistoric map of water supply system carved on a tufa platform in Kakavaberd, Aragatsotn (Photo by B. Gasparyan).

Figure 25. Water supply system of ‘12 bander/canals’ on the southern slopes of mount Aragats, used since prehistoric times until mid 20th century (Photo by A. Bobokhyan).

which is evidenced by epigraphic and manuscript data.31 Although no major water distribution construction works were carried out during the Persian rule, an attention was paid to the preservation of the existing canals (e.g. Dalma). After the annexation of Eastern Armenia to Russia, the Arazdayan and Shangire canals were built. Especially large canal construction works were carried out in Armenia during the Soviet period, when canals and reservoirs of ancient times were reused, such as the ones of Kari and Gegharda Lakes.32 The traditions of the water supply system have been quite stable over the centuries. Many water distribution structures have their origins in prehistoric times, with new engineering solutions.33 Among them are worth mentioning a part of the cyclopean dam found at the Kari Lake of Aragats, the same in the Geghama Mountains near Akna Lake,34 the Gegharda Lake reservoir of the Geghama mountains,35 the irrigation system of Ashtarak fed from Aragats, and especially Veri Aru,36 Karmir Sar / Tirinkatar (Kizil-Ziarat) of Aragats with its ‘12 bander/canals’ where the largest number of vishaps (dragonstones) are concentrated,37 as well as many Urartian dams (Figures 10-11, 25-26).

Figure 26. Menua Canal by Van and its modern transformation (Photo by A. Bobokhyan).

Concepts from Spiritual Culture Water worship and religious perceptions Water, as the source and a synonym-symbol of life, has a key place in the religious systems of the Near East (particularly the Armenian Highland), since the ancient times. In cosmogonic myths, primordial chaos was combined with dark waters (primordial sea, worldwide ocean, etc.), from which the universe, the cosmos originates.38 Usually the earth (the disk of the earth, the globe) rises from it. In these cosmogonic myths, a mythical sea monster in the shape of snake-fish or a dragon-fish, floating in the water from the immemorial times, plays an important role. In the Near-Eastern and Armenian beliefs, this dragon is also a symbol of infinity and eternity, as it is depicted as a ring with its head reaching for its tail and wrapped around the earth. Dragon-fish roaming in the world sea cause

Shirmazan 1962: 7-8, 147. Shirmazan 1962: 7-8; cf. Harutyunyan et al. 2005: 19-21. 33 Mints 1966: 247. 34 Gabrielyan 2000: 213. 35 Shirmazan 1962: 107; cf. also Mkrtchyan 1962: 523. 36 Saghumyan 1998: 46-47, 103-105. 37 Anonym 1985: 189; Lisitsyan 1940: 61. 31 32

38

50

Cf. Averintsev 1991.

Tork Dalalyan et al.: Water Management in Ancient Armenia earthquakes, floods, eclipse of lights,39 thus periodically disrupting the order established since the creation of the world, bringing it closer to the primordial chaos. Related to him are the eschatological myths of destruction of the world, in which the belief in the return of life to its original water form is expressed. This danger always hangs over the head of the world, the primordial dark sea continues to exist even after the creation of the Universe.40

trigger hero is determined by the Lake Van (myths of Hayk, Gagik and Tsovinar, Sanasun foundation). The river / water is an important component in the semi-historical and semi-mythological descriptions of the establishment of the Armenian royal centers and capitals. Armavir is built ‘on a hill on the river bank’, and the river is named after the grandson of the founder of the city.47 In the construction of VanShamiramakert the main place is occupied by the ‘river dam’, city streams and irrigation canal constructions.48 Yervandashat is built ‘on a single rocky hill, which surrounds Yeraskh, and the Akhuryan river flows from its front’.49 Bagaran is built ‘on the Akhuryan river’.50 The plan of Yervandakert and its surroundings is deliberately resembled the face of a ‘beautiful virgin’, in which ‘the river presented a mouth with its two lips at the height of its two banks’.51 Artashat is built ‘in the place where Yeraskh and Metsamor join… Yeraskh helps him with the wood of the forest’,52 Vagharshavan is built near the mixture of Murts and Yeraskh rivers.53 Vardgesavan and later in its place, Vagharshapat, were built near the Kasakh river, moreover the latter is mentioned twice with a poetic rhyme in a fragment of a verse, which supposedly described the building of that city through an epic song.54

The earliest mother goddesses, who become the foremother of the gods in mythological family trees, and later the source of life and everything in the world, at the same time personify the world sea and appear as its symbol (Inanna, Ishtar, Nanaya, Ninu). In the Armenian Highland that role played Tsovinar, who according to the reconstructions of various scientists, was the mother goddess of the ancient Armenian mythology. At the same time, she was considered the personification of the Purple sea.41 The word Tsovinar / Tsovian / Tsovin was not only a mythical name, but also a common name that meant ‘thunder’ and ‘rainbow’. Here the principle of unification of opposing pairs is important, according to which a mythological image, partly identified with water and with fire, plays a central role in the process of world creation.42 The root tsov in Armenian has an ancient sub-stratum origin and comes from the local tribal languages of ‘Nairi lands’, meaning ‘sea, lake’, ‘water basin, terrestrialreal or celestial-mythological’. It is not an accident that in the cuneiform system of the Kingdom of Van, this root is also adopted, meaning ‘sea’ or more often, ‘lake’. Among the 79 deities mentioned in the famous inscription of ‘Mher’s door’ there is a mythical name Ṣuininaúe, which is rightly associated with Tsovinar.43 However, like Ishtar of Mesopotamia and other goddesses parallel to her, Tsovinar could appear as a bisexual deity, appearing in male and female versions.44 This is explained by the fact that water had fertilizing properties.45

Festive system and folklore In the Armenian festive system, the two main waterrelated festivals, Jrorhnek (Blessing of the Waters) and Vardavar took place during opposing weather cycles, winter and summer. The first of these has mostly Christian symbolism, while the second is essentially a pre-Christian festival; it is obviously related to the cult of love, beauty, water goddess Astghik. The popularity and pan-Armenian nature of this festival is evidenced by the fact that it was widespread in all Armenian sub-ethnic groups with many dialect variants. On mythological level, the water in the Vardavar festival has a symbol of fertility, and is directly related to the myth of the love goddess – her ‘river bath’. In Christian context, it is related to the myth of the world flood, in which Astghik appears as Noah’s daughter. Thus, sprinkling with water during Vardavar, in Christian interpretation is explained as a memory of the world flood.

In Armenian genealogical myths and legends, the Van Lake is perceived as a real terrestrial model of the world mythological sea, which is at the same time the habitat of sun and dragons,46 i.e. contains both embryos of the chaos and of the cosmos. Due to this mythological significance, in the Armenian ethnogenetic legends the settlement direction of forefather / foremother culture-

In addition to these two festivals water plays a significant role also in other traditional holidays,

Lalayan 1896: 318; Nzhdehian 1902: 264. Harutyunyan 2000: 9-12; Nzhdehian 1902: 263. 41 Abeghyan 1966: 76-80. 42 Cf. Toporov 1977: 89. 43 Hmayakyan 1990: 31-32; Russell 1994: 43. 44 Abeghyan 1975: 70. 45 Abrahamyan 1983: 143-144. 46 Cf. Srvandztyants 1978: 69-70, 77; Ghanalanyan 1969: 82-83.

Khorenatsi I.12, 49. Khorenatsi I.16, 65-67. 49 Khorenatsi II.39, 195. 50 Khorenatsi II.40, 197. 51 Khorenatsi II.42, 199. 52 Khorenatsi II.49, 211. 53 Khorenatsi II.65, 241. 54 Khorenatsi II.65, 241.

39

47

40

48

51

Systemizing the Past such as the ritual of prediction though the songs of the Ascention celebration. In any case, apart from the fertility/sexual function, which is the general abstract concept of water worship, in agricultural communities such as the Armenians, water had a more specific ‘practical’ function, which was aimed at irrigating the land. Targeted, i.e. rain-making rituals that are not associated with any calendar celebration are performed to demonstrate this function; they are performed on demand in certain climatic conditions (drought or, conversely, showers) to cause or stop rain.

unifying role. The mountain range and the highest peak, the river that originates from it, the largest settlement in the valley of that river, the tribe that lived in that settlement and in the smaller villages around it were often called by the same name. The river and the river valley contributed to the formation of an isolated communication system, because especially in the winter months, due to the closure of mountain passes and gorges, the population of the river valleys was separated from other parts of the country. The main river with several mountain tributaries facilitated the formation of tribes, which were integrated like streams into the larger social units living around the main river. That is how tribal languages emerged, and later separate dialects and sub-dialects.

The mentioned ceremonies were accompanied by special ritual songs, which are characterized by the presence of ancient cultic layers, metaphorical, euphemistic, onomapoetic and ideophonic names of pagan deities. These genres of folklore, together with other genres (prayer, oath, proverb, etc.), represent the earliest examples of folklore that could have preserved the traces of ancient pagan narratives as purely as possible. Similarly, in the more conservative genres of the epic, the echoes of the cult of water sometimes have very deep roots.

The lakes played a similar role: in this case the lake catchment replaced the river basin in creating a common communication system. Moreover, if the rivers were largely unnavigable, the lake basins made it possible to use the waterway connections. The forts built in the coastal areas were intended to control the water areas. As landmarks, lakes have played a unique role in modeling of space. People in the Armenian Highland have been living on the plateaus between the mountains and the lakes. The ‘triangle’ of the three lakes, the mount Ararat (Masis) in its center, became the axis or ‘gravitational field’ around which almost all historical and cultural units of the Armenian Highland were formed having the same features (for early period, for example, pottery traditions, settlement systems, burial rites, worldviews, etc.). The more we move from this ‘identity-shaping triangle’ away, the more invariant cultural elements inherent to that key area diminish. In this sense, ancient Armenia can be defined as a ‘mountains- and lakes-oriented’ social space, in which the integrated tribes and clans joined the ‘heart’ of the country through the ‘pulses’ of water-rich rivers and streams flowing from the high mountains.

A separate group consists of topographic etymological legends, which explain the origin of springs, streams, rivers, lakes, and are often connected with specific ethnographic customs (pilgrimage, sanctuary cult, healing, fulfilment of dreams, etc.). One of the important functions of miraculous springs was the treatment of infertility. Such springs were called childbearing, and there were particular male and female springs that gave birth to a child of the appropriate sex.55 The purificating function of water is closely connected to transitional rituals (initiation), one of which is, e.g., the baptism. It was considered that a person baptized in water returns to his or her mother’s womb as a child and is reborn from there completely new and purified.56 Discussion

Due to its geographical location, the area under consideration turned into a buffer zone between various cultural traditions. Various interests clashed in this region rich in natural resources. Metals, stones and route control are often mentioned in the literature as causes for such clashes, but the issue of water and the struggle for it has never been the subject of a separate study.

The organization of water resources has to some extent contributed to the emergence of communication systems and the formation of identities. The riverbeds were the indicators of the main routes that formed the communication network also for the Armenian Highland. Of such rivers, the importance of the Araxes (cf. Figure 21) in particular was great not only as an internal communication route, but also as a culture shaping phenomenon (not accidentally referred to as ‘Mother Araxes’ in Armenian perception). The main tribal communities of the Armenian Highland were formed and existed in the valleys of particular rivers, with the mountain ranges surrounding the valleys as their natural border. Thus, the rivers played also a 55 56

The first political formations fighting for natural resources in the Near East were the Uruk colonizers at the end of the 4th millennium BC, when they established colonies in the Upper Euphrates-Tigris region, trying to exploit natural resources of mountainous regions. It is possible that they were intended to take over economically the lands from which these two rivers, which were vital to them, flowed. Perhaps the Assyrians were invading the Ararat Kingdom (Urartu) for the

Halajyan 1973: 41. Cf. Averintsev 1991.

52

Tork Dalalyan et al.: Water Management in Ancient Armenia same purpose. The desire to seize water resources is evident especially in later periods, up to the Ottoman Empire and Safavid Iran. Especially the Persian authorities had created a whole network of water structures in Armenia (construction of their prototypes started in the Sasanian period). Also in our days water seize playing important role in reshaping of political landscapes. The recent Artsakh war clearly indicates that the territorial struggle is essentially directed to gaining of water resources (cf. the Al Lakes or the Sev Lake, which have been used for irrigation since at least the Bronze Age, as evidenced by archaeological remains around them). Thus, the functions of water resources management have not changed much at present; only economic means have transformed, together with the logic of using of water energy.

Anonym, 1985. History of Irrigation and Drainage in the USSR. New Delhi, International Commission on Irrigation and Drainage. Asatryan, E. 2004. Monuments of Talin Region. Yerevan, Monument (in Armenian). Atrpet 1929. Chorokh Basin. Vienna, Mkhitaryan Printing House (in Armenian). Averintsev, S. 1991. Water. In S. A. Tokarev (ed.), Myths of the Peoples of the World, vol. 1: 240. Moscow, Soviet Encyclopedia (in Russian). Babayan, F. and Zhamkochyan, A. 2015. Outlines of the History and Archaeology of Ancient Yerevan. Yerevan, Service for the Protection of Historical-Cultural Environment (in Armenian). Barseghyan, L. A. 1967. ‘Vishaps’ of Geghama mountains. Patma-banasirakan handes (Historical-Philological Journal) 4: 181–188. (in Armenian). Bdoyan, V. 1972. Agriculture in Armenia. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Armenian). Belli, V. 1997. Urartian Irrigation Canals in Eastern Anatolia. İstanbul, Arkeoloji ve Sanat Yayınları. Bobokhyan, A., Gilibert, A. and Hnila, P. 2018. Karmir Sar: New Evidence on Dragon Stones and Ritual Landscapes on Mount Aragats, Armenia. In A. Batmaz, G. Bedianashvili, A. Michalewicz and A. Robinson (eds), Context and Connection: Essays on the Archaeology of the Ancient Near East in Honour of Antonio Sagona (Orientalia Lovaniensia Analecta 268): 255–270. Leuven, Paris and Bristol, Peeters. Colonge, D., Jaubert, J., Nahapetyan, S., Ollivier, V., Arakelian, D., Devilder, G., Fourloubey, C., Jamois, M.-H., Gasparian, B. and Chataigner, C. 2013. Le paléolithique moyen de la haute vallée du Kasakh (Arménie): caractérisation technologique et peuplement de montagne. Paléorient 39/2: 109–140. Dalalyan, T. 2004. Outlines of comparative analysis of the epics ‘Sasna tsrer’ and ‘Nartica’. In A. Yeghiazaryan and A. Petrosyan (eds), ‘Sasna tsrer’ and the World Epic Heritage, Materials of International Conference: 82– 87. Yerevan, Publishing House ‘Gitutyun’, NAS RA (in Armenian). Dalalyan, T. 2006. On the character and name of the Caucasian Satana (Sat‘enik). Aramazd: Armenian Journal of Near Eastern Studies 1: 239–253. Dalalyan, T. 2008. The first branch of the epic ‘Vipasank’ and the origin of the thunder heroic family. In A. Petrosyan (ed.), Shnorh i verust: Myth, Ritual, History (Festshrift for the 80th Anniversary of Sargis Harutyunyan): 19-50. Yerevan, Gitutyun (in Armenian). Dalalyan, T. 2011, Terms concerning the city life borrowed from Armenian and Georgian into Ossetic (Alanian). Revue des études arméniennes 33: 1–11. Dolukhanov, S. 1979. Paleogeography and primitive settlements of the Caucasus and Central Asia. Patma-

Bibliography Abeghyan, M. Kh. 1941. Monuments Called ‘Vishaps’ as Statues of the Goddess Astghik-Derketo. Yerevan, Armfan (in Armenian). Abeghyan, M. Kh. 1966. Works, vol. 1. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Armenian). Abeghyan, M. Kh. 1975. Works, vol. 8. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Armenian). Abrahamyan (Abramyan), L. A. 1983. Primitive Festival and Mythology. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Russian). Abrahamian, L. 2011. Yerevan Sacra: Old and new sacred centres in the urban space. In Ts. Darieva, W. Kaschuba and M. Krebs (eds), Urban Spaces after Socialism. Infrastructures, City Symbols and Everyday Life in Eurasia: 131–151. Chicago, University of Chicago Press. Abrahamian, L. 2012. Yerevan: Memory and forgetting in the organization of post-Soviet urban space. In A. Baiburin, C. Kelly and N. Vakhtin (eds), Russian Cultural Anthropology after the Collapse of Communism: 254–275. London, Routledge. Abrahamian, L. 2019. Dragon at the well: How the Australian Rainbow-Python, the Maori chthonic Great Mother and Indian mysterious serpent helped the Armenian Vishap. In A. Bobokhyan, A. Gilibert and P. Hnila (eds), Vishap between Fairy Tale and Reality: 119–130. Yerevan, Institute of Archaeology and Ethnography, NAS RA. Aghakhanyan, A. 1957. Development of irrigation in Soviet Armenia. In Proceedings of the Armenian Research Institute of Hydraulic Engineering and Meloration (Arm. NIIGiM), vol 2: 233–252, Yerevan, Arm. NIIGiM (in Russian). Aghakhanyan, G. A. 1960. Development of Water Economy in Armenia. Yerevan, Haypethrat (in Armenian).

53

Systemizing the Past banasirakan handes (Historical-Philological Journal) 2: 62–86 (in Russian). Gabrielyan, H. K. 2000. Armenian Highland, Yerevan, Yerevan State University Press (in Armenian). Gasparyan, A. 1968. ‘Nork’ journal. Herald of Archives of Armenia 1: 183–192 (in Armenian). Gasparyan, B. 2010. Landscape organization and resource management in the Lower Paleolithic of Armenia. Türkiye Bilimler Akademisi Arkeoloji Dergisi 13: 159–183. Ghanalanyan, A. T. 1969. Avandapatum: Collection of Legends. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Armenian). Ghapantsyan, L. 1952. On Stone Stelae on the Mountains of Armenia. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Russian). Ghazar Parpetsi, 1982. Armenian History and Paper by Vahan Mamikonyan. Yerevan, Yerevan State University Press (in Armenian). Gilibert, A., Bobokhyan, A. and Hnila, P. 2012. Dragon stones in context: the discovery of high-altitude burial grounds with sculpted stelae in the Armenian mountains. Mitteilungen der Deutschen Orient Gesellschaft 144: 93–132. Grekyan, E. 2002. Agricultural territories and irrigation system in north-western Iran according to cuneiform sources. Patma-banasirakan handes (Historical-Philological Journal) 2: 214-232. Grigoryan, A. 1990. Tombstones with snake depictions. Mshakuyt (Culture) 2-3: 99 (in Armenian). Hakobyan, T. Kh. 1981. Historical Geography of Armenia. Yerevan, Mitq (in Armenian). Halajyan, G. 1973. The Ethnography of the Armenians of Dersim (Armenian Ethnography and Folklore 5). Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Armenian). Harutyunyan, E. A. 1971. Some Fragments from the History of the Nutrition of Armenians. Yerevan, Press of the Academy of Scoences of the Armenian SSR (in Russian). Harutyunyan, E. V. 1964. Agriculture and Cattle Breeding of Urartu. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Russian). Harutyunyan, S. B. 2000. Armenian Mythology. Beirut, Vahe Setian Press (in Armenian). Harutiunyan, S., Kalantaryan, A., Petrosyan, H., Sargsyan, G., Melkonyan, H., Hobosyan, S., Avetisyan, P. and Gasparyan, B. 2005. Wine in Armenian Traditional Culture. Yerevan, Institute of Archaeology and Ethnography, NAS RA (in Armenian). Hasratyan, M. 1974. Shvanidzor medieval canal. In Armenian Art. Ancient and Middle Ages: 25–40. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Armenian). Hewsen, R. H. 2001. Armenia: A Historical Atlas. Chicago, University of Chicago Press.

Hmayakyan, S. G. 1990. The State Religion of the Kingdom of Van. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Armenian). Hnila, P., Gilibert, A. and Bobokhyan, A. 2019. Prehistoric sacred landscapes in the high mountains: The case of the vishap stelae between Taurus and Kaukasus. In B. Engels, S. Huy and Ch. Steitler (eds), Natur und Kult in Anatolien (BYZAS 24): 283–302. Istanbul, Veröffentlichungen des Deutschen Archäologischen Instituts Istanbul. Hobosyan, S., Gasparyan, B., Harutyunyan, H., Saratikyan, A. and Amirkhanyan, A. 2021. The Armenian Culture of Grape and Wine. Yerevan, Print Box (in Armenian). Hovhannisyan, K. 1996. Irrigation canals and water supply systems. In B. Arakelyan (ed.), History of Armenian Architecture, vol. 1: 184–189. Yerevan, Publishihg House ‘Gitutyun’, NAS RA (in Armenian). Hovhannisyan, N. A., Yesayan. A. A., Bobokhyan, A. A., Dallakyan, M. V., Hobosyan, S. G. and Gasparyan, B. Z. 2017. Armenian Vine and Wine. Yerevan, Antares. Hovsepyan, R. 2015. On the agriculture and vegetal food economy of Kura-Araxes culture in the South Caucasus. Paléorient 41/1: 69–82. Hovsepyan, R. 2017. The ceremony of ‘buka barane’ (‘bride of rain’) and other rainfall-inducing ceremonies of the Yezidis of Armenia. Etchmiadzin 74/1: 114–146. Hovsepyan, R. 2018. On the rain-related rituals of Tatev village, Armenia (ethnographic field-notes)․ Lraber hasarakakan gitut‘yunneri (Herald of the Social Sciences) 3(654): 347–356. Hovsepyan, R. 2021. Seeds from vishaps’ environment: Archaeobotanical findings from the high mountainous site of Karmir Sar (Tirinkatar, Mt. Aragats, Armenia). Quaternary International 579: 19– 28. Jalalbekyan, R. L. 1974. The dams of Mokhrablur. Patmabanasirakan handes (Historical-Philological Journal) 4: 157–162 (in Armenian). Kalantar, A. 1933. Aragats in History. Yerevan, Pethrat (in Armenian). Kalantar, A. 1937. An ancient water distribution system in Soviet Armenia. Bulletin of the Institute of History and Literature of the USSR 2: 171–194 (in Armenian). Karapetyan, L. 1979. Irrigation agriculture in Armenia of the Late Bronze Age. Patma-banasirakan handes (Historical-Philological Journal) 1: 59–74 (in Russian). Khachatryan, M. M. 1971. An engraved drawing of preUrartian water distribution system in Aragatsotn. Patma-banasirakan handes (Historical-Philological Journal) 3: 128–137 (in Armenian). Khanzadyan, E. V. 1969. Garni IV (Archaeological Excavations in Armenia 12). Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Armenian).

54

Tork Dalalyan et al.: Water Management in Ancient Armenia Khanzadyan, E. V. 1978. Cultic monuments of Metsamor. In R. V. Zaryan (ed.), Second International Symposium on Armenian Art, vol. 1: 112–123. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Russian). Lalayan, E. A. 1896. Javakhk. Azgagrakan handes (Ethnographical Journal), Book 1: 117–378 (in Armenian). Lisitsyan, St. 1940. Physical Geography of the Armenian SSR. Yerevan, Haypethrat (in Armenian). Marr, N. Ya. and Smirnov, Ya. I. 1931. Vishaps. Leningrad, OGIZ (in Russian). Meshchaninov, I. 1931. Irrigation in the Chaldian period of Transcaucasia. Reports of the State Academy of History of Material Culture 6: 12–15 (in Russian). Mints, A. (ed.) 1966. Geography of the Economy of the Transcaucasian Republics. Moscow, Nauka (in Russian). Montoya, C., Bãlãşescu, A., Joanin, S., Ollivier, V., Liagre, J., Nahapetyan, S., Ghukasyan, R., Colonge, D., Gasparyan, B. and Chataigner, C. 2013. The Upper Paleolithic site of Kalavan-1 (Armenia): an Epigravettian settlement in the Lesser Caucasus. Journal of Human Evolution 65: 621–640. Nzhdehian, G. 1902. Paravashunch: heaven and earth. Azgagrakan handes (Ethnographical Journal), Book 9: 263–265 (in Armenian). Ollivier, V., Fontugne, M., Hamon, C., Decaix, A., Hatté, C. and Jalabadze, M. 2018. Neolithic water management and flooding in the Lesser Caucasus (Georgia). Quaternary Science Reviews 197: 267–287. Petrosyan, A. E. and Bobokhyan, A. A. (eds) 2015. Vishap Stones. Yerevan, Publishihg House ‘Gitutyun’, NAS RA (in Armenian). Piotrovskij, B. 1939. Vishaps: Stone Statues in the Mountains of Armenia. Leningrad, Academy of Sciences Press (in Russian).

Piotrovskij, B. 1959. Kingdom of Van, Moscow, Oriental Literature Press (in Russian). Russell, J. R. 1994. Grace from Van: a micro-historiola. Journal of the Society for Armenian Studies 7: 35–45. Saghumyan, S. 1998. Ashtarak. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Armenian). Shirmazan, G. 1961. Visitors’ impressions of the construction of Lake Ayghr water pumping station. Banber Hayastani arxivneri (Herald of Archives of Armenia) 1: 143–155 (in Armenian). Shirmazan, G. 1962. Episodes from the History of Irrigation in Armenia. Yerevan, Haypetgyughhrat (in Armenian). Shirmazan, G. V. and Darbinyan, A. A. 1961. Waterworks in Soviet Armenia (1920-1940). Banber Hayastani arxivneri (Herald of Archives of Armenia) 1: 49–56 (in Armenian). Srvandztyants, G. 1978. Works, vol. 1. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Armenian). Toporov, N. 1977. On the reflection of an Indo-European myth in the ancient Armenian tradition. Patmabanasirakan handes (Historical-Philological Journal) 3: 88–106 (in Russian). Torosyan, R. M. 1976. The Early Agricultural Settlement of Teghut. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Armenian). Vardumyan, G. 2004. The Armenian Highland as the cradle of the Armenian faith. In E. Danielyan (ed.), Armenian Highland, the Cradle of Armenian and World Civilization: 142–150. Yerevan, Zangak (in Armenian). Wittfogel, K. A. 1957. Oriental Despotism: A Comparative Study of Total Power. New Haven, Yale University Press. Yepremyan, V. 1992. Water Supply of Ancient and Medieval Armenia. Yerevan, YerAsi (in Russian).

55

The Archaeological Site of Garni, Armenia. Pre-Arsacid Archaeological Evidence and an Urartian Inscription of Argišti on a Vishap Roberto Dan

Tuscia University and ISMEO – The International Association for Mediterranean and Oriental Studies, Rome, Italy

Arsen Bobokhyan

Institute of Archaeology and Ethnography, NAS RA and Yerevan State University, Yerevan, Armenia

Onofrio Gasparro

Ca’ Foscari University of Venice, Italy

Boris Gasparyan

Institute of Archaeology and Ethnography, NAS RA and Yerevan State University, Yerevan, Armenia

Artur Petrosyan

Institute of Archaeology and Ethnography, NAS RA and Yerevan State University, Yerevan, Armenia

Mirjo Salvini

ISMEO – The International Association for Mediterranean and Oriental Studies, Rome, Italy Abstract: The present article is aimed at a general archaeological re-evaluation of the pre-Arsacid archaeological evidence in the important site of Garni in Armenia, famous for its Ionic temple. The site is mostly known for the impressive Arsacid-Roman remnants, while the site has a millennial history starting from the Early Bronze Age. These pre-Arsacid archaeological remains are presented and discussed in this text as well as the Urartian inscription identified in the site. Keywords: Armenia, Garni, Pre-Arsacid archaeology, Urartian inscription, vishap

Introduction Garni is one of the most important archaeological sites not only of modern-day Armenia, but of the entire Armenian Highlands (Figure 1).1 The site has a millennial history, although it is generally remembered as one of the most important sites of the Arsacid-Roman era in the region. This is mainly due to the presence of an Ionic temple, the foundation date of which has long been debated, which underwent an integral anastylosis between 1969 and 1975, making it the bestknown building in all of pre-Christian Armenia. On the occasion of a new documentation of the archaeological site by means of aerial photography, it was decided to carry out a new study of the archaeological evidence regarding the phases preceding the Arsacid-Roman era and also a new study of the Urartian epigraph found on the site. The peculiarity of this inscription made The introduction was written jointly by the authors. Specifically, Mirjo Salvini wrote ‘The Urartian cuneiform inscription by Argišti I engraved on the Garni Vishap’ and Arsen Bobokhyan, Boris Gasparyan, Artur Petrosyan and Roberto Dan wrote ‘An overview of the pre-Arsacid archaeological evidence’. Onofrio Gasparro wrote ‘The Geographical location of Garni’ and prepared the orthophotos and the new graphic documentation of the site. 1

Systemizing the Past (Archaeopress 2023): 56–78

by the king Argišti (I), son of Minua (CTU A 8-12) lies in the fact that it is the only one among the hundreds attributable to the Urartians to have been made on an ancient Vishap. The article is substantially divided into two parts, a first where the archaeological evidence between the Early Bronze Age and the Early to Middle Classical Period is presented, and a second which includes a new study of the Urartian inscription. The Geographical Location of Garni Garni is situated about 30 km to the east of Yerevan, on the side of Geghama Mountains, at c. 1400 m above sea level (Figure 2). The well known fortified site with the Late Roman period temple is situated in the southwestern part of the village, on a high, rectangular shaped promontory that extends northwards, with a steep slope to the south. Immediately below the site flows the River Azat, which from the foot of the western slope of the Geghama Range crosses the modern village of Garni, then Lanjazat and Arevshat, joining the Araxes near Artashat. The erosive action of the river combined with the geological characteristics of the volcanic territory has created a

Roberto Dan et al.: The Archaeological Site of Garni, Armenia

Figure 1. Map showing the location of Garni and other relevant sites in the area (satellite picture after Google Earth).

Figure 2. Map showing the archaeological features in the Garni area (satellite picture after Google Earth).

57

Systemizing the Past

Figure 3. Ortophotograph of the Garni site (Kotayk Survey Project Archive, processing by Onofrio Gasparro).

deep gorge characterized by well-preserved rock walls of basalt columns. This portion of the Garni Gorge is often referred to as the ‘Symphony of the Stones’.

large volcanic plateau with three vertical sides directly overlooking the river while on the fourth side they are connected to the edge of the gorge.

At the confluence between the Azat and one of its tributaries (which the sources call either Garni or Goght from the name of another small village located about 5 km from Garni to the north-east), just over 500 m away to the east, a second rectangular promontory overlooks the gorge. On it, the remains of a settlement protected by a wall that blocked access to the promontory are still visible. The site has been dated to the LBA/EIA on the basis of the architecture and the pottery found on the site. Both sites are located in strategic positions, on a

An Overview of the Pre-Arsacid Archaeological Evidence Garni is a well known site in Medieval Armenian sources, and has been known to European travellers since the beginning of the 19th century. Excavations in Garni were conducted in 1909-1911 under the direction of Nikolay Marr and Yakov Smirnov (Russian Archaeological Society), and from 1949 under the direction of Babken Arakelyan (Director of the Institute 58

Roberto Dan et al.: The Archaeological Site of Garni, Armenia

Figure 4. Digital Elevation Model of the Garni site (Kotayk Survey Project Archive, processing by Onofrio Gasparro).

of Archaeology and Ethnography, Academy of Sciences of the Armenian SSR).2 In Garni village and in Garni Fortress in particular there are traces of habitation since the EBA.3 There also exist the ruins of an old hydraulic system that brought water from the top of Mount Vishapasar (3157 m)4 to Garni, consisting of a series of canals and artificial lakes that collect the water

from melting snow. A. Kalantar compared this system to a similar one discovered on Mount Aragats and dated its initial construction to the pre-Urartian era.5 As part of the activities conducted in Kotayk Region by the Kotayk Survey Project (KSP), an Armenian – Italian archaeological project, aerial documentation was performed with the aim of producing updated records of this important site (Figures 3-7).

For details of the geography, history, and history of investigation of Garni see Sahinyan 1983. 3 The description of remains of pre-Hellenistic Garni is based on Khanzadyan 1969; cf. also Khanzadyan 1967: 15-16. 4 In the name of this mountain survives the word vishap (dragon stone). According to Kalantar, vishaps were connected to the ‘cult of water’.

The Early Bronze Age occupation phase is present in different parts of the Hellenistic period Garni Fortress, comprising an area of c. 3 ha in the central part, near

2

5

59

Kalantar 1994: 34.

Systemizing the Past

Figure 5. South-west/north-east section of the Garni site (Kotayk Survey Project Archive, processing by Onofrio Gasparro).

Figure 6. North-west/south-east section of the Garni site (Kotayk Survey Project Archive, processing by Onofrio Gasparro).

60

Roberto Dan et al.: The Archaeological Site of Garni, Armenia

Figure 7. Aerial view of the Garni site from south, in the background is the LBA/EIA site in the gorge (Kotayk Survey Project Archive).

period material; no architectural remains have yet been discovered; they seem to have been destroyed by later constructions. MBA pottery is present in almost all trenches (especially in Area 2), as well as near foundations and towers of the Hellenistic fortress. According to E. Khanzadyan the MBA settlement was very large, extending beyond the fortress area. A group of MBA tombs were found by chance on the north-east edge of Garni village, and another group on the road leading to Artashat. The MBA is also present on another site in front of Garni Fortress - on a high and beautiful plateau (where Azat and Garni meet) under the LBA/ EIA settlement: in particular, MBA sherds were found under the bases of LBA/EIA houses. This is considered to be the Garni ‘settlement’.7 MBA painted and black incised pottery (Figure 10) known from the mentioned places in Garni dates to the Middle (MBA II – 2200/21002000/1900 Cal BC) and Late (MBA III – 2000/19001700/1650 Cal BC) phases of the MBA (associated with the Trialeti-Vanadzor, Sevan-Artsakh\Uzerlik, KarmirBerd and Karmir Vank traditions). Late Bronze-Early Iron Age is represented within the fortress only by a small quantity of potsherds. Some scholars maintain that traces of a pre-Urartian fortress were discovered during the excavation of the third and fourth northern towers of the Garni Fortress: in particular, the remains of a wall were unearthed

the temple and under the fortress walls. The main concentration is the 2nd Trench in the centre of the fortress, with three building horizons of house remains with mud-brick walls on stone bases. The houses are rounded in plan with rectangular additions. The EBA settlement was surrounded by defensive walls with ‘cyclopean masonry’, the remains of which have been found both under the Hellenistic period defensive walls (in the northern and north-eastern parts) and also within these walls. The second horizon is well known for its metallurgical contexts revealing a complete chaîne opératoire. At the southern end of the natural rock spur, overlooking the gorge, traces of rock-cut structures are visible, which are similar to those known from Agarak, Kakavadzor, Kosh and other Kura-Araxes sites, reflecting elements of the sacred landscape (Figures 8-9).6 Finds typical of the Kura-Araxes culture such as andirons and black-burnished pottery are well represented. The finds identified during the investigations can be attributed to the 3rd phase of the Kura-Araxes I (3100/3000-2900/2800 Cal BC) and the Kura-Araxes II (2900/2800-2600/2500 Cal BC; Figure 9). Middle Bronze Age remains are present within the fortress and outside of it, within the village, where tombs have been found. The MBA in the fortress is represented only by pottery finds. As a rule these appear above EBA layers, very often mixed with Hellenistic and Medieval For the Kura-Araxes sacred landscape see Avetisyan 2006; 2008; Avetisyan et al. 2017.

Arakelyan 1951: 22. Cf. Kushnareva 1997: 117, 189; Smith et al. 2009: 68, 72, Figure 22.o, q.

6

7

61

Systemizing the Past

Figure 8. Kura-Araxes period rock-cut structures close to the southern end of the natural rock spur (photographs by Levon Mkrtchyan).

in association with EIA and LBA pottery.8 However in the main publication of the results of the Garni excavations these wall remains are interpreted as EBA. The presence of an LBA/EIA fortification system in Garni is possible for two reasons: 1. there are no clear data (stratigraphy, profiles) demonstrating that the fortification wall described by Khanzadyan belongs to the EBA, 2. Fortifications are apparently not typical for bearers of Kura-Araxes culture: it is believed that they appear in Armenia mainly at the end of the MBA and are especially typical of the LBA/EIA. The absence of 8

EIA and MIA finds in Garni Fortress can be explained by the possibility of its having been a fort-post and not a settlement, as during EBA. The period under consideration is also attested by tombs found by chance (near the fortress and on the north-east edge of Garni village) and most importantly by the above-mentioned site in front of the Garni Fortress on a high plateau (where Azat and Garni meet). The last of these was partially excavated by H. Martirosyan: it was encircled by two walls and consisted of rectangular rooms with typical LBA/EIA pottery and a few other objects. These finds tend to date to the second half of the 2nd millennium BC, while material from settlements dates

Arakelyan 1957: 12-13; Kushnareva 1997: 54.

62

Roberto Dan et al.: The Archaeological Site of Garni, Armenia

Figure 9. Selected Early Bronze Age finds from the Garni complex (1949-1966 field seasons). 1-3) Kura-Araxes I/phase 3 (3100/3000-2900/2800 Cal BC); 4-10) Kura-Araxes II (2900/2800-2600/2500 Cal BC) (assembled after Khanzadyan 1969)

63

Systemizing the Past

Figure 10.Selected Middle Bronze Age finds from the Garni complex (1949-1966 field seasons). 1-7) MBA II (2200/21002000/1900 Cal BC); 8-20) MBA III (2000/1900-1700/1650 Cal BC) (assembled after Khanzadyan 1969)

64

Roberto Dan et al.: The Archaeological Site of Garni, Armenia

Figure 11. Selected Late Bronze Age and Iron Age pottery and Mitannian stone seal from the Garni complex (1949-1966 field seasons). 1-4) Late Bronze Age (1500/1350-1250/1200 Cal BC); 5-6) Early Iron Age (1200/1150-900/850 Cal BC); 7) Late Iron Age (750/700-600/550 Cal BC) (assembled after Arakelyan 1951 and Khanzadyan 1969)

65

Systemizing the Past to the LBA and the tombs to the EIA. Also four fortresses with cemeteries are reported 2-5 km north and east of Garni, the majority of which date to the LBA/EIA.9 The finds identified in the Garni complex pertaining to these periods can be attributed to the Late Bronze Age I-II (1500/1350-1250/1200 cal BC; Figure 11/14) and the Early Iron Age (1100/1050-900/850 Cal BC; Figure 11/5-6). Particularly important is the discovery of a Mitannian cylindrical seal (Figure 11/4) of black stone and with suspension hole and a depiction of a goat in front of a tree, which was found inside Tower V of the northern fortress wall (Figure 11/4). Arakelyan proposed its resemblance to cylindrical seals from Karmir-blur, which according to the scholar have an Assyrian-Babylonian origin, although he did not exclude the possibility that the seal could also have been made by the Urartians.10

the inscription of Başbulak (CTU A 5-26), as the site of Artashat,15 which is only 22 km away from Garni.16 The location of the site close to the Araxes suggests the possibility that Argišti I (785/780-756 BC) fought against the country of Siluni in the early stages of his expansion in modern-day Armenia.17 Traces of a preUrartian fortress (Figure 8) described before could be considered to be probably part of the fortresses destroyed by Argišti I. We must underline that up to now the excavations in Garni have not brought to light any reliable evidence that might suggest a stable Urartian occupation of the fortress. In any case, the valley where Garni is located was certainly used by the Urartians because there was a route for crossing the Gegham Mountain Range, which gave easy access to the southern coast of Lake Sevan.18 This is still the path used by shepherds for transhumance.19

Urartian period. Urartian artefacts have not been found in Garni Fortress (or its surroundings) but the Urartian inscription on a vishap (see below) clearly testifies to Urartian presence. Some scholars have suggested the possible existence of an Urartian outpost underneath the Hellenistic fortress, with the possible function of ‘summer residence’. Further investigation will clarify this question. Burney has suggested the possibility that the lowest courses of the Hellenistic fortifications were laid down during Urartian times,11 but this hypothesis has not been confirmed by archaeological investigations. Other indirect data on Urartian influences are of an architectural nature. The use of rectangular towers and ‘toothed’ walls in Hellenistic Garni was first recorded in Armenia during Urartian times (Ṭušpa, Argištiḫinili, Teišebai URU). The 51.8 cm unit of measurement used during construction work in Hellenistic Garni is first recorded in Armenia during the Urartian period.12 Also the structure seen in the Palace Hall, the inner space of which is divided in two equal parts by eight massive pillars in a longitudinal direction, was first used in Urartian architecture (Teišebai URU). Often cited are certain architectural details common to Muṣaṣir and Garni temples.13 From a historical perspective contacts between the area of Garni and the Urartians were probably directly connected with the first stages of Urartian expansion in the Araxes Valley. It is likely that the first territories conquered by the Urartians were relatively close to Garni. This becomes an unavoidable conclusion if we accept the identification of the fortress of Minuaḫinili, built by king Minua (c. 810-785/780 BC)14 during his northern campaign and referred to in

Middle-Late Iron Age. Relatively few finds can be attributed to the Middle-Late Iron Age (750/700The fortress of Artashat, the capital of the Artaxiad dynasty founded by Artaxes I in 176 BC (Hewsen 1987: 653-654), was probably built on an older Urartian fortress (Smith 2003: 175; Tiratsyan 1976: 135-136). The site is spread over several hills, known by the name of Khorvirap; on hill II, the highest one, there are remains of walls dating to Urartian times, mostly ruins reused in fortifications of a later period. The most interesting structure is a long stretch of wall exposed for a length of 400 metres and preserved for a height of 6-7 metres. The wall, 2.7 metres thick, has buttresses on the outer face that project for 1 metre and are spaced regularly at a distance of 1.7 metres. The wall was built with a stone base and a mud-brick superstructure. On the corners and at the centres of the buttresses there are decorative vertical recesses, 30 cm deep, which run for the entire surviving height of the wall. It has been estimated that the Urartian fortress would have covered an overall area of 2.5-2.6 hectares, elongated in an east-west direction (Arakelyan 1982; Kanecjan 2000: 104-105; 2001: 149-150; Smith 1996: 258-260; Tonikian 1992: 162, 172-173). Its foundation has been attributed to both Minua (Biscione et al. 2002a: 10; Khachatryan 1987: 41-42) and Argišti I (Kanecjan 2000: 104; 2001: 150; Kleiss and Hauptmann 1976: 27). Although Urartian pottery has been discovered on the site (personal communication from S. Kroll), no clear traces of Urartian structures have been identified up to now (Dan 2019: 1-16). 16 There are two main theories about the possible archaeological location of Minuaḫinili: according to one it was the fortress of Karakoyunlu, undoubtedly one of the most important Urartian fortresses in the region Iğdır (Marro and Özfirat 2005: 332; Özfirat 2010: 528), while the second associates it with the Urartian-level ruins of the fortress of Artashat on the other side of the Araxes, in modernday Armenia (Yervand Grekyan 2011, personal communication). The hypothesis advanced by Piotrovskij, that Minuaḫinili lies under the medieval ruins of the city of Tsolakert, seems unlikely (Piotrovskij 1969: 21). The first of these proposals is today considered the most reasonable because it is highly probable that the two building blocks discovered in Taşburun and Başbulak came specifically from the ruins of the fortress of Karakoyunlu, just 3-4 km distant from those places, rather than from Artashat, located 30 km away to the east and on the other side of the Araxes. For a different hypothesis regarding the location of Minuaḫinili, see Dan 2020: 108. 17 On the toponyms mentioned in the Garni inscription, see Dan 2020: 70, 134. 18 On main roads that connected the Ararat Depression with Lake Sevan, see Biscione and Dan 2011: 111-112 and Dan et al. 2021. 19 Personal communication from Raffaele Biscione, 2012. 15

Arakelyan 1951: 26. Arakelyan 1951: 27, Figure 15. Reanalysis and attribution of the seal to the Mitannian type was provided by N. Yengibaryan and P. Avetisyan to whom the authors express their thanks. 11 Burney and Lang 1971: 250-251. 12 On Urartian metrology, see Sivkov 1944. 13 Sahinyan 1969: 8, 11; 1983: 32, 40, 45, 48, 161, 209. 14 All the chronological references regarding Urartian kings are taken from Salvini 2008: 23 and 2018: 18. 9

10

66

Roberto Dan et al.: The Archaeological Site of Garni, Armenia

Figure 12. Fragments of painted carinated bowls from the Garni Complex (assembled after Khachatryan 1970: Plate III, IV/2)

600/550 Cal BC). Nevertheless, this pottery testifies to the continuity of occupation of the fortress in late/ post-Urartian times and the Achaemenid period, when the Yervandid dynasty was ruling on the Armenian Highlands (Figure 11/7). Evidence of this is also a series of fragments of painted carinated bowls discovered in the site (Figure 12).

Classic/Hellenistic Period. A consistent amount of finds discovered during the investigations in the site can be attributed with certainty to the Classic/ Hellenistic Period (4th – 1st centuries BC), when the Armenian Highlands was ruled by the local Yervandid (Yervanduni) and Artaxiad (Artashesid) dynasties. Finds which mainly originate from amphora (Figure 67

Systemizing the Past

Figure 13. Selected Classical period finds from the Garni complex (1956-1972 field seasons). 1. Amphora burial A; 2-9) Jars and vase; 10) Mirror; 11-13) Spatulas for applying cosmetics; Early (Yervanduni) and Middle (Artashesid) ages (4th1st centuries BC) (assembled after Khachatryan 1976)

68

Roberto Dan et al.: The Archaeological Site of Garni, Armenia

Figure 14. Selected Classical period finds from the Garni complex (1956-1972 field seasons). 14-18) Pendant, rings and earrings; 19-27) Bracelets; Early (Yervanduni) and Middle (Artashesid) periods (4th-1st centuries BC) (assembled after Khachatryan 1976)

69

Systemizing the Past

Figure 15. Selected Classical period materials from the Garni complex (1956-1972 field seasons). 28-30) Finger rings; 3137) Necklaces; Early (Yervanduni) and Middle (Artashesid) periods (4th-1st centuries BC) (assembled after Khachatryan 1976)

70

Roberto Dan et al.: The Archaeological Site of Garni, Armenia

Figure 16. Map of Garni during excavations, with pre-Hellenistic period remains (adapted from Arakelyan 1951: Figure 19)

plate turned out to be a fragmentary vishap with an Urartian inscription (Figure 19-21) on it, which was used as a supporting stone on which one of the eight pillars stood.21

13/1), stone-box type and simple grave-cut burials consist of vessels (Figure 13/2-9), mirrors and other objects for cosmetics (Figure 13/10-13), ornaments such as pendants, rings, earrings and bracelets (Figure 14), finger rings and necklaces (Figure 15), tools, weapons, coins and other types of grave goods.20

B. Arakelyan has made three important observations regarding the discovery of the Garni vishap. 1. The presence of a cuneiform inscription on a vishap demonstrates that it is earlier than the inscription: he dates the vishap ‘c. to the mid 2nd millennium BC, not earlier than 17-16th centuries BC’ (very precise, but a priori, without supporting evidence).22 2. He insists in that the vishap was not brought down from

The Vishap was found in 1963 within the Palace Hall of the 3rd century AD, west of the Temple, partly under the ruins of a church (Figures 16-18). This rectangular hall contained three square basalt plates and a fourth rectangular one, under the pillars of the hall, all on a line running down the middle of the room. The fourth For a description of Classical/Hellenistic Period finds from Garni see Khachatryan 1976.

21

20

22

71

Arakelyan and Arutjunjan 1966: 290-291. Arakelyan and Arutjunjan 1966: 293.

Systemizing the Past

Figure 17. Garni, plan of the pillared hall, church and Hellenistic temple showing the position of the Urartian inscription (modified after Sahinyan 1983: 42, Figure 15).

Figure 18. Garni, the rectangular hall with pillars. In the foreground, the broken vishap with the Urartian inscription, put in a vertical position near the place of discovery (after Sahinyan 1983: Table 13).

72

Roberto Dan et al.: The Archaeological Site of Garni, Armenia If the vishap of Garni is in loco (which is possible according to new data on foothill vishaps) we have archaeologically visible date termini indicating a time range between the EBA and EIA. The present data on vishaps suggest that the Garni vishap probably dates to the end of the MBA, a period which has left clear evidence in Garni. It is of note that some of the painted sherds from Garni dating to the same period feature birds and some kind of ‘mythical’ animals, and wavy lines which have been juxtaposed with motives seen on vishap stones.26 The Urartian Cuneiform Inscription by Argišti I Engraved on the Garni Vishap I will begin from CTU A 8-12 (vol. I, p. 351 and vol. V, p. 219). The inscription is carved on an isolated, irregular slab of rock, formerly the upper part of an ancient repurposed vishap (Figures 20-21), which was found in the site of the ancient Armenian fortress of Garni, 20 km east of Yerevan. The upper part of the vishap seems original, because of its round shape; it is broken diagonally on the left side. It seems clear that the Urartians had found the stone in this condition, since – taking account of the damage – the inscription is carved underneath the broken left part. Height of the stone: 153 cm. Width: 79 cm at the inscription level, 70 cm at the base. Thickness: 40-42 cm. The inscription begins 24 cm from the left edge of the stone, and it reaches and continues beyond, with the sign ni in the third line, the right edge.

Figure 19 The Garni vishap with the inscription (CTU A 8-12) of Argišti I (photograph by M. Salvini, 19.7.2006).

the Gegham mountains, where it might have stood beside a stream near Garni or by a canals made for the purpose of irrigating the valleys or supplying the needs of animal-breeding for the population of Garni, which has been inhabited since the end of the 4th millennium BC.23 3. He makes an important methodological point, writing that when investigating the function, nature and purpose of stone vishaps one should restrain from quick conclusions, taking into account ‘more important continuing archaeological investigations in the areas of discoveries of stone vishaps where there are not only water reservoirs and a net of irrigation canals but also rock carvings, cemeteries and other traces of life and activities of tribes inhabiting these regions, who created not only images of stone vishaps’.24

Height of the inscription: 29 cm. Width: 55 cm. Examined on July 26th 2006; it was lit by the sun until 14:30. According to the examiners Arakelyan and Harutyunyan,27 it is indeed a reused vishap. In the paper containing the editio princeps of that text, Arakelyan describes the circumstances of the vishap’s discovery, in a room of the 3rd century palace, and provides a review of the previous literature about both this inscription and the vishap issue. Since the vishap was found not in its original location, Arakeljan paid close attention to the problem of the stone’s provenience. In the Gegham mountains area, 25/30 km east of Garni, there are plenty of this kind of monument, which are located on the banks of an ancient water basin and some irrigation canals: our vishap could have been taken from here to Garni. But he highlights that plausibly some of these vishaps might also have been located around the big springs of Garni, which flow from underneath the rocks or near the canals dug to facilitate agriculture and pastoralism.

Considering the problem of dating vishaps, Khanzadyan relies upon a more mythological approach and without archaeological grounding differentiates between three stages of vishap history: 5-4th millennium - fishes, 3rd millennium - hybrids, 2nd millennium – bulls, and ascribes the vishap of Garni to the last stage.25 Arakelyan and Arutjunjan 1966: 291. Arakelyan and Arutjunjan 1966: 293. 25 Khanzadyan 1969: 142, 161.

Martirosyan 1978: 108-110. Arakelyan and Arutjunjan 1966: 290-297, with picture of the cast and autographed copy.

23

26

24

27

73

Systemizing the Past

Figure 20. A detailed image of the inscription of Argišti I (CTU A 8-12) carved on the vishap (photograph by M. Salvini, 19.7.2006).

Figure 21. Copy of the inscription (CTU A 8-12) of Argišti I in Garni (made by M. Salvini).

It is anyway clear, thanks to the following analysis of the Urartian inscription engraved on the vishap, that the text is not necessarily connected with the Garni area, whereas it mentions some mountains, or ‘mountainous lands’ (l.5), through which the Urartian king passed when he came back from his military expedition.

I offer a different transcription from that of Harutjunjan, and also from the one I published in CTU A 8-12. Furthermore, I include here a photograph taken on July 19th 2006, at 10 a.m., and a new autographic copy. I begin with the transcription, followed by an epigraphic and philological analysis of each line. 74

Roberto Dan et al.: The Archaeological Site of Garni, Armenia D ḫal-di-ni-ni al-su-ši 1 2 [m]ar-gi-iš-ti-še a-li 3 ḫa-˹ú-bi˺ [U]RU˹’a˺-ti-ni-a-˹ni˺ m 4 si-lu-ni-i ˹KUR˺-ni-e 5 šú-ia-ni KURba-[b]a-na-ni 6 a-li bi-di-i-a 7 ʾa-še ˹MUNUS˺lu pa-ru-b[i]

the rock. I exclude the possibility of the reading URUḫix-ri?, that I previously considered possible (see CTU I p. 351, corrected in CTU V p. 219). If we put together the signs ḪI+DINGIR, they compose the sign ʼA (Labat 397), and therefore the toponym starts with this sign, that we can also distinguish in line 7. Thanks to their shape and position, in the four wedges following to the right, I recognize the sign ti, similar to that in line 2. To conclude, the name of the city conquered by Argišti I should be ʼAtinia (ʼa-ti-ni-a), with the suffix -ni which marks the direct object of the verb. Hence, I cannot confirm the reading KUR[g]i(?)-[a]r-ni-a-[ni], proposed by Harutyunyan, and so the identity with the modern toponym Garni.

Line 1: in this line there is the only exception to the spelling of al-su-ši-ni, al-su-i-ši-ni and other variations, found a hundred times in Urartian inscriptions (CTU, vol. II, p. 30 and following). Here, it depends upon a wrong evaluation in the placement of the signs, also due to the roughness of the stone, which was not smoothed previously. The ši sign is found on a sharp edge of the rock, and it is evident that there is no vertical alignment in this inscription, an uncommon instance in Urartian epigraphy. Here the lapicide allowed himself to perform what may be called an ‘abbreviation’. The same occurred in the inscription CTU A 8-11, on the rocky spur of Lchashen. In both instances, urgency due to the conditions of the military campaign is clearly noted in the written text.

Line 4: the reference to a ‘land, region’ (KUR) connected with a people’s name means that the text is talking about a territory belonging to one of the multiple tribes living in the Armenian lands. We can find an example of this in the rock inscription of Gulidžan (Qulidjan), currently called Spandaryan, about 21 km south-east of Gyumri (CTU A 8-9, ll. 5-7): mar-gi-iš-ti-še a-li / ḫa-ú-bi URU du-ru-ba-ni / mqu-u-li-a-i-ni KUR-ni, ‘Argišti says: I conquered the city of Durubani in the land of Quliaini’. In this case, we have the toponym Siluni, that we have already read in the Annals of Argišti I, carved on the stele of Surb Sahak (CTU A 8-1 Vo, 10-11): mar-gi-iš-tii-še a-li-e ḫa-a-ú-bi (11) [U]RUqi-ḫu-ni KURsi-lu-ni-ni ṣui-ni-a bi-di-e, ‘Argišti says: I conquered the city of Qiḫu in the land of Siluni, on the shore (?) of the lake (or: near the lake)’. It is possible to infer that the land of Siluni can be mentioned in two ways, with two different determinatives. Moreover, the city of Qiḫu of the land of Siluni is located close to Etiuni, as we can deduce from CTU A 8-1 Vo, ll. 3-4, and it is sited overlooking Lake Sevan, as we read in CTU A 8-1 Vo, l. 11.28

Line 3: At the right of the ḫa sign and before the rest of the bi sign I could hardly read anything, and was not sure if the sign was u or ú, even if the presence of the verb ḫaubi was rather clear. The damage to the rock impedes a definite reading of the following signs and, since the toponym, complement of the verb ḫaubi ‘I conquered’, is certainly a hapax legomenon, it is impossible to complete the sentence with the help of other texts. But this toponym is certainly a city, since I read the logogram URU, and not KUR as Harutyunyan read instead, even if the first horizontal wedge is not well accentuated. Proceeding, we can easily read the two horizontal and two vertical wedges that compose the logogram URU. Following this sign, to the right, there is a sort of small horizontal wedge that I think might be a fortuitous line in the rock: if we consider it a horizontal wedge and, at the same time, we believe that the first horizontal wedge of the sign URU is damage to the rock, the resultant sign is therefore MAR. But that reading would not make sense, since in Urartian no words beginning with the sign mar are documented. We only find the name of the land KURmar-ma-a-ni, in an inscription of king Minua (CTU A 5-9, upper side, l. 9), but it refers to a geographical area located south of the capital Tušpa, far away from the Garni region. Furthermore, we do not find in this line the determinative, and it would not be coherent with what we read at line 4. The following sign is ḫi, and then – reading on the edge of the stone and inside the holes in the rock – I recognize the four wedges which compose the sign DINGIR. Regarding the position and shape of the wedges, I notice that they are carved at a higher level than the others, clearly because the writer wanted to avoid an already extant hole in

Line 5: šuia=ni KURbab=na=ni, ‘from all the mountains’, can also be interpreted as ‘from all mountainous lands’ and, for the following explanations, we have to think about the Gegham mountains (‘Geghamskie gory’). Line 6: the lapicide had, at the end of the line, plenty of room to carve the sign di, in order to complete the word bidia=di, ‘I came back’, that we have to integrate; as in line 1, the final sign is omitted, and in line 7 we have another abbreviation. With respect to this form, see the rock inscription of Habıbuşağı, near Kömürhan, on the west shore of the River Euphrates (CTU A 9-4, ll.18-20): gu-nu-ši-ni-ni bi-di-a-di ba-ad-gu-lu-ú-bi URUsa-si-i-ni URU MAN-nu-si mḫi-la-ru-a-da-a-i a-gu-nu-ni ma-nu gu-nu-šá-a ḫa-ú-bi, that I translate ‘I came back from the battle, I assaulted the city of Sasini, royal city of Ḫilaruda, (which) was fortified (and/but) I conquered 28 See also CTU II s.v. Qiḫuni and Etiuni, referring to the texts of the king Argišti I.

75

Systemizing the Past it with a battle’. Ḫilaruda was the king of Qumaḫa, the classical Commagene.

back, for example the Karagündüz stele and the rock inscription of Yazılıtaş. In this case, as in the others, the consideration involves consequences that affect our entire understanding of historical geography. The role of a rock inscription in the geographical localization of the ancient toponyms mentioned in its text is not as crucial as it was believed to be not so long ago. The consequences of an exaggerated trust in their value as evidence could be negative. The most significant example of this consideration is the beautiful rock inscription of king Minua near the village of Yazılıtaş (CTU A 5-3), discovered in 1827 by Schulz. It is generally held to constitute a fixed point showing the position of the land of Diaueḫi, corresponding, in the scholars’ opinion, to the Daiaeni mentioned in the Assyrian texts of Shalmaneser III. However, the subsequent discovery of another rock inscription of Argišti I, near Ortakent (CTU A 8-7), on the rock in the Morevdere canyon, a tributary of the River Kura, changed the perspective. The inscription mentions, once again, the land of Diaueḫi, the position of which was actually rather far north. In conclusion, the position of the Garni inscription marks a stage in a different path connecting the plain of Yerevan to Lake Sevan’s northern shore. The other path passed through Elar, where a rock inscription of Argišti I (CTU A 8-8), reporting other partial conquests of some places in the land of Etiuni, a target already defined from the previous kings, was found and destroyed. The Lake Sevan campaign mentioned in the Garni inscription happened in the fourth year of the reign of Argišti I (785/780-756 BC), therefore in approximately 780 BC.

Line 7: the wedges which compose the determinative MUNUS are not easily legible, but the integration is necessary, otherwise, instead of the word lu-, ‘woman’, we should read UDU, ‘sheep’. We find the same abbreviation in the inscription of the temple of Ayanis, CTU A 12-1, VI, l.10: pa-ru-ú-bi LÚ MUNUSlu, where we also notice the correspondence between the signs LÚ and ʼa-še. Fundamentally, this text shows many contractions, comparable with the contractions documented in the short texts on bullae (see CTU IV passim). The epigraphical data force us to locate the target of the military campaign, briefly described in this text, in an area distant from Garni, around the northern shore of Lake Sevan. The position of the city of Qeḫuni, the conquest of which is described in an inscription on the Lchashen rock (CTU A 8-11), suggests that the location of the military campaign may have been around the north-western corner of Lake Sevan. The city of Qeḫuni was in the land of Siluni, as we read in the Annals of Argišti I, carved on a stele from the church of Surb Sahak (CTU A 8-1, Vo, ll. 10-11): ḫa-a-ú-bi [U]RUqi-ḫu-ni KURsi-luni-ni ṣu-i-ni-a bi-di-e, ‘I conquered the city of Qeḫuni, in the land of Siluni, on the shore(?) of the lake (or: around the lake)...’. About this area in the Urartian period, see a specific paper concerning Lake Sevan.29 We may add to the documents previously published this Garni inscription – which, in my opinion, was written during the same military campaign that was described in the Annals of Argišti I. With the conquest of the cities of Qeḫuni and ʼAtinia, in the land of Siluni, the Urartians seem to have completed their occupation of the land of Etiuni, which covered the majority of Armenian territory.

Translation ‘Thanks to (or: through) the greatness of Ḫaldi, Argišti says: I conquered the city of ʼAtinia, of the land of Siluni; from all the mountains (or: from all mountainous lands), coming back (or: I came back and) I deported men and women’.

It is important to notice that, even from a strictly epigraphical point of view, the inscriptions on the rock of Garni (CTU A 8-12) and Lchashen (CTU A 8-11) have common features: the writing is unsteady and irregular, due to the roughness of the rock. Furthermore, these inscriptions refer to two particular conquests that occurred in the same area, in the north of Lake Sevan. Above all, the Garni inscription seems to have been carved on the rock during the return (bi-di-i-a) from the military campaign, far away from the scene of events: this fact also stands in the way of the attempt to identify in the modern name Garni the toponym of the place mentioned in the inscription, even if this hypothesis is tempting. I have had the opportunity to propose similar considerations regarding other texts referring to military campaigns, but in different circumstances – on the way to the target or on the way 29

Bibliography Arakelyan, B. N. 1951. Garni I. The Results of the Works of the Archaeological Expedition of the Institute of History of the Academy of Sciences of the Armenian SSR during the Years 1949-1950 (Archaeological Excavations in Armenia 3). Yerevan, Armenian SSR Academy of Sciences Publishing House (in Russian). Arakelyan, B. N. 1957. Garni II. The Results of the Works of the Garni Archaeological Expedition during the Years 1951-1955 (Classical Period) (Archaeological Excavations in Armenia 7). Yerevan, Armenian SSR Academy of Sciences Publishing House (in Russian). Arakelyan, B. N. 1982. Artashat I. Main Results of 1970-1977 Years of Excavations (Archaeological Excavations in Armenia 16). Yerevan, Armenian SSR Academy of Sciences Publishing House (in Russian).

Salvini 2002: 40-45.

76

Roberto Dan et al.: The Archaeological Site of Garni, Armenia Arakelyan, B. N. and Arutjunjan (Harutyunyan) N. V. 1966. Urartian cuneiform inscription from Garni. Patma-banasirakan handes (Historical-Philological Journal) 2(33): 290-297 (in Russian). Arakelyan, B. (ed.), 1987. Scientific Conference devoted to the Archaeological Fieldwork Results in 1985-1986 in the Armenian SSR to the 80 Anniversary of Famous Archaeologist Prof. K. Ghafadaryan (Abstracts of the Reports, 18-20 March, 1987). Yerevan, Armenian SSR Academy of Sciences Publishing House (in Armenian). Avetisyan, P. 2006. Sacred Landscape: Characteristics of Agarak Cult Constructions. In R. Büttner and J. Peltz (eds), Mythical Landscapes Then and Now. The Mystification of Landscapes in Search for National Identity: 256–277. Yerevan, Antares Publishing House (in Russian). Avetisyan, P. 2008. Recently found archaeological sites in Armenia. Aramazd: Armenian Journal of Near Eastern Studies 3/2: 39–50, 194–195. Avetisyan, H., Gnuni, A., Mkrtchyan, L. and Bobokhyan, A. 2017. An attempt at archaeological typologization of some monuments connected with the function of transition from the Bronze to the Iron Age in Transcaucasia. Epochs. Edition of the Department of History of 'St. Cyril and St. Methodius' University of Veliko Tarnovo 25/2: 316–332 (in Russian). Biscione, R., Hmayakyan, S. and Parmegiani, N. 2002a. Armenian - Italian archaeological survey. Campaigns 1994-2000. In R. Biscione, S. Hmayakyan and N. Parmegiani (eds), The North-Eastern Frontier. Urartians and non-Urartians in the Sevan Lake Basin. I, The Southern Shores (Documenta Asiana 7): 9–18. Rome, CNR - ICEVO. Biscione, R., Hmayakyan, S., Parmegiani, N. (eds) 2002b. The North-Eastern Frontier. Urartians and nonUrartians in the Sevan Lake Basin. I, The Southern Shores (Documenta Asiana VII). Rome, CNR - ICEVO. Biscione, R. and Dan, R. 2011. Dimensional and geographical distribution of the Urartian fortifications in the Republic of Armenia. Aramazd: Armenian Journal of Near Eastern Studies 6/2: 104– 120. Bobokhyan, A. and Badalyan, M. (eds) 2021. The Archaeological Heritage of the Land Uduri-Etiuni. To the Memory of Armenian Famous Archaeologist and Intellectual Ashkharbek Kalantar. Yerevan, Institute of Archaeology and Ethnography Press. Burney, Ch. A. and Lang, D. M. 1971. The People of the Hill. Ancient Ararat and Caucasus, London, Weidenfeld & Nicolson. Büttner, R. and Peltz, J. (eds) 2006. Mythical Landscapes Then and Now. The Mystification of Landscapes in Search for National Identity. Yerevan, ‘Antares’ Publishing House. CTU = Salvini 2008-2018. Dan, R. 2019. Between Urartian and Achaemenid Architectural Traditions: Considerations on the so-

called ‘Urartian wall’ of Artašat. Iranica Antiqua 54: 1–16. Dan, R. 2020. A Study on the Toponomastic of the Kingdom of Bia/Urartu (Serie Orientale Roma Nuova Serie 19). Roma, Scienze e Lettere. Dan, R., Petrosyan, A., Vitolo, P. and Gasparyan, B. 2021. From the Ararat Depression to Lake Sevan. An analysis of Urartian military roads in the territory of modern-day Armenia (late 9th to mid-8th centuries BC). In Bobokhyan, A. and Badalyan, M. (eds), The Archaeological Heritage of the Land Uduri-Etiuni, To the Memory of Armenian Famous Archaeologist and Intellectual Ashkharbek Kalantar: 288–300. Yerevan, Institute of Archaeology and Ethnography Press. Hewsen, R. H. 1987. Artaxata. In E. Yarshater (ed.), Encyclopaedia Iranica, vol. 2: 653-654. London, Boston, and Henley, Routledge & Kegan Paul. Kalantar, A. 1994. An ancient irrigation system in Armenia. In G. Karakhanian (ed.), Ashkharbek Kalantar, Armenia from the Stone Age to the Middle Ages, Selected Papers (Civilizations Du Proche-Orient, Serie I, Archaeologie Et Environnement 2): 29–35. Neuchâtel and Paris, Recherches et Publications. Kanecjan, A. G. 2000. Beziehungen zwischen urartäischer und achaimenidischer Baukunst in Armenien. Archäologische Mitteilungen aus Iran und Turan 32: 103–110. Kanetsyan, A. 2001. Urartian and Early Achaemenid palaces in Armenia. In I. Nielsen (ed.), The Royal Palace Institution in the First Millennium BC, Regional Development and Cultural Interchange between East and West (Monographs of the Danish Institute at Athens 4): 145–153. Aarhus, Aarhus University Press. Karakhanian, G. (ed.) 1994. Ashkharbek Kalantar, Armenia from the Stone Age to the Middle Ages, Selected Papers (Civilizations Du Proche-Orient, Serie I, Archaeologie Et Environnement 2). Neuchâtel and Paris, Recherches et Publications. Khachatryan, Zh. D. 1970. A typical style of 7th-1st centuries BC pottery of Armenia. Patma-banasirakan handes (Historical-Philological Journal) 1: 269–278 (in Armenian). Khachatryan, Zh. D. 1976. Garni V, Classical Period Necropolis (The Results of the Works during the Years 1951-1955) (Archaeological Excavations in Armenia 15). Yerevan, Armenian SSR Academy of Sciences Publishing House (in Russian). Khachatryan, Zh. D. 1987. Excavations of Artashat. In B. Arakelyan (ed.), Scientific Conference devoted to the Archaeological Fieldwork Results in 1985-1986 in the Armenian SSR to the 80 Anniversary of Famous Archaeologist Prof. K. Ghafadaryan (Abstracts of the Reports, 18-20 March, 1987): 40–42. Yerevan, Armenian SSR Academy of Sciences Publishing House (in Armenian). Khanzadyan, E. V. 1967. The Culture of the Armenian Highlands during the 3rd Millennium BC. Yerevan, 77

Systemizing the Past Armenian SSR Academy of Sciences Publishing House (in Armenian). Khanzadyan, E. V. 1969. Garni IV, The Results of the Excavations, Years 1949-1966 (Including from the Early Bronze Age to the Uraratian Era) (Archaeological Excavations in Armenia 12). Yerevan, Armenian SSR Academy of Sciences Publishing House (in Armenian). Kleiss, W. and Hauptmann, H. 1976. Topographische Karte von Urarṭu (Archäologische Mitteilungen aus Iran, Ergänzungsband 3). Berlin, Dietrich Reimer Verlag Berlin. Kushnareva, K. Kh. 1997. Southern Caucasus in Prehistory, Stages of Cultural and Socioeconomic Development from the Eight to the Second Millennium B.C. Philadelphia, Published by the University Museum, University of Pennsylvania. Marro, C. and Özfirat, A., 2005. Pre-classical survey in Eastern Turkey. Third preliminary report: Doğubayazıt and the eastern shore of Lake Van. Anatolia Antiqua 13: 319–356. Martirosyan, H. 1978. The Science Begins in Primitive Society. Yerevan, Sovetakan Grogh Publishing House (in Armenian). Matthiae, P., Pinnock, F., Nigro, L. and Marchetti, N. (eds) 2010, Proceedings of the 6th International Congress of the Archaeology of the Ancient Near East 5 May - 10 May 2009, "Sapienza", Università di Roma, Volume 2: Excavations, Surveys and Restorations: Reports on Recent Field Archaeology in the Near East. Wiesbaden, Harrassowitz Verlag. Nielsen, I. (ed.) 2001. The Royal Palace Institution in the First Millennium BC, Regional Development and Cultural Interchange between East and West (Monographs of the Danish Institute at Athens 4). Aarhus, Aarhus University Press. Özfirat, A. 2010. Archaeological investigations in the Mt. Ağrı region: Bronze and Iron Ages. In Matthiae, P., Pinnock, F., Nigro, L. and Marchetti, N. (eds), Proceedings of the 6th International Congress of the Archaeology of the Ancient Near East 5 May - 10 May 2009, »Sapienza«, Università di Roma, Volume 2: Excavations, Surveys and Restorations: Reports on Recent Field Archaeology in the Near East: 525–538. Wiesbaden, Harrassowitz Verlag. Piotrovskij, B. B. 1969. The Ancient Civilization of Urartu. Geneva, Nagel.

Sahinyan, A. A. 1969. Architectural Monuments of Garni and Geghard. Yerevan, ‘Hayastan’ Publishing House (in Armenian). Sahinyan, A. A. 1983. Architecture of Antique Buildings of Garni. Yerevan, Armenian SSR Academy of Sciences Publishing House (in Armenian). Salvini, M. 2002. The historical geography of the Sevan Region in the Urartian Period. In R. Biscione, S. Hmayakyan and N. Parmegiani (eds), The NorthEastern Frontier. Urartian and Non-Urartians in the Sevan Lake Basin. Volume I: The Southern Part (Documenta Asiana 7): 30–60. Roma, CNR - ICEVO. Salvini, M. 2008. Corpus dei Testi Urartei, Vol. I, Le iscrizioni su pietra e roccia (Documenta Asiana 8/1). Roma, CNR - ICEVO. Salvini, M., 2012. Corpus dei Testi Urartei, Vol. IV, Iscrizioni su bronzi, argilla e altri supporti nuove iscrizioni su pietra paleografia generale (Documenta Asiana 8/4). Roma, CNR - ICEVO. Salvini, M. 2018. Corpus dei Testi Urartei, Vol. V, Revisione delle epigrafi e nuovi testi su pietra e roccia (CTU A). Dizionario urarteo. Schizzo grammaticale della lingua urartea. Paris, De Boccard. Sivkov, A. V. 1944. About the main linear measures of Urartu and Ancient Armenia. Proceedings of the Academy of Sciences of Armenian SSR 1-2: 83–88 (in Russian). Smith, A. T. 1996. Imperial Archipelago: The Making of the Urartian Landscape in Southern Transcaucasia. Unpublished PhD thesis, University of Arizona. Smith, A. T. 2003. The Political Landscape. Costellations of Authority in Early Complex Polities. Berkeley, Los Angeles and London, University of California Press. Smith, A. T., Badalyan, R. S. and Avetisyan, P. S. 2009. The Archaeology and Geography of Ancient Transcaucasian Societies I: The Foundations of Research and Regional Survey in the Tsaghkahovit Plain, Armenia (Oriental Institute Publications 134). Chicago, Oriental Institute of the University of Chicago. Tiratsyan, G. 1976. Artashat. In V. H. Hambardzumyan (ed.), Armenian Soviet Encyclopedia, Volume II: Argishti – Gegharvan: 135–136. Yerevan, Academy of Sciences of the Armenian SSR, Armenian Soviet Encyclopedia’s Main Board Publishing House (in Armenian). Tonikian, A. 1992. The layout of Artashat and its historical development. Mesopotamia 27: 161–187.

78

The Kurtan Belt Ruben Davtyan

Martin Luther University Halle-Wittenberg; Max Planck Institute for Social Anthropology, Halle, Germany

Michael Herles

Ludwig Maximilian University of Munich, Germany Abstract: The aim of this article is the discourse on an unpublished bronze belt originating from a tomb in the village of Kurtan in northern Armenia. The edges of the front side of the richly ornamented belt are embellished with geometric patterns. The patterns connect in the shape of a frame. The central scene inside the smaller frame comprises depiction of six cantering horses from the right side. The mentioned images are comparable with a number of belts found in the South Caucasus. However, some important details (interruption of mane, structure of forelegs, and articulation of the left ear) reveal close stylistic connection with artistic objects from the necropolis of Lori Berd. The Kurtan belt was certainly a prestigious and high-valued object to the Iron Age community of Kurtan. Keywords: Iron Age. Kurtan. belt. iconography. horse. geometrical patterns. mountings

Introduction The village of Kurtan is situated in the province of Lori (Armenia), on the riverbank of Gargar. The landscape of the village is characterised by gorges created by Gargar and its inflows. The territory of Kurtan is rich with archaeological sites, examined by the authors of the given article in 2016. Archaeological Sites in Kurtan H. Martirosyan reported a necropolis with a clear row of cromlechs in the southern part of the village.1 He investigated metal objects such as three iron spear heads, iron sickle, bracelets, horse bit and iron axe coming from a destructed tomb, and dated them to the 6th-5th century BC.2 A part of the necropolis was examined in the northwestern part of the village (Figure 1). The excavations in 2004 led by S. Devedǰyan revealed four tombs. Tombs 1-3 possessed adjacent cromlechs around the cist chambers. A male skeleton was found in tomb 1, whereas no skeleton was detected in the others. All three date back to the 11th-9th century BC.3 Tomb 4 with its cromlech lied c. 40 m south from others. The chamber type was a soil pit, where a 50-55 years old man was buried. The accompanying pottery unveiled that the tomb dates back to the 15th century BC.4 The fortified settlement of Kurtan is located on an elevated platform, on the other side of the excavated Martirosyan 1964: 289. Martirosyan 1964: 289-291, Plate XXXII․ 3 Devejyan and Hobosyan 2008: 113-116. 4 Devejyan and Hobosyan 2008: 116-117. 1 2

Systemizing the Past (Archaeopress 2023): 79–87

part of necropolis, in an area between the river Gargar and its tributary Arzakan. The site is marked with its cycloptic war structures. A joint Armenian-Russian team worked here from 2006 to 2007. They focused on the citadel parts and the main settlement.5 Excavations resulted in revealing two main periods of occupation: the first refers to the Late Bronze Age and the second to the 7th-6th century BC.6 In 2016, we collected pottery sherds from the territory of the settlement, dated them to Late Bronze and Early Iron ages, corresponding with the dating in excavation reports.7 Finally, a row of pottery and metallic objects were found during construction works in the northern part of the village. They are kept in the Kurtan village museum, where objects of archaeological and ethnographic  value are exhibited. The village was visited during the mentioned survey campaign. The bronze belt documented in the museum is the object of study of this article.8 Tomb with the Belt (Figure 2) Throughout the construction of the new school in Kurtan, a tomb with big covering stones was unveiled. Though no systematic excavations were conducted, some information was recorded: the chamber was a soil Aslanyan and Vasilev 2008: 380-381. Aslanyan and Vasilev 2012: 156-157. 7 Herles and Davtyan 2017: 29, Figure 10. 8 We would like to express our sincere gratitude to P. Avetisyan, who supported us in organizing the Lori Project and as well as the restoration of the belt. Our sincere thanks also go to H. Kirakosyan, the head of the Kurtan museum for allowing us to examine the belt and for giving valuable information about its context. 5 6

Systemizing the Past

Figure 1. Map of Kurtan’s archaeological sites (L. Mkrtchyan).

pit in a length of more than 1.5 m and depth of more than a meter. The human skeleton was preserved. From the burial goods a jug, a spear and a bronze belt which are kept in Kurtan museum are ascribed to the tomb. Description (Figure 3) The Kurtan belt represents a thin bronze sheet of 76.9 cm in length, 11 cm in width and 0.1-0.2 cm in thickness. The sheet is rounded at corner edges. Both edges are furnished with three perforations. According to the

Figure 2. Photo of the tomb with bronze belt (Museum of Kurtan).

Figure 3. Photo of the belt (V. Hakobyan).

80

Ruben Davtyan and Michael Herles: The Kurtan Belt

Figure 4. Drawing of the belt (A. Sahakyan).

next to each other. The quadrupeds’ contours are almost identical: they show elongated legs ending in hoofs, thin torsos and relatively abruptly descending heads. Hind legs are stretched, whereas the front legs are spread out in different sides. The shapes of the figures, hoofs as well as the presence of mane on the necks and tapered ears leave no doubt for identification of those figures as horses.

typology of R. Virchov and J. de Morgan the belt can be classified into the category of wide belts.9 Traces of mountings, most likely at some time after its final production, can be observed on the bottom of the sheet. Here, three bronze strips of c. 1 cm in width are attached to the belt sheet by two rivets respectively. The fittings stand out on the flat surface of the belt and represent a result of a secondary activity. The level of preservation is almost exceptional. Except for some small fragments, the belt is complete.

All six horses are in a motion of galloping to the right. They have small differences in fillings. The dotted fillings are prevalent: they comprise most of the surface of horses’ bodies. Circular arrangement of dots can be seen on the shoulders of horses 1-310 (Figures 6-8) and 5 (Figure 10).

Design (Figure 4) The belt is richly incised with different motifs. The edge is ornamented with a dotted line followed by two rows with hemispheric motifs. These lines stop at the area of perforations which are symmetrically connected through engravings of vertical net-like stipe.

Hoofs are not filled. Heads and torsos (backs) are marked differently. In case of horses 4-6 (Figures 9-11) the area of heads is not filled, but marked with dotted edge lines. Horse 1’s head is filled with dots, whereas those of horses 2-3 show diagonal lines. Heads of horses 2, 3, 4 and 6 appear to be marked with small notches at the bottom right. In case of horse 2, it can be seen on both sides. No facial features are indicated.

The upper and lower parts of the inner side are framed by multiple triangular motifs. The same motif covers three chevron patterns slid in each other and engraved at both ends. The chevron patterns have net-like filling and end with triangular motifs. Wave-like and dotted lines are placed between them.

Linear patterns can be observed at the torsos (backs) of horses 3-6. Also, the torso of horse 1 is only partially filled.

In the centre of the belt, there is a rectangle with triangular motifs at the corners. The edges are filled with herringbone patterns. The rectangular is framed with dotted lines. Above the rectangle, a line with hemispherical motifs is engraved, similar to the rim decoration of the belt. The same line is engraved below the rectangle.

Except for horse 5, all others show one or two tiny stripes on their left ear. Differences appear also in the way of depiction of tails, which end either in rhombi or have angular shape. The tails of horses 2, 3, 4, 6 are also marked with herringbone patterns, which symbolise horsehair on the tail.

The main scene of the belt is depicted inside the described rectangle: it is a row of six figures depicted 9

Yesayan 1984: 203.

10

81

The numeration is from left to right.

Systemizing the Past

Figure 5. Triangular pattern. Figure 6. Horse 1.

Short diagonal notches are depicted on the neck and signify horse mane. It is noteworthy that these notches run from the top of the neck to its end and after a short interruption continue at the withers of horses. The Kurtan Belt in the Context of South Caucasian Belts The iconography of the Kurtan belt can be conditionally divided into rim motifs, lateral triangles and the central scene. This composition coincides with the South Caucasian tradition of belt design. Most of the ornamented belts show geometrical rim motifs. Decorating the rims was most likely an established standard of belt designing. More elaborated belts with figurative images are marked with lateral triangles (Figure 5). This can be observed at several belts from Northern Armenia and Georgia.11 With their cross-shaped tip, the triangles on the Kurtan belt resemble one of the belts from Samtavro.12 Figure 7. Horse 2.

In the open-up position of belts, the triangles flank the main scene and point to the main scene. While bound, the triangles come together and create a lozenge.

Horses running in the same direction seldomly occur as a central scene. Belts with rows of similar animals are known from Tli,13 Samtavro,14 Agdashdyuzi.15

The rectangular frame around the horse row is particularly typical for the Kurtan belt. The matter is that rectangular fields filled with animal rows or geometrical patterns are known from other examples, however, those show mostly secondary scenes. In our case, this frame articulates the central scene and enhances its visual appeal.

In contrast to most of the other figurative belts where the figures extensively fill in the surface, the Kurtan belt has blank spaces. The latter and rows of running animals are very typical for Urartian belt’s design. Urartian influence on South Caucasian belts was also Tekhov 1980: Plate 97.10. Chidašeli 1986: nos 2, 8. 15 Geyusheva 2009: 172. 13

See Chidašeli 1986: 8, 16 etc.; Yesayan 1984: nos 40, 49, 61. 12 Chidašeli 1986: no. 1. 11

14

82

Ruben Davtyan and Michael Herles: The Kurtan Belt

Figure 8. Horse 3.

Figure 10. Horse 5.

Figure 9. Horse 4.

Figure 11. Horse 6.

suggested for pointed and interlaced ornaments.16 Some Urartian impact on the arrangement of motifs at the Kurtan belt should not be excluded. However, the design of motifs and patterns goes entirely in line with long standing traditions of South Caucasian art.

as Lori Berd,18 Odzun,19 Musieri,20 Sanahin,21 Lorut,22 Lshashen,23 and Astghi Blur.24 In Georgia images of horses are known from Tli,25 Narekvavi, Samtavro,26 Mtskheta.27 In Azerbaijan it is

Horse Iconography on South Caucasian Belts The depiction of separate horses were identified on a number of South Caucasian belts.17 In Armenia they were found in the north and north east of Armenia such

16 17

Devejyan 1981: Plate XVIII. Yesayan 1984: no. 40. Castellucia 2017: no. 304. 21 Martirosyan 1964: Figure 58. 22 Yesayan and Avetyan 1981: Figure 1. 23 Yesayan and Mnatsakanyan 1977: Figure 1. 24 Yesayan 1967: Plate II. 25 Chidašeli 1986: 10; Tekhov 1980: Plate 97.10. 26 Chidašeli 1986: no. 8. 27 Chidašeli 1986: no. 20. 18 19 20

Yesayan 1984: 108. Castellucia 2017: 56-57, Figure 73.

83

Systemizing the Past evidenced in Khachbulag.28 Other images show human figures on the horseback.29 Stylistically relatively similar engraved image of horse or dog occurs on the bronze belt found accidently in Sochi, which was rightly compared with the images from Armenia and Georgia.30 That no concrete parallels to Kurtan horses were found is not much surprising, since each belt has its own features. Images of horses show different shapes. Yet most of them show curved shapes of the body and relatively short legs, which are not characteristic for the horses in our case. However, some other features are well comparable. The contour of the horse bodies, especially the heads are similar to those from Tli. Moreover, some of the horses are completely or partially filled with dots (e.g. Lori Berd, Astkhi blur, Odzun, Samtavro, Tli). Except for that, the belts from Lori Berd and Astghi Blur correspond with tapered ears, short tail and mane. The position of legs cannot be found on any other belt: the horses are in a running motion.

Figure 12. Bone inlay from tomb 63 in Lori Berd (R. Davtyan).

a special horse breeding. The numerous horse bones discovered in Lori Berd confirm that the region used to be a centre of horse breeding at least during the Iron Age.35 However, the examples are restricted to Lori Berd and Kurtan, and the data is highly insufficient for assuming that there was an independent horse breed.

Horse Iconography on Other Archaeological Objects The horse images closest to the one on the Kurtan belt, which can be put in parallel to the mentioned, originate from Treli and Lori Berd.

It is more likely that the discussed feature was a result of horse mane styling which corresponded to the aesthetic standards of the given time.

A bronze buckle of a rider found in tomb 63 at Treli shows a horse with waisted body and head. It dates to the end of 2nd millennium BC.31

As mentioned before, almost all six horses on the belt have forelegs marked with hocks rather than knees, which the front legs of a horse are expected to have. This ‘mistake’ can be found in many horse images on South Caucasian belts, as well as on the buckle (or standard) found in tomb 63-2 in Lori Berd.36

Geographically and chronologically to Kurtan is Lori Berd which lies only 13.5 km (linear distance) to the west far from Kurtan. The excavations of richly furnished tomb 63 belonging to the 8th-7th centuries32 BC resulted in finding a row of bone inlays of different motifs. Four fragmentary preserved inlay elements are carved in the form of horses (Figure 12). The main features such as the thin back, somewhat suspended state of the head and especially the position of legs correspond with those from the Kurtan belt.

Interpretation of Motifs One might go no wrong in culturally attributing the bronze belt of Kurtan to South Caucasian belts. However, many details show differences and as the saying goes, ‘the devil is in the details.’ For instance, relatively enigmatic features represent the stripes on the left ears. For some of them, it seems even the stripes wave while horses gallop. This is not characteristic for any other horse images known to us. It is relatively hard to explain their meaning. The stripes can be a sign of belonging to a person or group of people. Alternatively, they can simply denote decorations.

Despite the interruption of mane at horse withers is a relatively small detail, it is evidenced at all six horses. This feature can be also observed at a ferrule from tombs and at bits from tomb 64 in Lori Berd with double and triple horse heads.33 The connection between mane depiction manners in both sites is striking. Also, the horse bit with horse-head cheekpiece found in Sarychoban and dated back to 12th/11th century BC share the same characteristic.34 It might indicate

Anyhow, we would like to draw attention to the fact that this feature is typical solemnly for the left ear. The point is that the left ears of the horses are also marked at the horse bits and other horse depictions from Lori Berd. Although in the case of Kurtan the ears are rather

Yesayan 1984: no. 47. Castellucia 2017: 67-70. 30 Catalogue St. Petersburg – Berlin: 2020: no. 206.2. 31 Abramishvili 1978: color Plate VII. 32 Information on the tomb is published in Devejyan 2007: 138. 33 Catalogue History Museum of Armenia 2018: 166. 34 Dzhafarov 1993: 203-204, Figure 7.1. 28 29

35 36

84

Mezhlumyan 1991: 45-46, Devejyan 2006: 143-144. Devejyan and Davtyan 2017: Figure 1.2; 2.2.

Ruben Davtyan and Michael Herles: The Kurtan Belt

Figure 13. The reverse of the belt sheet and the metal strip.

Mares with a foal are depicted in a concentric circle on the belt from Chabaruchi and stallions with circles or rosettes.39 Frequently, they are shown in the form of solar motifs and occur on the same body parts of the animals like on the belt discussed.

decorated than cut, the special attention paid to the left ear can be found on the items of both sites. In this sense, it is important to mention that differentiated horse ear marks have been detected in kurgans 1 and 5 at Pazyryk. Rudenko suggested that they are marks of ownership and show that they originated in diverse herds of animals from different areas but had become part of one herd belonging to the deceased during his lifetime.37 In case of Kurtan, however, the ear marks are similar, which might be an evidence they originate from one herd.

These are mostly horses or quadruped animals sharing morphological features with horses.40 The existence of sun discs on horses and their semantic connection on South Caucasian belts was mentioned by Israelyan.41 She postulated her theory with a comparative  study, encompassing the fields of archaeology (e.g. the Trundholm sun chariot), historiography (records of Xenophon about horse scarification for Mithra) and folkloristics (fire-horses in Armenian tales).42 The iconographic link between the horses and the sun is also mentioned by Yesayan and Chidašeli.43 At this point it should be mentioned that the very word of belt is linguistically connected with words like ‘(rain) bow’, ‘arc’, ‘firmament’ in Ancient Armenian, Georgian and Iranian languages. The linguistic connection associating belts with the sky is an important argument for ascribing cultic meaning to them.44

The design of torsos differs. The vertical lines are interrupted by dots, which, in most cases, fill the rest of horse bodies. It should be considered whether the linear areas signify saddles or horse clothing. The point is that no image of horses or horsemen on South Caucasian belts show any clear depiction of a saddle or any other piece of horse harness. The riders on the silver goblet from tomb 56-2 in Lori Berd clearly indicate rug appliances.38 The absence of comparative images on other South Caucasian belts makes the identification of the mentioned motifs as saddles somewhat complicated, nevertheless, the possibility cannot be excluded.

In this context it can be argued that the circular arrangement of the dots on the Kurtan belt was intentional and can refer to sun motifs, as it is the case on other South Caucasian belts.

The circular arrangement of the filling dots on horse hips and shoulders is striking. There are other belts known to us which contain animal body outlines with circular motifs inside them.

Chidašeli 1986: 36, Plate 9.14. See Castelluccia 2017: 66. Israyelyan 1967: 87. 42 Israyelyan 1967: 87-88. 43 Chidašeli 1986: 36; Yesayan 1984: 124. 44 Piotrovskij 1949: 95. 39 40 41

37 38

Rudenko 1953: 147-148. Devejyan 2010: Plate XVI.2.

85

Systemizing the Past Georgia. Some details reveal links to the well-known necropolis of Lori Berd.

Metal Strips Three metal strips are riveted to the bottom edge of the Kurtan belt. A closer look will unveil vertical slots on the edges which the strips cover (Figure 13).45

Although the exact finding circumstances are unknown, it is very likely that the belt was a burial good in a rich burial. Undoubtedly, it belonged to a member of the local elite in the ancient community of Kurtan, it was an object of high value and of particular importance in regard to its representative aspect.

The mounting measures are known from metal belts found in tomb 1 at Shikahogh,46 in tombs 41 and 49 at Beshtasheni47 and in Stepantsminda.48 Also, many of the Urartian belts were restored through metal strips.49 One of the belt sheets seems to have been torn apart and later mended with metal strips and rivets over its whole width.50

The Kurtan belt contributes to the group of figurative belts coming from the Lori province and the adjusting areas. Moreover, it shows new aspects of the skills of artists of the given time as well as the taste of the Iron Age community in Lori.

Most of the strips are attached to the bottom rim of the belts (such as in case of the Kurtan belt). Perhaps, the lower edge tends to obtain defective slots more easily due to wearing.

Bibliography Abramishvili, R. 1978. Tbilissi. Sites Archaéologiques I. Tbilisi, Metsniereba. (in Georgian). Aslanyan, S. A. and Vasilev, S. A. 2008. Elements of urbanization in northern Armenia during the Urartian period (excavations of the Kurtan settlement). In A. P. Derevyanko and N. A. Makarov (eds), Trudy Vserossiyskogo arxeologič‘eskogo sezda v Suzdale II (XVIII): 380-381. Moscow, Institute of Archaeology (in Russian). Aslanyan, S. A. and Vasilev, S. A. 2012. Excavations of the Kurtan settlement in northern Armenia. In The Latest Archaeological Discoveries in the North Caucasus: Researches & Interpretations. The XXVII-th Krupnov’s Readings. Proceedings of the International Scientific Conference, Makhachkala, April 23-28, 2012: 156-157. Maxačkala, Mavraev’ (in Russian). Catalogue of History Museum of Armenia 2018. Urartu. The kingdom of Van: Findings from Urartian sites of Armenia (860-585 BC), Yerevan, History Museum of Armenia. Catalogue of St. Petersburg – Berlin 2020. Ausstellungskatalog „Eisenzeit. Europa ohne Grenzen. 1. Jahrtausend v. Chr.“ Sankt Petersburg, Staatliche Eremitage, vom 10. November 2020 bis zum 28. Februar 2021 Moskau, Staatliches Historisches Museum, vom 15. April bis zum 15. Juli 2021. St. Petersburg, Tabula Rasa. Castelluccia, M. 2017. Transcaucasian Bronze Belts. Oxford, BAR Publishing. Chidašeli, M. S. 1986. Die Gürtelbleche der älteren Eisenzeit in Georgien. Beiträge zur Allgemeinen und Vergleichenden Archäologie 8: 7–72. Devedjyan, S. H. 1981. Lori Berd 1. Results of the Excavations from 1969 to 1973. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Russian). Devejyan, S. H. 2006. Lori Berd II. Bronz Moyen. Yerevan, Gitutyun (in Armenian and in French). Devejyan, S. H. 2007. Lori Berd. In A. Kalantaryan (ed.), The Gold of Ancient Armenia (III millennium BC – 14 century AD): 135–150. Yerevan, Publishing House of the Armenian Academy of Sciences (in Armenian).

Many examples of belt restoration measures reveal that these objects were treated in a special way. These examples also show the particular importance of the belts for their owners and suggest that they were worn by more than one generation. Dating Bronze belts were used in the South Caucasus since the beginning of the second half of the 2nd millennium BC and continued to be in use in the Achaemenid period. The majority of the figurative ornamented belts dates back to the first centuries of 1st millennium BC.51 The ones from Northern Armenia are considered to originate from 9th to 7th/6th century BC.52 The tomb inventory found with the bronze belt and shown on the excavation photo can date back to 9th/8th-6th century BC. The mentioned iconographical comparisons with the objects from Lori Berd dating from 8th to the first half of the 7th century BC is obvious. Thus, we find the dating of the Kurtan belt to 8th-7th century BC reliable. Summary The Kurtan belt represents a unique example among South Caucasian belts. With its rich iconography it is connected with other belts from Northern Armenia and 45 It has been also suggested that they have both functional and decorative meaning (Catelluccia 2017: 24). 46 Xnkikyan 2002: Plate 81.4. 47 Narimanishvili et al. 2017: Figure 4.3, 4.4. 48 Castelluccia 2017: 291. 49 See Kellner 1991: 40, nos 12, 57, 76, 105, 175, 178, 182, 237-239, 241, 269, 397, 412; Tarontsi 2018: Figure 3. 50 Kellner 1991: no. 269. 51 Chidašeli 1986: 20-24. 52 Chidašeli 1986: 22-23.

86

Ruben Davtyan and Michael Herles: The Kurtan Belt Devejyan, S. H. 2010. Some Urartian objects from the tombs of Lori Berd. Aramazd: Armenian Journal of Near Eastern Studies 5/2: 76–89. Devejyan, S. H. and Davtyan, R. H. 2017. Elaborate harness buckles from Lori Berd. Aramazd: Armenian Journal of Near Eastern Studies 11/1-2: 186–205. Devejyan, S. H. and Hobosyan, S. G. 2008. Excavations in the village Kurtan of the Lori province. In P. A. Avetisyan, A. A., Kalantaryan, and R. S. Badalyan (eds), The Culture of Ancient Armenia 14: 113–127. Yerevan, Gitutyun (in Armenian). Dzhafarov, G. F. 1993. A kurgan of the Bronze Age period near Sarychoban. Rossijskaja arxeologija 4: 191–207 (in Russian). Geyusheva, T. H. 2009. On the bronze belt from Apsheron. In Caucasus. Archaeology and Ethnology. International Scientific Conference, 11-12 September, Materials: 167– 172. Baku, National Academy of Sciences. Herles, M. and Davtyan, R. 2017. Neue Untersuchungen in der Provinz Lori (Armenien). Mitteilungen der Deutschen Orient-Gesellschaft 149: 9–56. Israyelyan, H. R. 1967. Traces of sun worship in Armenia during the Bronze Age. Herald of the Social Sciences 4: 77–88 (in Armenian). Kellner, H.-J. 1991. Gürtelbleche aus Urartu. Stuttgart, Steiner. Martirosyan, H. A., 1964. Armenia in the Bronze and Early Iron Ages. Yerevan, Academy of Sciences (in Russian). Mezhlumyan, S. K. 1991. Zooarchaeological materials of Lori Berd. In G. A. Tiratsyan (ed.), Archaeological Field Work in the Republic of Armenia in the Years 19891990. Abstracts of Papers Delivered at the Conference in May 1991: 45-46. Yerevan, Publishig House of the Armenian Academy (in Armenian).

Narimanishvili, D., Vadachkoria, R., Tamazashvili, K., Juszczyk, K. and Tomczyk, W. 2017. Archaeological excavations at Beshtasheni Late Bronze Age early Iron Age Cemetery (2015-2016 results). Aramazd: Armenian Journal of Near Eastern Studies 11/1-2: 120– 144. Piotrovskij, B. B. 1949. Archaeology of Transcaucasia. Leningrad, Leningradskij Gosudarstvennij Universitet (in Russian). Rudenko, S. I. 1953. The Culture of Mountainous Altay during the Scythian Period. Moscow and Leningrad, Izdatelsvo akademii nauk SSSR (in Russian). Tarontsi, S. 2018. The mythical images of lion and bull in the Urartian bronze belt iconography. History and Culture. Armenological Journal of Yerevan State University: 350–365 (in Armenian). Tekhov, B. V. 1980. The Tli Necropolis I. Tbilisi, Mecniereba (in Russian). Xnkikyan, O. S. 2002. Syunik During the Bronze and Iron Ages. Yerevan, Mayreni. Yesayan, S. A. 1967. Tomb 14 of the Astghiblur necropolis. Patma-banasirakan handes (HistoricalPhilological Journal) 1: 221–226. (in Russian). Yesayan, S. A. 1984. Gürtelbleche der älteren Eisenzeit in Armenien. Beiträge zur Allgemeinen und Vergleichenden Archäologie 6: 97–198. Yesayan, S. A. and Avetyan, V.V. 1981. Bronze belts discovered in Lori. Patma-banasirakan handes (Historical-Philological Journal) 4: 314-318. Yesayan, S. A. and Mnatsakanyan, H. H. 1977. Bronze belts from Lchashen and Stepanavan. Patmabanasirakan handes (Historical-Philological Journal) 3: 276–281 (in Russian).

87

Achaemenid Habitats in Beniamin II (Shirak, Armenia) from the End of the 6th Century BC to the End of the 4th Century BC Stéphane Deschamps

Ministry of Culture, Paris, France

François Fichet de Clairfontaine Ministry of Culture, Paris, France

Felix Ter-Martirossov† Abstract: At the foot of a hill on which was built an Achaemenid palace, several semi-excavated constructions from the 6th4th centuries BC were found. Welcoming metallurgy workshops at their origin, they were later the seat of a breeding activity while sheltering a domestic space. From artisanal iron workshop to long-house, they offer a triple interest: They interest the economy of an Achaemenid palatial domain and present a typical architecture of the Armenian plateau. Finally, they seem to be very similar to the houses described by Xenophon in the 5th century BC, in the Anabasis and the epic of the Ten Thousand. The excavations have thus uncovered constructions presenting similar similarities with the traditional architecture of the ‘glkhatoun’ type and its corbelled roof called ‘hazarashen’. They thus demonstrate the antiquity of the architectural system, dating as early as the second half of the 6th century BC. Keywords: Achaemenid, semi-excavated habitat, metallurgy, breeding, Xenophon, glxatoun, hazarashen

The Site and Periodization of Its Occupations Works following the earthquake that affected the town of Gyumri (former Leninakan) in december 1988 led to the discovery of the site of Beniamin, occupied mainly during the Achaemenid and Hellenistic periods (Figure 1). Located about 15 km south of Gymri, it extends on a succession of small hills and a plain that they dominate, west of Gyumri-Yerevan road. Numerous archaeological excavations were carried out under the direction of Prof. Felix Ter-Martirossov, Mr. Hamazasp Khachatryan and Ms. Larisa Yeganyan. These researches were concentrated on two large groups of vestiges: a palace of the Achaemenid period located on one of the hills dominating the site to the north, then a monumental complex (basilica) of Hellenistic period, associated with habitats and burials (Figures 2-3). From 1999 to 2007, under Armenian-French scientific cooperation project, new researches were carried out on the Achaemenid palace in order to complete the study. Then they continued below the hill, West, in order to study a habitat punctually recovered during surveys carried out few years ago. Four excavations were devoted to the study of habitat, the subject of this article.1 This program was completed on the ground in 2007.2 1 Archaeological excavation was co-directed by a joint team - F. TerMartirossov and S. Deschamps, F. Fichet de Clairfontaine, and V. Mutarelli for architectural surveys and analyzes. These operations were financed by the Ministry of Culture (1999-2005) and the Ministry of Foreign and European Affairs (2006-2007). 2  This research was carried out in close collaboration with

Systemizing the Past (Archaeopress 2023): 88–115

Part I. Outbuildings of the Palace ? : Habitats and Workshops General Constructives Data Below the palace, two vast buildings of quadrangular plan and the beginning of a third, constituting or not one or several houses, have been discovered, each one of an average surface of 150 m2 (Figures 4-5). Established on a low north-south slope, these constructions have the peculiarity of being partially excavated. The design principle of construction is particularly visible in building A, located in the upper part of the slope. Its right-of-way was excavated and then the inner walls of this excavation were simply covered with a clad of large blocks of basalt (conserved to a maximum height of 2.50 m in the north of the building) assembled with a clay binder (Figure 6). Some blocks may also be arranged to reinforce the upper levels of the walls, but the inner walls of the excavated parts have only one stone sidding. The interior walls were covered with a outside archaeologists Hamazasp Khachatryan, director of the Shirak Regional Museum, and Larissa Yeganyan, who conducted the first research on the site. We would like to thank them warmly for their advice and valuable collaboration during all this work. This program also benefited from the collaboration of Mrs. Armine Gabrielyan, Armine Hayrapetyan, and Varduhi Melikyan (Institute of Archaeology and Ethnography, NAS RA, Yerevan, Armenia) and Anna Azizyan (Institute of Archaeology and Ethnography, NAS RA and Erebuni Historical and Archaeological Museum-Reserve, Yerevan, Armenia) whom we would like to thank for their effective participation in this research.

Stéphane Deschamps et al.: Achaemenid Habitats in Beniamin II

Figure 1. Achaemenid sites in Shirak – Armenia (drawing by F. Fichet de Clairfontaine).

coating of smoothed clay, so that the basalt blocks were not visible at all. The use of a double sidding system surrounding an internal blockage of clay and stone is attested only in the case of joint ownership with a space already excavated or when the difference in level between the floor of the house and the outside requires it (Figure 7). Such is the case of the western part of the north wall of building B, in contact with the excavated parts of building A, the west wall of the same building (adjoining building D), and the strong south facade wall of building B, joint to a likely yard, to the south. Access to the exterior is to the south (building A, period Ia and

building B, period Ib), in the form of a fairly narrow and relatively long corridor, clearly prolonged beyond the door by low walls If only a remain of a low wall was found for building A (base of wall M1081, period Ia), the exit device was better preserved for building B (period 1b). After having crossed a narrow corridor of an average length of 4 m (corresponding to the width of the south wall of the building, Figure 8), an outer space was reached, the lateral parts (east and west) of which were maintained by low walls, enable to be held in place the exterior soil. It is not known if such a device existed in building A for period Ia, but the discovery of the 89

Systemizing the Past

Figure 2. Aerial view of Beniamin. At the top of hill, Achaemenid palace, at left below the outbuildings, and rigt near the road, Helenistic palace (photo by A. Furwangler).

Figure 3. Map of Achaemenid sites in Beniamin (CAD S. Deschamps).

90

Stéphane Deschamps et al.: Achaemenid Habitats in Beniamin II

Figure 4. Semi-excavated buildings A and B - Ve-IVe BC (photo by S. Deschamps).

Figure 5. Overall plan of mixed-houses in Beniamin (drawing by V. Mutarelli and S. Deschamps).

91

Systemizing the Past

Figure 6. Facing of large basalt blocks. Building B. Beniamin (photo by S. Deschamps).

Figure 8. Long corridor and door south. Building B. Beniamin (photo by S. Deschamps).

The analysis of the levels of destruction clearly indicates the presence of a soil roof, probably vegetalized, that was to be supported by a wooden framework. Careful investigation of a cross-section in the northeast corner of building B, while showing the dynamics of collapse of the roof (Figure 12), made it possible to restore its roof and suggests the existence of a roof covering structure of a cantilevered pillar system forming a central dome covered with clay and earth. Four stratigraphic units succeed each other (Figure 13). At the base is a homogeneous and pure yellow clay layer (no. 1102), whose thickness varies from 0.20 to 0.30 m on average. Although it is a level of pure clay, we will note the presence of some ash passages suggesting the disintegration of plant elements. This first layer is surmounted by a level (no. 1101), the thickness of which varies from 0.23 to 0.30 m, consisting of a fine matrix, rather ashy, evoking more clearly the presence of plant elements. It is then surmounted by a new clayey layer consisting of a very fine matrix, similar to the layer no. 1102 constituting the base of this succession. The upper interface of this layer reveals the base of a new layer whose texture is very close to no. 1101 but whose upper part is truncated. The dip of these different layers clearly indicates them in a dynamic collapse of a

Figure 7. South Wall and long corridor to access the building B. Beniamin (photo by S. Deschamps).

foundations of a low wall and then of a basin (no. 1117, Figure 9) further to the east also suggest the presence of an outer courtyard house, partially excavated (Figure 9). From this viewpoint, building B seems to reproduce the same architectural layout as building A against which it is leaning during the period Ib (Figure 10). Each building is provided, in its central part, with four stone bases intended to receive wooden columns which supported the roof (Figure 11 – detail of a column base). 92

Stéphane Deschamps et al.: Achaemenid Habitats in Beniamin II

Figure 9. Basin 1117. Period IA. Beniamin (photo by S. Deschamps).

covering material that has collapsed suddenly from the periphery to the center of the building. As a hypothesis, we propose to identify two constructive phases for the preparation or repair of the covering corresponding to the layers no. 1102/1101, and then no. 1098 and the base of the layer which overcomes it. In this case, the first level of covering would be ensured, above the wooden elements of the framework, by a homogeneous clay layer intended to ensure a leaktightness.This level of preparation would then be surmounted by a layer closer to the texture of the topsoil, preparing a vegetalisation of the roof. Similarities with the traditional architecture of the Armenian continental shelf suggest that the constructive mode of the cover of building B can be compared with the ‘glkhatoun’ houses with a progressive cantilever beam covering system (‘hazarashen’, Figure 14), visible in traditional Armenian architecture until the 19th or early 20th centuries, by examples in Talin (Figures 15-16) or Guerakar (Figure 17).3

Period Ia – A First Habitation? The first period of occupation (Period Ia) concerns the construction of building A, in north (Figure 18). With a surface area of nearly 130 m2, it is largely excavated in the natural terrain and has walls with simple facing of large blocks of basalt. It has access to the south, the location of which will be taken over later by the communication door with building B (period Ib) (Figure 19). This first period of occupancy left little evidence of occupation due to the many subsequent modifications. However, one notes that there is a facing stone pit located almost in the center of the building (Figure 20), and a second basin to the south, located outside this first habitat, probably in the right-of-way of a courtyard, which will then be occupied by building B. The stone pit F2171, uncovered in the center of the building, is a pit with an average length of 1,30 m and a width of 0,80 m. With a depth of 1,10 m, this pit is completely paving with small slabs. The filling, homogeneous, consisted of topsoil containing no particular artefact and no trace of rubefaction. The existence of this central structure, which was filled before the introduction of the metallurgical activities, confirms the existence of pre-occupation phase, without our being able to assess the duration of this first phase (probably fairly brief).4 This first phase is also confirmed by the discovery of

Six main periods of occupancy were highlighted by the analysis of stratigraphic data and archaeological artefacts discovered on soils.

The covered surface and the spans between the wooden pieces are difficult to associate with the principle of a flat roof. It would have been difficult to bear the loads of clay and earth and the weight of snow in winter. 3

It will be noted that this structure is located in the central space of the building but slightly set back from the supposed location of the zenithal illumination erdik.

4

93

Systemizing the Past

Figure 10. Periods 1A & B (drawing by V. Mutarelli and S. Deschamps).

94

Stéphane Deschamps et al.: Achaemenid Habitats in Beniamin II

Figure 11. Column base. Building B. Beniamin (photo by S. Deschamps).

Figure 12. Stratigraphic section showing the collapse of the roof. Building B (photo by S. Deschamps).

95

Systemizing the Past

Figure 13. Stratigraphic section. Building B (drawing by F. Fichet de Clairfontaine).

a structure morphologically close enough, southeast of building A, in the northeast corner of building B. It is also a dry stone structure, filled with a homogeneous humic sediment, with remains of a small oven, on the summit of its filling, contemporary to first metallurgical activities in building B (see Figure 9). The relative chronology of this structure is unambiguous: the southern facing of wall M2024 rests on the wall of the northern part of the pit, thus demonstrating that the wall of building B was built after the construction of this pit. If it offers a morphology similar to F2171 (dry stone facings and identical filling), it differs by its proportions : a length of 3m for an average width of 1.90 m and a conserved depth of 0.43 m. There is no element to assign a particular function to this structure (basin?). Located south of building A, and thus outside, it also raises the question of the surroundings of this building before the construction of building B. It suggests the original presence of a courtyard which will then be replaced by building B. In view of these different aspects, it is difficult to specify the function of this first construction, but nothing seems to oppose domestic use. This first occupation is dated from the end of the 6th century and the beginning of the 5th century BC.5

Two new buildings are implanted south of building A (Figure 21). This concern building B which has a surface area of 148 m2 and a other building ( building D), which extends largely outside the archaeological excavation, built to the west of building B. According to the same architectural principle, building B (Figure 22) is partially excavated in the natural ground, and some walls have a double facing where the difference in level with the exterior ground requires it.6 An access door was built in the middle of the south facade of the hall, in the form of a gently sloping corridor on the ground, making it possible to compensate the difference in level with the exterior. This door opens to a small courtyard bounded to the east and west by low walls. The second important modification concerns the function of these different spaces. Indeed, the three buildings uncovered seem exclusively destined for an activity of iron metallurgy from the start of the 5th century BC (Figure 23). The metallurgical evidence points us towards a postreduction activity clearly comprising a forging activity and may also be of metal refining. On the other hand, no evidence has been found to identify any ore reduction activity. The best preserved elements of this craft were found in the central part of building A (Figure 24) in form of a large oval pit containing almost exclusively ashes and charcoal and two wooden subfoundations blocked with stones. By comparison with other similar structures, it is possible to identify these two structures like anvil bases characteristic of forges (Figures 25-

Period Ib: Specialized Activities in the Iron Metallurgy Two main events characterize the following period (period Ib): development of buildings and introduction of specialized activities.

Such is the case of the south wall which constituted the front of the housen opening onto the courtyard and whose average width is about 4 m.

6 5

Fichet de Clairfontaine and Deschamps 2012.

96

Stéphane Deschamps et al.: Achaemenid Habitats in Beniamin II

Figure 14. Proposition of restitution of buildings and their frames (drawing by V. Mutarelli and S. Deschamps).

26). This hypothesis was confirmed by the discovery of many iron scalings, characteristic of this phase of metalworking.7 Moreover, the very numerous small combustion structures (Figure 27), most often placed

on the ground, have delivered this type of scalings and show an intense activity of forging in the two buildings A and B. A test was also carried out on the soil of building D, showing similar activity, although the small excavation surface in this area does not make it possible to specify the nature or organization of the activity in building D. It should be noted that the anvils occupy a central position in building A, which, unlike combustion structures, require natural lighting, where

7 The presence of two types of scales, in plates and balls, suggests that we are not in the presence of the last finishing phase of the objects, but that more important firing operations have also been carried out (refining, welding). Information provided by Ms C. Le Carlier de Veslud, archeometallurgist, UMR 6566 CReAAH, Rennes.

97

Systemizing the Past

Figure 15. Talin Hazarashen Plan. (photo by F. Fichet de Clairfontaine).

Figure 16. Talin. Hazarashen. Frame and erdik (photo by F. Fichet de Clairfontaine).

98

Stéphane Deschamps et al.: Achaemenid Habitats in Beniamin II

Figure 17. Hazarashen plan in Guerakar (drawing by A. Rideau).

Figure 18. Building A. Period 1a (drawing by V. Mutarelli and S. Deschamps).

99

Systemizing the Past

Figure 19. Access door between buildings A & B (photo by F. Fichet de Clairfontaine).

we propose to place a central zenith opening (the erdik of hazarashen type blankets), while the hearths appear to be more systematically located on the periphery of the building. These activities occupied three buildings, clearly excluding any other domestic activity. So we’re probably in front of with a specialized activity. Period IIa / IIb: New Functions Linked to Breeding

Figure 20. Pit 2171. Building A (photo by S. Deschamps).

100

During the period II, probably before the middle of the 5th century BC, the metallurgical craft is abandoned in both buildings in favor of a new activity related to the breeding. This change in function is suggested by development of stone pavements and mangers for cattle and, to a lesser extent, for sheeps (Figure 28). In building A, these pavements occupy the northern and southern parts of the space. These surfaces, consisting of wellarranged basalt slabs, have a fairly constant width of 1.60 m with respect to the foot of each manger. Their length varies from 9 m to 9.50 m (north of building B, north and south of building A) to 6,40 m (west of building B) (Figure 29). In building B, the architectural developments, better preserved, were highlighted along the north and west walls, by example, the paving 1075, which was adossed against the north wall of building B (Figure 30). The construction of this device first required the filling

Stéphane Deschamps et al.: Achaemenid Habitats in Beniamin II

Figure 21. Buildings A & B. Period Ib (drawing by V. Mutarelli and S. Deschamps).

Figure 22. View of building B from south (photo by F. Fichet de Clairfontaine).

101

Systemizing the Past

Figure 23. Building A and métal working. Period 1a. Plan (drawing by V. Mutarelli and S. Deschamps).

Figure 24. View of metalworking activities in building A (photo by S. Deschamps).

102

Figure 25. Anvil in building A. period 1b. (photo by S. Deschamps).

Stéphane Deschamps et al.: Achaemenid Habitats in Beniamin II

Figure 26. Second anvil in building A. Period 1b (photo by S. Deschamps).

Figure 27. Small fireplace in building A. period 1b (photo by S. Deschamps).

103

Systemizing the Past

Figure 28. Buildings A & B during phases IIa - IIb (drawing by V. Mutarelli and S. Deschamps).

As for the previous period, we are confronted with the question of the functions of these different buildings: a unique function linked to breeding ? Mixed activities combining domestic (habitat) and breeding functions? If we look carefully the surface ratio between the paving / feeding system and the total area of each building, is 23.50% for building B and 33.80% for building A. In building A, taking into account the spaces for the clearances and circulation of cattle, there is very little building for other activities, for example related to domestic functions. It is also that in this building that the most important litter levels have been observed. In building B, on the other hand, aside, apart from the fact that these spaces occupy a smaller area, they are distributed north of a diagonal linking the north-east and south-west corners of the hall. Thus, all the space east of the bases of oriental columns could be allocated to another function, on an average surface of 60 m2 (about 40% of the total area). But there is no evidence that building B already receives a habitat from IIa / Iib.

of the old communication door between buildings A and B. A siding was then constructed in front of the wall at an average distance of 0.50 m to 0.60 m from the south facing of the wall and on a preserved height of 0.35 m (Figure 31). The space between these two sidings was filled by a succession of small clay layers. A particular device was found within this bench, at a distance of 4.40 m from the west wall of the building. It is a vertically laid basalt slab with, on its western side, a circular hole of 9 cm diameter to receive a wooden element (Figure 32). A second hole, observed at the level of the facing of the west wall, allowed for the installation of a woodden stall of a length of 4.40 m, located at an average altitude of 0.35 m compared to the level of the paving. The association between this device and the pavement, the levels of litter revealed (Furthermore, the researches in building A identified litter deposits with many grass or straw residues) and the ethnological comparisons with similar architectural devices still preserved in the surrounding villages leaves little doubt (also for example in Haykadzor during Middle Age (Figure 33): this is a system of feeders/mangers suitable for cattle, build for stables.... 104

Stéphane Deschamps et al.: Achaemenid Habitats in Beniamin II

Figure 29. General view of building B. In the foreground, a base, and in the background paving along manger. Period IIa (photo by S. Deschamps).

Figure 30. Paving in building B along the north wall (photo by S. Deschamps).

105

Systemizing the Past

Figure 31. Detail of mangers in building B (photo by S. Deschamps).

Figure 32. Hole for wooden pièce in stable building B (photo by S. Deschamps).

Figure 33. Medieval stable in Haykadzor. Archaeological excavation by H. Khachatryan (photo by F. Fichet de Clairfontaine).

106

Stéphane Deschamps et al.: Achaemenid Habitats in Beniamin II connection between buildings A and D (see Figure 34). This new communication thus indicates a working relationship between the three buildings. The small surface excavated in building D unfortunately does not allow us to specify its function. Let us note, however, existence of levels of litter that again underline a cattle breeding or feeding activity in the northeast part of this building.

Period IIc: Domestic Activities During Phase C of Period II (Period IIc), the eastern part of building B has particular interior installation with presence of an adobe bench of width of 0.30 m adossed to edge of the eastern wall, and probable decoration created in clay soil in the form of broken lines (Figure 34) The footprint with an average width of 10 cm may may correspond to the location of a missing wooden element. Located in front of the brick bench, it appeared as series of irregularly broken lines with no particular symmetry (Figure 35). If the precise function of this arrangement eludes us, there is no doubt that it testifies to a particular care given to this space which we propose to identify as the part reserved for habitat and domestic functions. The analysis of the ceramic reinforces this hypothesis, with a richer ceramic catalog than for previous periods and the use of fairly thin open forms more in use in the habitats.

In view of these observations, we suggest to restituate an activity, probably strictly reserved for breeding or feeding in building A, and a mixed activity associating housing (60%) and breeding functions (40%) in building B, during this period. Period III: A Brutal Destruction Following an Earthquake? Period III is characterized by a brutal destruction of buildings A and B.8 The most obvious evidence of this event was uncovered in the western half of building A, as well as in the northeast corner of building B.

Circulation between buildings was also modified during period II. After filling of the old communication door between buildings A and B, made necessary by creation of the feeder, a new door was created to obtain a new

The investigated surface, too small for building D, does not confirm whether this destruction also affects this building.

8

Figure 34. Plan of mixed-houses during period IIc (drawing by V. Mutarelli and S. Deschamps).

107

Systemizing the Past disorders in the wall sidding M2024 and M2014 (north wall of building A). At the end of period IIIb, habitats are abandoned, for duration which cannot be estimated precisely. Period IV: Partial Reoccupation of Buildings The fourth period, poorly represented on stratigraphic level, are reflected by partial reoccupation of buildings A and B. This reoccupation is particularly clear at the northwest corner of building A and northeast of building B, which is significantly affected by the destruction of period III. Walls, rather coarsely built, have come to consolidate the angles of the two buildings; On the other hand, the levels of collapse of the cover of the buildings have been re excavated to implant new occupations. This ‘repair’ is synchronous with level of homogeneous and humic earth, the presence of which has been verified at both the southeast corner of building A and the right-of-way of building D. This backfill also indicates that building D was fully filled after its destruction. For building A, the consolidation of the corner of the building was accompanied by closing old access and a new door was build in the north-west corner of the building, framed by two walls conforting the corner. Generally speaking, although some abandonment sequences have been identified punctually, it seems to us that this phase of reoccupation follows only briefly the destruction of buildings during period III. The contemporary occupancy levels of this period have been preserved in a too incomplete manner to allow us to identify the function of the spaces thus reused. It should be noted that no evidence of livestock-related activities was identified for this period, more marked by purely domestic occupation confirmed by ceramic assemblages.

Figure 35. Detail of the mud-brick bench in building B, period IIC (photo by S. Deschamps).

The first levels of destruction rest directly on the last periods of occupancy of period II, which tends to prove that this destruction was not preceded by an episode of abandonment, even a relatively brief one, which would inevitably have left a testimony stratigraphic. The dip of the layers of destruction observed in buildings A and B indicates a collapse of the central part of the buildings which leaded to general subsidence of the roof of the two buildings.9 Walls appear relatively unaffected by this destruction, by exception of two areas: the northeast corner of building B and the southwest corner of building A (Figure 36). The observation of architectural survey of houses A and B shows a deformation of a parallelogram square plane, characteristic of what can be observed after an earthquake on constructions not unlinked.10 The south wall of building A has been deformed by compression and stretching to corners under the pressure of perpendicular walls. This destruction was not limited to the roof of buildings A and B, but also caused the wall M2024 to tilt southward in the northeast corner of building B, the collapse of the southern angle West of building A and a number of 9

Period V: New Habitats This period concerns essentially a new building (building C), which is located on the right-of-way of the old building D, where two fireplaces and pits have been uncovered (Figure 37). It should be noted that this new occupation, of which archaeology has only yielded a partial plan, has no relation with the previous constructions. Of a clearly different orientation, this construction covered the old buildings destroyed and no doubt very widely filled. Building C forms a construction resting on stone walls with a maximum width of 7 m. The eastern wall of this building has been identified (M1121), consisting of a simple siding of basalt blocks facing north-west / south-east. This wall extended to the level of the old P1080 door right-of-way. A small wall in return has been identified (1122) but we can not specify in the present state of the research whether it is really a wall or a threshold arranged at

Deschamps et al. 2006. Helly and Rigaud 2011.

10

108

Stéphane Deschamps et al.: Achaemenid Habitats in Beniamin II

Figure 36. Collapse during earthquake in building A (photo by S. Deschamps).

the level of a possible entrance. The northern corner of this building was also uncovered, as well as fairly coarse wall consisting of an accumulation of basalt blocks, delimiting Building C to the northwest. In the eastern part of Building C, a clayey soil was discovered, preceded by a particularly careful paving occupying the western part of the hall. The discovery of three silos located in the building, west of pavement 1135, must be noted. These silos have a perfectly characteristic pear profile. In the upper part of the filling of the silo 1160, the grave of a child placed in a ceramic was discovered (Figure 38). Three pits were also discovered in the vicinity of this construction, one in the rightof-way of the former building A, and the other two in the old building B. The pit F1128 is dug partly on the right-of-way of the old wall M1126, without however altering the siding of this wall. This shows both that this pit is not linked to a material recovery, and that the relatively poorly preserved M1126 wall was already strongly levelled before the V period. Similarly, paving 1085, bordering west wall of the building (period II) is reached in the bottom of the pit without being altered. The filling of this pit did not provide any particular piece of information informing us about its function. It has a roughly circular shape, with an average diameter of 1.70 m. The pit F1130 has an oval shape (1,30 m × 1,10 m) and is located in the same stratigraphic context as the pit F1128. However, its filling differs considerably. It presents a filling comprising of many animal bones among which one noted the presence of bones of horse

and dog. More singular is the presence of human bones: a fragment of the skull cap and, more interestingly, the lower part of a human skeleton sectioned at the level of the third lumbar vertebra and placed in anatomical connection to the level of the base of the femurs. No other element has been bring to light and the hypothesis of a subsequent burial incorporated in the pit following stratigraphic compressibility must be excluded, we must admited that this part of the skeleton was deposited in the filling of the pit (relegation burial?). The presence of these conservation structures encourages us to interpret this construction as a small domestic building linked to an agricultural unit. The surface studied is too restricted for us to attempt a restitution of building C, but it will be remembered that it uses radically different construction methods than those of earlier periods. The ceramics, which still similar to achaemenid productions while incorporating new forms, could be dated from the end of the 4th century BC. Period VI: A Necropolis of the 3rd-2nd Centuries BC (?) This last period after the Achaemenid phase put and end to domestic occupation of this area of habitat to the benefit of a funerary space (Figure 39). Three graves were found at the eastern edge of the excavated area. Burial 1050 is located near the southeast corner of building B, placed into a narrow pit with an average length of 1 m for a width of 0.75 m and a conserved 109

Systemizing the Past

Figure 37. New occupation during period V (drawing by V. Mutarelli and S. Deschamps).

110

Stéphane Deschamps et al.: Achaemenid Habitats in Beniamin II

Figure 38. Grave n.1160. Period V (photo by S. Deschampss).

depth of 0.60 m. The whole of the pit is in stone clad and the burial is in a bent position, head to west. The accompanying funeral objects consists of two ceramics, a iron and glass paste bracelet. The second grave (Gr.1124) is located partly on the right-of-way of the eastern wall of building B (M1118). The pit is underlined by a stone wedging preserved over a height of 0.35m and the top of the blocks appears at 9 cm below the current soil. The body is placed in a bent position on the left side. The funeral objects of the deceased are not very numerous and limited to two bronze bracelets and an iron bracelet, as well as some small pearls (Figure 40). A little further north, a third grave was discovered, also marked by a blocking stones constituting the limits of the tomb. Three slabs of larger size ensure the cover of the pit. The body is also placed in a flexed position on the left side, with the forearms bent under the chin. Two bronze bracelets accompany the deceased (a bracelet on each forearm), as well as a small terracotta flask placed at the back, practically at the level of the blocking stones. The head is oriented to the north-west.

Part II – Conclusions - General Characteristics of Achaemenid Habitat in Beniamin The general presentation of the archaeological data calls some remarks, on the one hand on the function of the successive occupations, and on the other hand on the modes of construction adopted. As regards the function of the studied areas, two elements dominate. In addition to an poorly characterized first occupation (period Ia), the dominant element for the oldest occupations concerns an important activity geared on iron metallurgy, and more particularly on the last phase of metalworking. The most surprising element is undoubtedly the importance devoted to this activity which occupies the two (or perhaps three?) buildings uncovered, to the exclusion, it seems, of any other activity of the domestic type. The remains of this activity are too altered by the later occupations to be able to attempt a fine spatial analysis. Nevertheless, the number of small combustion structures, most often laid on the ground, the density of scalings near all heating zones and the evidence of two anvil bases in the central part of building A confirm an intense forge activity. The absence of slags which could reflect a metal reduction activity should be noted, but it can not be excluded that these elements should be gradually removed from the workshops. Finally, the direct communication between these three buildings leaves little doubt and emphasizes that we are in the presence of an important mettalurgy workshop which brings together at least

The objects uncovered in these three graves can be dated from the 3rd century BC. This funeral space, of which archaeological excavation have probably only reached the western limit, consecrates therefore the exodus from habitat in favor of a displacement towards the east of the site of Beniamin at the end of the 3rd century BC, towards the monumental center of the Hellenistic period.

111

Systemizing the Past

Figure 39. Period VI - Hellenistic necropolis (drawing by V. Mutarelli and S. Deschamps).

three buildings, all on a surface of more than 280 m2 (about 400 m2 if the surface of building D is added to the average area of the other two buildings. This period could fall between the end of the 6th and the first quarter of the 5th century BC, which places these contemporary occupations of the first period of the Achaemenid palace located on the nearby hill. More than a place of power with a palace included in the context of the ‘column halls’ of the Achaemenid Persian empire, Beniamin would also be a place of specialized economic activity, probably gradually concentrating a habitat. Only further research would make it possible to specify the importance and the density of the habitat of this period from the end of the 6th century BC.

activities indicates a mixed function for building B, the eastern part of which appears to be well reserved for domestic activities. The apparent absence of domestic activities in building A and the circulations changes that now require access to this building through building D confirm that these three buildings continue to form a single unit, now focused on breeding. On the other hand, the following periods, IV and V, brought a marked decline of the habitat towards individual structures at the end of the Achaemenid period. This decline will be definitely consecrated by the emergence of a necropolis whose archaeological excavations has approached only the western limit. The construction methods used constitute a second important contribution of this research. Firstly, it should be noted that the constructive principles are totally different from those used for building the palace. For the palace, the first two periods of occupation use adobe brick walls resting on foundation walls with one or two courses of stone. Special internal dispositions, such as the presence of adobe brick benches, are part of construction methods which seem to appear after the

During a second period, the metallurgical activity is abandoned in favor of an activity related to livestock farming. This change in activity affects all three buildings which maintain a functional relationship, with a main access located on the southern part of building B.11 As we have mentioned, the spatial distribution of 11 Of course, at this stage of the study, direct access can not be excluded from building D, in the south.

112

Stéphane Deschamps et al.: Achaemenid Habitats in Beniamin II

Figure 40. Hellenistic grave n.1124 (photo by S. Deschamps).

Urartian period.12 The habitats, on the other hand, are part of an earlier architectural tradition, mostly using walls built with large blocks of basalt put together with sharp joints or with a clay binder. We are therefore being in présence of two distinct architectural traditions. The most decisive factor is undoubtedly the use, from the end of the 6th century BC to a covering system resting on four circular slabs of basalt, placed in square in each building. These four bases were receive wooden 12

columns themselves supporting the roof structure of the hall. Such supports were obviously necessary because of the distance between the walls, always greater than 10 m. The type of roof structure used is not unlike the glkhatoun type house of the traditional architecture of the Armenian plateau, with its roof type hazarashen, primitive form of the dome covered with earth resting on a frame of wood built with corbelling from the centre square.13

For this, see Deschamps et al. 2012.

13

113

We do not refer to the many publications available on this subject.

Systemizing the Past Two facts have here supported this hypothesis. The stratigraphic evidence is the first which indicates a mode of covering consisting of earth and wooden corbelled-structure. The second concerns more particularly the spatial analysis of the activities of the forge, and more particularly the central position of the two anvil bases. Unlike hearths for forging, this phase of the work requires lighting, if possible natural. In the case of a mode of coverage close to the hazarashen principle, this lighting could thus be provided by the erdik (see Figures 15-16 for Talin), the zenithal skylight present in this type of cover. It should be noted that this is the only lighting available for these buildings.

Underground dwellings: here, excavated constructions. Opening looking like a well: this is probably a description of the erdik (see below). The interior was spacious: contrast between the access (erdik, see above) and the internal space, here between 130 and 150 m2. Cows, goats, sheep ...: Device for stalls of cattle and sheep with the association of pavings / feeding troughs and the important levels of litter. For the cattle entrances dug in the ground: the access corridor to the building B seems to us perfectly correspond to this element, here preceded by a forecourt of which it is noted that the lateral parts are well adapted to channel the cattle towards this access

These few elements, from which we suggest a restitution (see Figure 14), seem to confirm the use of hazarashen type cover as early as the end of the 6th century BC. Among the various possible options, we chose for this restitution proposal the simplest solution consisting of an assembly of wooden beams resting on the top of the walls.14 This one formed a frame on which we could placed longitudinal beams, the central square supporting them as well. The latter was naturally supported by the four wooden columns of the central space. The domed roof device could then be easily obtained by superimposing beams forming wooden frames of progressively reduced width, providing in the summit the space for the yerdik. The whole was then covered with a layer of clay found in stratigraphy and whose function was ensure waterproof before the laying of a second layer of soil, vegetated on the surface.

People were descending on a ladder. This precision might seem curious if one omits that the Ten Mile army joins this Armenian village in the middle of winter under abundant snowfall. Access by the erdik using a ladder was probably more appropriate than using the access corridor, used for cattle and cleaning litter and probably impracticable during this season. It seems fairly clear to us that the description of Xenophon, a Greek unfamiliar with this type of construction, convey us the probably fairly faithful picture of the houses of the period II.16 These are mixed-houses, as we have tried to demonstrate, where domestic and livestock activities can be associated in the same space. Moreover, these constructions were well suited to the rigorous climate of the Armenian plateau, alternating rigorous winters and hot summers, for which underground structures permitted a controlled temperature to be maintained. The hazarashen type of construction may also be a form of response and adaptation to seismic constraints.The roof could thus collapse without necessarily bringing about the collapse of the whole house, as we have seen that only two angles were significantly deteriorated during the period III, which did not prevent, after some repairs, to reinvest the same spaces during period IV.

This restitution clarifies a passage from the epic of the Ten Thousand in the Anabasis, during which Xenophon relates the expedition against the king of Persia Artaxerxes by his brother, Cyrus the Younger, with the help of Spartan troops and Greek mercenaries. The retreat of this routed army takes them to villages on the Armenian plateau. Xenophon thus describes the interior of an Armenian house: ‘The dwellings were underground. Their opening looked like a well, but the interior was spacious. The cattle had entrances dug into the ground; The people were descending by a ladder. In these dwellings, there were goats, sheep, cows and poultry and the little ones of these animals. All the cattle were fed in hay on the inside’.15 If this passage has given rise to various interpretations that may suggest troglodytic habitats (underground dwellings), the comparison between the text and the house (s) of Beniamin is quite eloquent.

Bibliography Badalyan, R. S., Edens, Ch., Kohl, Ph. L. and Tonikyan, A. V. 1992. Archaeological investigations at Horom in the Shirak Plain of the Northwestern Armenia, 1990. Iran 30: 31–48. Badalyan, R. S., Edens, Ch., Gorny, R., Kohl, Ph. L., Stronach, D., Tonikyan, A. V., Hmayakyan, S. and

Let us retain the essential elements: See Donabedian 2008: 50-51. 14 In this regard, for the north wall of building A, we can noted the presence of a bed of stone blocks at the top of the conserved part of the wall. It may have contributed to adjust of the horizontal wooden beams. 15 Xenophon, Anabasis, book IV, chap. 5, 25.

16 The epic of the Ten Thousand (Anabasis) is dated towards 401-400 BC, therefore shortly before the period III which consecrates the destruction of the buildings.

114

Stéphane Deschamps et al.: Achaemenid Habitats in Beniamin II Zardaryan, M. 1993. Preliminary report on the 1992 excavations at Horom, Armenia. Iran 31: 1–24. Badalyan, R. S., Kohl, Ph. L., Stronach, D. and Tonikyan, A. V. 1994. Preliminary report on the 1993 excavations at Horom, Armenia. Iran 32: 1–29. Badalyan, R. S., Kohl, Ph. L. and Kroll, S. E. 1997. Horom 1995. Bericht über die amerikanischdeutsche archäologische Expedition in Armenien. Archäologische Mittelungen aus Iran und Turan 29: 191–228. Deschamps, St., Fichet de Clairfontaine, F. and Mutarelli, V. 2012. Beniamin (5-4th centuries BC): A palace and its dependencies during the Achaemenid period. in P. S. Avetisyan and A. A. Bobokhyan 2012 (eds), Archaeology of Armenia in Regional Context. Proceedings of the International Conference Dedicated to the 50th Anniversary of the Institute of Archaeology and Ethnography, Held on September 15-17, 2009 in Yerevan: 197–207. Yerevan, Gitutyun. Donabedian, P. 2008. L’âge d’or de l’architecture arménienne. Nîmes, Éditions Parenthèses. Fichet de Clairfontaine, F. (ed.) 2007. Dans les Montagnes d’Arménie, 500 000 ans d’histoire avant notre ère. Catalogue d’exposition, collectif d’auteurs. Musée archéologique de Saint Raphaël (12 janvier-19 mars 2007), Musée départemental des Antiquités - Rouen (3 avril - 23 juillet 2007), Exposition organisée

avec le soutien d’« Arménie mon amis », année de l’Arménie en France, 2007: 23–28. Rouen. Fichet de Clairfontaine, F. and Deschamps, S. 2012. La céramique ourartéenne et post-ourartéenne du secteur du temple de Haldi de la forteresse d’Erebuni (milieu VIIe – début VIe siècle avant J.-C.). Aramazd: Armenian Journal of Near-Eastern Studies 7/1: 105–143. Fichet de Clairfontaine, F. and Deschamps, S. 2022. Achaemenid ceramics from the site Beniamin II (Shirak, Armenia): the end of the 6th to the mid 4th centuries BC. Aramazd: Armenian Journal of NearEastern Studies XV/1-2: 202-250. Karapetyan, I. A. 2003. Material Culture of Armenia in the VI-IV Centuries BC (Archaeological Monuments of Armenia 19). Yerevan, ‘Gitutyun’ Publishing House of the NAS RA (in Armenian and Russian). Khatchadourian, L. 2008. Social Logics under Empire : The Armenian Highland Satrapy and Achaemenid Rule ca. 600-300 BC. Unpublished PhD Dissertation, University of Michigan. Khatchadourian, L. 2014. Empire in the every day: A preliminary report on the 2008-2011 excavations at Tsaghkahovit, Armenia. American Journal of Archaeology 118/1: 137–169 Ter-Martirossov, F. 1999. Shirak in the epoch of Classical Antiquity. Research Papers – Shirak Centre of Armenian Studies 2: 32–46 (in Russian).

115

New Findings on Urartian Rock-Cut Tomb in Mazgirt/Kaleköy Fortress* Serkan Erdoğan

Yozgat Bozok University, Yozgat, Turkey Abstract: Kaleköy Fortress, located near to the Mazgirt District, is 26 km away from the city center of Tunceli. The fortress has been established on an oval-shaped hill with the height of 80 m and the diameter of 750 m. There is a two-roomed rock-cut tomb in Kaleköy Fortress on the southern façade of fortress citadel, on rocky podium. Architectural installation and decorations give this tomb a unique place among the other examples of Urartian rock-cut tombs. Architectural decoration elements, wall paintings and marks of inner decoration have qualities that are comparable with the royal tombs in Tušpa, the capital city of Urartu. The bodies of people buried in this rock-cut tomb have not been found. However, it can be assumed that they were members of the Urartian ruling house, perhaps, high-ranking officials, who governed over those distant regions. Thus, the rock-cut tomb in Kaleköy Fortress should be evaluated within the category of royal Urartian rock-cut tombs. This study aims to reveal the unknown features of this tomb within the light of new observations. Keywords: Urartu, Kaleköy, rock-cut tomb, local architecture, wall paintings

Introduction Kaleköy Fortress is located in the District of Mazgirt in the southeast of Tunceli, 26 km away from the city centre (Figure 1). The exact location of the fortress is 11 km away from the east of Mazgirt District centre, 4.5 km by direct line (Figure 2). The fortress takes its name from the Kaleköy village that is settled on its southern and south-eastern foothills. Kaleköy Fortress was built on a hill which is about 750 m in diameter and 80 m in height (Figure 3). The fortress is constructed on a steep hill dominated the valley spreading to north and east. This hill has been terraced to create eligible area for the construction of the structures of the fortress. The fortress construction technique with terracing method is familiar to the Urartian settlement model, as this construction technique is also known from other Urartian fortresses such as Mazgirt,1 Kancıklı near Ağrı/Patnos,2 and Bastam3 in north-east Iran. The Kaleköy Fortress contains many structures that are peculiar to Urartian architecture, including tworoom rock-cut tomb, an open-air sanctuary, an openair worship area, a stepped rock-cut tunnel, a chamber (rock-cut tomb?) in the middle of which there is a well, and rock platforms. It is a great pleasure to take part in this work honouring Pavel Avetisyan. This brief article, related with an Urartian rock-cut tomb that is located in Upper Euphrates Region is based on findings of a survey examination carried out in the region. 1 Huntington and Lehmann 1901: 180. 2 Burney and Lawson 1960: 189-192. 3 Kleiss 1980: 301; Kroll 1972: 294. *

Systemizing the Past (Archaeopress 2023): 116–126

Among the ceramics, Iron Age local featured grooved or ungrooved, mahogany or buff coloured ceramics of the Upper Euphrates region are predominant. However, a small number of red-slipped, thin-walled ‘palatial’ (Biainili) and low-quality Urartian ceramics have also been detected. General Features of the Rock-Cut Tomb The rock-cut tomb is located in the south of the citadel of Kaleköy Fortress that is ascending on a rock podium. The dromos of the rock cut-tomb is accessible by stairs, three of which are still preserved. The dromos in the shape of barrel vault is 1 m long, 2.05 m wide and 2.70 m high (Figure 4). The front facade architecture of the splendid entrance structure, which ends in an arch,4 is peculiar for the western regions of Urartu and exhibits the general appearance of the tomb (Figure 5). On the lower part of the left (western) wall of the dromos, towards the door frame there is a cuneiform inscription left by the Urartian king Rusa, son of Argišti (first half of the 7th century BC). The festoon ornamented cornice overhanging a 0.10 m bulge on the top of the inscription which is 1.31 m over the dromos facet lengthens to the doorway of flat ceiling. A cornice that has an apparent bulge divides the upper structure of the dromos from the lower structure of the wall. Right of this protruding cornice is formed by a wholly festoon ornamented cornice extension. The flat ceiling doorway is 1.75 m high and 1.10 m wide. Gateway of the rock-cut tomb is formed by three interior steps and an interior step in the chamber. Çevik 2000: 32; Köroğlu 1996: 36; Lehmann-Haupt 1910: 468-469; 1931: 627; Sevin 1994: 61.

4

Serkan Erdoğan: New Findings on Urartian Rock-Cut Tomb in Mazgirt/Kaleköy Fortress

Figure 1. Location of Kaleköy Fortress.

Figure 2. Satellite image of Kaleköy Fortress.

117

Systemizing the Past

Figure 3. A view of Kaleköy Fortress from the East.

Figure 4. General view of Urartian rock-cut tomb in Kaleköy Fortress.

settings of the door leaf are inside it is understood that passing is made through an one-winged inward door. Doorway is formed by two frames both inward and outward.

There is an oblong hallow that must be related with the door stay mechanism (open-close) on the left of the second interior step pedestal of the doorway. On both side of the third step there are two round holes burrowed to strengthen door leaf and there has been a comparatively deep round hole drilled to the upper ceiling panel of the door frame to support the door from above. On the left wall of the doorway there are two side-to-side mortises in the form of arch. Since the

The rock-cut tomb is two-chambered (Figure 6). The total area that it covers is 31 square m. The rough rectangular main chamber is in the size of 5.95-5.65 m × 3.60-3.10 m. The ceiling height of the chamber is 2.50 m. 118

Serkan Erdoğan: New Findings on Urartian Rock-Cut Tomb in Mazgirt/Kaleköy Fortress

Figure 5. Frontal view of Urartian rock-cut tomb in Kaleköy Fortress.

Figure 6. Plan of Urartian rockcut tomb in Kaleköy Fortress (after Çevik 2000: Plate 31a).

119

Systemizing the Past

Figure 7. Festoon ornamented cornices of the rock-cut tomb.

Figure 8. Niche in the main chamber of rock-cut tomb.

Just in front of the anteroom there is a 0.60 m diameter rock niche in the floor of the main chamber. Another 0.30 m diameter rock niche also has been detected on the west side of the anteroom. These rock holes must have been used to place the libationary bowls. Inside of this well-crafted chamber 0.6 cm or 12-13 cm festoon ornamented cornices are the main architectural attractions. These festoon cornices start from above the interior surface of the doorway and stretch to the side-to-side walls of the main chamber. These cornices ornamented with festoon ends linked with a festoon cornice on a jut which gets through cascading the height of the ceiling coat (Figure 7). This ceiling decorations that has a 25 cm jut ornamented with festoon has divided the south part of the ceiling of the main chamber from the north part. A huge rectangular niche 1 m deep and 1.55 × 0.95 m in size has been cut into the north wall of the main chamber (Figure 8). Considering the size of the niche, it is possible that it has been built to serve as a sarcophagus. Thus, the corpse must be reposed in full or half hocker positions. In front of this big niche there are two oval bowls or urn5 collets, one which is 0. 60 m and the other 0.40 m in diameter, burrowed on the chamber floor. Passage from main chamber to the adjoining chamber is provided by a 1.80 m × 1.10 m sized doorway. The doorway of the adjoining chamber on the east wall of the main chamber has both an interior and exterior door frame (Figure 9). The same festoon ornaments on both door frames except their sills are served as architectural decoration of the door cornice. On the door saddle and top rail of the doorway of the adjoining chamber there are to opposite burrows that must be related with openclose mechanism. The adjoining chamber, 3.55 m × 3.50 m in sizes, has a roughly rectangular plan. A niche 1.50 m × 1.30 m in sizes and 1 m deep has been carved on the west wall of the adjoining chamber. On the east wall

Figure 9. Door frame of the main chamber of the rock-cut tomb.

there is another amorphous niche 0.90 m × 0.85 m in sizes and 0.72 m deep. This niche is relatively small and carved imprecisely. In the adjoining chamber no other architectural furnishing has been detected except the niches and the doorway ornamented with festoon-like cornices. The adjoining chamber looks more plain and simple compared to the main chamber.

5 On these collets in the tomb are burrowed to install urn, see Çevik 2000: 55.

120

Serkan Erdoğan: New Findings on Urartian Rock-Cut Tomb in Mazgirt/Kaleköy Fortress

Figure 10. Gate with well.

Acoustics is another significant peculiarity of the rockcut tombs. Light sound reverberation creates a different acoustic space. The light resonance of the sound in the main and adjoining chambers might help to create an ethereal environment and a motivating impact.

seems to be related with afterlife as a part of tomb architecture. Wall Painting of Rock-Cut Tomb On the northern part of the ceiling of the main chamber three ribbon-like red stains, each of which are 0.80–0.90 m long and 0.40 m wide, have been detected. These colour bands have not been preserved on the upper south wall of the main chamber, near the entrance due to external factors like sun rays.

The Gate with Well Steps climbing to the dromos of the rock-cut tomb has been built by rock carving and provided passage to another door (Figure 10). This vaulted gate just on the north (left) of the rock-cut tomb has been named as ‘gate with a well’ because of its relation with the oval well in front of it. The upper right part has not preserved well, passage to the hypethral gate with the well has been provided by a step which is divided from the steps going to the dromos of the rock-cut tomb. Therefore, using the same foreground with the rock-cut tomb entrance, and having the same elevation indicates that both structures have direct relations and they both have been planned accordingly with each other. Detecting two conic rock hollows serving as mortise on the south side of the casing corroborates opinions on depicting here as a gate. The opening of the well is approximately 4 m in diameter and the visible depth to the earth fill is 2 m. There is a half round vaulted niche carved on a rock façade in the south over the gate with a well.

On the upper part of the exterior and interior door frame of the doorway of the adjoining chamber of the rock-cut tomb there are semicircular wall painting ornaments combining with each cornice that are ornamented with festoon (Figure 11). Although these colour remarks are more unobtrusive on other festooned cornices of the door frame, it is understood that interior and exterior door frame is entirely covered with these wall painting decorations. It can be observed that another red motif has been refurbished into the half-round red form in these colour streaks. Since these wall paintings have been highly damaged it is not possible to make an exact inference on the shape of these motifs. However, existing streaks can be resembled to ring motifs (Figure 12). Similar motifs can also be seen on the wall paintings of the Urartian site at Arin-Berd (Erebuni).6 These red wall paintings observed on the adjoining chamber doorway of the rock-cut tomb, is also seen to be festoonery cornices on the upper wall of the main chamber. Unfortunately,

It is a plausible inference that the dromos of the tomb and the door system which opens to the well, both of which are unique to Urartian architectural style, might serve to a cultic ritual. This structure defined as a well

6

121

Hovhannisyan 1961: Figure 28; 1973: 7, 16-17, 27.

Systemizing the Past

Figure 11. General view of the wall paintings of the door frame of the adjoining chamber.

Figure 12. Wall paintings on the festoon cornices of the door frame of the adjoining chamber.

wall paintings on the festoonery cornice in the main chamber are too unobtrusive to be detected. It is doubtless that these should be combined with another wall painting in different colour that is not preserved well till present.

except the north wall, which is decorated with niche of the main chamber (Figure 13). These application mortises can be claimed to be drilled for the mushroomheaded iron nails, siqqatu in Akkadian, which were used to hang some objects. The use of these types of nails found in archaeological excavations is understood to hang dedicatory objects in Urartian cult structures.7 It has been claimed that plates that have concave square corners for either decoration or sacrifice functions were hanged via mushroom-headed nails on Argišti’s (Horhor) rock-cut tomb in Van Fortress.8 The existence of concave square traces on the spots of application mortises on the walls of Argišti’s tomb is the most significant evidence of this.

The red colour streaks in the Kaleköy rock-cut tomb is the same with the colour streaks lined northwardly in three large rows on the ceiling of the main chamber that is opposite of the door of adjoining chamber. The exact tone of the colour is claret red. These red decorations which follows the ceiling of the main chamber and festoonery door cornice of adjoining chamber and festoonery wall cornice over it and ceiling cornice ornamented with festoon indicates that it is an entire wall painting decoration. This wall paintings could be accepted as the obscure extant marks of a rich interior decoration.

Considering the Argišti’s rock-cut tomb, it is plausible to suggest that concave square bronze plates was hanged on the application mortises on the walls of the main chamber of the Kaleköy rock-cut tomb (Figure 14).

Application Mortises of the Main Chamber of the Rock-Cut Tomb

Belli 2003: 24-25; Bilgiç and Öğün 1965: 16, Plate XXI; Burney 1966: 95, Figure 19, Plate XIXa; Çilingiroğlu 2001: Figures 25-26; Dan 2012: 303; Sağlamtimur et.al. 2001: 220, Plate II/19-23, III/51-52. 8 Dan 2012: 304-305, Plate LVIII/7. 7

There are application mortises that have been drilled with 0.40 m to one another on all three side walls 122

Serkan Erdoğan: New Findings on Urartian Rock-Cut Tomb in Mazgirt/Kaleköy Fortress

Figure 13. Application mortises in the main chamber.

Figure 14. Illustration of the concaves square plates related with application mortises.

However, no traces of the objects that are thought to be hanged on the walls of the tomb have been detected.

encountered in the Umudumtepe rock-cut tomb near Erzurum10 and the Doğu Odaları rock-cut tombs in Van Fortress.11

Wooden Architectural Imitation

That bulges of the festoonery cornices that are the most significant architectural ornaments of the rockcut tomb resemble timber peripheral ties, in other

It is known that beside the main chamber of the Kaleköy Fortress rock-cut tomb, festoonery cornices have also been used in Neftkuyu and İçkale rock-cut tombs in Van Fortress.9 Moreover, plain corniche decorations are also 9

Figures 15-18; van Loon 1966: 61. 10 Çilingiroğlu 1980: 198; Sevin 1987: 46. 11 Sevin 1987: 45.

Çevik 2000: 70-71, Figure 10a-c; Forbes 1983: 103; Sevin 1987: 45-47,

123

Systemizing the Past

Figure 15. A house pattern in wooden masonry wall (Çantı) technique (after Güneş 2014: 30).

words, to half round beam ends, so the rock-cut tomb has a wooden imitated style is the common view of many researchers.12 When these festoonery cornices is scrutinized carefully it is seen that not only the ceiling coat of the modelled wooden structure for the rock-cut tomb, but also the side walls also decorated with timber beams or peripheral ties. Such traditional architecture is known as Wooden Masonry Wall (Ahşap Yığma Duvar/ Çantı) construction technique in Anatolia (Figure 15). This structure technique that is generally seen in the Eastern Black Sea region, on stone foundation wooden beams bounded with each other is superposed by using a technique called ‘boğaz geçme’ (bounded beam). In this construction technique, it has been taken into consideration that building sections of the wooden beams and logs are in the same side lengths as in Kaleköy rock-cut tomb. The reason why the ceiling coat of the main chamber of the rock-cut tomb is cascaded can be explained with the inclination of showing the wooden beams, which are called cisir or keran, with festoonery cornices in traditional architecture of Eastern Turkey. If these cornices ornamented with festoon are assumed to be adapted from timber beam ends, it can be said that there was a construction method based on timber in

Eastern Turkey in housing construction in Middle Iron Age. On this sense, Kaleköy rock-cut tomb is a unique structure model that reflects a different construction system for its time. Assessment and Conclusion The rock-cut tomb with its installation and decorations is one of the most astounding examples in Urartu. It is surprising that such a magnificent rock tomb is located in a small border settlement approximately 450 km west of Tushpa, the capital city of the Urartian kingdom. This surprising situation related to the rockcut tomb can be explained by the fact that it is a part of the construction projects that is implemented in the 7th century BC by Rusa, son of Argišti (Figure 16). On the other hand, it should be noted that the settlement structure of Kaleköy Fortress differs from planned cities such as Ayanis, Kefkalesi, Karmir Blur, and Toprakkale built by Rusa. Although the rock-cut tomb and some structures were built during the time of this king, we do not have enough evidence to suggest that Kaleköy Fortress was built completely in the same period. With its peculiarities – the dromos-like entrance, on which there is an Urartian inscription, the door system opening to the well/cistern associated with the tomb, and the decorations of festoonery cornices and

Çevik 2000: 90; Piotrovskij 1966: 306; Schäfer 1977: 253; van Hulsteyn 1981: 18; van Loon 1966: 62. 12

124

Serkan Erdoğan: New Findings on Urartian Rock-Cut Tomb in Mazgirt/Kaleköy Fortress

Figure 16. Urartian inscription at the main entrance of the rock-cut tomb in Kaleköy Fortress.

symmetric niches, this rock-cut tomb become a unique structure among Urartian tombs.

rather than adobe or stone can be explained by the rich forest cover of this region.

The fact of local/regional production of paints in Urartian kingdom is one of the significant results of recent research.13 After chemical analyses it is understood that Urartian wall paints are both inorganic14 and herbal based.15 However, no archaeometric analysis has been done to answer questions on the production and origin of the wall paints discovered in rock-cut tomb of the Kaleköy Fortress.

Burial chambers built of timber side walls and roof is a well-known tomb technique used in Phrygian tumuli (MM, Körte II – III- IV, A, N, KY, P, W, X and Y). The Kaleköy rock-cut tomb which looks like a rock replica of a wooden structure shares the same typological characteristics with the two-chambered Midas Mound Tomb (Great Tumulus),16 the burial chambers of which were orderly constructed on ground level by imbricating cedar and juniper logs.

Festoon architectural decorations on the cornices that are generally 6-8 cm semi-circles give the impression that it has been adapted from wood timber beam ends. So, the rock-cut tomb must be a replica of a wooden construction model. This model structure of the rockcut tomb points to a place ceiling structure and side walls of which were decorated with wooden beams and peripheral ties. Considering the semi-circles of the modelled beam ends of the main chamber of the rockcut tomb it is understood that they are made of thin trunk trees, such as poplar trees. If this construction technique is taken as the criterion, it can be claimed that wooden structures are the dominant building type in Middle Iron Age settlements in Dersim Mountains. The reason why wood is preferred as a building material

Architectural ornaments, wall paintings and remaining patterns of interior decoration show that Kaleköy rockcut tomb is as magnificent as the royal tombs in Tušpa, the capital city of Urartu. Although it is hard to say who was buried in this tomb, it seems logical to think about a kinship of the deceased(s) with the ruling dynasty of Urartu. Bibliography Belli, O. 2003. Van – Aşağı ve Yukarı Anzaf Urartu Kaleleri Kazısı: Bir Ara Değerlendirme (1991–2002)/ Excavations at Van – Lower and Upper Anzaf Urartian fortresses: An intermediary evaluation (1991–2002). Colloquium Anatolicum 2: 1–49.

Keheyan et al. 2018: 89; Ingo et.al. 2013: 4284, 4289; Nunn 2012: 336. Ingo et al. 2013: 4283-4290. 15 Keheyan et. al. 2018: 88. 13 14

16

125

Young 1958a: 7-9; 1958b: 139-154.

Systemizing the Past Bilgiç, E. and Öğün, B. 1965. Second season excavation at Kef Kalesi of Adilcevaz (1965). Anatolia 9: 11–19. Burney, Ch. A. and Lawson, G. R. J. 1960. Measured plans of Urartian fortresses. Anatolian Studies 10: 177–196. Burney, Ch. A. 1966. A first season of excavation at the Urartian citadel of Kayalıdere. Anatolian Studies 16: 55–111. Çevik, N. 2000. Urartu Kaya Mezarları ve Ölü Gömme Gelenekleri. Ankara, Türk Tarih Kurumu Yayınları. Çilingiroğlu, A. A. 1980. An Urartian fortress in Diauehi: Umudum Tepe (Kalortepe). Anadolu Araştırmaları 8: 195–198. Çilingiroğlu, A. 2001. Temple area. In A. Çilingiroğlu and M. Salvini (eds), Ayanis I. Ten Years’ Excavations at Rusaḫinili Eidurukai 1989–1998 (Documenta Asiana 6): 37–65. Roma, Istituto per gli Studi Micenei ed Egeo– Anatolici. Dan, R. 2012. Nails in the Wall: The sikkatu in Urartian toreutic production. In G. P. Basello and A. V. Rossi (eds), Dariosh Studies II. Persepolis and its Settlements: Territorial System and Ideology in the Achaemenid State: 301–311. Napoli, Universitá degli Studi di Napoli L’Orientale. Güneş, M. E. 2014. Geleneksel Ahşap Yapılarda Taşıyıcı Sistem Kurgusunun İncelenmesi: Safranbolu Örneği. Unpublished M.A. thesis. Yıldız Technical University. Forbes, T. B. 1983. Urartian Architecture (British Archaeological Reports, International Series 170). Oxford: British Archaeological Reports. Hovhannisyan, K. L. 1961. Arin-Berd I. Architecture of Erebuni according to 1951-1959 Excavation Materials (Archaeological Excavations in Armenia 9). Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Russian). Hovhannisyan, K. L. 1973. The Wall-Paintings of Erebuni. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Armenian, Russian and English). Huntington, E. and Lehmann, C. F. 1901. Weitere Berichte über Forschungen in Armenien und Commagene. Zeitschrift für Ethnologie 33: 173–209. Ingo, G. M., Çilingiroğlu, A., Di Carlo, G., Batmaz, A., De Caro, T., Ricucci, C., Parisi, E. I. and Faraldi, F. 2013. Egyptian Blue cakes from the Ayanis fortress (Eastern Anatolia, Turkey): micro-chemicaland – structural investigations for the identification of manufacturing process and provenance. Journal of Archaeological Science 40: 4283–4290.

Keheyan, Y., Badalyan, M. and Dan, R. 2018. Some analyses of pigments of Erebuni. Aramazd: Armenian Journal of Near Eastern Studies 12/2: 83–96. Kleiss, W. 1980. Bastam, an Urartian citadel complex of the seventh century BC. American Journal of Archaeology 84/3: 299–304. Köroğlu, K. 1996. Urartu Krallığı Döneminde Elazığ (Alzi) ve Çevresi. İstanbul, Arkeoloji ve Sanat Yayınları. Kroll, S. E. 1972. Excavations at Bastam, Iran. Archaeology 25/4: 292–297. Lehmann-Haupt, C. F. 1910. Armenien Einst und Jetzt I. Berlin, B. Behr’s Verlag. Lehmann-Haupt, C. F. 1931. Armenien Einst und Jetzt II/2. Berlin: B. Behr’s Verlag. Nunn, A. 2012. Wandmalerei in Urartu. In S. Kroll, C. Gruber, U. Hellwag, M. Roaf and P. Zimansky (eds), Biainili–Urartu. The Proceedings of the Symposium held in Munich 12-14 October 2007 (Acta Iranica 51): 321– 337. Leuven, Peeters. Piotrovskij, B. B. 1966. Il regno di Van: Urartu (Incunabula Graeca 12). Roma, Edizioni dell‘Ateneo. Sağlamtimur, H., Kozbe, G. and Çevik, Ö. 2001. Small Finds. In A. Çilingiroğlu and M. Salvini (eds), Ayanis I. Ten Years’ Excavations at Rusaḫinili Eidurukai 1989–1998 (Documenta Asiana 6): 219–250. Roma, Istituto per gli Studi Micenei ed Egeo–Anatolici. Schäfer, H. P. 1977. Die Inschrift Rusa II. Argištehinis in Mazgirt-Kaleköy (UKN 279 = HChl 127). Studi Micenei ed Egeo-Anatolici 18: 249–268. Sevin, V. 1987. Urartu Oda Mezar Mimarisinin Kökeni Üzerine Bazı Gözlemler. In A. Çilingiroğlu (ed.), Anatolian Iron Ages 1. The Proceedings of the First Anatolian Iron Ages Colloquium held at İzmir, 24-27 April 1984: 35–55. İzmir, Ege Üniversitesi Edebiyat Fakültesi Yayınları. Sevin, V. 1994. Three Urartian Rock-Cut Tombs from Palu. Tel Aviv 21/1: 58–67. van Hulsteyn, Y. 1981. Urartian Built and Rock– Cut Tombs, Ph. D. thesis, Bryn Mawr College. van Loon, M. N. 1966. Urartian Art. Its Distinctive Traits in the Light of New Excavations. İstanbul and Leiden, Nederlands Historisch-Archaeologisch Instituut. Young, R. S. 1958a. The Gordion Tomb. Expedition 1: 3–13. Young, R. S. 1958b. Gordion Campaign of 1957: Preliminary Report. American Journal of Archaeology 62: 139–154.

126

Dalarik-1: A New Lower Paleolithic Cave Site in the Republic of Armenia Boris Gasparyan

Institute of Archaeology and Ethnography, NAS RA / Yerevan State University, Yerevan, Armenia

Artur Petrosyan

Institute of Archaeology and Ethnography, NAS RA, Yerevan, Armenia

Phil Glauberman

Institute of Archaeology and Ethnography, NAS RA, Yerevan, Armenia / Catalan Institute of Human Paleoecology and Social Evolution (IPHES), Universitat Rovira i Virgili, Tarragona, Spain / Karls Eberhard University of Tübingen, Tübingen, Germany

Ani Adigyozalyan, Hayk Haydosyan and Soseh Aghaian Institute of Archaeology and Ethnography, NAS RA, Yerevan, Armenia

Makoto Arimura

Tokai University, Kanagawa, Japan

Ellery Frahm

Yale University, New Haven, USA

Samvel Nahapetyan

Yerevan State University, Yerevan, Armenia

Dmitri Arakelyan

Institute of Geological Sciences, NAS RA, Yerevan, Armenia

Jennifer Sherriff

King’s College London, Great Britain

Teo Karampaglidis

MONREPOS, Archaeological Research Centre and Museum for Human Behavioural Evolution, Neuwied, Germany

Masha Krakovsky

Israel Antiquities Authority, Jerusalem, Israel

Ariel Malinsky-Buller

The Hebrew University of Jerusalem, Israel Abstract: Stratified Lower Paleolithic sites are very rare in the Armenian Highlands and the Caucasus, and their fragmentary record limits our knowledge of the initial hominin occupation phases and lifeways in the area during the Early and Middle Pleistocene. The recently discovered Dalarik-1 Cave in Eastern Armenia will add some knowledge to the LP study of the region, containing an in situ and well preserved Acheulian assemblage together with fossilized faunal remains. These factors make Dalarik-1 ideal for ideal for the study of hominin adaptive behaviors as well as timing, duration and regional occupational pattering, which are among the most significant and debated aspects of human evolution in Eurasia. Keywords: Armenian Highlands, Ararat Depression, Mastarahegheghat River, Cave sites, Lower Paleolithic, Acheulian assemblages, Raw material sources

Introduction The earliest evidence currently known of hominin expansions outside Africa was found at the open-air site of Dmanisi in Georgia, dated to c. 1.85–1.77 Ma.1 The 1

Ferring et al. 2011.

Systemizing the Past (Archaeopress 2023): 127–141

later part of the Lower Paleolithic (LP hereafter) record, and mainly the period of the Middle Pleistocene (780 – 300 Ka), is represented within the southern Caucasus and the Armenian Highlands by a large number of open-air and a few cave sites that, for now, remain incompletely chronologically and environmentally

Systemizing the Past

Figure 1. Location map of Lower Paleolithic sites in the Armenian Highlands and the southern Caucasus with the obsidian sources utilized by hominins at Dalarik-1.

contextualized (Figure 1).2 A limited number of openair LP sites are securely dated to this time period.3 The fragmentary record of LP occupation in this region raises fundamental research questions: what were the nature and timing of hominin dispersals into the region? Did environmental constraints condition occupations in the region? What were the adaptive responses of these early hominins, and how did they manifest in their subsistence strategies and mobility patterns?

of High Medieval period cultural remains (ceramic fragments and a fireplace) under the rock shelter show that the cave served as a short-term shelter probably for shepherds or transhumance. The Site of Dalarik-1 Dalarik-1 (40°13’8.82”N, 43°53’38.04”E, 930 masl) is located adjacent to the Ararat Depression in the Mastarahegheghat River canyon, a tributary of the Araxes River, in the Armavir Province of Armenia. The area is part of Mt. Aragats volcanic province, which has been formed by a suite of several mafic lava flows and pyroclastic deposits which can be traced along the Mastarahegheghat River canyon at its southernmost fringes (Figures 2 and 3).4 Dalarik-1 is a small, corridortype cavern (7.6m long, 3.8m wide and 2.3 m high) around 19 m2 in surface area under the rock shelter, which was formed as a result of mechanical weathering of the lava flow. The opening looks toward the southeast. Nowadays it is dry inside of the cave, but at the

The recently discovered Dalarik-1 Cave in Eastern Armenia will add some knowledge to the LP study of the region, containing an in situ and well preserved Acheulian assemblage together with fossilized faunal remains, representing an early occupation phase of the Mt. Aragats volcanic province. Dalarik-1 has also yielded a limited amount of obsidian artefacts, which can be attributed to the Upper Paleolithic or at least to the Early Holocene, and can also be of great interest. Traces 2 E.g. Doronichev and Golovanova 2010; Gasparyan et al. 2020; Lyubin 1989; Mgeladze and Moncel 2016. 3 E.g. Adler et al. 2014; Gasparyan et al. 2014; 2020.

4

128

Avagyan et al. 2018; Gevorgyan et al. 2018.

Boris Gasparyan et al.: Dalarik-1

Figure 2. Geological map of the Mastarahegheghat River canyon and location of the Dalarik-1 Cave.

129

Systemizing the Past very end of its opening, traces exist of a small spring, which was probably active during the Late Pleistocene or Early Holocene. There is a large (app. 50 m2) platform in front of the cave, which is one of the terraces of the

Mastarahegheghat River. The terrace is at a hight of 6m above the current riverbed. Traces of a stone-built fence or enclosure are noticeable at the beginning of the platform (Figures 3-4 and 6/1).

Figure 3. 1. Aerial view of the Mastarahegheghat River canyon; 2. Location of Dalarik-1 Cave; 3. Main view of the site.

130

Boris Gasparyan et al.: Dalarik-1

Figure 4. Topographic map with location of Dalarik-1 Cave along the Mastarahegheghat River canyon.

131

Systemizing the Past

Figure 5. The grid locating the test Trenches 1–3, excavated during the 2017 and 2018 field seasons.

132

Boris Gasparyan et al.: Dalarik-1 The site was discovered by an Armenian-Japanese joint expedition in 2017 (Institute of Archeology and Ethnography, National Academy of Sciences, Republic of Armenia and Tokai University). Two test trenches (Trenches 1 and 2) were excavated inside the cave during the 2017 season. Before the excavations, a 1 x 1 m grid was set up on the surface of the site. Trench 1 includes Squares E3-G3 (Figures 4 and 5). After removing the aeolian sediments, this trench was excavated to a depth of 10-25cm, exposing basaltic bedrock. In the limits of Squares F3 and G3, a medieval fireplace constructed with basalt slabs was uncovered (Figures 4-5, 6/2-3 and 7). Ash and charcoal remains, mixed with aeolian sediments, zoogenic humus, and cave earth, were recorded around and inside of the feature. As a whole, from both squares nine artifacts were found: two medieval ceramic fragments and seven obsidian pieces.

at the top (represented by zoogenic humus mixed with aeolian sediments). Level 2 also contained ash, charcoal remains, and ten obsidian artefacts (Figures 4-5, 6/4 and 7). A charcoal sample taken from the depth of 35cm yielded a 14C date of 1050 ± 20 BP (IAAA-171292) which, after calibration, appears in the range of 969 – 1024 cal AD at 95.4% confidence. This calibrated age fits very well with the medieval fireplace and the ceramic fragments, showing that the cave served as a short-term shelter at the beginning of the High Medieval period (10th-11th centuries AD). The 19 obsidian finds which are mainly represented by knapped nodules, debris, and chunks also contain four diagnostic pieces: an obsidian blade, a resharpening flake, a burin, and a carinated scraper (Figure 8). These finds can be attributed to the Upper Paleolithic or at least to the Early Holocene, which might represent the second occupational phase of the site.

Trench 2 was excavated directly at the entrance of the cave within the limits of Square I4. Bedrock, represented by basaltic blocks and debris, was reached at the depth of 35-40cm, and it underlies Level 2 (cave earth composed of yellowish-brown silt and clay) and Level 1

Test excavations at the site were continued in 2018 on a terrace in front of the small rock shelter between the two preserved portions of the fence (Trench 3) covering an area of 4m2 (Squares N6-N7, and O6-O7). 40-cm-

Figure 6. 1. The entrance of Dalarik-1 Cave and stone built fence; 2. Trench 1 excavated inside the cave (Squares E3-G3); 3. Medieval fireplace constructed with basalt slabs uncovered in the Squares F3 and G3; 4. Trench 2 excavated within the limits of Square I4.

133

Systemizing the Past

Figure 7. Plan and section of the 2017 excavation.

134

Boris Gasparyan et al.: Dalarik-1

Figure 8. Obsidian artifacts from 2017 excavation (Trenches 1 and 2, Squares G3 and I4): 1-1a. Blade; 2-2a. Burin; 3-3a. Resharpening flake; 4-4a. Carinated scraper (Upper Paleolithic or Early Holocene).

135

Systemizing the Past

Figure 9. 1-2. Trench 3 exposed within the limits of the Squares N6-N7, and O6-O7; 3-4. Massive basalt bifaces upon excavation in Trench 3.

thick deposits were exposed, which overlie the basaltic bedrock (Figures 4-5 and 9-10).

modification on their ventral surfaces (Figures 11 and 12). Among the flake tools, side-scrapers and transversal scrapers are the most common types in the assemblage (Figures 12 and 13). The scrapers were intensively reduced, as indicated by the semi-stepped retouch that displays many generations of modifications and rejuvenation. As a whole, the Dalarik-1 assemblage from Trench 3 can be attributed to the Acheulian, representing the initial occupation phase of the site.

A small artifact assemblage was unearthed from this test excavation, comprising a total of 191 lithic artefacts produced from three raw materials: obsidian (n=150), basalt (40), and dacite (1). The lithic assemblage contains diagnostic LP artefacts (bifaces) (Figures 11 and 12). In addition, a few fossilized faunal remains were found, including fragments of large mammal bones, possibly horse taxa.

The entire obsidian artifact assemblage was analyzed using portable X-ray fluorescence (pXRF) to find elemental correlations among the obsidian artifacts excavated at Dalarik-1 and their matching volcanic sources. This enabled the reconstruction of the mobility of the site’s occupants. The results demonstrate that 96.7% (n = 146) of the obsidian artefacts derive from the sources of Pokr Arteni, c. 18 linear km from the site. Three chips are made of Gutansar obsidian. One appears to match the Kars-Digor source (located in Eastern Turkey), and a flake (larger than 2cm) has no immediate geochemical match. The distances from Dalarik-1 to the Kars-Digor and Gutansar sources are about 45 and 75 linear km, respectively (Figure 1). However, on foot,

The lithic assemblage from Dalarik-1 is well preserved, with no indications of rolling or other secondary modifications, despite the proximity of the artefacts to the modern surface. Moreover, almost half of the excavated lithic sample comprises small flakes (1-1-7%), rounded to sub-angular feldspar (mainly plagioclase and microcline, accompanied by orthoclase) and quartz grains up to 300-600 μm (1>5%), sub-rounded to rounded sandstone up to 700-950 μm (>1-3%).

2. Group B: Group B fabrics are heterogeneous, sometimes lumpy, mainly sandy and porous or very porous.

Fabric Kg A22-3. Typical A clay with angular grog grains (220-1300 μm, 5%) and unidentified organic material ‘ghosts’ (300-700 μm, %). Some of the vegetal material ‘ghosts’ contain phytolith traces.

Fabric Kg B. The matrix is grey to dark grey (PPL), reddish to dark brown or almost opaque (XPL), marly and sometimes exhibits weak optical orientation. The silty component (>3-5%) includes mainly feldspar, accompanied by hornblende, mica, zircon and opaque minerals. The non-plastic component (25%) includes mainly badly sorted sand of sub-angular to rounded basalt (300-2600 μm, 15%) and andesite (300-1400 μm, 5%), angular to rounded feldspar (plagioclase, microcline, anorthoclase and rare orthoclase; 3->7%) grains up to 530 μm. The temper also includes sub-angular to sub-rounded clinopyroxene and orthopyroxene (3-20%), angular to sub-rounded feldspar grains up to 500 μm (rarely up to 1000 μm, 3->7%) and rounded basalt and andesite grains up to 500 μm (very rare up to 2000 μm; 7%), accompanied by sub-angular to sub-rounded pyroxene up to 1000 μm (>1-3%), rounded sandstone up to 900 μm (1%), rounded quartz up to 200 μm (1%), mica and hornblende up to 100 μm (1-3%) and rare rounded basalt grains up to 700 μm.

Fabric Kg BB3. BB matrix, silt and temper, and angular to sub-angular grog fragments (600-1300 μm, 7%). Fabric Kg BB22-3. The temper includes angular to subangular shale up to 1200 μm (>7%), sub-rounded to rounded sandstone (230-1400 μm, 3%), angular grog (600-1200 μm, 5%), angular to sub-rounded feldspar up to 490 μm μm (3%), limestone up to 1500 μm (3%), sub-angular to sub-rounded pyroxene up to 400 μm ( eiš (ēš), and *epa > ewa, see Petrosyan 2012: 147-151, with references. For the closeness or even identity of the images of Teššub/Teišeba and Armenian thunder god Vahagn, see Petrosyan 2020a: 47-51. 6 Martirosyan 1973: 22-24. 7 Cf. Khachatryan 1975: Figures 53-54, 57, 61-62. See the Figures 6-7 of L. Abrahamian’s article in this volume. 8 Petrosyan 2015. 9 For the fish-like serpent, adversary of the thunder god in some IndoEuropean (Hittite, Armenian, Norse) traditions, see Petrosyan 2015: 33-36, 38. 10 For the K‘aǰs as the Männerbund of Vahagn, see Petrosyan 2018b. 11 Emin 1873; Ivanov 1969; Petrosyan 2020b, with references. 12 Petrosyan 2018b, cf. also the figure of the patriarch Aram of the ethnogonic myth, the epicized version of the thunder god, who performs his exploits as the head of an army of norati ‘youthful’ warriors, see Petrosyan 2011: 343.

13 For those paleo-demographic and inventory data of the early Iron Age societies in Armenia, see Avetisyan 2014: 57, 65; Avetisyan and Mkrtchyan 1994; Badalyan and Aghekyan 1993; Gevorgyan 2010: 203; Mkrtchyan 2001: 54-56. 14 Petrosyan 2009. 15 Petrosyan 2018a: 158-174.

387

Systemizing the Past echo of the times when the Hurro-Urartians appeared in Syria and adjacent regions populated by the West Semites. This is comparable with the Armenian myths of ethnogonic patriarchs Hayk and Aram, epicized thunder gods, Armenian doubles of Teššub, who kill their Babelonian and Syrian adversaries Bel and Baršam (i.e., the gods Bel-Marduk and Baal-Šamin), respectively.16 It is interesting that at the end of the Bronze Age a Greek-Achaean population probably appeared in Cilicia.17 Cilicia is one of the locales where Teššub’s myths were concentrated,18 which may indicate where and when the mythologem of Theseus was borrowed by the Greeks.

Gevorgyan, A. Ts. 2010. Metalwork in Armenia in the pre-Urartian period. In Avetisyan P. S. and Petrosyan A. Y. (eds), Through Haldi’s Power… Studies in Honour of the 100th Anniversary of the Birth of Boris Piotrovsky: 191–211. Yerevan, Publishing House ‘Gitutyun’, NAS RA (in Russian). Ivanov, V. V. 1969. Using for etymological studies onestem words in the ancient poetry in Indo-European languages. In O. N. Trubachev (Editor-in-Chief), Zh. Zh. Varbot, L. A. Gindin, G. A. Klimov, V. A. Merkulova and V. N. Toporov (eds), Etymology 1967: 40–56. Moscow, Nauka (in Russian). Ivanov, V. V. and Toporov V. N. 1974. Studies on Slavic Antiquities. Moscow, Nauka (in Russian). Karakhanyan, G. H. and Safyan, P. G. 1970. The Rock Paintings of Syunik. Yerevan, Press of the Academy of Scienses of the Armenian SSR (in Armenian). Khachatryan, T. S. 1975. Ancient Culture of Shirak (III-I Millennia BC). Yerevan, Yerevan University Press (in Russian). Martirosyan, H. A. 1973. The Prehistoric Hieroglyphs of Armenia and Their Urarto-Armenian Doubles. Yerevan, Press of the Academy of Scienses of the Armenian SSR (in Armenian). Martirosyan, H. A. 1978. Sciеnce Begins in Prehistory. Yerevan, Soviet Writer (in Armenian). Mkrtchyan, R. A. 2001. Paleoanthropology of the Hоrom Burial Ground. Yerevan, Zangak-97 (in Russian). Laroche, E. 1976-1977. Glossaire de la langue hourrite (Revue hittite et asianique 34-35): 13–161. Paris, Klincksieck. Petrosyan, A. Y. 2009. Forefather Hayk in the light of comparative mythology. Journal of Indo-European Studies 37/1-2: 155–163. Petrosyan, A. Y. 2011. Armenian traditional Black Youths: The earliest sources. Journal of Indo-European Studies 39/3-4: 342–354. Petrosyan, A. Y. 2012. The cities of Kumme, Kummanna and their god Teššub / Teišeba. In M. Huld, K. JonesBley and D. Miller (eds), Archaeology and Language: Indo-European Studies Presented to James P. Mallory (Journal of Indo-European Studies, Monograph 60): 141–156. Washington, DC, Institute for the Study of Man. Petrosyan, A. Y. 2015. Thirty years later: the dragon stones and the dragon-fighting myth. In A. Y. Petrosyan and A. A. Bobokhyan (eds), The Vishap Stelea: 13–52․ Yerevan, Publishing House ‘Gitutyun’, NAS RA (in Armenian). Petrosyan, A. Y. 2016. Indo-European *wel- in Armenian mythology. Journal of Indo-European studies 44/1-2: 129–146. Petrosyan, A. Y. 2018a. The Problem of Armenian Origins (Journal of Indo-European Studies, Monograph 66). Washington, DC, Institute for the Study of Man. Petrosyan, A. Y. 2018b. From Armenian demonology: the K‘aǰs. Journal of Indo-European Studies 46/1-2: 206–218.

Bibliography Abrahamian, L. 2004. ‘Axe-bull’ in the labyrinth: towards clarification of one Broze and Iron-Age composition. In S. Hayrapetyan (ed.), Historical-Cultural Legacy of Shirak, Proceedings of the 6th Republican Scientific Conference: 12–14. Gyumri, The Shirak Center for Armenological Studies (in Armenian). Avetisyan, P. S. 2014. Armenian Highland in the 24th-9th centuries BC (Dynamics of Socio-Cultural Transformations according to Archaeological Data). Scientific report for the degree of Doctor of History. Yerevan, Institute of Archaeology and Ethnography, NAS RA (in Armenian). Avetisyan, P., Dan, R. and Petrosyan, A. 2018. Axes, labyrinths and astral symbols: bronze pendants and pins from the Armenian Highlands. Ancient West and East 17: 27–63. Avetisyan, P. S. and Mkrtchyan, R. A. 1994. Experience of socio-demographic analysis of Early Iron Age Talin burials. In Scientific Session Devoted to the Results of Archaeological Fieldworks in 1991-1992: 24–27. Yerevan, Institute of Archaeology and Ethnography, AS RA (in Armenian). Badalyan, R. S. and Aghekyan, O. K. 1994. Еxcavations of the Horom necropolis. In G. A. Tiratsyan, A. A. Kalantaryan and G. Y. Areshian (eds), Аrchaeological Works in New Buildings of Armenia: 67–74. Yerevan, Academy of Sciences, RA (in Armenian). Bryce, T. 2016. The land of Hiyawa (Que) revisited. Anatolian Studies 66: 67–79. Diakonoff, I. M. 1990. Archaic Myths of East and West. Moscow, Nauka (in Russian). Emin, N. O. 1873. Vakhagn-Vishapakah of Armenian mythology is Indra-Vritrahan of the Rig-Veda. St. Petersburg, Academy of Sciences (in Russian). Gander, M. 2012. Ahhiyawa – Hiyawa – Que: Gibt es Evidenz für die Anwesenheit von Griechen in Kilikien am Übergang von der Bronze- zur Eisenzeit?  Studi Micenei ed Egeo-Anatolici 54: 281–309. Petrosyan 2009: 160-161. Bryce 2016; Gander 2012; Yakubovich 2015. 18 See, e.g., Laroche 1976-1977: 279. 16 17

388

Armen Y. Petrosyan: ‘Axe-Bull’ – Order of the Thunder God Petrosyan, A. Y. 2020a. The Oldest Sources of the Armenian Epic 2. Yerevan, Institute of Archeology and Ethnography, NAS RA (in Armenian). Petrosyan, A. Y. 2020b. Saussurean anagrams of the ‘song of Vahagn’. Indo-European Linguistics and Classical Philology 24: 181–189 (in Russian).

Yakubovich, I. S. 2015. Phoenician and Luwian in Early Iron Age Cilicia. Anatolian Studies 65: 35–53.

389

Overlooking the River Hrazdan Valley: The Fortified Site of Tghit in the Tsaghkunyats Mountains, Kotayk Region, Armenia* Artur Petrosyan

Institute of Archaeology and Ethnography, NAS RA, Yerevan, Armenia

Roberto Dan

Tuscia University and ISMEO – The International Association for Mediterranean and Oriental Studies, Rome, Italy

Priscilla Vitolo

University of Naples L’Orientale and ISMEO - The International Association for Mediterranean and Oriental Studies, Rome, Italy

Onofrio Gasparro

Ca’ Foscari University of Venice, Italy

Boris Gasparyan

Institute of Archaeology and Ethnography, NAS RA and Yerevan State University, Yerevan, Armenia Abstract: During eight years of activity carried out by the Armenian-Italian archaeological mission, the Kotayk Survey Project, 174 archaeological sites have been investigated in detail. Among the pre-medieval sites, one of the most important is undoubtedly the Tghit complex. This is characterized by a fortified structure of impressive dimensions associated with an artificial lake. Tghit controlled the entire valley of the River Hrazdan from its highly strategic position. The goal of this contribution is to describe the structures investigated and discuss their chronology and function. Keywords: Armenian Archaeology, Kotayk Region, River Hrazdan Valley, fortified shelters, Iron Age

Introduction During eight years of activities (2013-2020), the Kotayk Survey Project (KSP), an Armenian–Italian archaeological expedition, has developed an extensive survey along with several excavations, of which the most important is that of the multi-period site of Solak-1/Varsak (KSP016), discovered during the mission’s activities.1 The survey has been conducted *

We wish to express our gratitude and admiration to Pavel Avetisyan, director of the institute and colleague. The Armenian-Italian Kotayk Survey Project, within which the site discussed in this article was studied, was born, developed and grew with his help and constant support. Many of the results we have achieved have been possible thanks to his institutional and expert assistance. We therefore wish to dedicate this paper to him on the occasion of his 65th birthday. We would also like to remember here Gregory Areshian (1949-2020), who made the initial studies of the site described in this paper. 1 The activities are conducted under a scientific cooperation agreement between the Archaeology and Ethnography Institute of the Academy of Sciences of the Armenian Republic and ISMEO-The International Association for Mediterranean and Oriental Studies. The project is also funded by the Ministry of Foreign Affairs and International Cooperation (MAECI) of the Italian Republic and by Progetto MIUR ‘Studi e Ricerche sulle culture dell’ Asia e dell ‘Africa’. The authors express their deepest thanks to Prof. Pavel Avetisyan, Director of the Institute of Archaeology and Ethnography of the Academy of the Sciences NAS RA. From the Italian side, conduction of the work has been possible thanks to the unstinting assistance of the Italian institutions that support and finance the project. In

Systemizing the Past (Archaeopress 2023): 390–405

in particular along the River Hrazdan and the River Marmarik, the most important of its tributaries.2 During this time the mission has investigated 174 archaeological sites that span from the Palaeolithic to the Late Medieval period. They are of various types, from open-air prehistoric sites to fortified structures, settlements and cemeteries dated to the protohistoric period, as well as fortified sites from the Urartian era, medieval castles and monastic complexes. Among the several fortified structures that can be attributed to the protohistoric period, one of the most significant is the Tghit fortress that is examined in detail in this this context, we would like to thank in particular the president of ISMEO, Prof. Adriano V. Rossi, the Italian Ministry of Foreign Affairs and International Cooperation (MAECI), the Italian Embassy and the Ambassador Vincenzo Del Monaco. We would like also to thank Artak Avetisyan, member of the IAE NAS RA and irreplaceable mission member and travel companion with whom the site of Tghit discussed here has been visited. The contents of this article have been developed jointly by the authors. Specifically, A. Petrosyan wrote ‘History of Studies’ and ‘Geographical Location’, R. Dan wrote ‘The fortified complex’, P. Vitolo wrote ‘The pottery’, and B. Gasparyan wrote ‘The Artificial Lake’. Introduction and Conclusions were written jointly by all authors. O. Gasparro is responsible for the plan of the site, the ortophotographs and the digital elevation models (Figures 3, 6, 7 and 14). R. Dan is the author of Figures 1, 2, 9. All the ground and aerial pictures are from the Kotayk Survey Project Archive. 2 On the survey activities, see Petrosyan et al. 2020a: 211-212 and Petrosyan et al. 2020b.

Artur Petrosyan et al.: Overlooking the River Hrazdan Valley

Figure 1. Map of the central part of the Kotayk Region, with the position of the principal archaeological sites and modern centres (prepared by R. Dan from a satellite picture taken from Google Earth).

information as part of the preparation of an inventory of historical and cultural monuments of the peoples living in the USSR. Activities were conducted in AugustSeptember 1975 with a surface survey in Ashtarak and Nairi4 regions. The several sites investigated there included Tghit, one of the highest-altitude fortified structures in the area. The investigators identified a complex composed of a fortress, an artificial lake, a burial ground and traces of Late Medieval settlement. The fortress was tentatively dated on the basis of its architectural features to the Late Bronze Age and Early Iron Age, a proposal that corresponds to the attribution given in the National List of Monuments. The fortified complex was said to possess three rows of fortifications. The upper defensive alignment (Wall-1) stands out for its impenetrability; it runs around the edges of the flat hilltop and is strengthened by four massive rectangular towers. On the west side of this wall, between the two towers there is a 6 m wide gate. The dimensions of the central part of the upper fortified structure are about 130 m in length (north-south) by about 60 m wide (east-west). The other two walls, which are placed on the slope of the hill, in addition to their protective functions acted as retaining walls for artificial terraces where there are also ruins of residential buildings. An artificial lake had been created not far away and was still

paper. Its uniqueness is mainly linked to two factors, its geographical location and its extraordinary state of preservation. The site was visited and studied twice during KSP activities and given the code KSP036. The preliminary visit was performed in 2014, to establish the overall characteristics of Tghit and collect the initial data on the architectural and environmental context. On this occasion a very small test excavation was performed to identify diagnostic archaeological finds that could help us to date the archaeological structures. A more extensive study mission was conducted in 2018, when a detailed study of the architectural remains and the aerial documentation of the entire site was made in order to produce the plans, sections etc. that accompany this text. In general, the site has never been systematically studied; sporadic references can be found in previous studies3 and some information has already been obtained as part of the Kotayk Survey Project. Tghit is in the national monuments list (Code: 6.32.1), where it is recorded as a fortified site dating from the 13th - 9th century BC to the Middle Ages. History of Studies Earlier studies were carried out by a joint archaeological expedition of the Institute of Arts and the Institute of Archaeology and Ethnography of the Academy of Sciences of Armenia SSR, and the Centre for Armenian Studies of Yerevan State University, which collected 3

The district of Nairi was an administrative-territorial unit within the Armenian SSR that existed from 1972. It was abolished in 1995 during Armenia’s transition to a new administrative-territorial division. The Nairi region was located on the Yeghvard plateau, with an altitude of 900-2577 m a.s.l.

4

Areshian et al. 1977: 83-84; Simonyan 2012: 35.

391

Systemizing the Past

Figure 2. View of the central part of Kotayk Region from the site of Tghit, indicating the most significant mountains visible (prepared by R. Dan).

functioning. Traces of late medieval settlement can be seen in the central part of the upper fortified complex. In conclusion, it was said to be the best-preserved preUrartian monument encountered by the expedition.5 Evidently no diagnostic material such as pottery was collected during the expedition, and the date was attributed on the assumption that so-called ‘cyclopean’ fortifications may usually be assigned to the Bronze or Iron Ages. In any case, after this investigation, the site substantially disappeared from specialist literature until the start of the Kotayk Survey Project, which studied it in 2014 and 2018.6 Geographical Location The site is located on a rock outcrop on one of the southern Tsaghkunyats Range peaks, on the border between Kotayk and Aragatsotn regions (Figure 1). It lies on the flat top of a conical hill surrounded by oak forests,7 in a landscape substantially untouched by anthropic exploitation, except for pastoral activities such as transhumance. The position is highly strategic due to the fact that the site overlooks the entire Hrazdan Valley, with a perfect view stretching from the Gegham Range to the Ararat Depression (Figure 2). The area is naturally protected on the west by the last part of the relief. The site is located 2 km north-west of the village of Teghenik, formerly called Tghit, from which a poorly transitable dirt road leads to a path through the forest, reaching the site after 3.35 km. The structures identified here are distributed on a northsouth oriented plateau measuring about 900 × 200 m. Due to the considerable elevation of the small plateau, which slopes 50 m from north (2180 m a.s.l.) to south (2130 m a.s.l.), the climatic conditions in the area are particularly harsh, with winter temperatures reaching many degrees below zero and heavy snowfalls. The geological landscape is characterized by outcrops of igneous rocks such as trachyte and rhyolite, with which the walls of the fortified structure were built. The Fortified Complex The site consists of a fortified complex and an artificial lake (Figure 3). The fortified complex of Tghit (coordinates: 40°26’13.56”N 44°33’17.69”E; elevation: 2185 m a.s.l.) is located on an isolated rock spur in the south-east part of the small plateau, which rises about 15 m above the level of the plateau (Figures 4-8). The entire complex measures8 about 140 × 120 m and covers Areshian et al. 1977: 83-84, Figure 4. On the investigations in Tghit in the frame of the Kotayk Survey Project, see Petrosyan et al. 2015: 62, 64, Plate XXII/2; Petrosyan et al. 2020a: 219-220, Figures 8-11. In this paper, some of the information and the interpretation previously provided have been revised and updated. 7 Areshian et al. 1977: 83. 8 These measurements have been rectified from the ones presented in the previous publications. 5 6

392

Artur Petrosyan et al.: Overlooking the River Hrazdan Valley

Figure 3. Orthophotograph and digital elevation model of the fortified complex and the artificial lake in Tghit (produced by O. Gasparro).

Figure 4. Vertical aerial view of the Tghit fortified complex (Kotayk Survey Project Archive).

393

Systemizing the Past

Figure 5. Aerial view of the Tghit fortified complex with the River Hrazdan Valley in the background (Kotayk Survey Project Archive).

Figure 6. Rectified photograph of the fortified complex of Tghit (produced by O. Gasparro).

394

Artur Petrosyan et al.: Overlooking the River Hrazdan Valley

Figure 7. Plan of the fortified complex of Tghit: in black the three fortification lines; in red later structures (drawn by O. Gasparro).

an area of 1.25 ha. The fortified complex is composed of three lines of fortification placed at different heights, built on the west side, the most naturally exposed. The highest walled part (Wall-1), the first to have been built, protects the west, south and east sides of the upper part

of the natural rock outcrop. The landform itself affords protection to the northern side. The size of this upper enclosure is 136 × 75 m, with an area of 0.66 ha. The perimeter of this area is about 325 m. The total length of the western section of Wall-1, the biggest one that 395

Systemizing the Past

Figure 8. Aerial picture of the eastern side of the Tghit fortified complex, with mostly collapsed walls (Kotayk Survey Project Archive).

is slightly curved, is 136 m; the southern portion of Wall-1, which is linear without buttresses, is about 38 m long. The eastern part of Wall-1, the worst preserved and indeed mostly collapsed, is about 100 m in length (Figure 8). The width is irregular, differing according to the protection offered by the natural formation. The walls were built in complete conformation to the shape of the rock outcrop; it seems that the bedrock was not removed to host them. The foundations rest directly on the bedrock, a circumstance that has enabled the excellent degree of preservation notwithstanding the frequent earthquakes in the region. All the walls of the complex were built using local igneous stone. They are double-faced and dry-stone; no mortar has been identified in the complex, with the use of large stones for the outer faces and an infill of medium and smallsized loose stones. As mentioned, they vary in width. The walls on the western side, that most naturally exposed, have various widths generally between 3 and 4 m; in the points where there are the buttresses this may reach more than 10 m. The south and east walls have an average width of between 2 and 3 m. Both these sides are naturally protected by the very steep slope which overlooks the village of Teghenik and the entire Hrazdan River Valley. The western wall of the upper line of fortifications (Wall-1) is the best preserved and in some points is 6 m high (Figure 9). The upper fortifications (Wall-1) are equipped on the western side with a total of five large, 13-15 m long buttresses, which protrude about 3 m from the face of the wall. These buttresses are irregularly spaced, between 13 and 18 m apart. Two

of them (Buttress-3 and Buttress-4) stand on either side of the main gate. In some parts of Wall-1 clear evidence of rebuilding is visible. For example in the southern corner of Buttress-3 and in the southern part of the main gate, small stones have been used to repair the wall. The contrast between the large stones used for the original masonry and this later intervention is evident (Figures 10-11). Some sections of Wall-1 have collapsed, for example between Buttresses-2 and 3, and between Buttress-3 and 4, as well as the north-western corner of Buttress-3. In most cases these collapses involve not the original masonry but the later rebuilds. A gate 6 m wide is present in the middle of the upper fortification on the western side (Figures 12-13). The defence of this gate was facilitated by the different position—some metres further south—of the second fortification gate (in Wall2), which would have forced enemies to make a sharp 180° degree turn. On the southern side, the main gate was reinforced by one buttress (Buttress-4). Despite the presence of dense vegetation, which precludes the complete comprehension of the structures, in the southern part of the upper enclosure wall remains are clearly discernible. These rooms, at least thirteen, are mostly attached to the western upper fortification walls, and most of them are contiguous rooms, square (2 × 2 m) or rectangular (4.50 × 2.50 m; 2.50 × 2 m), located just south and north of the main gate (Figure 9). An isolated oval building (4 × 3 m) is recognisable in the central part of this area. All these structures seem to have reused stones that have fallen from the main fortification. The total perimeter of the upper fortified line, including the 396

Artur Petrosyan et al.: Overlooking the River Hrazdan Valley

Figure 9. Orthophotograph of the western side of the Wall-1 of Tghit fortified complex showing the five buttresses and gate. Note on the right, to the south, the rectangular and elliptical buildings mostly abutting the walls (Kotayk Survey Project Archive; graphic processing by R. Dan).

Figure 10. View from the ground of Buttress-3 on the western side of Wall-1 in Tghit. Note on the right corner of the buttress, to the south, the small stones used for later rebuilds (Kotayk Survey Project Archive).

397

Systemizing the Past and at the southern corner of the southern buttress (Buttress-5) of Wall-1. As mentioned above, the gate of the second fortification wall stands some metres south of the gate of the upper fortification, in order to better control access to the top of the fortress. It proved impossible to obtain exact measurements due to the large amount of stone debris fallen from the collapsed walls. The lower fortification wall (Wall-3), was built directly joined to Wall-2, and has a total length of about 170 m. It was not possible to identify the gate due to the presence of stone debris and vegetation. Scattered, irregular structures are visible just outside the fortress, in correspondence to the junction between Wall-2 and Wall-3. Due to the use of the same building materials it is difficult to frame chronologically the epoch of construction of the three walls. It is evident that Wall-1 was built before Wall-2, since the latter is built over the previous one. Similarly, Wall-3 is on a side laying on the Wall-2. How much time may have elapsed between the construction of these walls is not known; Walls-2 and 3 seem slightly coarser in construction technique and have no real buttresses. The Artificial Lake

Figure 11. View from the ground of the southern corner of Buttress-4 (Kotayk Survey Project Archive).

An elliptically shaped artificial water reservoir is located about 150 m north of the fortified complex (coordinates: 40°26’23.15”N 44°33’14.17”E; elevation: 2166 m a.s.l.), in the central part of the natural plateau (Figures 3, 14). It measure 110 × 200 m, with a surface area of 1.45 ha. It was obtained by the construction of an earth embankment reinforced by large stone blocks; the embankment is 475 m long, has an average height of 1 m and a width of about 5 m. It must be borne in mind that the basin has been partly filled by sediments carried by the water, so the original dimensions of the embankment, as well as the depth of the basin, are not verifiable. The reservoir, which is still in operation today, is mainly fed by the melting of snow from the surrounding hills. The water was largely channelled and carried into the basin by an artificial channel, still clearly visible on the ground, coming from the north. The canal is about 1 km in length, from the upper part of the hill above the basin, 1 to 2 m wide, with a depth of about 0.50-0.70 cm. This too has been largely infilled with water-borne sediments. The canal starts form a plateau about 80 m higher than the reservoir level, which is where most of the water generated by the melting snow must have come from. The canal enters the basin at its northern edge, but at that point it forks, with a branch that runs along the basin on the east and then continues southwards, passing close to the fortified complex and down towards the hills beneath. The quantity of water produced was greater than that which could be contained by the artificial lake, so a part was directed to the land beneath. It is clear that the canal and artificial basin are absolutely not datable

northern natural cliff, is about 307 m, for an enclosed area of 0.6 ha. It seems quite evident that these structures were built much later that the external walls and that there were originally no structures inside this area, a circumstance confirmed by the outcropping in most places of the natural bedrock. Very few potsherds, unfortunately not diagnostic, were collected here, a situation due to the dense plant cover at the times of our visits. The less protected western side of the natural hill descends gradually toward a large depression. In this part two more lines of fortifications were built. The complex is protected by a total of 630 m of walls (Wall-1, 2 and 3). The second (Wall-2) and third (Wall-3) defensive walls are thinner, about 2-3 m in width. These, again double-faced dry-stone walls, seem to be linear, although in some sections there are curved protrusions presumably dictated by topographic rather than defensive needs. These curved protrusions resemble a similar situation seen in some protohistoric fortified complexes in Samtskhe-Javakheti Region, Georgia.9 Judging by their construction technique, these walls seem to be contemporary with the upper wall (Wall-1). The median wall (Wall-2) has a length of about 187 m and is joined to the upper western wall at the northern corner of the northern buttress (Buttress-1) of Wall-1 For example the sites of Abuli (Dan et al. 2019; Licheli et al. forthcoming) and Didi Khanchali (Licheli et al. 2022) and the site KSP084 identified and studied in the Kotayk Survey Project (Petrosyan et al. 2020a: 138-139, Figure 4).

9

398

Artur Petrosyan et al.: Overlooking the River Hrazdan Valley

Figure 12. Aerial view of the gate in the western part of Wall-1, between Buttresses-3 and 4 (Kotayk Survey Project Archive).

Figure 13. Vertical overhead view of the gate in Wall-1 (Kotayk Survey Project Archive).

399

Systemizing the Past

Figure 14. Orthophotograph and digital elevation model of the artificial basin in Tghit (produced by O. Gasparro).

are quite common in the Late Bronze Age but also closely resemble material from the Solak-1 site where the assemblage belongs mostly to the earlier phase of the Middle Iron Age.11 In this case, the impossibility of deducing whole vessel forms limited their attribution to a more precise period. All the pottery collected was in every-day use, mainly common or kitchen ware. With regard to the medieval period, a fragment of glazed pottery could be dated to the 13th-14th centuries AD.12

from their structural characteristics. Their construction must have been associated with that of the fortified complex. The latter, however, undoubtedly had various phases of use ranging from the protohistoric era to the Middle Ages and currently it is not possible to link the artificial lake to one specific period of use of the fortified complex. The Pottery In 2014 a small trial trench was dug in the eastern part of the fortified complex so as to obtain a preliminary date for the complex. The stratigraphy consisted only of colluvial deposits, and the few pottery sherds were mixed (Figure 14). Preliminary information on this pottery was published in 2015.10 The 23 fragments collected (of which 11 are diagnostic) pertain mostly to the Late Bronze/Iron Age, but occupation during the medieval period is also evident (Figures 15-16). The Late Bronze/Iron Age horizon is represented by two medium-sized neckless jars and one kitchen pot, all with smoothed external surfaces, with colours ranging from dark grey to brownish. Some body sherds present decorations, incisions or impressions creating grooved bands, wavy lines, or notches in one case, and the external surfaces are mostly simply smooth and in some cases burnished. This long chronological attribution is due to the fact that these decorations 10

KSP36.113 - neckless jar; Simple Ware; rim (Figure 15/1) Small fragment of a jar rim with rounded trapezoidal section, outer envelope. The external surface is smooth. Traces of fast-wheel working are visible on the inner part. The section is quite compact with a low density of inclusions. Thickness:14 0.7 (rim); diameter: 14; height: 2.4; width: 4.5 11 Grooved pottery occurs from the Late Bronze Age, see e.g. Gegharot (Badalyan et al. 2008) or Agarak, Aragatsi Berd, Kuchak I, Mastara and others (Badalyan and Avetisyan 2007). For the Iron Age specimens from Solak-1 see Vitolo et al. forthcoming. 12 The medieval fragments were analysed by the archaeologist Astghik Babajanyan, whom we would like to thank for her professional assistance. 13 This code (KSP) is the abbreviation of the Kotayk Survey Project; 36 is the identification number of the site of Tghit. 14 All the measurements are expressed in cm. The height and width refer to the fragment itself.

Petrosyan et al. 2015: 64.

400

Artur Petrosyan et al.: Overlooking the River Hrazdan Valley

Figure 15. Photograph of the pottery collected from the survey and test trench in Tghit (Kotayk Survey Project Archive).

Colour: external, internal and section 10YR 5/2 (greyish brown)

upper part of the rim there is a groove, possibly for holding a lid. The external surface is finely smoothed, brown-grey in colour, with a high density of projecting medium sized inclusions. Traces of fast-wheel working are visible on the inner part. The section is rough with a high density of small mineral inclusions.

Period: Late Bronze Age/Iron Age KSP36.2 - jar; Common Ware; rim (Figure 15/2) Small fragment of a jar rim with trapezoidal section, slightly enlarged externally. The outer surface is finely smoothed. Fast-wheel traces are visible on the inner part. The section is not so compact, brown in colour, with visible mineral inclusions.

Thickness: 0.7 (wall); 1.2 (rim); diameter: 12; height: 2.8; width: 4

Thickness: 0.6 (wall); 0.9 (rim); diameter: 16; height: 1.8; width: 2.6

Period: Late Bronze Age/Iron Age

Colour: external 10YR 5/1 (grey); internal and section 10YR 6/2 (light brownish grey)

KSP36.4 - jar; Common Ware; rim (Figure 15/4)

Colour: external 5Y 3/1 (very dark grey); internal 7.5YR 4/1 (dark grey); section 5Y 5/1 (grey)

Small fragment of a roughly made closed form. The rim is oval in section, highly everted, flattened above and enlarged outwards. On the inside a junction line between rim and neck is visible, obtained the by columbine technique. At the bottom of the neck it is possible to perceive a change of orientation of the walls. The surface are barely hand smoothed. The section is rough, with frequent mineral inclusions.

Period: Late Bronze Age/Iron Age KSP36.3 - neckless pot; Simple Ware; rim (Figure 15/3) Small fragment of a neckless pot rim. It is outer highly envelop, with externally rounded oval section. On the 401

Systemizing the Past

Figure 16. Drawing of the pottery collected in the survey and test trench in Tghit (drawing by A. S. Bonfanti and graphic processing by A. Cesaretti).

Thickness: 1 (wall); 1.1 (rim); diameter: 10; height: 3; width: 5.1

Period: probably Middle Ages KSP36.6 – closed form; Simple Ware; wall (Figure 15/6)

Colour: external 5YR 7/2 (pale red); internal 5YR 6/2 (pinkish grey); section 10YR 5/2 (greyish brown)

Body sherd of a closed form with grooved decoration. It is decorated with multiple incised horizontal lines, 0.1-0.2 high, c. 0.5 cm apart. Just a portion of the whole decoration is visible, with five incisions, obtained by different pressures; two of them are barely visible. The external surface is smoothed with traces of burnishing, dark grey in colour, with some projecting calcareous mineral inclusions. Wheelmade. The fabric is quite coarse, with frequent mineral inclusions.

Period: Middle Ages KSP36.5 - close shape; Simple Ware; base (Figure 15/5) Flat base of a closed form. The external surface is finely smoothed. Traces of fast-wheel are visible on the inner part. The fabric is compact, light pinkish-brown, with scarce mineral inclusions.

Thickness: 0.6; height: 4.4; width: 2.3

Thickness: 0.7 (wall); 0.8 (base); diameter: 18; height: 3; width: 3.1

Colour: external 2.5Y 4/1 (dark grey); internal 5Y 5/1 (grey); section 5Y 6/1 (grey)

Colour: external 5YR 7/4 (pink); internal and section 5YR 7/3 (pink)

Period: Iron Age 402

Artur Petrosyan et al.: Overlooking the River Hrazdan Valley KSP36.7 – closed form; Simple Ware; wall (Figure 15/7) Body sherd of a closed form with grooved decoration. Decorated by multiple incised horizontal lines, 0.15 cm high, close by to each other. Just a portion of the decoration is visible, with eight lines incised after burnishing. The external surface is burnished, dark grey in colour, with some projecting calcareous mineral inclusions. Wheelmade. The fabric is quite compact, brownish, with some mineral inclusions.

Small body sherd from the upper part of a jar. Decorated with two different incisions: a band composed of closely spaced multiple horizontal lines, total height 0.5 cm; at a distance of 0.1 cm there is a single wavy decoration, with curved lines 0.2 cm thick. Each wave is 0.4 cm high and 1.4 cm wide at the base. The external surface is light brown, finely smoothed, while the inner surface is dark grey with visible traces of fast-wheel working. The fabric is quite coarse, dark brownish, with frequent mineral inclusions.

Thickness: 0.7; height: 2.7; width: 2.8

Thickness: 0.9; height: 4.2; width: 4.3

Colour: external, internal and section 10YR 5/2 (greyish brown)

Colour: external 10YR 7/2 (light grey); internal 10YR 5/1 (grey); section 5Y 5/1 (grey)

Period: Iron Age

Period: Iron Age

KSP36.8 – closed form; Simple Ware; wall (Figure 15/8)

KSP36.11 - jar; Fine Ware; wall (Figure 15/11)

Wall fragment of a closed form with grooved decoration. Decorated with multiple incised horizontal lines, 0.2 cm high, closely spaced. Just a portion of the decoration is visible, with six incised lines obtained by a tool under pressure. The external surface is smoothed and brown in colour, while the inner surface is finely smoothed and greyish-brown in colour. Wheelmade. The fabric is quite compact, brownish, with some mineral inclusions.

Small fragment of a closed form. Just one portion of a single incised wave is visible. The line is thin, the external surface is smoothed and dark grey. Wheelmade. The fabric is compact, with few mineral inclusions. Thickness: 0.6; height: 2.8; width: 2.9 Colour: external 5Y 4/1 (dark grey); internal 2.5Y 5/1 (grey); section 7.5YR 6/2 (pinkish grey)

Thickness: 0.8; height: 2.9; width: 2.6

Period: Iron Age

Colour: external and section 10YR 6/2 (light brownish grey); internal 10YR 5/1 (grey)

Conclusions The site of Tghit is undoubtedly one of the most interesting and spectacular sites on the entire Hrazdan-Kotayk Plateau. As mentioned, it is unusual due to its extraordinary geographical location and the exceptional degree of preservation of a good part of the wall structures. In addition, this is one of the highest altitude fortresses currently known in the region. The site needs to be discussed in terms of both its history and its function. As regards its construction, there is no doubt that the foundation of the oldest nucleus of the fortified complex, recognisable in Wall-1 of the upper enclosure, must be dated to the protohistoric period. The pottery and architecture suggest a pre-Urartian Iron Age foundation. In any case, in the absence of more in-depth excavations within the complex, it is not possible to exclude an older phase. As observed, the pottery identified in the site, mostly common ware, can be framed chronologically between the Late Bronze/Iron Age and the Middle Ages. In general, as already noted, Walls-2 and 3, built on the west side of the rocky hill that hosts the fortified complex, post-date the foundation of Wall-1 for reasons of construction technique and physical relationships. It is not possible to know how much more recent they might

Period: Late Bronze Age/Iron Age KSP36.9 - pot; Kitchen Ware; wall (Figure 15/9) Small fragment of the upper part of a pot used as Kitchen Ware. It is coarse, slightly smooth on the outside with an irregular grooved decoration involving varying pressure on the same line, barely visible. Incisions c. 0.3 cm high. On the upper part there is a series of impressed horizontal notches. The external surface is light brown in colour. Handmade. The fabric is coarse, light brown in colour, with a high density of mixed mineral inclusions. Thickness: 0.7; height: 4; width: 1.9 Colour: external 10YR 5/2 (greyish brown); internal 5Y 7/1 (light grey); section 10YR 6/2 (light brownish grey) Period: not identified KSP36.10 - jar; Common Ware; wall (Figure 15/10)

403

Systemizing the Past be. In general these walls, in which mortar is entirely absent and protohistoric construction techniques were evidently used, cannot in any way be dated to the medieval era. The walls have buttresses designed more for mechanical functions than military ones (the most evident collapses have occurred between the buttresses), and with regard to form and dimension are quite similar to other specimens of pre-Urartian fortresses in the region. From this perspective, it should be noted that the buttresses appear exclusive to Wall-1, while in Wall-2 for example, there are curved protrusions, which can be compared to a series of protohistoric fortified structures located in the territory of southern Georgia and Armenia itself. Inside the upper enclosure in the first construction phase there would seem to have been no buildings of any kind, given that, as we have noted, bedrock outcrops almost everywhere. All the rectangular and elliptical structures found within the upper wall circuit appear to be attributable to later occupation phases that cannot be dated with exactness, but probably belong to the medieval era in general, as some potsherds seem to suggest. These small rooms, built mainly using stones taken from the perimeter walls, cannot be defined as structures with specific functions, but were probably non-residential, and appear more like small shelters probably used over the centuries by shepherds who carried out transhumance in these areas. Similarly, there are also considerable uncertainties regarding the chronology of the artificial lake. This cannot be dated on the basis of construction techniques, but it was certainly connected to the fortified complex. Given the surface area and the workforce necessary for the lake’s construction, it is unlikely that its construction was associated with particularly late stages, or when the place, by then downsized with respect to its original function, became mainly frequented by the shepherds who used the area for transhumance. A final reflection must be made regarding the function of the first fortified complex. The absence of buildings referable to the first construction phase suggests that the complex may possibly have been a fortified refuge. This is also suggested by the environmental context: the site is difficult to reach from the valley bottom and the climate on the small plateau that houses the structures is much more rigid and exposed to the elements than the surrounding areas. It must certainly have been frequented only sporadically. It was most likely a refuge for the population of some villages in the valley bottom who, in moments of great danger — when warlike groups or armies crossed the area — took refuge on high ground.15 This is a phenomenon that is beginning to be noted extensively in the region, with the presence of numerous fortified structures

that served as refuges. Striking examples are those of Abuli and Shaori in the Samtskhe-Javakheti, built at altitudes of almost three thousand metres, in positions that are strategically not visible from the territories below.16 It is possible, for example, that the population who resided in the fortified settlement of Aghzibir,17 with which is associated the extensive burial ground of Karashamb, used both in the Middle Bronze Age and in the Iron Age, utilized this place as a temporary refuge, as well as other sites in the area. After all, these settlements were located at the foot of the rocky spur which hosts Tghit. At the same time, the highly strategic nature of the latter leaves open the possibility, especially in historical times and in subsequent phases, that this fortified complex was home to small garrisons that used it to control transit in the Hrazdan Valley. It is clear, as ascertained for all the structures referred to as ‘cyclopean’, or characterized by impressive stone fortifications, that these were reused over the millennia, essentially by all the communities that lived in these territories. This was due to the desire to reuse the huge pre-existing structures, which were also very often built in highly strategic places. Tghit possessed both of these requirements. Bibliography Areshian, G., Kafadaryan, K., Simonyan, H., Tiratsyan, G. and Kalantaryan, A., 1977. Archaeological investigations in Ashtarak and Nairi districts of the Armenian SSR. Lraber hasarakakan gitut‘yunneri (Herald of Social Sciences) 4: 77–93 (in Armenian). Badalyan, R. S. and Avetisyan, P. S. 2007. Bronze and Early Iron Age Archaeological Sites in Armenia. I, Mt. Aragats and its Surrounding Region (BAR International Series 1697). Oxford, Archaeopress. Badalyan, R. S., Smith, A. T., Lindsay, I., Khatchadourian, L. and Avetisyan, P. 2008. Village, fortress and town in Bronze and Iron Age Southern Caucasia: A preliminary report on the 2003-2006 investigations of Project ArAGATS on the Tsaghkahovit Plain, Republic of Armenia. Archäologische Mitteilungen aus Iran und Turan 40: 45–105. Dan, R., Licheli, V., Vitolo, P., Chogovadze, T., Cesaretti, A. and Chilingarashvili, T. 2019. Abuli, un complesso fortificato nella regione di Samtskhe-Javakheti, Georgia. Bollettino Unione Storia ed Arte 14: 9–20. Licheli, V., Dan, R., Vitolo, P., Chigivadze, T., Cesaretti, A., Chilingarashvili, T. and Saccone, T. 2022. Cyclopean fortresses, royal cities or mountain shelters? The Abuli and Shaori Complexes in Southern Georgia in the light of recent archaeological investigations. Ancient Civilizations from Scythia to Siberia 28: 148–176. Petrosyan, A., Dan, R., La Farina, R., Raccidi, M., Castelluccia, M., Gasparyan, B. and Babajanyan, A.

15 There are several other fortresses in the territory of Armenia with similar features; for example Dzyanberd and Lazaravan on the slopes of Mount Aragats and Asni in the Ararat region.

16 17

404

On this topic, see Licheli et al. 2022. On this site, see Petrosyan et al. 2020b: 145, Figure 16.

Artur Petrosyan et al.: Overlooking the River Hrazdan Valley 2015. The Kotayk Survey Project (KSP): Preliminary report on 2014 fieldwork activity. Aramazd: Armenian Journal of Near Eastern Studies 9/1: 58–68. Petrosyan, A., Dan, R. and Vitolo, P. 2020a. The Kotayk Survey Project (KSP): An overview of the first six years of activities (2013-2018). Aramazd: Armenian Journal of Near Eastern Studies 10/1-2: 208–233. Petrosyan, A., Dan, R., Vitolo, P., Melikyan, V., Gasparro, O. and Nahapetyan, S. 2020b. Archaeological investigations in Kotayk Region as part of the Kotayk Survey Project (KSP). A glance at a selection of fortresses (2013 – 2019). In P. Avetisyan and A. Bobokhyan (eds), Archaeology of Armenia in Regional Context, Proceedings of the International Conference

dedicated to the 60th Anniversary of the Institute of Archaeology and Ethnography Held on July 9-11, 2019 in Yerevan: 133–151. Yerevan, Gitutyun. Simonyan H. 2012. Armenia at the Late Bronze and Early Iron Ages (15th-9th centuries BCE). In H. Simonyan (ed.). Archaeological Heritage of Armenia: 33-39. Yerevan, Hushardzan. Vitolo, P., Dan, R. and Petrosyan, A. forthcoming. A glance at the ‘black burnished ware’ in Armenia: a preliminary revaluation of the data from Room 1 in the Solak-1/Varsak site. In S. Heinsch, K. Jakubiak, R. Dan, P. Vitolo, M. Iskra and W. Kuntner (eds). The Long 7th Century: A Time of Transformation and Collapse in Urartu. Araxes. Turnhout, Brepols.

405

From Landjik to Dvin: Armenian Evidence of Decapitation from Prehistory to the Mediaeval Era Daniel Thomas Potts

Institute for the Study of the Ancient World, New York University, USA Abstract: Decapitation is a rare but important feature in Armenia’s archaeological record. Examples ranging in time from the 4th millennium BC to the mediaeval are briefly discussed, and motivations behind decapitations in the history of the Caucasus and the broader Near East are presented. The difficulty of revealing motivation in circumstances lacking epigraphic or literary sources is stressed. Keywords: Decapitation, Armenia, Kura-Araxes Culture, Bronze Age, Iron Age, Caucasus

Introduction In 1986 Paul-Henri Stahl (1925-2008), Directeur d’études at the EHESS (École des hautes études en sciences sociales) in Paris, published his masterful study of decapitation, one that ranges across a wide array of issues and subject matter, from antiquity to the reign of Timur Leng.1 More recently, in 2010, a conference on the subject of ‘Crânes trophées, crânes d’ancêtres et autres pratiques autour de la tête: problèmes d’interprétation en archéologie,’ was held in France, at which archaeological evidence of decapitation was carefully examined and discussed.2 In the Armenian Christian tradition, there is surely no more prominent case of decapitation than that of John the Baptist (Matthew 14:6, Mark 6:26, Luke 9:9; Figure 1), commemorated in Canon 58 of the Armenian hymnal.3 Other instances of decapitation are scattered throughout the Hebrew Bible, from David’s removal of Goliath’s head (1 Samuel 17:51, 54; Psalm 151 LXX:7) to Saul’s beheading by the Philistines (1 Sam 31:9) and, most dramatically, the beheading of Ahab’s seventy sons, on Jehu’s order (2 Kings 10:7-8). Apart from these cases, perhaps the most well-known Biblical example, one that was depicted endlessly by artists and memorialized in Vivaldi’s sacred military oratorio Juditha triumphans devicta Holofernis barbarie (RV 644; composed in 1716), is the decapitation of Nebuchadnezzar’s general Holofernes by Judith (Judith 13:8).4 That decapitation was a reality in Armenia’s own past is borne out by a number of bioarchaeological analyses of skeletal remains ranging in time from the 4th millennium BC to the mediaeval era.5 Bearing in mind that the reconstruction of motive and circumstance based on mute archaeological evidence is Stahl 1986. Boulestin and Gambier 2012. Ter-Mikaëlian 1905: 45. 4 See the papers in Brine et al. 2010. 5 E.g. Khudaverdyan 2014a; 2014b; 2015; 2017; Khudaverdyan and Hobossyan 2017; Khudaverdyan and Babayan 2018. 1 2 3

Systemizing the Past (Archaeopress 2023): 406–410

Figure 1. The beheading of St. John the Baptist, from the Walters Art Gallery Ms. W.547 hymnal, completed by Yakob Pēligratc‘i on 15 August 1678 in Constantinople (Creative Commons Attribution-Non Commercial-Share Alike 3.0 Unported Access Rights; The Walters Art Museum, 600 North Charles Street, Baltimore MD 21201, USA).

Daniel Thomas Potts: From Landjik to Dvin being ritually murdered but also that violence…perhaps played a very important role in such sacrifices’.11

often impossible, it is nevertheless the contention here that the Armenian data can be better understood via comparisons with the practice of decapitation in both earlier and later periods, both within the southern Caucasus and in neighboring regions. It has been said that, ‘Very little research…has focused on the notion of decapitation throughout the prehistoric Caucasus’.6 To that end, this short study was written to explore this important topic. It is offered to Pavel Avetisyan in great admiration of all that he has done to further our understanding of Armenia’s extraordinarily rich past.

4. At the Late Bronze Age and Iron Age site of Bakheri chala, a male cranium in burial 9 (age 40-55) displayed damage to the mandible, the left occipital bone and the mastoid process, consistent with beheading by a right-handed person using a sharp instrument, such as a sword.12 5. A female (age 45-49) cranium from Beniamin, Shirakavan I, dating to sometime between the 1st century BC and the 3rd century AD, displays damage to the mastoid process and foramen magnum consistent with decapitation.13

Archaeological Evidence from Armenia Published examples of decapitation from archaeological contexts in Armenia date from at least the 4th millennium BC although for the most part little contextual information is available. These examples fall into two categories. The first may be broadly considered hostile or punitive, while the second may be considered positive, sacrificial or reverential. To the first category belong crania from seven sites.

6. Burial 4 at Shirakavan, which dates to late Antiquity, contained the cranium of a female (age 40-44) who was probably decapitated using a sword.14 7. At mediaeval Dvin, two females (age 18-20, 20-25) and one male (age 20-25) show signs of decapitation. As in older cases, these include mechanical damage to the occipital condyles as well as signs of damage on the left mastoid process, suggesting decapitation by an instrument such as a sword. Mandibular damage has been interpreted as an indication that the individual’s head was lying on its side, e.g. on a chopping block, in ‘the traditional pose adopted for judicial decapitation’.15 The precise dates of these individuals are uncertain.

1. At Landjik, a Kur-Araxes site on the Shirak Plateau, one male (skeleton 1, age 60-65) and two female crania (skeleton 3, age 45-49; skeleton 4, age 50-59), display evidence of possible decapitation. In the case of skeleton 1 this includes ‘probable peri-mortem sharpforce weapon injury’ as well as damage to the occipital condyles, foramen magnum, mastoid process, and inferior surface of the occipital bone,7 consistent with the individual having been beheaded while in a vertical position, e.g. with a sword or axe. In the case of skeleton 3, the occipital condyles display mechanical damage, as did skeleton 4, which also showed perimortem fractures on the right mastoid process and right occipital condyle, consistent with decapitation.8 In all three cases decapitation has been attributed to ‘intentional ritual violence’.9

To the second category belong crania from at least two sites. At Late Bronze Age (14th-12th centuries BC) and Iron Age (6th-5th centuries BC) Lori Berd, two male crania (burial 105, age 18-20; burial 107, male 50-55) were found, oriented towards the east, in association with funerary offerings.16 Contextual evidence is, however, limited. At the Iron Age site of Shirakavan, datable to the 9th-6th centuries BC, the situation was very different. There two isolated female crania (age 16-20, 20-25) and one male (age 50-55) were discovered in a burial context that included funerary objects. The crania all lacked their mandibles and showed signs of tooth extraction. These have been interpreted as ‘part of an elaborate ritual complex within Shirakavan society’ and as ‘human sacrifices rather than venerated ancestors’,17 illustrating ‘sacrifice as cosmological transition’.18 It has been further speculated that human sacrifice, if it indeed occurred at Shirakavan, may have

2-3. Two sites of Late Bronze and Iron Age date near Lake Sevan have yielded evidence of decapitation. At Noraduz (11th-6th cent BC) the cranium of a female in burial 21 (age 40-45) displayed signs of damage on the mastoid process suggestive of decapitation in a vertical posture. Similarly, the cranium of a female (age 20-25) from Karmir, burial 3, displayed breakage of the left mastoidal process and left occipital condyle, also suggestive of decapitation in a vertical posture.10 The fact that this latter individual had also suffered an antemortem injury, it is observed, ‘provides a good case for the assertion that not only were individuals such as this

Khudaverdyan 2014a: 1561. Khudaverdyan and Hobossyan 2017: 327. 13 Khudaverdyan 2015: 220. 14 Khudaverdyan 2014c: 226. 15 Khudaverdyan and Babayan 2018: 2. 16 Khudaverdyan 2014b: 174. 17 Khudaverdyan 2014b: 178. 18 Khudaverdyan 2017: 490. 11 12

Khudaverdyan 2017: 488. Khudaverdyan 2017: 501. 8 Khudaverdyan 2017: 504, 506. 9 Khudaverdyan 2017: 508. 10 Khudaverdyan 2014a: 1560. 6 7

407

Systemizing the Past been a form of divine offerings, perhaps to ensure fertility.19

— culminating in the performance of a ritual offering or final, destructive gesture. Disentangling these concurrent, often overlapping motivations is often impossible and may only serve to impose unjustified, artificial strictures on what is surely a multi-facted sphere of human behavior.24

Circumstances of Decapitation Although the material reviewed here in no way constitutes a statistically valid sample upon which clear deductions can be based, it does not appear as though decapitation was common in ancient Armenia. This is not to say, however, that instances of decapitation may not have been hugely significant to all involved, and not just the deceased. Having said that, without literary or epigraphic evidence it is always going to be difficult to determine the motivation behind an individual case of decapitation. For purposes of discussion, those few cases in which isolated crania were discovered in association with funerary offerings, showing what might be termed positive or reverential treatment, will not be further discussed here. Rather, the focus is on those apparently hostile or punitive cases that have been documented.

In the case of rebel leaders withholding tribute or seeking to assert independence, their subjugation and capture was often followed by decapitation.25 It is therefore untrue to infer, from the lowly circumstances of some decapitated individuals, that decapitation was reserved for low-status offenders, such as thieves.26 As Charles Macfarlane observed in 1828 when he was in Constantinople Though a trifle, in reference to Turkish customs, it may be as well to rectify a mistake which prevails, as to the manner in which the heads of those who have received the reward of their crime, or (cases of more frequent occurrence) have fallen under the hate or suspicion of the Porte, are disposed of in the Serraglio. It has generally been supposed that those heads were stuck on spikes on the summit of the gates or on the edges of the Serraglio walls, and there exposed in horrid rows, to the gaze of the public. This popular error has sanctioned the flights of fancy, the Sultan’s palace has been converted into a Golgotha,

As Stahl noted, decapitations often constitute a form of human sacrifice, even if this is not in the ritualized sense often attributed to them. The decapitation of enemy combatants during or prisoners-of-war following combat,20 e.g. before a warlord or ruler, or the head of a mortal enemy sent from the battlefield back to the capital,21 may take on aspects of an offering. Simultaneously a political as well as a ritual act, decapitation is rarely a straightforward event with a singular explanation, but rather a result of complex motivations involving a combination of aims — the annihilation of an enemy, the exaction of justice or retribution, the instillment of fear in an adversary, the elimination of a political opponent,22 the advancement of one’s position in the eyes of a commander or ruler23

(Bakar), et qu’on lui a coupé la tête.’ See Brosset 1847: 344. In 24 Stahl 1986: 188-189 wrote, ‘même si on ne peut point parler d’un sacrifice humain caractérisé, certaines décapitations le rappellent étrangement. Ainsi, la mise à mort solennelle des prisonniers après un combat, devant la tente du sultan; ou encore, les têtes envoyées au centre politique et religieux du pays, qui ont un caractère évident d’offrande. Phénomènes politiques en même temps que rituels, la décapitation et les pratiques en rapport avec le crâne peuvent rarement être interprétées exclusivement comme politiques ou comme rituelles, car dans la grande majorité des cas il s’agit de motivations complexes; une même tête est coupée pour se défendre d’un ennemi, pour accomplir un acte de justice, pour éliminer un adversaire politique, pour effrayer l’ennemi, mais dans le cadre d’un rituel, à chaque fois que cela est possible, et finit comme offrande auprès du centre spirituel de l’Empire ou de pays. Des aspects en rapport avec la mort et son rituel viennent s’y ajouter et il est bien souvent malaisé de démêler dans cet ensemble l’interprétation la plus proche de la vérité.’ Cf. Cazacu 1996: 246-247. 25 See e.g. Cazacu 1996: 247 discussing ‘les cas de décapitation et d’envoi à Istanbul de têtes de princes roumains de Valachie [in 1799] et de Moldavie [in 1777]…et d’exposition publique de ces têtes. Le caractère mixte, politique et rituel, de cette cérémonie ressor très clairement des soins prodigués à la conservation et au transport de la tête, de son identification par le sultan (quand il s’agit d’un prince roumain), de la place qui lui est assignée pour être exposée devant le palais impérial, de la durée de cette exposition, du sort ultérieur de la tête et du corps.’ These were vassals who revolted against the Porte. 26 E.g. in Constantinople during the mid-18th century Guer observed, ‘Avoir la tête tranchée est une mort infâme, mais la plus commune; elle n’est en usage que pour les gens de néant & des Esclaves. Le Roi de Maroc se fait un jeu de cette sorte d’exécution; on diroit qu’il aime à couper des têtes, pour montrer sa dextérité.’ See Guer 1747: 160161. The same low status probably characterized the decapitated individuals whom the French ambassador to the Court of Fath ‘Ali Shah saw in 1808 outside of the Shah’s palace in Tehran: ‘Quatre cadavres dont on avait coupé la tête, sont étendus devant la porte du palais du Roi. Chaque passant ne manque pas de donner un coup de pied à ces têtes.’ See Gardane 1809: 52.

Khudaverdyan 2014a: 1562. In 1609 in Georgia ‘Zaza Tzitzichwili renverse le Pacha et lui coupe la tête.’ See Brosset 1845: 253. In 1615, ‘les Persans avaient coupé la tête aux Cakhes soumis.’ See Brosset 1845: 275. For a survey of the decapitation of slain enemies following battle see Zouache 2010. For the practice in prehistory see Testart 2008. 21 The head of the Elamite rebel Te’umman, sent to the Assyrian king Assurbanipal, is a case in point. See e.g. Bonatz 2004. In 1605, when Constantine I was fighting the Kakhetians in Georgia, ‘lorsque les deux armées furent en bataille, Costantiné consulta les astrologues, et que ne voulant plus combattre, il ordonna de faire retirer ses soldats; ce que voyant les Cakhes, Dawith, Thamaz et Bébour fondirent sur lui, le précipitèrent de son cheval, et lui ayant coupé la tête, l’apportèrent à la reine Kéthéwan.’ See Brosset 1856: 158. 22 In 1717 ‘Abdallah, chef des Afghans du Candahar, veut soumettre aux Persans ses états pour pouvoir vivre tranquille: Mahmud, fils de Mirweis, apprend ce dessein de son oncle, se met à la tête d’une quarantaine d’amis de son père, se rend maître de son palais, coupe la tête à Abdallah, et se fait proclamer roi.’ See Buret de Longchamp 1825: 33. 23 Thus Charpin 1994 noted several examples in the Mari letters, e.g. when the inhabitants of Talhayûm decapitated the head of their king, Hammî-epuh, and sent it to Šubat-Šamaš; or Tarim-Šakim wrote to Yasmah-Addu, boasting that he had decapitated ten enemies that he was going to bring to his father, Samsî-Addu. In 1723 ‘Othar a été traitreusement livré aux Turks, par les soins du maître de la Géorgie 19 20

408

Daniel Thomas Potts: From Landjik to Dvin wish to frighten their children. Having trusted to the word of the khan of Badkoo [Baku], he advanced to that fortress for the purpose of receiving the keys, being accompanied only by one Georgian, the Colonel Prince Eristow, and a few Cossacks; when they fell into a snare, and he was shot by order of the khan. They immediately cut off his head and that of Prince Eristow, and sent them to the shah of Persia: the bodies were buried on the spot.29

and to speak only of recent pictures drawn of the palace, a French poet [Victor Hugo] describes the walls as ‘decorated with six thousand heads,’ while an English writer [Tales of the Great St. Bernard] represents its gate as ‘hung with ranges of immense bones, looking ghastly in the illumination.’ Now, the truth is, there is nothing of all this; the heads of delinquents or victims, if of common condition, are thrown on the ground by the side of the Serraglio outer gate; if of rank, as Pashas, &c., ‘they are placed in a dish, as Doctor Walsh correctly describes in the case of Ali Pasha, of Yanina, ‘on a low marble pillar, between the first and second gates of the Serraglio.’ On common occasions the heads are exposed only a few hours, but on more important ones, when government wishes to impress the people, they are left for three days, but seldom longer. After the exposure, they are thrown away, or purchased and buried by relations or friends, but are never kept to fringe walls, and decorate gates. During the horrid exhibition at the gate, or within the court-yard of the Serraglio, yaftas or paper scrolls, setting forth (truly or falsely) the offences for which those heads are there, are suspended over them….Headless trunks, and strangled men, are often seen floating down the Bosphorus, and round the Serraglio-point, but I never could learn, even from the oldest people at Stambool, that skeletons, heads, and bones, were ever used to decorate the summits of the Serraglio walls.27

It is clear, of course, that the rule regarding not shedding the blood of a fellow Muslim did not always prevail, particularly in conflicts involving Sunni vs Shi‘a. On plumes worn in the helmets of Kurdish warriors, Moritz von Kotzebue observed, ‘each of which, it was said, had been granted to them in return for an enemy’s head. I counted as many as five on some horsemen’.30 Some Conclusions Regarding the Armenian Evidence It has been suggested that decapitated crania from archaeological sites in Armenia may have been the results of corporal punishment; trophies taken in battle; relics for curation and veneration; and the by-product of ritual bloodletting.31 It has also been suggested that the severed crania themselves may have had a variety of functions. These include serving as apotropaic devices; representations of slain and defeated enemies; vessels of supernatural or magical properties; mnemonic devices facilitating ancestor veneration; and triggers to help forget, erasing social memory of the deceased.32

Where religious prohibition intrudes decapitation practices may be modified. It has frequently been noted, for example, that Islam forbids shedding the blood of another Muslim. For that reason, Muslims destined for decapitation, such as enemies of the state, were first strangled so that they were dead prior to decapitation.28 Conversely, in the case of Muslim vs Christian, no such inhibition exists. A striking case is provided by the decapitation of the Georgian Prince Pavel Tsitsianov (Tsitsishvili), in Russian service, who was betrayed and decapitated by the Khan of Baku in 1806. As the Russian diplomat Wilhelm von Freygang wrote,

With no information on the internal political organization of most of the societies in which instances of decapitated crania were found, we are unable to confirm or reject these alternatives. Political vengeance and the annihilation of enemies have certainly been powerful motivations behind decapitation through time, as the sources quoted above clearly demonstrate. Yet whether the individuals whose remains have been recovered in archaeological context were executed, killed in battle, put to death as punishment for an offense, whether theft or witchcraft or sorcery, or ritually killed, we do not know. The small number of decapitated individuals indentified through archaeological excavation in Armenia, however, would strongly suggest that decapitation was an exceptional treatment for transgressors, enemies or combatants. As such, the praxis may have been all the more powerful for its living witnesses because it was so rare. Decapitation was probably one of the closest things to punishment by divine sanction that human societies in the Caucasus ever devised, and it would seem to have

The Prince Czizianow, descended from a Georgian family, after having distinguished himself during many years in the Russian service, was nominated in 1803 military governor of Georgia. He it was who gained by assault the fortress of Ganja, the name of which he changed to that of Elizabetpol: he took possession of several khanships, such as those of Schoucha and Karabagh, and was held in awe by the Persians, Turks and Lesgees, and by all the other tribes of these countries. To this very day his name is repeated as something dreadful, when the Persians 27 28

Freygang and Freygang 1823: 159-160. Kotzebue 1819a: 248-249. 31 Khudaverdyan 2015: 225. 32 Khudaverdyan 2017: 488-493. 29 30

Macfarlane 1829: 388-392. Cazacu 1996: 248.

409

Systemizing the Past been much rarer than, e.g. martyrdom by decapitation after the advent of Christianity in the region. ‘La mort infâme,’ as Cazacu called decapitation,33 must have been a form of human annihilation that occurred only very rarely, and left an indelible impression on the minds of all concerned. Intra-societal decapitation — inflicting this sort of punishment on members of one’s own social group — was probably less frequent than inter-societal decapitation, once warfare between rival neighbors and standing armies became more common. New excavations, and careful attention to human skeletal remains, will hopefully shed more light on this fascinating subject in the years ahead.

of a Journey into Persia in 1812, and An Abridged History of Persia since the time of Nadir Shah. London, John Murray. Gardane, A. de. 1809. Journal d’un voyage dans la Turquied’Asie et la Perse, fait en 1807 et 1808. Paris, Chez Le Normant. Guer, J.-A. 1747. Mœurs et usages des Turcs, leur religion, leur gouvernement civil, militaire et politique, vol. 2. Paris, Chez Merigot & Piget. Khudaverdyan, A. Y. 2014a. Decapitations in Late Bronze Age and Iron Age sites from Sevan region (Armenia). Journal of Siberian Federal University, Humanities & Social Sciences 9: 1555–1566. Khudaverdyan, A. Y. 2014b. Ritual and ceremonial dismembering bones in a burials in Bronze and Iron centuries from Armenian Plateau. Journal of Antropologija 14: 163–195. Khudaverdyan, A. Y. 2014c. Les inhumations des cimetières de la plaine Chirak (Arménie): Approche biologique et sociale. Issues in Ethnology and Anthropology NS 19: 219–242. Khudaverdyan, A. Y. 2015. Palaeopathology of human remains of the 1st century BC-3rd century AD from Armenia (Beniamin, Shirakavan I). Anthropological Review 78: 213–228. Khudaverdyan, A. Y. 2017. Decapitation of humans and anthropomorphic figurines in the Kura-Araks Culture from Armenia. Mankind Quarterly 57: 487– 518. Khudaverdyan, A. Y. and Babayan, F. S. 2018. The human skeleton from Dvin, Armenia: detailed anthropological and paleopathological analysis. Bulletin of the International Association for Peleodontology 12: 1–17. Khudaverdyan, A. Y. and Hobossyan, S. G. 2017. Bioarchaeological evidence for the health status of a Late Bronze Age and Early Iron Age Bakheri Chala population (Armenia). Anthropologie 55: 319–336. MacFarlane, C. 1829. Constantinople in 1828: A residence of sixteen months in the Turkish capital and provinces, 2nd ed., vol. 2. London, Saunders and Otley. Stahl, P.-H. 1986. Histoire de la décapitation. Paris, Presses Universitaires de France. Ter-Mikaëlian, N. 1905. Das armenische Hymnarium. Studien zu seiner geschichtlichen Entwicklung. Leipzig, J.C. Hinrichs’sche Buchhandlung. Testart, A. 2008. Des crânes et des vautours ou la guerre oubliée. Paléorient 34: 33–58. Zouache, A. 2010. Têtes en guerre au Proche-Orient, mutilations et décapitations, Ve-VIe/XIe-XIIe siècles. Anisi 43: 195–244.

Bibliography Bonatz, D. 2004. Ashurbanipal’s headhunt: An anthropological perspective. Iraq 66: 93–101. Boulestin, B. and Henry-Gambier, D. 2012. Crânes trophées, crânes d’ancêtres et autres pratiques autour de la tête: problèmes d’interprétation en archéologie (British Archaeological Reports International Series 2415): 17–19. Oxford, British Archaeological Reports. Brine, K. R., Ciletti, E. and Lähnemann, H. 2010. The Sword of Judith: Judith Studies Across the Disciplines. Cambridge, Open Book Publishers. Brosset, M.-F. 1845. Matériaux pour servir à l’histoire de la Géorgie depuis l’an 120 jusqu’en 1755. Mémoires de l’Académie impériale des Sciences de St.-Pétersbourg 6th ser. 5: 165–315. Brosset, M.-F. 1847. Notice historique sur les trois dernières années du règne de Wakhtang VI et sur son arrivée en Russie, d’après des documents authentiques. Bulletin de la Classe des Sciences Historiques, Philologiques et Politiques de l’Académie Impériale des Sciences de Saint-Pétersbourg 3: 322–345, 354–375. Brosset, M.-F. 1856. Histoire de la Géorgie depuis l’Antiquité jusqu’au XIXe siècle, vol. 2, part 1. St. Petersburg, Imprimerie de l’Académie Impériale des Sciences. Buret de Longchamp, P.-N. 1825. Les fastes universels, ou tableaux historiques, chronologiques et géographiques…, vol. 8. Brussels, J.-B. Dupon. Cazacu, M. 1996. La mort infâme. Décapitation et exposition des têtes à Istanbul (XVe-XIXe siècles), in G. Veinstein (ed.), Les Ottomans et la Mort. Permanences et mutations: 245-289. New York/Cologne, Brill. Charpin, D. 1994. Une décollation mystérieuse. Nouvelles Assyriologiques Brèves et Utilitaires 1994/3: no. 59. Freygang, F. and Freygang, W. von. 1823. Letters from the Caucasus and Georgia; to which are added, The Account

33

Cazacu 1996.

410

Reconstructing the Lifeways of the Kura-Araxes* Mitchell S Rothman

Widener University/Penn Museum Abstract: The emphasis of Armenian archaeology in the last few decades has been on establishing the chronology of named archaeological ‘cultures’ that define specific periods of the Neolithic, Chalcolithic, Bronze, and Iron Ages. This effort is an essential first step in any analysis of Armenia’s archaeological past. As that chronology becomes clearer, we can move on to other issues. One such issue is to understand the society and cultural traditions of the Kura-Araxes better and to discover why they transitioned into the societies like those of the Early Kurgan Period. To do this we need to expand the data we collect and how we use it to understand societal and cultural issues better. We need to re-assess our analytical stance and widen and enrich the publication of details of our excavations’ and surveys’ archaeological remains if we are to discover what the lifeways of the populations in Armenia’s deep past were, and why they changed. This contribution in Pavel Avetisyan’s honor proposes some ideas of how we can do that. Keywords: archaeological analysis, Kura-Araxes, cultural transitions, theory, Armenia

The Questions We Ask and Approaches to Answering Them Archaeology has the ability to answer a number of questions about the past. The basic data we analyze addresses questions of ‘where’ and ‘when.’ From these questions we construct the basic space-time framework of our datasets and their content. This is a foundational first level of analysis. From these largely descriptive data we begin the search for the next level of answers to the questions ‘what’ and ‘how.’ Many of the what questions are being answered using new scientific methods. For example, what plants did they grow, what animals did they raise, what was their landscape and climate like? This is not to mention more traditional archaeological questions: what goods did they produce, what kinds of technology and tools did they have, what kinds of buildings did they construct, what rituals did they practice for this and the next life, etc.? The how questions take us further into their world of traditional behaviors and social structures. For example, how did they produce various goods, how did they use their tools and technologies, how did they use symbols and rituals to re-affirm and communicate their ideology and values, how did they structure the spaces in which they lived? In another important sense, the how questions take us into their social, economic, and political organization: how did they structure who made decisions and how much, if any, authority the decision-makers had, how were their relationships and their status defined in the way they conducted production, distribution and who I first met Pavel jan in 2009, when I joined Hakob Simonyan for the Shengavit excavation and, more importantly, publication project, which is moving ahead. He was always gracious and interested in the new ideas I was bringing. This despite his own strong intellectual tradition and encyclopedic knowledge of thousands of years of Armenian pre- and early history. In thanks for his always warm welcome, and frequent help, I offer the following article as a way to build on the foundations he and others laid to take our field further.

*

Systemizing the Past (Archaeopress 2023): 411–423

got to consume what goods, how did they use mental ideas (symbols, perceptions, and traditions) to define the current social order and to permit them to adapt or fail to adapt to new circumstances in their societies and in their particular human and natural environments, how did these determine family and group structure? At its best eventually we can answer ‘why’ questions. Why did these evolutionary paths take the shape they did? For me, again, the two why questions that fascinate me are: why did the Kura-Araxes cultural tradition and societal system originate and change, and why did it transform into a very different society and cultural tradition in the Early Kurgan Period? Implicit in these questions are two very different approaches to knowing the past. Those approaches can be understood as either ‘emic’ or ‘etic’.1 An emic approach looks at the world from the native or, in this case, ancients’ point of view. An etic approach sees the world from the modern analyst’s point of view. At first, this distinction would seem pretty far afield from archaeology, but if we think about how we do typology— that is, what criteria we choose to differentiate one type from another—we use an etic approach. We assume that elements of the shape of the rim, for example, are meaningful, because they are meaningful to us. Were they meaningful from an emic point of view for people in the past? This is one reason I always encouraged my students to take a class in pottery making to see how the process affects the potters’ point of view as to which variations are possible, and which are important. The same problem often affects our understanding of change. From an etic point of view, we tend to look from the final outcome back to its predecessors and often assume somehow that the process was headed 1

Hayden 1984.

Systemizing the Past that way all along. But did the emic viewpoint of the native, which determines how they behave, in any way presage what would happen? They are living in the moment, trying to meet their needs and desires within a world defined for them by their history, traditions, technical capabilities, and the local circumstances in their human and natural environments; in other words, their lifeways. If we are to understand what happened, much less why it happened, we need to ‘look up’ with the emic point of view.2 No hunter-gatherer or Neolithic farmer wakes up one day, and asks, ‘how can we make our society change to evolve kingly authority?’ A constellation of events, trends, and perceptions put stress on the system and offered alternatives in order to catalyze the creation of new organizations ultimately like those we call a state. The definition of what defines a state is etic, but the actual process of it becoming one is emic. Those alternatives are not infinite.3 Despite significant differences among societies in different regions, similar patterns of change are visible again and again in different regions and over different time spans.

and base construction, the more likely potters made the pots in a centralized workshop. Even then, the real measure of a centralized system is how and to which consumers these pots were exchanged. When few archaeologists even mention pots outside the typical, and rarely publish a detailed set of measurements from a large sample of pots, how can we do this? In general, many other lifestyle questions never get asked or answered, because archaeologists think that listing general categories of typical artifacts‘archaeological cultures’- is the endpoint to be achieved. Proposal for a New Approach to Discovering Lifestyles If our goal is (1) to determine in detail the activities that made up the ancients’ lifestyles, (2) the culture that set traditions and conveyed meaning to them, each in an emic way through the what and how questions, then (3) the organization of their societies in an etic way, and (4) patterns of change explained by asking and answering the why questions, what do we need to do to achieve it?

The problem is that our emphasis on chronology does not make it easy to find those critical patterns that set the stage for change. Again, developing chronologies is an essential task for archaeologists in Armenia and the adjoining territories.4 However, chronology-making, as we have done it, is an essentially etic pursuit. As Avetisyan describes the method, ‘These [Early Bronze to Iron Age] archaeological cultures strongly discerning from each other by their characteristic features differ also by sustainability, chronology and areas of distribution’.5

I propose that first we need to adopt a Social Science definition of Culture to understand the lifestyles (the nature) of the Kura-Araxes and why in the end the adaptation it represented did not last. In Social Science, Culture is the skeleton of societies’ beliefs, values, traditions, actions, and structure. For archaeologists, it is probably best to split Culture into two inter-related halves: culture and society.7 According to this method, culture represents its mental aspects. So, they are the values, beliefs, ideology, symbols, rituals, and traditions that define aspects like identity, beliefs, and values in the way this and the divine world work and how people are expected to behave. Reality is how they actually do behave within these guidelines. Society, on the other hand, is the relations of individuals and groups. These emic relationships help us generate an etic definition of how societies functioned and adapted to changing circumstances. They help us understand organizations we classify as economic, political, and social in an etic way. In reality the two halves are part of one whole. However, given our need to discover these elements from artifacts that were left and happen to survive, and to understand how culture and society interact, it is better analytically to separate them. The Russian school had a period when it was interested in a culture ecology perspective that is related to this approach of adaptation of cultures and societies.8

This is especially true because of archaeologists’ emphasis on pottery style. An emphasis on pottery style creates two problems. First, it takes the focus off many important elements of lifestyle. As Adam Smith6 has written, ‘To date, we know frustratingly little about the social world of Kura-Araxes village communities,’ the emic elements. Second, to make a typology, one literally needs to focus on the typical or normal. What, however, if we want to find out whether pottery was made in household contexts for use by a narrow group of consumers or in workshops for a much wider set of consumers exchanged through formal distribution networks? The best measure of that is whether there was standardization. A standardized, workshop-made pottery corpus requires the same technique be applied again and again with the same result in order to make the process efficient for the potters. The less the variation in key measurements like body thickness, rim

Another problem is the way we define distinctive ‘cultures.’ Again, Sagona9 and others maintain that

Vitelli 1998. Steward 1955. Avetisyan and Bobokhyan 2008; Avetisyan 2017; Marro and Hauptmann 2000; Sagona 1984; 2000. 5 Avetisyan 2017: 57. 6 Smith 2005: 260. 2 3 4

Rothman 2017. Sagona 2010. 9 Sagona 2014. 7 8

412

Mitchell S Rothman: Reconstructing the Lifeways of the Kura-Araxes pottery style is the way to do this.10 So, Sagona claims that the Kura-Araxes period extends well past 2500 BC, as late as 1600 BC. 2500 BC when most Kura-Araxes sites were abandoned. Yet, our understanding of these two general periods is that the Kura-Araxes was typified by largely peaceful, inward-looking, settled, farmer/ herders, who over the period between 3500 and 2500 BC saw dramatic increases in the numbers of sites, in varying size of settlements, and the growth of more centralized societal systems.11 They never reached the level of the state or perhaps even chiefly society.12 The Early Kurgan Period is one of mobility, militarism, and a very different political structure with ranked or even authority-based leadership. That is, institutionalized differences in the status of some people to the point where individuals or groups have the power to make decisions and plans and enforce compliance by the majority of others. Archaeologists have found few Bedeni or Martkopi (Early Kurgan Period) settlements. The settlements they did find, like Berikldeebi, had architecture not typical of the Kura-Araxes.13 How then can the lifeways of the people of the Early Kurgan Period be seen as the same as those of the Kura-Araxes? Even if some, not all, elements of style in pottery of the Early Kurgan Period are superficially the same as the Kura-Araxes, does that mean they are the same? To me, the answer is no.

individual behaviors. Each site leaves material evidence of the attempt by ancient people to meet their needs and desires, defined by cultural standards. Comparing these individual cases helps us see what larger patterns were developing and what factors likely led to those patterns without sacrificing the details of daily life. Sometimes, the very differences among sites or households will reveal those factors most clearly as the ancients lived them. Looking at how this changed level by level made the processes of change clearer. Achieving this end means adopting a more Social Science definition of culture and being clear when we are using an emic and when we are using an etic approach. Understudied Artifacts and Analytical Techniques For this analytical stance to work, first, we need to study, and publish, a variety of artifacts that we tend to ignore. We need to ‘publish’ much more detail on all of our artifacts. The easiest way to convey this information is on the worldwide web. No publisher wants to put that many pages of data in a printed book. Again, one reason to do this is that greater collection and publication permits us to reconstruct the activity areas of each building and open space. For example, for Tepe Gawra I mapped the distribution of artifacts in buildings and open spaces through time.17 I could then compare the activities and access to goods house by house in each level, and then level compared with level. This distribution of functions gave me a clear picture from the emic point of view of what different groups did, in what ways they cooperated, how they structured relations of production, distribution, and who got to consume them, and in that case with the help of seals and sealings how they organized decision-making.

This is not to say that pottery style is not useful. Like a dialect of language, pottery style reflects the identity of particular groups of producers. The more they exchange and share techniques and ideas about what pots were to be used for, as well as what shapes and decorative elements should be applied, the more likely they will have a common corpus of style. Their consumers see these as the norm. This local style differs from other areas even within generally the same cultural tradition.14 So, Kura-Araxes KA1 pottery is largely uniform, whereas the KA2 pottery reflects the balkanization of groups within the Kura-Araxes cultural tradition.15 These further reflect elements of economic organization. The style zones appear to match the geographical areas within which most of economic exchange existed.16

In Armenia we lack a seal/sealing locking mechanism, typical in Mesopotamian administration. Similarly, we lack the seals on mass-produced pottery of the Early Bronze II mass-produced pottery exchange network of the Levant.18 The few objects archaeologists in the South Caucasus call seals cannot be administrative hardware without both the seal and what was sealed. Very likely this means that the kind of administration that typified Mesopotamian societies simply did not exist among the Kura-Araxes societies, or certainly the mechanisms to achieve some kind of controls were handled in a completely different way that left few physical remains.

So, our goal must be to come as close as we can to understanding the lifeways of the ancients, household by household, level by level, and site by site, over time. Like biological evolution the focus of cultural and societal evolution is the sum of individual modifications that reflect cultural and historical traditions and

One of those data sets that need greater analysis and publication are bone tools (Figure 1). Bone tools have changed amazingly little since the Mesolithic. In Figure 1, A to F are from Shengavit, a site of the KA 2

Sagona 2014. Rothman 2015. 12 Rothman 2021. 13 Dshawachischvili 1998; Sagona 2018. 14 Rothman 2014. 15 Badalyan 2014. 16 Batiuk et al. in press. 10 11

17 18

413

Rothman 2002. Rotem et al. in press.

Systemizing the Past

Figure 1. Bone Awls from Shengavit (after Rothman, Forthcoming 1, Figure 5g.2).

(3000 to 2500 BC) (Rothman forthcoming 1). G is from an American historical house of the 19th century AD. Clearly, they had similar functions. We can therefore use information from traditional societies to give us an idea of their function. Figure 1 illustrates the most common category: awls. They are characterized by pointed ends, but there are a number of functions represented. Types C and D appear to be used primarily for basket and leather making, two products we almost never read about for the Kura-Araxes, but given the number and percentage of these tools, clearly it was an important craft.19 Type E looks very much like a spacer used in weaving. The spacer lifts the warp to allow easier transit of the shuttle cock, probably made of wood and therefore gone in archaeological contexts, carrying the weft. Types B, F, and G are too crude for the work of C and D. Very similar tools, however, are attested ethnographically for scraping and removing hair from animal skins for leather making. Type A is generalized tool that has many possible functions. In addition to those illustrated are needles for closing clothing (the same function as fibulas), bone arrows, flint retouchers, spindle whorls, and some red deer horns with a cut-out for hafting or hanging. Their function can be anything from digging to use as ritual symbols. Rarely do publications even mention this group of important tools, or if they do, rarely is there a complete list of them with their proveniences.

category is also rarely published, much less analyzed. Yet, it is probably more important than lithics in terms of the variety of its uses.20 The way I have proposed to classify them is by their working surface. Querns, one of the few types ever mentioned, and tools like shaft straighteners, pecking stones, and anvils have their working surfaces on the top. The grinder used with querns, abraders for a variety of functions from leather making to plant and metal ore processing, polishers for plastering and smoothing bone have working areas on the bottom. Harrows, axes, and pestles’ working areas are on their edges. Hammers, pestles, chipped stone retouchers are ones with their working edges on the end. Last are a wide variety of tools whose whole is the working surface. These include maceheads, digging stick weights, measuring weights, spindle whorls, burnishing stones, sinkers, loom weights, etc. So, this ground stone category represents functions like agricultural work, agricultural product processing, fishing, ceramic, cloth, leather, arrow, bead (jewelry), and tool making, building, etc. The materials used may be important. The maceheads are the only category at Shengavit of which a majority were made from marble. This may relate the hardness of marble and therefore their military use, or the visual characteristics of marble as symbols of some kind of status. This is a rich data set that almost all of us do not carefully collect or adequately publish.

Another category of tools is ground stone tools (Figure 2). This amazingly varied and information-rich

Pottery as a representation of style, as already mentioned, is one of the most discussed artifact

19

Rothman, forthcoming 1.

20

414

Rothman, forthcoming 2.

Mitchell S Rothman: Reconstructing the Lifeways of the Kura-Araxes

Figure 2. Ground Stone Tools from Shengavit (after various pieces from Rothman, Forthcoming, Figures 5f. 2, 6, 7, 12, 18).

categories for Armenian archaeology (Figure 3). While pottery style and typology are important topics, as already discussed, they tend to reduce the actual variability of pottery characteristics, and rarely are their full data reported. Wilkinson21 and others discuss the functions of pottery as another, equally important set of data. How they cooked, ate, and, I would add, used pottery for other purposes is equally of interest. Anthropological research shows that food choice and preparation are among the longest lasting cultural traits in every society.

in different categories and different levels of the site (Table 1).22

At Shengavit we classified all the pots, including nondiagnostic body sherds as fine, cooking, or utility wares. The tempering of each of the categories is different, as reflected in petrography. We then made typologies not on etic categories of bowls, jars, etc., but the culturally less biased emic categories of open, closed, and intermediate vessels (Figure 4). With a sample of 1430 diagnostic sherds from good proveniences we were able to analyze whether the potters were working in workshops (standardized measures) or not, factoring

Not specifically discussed here, but equally important, are chipped stone tools, metal objects, jewelry, figurines, ritual (including mortuary) items. Metals receive more attention than most other artifact categories, but even there whether the importance of metals in the KuraAraxes is because of their use or as status items because of their rarity has yet to be established.23

21

As Table 1 based on typology Figure 3 illustrates, the variability of different measurements of circumference, body thickness, and rim thickness suggest a lack of standardization. Not even the sizes of pot types are standardized. At the same time, these data imply that there was a shared tradition of general shapes among Shengavit potters, who were likely part of extended family or neighborhood production units.

22

Wilkinson 2014.

23

415

Rothman forthcoming 3. Stork 2013.

Systemizing the Past

Figure 3. Sample of Ceramics from Shengavit (after Rothman forthcoming 3, Figure 1. Figure 5e.1. Varieties of Shengavit Kura-Araxes pottery A. intermediate vessel with inscribed design, B. Closed storage vessel from under K6 building 3 floor, C. intermediate vessel, D. large closed storage vessel, E. small intermediate eating/ serving pot, F. Cooking pot, G. small intermediate eating pot, H. open vessel, I. intermediate eating pot, J. open vessel, strainer.)

416

Mitchell S Rothman: Reconstructing the Lifeways of the Kura-Araxes

Figure 4. Typology of Intermediate 1 Ceramics from Shengavit (Rothman and Manouukian Forthcoming, Figure 5e.23).

The pottery, as described above, is divided two ways. One is by whether it is fine, utility, or cooking wares. These are to some degree etic decisions, but the tempering indicates that ancient potters categorized them the same from an emic point of view. A second level is by type, which is mostly etic, although the type groups are probably related to their function. Types are either recognizable types or types of rims, bases, or bodies that are somewhat diagnostic. To some degree, the rims and profiles can be attributed to a functional category, but to be safe, I only used the clearly recognizable types.

Mapping Lifeways To pull together all these elements, I believe, we need to map the artifacts back into the best proveniences. In this case, I chose three buildings from Shengavit as examples. Two are from the latest Level I (near the end of the Shengavit Kura-Araxes occupation): Building 1 in square I14 and the room with a ritual emplacement in M5. The third is a square building in stratum III, Building 2 in square K6. The first two are complete, which is the best circumstance. The third was cut by a modern pit, but it is mostly intact. 417

Systemizing the Past Table 1.

Type

Circumference avg. cm.

Circumference range cm.

Circumference median cm.

Rim thickness avg. mm.

Rim thickness range mm.

Rim thickness median mm.

Body thickness avg. mm.

Body thickness range mm.

Body thickness median mm.

N

XX

9.39

3.6-9.37

10

6.2

3.1-9.37

4.9

7.9

3.9-12

8

38

XXI

11.95

7-24

12

5.79

4.9-9.36

5.4

8.25

5.2-14.6

7.4

21

XXII

7.8

4-43

12

4.33

3.2-11.5

8.3

6.3

4.2-19.6

8.3

46

XXIII

13

6-20

14

6.82

6.5-8.7

7.6

9.1

7.7.-12

8.4

12

XXIV

16.17

7-24

13

7.95

4.8-11.9

13

10.1

6-14.1

10

6

XXVI

12.75

5-34

9

6.5

3.7-12.4

7.2

8.6

5.3-14.1

8.9

9

The building in square I14 (Figure 5) is one of the larger square buildings with a small annex of the latest phase of the Kura-Araxes. We found another in square K6/ L6/L7, and Bayburtian mentions such a construction in his field notes. That building in square K6/L6/L7 is very similar to the building to the square I14 building. Unfortunately, being closer to the surface, its fill is mixed with modern glass and metal items, making it a less reliable comparison. So, I chose a building under the just mentioned late building in K6/L6/L7. It is a square building, as mentioned, cut by a modern pit. The third example is square M5, a room that might be part of a larger building or a separate one. It looks very much like one of the long two-room buildings from Pulur Sakyol with a ritual emplacement in the back room entered through a set of steps down into a somewhat subterranean space.24 Whether it is like Pulur Sakyol, one room of a house, or a special function room is much debated.25

and burnishing stones), agriculture (sickle blades, harrows), and maybe woodworking (hammers). The ratio of fine to utility to cooking pots sherds is 3:4:1. Counting only non-diagnostics, it is 1.4:2:1. This, in my opinion, describes a household with subsistence and craft making activities. Square I14 also has evidence of cooking and eating. It also had a significant number of querns and abraders (manos and metates), and a sickle blade. The ratio of fine to utility to cooking pots, however, is significantly different: 1:2:1. We recovered no storage pot sherds. Nor were there any evidence of weaving tools. Unlike square K6, however, there are many wedges for working bone or wood, as well as gravers, fine hammers (perhaps flint retouching tools), and arrow points. There are few remains of wild animals that the arrow points may have been used to hunt. To me this all suggests that this, too, was a domestic space. However, the craft activities engaged in by the residents of square I14 may well have been different than those of square K6, Building 2. As we complete the mapping of other buildings, this pattern may become clearer.

Comparisons of the three buildings presents an interesting emic perspective on ancient Kura-Araxes Shengavit. All show a consumption of similar animals. In all the most common are sheep/goat and cow. The ratio, however, is somewhat different. In square I14 the ratio is 1:2.6 cow to sheep/goat, in square K6 it is 1:2.28, and in M5 1:3.8. Each has a few pig, dog, and red deer bones. Interestingly, square M5 has many more fish. The disparity between square I14 and K6, on the one hand, and square M5 is interesting, perhaps relating to the sacrifices on the small altar.26 K6 Building 2 has evidence of storage, cooking, and many of the pot types I have called eating pots. Evidence of food processing (querns and grinders), exist alongside tools for weaving (spindle whorl), pottery making (gravers

Square M5 has some of the same cooking, and eating vessels as the other two places, plus more storage jars. However, the ratio of fine to utility to cooking at 2.5:2:1 has many more fine wares than the other two buildings. This pattern is somewhat contradicted by the nondiagnostics, which emphasize cooking wares. In either case the preparation and consumption of food, some prepared with querns and abraders is evidenced. There are production tools like a sickle, a few gravers, and hammers or harrows. However, only in this context did we recover three different figurines, ground stone pot stands, jewelry (a bead), and anvils. So, there are signs of some domestic activities, but more remains of symbolic, ritual behaviors than in the other two locations. This is not a space totally reserved for ritual activities, but it is a place where ritual activities

Koşay 1976. Simonyan and Rothman 2015. 26 Simonyan and Rothman 2015. 24 25

418

Mitchell S Rothman: Reconstructing the Lifeways of the Kura-Araxes

Figure 5. Symbol Set for Distribution maps (after Simonyan and Rothman Forthcoming, Chapter 7).

419

Systemizing the Past

Figure 6. Distribution of Artifacts from Building 1 Square I14, Shengavit (after Simonyan and Rothman Forthcoming, Chapter 7).

missing from the others occurred. With all the cooking pots, this could be a place of public feasting or perhaps preparing the animals before their sacrificing. The ritual hearth contained wheat and barley seeds, while the two ash-filled bins contained only clean wheat and barley remains.

Bibliography Avetisyan, P. 2017. On the issues of chronology and periodization of the Armenian Middle Bronze Age archaeological cultures. Aramazd: Armenian Journal of Near Eastern Studies 11/1-2: 57-87. Avetisyan, P. and Bobokhyan, A. 2008. The pottery traditions of the Armenian Middle to Late Bronze Age ‘Transition’ in the context of Bronze and Iron Age periodization. In K. Rubinson and A. Sagona (eds), Ceramics in Transition. Chalcolithic Through Iron Age in the Highlands of the Southern Caucasus and Anatolia (Ancient Near Eastern Studies, Supplement 27): 123–183. Leuven, Peeters. Badalyan, R. 2014. New data on the periodization and chronology of the Kura-Araxes culture in Armenia. Paléorient 40/2: 71–92.

Conclusion With just these three examples, the richness and variability of the ancient past should be clear. Once we begin to look at the ancient reality from the point of view of the ancients themselves, once we take the time to look at all of the kinds of artifacts and analyze them all, I believe we can make great progress in understanding the nature of Kura-Araxes societies, traditions, and trajectories of change that created the Armenian past. 420

Mitchell S Rothman: Reconstructing the Lifeways of the Kura-Araxes

Figure 7. Distribution of Artifacts from M5, Shengavit (after Simonyan and Rothman Forthcoming, Chapter 7).

421

Systemizing the Past

Figure 8. Distribution of Artifacts from Building 2 Square K6, Shengavit (after Simonyan and Rothman Forthcoming, Chapter 7).

422

Mitchell S Rothman: Reconstructing the Lifeways of the Kura-Araxes Badalyan, R., Smith, A. and Avetisyan, P. 2003. The emergence of socio-political complexity in Southern Caucasia. In A. Smith and K. Rubinson (eds), Archaeology in the Borderlands. Investigations in Caucasia and Beyond (Cotsen Institute Monograph 47): 144–166. Los Angeles, Cotsen Institute of Archaeology. Batiuk, S., Rothman, M. S, Samei, S. and Hovsepyan, R. 2022. Unraveling the Kura-Araxes Cultural Tradition Across Space and Time. Ancient Near Eastern Studies 59: 235–325. Dshwachischvili, A. 1998. Die Grabungen vom Berikldeebi in Georgian. Georgica 21: 7–21. Hayden, B. 1984. Are emic types relevant to Archaeology? Ethnohistory 31/2: 79–92. Koşay, H. 1976. Keban Project Pulur Excavations 1968-1970. Ankara, Middle East Technical University. Marro, C. and Hauptmann, H. (eds), Chronologies des pays du Caucase et de l’Euphrate aux IVe-IIIemillenaires. Istanbul, IFEA and Paris, De Boccard. Rotem, Y., Iserlis, M, Rosenblum, A. and Rothman, M. S. in press, A Late 4th Millennium BCE Cylinder SealAmulet from Tel Yaqush and its Contribution to the Understanding of EB I-II Communities in the Central Jordan Valley. Levant 53/1: 13–29. Rothman, M. S 2002. Tepe Gawra: The Evolution of a Small, Prehistoric Center in Northeastern Iraq. Philadelphia, University Museum Press. Rothman, M. S 2004. Studying the development of complex society: Mesopotamia in the late fifth and fourth millennia BC. Journal of Archaeological Research 12/1: 75–119. Rothman, M. S 2014. Kura-Araxes culture areas and the late 4th and early 3rd millennia pottery from Veli Sevin’s surveys in Malatya and Elazığ (Turkey). Origini 36: 37–91. Rothman, M. S 2015. The changing organization of Kura-Araxes culture. In M. Işıklı and B. Can (eds), International Symposium on East Anatolia – South Caucasus Cultures, Vol. 1: 121–131. Cambridge, Cambridge Scholars Publishing. Rothman, M. S 2017. Explaining the Kura Araxes. In K. Weber, E. Hite, L. Katchadourian and A. Smith (eds), Fitful Histories and Unruly Publics: Rethinking Temporality and Community in Eurasian Archaeology: 217–257. Leiden, Brill. Rothman, M. S 2021. Approaches to the nature of the Kura-Araxes societies in their homeland. In F. Balossi Restelli, A. Cardarelli, G.-M. Di Nocera, L. Manzanilla, L. Mori, G. Palumbi and H. Pittman (eds), Pathways through Arslantepe, Essays in Honour of Marcella Frangipane: 163–172. Rome, Universitá di Roma Sapienza.

Rothman, M. S Forthcoming 1. Chapter 5g, The Bone Tools. In H. Simonyan and M. S Rothman (eds), Shengavit: A Kura-Araxes Center in Armenia. Costa Mesa, CA: MAZDA Press. Rothman, M. S Forthcoming 2. Chapter 5f, The Ground Stone Remains. In H. Simonyan and M. S Rothman (eds), Shengavit: A Kura-Araxes Center in Armenia. Costa Mesa, CA: MAZDA Press. Rothman, M. S Forthcoming 3, Chapter 5e: The Ceramics. In H. Simonyan and M. S Rothman (eds), Shengavit: A Kura-Araxes Center in Armenia. Costa Mesa, CA: MAZDA Press. Sagona, A. 1984. The Caucasian Region in the Early Bronze Age. 3 vols (British Archaeological Reports. International Series 214). Oxford, B.A.R. International Series. Sagona, A. 2000. Sos Höyük and the Erzurum region in late prehistory: a provisional Chronology for northeast Anatolia. in C. Marro and H. Hauptmann (eds), Chronologies des pays du Caucase etde l’Euphrate aux IVe-IIIemillenaires: 329–337. Istanbul, IFEA and Paris, De Boccard. Sagona, A. 2010. Past and present directions in the archaeology of the Transcaucasus. Türkiye Bilimler Akademisi Arkeoloji Dergisi 13: 143–157. Sagona, A. 2014. Rethinking the Kura-Araxes. Paléorient 402: 23–46. Sagona, A. 2018. The Archaeology of the Caucasus. From Earliest Settlements to the Iron Age (Cambridge World Archaeology). New York, NY, Cambridge University Press. Simonyan, H. and Rothman M. S. 2015. Regarding ritual behaviour at Shengavit, Armenia. Ancient Near Eastern Studies 52: 1–46. Smith, A. 2005. Prometheus unbound, South Caucasia in prehistory. Journal of World Prehistory 19: 229–279. Steward, J. 1955. Theory of Cultural Change. Bloomington, University of Indiana Press. Stork, L. 2013. The social use of metal at the beginning of the Early Bronze Age in the Upper Euphrates Valley. In L. Bombardieri, A. D’Agostino, G. Guarducci, V. Orsi and S. Valentini (eds), SOMA 2012. Identity and Connectivity. Proceedings of the 16th Symposium on Mediterranean Archaeology, Florence, Italy, 1-3 March 2012 (British Archaeological Reports. International Series 2581/I): 25–32. Oxford, Archaeopress. Vitelli, K. 1998. ‘Looking up’ at early ceramics in Greece. In J. Skibol and G. Feinman (eds), Pottery and People. A Dynamic Interaction: 184-198. Salt Lake City, University of Utah Press. Wilkinson, T. C. 2014. The Early Transcaucasian phenomenon in structural systemic perspective: Cuisine, craft and economy. Paléorient 40/2: 203–229.

423

Middle Bronze Age Ceramics in Macro and Micro Perspectives* Karen S. Rubinson

Institute for the Study of the Ancient World, New York University, USA Abstract: Pavel Avetisyan has recently presented a chronology and proposed periodization of the ceramic groups of the Middle Bronze Age South Caucasus based on a detailed review of the ceramics, site stratifications where available, and scientific dates from a broad range of sites.Avetisyan divides the period from the third quarter of the 24th century BCE through the third quarter of the 16th century BCE into three stages which he identifies as post-Kura-Araxes I, II and III. Thus the ‘cultures’ of the Soviet tradition, summarized by Kushnareva, have been placed in a longitudinal regional development, where the divisions based on detailed ceramic typologies are designated by a less micro-geographical affiliation than in the past. In this paper, I address the issue of the traditions of naming of the ceramic groups in the last of the three phases proposed by Avetisyan, exploring how the heuristic tool of naming influences the perspective and interpretation of these ceramics in the landscape, and propose a model for interpreting the meanings of the typological groups. Keywords: Middle Bronze Age, ceramics, typology, identity, Araxes/Aras/Araks, Verin Naver, Iran, Nakhichevan

‘[T]he world of humankind constitutes a manifold, a totality of interconnected processes, and inquiries that disassemble this totality into bits and then fail to reassemble it falsify reality. Concepts like ‘nation,’ ‘society,’ and ‘culture’ name bits and threaten to turn names into things. Only by understanding these names as bundles of relationships, and by placing them back into the field from which they were abstracted, can we hope to avoid misleading inferences and increase our share of understanding.’1 Pavel Avetisyan has recently published a significant article about the Middle Bronze Age in the South Caucasus2 that was developed from a substantial study.3 Based on detailed review of the ceramics, site stratifications where available, and scientific dates from a broad range of sites, he presented a chronology and proposed periodization of the ceramic groups. Avetisyan divides the period from the third quarter of the 24th century BCE through the third quarter of the 16th century BCE into three stages which he identifies as post-Kura-Araxes I, II and III.4 Thus the ‘cultures’ of the Soviet tradition, summarized by Kushnareva,5 have been placed in a longitudinal regional development, where the divisions based on detailed ceramic typologies are designated by a less micro-geographical affiliation than in the past. Charles * It is with great gratitude that I present this paper in honor of Prof. Dr. Pavel Avetisyan. Since I first met Pavel at Horom in 1993, he has generously contributed to my research. His knowledge of South Caucasus ceramics was fundamental to the success of the South Caucasus ceramics workshop at Barnard College in 2003 (Rubinson and Sagona 2008). His scholarship is foundational to my own work on this subject. 1 Wolf 1982: 3, cited in Kohl 2008: 495. 2 Avetisyan 2017. 3 Avetisyan 2014. 4 Avetisyan 2014: 83-85; 2017: 65. 5 Kushnareva 1993: 92-171; 1997: 81-149.

Systemizing the Past (Archaeopress 2023): 424–427

Burney noted the terminological challenge fifty years ago: ‘In Trans-Caucasia the Soviet archaeological terminology has been developed without any regard to terms used in Turkey and Iran, and vice versa. With no possibility of applying the Soviet periods to the whole zone dealt with in this book, and with the manifold regional variations and local cultural provinces, new terms have been applied wherever justifiable by the context, as in the third millennium BC. Elsewhere existing terms have been as much as possible synthesized into a comprehensible whole, through complete consistency cannot be achieved.’6 This issue of terminology is a matter that I’ve addressed before7 and Avetisyan’s suggestion is welcome, since the Kura-Araxes tradition has been found in much of the South Caucasus and immediately adjacent regions,8 so that his proposal has a regional referent with some geographical, as well as archaeological, reality. For the Middle Bronze Age, Avetisyan’s proposal disentangles various ceramic traditions from the locations where they were first named, which were sometimes at the outermost locations of distribution.9 Other efforts to define these ceramic traditions have agglutinated some of the styles with site specific names under broader terms, such as Edward’s Urmia Ware,10 which, although stratified at Haftavan, Iran, with polychrome painted Burney and Lang 1971: 13. Rubinson 2004; 2005; Rubinson and Sagona 2008: 1-3. I would note that Burney’s suggestion of the Early Transcaucasian Culture for this material didn’t take widespread hold. There are other terms that have also been used, see, for example Smith 2015: 102, n. 1 and Sagona 2018: 215-216. 9 The Middle Bronze ceramic traditions as listed by Avetisyan are Early Kurgan, Trialeti-Vanadzor, Sevan-Artsakh, Karmirberd, and Karmirvank or Van-Urmia (Avetisyan 2017: 57). 10 Edwards 1981. 6 7 8

Karen S. Rubinson: Middle Bronze Age Ceramics in Macro and Micro Perspectives characterizes some of the burials he has analyzed as containing a ‘mixed complex.’ In some cases that includes ceramics typical of differently named traditions, sometimes with ornamentation typical for more than one of the traditions, and sometimes with unpainted pottery types typical for more than one of the defined traditions. Some burials contain pottery identified with two traditions and some, three traditions, such as tomb 12 at Verin Naver.20

wares occurring above black-on-red painted wares, was named without a diachronic component. This material has been called the Van-Urmia tradition and then VanUrmia ware,11 as well as Araxes/Aras Ware;12 none of these terms distinguish the diachronic differences in decoration and forms. Naming is a heuristic tool. Avetisyan notes the following ceramic traditions as chronologically part of his postKura-Araxes III stage: Karmir-Berd, Sevan-Artsakh (late complexes), Karmir-Vank, late Trialeti-Vanadzor.13 With these recent publications, Avetisyan is thus recognizing both the macro-regional picture and the micro-regional ones and untying the macro-regional picture from the differing periodization terminologies beyond the modern South Caucasus borders.

So how should we consider these ceramic groups, since burials are charged places and ceramics of these variously named styles have been found together in a single burial? I see no way to interpret the various styles as markers of identity that have strongly defined boundaries. Although there are clearly some distinctive groups of painted wares, such as the black on red jars with hanging triangles together with bowls with concentric half-circles under the rim,21 and polychrome painted wares that are late in the sequence and continue into the Late Bronze Age,22 are we working too hard to separate things that weren’t in the past, as SevanUzerlik and Karmirberd (Tazakend)? It is one thing to note the chronological change of monochrome painted wares earlier than polychrome ones, but how should we think about these other typological distinctions of ceramic decoration? It may indeed be that the diverse ceramic styles are identity markers of some kind, but I suggest that if that is the case, they must represent groups with socially established regular interactions, perhaps indeed reflecting a web of interactions based on habitual patterns of movement as Kohl suggests for other phenomena.23 Fracetti’s hypothesis of a zone of interactions he calls the Inner Asian Mountain Corridor, which he suggests is also a set of loci along which technology and ideology are transferred and exchanged,24 is useful for understanding the South Caucasus Middle Bronze Age ceramic distribution patterns. What is pertinent is his model of zones of interaction both within summer pastures among groups that come from many different lowland sites as well as more directly between the nodes of the lowland sites themselves, essentially a web of contact.

The question I raise here is how we can interpret the micro-regional terms. Kushnareva provided maps that showed the Karmirberd and Kizylvank (Karmirvank) ‘cultures’ occupying unbounded discrete geographical regions14 and she notes that as she isolated the Sevan-Uzerlik ‘culture,’ it required her to revise her definition of Karmirberd, since there are some shared characteristics between them.15 A subsequent map by Özfırat draws boundaries around the various traditions and shows the region of concentrated overlap.16 Recently Wilkinson once more placed the individual traditions in discrete physical locations, although without linear borders.17 M. Smith argues that bounded and shaded areas on maps project a certainty about boundaries and express the maximal extents of entities rather than the changes that occur over time that at some point reach these outer limits.18 Further, Kushnareva’s and Wilkinson’s maps do not convey one aspect of these ceramic distributions, that some of the defined traditions occur together within single cemeteries and individual graves. Özfırat’s map does convey regional overlap in a general way, without visually detailing the characteristics of this intersection. A. Smith wrote that it remains unclear ‘exactly what the typological distinctions in ceramics that define the end of the Middle Bronze Age represent in terms of time, space or social difference.’19 However, Avetisyan’s research provides information that I find suggestive for a tentative interpretation that I present here. Avetisyan

Most of the deeply stratified sites yielding these ceramics lie along the middle Araxes valley in Armenia and Nakhichevan,25 including the recently excavated Kızkala.26 Along the Araxes on the Iranian side, results of a recent survey between Jolfa and the KhodaAfarin Plain yielded no recognizable Middle Bronze

Rubinson 2005. Özfırat 2002, where she notes that the broad term can be subdivided into regional categories; Özfırat 2008: 106. 13 Avetisyan 2014: 82; 2017: 64. The ceramic traditions are also known by other names, see notes 6 and 11 above. 14 Kushnareva 1993: 98; 1997: 85. 15 Kushnareva 1993: 151-152; 1997: 129. 16 Özfırat 2001: 57. 17 Wilkinson 2014: 273. 18 M. Smith 2005. Smith discussed political entities, but the principle is relevant here regardless of what socio-political phenomena the Middle Bronze ceramic categories represent. 19 Smith et al. 2009: 29. 11 12

20 See, for example, Avetisyan and Bobokhyan 2008, Figures 21-22 and 25. 21 See, for example, Bahşeliyev et al. 2019: Figures 12-13; Swerida and Nugent 2019: Figures 1.7, 1.8. 22 Bahşeliyev et al. 2019: 57-58; Özfırat 2008: 105. 23 Kohl 2008: 498. 24 Fracetti 2013. 25 Özfırat 2005: 146, n. 22; Ristvet et al. 2011: 9. 26 Bahşeliyev et al. 2019.

425

Systemizing the Past pottery,27 although not far from the Araxes River basin, Urmia Ware is found at the site of Kul Tepe Jolfa (Hadishahr).28 This site might define the region through which Haftavan Tepe is connected to this Araxes basin phenomenon, since it is near the lowland river basins that run through Khoy and other sites where Urmia Ware has been found.29

archaeological cultures. Aramazd: Armenian Journal of Near Eastern Studies 11/1-2: 57–87. Avetisyan, P. and Bobokhyan, A. 2008. The pottery traditions of the Armenian Middle to Late Bronze Age ‘Transition’ in the context of Bronze and Iron Age periodization. In K. Rubinson and A. Sagona (eds), Ceramics in Transition. Chalcolithic Through Iron Age in the Highlands of the Southern Caucasus and Anatolia (Ancient Near Eastern Studies, Supplement 27): 123–183. Leuven, Peeters. Bahşeliyev, V., Ristvet, L., Gopnik, H., Nugent, S. and Swerida, J. 2019. Kızkale settlement and its necropolis (Nakhchivan, Azerbaijan). Türkiye Bilimler Akademisi Arkeoloji Dergisi 25: 47–61. Burney, Ch. and Land, D. M. 1971. The Peoples of the Hills: Ancient Ararat and Caucasus. London, Weidenfeld and Nicholson. Edwards, M. R. 1981. Pottery of Haftavan VIB (Urmia Ware). Iran 19: 101–140. Edwards, M. R. 1986. ‘Urmia Ware’ and its distribution in North-Western Iran in the second millennium BC: A review of the results of excavations and surveys. Iran 24: 57–77. Frachetti, M. 2013. Bronze Age pastoralism and differentiated landscapes along the Inner Asian Mountain Corridor. In S. A. Abraham, P. Gullapalli, T. P. Raczek and U. Z. Rizvi (eds), Connections and Complexity: New Approaches to the Archaeology of South Asia: 279–298. Walnut Creek, CA., Routledge. Kohl, Ph. 2008. Shared social fields: evolutionary convergence in prehistory and contemporary practice. American Anthropologist 110/4: 495–506. Kushnareva, K. Kh. 1993. The Southern Caucasus in the IX-II Millennia BC: Stages of Cultural and Socio-Economic Development). Saint Petersburg, Peterburgskoe vostokovedenie (in Russian). Kushnareva, K. Kh., 1997. The Southern Caucasus in Prehistory: Stages of Cultural and Socioeconomic Development from the Eighth to the Second Millennium BC. Philadelphia, The University Museum, University of Pennsylvania. Marro, C., Bakhshaliyev, V., Berthon R. and Thomalsky, J. 2019. New light on the Late Prehistory of the South Caucasus: Data from the recent excavation campaigns at Kültepe I in Nakhchivan, Azerbaijan (2012-2018). Paléorient 45/1: 81–113. Maziar, S. and Zalaghi, A. 2021. Exploring beyond the river and inside the valleys: Settlement development and cultural landscape of the Araxes River Basin through time. Iran 59/1: 36–56. Özfırat, A. 2001. Erste Betrachtungen zum ostanatolischen Hochland im 2. Jt. V. Chr. Istanbuler Mitteilungen 51: 27–60. Özfırat, A. 2002. Van-Urmia painted pottery from Hakkari. Archäologische Mitteilungen aus Iran und Turan 34: 209–228.

As has long been noted, many of the cemeteries that contain burials with the Middle Bronze painted potteries discussed here are located in summer pasture areas. And many of the deeply stratified sites that contain this pottery are in the Araxes River basin or connected to it via tributary waterways. So the model suggested by Frachetti, of groups interacting both during seasonal movement to upland pastures as well as via interconnected valleys might explain the later Middle Bronze South Caucasus patterns of ceramic distribution. The distinctively named ceramic groups, when they are not marking chronological difference, may be displays of some kinds of group identities, perhaps kinbased. And the overlapping of decorative patterns, paint palettes, and occurrence of unpainted wares, both at some settlement sites but most especially in closed burial contexts, may reflect the social interactions of these groups facilitated by their movements through and within the landscapes. The terminology proposed by Avetisyan, that is post-KuraAraxes III consisting of several variously named groups of ceramics, would fit this model, the distinctively named types as parts of ‘bundles of relationships,’ that, when placed into the landscape can be explained as a series of contemporary social interactions. I could not have proposed this hypothesis without the detailed research of Avetisyan, as well as his framing the results of his work in a time-slice broadly defined. Bibliography Abedi, A, and Omrani, B. 2015. Kura-Araxes Culture and North Western Iran: New perspectives from Kul Tepe Jolfa (Hadishahr). Paléorient 41: 55–68. Avetisyan, P. 2014. Armenian Highland in the 24th9th Centuries BC (Dynamics of Socio-Cultural Transformations according to Archaeological Data). Yerevan, Institute of Archaeology and Ethnography, NAS RA (in Armenian with Russian and English summaries). Avetisyan, P. 2017. On the issues of chronology and periodization of the Armenian Middle Bronze Age

Maziar and Zalaghi 2021: 44. Abedi and Omrani 2015; Maziar and Zalaghi 2021: 43. 29 Edwards 1986: 68. As Abedi and Omrani note, Kul Tepe Jolfa ‘lies in a strategic position, next to the broad middle Araxes valley, at the crossroads of major routes linking the Iranian plateau to Anatolia, the South Caucasus and North Mesopotamia’ (2015: 56). And see Marro et al. 2019: 82-83 for these same interconnected routes in earlier periods. 27 28

426

Karen S. Rubinson: Middle Bronze Age Ceramics in Macro and Micro Perspectives Özfırat, A, 2005. Transhumance on the Eastern Anatolian High Plateau in the 2nd mill. BC. Archäologische Mitteilungen aus Iran und Turan 37: 139–152. Özfırat, A. 2008. Highland plateau of Eastern Anatolia in the second millennium BCE. In K. Rubinson and A. Sagona (eds), Ceramics in Transition. Chalcolithic Through Iron Age in the Highlands of the Southern Caucasus and Anatolia (Ancient Near Eastern Studies, Supplement 27): 101–121. Leuven, Peeters. Ristvet, L., Baxşəliyev, V., Aşurov, S., Earley-Spadoni, T. 2011. Settlement and society in Naxçivan Archaeological Project. Iranica Antiqua 46: 1–53. Rubinson, K. S. 2004. Dinkha Tepe, Iran, and so-called Urmia Ware. In A. Sagona (ed.), A View from the Highlands: Archaeological Studies in Honour of Churles Burney (Ancient Near Eastern Studies, Supplement 12): 661–676. Louvain, Peeters. Rubinson, K. S. 2005. Second millennium BC painted potteries and problems of terminologies. Archäologische Mitteilungen aus Iran und Turan 37: 133–138. Rubinson, K. S. and Sagona, A. 2008. Introduction: A question of nomenclature. In K. Rubinson and A. Sagona (eds), Ceramics in Transition. Chalcolithic Through Iron Age in the Highlands of the Southern Caucasus and Anatolia (Ancient Near Eastern Studies, Supplement 27): 1–7. Leuven, Peeters.

Sagona, A. 2018. The Archaeology of the Caucasus. From Earliest Settlements to the Iron Age (Cambridge World Archaeology). New York, NY, Cambridge University Press. Smith, A. T. 2015. The Political Machine. Assembling Sovereignty in the Bronze Age Caucasus. Princeton and Oxford, Princeton University Press. Smith, A. T., Badalyan, R. and Avetisyan, P. 2009. The Archaeology and Geography of Ancient Transcaucasian Societies, Vol. 1: The Foundations of Research and Regional Survey in the Tsaghkahovit Plain, Armenia (Oriental Institute Publications 134). Chicago, Oriental Institute of the University of Chicago. Smith, M. 2005. Networks, territories and the cartography of ancient states. Annals of the Association of American Geographers 94/4: 832–849. Swerida, J. and Nugent, S. 2019. Fashioned identity in the Şərur valley, Azerbaijan: Kurgan CR8. In M. Cifarelli (ed.), Fashioned Selves: Dress and Identity in Antiquity: 11–26. Oxford and Philadelphia, Oxbow Books. Wilkinson, T. C. 2017. Tying the Threads of Eurasia: TransRegional Routes and Material Flows in Transcaucasia, Eastern Anatolia and Western Central Asia, c. 3000-1500 BC. Leiden, Sidestone Press. Wolf, E. R. 1982. Europe and the People Without History. Berkeley, University of California Press.

427

Monitoring Heritage At Risk: Caucasus Heritage Watch and the Armenian Monuments of Nagorno-Karabakh Adam T. Smith

Cornell University, Ithaca, NY, USA

Lori Khatchadourian

Cornell University, Ithaca, NY, USA

Ian Lindsay

Purdue University, West Lafayette, IN, USA Abstract: Caucasus Heritage Watch (CHW) was founded in 2020 in the wake of the second Nagorno-Karabakh war, to monitor and document endangered and damaged cultural heritage using high-resolution satellite imagery. The proliferation of highresolution commercial satellite imagery has transformed the ways in which archaeologists can respond to cultural heritage crises. CHW introduces the techniques of satellite surveillance of endangered heritage to the South Caucasus and pursues a model of engaged, public archaeology that provides authoritative data for local heritage management decisions, and forensic evidence for international preservation efforts. On a regular monitoring cycle, we have hundreds of discrete cultural heritage sites under satellite observation, including churches and mosques, cemeteries and khachkars, bridges and fortresses, and other cultural properties that tell the dynamic story of centuries of life in the region. In this article, composed in honor of Pavel Avetisyan, we outline the steps in establishing our monitoring process, share some of our early findings, and narrate our workflow for the regular observation of heritage sites at risk. Keywords: Cultural heritage, remote sensing, satellite observation, heritage monitoring, Nagorno-Karabakh, South Caucasus

Introduction The preservation of prehistoric and historic sites has long been fundamental to Pavel Avetisyan’s vision of archaeological practice. From his leadership in establishing the Agarak Historic-Cultural Preserve,1 the first prehistoric site to attain such a designation in Armenia, to his pioneering support for digitizing the Institute of Archaeology and Ethnography’s archives, Pavel has brought a 21st century vision of preservation and Digital Humanities scholarship to Armenia’s Academy of Sciences. In this contribution, we offer a summary of the developing work of Caucasus Heritage Watch (CHW) in recognition of Pavel’s deep commitment to the archaeological resources of the region. Archaeology from Space CHW was initiated in late 2020 by the co-authors to monitor endangered cultural heritage sites in Nagorno-Karabakh in the wake of the recent war using spaceborne remote sensing technologies (Figure 1).2 Avetisyan 2008. We are grateful to the Armenian General Benevolent Union (AGBU) and the Aragats Foundation for emergency funding in support of this research. 1 2

Systemizing the Past (Archaeopress 2023): 428–439

Broadly speaking, archaeological remote sensing entails prospecting methods that employ terrestrial, airborne, or spaceborne sensors to detect archaeological features by their physical, chemical, or magnetic properties. The technology has come a long way since the Royal Engineers’ Balloon Section captured the first aerial photos of Stonehenge in 1906, revealing traces of ancient human occupation via subtle landscape features—crop marks, soil marks, and shadow marks—only visible when viewed from above.3 This work demonstrated the extraordinary power of aerial imaging to aid in the discovery and interpretation of archaeological sites. While aerial photography continues to play an important role in the study of historic landscapes, archaeologists began exploiting the advantages of spaceborne earth observation soon after the Landsat satellites launched in 1972.4 The subsequent appearance of high-resolution datasets with the IKONOS program in 1999 only broadened interest in these tools and their applications to a range of research problems focused on site discovery, land use classification and change, as well as monitoring and heritage conservation.5 Recently declassified Cold War-era reconnaissance Crawford 1923. Parcak 2009. 5 Agapiou 2017. 3 4

Adam T. Smith: Monitoring Heritage At Risk

Figure 1. Map of the Nagorno-Karabakh region.

Some of these efforts are the result of academic collaborations, such as the Endangered Archaeology in the Middle East and North Africa (EAMENA) project led by Oxford, Leicester, and Durham Universities,10 as well as partnerships between scholars and government (e.g. the Syrian Cultural Heritage Initiative). Satellitebased monitoring of endangered cultural heritage sites not only provides vital data to researchers concerning changes to archaeological resources, but authoritative information for international and governmental agencies involved in the protection of cultural heritage.

imagery, very high-resolution (< 1 m) commercial satellite data (e.g., Maxar WorldView, Planet SkySat), new high- to medium-resolution open access resources such as Sentinel (10 m) and PlanetScope (3 m) data, and the rising popularity of affordable Unmanned Aerial Vehicle (UAV) technologies, have all provided a flood of data-rich 2D and 3D products, including georeferenced orthoimagery, near-surface geophysical survey data, digital elevation models (DEMs), and Light Detection and Ranging (LiDAR) imagery. This new generation of imagery and data products is rapidly transforming archaeological research and heritage management.6

As a result of the Russia-brokered ceasefire to the Second Nagorno-Karabakh War, new threats to the region’s heritage sites now loom, as roughly 1000 medieval and early modern monuments now fall under Azerbaijani jurisdiction. The troubling history of willful, state-sponsored cultural erasure in Nakhichevan, from the appropriation of churches into the pseudohistory of Caucasian Albania to the wholesale destruction of the medieval Armenian cemetery at Djulfa,11 gives cause for concern regarding the fate of these monuments (see CHW’s September 2022 Special Report #1 and accompanying StoryMap web platform entitled ‘Silent Erasure: A Satellite Investigation of

In the 21st century, satellite imagery has proven an effective tool for monitoring cultural heritage at risk. Researchers have used it to document the looting and destruction of cultural heritage sites in Syria, Iraq, Egypt, Afghanistan, Yemen, and China.7 Earth observation using satellite imagery has also been used for cultural heritage disaster risk management8 and assessing threats posed by development and resource extraction.9 6 Adamopoulos and Rinaudo 2020; Luo et al. 2019; Opitz and Hermann 2018. 7 e.g., Casana 2015; Casana and Laugier 2017; Fradley and Hardouin 2019; Hammer et al. 2018; Lauricella et al. 2017; Parcak et al. 2016; Ruser et al. 2020. 8 e.g., Agapiou et al. 2020. 9 Rayne et al. 2020.

10 11

429

Zerbini 2018. AAAS 2010; Maghakyan 2019; see also Maghakyan 2021.

Systemizing the Past

Figure 2. Screenshot of the Caucasus Heritage Watch Dashboard from October 17, 2022.

international agencies and publics; c) provide a forensic resource so that publics can hold responsible parties accountable for harms, including their own leaders; and d) innovate new and transferrable methodologies, and disseminate workflows that can empower our partners in the region and assist researchers working in other conflict zones.

the Destruction of Armenian Cultural Heritage in Nakhchivan’ https://storymaps.arcgis.com/stories/ 48703f664f2f467b8f4f42008d8c75da; https://indd.adobe .com/view/2a6c8a55-75b0-4c78-8932-dc798a901 2fb12).13 On-the-ground monitoring of these sites is not possible; as of June 2021, the government of Azerbaijan has declined to permit UNESCO’s mission to access sites, and scholars with expertise in the region’s cultural heritage cannot enter Azerbaijan. As stewards of the human past, archaeologists of the South Caucasus are faced with the challenge of responding to a cultural heritage crisis in which the possibilities for action are severely constrained.

Such ambitious goals require working in two directions. First, we seek to closely observe changes to atrisk sites by tasking earth observation satellites to capture images on a regular cycle. These changes are documented in real time on our web dashboard (Figure 2) and compiled in our seasonal reports published on our website (http://caucasusheritage.cornell.edu). Second, we are committed to looking back in time in order to study past episodes of heritage destruction that can help us document the ways in which war and ethnic conflict have impacted the region’s heritage resources over time.

CHW’s mission is to monitor endangered cultural heritage in the region using high-resolution satellite imagery. Our goals are to a) provide reliable information on the condition of sites to counter state denialism, falsification, and other abuses that place heritage sites at the center of political conflict; b) discourage or restrain state actors from intentional erasure both through the act of bearing witness, and by the dissemination of authoritative research to relevant national and

Monitoring the Present In order to regularly monitor the condition of regional heritage sites, the first task was to build a comprehensive database with verified geographic coordinates.

Khatchadourian et al. 2022. 13 See Petrosyan 2020 for a historical overview of threats to Armenian heritage in Azerbaijan. 12 

430

Adam T. Smith: Monitoring Heritage At Risk

CHW monitoring pipeline & workflow

Vendors

Maxar Vivid

Cornell

ArcGIS Online

Purdue

Planet Date storage

Monitored sites web layer

ArcGIS Pro

local net

Data storage local net

download

sync

Imagery folders

ArcGIS Pro

Monitoring

sync

CHW web map

Monitoring

Imagery folders

Dashboard

Globus file transfer

Figure 3. Diagram of the CHW monitoring workflow.

Sarhat Petrosyan assisted with refining geographic coordinates for sites in Shushi, while independent scholar Dr. Husik Ghulyan provided a shapefile of the post-war borders of Nagorno-Karabakh, helping us discern sites located on or near the line-of-contact. The expertise of all of these partners has been vital.

The official monument lists for Nagorno-Karabakh, covering the former districts of Askeran, Hadrut, Karvachar, Martakert, Martuni, Qashatagh, and Shushi, were used as the starting point, along with a site list provided by the Armenian Apostolic Church. These lists were translated and re-formatted in preparation for inclusion in a geodatabase, with each site classified by type and date, and further described based on location to nearest village, elevation, data source, and priority (sites were designated ‘vital’, ‘high’, ‘medium’ or ‘low’ based on their architectural, cultural, or art historical importance). Several months were devoted to verifying geographic coordinates, and in many cases, identifying more precise locational information than that provided on the available lists through careful use of Google Earth and consultations with our partners. We have been fortunate to have the team at Monument Watch and Research on Armenian Architecture as partners in this database building effort. In addition, architect

After assembling and aligning site information from existing datasets and fixing their geographic locations, the second phase of our pre-monitoring work was to design a GIS geodatabase to manage these data. We elected to build our geodatabase in Esri’s commercial ArcGIS environment due to the interoperability of webbased, desktop, and mobile interfaces that facilitate seamless collaborative workflows between our institutions at Cornell and Purdue. In addition, GIS file formats are easily shared with open-source platforms, such as QGIS or GRASS, which are seeing more regular use in the South Caucasus and can be used in the 431

Systemizing the Past

Figure 4. Impact Assessment for Mets T’agher Cemetery

future to integrate workflows between CHW and our partner organizations. With the GIS environment in place, we were ready to define Areas of Interest (AOIs) for satellite image capture, embracing as many sites as possible (emphasizing ‘vital’, ‘high’, and ‘medium’ priority sites).

for the region came from the constellation of satellites operated by Maxar Technologies, including WorldView 1 and 2 and GeoEye1. These platforms provide orthorectified optical imagery with a resolution of approximately 0.5 m. We opted to utilize Maxar’s Vivid product for pre-war baseline as these tiled images are assembled out of the best images from the last 1-2 years from all of the available Maxar platforms. The images are stitched together to provide a seamless basemap but metadata still allows us to clearly define the dates of original capture.

With the database taking shape, our next task was to secure reliable baseline data to provide a clear sense of the condition of heritage monuments in the region prior to the 2020 war. Here, the most reliable imagery 432

Adam T. Smith: Monitoring Heritage At Risk

Figure 5. Impact Assessment for Shushi/Shusha’s Northern Cemetery

constellation consists of 21 satellites orbiting the Earth and capturing imagery 5-7 times per day, providing us the data we need to regularly assess site conditions and inform regional stakeholders in a timely manner.14

Once baseline imagery was in hand, the next challenge was securing a satellite platform for tasking regular image captures of the AOIs. Although the commercial satellite market has expanded rapidly over the last decade, the most obvious vendor for our needs in terms of cost and scheduling ability was Planet Labs. Planet’s SkySat platform provides us the ability to ‘task’ their satellites to provide their highest resolution (50 cm), multispectral (4-band) imagery of specific at-risk locations essentially on-demand. The SkySat

With baseline and tasked imagery in hand, the monitoring process could begin. During spring 2021, our team included the co-authors and Salpi Bocchieriyan. 14

433

Planet Labs 2021.

Systemizing the Past

Figure 6. Impact Assessment for Ghazanchets’ots Cathedral.

Monitoring proceeds by comparing the before and after imagery for each site and the surrounding landscape. Observations of each site are then entered into the GIS using a standardized set of assessments (e.g., no impact, destroyed, damaged, threatened). The system was designed such that observations made by CHW team members working in different locations could be synchronized in real time, and visualized on a shared web map that feeds our results to an online Dashboard accessible to the public via the CHW website. When an analyst observes a concerning change, consultations

are held with our partners and as a team to work toward a strategic response. When CHW determines that public scrutiny might blunt further intentional or accidental damage to a site, we use social media to broadcast the threat and to help focus the attention of relevant organizations, analysts, journalists and authorities. Results of the Spring 2021 Monitoring Cycle The work of site monitoring in spring of 2021 engaged our full team of researchers in comparing baseline and 434

Adam T. Smith: Monitoring Heritage At Risk

Figure 7. Impact Assessment for KanachZham Church.

recently captured imagery to detect changes in sites indicative of destruction, damage, or an impending threat. We monitored 275 distinct sites, assessing change and defining the parameters for our categories of destruction, damage, and threat in an iterative and collaborative process. Of the three categories of impact we observed, site destruction was typically the easiest to define since the force necessary to demolish a structure or erase a cemetery (e.g., from bulldozers and other heavy equipment) leaves significant scars on the landscape large enough to be seen in satellite

imagery. For example, Figure 4 juxtaposes baseline imagery from the cemetery at Mets T’agher (in the former district of Hadrut) with post-conflict captures dated, 8 April 2021, and 16 May 2021. The destruction of standing monuments and extensive grading of the earth is clearly visible. In a subsequent image from 16 May 2021, additional destruction is in evidence as the graded soil has been piled alongside the road path. Recent terraforming, road building, and other landscape alterations that have stripped existing turf or trees are rendered particularly visible in false color 435

Systemizing the Past

Figure 8. Threat Assessment for Vank’asar Church released April 19, 2021.

most challenging in the case of Kanach Zham church (Figure 7), also in Shushi/Shusha. Here, clear evidence of extensive destruction was visible in a comparison of our baseline imagery with a Google Earth image from 14 February 2021. The western cupola is completely destroyed and the roof has sustained heavy damage. An image analyzed from social media during the weeks just after the ceasefire clearly confirms this damage. However, the status of the eastern cupola was more uncertain. The profile of the church’s shadow in the 14 February image appears to show a surviving remnant of the cupola. In our April SkySat image, the shadow line is not sufficiently well-defined to provide clarity as to whether that part of the structure still remained or rather had instead been destroyed, evidence of a post-conflict attack on the structure. We will of course continue to monitor Kanach Zham for evidence of continuing attack. Such cases also vividly illustrate how the seasonal and diurnal lighting conditions on the ground at the moment satellite imagery was captured can impact data interpretation.

composites, or band combinations that utilize the nearinfrared band to highlight the density of vegetation (or its tell-tale removal) near a site. Damage to heritage sites had a wider range of image signatures. The most obvious were locations where new construction impinged on heritage sites. In Shushi/ Shusha, the historic northern cemetery was damaged by the construction of a new building and enclosure fence (Figure 5). A published map of the cemetery15 provided to us by Monuments Watch suggests that the construction likely destroyed numerous individual plots and their associated funerary monuments. But damage also could be less directly visible. In Shushi/ Shusha’s southern cemetery, for example, imagery suggests the vandalism of a covering over a single plot. These smaller impacts are individually of less significant impact, but over time can add up to severely damaged heritage sites. For this reason, CHW’s longterm approach to monitoring will be able to regularly assess the cumulative effects of both large and small scale impacts.

Defining threats to heritage sites has engaged the team in extensive discussion. In some respects, we understand all of the sites in our database as at risk. This is why we attempt to monitor them. But the assessment of heritage as ‘threatened’ within our monitoring classification system means something more imminent, specific, and actionable. Imminence is generally defined by the spatial proximity of potentially destructive equipment (e.g., earth movers,

Moreover, damage assessments conducted so soon after the recent conflict also clearly described changes to sites that occurred during the war itself. The most obvious of these is the hole in the roof of Ghazanchet’tsots Cathedral (Figure 6). Separating out damage during conflict from damage afterwards was 15

Harutyunyan 2008.

436

Adam T. Smith: Monitoring Heritage At Risk In order to transform an assessed threat into action that might reduce the possibility of damage or destruction, it is important that our observations be shared with a wide public. We publicize our observations via our Twitter feed, our website, and regular summary reports. The Twitter feed has been a particularly interesting lens onto the role of social media in raising awareness about the threats to heritage within situations of sustained conflict. CHW’s first alert (issued on 19 April 2021) called attention to the potential threat at Vank’asar church (Figure 9). This initial alert had over 125,000 ‘impressions’ (a tally of all the times the tweet has been seen) and over 5,500 ‘engagements’ (the total number of times users interacted with a tweet). Our subsequent documentation of the destruction of the cemetery at Mets T’agher gained over 377,000 impressions with over 13,000 total engagements. A tweet reporting damage to the Shushi/Shusha Northern Cemetery received over 280,000 impressions and almost 10,000 engagements. Although early in the process of developing a public face for our monitoring process, it is clear that Twitter’s reach is far beyond anything we could achieve through traditional means of media engagement (much less the reach of academic journals). CHW also has a Facebook page, but our data on activity there is still too preliminary to offer any insights into the impact of social media on heritage monitoring programs. Heritage Destruction in the Past CHW’s long-term goal is not only to monitor current threats to Armenian cultural heritage, but to document the impact of ethnic conflict between Armenians and Azerbaijanis on both communities’ cultural heritage. This archival dimension of our work requires building a second geodatabase (of Islamic sites in NagornoKarabakh and Armenian sites in Nakhichevan), and identifying and analyzing historical satellite imagery in order to chart changes to (or destruction of) Armenian and Azerbaijani cultural heritage sites in the years between the First and Second Nagorno-Karabakh wars. As an initial step forward in this research, CHW assisted in the satellite image procurement and analysis for a recent investigative study by Simon Maghakyan (2021), published in The Art Newspaper, documenting the erasure of churches in the village of Agulis, Nakhichevan (Figure 6). CHW utilized imagery from the declassified KH-9 Hexagon photographic reconnaissance satellite, available from the USGS Earth Resources Observation and Science (EROS) archives, to confirm the obliteration of seven churches in this one village alone.

Figure 9. Hexagon Satellite Image of the Town of Agulis (Nakhichevan A.S.S.R.), October 8, 1977 (USGS EROS).

large military vehicles, or construction equipment). In the case of Amenap’rkitch chapel in Mets T’agher, the bulldozing of an immediately adjacent space to create a staging area for road construction vehicles presented clear evidence of a real threat. In the case of Vank’asar Church, the presence of heavy equipment in the parking area along with a possible temporary structure also suggested a threat (Figure 8). Indeed, in this case the remoteness of the site and the lack of other potential explanations for the presence of equipment made the threat assessment particularly acute. In other cases, such as around the monastery of Tsitserrnavank and the nearby Palace of Melik’ Haykaz, large vehicles were observed in the vicinity of the sites, but they were parked and not actively engaged in activity risking the safety of these properties. As a result, such conditions were noted in our database for future reference, but they did not elevate the status of the sites to ‘threatened.’

Archival image analysis necessarily follows a different workflow than our regular monitoring activities. For our regular monitoring, we are keen to narrow our time windows as much as possible, ideally resolving impact events to a span of a few weeks. With archival imagery 437

Systemizing the Past serving as baseline data, our temporal resolution is obviously much coarser. Typically, the available baseline imagery will provide a pre-independence date from the 1970s or early 1980s. Following that, the next available imagery of sufficient resolution is typically from the early 2000s, when high-resolution imagery was just starting to become available. The sources for the early images vary but can include CORONA, Gambit, Hexagon, the European Space Agency’s Copernicus platform, NASA’s Landsat 7 sensor, and QuickBird, a private satellite managed by DigitalGlobe (now Maxar Technologies). Of these, QuickBird’s panchromatic camera boasted the highest resolution of approximately 0.61 m. However, it was not until the 2010s that the pace of image captures in the region became sufficiently regular, and their release on Google Earth became sufficiently timely, to allow for more sophisticated time-series analysis. In combination with archival and ethnohistoric research, these earlier satellite resources will be key to the archival dimension of CHW’s work to document the condition of not only medieval Armenian monuments, but also mosques, mausolea, and Azerbaijani cemeteries.

heritage, and bring the safeguarding of sites into processes of reconciliation. CHW asks whether spacebased imaging technologies can: a) reduce state-sponsored aggression against tangible cultural heritage through sustained, systematic, non-partisan, publicly-visible surveillance; b) create documentation that can provide a forensic resource for holding parties accountable; and c) engage stakeholders, from minorities who are the targets of erasure to majority populations who might be allies in preservation. In seeking to harness 21st century technologies to help take heritage sites off of the front lines of regional conflicts, we are not naive as to the seriousness of the obstacles we face, nor the authoritarian ideologies that frame the heritage of the ‘Other’ as a political threat to national integrity. Hence our documentation is also intended to provide a detailed forensic record of attacks on cultural sites that can be used by organizations from UNESCO to the International Criminal Court. Only organizations with global standing can truly provide a strong deterrent to heritage erasure. But as is evident from prior cases of heritage erasure ignored, the vigilant interest of concerned parties is vital to keeping the attention of the global community. CHW’s longterm commitment to its monitoring program is just one part of what will need to be a concerted effort to raise and maintain awareness, a project where our partners are vital to both research and community engagement. We look forward to building and strengthening these partnerships, including with the Institute of Archaeology and Ethnography, which Pavel Avetisyan has led so ably.

Heritage Past and Future CHW poses the question as to whether 21st century technologies can help to deter cultural erasure. Does the proliferation of satellite imaging platforms offer the possibility to not just document damage to heritage sites but significantly reduce the threat of erasure? Existing international conventions for the protection of cultural heritage were not designed to respond to current threats to humanity’s fragile archaeological record. The devastation of monuments in World War II gave rise to important UNESCO protocols for safeguarding heritage in armed conflict. These conventions rely on the cooperation of signatory state parties in their implementation and presume the promulgation of national laws for safeguarding heritage from domestic threats, such as development and looting. But during the first decades of the 21st century, cultural heritage has been moved to the front lines of battles between repressive governments and minority communities. And as a result, cultural erasure has emerged as a distressing new weapon of state actors that seek to suppress minority claims to civil rights by eradicating their pasts. From Afghanistan to Syria, Iraq to China, Yemen to Azerbaijan, cultural genocide now looms as the most pressing threat to human heritage, and we have no effective tools to deter it.

Bibliography AAAS (American Association for the Advancement of Science) 2010. High-Resolution Satellite Imagery and the Destruction of Cultural Artifacts in Nakhichevan, Azerbaijan. Report of the Geospatial Technologies and Human Rights Project. Adamopoulos, E. and Rinaudo, F. 2020. UAS-based archaeological remote sensing: review, metaanalysis and state-of-the-art. Drones 4/3: 46. Agapiou, A. 2017. Remote sensing heritage in a petabyte-scale: satellite data and heritage earth engine applications. International Journal of Digital Earth 10/1: 85–102. Agapiou, A., Lysandrou, V. and Hadjimitsis, D. 2020. Earth observation contribution to cultural heritage disaster risk management: Case study of eastern Mediterranean open air archaeological monuments and sites. Remote Sensing 12(1330): 1–12. Avetisyan, P. 2008. Protection of archaeological monuments in Armenia, the excavations in Agarak.

Although attacks on cultural heritage are broadly deplored, there has been little experimental work on how to prevent it. Despite advances in our ability to document, reconstruct, and recreate damaged sites, this work comes too late. We need tools that can help deter cultural genocide, blunt its impact on tangible 438

Adam T. Smith: Monitoring Heritage At Risk In Protection of Armernian Cultural Heritage in Armenia and Abroad: 124–128. Yerevan, Ministry of Culture. Casana, J. 2015. Satellite imagery-based analysis of archaeological looting in Syria. Near Eastern Archaeology 78/3: 142–151. Casana, J. and Laugier, E. J. 2017. Satellite imagerybased monitoring of archaeological site damage in the Syrian civil war. PLoS ONE 12(11): 1–31. Crawford, O. G. S. 1923. Air Survey and Archæology. The Geographical Journal 61/5: 342–60. Fradley, M. and Hardouin, S. 2019. Remote sensing of endangered archaeology on Gebel Ataqah, Egypt. Palarch’s Journal of Archaeology of Egypt/Egyptology 16/2: 1–21. Hammer, E., Seifried, R., Franklin, K. and Lauricella, A. 2018. Remote assessments of the archaeological heritage situation in Afghanistan. Journal of Cultural Heritage 33: 125–144. Khatchadouran, L., Smith, A. T., Ghulyan, H. and Lindsay, I. 2022. Silent Erasure: A Satellite Investigation of the Destruction of Armenian Cultural Heritage in Nakhchivan, Azerbaijan. Ithaca: Caucasus Heritage Watch. Lauricella, A., Cannon, J., Branting, S. and Hammer, E. 2017. Semi-automated detection of looting in Afghanistan using multispectral imagery and principal component analysis. Antiquity 91: 1344– 1355. Luo, L., Wang, X., Guo, H., Lasaponara, R., Zong, X., Masini, N., Wang, G., Shi, P., Khatteli, H., Chen, F., Tariq, S., Shao, J., Bachagha, N., Yang, R. and Yao, Y. 2019. Airborne and spaceborne remote sensing for archaeological and cultural heritage applications: A

review of the century (1907–2017). Remote Sensing of Environment 232: 111280. Maghakyan, S. 2019. A Regime Conceals Its Erasure of Indigenous Armenian Culture. Hyperallergic February 18, 2019. Maghakyan, S. 2021. Special investigation: Declassified satellite images show erasure of Armenian churches. The Art Newspaper 1 June 2021. Opitz, R. and Herrmann, J. (2018). Recent trends and long-standing problems in archaeological remote sensing. Journal of Computer Applications in Archaeology 10(10): 1–24. Parcak, S. 2009. Satellite Remote Sensing for Archaeology. New York, Routledge. Parcak, S., Gaithings, D., Childs, C., Mumford, G. and Cline, E. 2016. Satellite evidence of archaeological site looting in Egypt: 2002-2013. Antiquity 90: 188– 205. Petrosyan, H. L. 2020. Ethnocide in Artsakh: the Mechanisms of Azerbaijan’s Usurpation of Indigenous Armenian Cultural Heritage. In S. Maghakyan (Trans.) Cultural Heritage Experiences and Perspectives in International Context: 79–90. Yerevan, National Gallery of Armenia. Planet Labs. 2021. Real-Time Satellite Monitoring with Planet. Ruser, N., Lebold, J., Munro, K. and Hoja, T. 2020. Cultural Erasure: Tracing the Destruction of Uyghur and Islamic Spaces in Xinjiang. Report from the Australian Strategic Policy Institute. Zerbini, A. 2018. Developing a Heritage Database for the Middle East and North Africa. Journal of Field Archaeology 43 (sup1): S9–18.

439

Tigran the Great and Mithradates Eupator: Two Parallel Kings of Kings? Giusto Traina

Sorbonne University, Paris, France Abstract: Although Mithradates Eupator and Tigran the Great were joint allies in their conflict against Rome, the Roman sources (and subsequently modern scholarship) has prioritized the portrait of Mithradates, considered the main enemy. The most important element of comparison is the Oriental title of ‘king of kings’, attested for both monarchs. In any case, as a result of the development of the Eastern policy between the heyday of Roman imperialism and the pax Augusta, the Roman sources focus on the Mediterranean theatre of operations. Keywords: Ancient Armenia, Roman history, Ancient kingship, Mithridates, Tigranes the Great

Although Max Weber’s definition of charismatic rulership cannot be indiscriminately applied to all ancient societies,1 nevertheless we can deploy charismatic elements to understand the geopolitical balance of power in several historical moments. This is particularly interesting for the historical context of Mithradates Eupator’s conflict with Rome, where different political experiences were at stake. Tigran ‘the Great’ of Armenia represents an interesting case.2 Mithradates’ main ally in the coalition against Rome ruled Armenia from 95 to c. 55 BCE and tried to transform his state into a Hellenistic power, is somewhat neglected by modern scholarship. Endorsing the bias of Roman sources, modern scholarship has prioritized the portrait of Mithradates. Even Théodore Reinach, who harshly criticized Mithradates as an Oriental despot, nevertheless respected the king’s bravery; by contrast, he considered Tigran a coward and an opportunist.3 In subsequent years, the judgement has not changed.4

History, Adrian Nicholas Sherwin White claimed that before Lucullus’ campaigns ‘Tigranes had no previous connections with Rome’.5 This may be true only until 95 BCE, when Tigran mounted the throne of Greater Armenia, after several years spent as a hostage in Parthian Babylon,6 but it is hard to believe that the Roman Senate ignored the kingdom of Greater Armenia as an important element of its eastern policy.7 As a result, Roman historians selectively focused on one main enemy, that is, Mithradates Eupator. The vulgate of the Mithridatic wars gives him the role of chief villain and antihero of the story. This is clearly expressed by Justin 37.1.7: cuius ea postea magnitudo fuit, ut non sui tantum temporis, uerum etiam superioris aetatis omnes reges maiestate superauerit bellaque cum Romanis per XLVI annos uaria uictoria gesserit ‘whose greatness was afterwards such that he surpassed all kings, not only of his own but of preceding ages, in glory, and carried on war against the Romans, with various success, for forty-six years’.8

Ancient historians have usually tended to marginalize the political and diplomatic role of Armenia: for example, in his chapter in the Cambridge Ancient

This logic is also reflected by the periodization of the Mithridatic wars. According to Florus (Epitome 1.40.2) and Eutropius (Breviary 6.12; see Orosius, Histories against the Pagans 6.1.30), the conflict lasted forty

1 Literature in Gotter 2008; Linke 2015; Sommer 2011. For the evolution of the term charisma in the ‘vernacular’ dimension of the social sciences, see Derman 2012: 212-215. 2 On the title of megas basileus see Muccioli 2013: 395-417. 3 See, e.g., the harsh verdicts in Reinach 1890: 344-345. 4 The only exception came from Armenian scholars, but the only work translated in a Western language was Hakob Manandyan’s valuable biography of Tigran (Manandian 1963; the first Armenian edition dates from 1940). When Jérôme Carcopino, the doyen of Roman studies in France, wrote a foreword of the French translation of the book (1963), although he generally appreciated it, nonetheless criticized Manandyan’s excessive indulgence towards Oriental monarchs. The other scholars reacted more critically, so that, despite the availability in a Western language (which rarely occurs, even for the best examples of Armenian scholarship), Manandyan’s book is seldom mentioned. An English translation by G. A. Bournoutian, with slight complements in the footnotes, is now available: Manandyan, H. 2008, Tigranes II and Rome. A New Interpretation Based on Primary Sources, Costa Mesa, Ca., Mazda Publishers.

Sherwin-White 1992: 238. Geller and Traina 2013. On the historical context, see Dąbrowa 2021: 43-45. 7 After all, although Hannibal’s activity as a military adviser for the foundation of Artaxata may have been considered an affront (Strabo, Geography, 12.14.14; Plutarch Lucullus, 3-4; see Khachikyan 2006), the full independence of the kingdom with Artašēs (Artaxias) is a side effect of the peace of Apamea in 188 BCE (Traina 2017a). Moreover, Artašēs’s successor Tigran I did support Rome in the Third Punic war with a cavalry unit. Ampelius 32.1 says on the kings of Armenia and Cappadocia: Tigranes, qui iam scriptus est, qui tertio Punico bello populo Romano iuvit sub Manilio consule [149 bc] et Scipione Aemiliano. See also Ampelius 46, 7, referring to the destruction of Carthage by Scipio Aemilianus ‘together with Tigranes’ (una cum Tigrane). Pierangelo Buongiorno (2021: 90) rightly speaks of ‘a dialogue between Rome and Armenia, often conflicting, but nevertheless fruitful’. 8 In his commentary, Ballesteros Pastor 2013: 114-116 considers it an obituary of the king. On Justinus’ judgement on Mithradates see also Ballesteros Pastor 2016: 86-90; Traina 2016b: 107-109.

Systemizing the Past (Archaeopress 2023): 440–445

5 6

Giusto Traina: Tigran the Great and Mithradates Eupator to Plutarch, Lucullus 14.7, the Roman general seems to acknowledge this title in his speech to the troops by 71 BCE. This might be regarded as a re-elaboration post quem, but in another passage of the Life of Lucullus (21.7) Plutarch shows that the Romans refused to use this title in a letter addressed to Tigran, which eventually provoked the anger of the king, who did not recognize Lucullus as autokratōr.

years and started with Mithradates’ intervention in Cappadocia around 102/100 BCE. Otherwise, for Appian (Mithridatic Wars, 17; see also Granius Licinianus 35, fr. 89-90), these wars start around 88 BCE.9 According to the first periodization, possibly depending on Livy, Mithradates is the only actor of the conflict, whereas the second tendency points to a starting point of the real conflict after the alliance with Tigran. This does not necessarily imply a more positive evaluation of the Armenian king by the historians who shared the thirty year-periodization: in fact, the main source of the first part of Appian’s Mithridatic book depends on a source highly favourable to Sulla (possibly his own autobiography), the protagonist of the military operations in Cappadocia.10

On the other hand, the numismatic evidence seems to suggest a later date for Tigran’s assumption of the title of ‘king of kings’: such are the conclusions of the Armenian numismatist Ṙuben Vardanyan, who thinks that Tigran did not claim the title before 61 BCE. According to this revision of the historical context, after the agreement imposed by Pompey on Tigran in 64, the Romans tried to increase the rivalry between Armenian and Parthians (as Cassius Dio 37.6.2 argues).16 The same later date is proposed, on the basis of different evidence, by R. Shayegan, who argues that the Parthians did not really claim Achaemenid origins, and that the literary evidence reflects a Roman point of view.17 One of his main arguments is that the title of ‘king of kings’, stemming from the Achaemenid titulature, was only adopted by Mithradates II, after the conquest of Babylonia, under the influence of the local priests.18 According to Shayegan, the use of this title by non-Iranian kings like Mithradates and Tigran proves that it did not point at an Achaemenid tradition.19 According to this view, the title would be less important in the East than it is usually held. From a different point of view, Luis Ballesteros Pastor observed that in the first century BCE, the title of king of kings was somehow debased, and claimed by different kings at the same time.20

If we turn away from the Roman source tradition to material produced by the kings themselves, a different picture emerges. It is worth considering the ideological elements of the propaganda of Mithradates and Tigran. Both kings used different codes with their subjects and potential allies, according to their identities. This is clearly shown by the complex elaboration of Tigran’s tiara in the portraits of his coins, mixing Iranian and Macedonian symbols.11 We might also be tempted by a recent interpretation of some coins of Tigran, with a tiara allegedly bearing the image of Halley’s comet, which could somehow recall a coin of Mithradates, possibly connected with the passage of the comet announcing his birth in 135 BCE.12 At any rate, the most important element of comparison is the Oriental title of ‘king of kings’, which was recently studied by the late Federicomaria Muccioli in his massive study of Hellenistic royal titulature,13 and attested for both monarchs. For Mithradates, the sole evidence of a title of ‘king of kings’ is an inscription from Nymphaeum.14 If the king bore this title, he seems to have done it in an area limited to the Black Sea, as did his son and successor Pharnaces II.15 This may be detected in a passage of Florus (3.5.1), who names Mithradates omnium [scil. regum] longe maximus, in a context that does not concern all the kings of Asia, but merely the Ponticae gentes. On the other hand, there is evidence for a title of king of kings borne by Tigran. According to the literary sources, Tigran already held this title at the time of his war with Lucullus: according

Do we need to accept entirely a revision of the evidence? In fact, after his occupation of Syria by 83 BCE, or at least after his invasion of Upper Media and Northern Mesopotamia, Tigran -pace Shayegan- was de facto a king of kings, and a proof of this authority is the absence of See Vardanyan 2011. On the problems of Tigran’s coinage see also Duyrat 2012; Shayegan 2011: 241, n. 768. 17 Shayegan 2011; but see the criticism of Dąbrowa 2012. 18 For an earlier occurrence of the title by the Arsacids, see Muccioli 2013: 405. In fact, Gōdarz I, who does not display the title of king of kings on the coins (Shayegan 2011: 225-226; 235-236). This change of title was interpreted as a side effect of the territorial expansion of Tigran, who, at one point, had indeed claimed the title of king of kings, as indeed Mithradates Eupator did (Ballesteros Pastor 2018: 144-149). See, however, Sinisi 2021: 19-21. 19 Del Monte 1997: 153, 250: Shayegan 2011: 316-319. More recently, Rolf Strootman (2020: 125-127), who ignores Muccioli 2013, argued that ‘the use of the title Great King by late Hellenistic rulers such as Mithradates I of Parthia, Mithradates VI of Pontos, Tigranes I [sic] of Armenia, Artavasdes I of Atropatene, and Antiochos I of Kommagene was not the symptom of an ‘Achaemenid revival’ but that this referred first of all to the Seleukid Dynasty.’ Accordingly, Tigran’s claim of the title is merely a side effect of the result of political change and uncertainty (Strootman 2020: 153). 20 Ballesteros Pastor 2016. 16

Mazzarino 1966: II.2, 325-326. Mastrocinque 1999. On Sulla’s autobiography see Smith 2013. See recently Traina 2021: 146-149. 11 Invernizzi 1998: ix-xviii. 12 Ramsey 1999, on Mithradates’ comet. Gurzadyan and Vardanyan 2004 dated Tigran’s coins to 87 BCE, the year of Halley Comet’s passage by the Earth. But see the skepticism of Panaino 2018. 13 Muccioli 2013: 52-61, 395-417. 14 Vinogradov et al. = SEG 38.668; see Muccioli 2013: 410. Before the finding of this inscription, the common opinion followed Reinach 1890: 251-252. 15 Ballesteros Pastor 2017. 9

10

441

Systemizing the Past this title in the titulature of the Parthian kings.21 Muccioli argues that his use of the title reflects ‘il tentativo di fondere, anche nelle elaborazioni genealogiche, il ramo iranico e quello greco-macedone’.22 In a way, the title still retained a charismatic relevance. Its adoption may have been suggested by the circle of his philoi, which, in the Hellenistic courts, could influence the ideological moves of the king, such as the adoption of a particular epithete.23 Tigran’s court was frequented by Greek intellectuals such as Amphikrates, a protégé of his wife Kleopatra, or Metrodoros of Skepsis, whose Greek work On Tigranes can be detected at least as an indirect source of Pompeius Trogus (through Strabo’s lost Historical commentaries?), at least until 70/69, when he was put to death.24 Metrodoros, whose nickname expressed his hatred of the Romans,25 was one of those Greek historians ‘humouring barbarian kings who detested Rome’s supremacy, — princes to whom they were ever servilely devoted and with whom they associated as flatterers, — by presenting them with histories which were neither just not true’ (Dionysius of Halicarnassus 1.4.3);26 unfortunately, we have only a very slight idea of the content of his work on Tigran.27 These intellectuals would have rubbed shoulders with the Armenian bards charged to sing the king’s chanson de geste (maybe not in the case of Amphikrates, however, who, according to Plutarch, Lucullus 22.6-8, committed suicide when Tigran forbad him to contact other Greeks).28

Moreover, although later sources accused Tigran of oppressing the Greeks,29 the king was not responsible for massacres of Roman citizens as was the case with Mithradates. The marginalization of Tigran can be detected rather early, as is shown in Sallust’s Epistula Mithridatis (Histories 4.69), where Mithradates allegedly asks for help from the Parthian king Frahat III, before the battle of Tigranocerta. In this letter, Mithradates criticizes the ambiguous attitude of the Armenian king, who asked him too late to refresh their former alliance, and finally exonerates ‘his own conduct at the expense of Tigran’.30 In the structure of Trogus’ Historiae Philippicae (Justin 38.3-40.1), the kingdom of Greater Armenia is not treated in a separate book, but embedded in the books on the Parthians and the later Seleucids (significantly enough, there is no trace of Tigran and Armenia in book 38, on Mithradates). Similarly, the structure of Appian’s Roman History presents a succession of the Roman wars with shifts from the Syriaka to the Mithridatika. As far as we may argue from Livy’s Periocha 97 (see also Eutropius, Breviarium 6.8; Orosius, Histories against the Pagans 6.3.5), Tigran enters the scene only when Lucullus forces Mithradates to flee to Armenia. Plutarch (Pompey 45.2) and Appian (Mithridatic Wars 116-117) give Armenia a secondary role in the list of kings and peoples defeated by Pompey.31 Possibly, Tigran was marginalized very early. After all, when Lucullus forced him to withdraw from Asia Minor and Syria, not only his imperial dream, but also the memory of his empire vanished.32

This may be only partly hinted at in the literary evidence, which mostly highlights the role of Mithradates. For instance, Cicero (Academica Priora 2.1.3) praises Lucullus for his victory over Mithradates, ille rex post Alexandrum maxumus. This passage, written in 45 BCE, contains an implicit criticism of Pompey’s memory: shortly after, Cicero attributes to Lucullus the reordering of the Roman East (ut hodie stet Asia Luculli institutis servandis et quasi vestigiis persequendis). In any case, both Lucullus and Pompey preferred to highlight their victory on Mithradates Eupator.

This may also explain a passage in Strabo’s Prolegomena (1.2.34) on the similarity of the Armenians, at least the frontier Armenians, to Syrians and Arabs. This claim is a part of the critical digression on the ethnos of the Erembians, whom Homeric exegesis considered as the ancestors of the Arabs (Odyssey 4.84).33 This does not necessarily mean that Strabo was trying to deprive the Armenians of an Iranian identity: in fact, the geographer (11.13.9) duly considers the ethnographical and religious connections of Armenians with the Iranian peoples. However, in this context he seems to display some bias against an Armenian identity, also quoting the authority of Posidonius, fr. 280 Edelstein-Kidd, who did not mean the Armenians, but the Arameans34).

Muccioli 2013: 408. Muccioli 2013: 136, and 142-143, for a discussion on the cult of Zeus Stratios. 23 Literature in Muccioli 2013: 16, n. 19. 24 Mazzarino 1966: II.1, 486-488; Salomone 1973. See also Ferrary 1988: 228-229 and Fromentin 1998: xxix-xxx. 25 Pliny, Natural History. 34.34 = FGrHist T 6a (cui cognomen a Romani nominis odio inditum est). Pliny does not mention this nickname, but it is very likely Misoromaeus, in opposition to the Greek epithet Philorhōmaios, held by the so-called client kings: see Mazzarino 1966, II.1, 208-209. See also Pédech 1991: 67, ‘on s’est demandé si son Histoire de Tigrane était un véritable récit historique ou un éloge à la manière de l’Agésilas de Xénophon. Le sujet du seul fragment qui en reste fait pencher vers la première alternative ; la référence à un livre premier le confirme : un éloge ne comportait pas plusieurs livres’. 26 See also the famous levissimi ex Graecis in Livy 9.18.6 (a status quaestionis in Muccioli 2007; see also Muccioli 2018: 293-294, n. 79). 27 See Marastoni 2007; Pédech 1970: 67; Traina 2016b. 28 Ferrary 1988: 481; Savalli 1998: 199. On Tigran’s Armenian epic see Russell 1999. 21 22

In any case, possibly after Tigran’s submission to Pompey (and in any case after Crassus’ defeat at Carrhae), the Roman sources focus on the Mediterranean theatre of operations and usually prefer to speak of a kingdom of ‘Syria’ rather than of a Seleucid empire.35 Such images See Breglia Pulci Doria 1977. See Adler 2006 2011: 17-98. 31 See Östenberg 2009: 147. 32 This is highlighted by Asdourian 1911: 51. On the memory of Tigran’s empire, see Traina 2004. 33 See Traina 2017b: 96. 34 See Traina 2001: 144. 35 Muccioli 2004: 144-145, who rightly supposes that this transformation is a sequel of the peace of Apamea; see now Coloru 29 30

442

Giusto Traina: Tigran the Great and Mithradates Eupator universale e la successione degli imperi nell’antichità: 139–170. Roma, ‘L’Erma’ di Bretschneider. Breglia Pulci Doria, L. 1977. Plutarco e Tigran II φιλέλλην. Annali Facoltà Lettere Napoli 16: 37–67. Breglia Pulci Doria, L. 1979. Tigrane il Grande d’Armenia in Mosé di Corene. Tradizione classica e tradizione locale. Dialoghi di Archeologia n.s.1: 95–108. Buongiorno, P. 2021. The Roman Senate and Armenia (190 BC-AD 68). Electrum 98: 89–104. Carcopino, J. 1963. Avant-propos. In H. Manandian 1963. Tigrane II et Rome. Nouveaux éclaircissements à la lumière des sources originales: v–x. Lisboa, Imprensa Nacional. Coloru, O. 2021. Les Séleucides : rois d’Asie, rois de Syrie. Évolution d’un concept géopolitique dans les dernières phases de l’histoire sélucide. In C. Feyel and L. Graslin-Thomé (eds), Les derniers Séleucides et leur territoire: 39–52. Paris, de Boccard. Dąbrowa, E. 2012. Review of Shayegan 2011. Sehepunkte 12, 4, viewed 2 July 2021, . Dąbrowa, E. 2021. Parthian-Armenian relations from the 2nd century BCE to the second half of the 1st century CE. Electrum 28: 41–57. Del Monte, G. 1997. Testi dalla Babilonia ellenistica. Volume I. Testi cronografici. Pisa and Roma, Istituti editoriali e poligrafici internazionali. Derman, J. 2012. Max Weber in Political and Social Thought. From Charisma to Canonization, Cambridge, Cambridge University Press. Duyrat, F. 2012. Tigrane en Syrie : un prince sans images, Cahiers des études anciennes 49: 167–209. Ferrary, J.-L. 1988. Philhellénisme et impérialisme. Aspects idéologiques de la conquête romaine du monde hellénistique, de la seconde guerre de Macédoine à la guerre contre Mithridate. Roma, École française de Rome. Forni, G. and Angeli Bertinelli, M. G. 1982. Pompeo Trogo come fonte di storia. In H. Temporini (ed.), Aufstieg und Niedergang der römischen Welt. Teil II, Principat. Band II. 30.1, Sprache und Literatur (Literatur der augusteischen Zeit: Allgemeines, einzelne Autoren): 1298–1362. Berlin and New York, De Gruyter. Fromentin, V. 1998. Denys : l’homme et l’œuvre. In Ead. (ed.), Denys d’Halicarnasse. Antiquités romaines. Tome I. Introduction générale et livre I: ix–xcix. Paris, Les Belles Lettres. Geller, M. and Traina, G. 2013. ‘Tigranu, the crown prince of Armenia’: evidence from the Babylonian astronomical diaries. Klio 95: 447–454. Gotter, U. 2008. Die Nemesis des Allgemein-Gültigen. Max Webers Charisma-Konzept und die antiken Monarchien. In P. Rychterová, S. Seit, and R. Veit (eds), Das Charisma. Funktionen und symbolische Repräsantationen: 173–186. Berlin, Akademie Verlag. von Gutschmid, A. 1894. Die beiden ersten Bücher des Pompeius Trogus. In A. von Gutschmid (ed.), Kleine Schriften V: 9–217. Leipzig, Teubner.

are the result of the development of the Eastern policy between the heyday of Roman imperialism and the pax Augusta: the Roman generals began to consider themselves as the true successors of Alexander, whereas the Hellenistic monarchies were regarded as degenerate, and so unfit to inherit the legacy of the Macedonian power.36 Did the title of ‘king of kings’ play a role in the comparison of Mithradates and Tigran? At least a passage of Velleius Paterculus (2.33.1) seems to hint at a Roman debate on this issue. Velleius introduces Tigran as ‘the most important of the kings’ (regum maximum) in his narration of Roman history. This expression is less than trivial, as the Romans were very sensitive to the honorific meaning of the superlative maximus.37 On the other hand, Velleius seems to be the only classical author giving Tigran a higher status than his ally and father-in-law.38 Bibliography Adler, E. 2006. Who’s anti-Roman? Sallust and Pompeius Trogus on Mithridates. Classical Journal 101: 383–407. Adler, E. 2011. Valorizing the Barbarians: Enemy Speeches in Roman Historiography. Austin, University of Texas Press. Asdourian, P. 1911. Die politischen Beziehungen zwischen Armenien und Rom von 190 v. Chr. Bis 428 n. Chr. Venezia, Mechitaristenbuchdruckerei auf San Lazzaro. Ballesteros Pastor, L. 2013. Pompeyo Trogo, Justino y Mitrídates. Comentario al ‘Epítome de las Historias Filípicas’ (37, 1 ,6-38, 8, 1). Hildesheim, Zürich, and New York, Olms. Ballesteros Pastor, L. 2016. Mitrídates Eupátor, un enemigo de Roma en el Epítome de Justino. In A. Galimberti and G. Zecchini (eds), Studi sull’Epitome di Giustino. III. ll tardo ellenismo. I Parti e i Romani: 63–97. Milano, Vita e Pensiero. Ballesteros Pastor, L. 2017. Pharnaces II and his title ‘King of Kings.’ Ancient West and East 16: 297–303. Ballesteros Pastor, L. 2018. De Rey del Ponto a Rey de Reyes. El imperio de Mitrídates Eupátor en el contexto del Oriente tardo-helenístico. In L. R. Cresci and F. Gazzano (eds), De imperiis. L’idea di impero 2021. 36 See, e.g., L. Manlius Vulso’s speech to the troops before a battle against the Galatians in 187 BCE in Livy 38.17.11: Macedones, qui Alexandriam in Aegypto, Seleuciam ac Babyloniam, vel sparsas per orbem terrarum colonias habent, in Syros, Parthos Aegyptiosque degenerarunt ‘the Macedonians who hold Alexandria in Egypt, who hold Seleucia and Babylonia and other colonies scattered throughout the world, have degenerated into Syrians, Parthians, Egyptians’. See also Traina 20112012: 121. 37 In his discussion of Pompey’s appellative of Magnus, Plutarch (Pompey 13.7-11) highlights this exceptional value. See Muccioli 2013: 52. 38 Breglia Pulci Doria 1979: 101 thought that the only exception to the ‘rapporto Mitridate-Tigrane quale è a noi consueto’ can be only found in the local tradition in Movsēs Xorenac‘i.

443

Systemizing the Past Gurzadyan, V. G. and Vardanyan, R. 2004. Halley’s Comet of 87 BC in the coins of Armenian King Tigranes? Astronomy & Geophysics 45: 4–6. Habinek, T., 2012, Metrodoros (184). In Brill’s New Jacoby, General Editor: Ian Worthington (Macquarie University). Consulted online on 02 July 2021

Invernizzi, A. 1998. Regalità armena. Introduzione all’Armenia ellenistica. A. Invernizzi (ed.), Ai piedi dell’Ararat. Artaxata e l’Armenia ellenistico-romana: xix–xxix. Firenze, Le Lettere. Jacoby, F. 1958. Die Fragmente der griechischen Historiker (FGrHist). Dritter Teil. Geschichte von Städten und Völkern (Horographie und Ethnographie). C. Autoren über einzelne Länder. Nr. 608a-856 (Erster Band: Aegypten-Geten Nr. 608a-708). Leiden, Brill. Jacoby, F. 1962, Die Fragmente der griechischen Historiker (FGrHist). Zweiter Teil. Zeitgeschichte. B. Spezialgeschichten, Autobiographien und Memoiren, Zeittafeln. Zweite und vierte Lieferung. Kommentar zu Nr. 106-261. Leiden, Brill. Khachikyan, A. 2006. Hannibal in the East and in Armenia. Patma-banasirakan handes (HistoricalPhilological Journal) 1: 285–296 (in Russian). Linke, J. 2015. Das Charisma der Könige. Zur Konzeption des altorientalischen Königtums im Hinblick auf Urartu. Wiesbaden, Harrassowitz. Manandian, H. 1963. Tigrane II et Rome. Nouveaux éclaircissements à la lumière des sources originales. Lisboa, Imprensa Nacional. Manaseryan, R. 1987. Tigran the Great. Armenia’s Struggle against Rome and Parthia. Yerevan, Hayastan (in Armenian). Marastoni, S. 2007. Metrodoro di Scepsi. Retore, filosofo, storico e mago. Alessandria, Edizioni dell’Orso. Mastrocinque, A. 1999. Studi sulle guerre mitridatiche. Wiesbaden, Franz Steiner. Mazzarino, S. 1966. Il pensiero storico classico, II.1-2. Bari, Laterza. Muccioli, F. 2007. La rappresentazione dei Parti nelle fonti tra II e I secolo a.C. e la polemica di Livio contro i levissimi ex Graecis. In F. Muccioli and T. Gnoli (eds), Incontri tra culture nell’Oriente ellenistico e romano: 87– 115. Milano, Mimesis. Muccioli, F. 2013. Gli epiteti ufficiali dei re ellenistici. Wiesbaden, Franz Steiner. Muccioli, F. 2018. The Ambivalent Model: Alexander in the Greek World between politics and literature (1st Century bc / beg. 1st Century ad). In K. R. Moore (ed.), Brill’s Companion to the Reception of Alexander the Great: 275–303. Leiden and Boston, Brill. Östenberg, I. 2009. Staging the World. Spoils, Captives, and Representations in the Roman Triumphal Processions. Oxford, Oxford University Press. Panaino, A. 2018. Astral omina and their ambiguity: The case of Mithridates’ comets. Iran and the Caucasus 22: 232–256.

Pédech, P. 1991. Deux Grecs face à Rome au Ier siècle av. J.-C. : Métrodore de Scepsis et Théophane de Mytilène. Revue des études anciennes 93: 65–78. Ramsey, J. T. 1999. Mithridates, the Banner of Ch’ihYu, and the Comet Coin. Harvard Studies in Classical Philology 99: 197–253. Reinach, T. 1890. Mithridate Eupator, roi de Pont. Paris, Firmin-Didot et c°. Russell, J. R. 2004. The lost epic of Tigran: A reconstruction based upon the fragments. In J. Russell, (ed.), Armenian and Iranian Studies: 1031– 1050. Cambridge, MA, Harvard University Press. Salomone, E. 1973. Fonti e valore storico di Pompeo Trogo (Iustin., XXXVIII 8, 2-XL). Genova, Istituto di Storia antica e scienze ausiliarie. Savalli Lestrade, I. 1998. Les philoi royaux dans l’Asie hellénistique. Genève, Droz. Sciandra, R. 2008. Il ‘Re dei Re’ e il ‘Satrapo dei Satrapi’: note sulla successione tra Mitradate II e Gotarze I a Babilonia (ca. 94–80 a.e.v.). In B. Virgilio (ed.), Studi Ellenistici 20: 471–488. Pisa and Roma, Fabrizio Serra. Shayegan, R. 2011. Arsacids and Sasanians. Political Ideology in Post-Hellenistic and Late Antique Persia. Cambridge, Cambridge University Press. Sherwin-White, A. N. 1992. Lucullus, Pompey, and the East. In J. A. Crook, A. Lintott, and E. Rawson (eds), The Cambridge Ancient History, Volume 9: The Last Age of the Roman Republic, 146–43 BC, 2nd edition: 229–273. Cambridge, Cambridge University Press. Sinisi, F. 2012. Sylloge Nummorum Parthicorum. Volume VII. Vologases I – Pacorus II. New York, Paris, London, Vienna, Tehran and Berlin, Austrian Academy of Sciences Press. Smith, C. J. 2013. L. Cornelius Sulla. In T. J. Cornell (ed.), The Fragments of the Roman Historians, 1: 282–286; 2: 472–491; 3: 289–299. Oxford, Oxford University Press. Sommer, M. 2011. Empire of glory. Weberian paradigms and the complexities of authority in imperial Rome. Max Weber Studies 11: 155–191. Strootman, R., 2020. The Great Kings of Asia: Imperial titles in the Seleukid and post-Seleukid Middle East. In R. Oetjen (ed.), New Perspectives in Seleucid History, Archaeology and Numismatics. Studies in Honor of Getzel M. Cohen: 123–157. Berlin, De Gruyter. Traina, G. 1999-2000. Épisodes de la rencontre avec Rome (IIe siècle av. J.-C.-IIIe siècle ap. J.-C.). Iran and the Caucasus 3: 59–78. Traina, G. 2001. Strabone e le città dell’Armenia. I G. Traina (ed.), Studi sul XI libro dei Geographikà di Strabone: 141–154. Galatina, Congedo. Traina, G. 2004. Mythes fondateurs et lieux de mémoire de l’Arménie pré-chrétienne (i). Iran and the Caucasus 8: 169–181. Traina, G. 2010, Teatro greco nell’Armenia antica. In E. Migliario, L. Troiani and G. Zecchini (eds), Società indigene e cultura grecoromana: 95–103. Roma, ‘L’Erma’ di Bretschneider. 444

Giusto Traina: Tigran the Great and Mithradates Eupator Traina, G. 2011-2012. Letteratura classica e spazio geografico partico: alcune osservazioni. Geographia antiqua 20-21: 119–122. Traina, G. 2012. Tigran il Grande d’Armenia e la Giudea. In G. Urso (ed.), Iudaea socia - Iudaea capta: 79–88. Pisa, ETS. Traina, G. 2016a. Traditions on Armenia in Submerged Greek Literature: Preliminary considerations. In G. Colesanti and L. Lulli (eds), Submerged Literature in Ancient Greek Culture. Vol. 2. Case Studies: 111–123. Berlin and New York, De Gruyter. Traina, G. 2016b. L’impero di Tigran d’Armenia nella versione di Trogo-Giustino. In A. Galimberti and G. Zecchini (eds), Studi sull’Epitome di Giustino. III. ll tardo ellenismo. I Parti e i Romani: 99–116. Milano, Vita e Pensiero. Traina, G. 2017a. Rois ou dynastes ? Les territoires arméniens à l’époque d’Antiochos III. In C. Feyel

and L. Graslin-Thomé (eds), Antiochos III et l’Orient: 377–388. Paris, de Boccard. Traina, G. 2017b. Strabo and the history of Armenia. In D. Dueck (ed.), The Routledge Companion to Strabo: 93–101. London and New York, Routledge. Traina, G. 2021. Les Romains et les Parthes avant Carrhes. In M. T. Schettino, S. Pittia, and G. Zecchini (eds), Héritages de Sylla: 147–162. Roma and Bristol (CT), ‘L’Erma’ di Bretschneider. Vardanjan, R. 2011. The struggle on the title ‘king of kings’ in the context of Roman Eastern policy of the I century BCE. Patma-banasirakan handes (HistoricalPhilological Journal) 1: 230–251 (in Russian). Vinogradov, Ju. G., Molev, E. A. and Tolstikov, V. P. 1985. New epigraphical sources for the history of the age of Mithradates. In O. Lordkipanidze (ed.), The Black Sea Litorial in the Hellenistic Age: 589–600. Tbilisi, Mecnieriba (in Russian).

445

Modelling of Bronze and Iron Age Monuments at the Northwestern Slopes of Mount Aragats based on a Case Study of Lernakert* Benik Vardanyan and Levon Mkrtchyan

Institute of Archaeology and Ethnography, NAS RA, Yerevan, Armenia Abstract: Archaeological investigations in Lernakert (Shirak province, Armenia) were carried out in 2019-2021. The site consists of two cyclopean fortresses, remains of settlements, necropolises, and megalithic monuments situated at the northwestern slopes of the mount Aragats. Most of the monuments date to the Bronze and Iron Age. New approaches, such as remote sensing through digitalization, aerial photography, and 3D-modelling and mapping are an integral part of our research. These methods, along with archaeological excavations (and sometimes even without) reveal architectural patterns of fortresses, the location and regularity of necropolises, the phases of their developments, the outlines of ancient roads and pathways, as well as define monuments’ boundaries. Based on aerial photographs and data modelling, a database was created, which serves as a methodological example for further investigation of the slopes of mount Aragats. Keywords: Armenia, Shirak, Lernakert, remote sensing, aerial photography, digitalization, 3D-modelling, mapping, database

Introduction Modern-day devices and technology play a key role in investigations of archaeological monuments. It is commonly known that archaeological excavations result in the stratigraphy and architecture of the monument. It is no accident that excavations are usually described as unrepeatable experiment. Currently, the need for non-destructive methods aiming to avoid any physical disturbance of the archaeological and historical monuments is more important than ever before. These methods enable the conservation and inheritance of the monument for future generations. In this regard, remote sensing is of particular importance for meeting the mentioned requirements. The current contribution aims to illustrate the successful application of remote sensing based on data from drone photography. The case study of Lernakert will exemplify the possibilities and advantages of the mentioned methods.

The elevation maps were constructed for each monument or section separately accordingly with scale and area outline. The maps also give information of the coordinate system (each unit = 1 m) and forms in section with necessary coordinates (latitude/longitude/ height). 2D and 3D maps of the archaeological site were created on the base of above-mentioned data. From 2019 to 2021, archaeological monuments of Lernakert were extensively investigated by our team. Six campaigns resulted in documenting and examining two prehistoric fortresses (Veri Berd and Vari Berd), settlements with necropolises around them, structures of megalithic character and other types of monuments. The results were registered in our database, which became a solid methodological and informative basis for investigation of archaeological sites in northwestern slopes of the mount Aragats and generally for remote sensing of other Bronze and Iron Age sites of Armenia.

Site location and research

Digitalization

Since 2019, in conjunction to archaeological field work, remote sensing has been used regularly in Lernakert. In general, a relatively wide area of the site was documented through drone photography: particularly two fortresses (Veri Berd and Vari Berd), six necropolises, stones put in a line (the purpose of which is unclear today), roads, etc. Through fixed photos, we created orthophotograph  and elevation maps. Based on these components, digital and 3D models of the site were constructed.

The aerial photos were obtained using an Inspire 1v2 drone. Accordingly, polygons with 3D settings were prepared beforehand for each section. Depending on the relief of a given area, aerial photos were taken at different heights. For instance, the fortresses were documented at a height of 50-70 m, whereas the necropolises at from 30-50 m (Figures 2-9). Approximately 50-70 fixed photos served as the main sources for the digitalization of the monuments.

The research was funded by the State Committee for Science of the Ministry of Education and Science of the Republic of Armenia under the scientific topic code 20TTSH-009.

* 

Systemizing the Past (Archaeopress 2023): 446–453

Data processing started right after field work and aerial photography. With the help of fixed photos, we created orthophotos already with a coordinate basis (Geo TIF) and a digital model of the surface (DSM) based on numerous coordinate and height units. Besides

Benik Vardanyan and Levon Mkrtchyan: Modelling of Bronze and Iron Age Monuments

Figure 1. The northwestern Slopes of Mount Aragats and the location of Lernakert.

Figure 2. Orthophoto plan of Vari Berd.

447

Systemizing the Past

Figure 3. Elevations, outlines, and section of Vari Berd.

the fixed photos, auxiliary contribution was gained through GIS, which was used throughout the entire process. Using orthophotos and the DSM, we were able to create a geodesic map of the Lernakert monuments with horizontal lines of 1 m height coordinate system and relevant scale, as well as the section of monuments, mostly with east-west orientation.

which can be observed in architecture and its details, as well as in the principles of space organization and landscape exploitation. Fortresses The fortresses of Lernakert are located on the slopes of mount Aragats, separate from the mountain ranges stretching from east to west, but in close proximity to them (Figure 1). The fortified hills represent continuations of mountain ranges, which operate as natural protection for the fortresses. In case of both fortresses the fortification, walls are built into the natural rock on the west. Thus, the fortified walls of Vari Berd and Veri Berd were adapted to the edges of the hill, whereas the slopes (northeast slopes of Vari Berd, northeast slopes of Veri Berd) were turned into artificial cascade terraces. (Figures 2- 4). The slopes of both fortified hills were used also as settlements.

3D models of all digitalized monuments and cultural landscapes are of particular importance for documentation, further reconsideration and computer reconstruction of those monuments. Currently, Lernakert is considered as one of the most densely and comprehensively digitalized areas of Armenia, which consists of almost all known monument types of the Bronze and Iron Ages. In the near future, after digitalizing some of newly explored parts, the 3D map of the entire cultural landscape of Lernakert will be ready. Along with archaeological excavations and investigations, it will contribute to our understanding about the emerging, development and final phase of social-cultural practices in Lernakert in regard to different archaeological layers.

Judging by the surface pottery finds Vari Berd was inhabited in several periods, the earliest of which dates back to the 3rd millennium BC. Remains of Early Bronze Age structures have been found along the northern foothills. The Early Bronze Age settlement extends northward, reaching the southern and southwestern outskirts of the modern village of Lernakert. The next stage of settlement of the area belongs to the Late Bronze Age, evidenced by the Late Bronze Age burials within the Early Bronze Age settlement on the western slope.

Modelling of the monuments of Lernakert Two fortresses among Bronze and Iron Age monuments of Lernakert, have outstanding role. They are located on natural hills and occupy a territory of artificially created platforms. Both fortresses show similarities, 448

Benik Vardanyan and Levon Mkrtchyan: Modelling of Bronze and Iron Age Monuments

Figure 4. Orthophoto plan of Veri Berd.

Thanks to aerial photography and three-dimensional modeling it was possible to approximate the boundaries of the Early Bronze Age settlement, to trace the forms of the still visible above-ground architectural structures and document their transformations (Figures 2-3).

attacking force. Stone lined road begins from the southwestern slope of the fortress with parallel walls, which in some places has been interrupted by recent land amelioration. This road, according to the aerial photos, was supposed to connect the network of roads leading to the tombs (Figures 4-5).

The eastern slope of Vari Berd was also inhabited. In this section, remains of various constructions were observed, the most prominent of which is a wheelshaped structure located east of the third platform. Its documentation was made possible only by using the above-described methods (Figure 2).

Necropolises Extensive necropolises dating to Bronze and Iron Ages, surround the two fortresses of Lernakert. Considering chronological differences, construction similarities, and segregation of certain burials by walls, six distinct cluster groups can be identified at the necropolis: I-VI.1

The Veri Berd settlement is located on the north-eastern slopes of the fortress, where maximum protection appears to have been guaranteed; it is no coincidence that the northern entrance to the fortress passes through the settlement. Remains of various structures are visible in the areal images. The tombs are not visible on the surface in the lower settlement area (Figure 4).

The first group is situated on the eastern side from Vari Berd fortress. It is on the upper side of the hill, where the slopes have been shaped by two seasonal streams (Figures 6-7). Its ravines are bonded by a row of vertical (orthostatic) small and middle-sized rough stones. The tombs in this necropolis are built on the natural rock. The burials extend for 7-15 m in diameter (cromlech) and 0.5-1.5 in height (mound). In general, the slab-stone tomb chambers are constructed on the north-south axis. However, three tombs on the slopes of the western mound have both east-west and northeast-southwest orientations which include half pit and half slab-made chambers. The slab-stone tombs on the eastern side

In southwest, the fortified hill gradually merges with the plain (the fortress is at 2153 m asl, the south-west slope is at 2145 m asl, east-north slope at 2099 m asl, 2117 m), so it was probably considered as the most vulnerable part of the defense system. Walls built parallel to the fortified wall, irregular structures formed by them, pits create a network of adjacent complexes, which resemble a system of barriers, the purpose of which, in our opinion, was to break the momentum of potential

1

449

Vardanyan et al. 2021.

Systemizing the Past

Figure 5. Elevations, outlines, and section of Veri Berd.

Figure 6. Orthophoto plan of tomb group no. 1 of Veri Berd.

450

Benik Vardanyan and Levon Mkrtchyan: Modelling of Bronze and Iron Age Monuments

Figure 7. Elevations, outlines, and section of tomb group no. 1 of Veri Berd.

of the necropolis are oriented from east to west. The differences between the orientation of the chambers are caused by chronological reasons.

eastern edge of the first circle of the ring kurgan. The first circle connects to the second one by its outer medial row. In its layout, it has the shape of a multipetal flower (Figure 8). Massive slabs cover the entrance of the chamber. In some places, the southern part of the chamber is built in two rows. In the same manner, in the southern part of the surrounding ring a two-row wall composition can be observed. The covering slab has been moved from the northwestern side to south, and, most probably, looters entered from this side.

Veri Berd’s second necropolis is located on the mound of the fortress, on the slopes of the northern ledge. In this group, the tomb chamber is a simple earthen pit surrounded by stone circle. Combined, they form a dense network of tombs and date back to early stages of the Late Bronze Age.2 The third tomb group has the highest elevation among tombs documented thus far. It is located at 2182 m above sea level. The group is situated on the so-called ‘Kaskei Č‘ayir’ site, on the top of a high terraced rock. A line of 2-3 vertical stone lines (walls) outlines the plane surface of the hill’s top from the North and the South. It continues for 700-800 m. Among the stone routes tombs with 8-15 m diameter and high earth embankment can be observed. The mentioned rows of standing stones on the western side of the ravine join a building, which proves this tomb to have been a complex architecture monument (Figures 8-9).

At the southern edge of the hill, there is a construction with rounded corners. The construction has a rectangular layout with its long axis oriented to the east-west. The structure is preserved to a height of 2-3 block-rows (1.2 - 1.5 m). The southern side of the building merges with the natural rock. The other three walls of the building repeat the structure of the tomb’s base-wall and the similar construction setup. The latter plays a significant role in the matter of the chronology and probably, the function.

The slab-stone chamber of the burial mound is bonded by an oval stone circle (21 m North-South, 16 m EastWest). It joins the fence passing through North-South, the height of which reaches 1.2 - 1.6 m. This is the

As a result of a three-year study of the Bronze and Iron Age monuments of Lernakert issues of dating were clarified. Thus, the earliest stage of settlement in the area belongs to the Early Bronze Age (29th-26th centuries BC), which is evidenced by the materials

2

Dating

Vardanyan et al. 2021: 199.

451

Systemizing the Past

Figure 8.

Figure 9.

452

Benik Vardanyan and Levon Mkrtchyan: Modelling of Bronze and Iron Age Monuments The last stage of the Veri Berd settlement is represented by the cultural layer of the post-Urartian period, which in certain places has a thickness of 1.80 meters.

found on the slopes of Vari Berd and in the southern and southwestern edge of the modern village of Lernakert. The final stage of the Middle Bronze Age is represented in the Group I and II tombs of Veri Berd.

Thus, the megalithic landscape of Lernakert is formed in the Early Bronze Age, but reaches its peak of development in the Late Bronze Age and partly in the Iron Age.

The Late Bronze Age is represented in both fortresses and tombs. The construction of the cyclopean fortress of Veri Berd (or the settlement in the area of the fortress) dates back to the 14th-13th centuries BC, whereas the main cultural layer of the fortress dates back to the 5th/4th-3rd centuries BC.3

Bibliography

The research conducted from 2019-2021 in the Veri Berd necropolis revealed 19 tombs, 12 of which belong to different sub-types of the Late Bronze Age. In Vari Berd Late Bronze Age is represented by a mound on the western slope of the fortress and as a mound pottery in the territory of the settlement.

Vardanyan, B., Davtyan, R., Manukyan, S. and Saribekyan, M. 2021. Lernakert: ein neues archäologisches Projekt in Shirak. In P. Avetisyan and A. Bobokhyan (eds), Archaeology of Armenia in Regional Context, Proceedings of the International Conference dedicated to the 60th Anniversary of the Institute of Archaeology and Ethnography, Held on July 9-11, 2019 in Yerevan: 189–200. Yerevan, Institute of Archaeology and Ethnography Press.

The next stage of settlement refers to the Iron Age, which is represented in the lower layers of the fortress and is ubiquitous in Veri Berd. Iron Age complexes were also recorded in the first cluster of the Veri Berd necropolis.

Zarikyan, N., Vardanyan, B., Davtyan, R., Manukyan, S. and Saribekyan, M. 2020. A Preliminary Report on the Faunal Remains from Lernakert (2019 excavations) (Shirak, Armenia). Scientific Works 23/2: 58–73 (in Armenian).

3

Zarikyan et al. 2020: 61.

453

The Early Medieval Complex of Agarak Nora Yengibaryan and Lilit Ter-Minasyan

Institute of Archaeology and Ethnography, NAS RA, Yerevan, Armenia Abstract: The article presents the early medieval complex of the Agarak multilayered archaeological site. The original appearance of the structure, as well as the reconstructions carried out due to the change of the function of the complex over time. The remains of material culture discovered in the complex are discussed in the context of the findings of contemporary sites. Keywords: Agarak, Middle Ages, early medieval complex, columned hall, Annex, glkhatun, fire temple, ‘Midis’ laying, base, 3D restoration

Introduction The Agarak archaeological site is located in the Aragatsotn region of the Republic of Armenia, on the south-eastern slopes of Mount Aragats. Only the rocky platform of the site, which spreads between Agarak and Voskehat villages, has been excavated.1 The YerevanGyumri (formerly Leninakan) highway, built in 1978, divided the rocky platform into two parts, destroying a significant part. Initially, excavations were carried out only on the left side of the road, which established that the settlement had seven stages of habitation starting from 29th century BC to the Late Middle Ages.2 Rock-cut pits, Early Bronze Age structures, retaining walls, rockcut winepresses, and other structures were discovered. The excavations also uncovered an Urartian rock-cut tomb, as well as classical and early medieval tombs. The section to the right of the rock platform, discussed in detail in this article, was excavated in connection with the construction of the ‘North-South’ road corridor.3 Historical Overview The Aragatsotn province of Ayrarat region of Greater Armenia spread on the southern slopes of Mount Aragats and in histroical literature is found under the names Votn Aragatsoy (Foot of Aragats), Aragatsotan, Koghmn Aragatsa (Aragatsu, Aragatso) (side of Aragats).4 Due to the instability of the country’s political situation, the borders of the province were constantly changing. The province of Aragatsotn was the property of the Arsacid kings, where the royal winter-military bases 1 This part of the site was excavated for six years under the direction of P. Avetisyan (2001-2008, with intermissions in 2003 and 2006). 2 Avetisyan 2003: 52-57. 3 Excavations were carried out in the right part of the rock platform in 2012-2014. The complex was excavated in 2013: Head of the expedition P. Avetisyan, archaeologist N. Yengibaryan, arcitect L. Ter-Minasyan. The photos used in the article were made by V. Hakobyan (Figures 15-18), B. Gasparyan (Figure 34), I. Kalantaryan (Figures 35-42), L. Ter-Minasyan (Figure 43). Drawings of artifacts were done by N. Yengibaryan (Figures 1-13), dimensions and 3D restorations are by L. Ter-Minasyan (Figures 14-33). 4 The province of Aragatsotn included the regions of Ashtarak and Talin and some parts of Yeghvard, Vagharshapat, and Armavir.

Systemizing the Past (Archaeopress 2023): 454–484

were located. For the services provided the Arsacid kings donated domains in Aragatsotn to the ministerial houses of Amatuni (in Oshakan), Mamikonyan (in Aruch), and Gnuni (in Mastara) in the 4th century, which expanded their domains after the fall of the Arsacid dynasty.5 One of the oldest trade routes leading to the Black Sea passed through the territory of Aragatsotn. This is evidenced by the presence of medieval caravanserais (Aruch, Mastara, Dashtadem). Among the defensive structures are Zakari Berd, Ashnak (Ashnak 2 and 3), Kakavadzor, Nerkin Sasnashen, Davtashen, and Bazmaberd fortresses. In Aragatsotn are present all types of famous churches of the 4th-7th centuries: single-nave and three-nave basilicas, central domed and multi-arched churches, including famous monuments of early medieval Armenian architecture (Mastara, Garnahovit, Shenik, Aragats, Avan, Parpi, Tsiranavor of Ashtarak). The sculptured four-sided stelae, dated between 4th-7th centuries are known from almost all settlements of the region.6 The abundance of churches and early medieval sites is indicative of the strong foundations of the Christian religion. In the vicinity of the studied complex is the clearly dated early medieval settlement of Aghdzk7 where in the sixties of the 4th century the remains of the Arsacid kings taken back from the Persians were reburied. The Aghdzk monument consists of a three-nave basilica, a tomb, and a stela ending in a winged cross. Other monuments found in the area of ​​the settlement are water pipes, roof tiles, and ornamented bricks, which indicate that Aghdzk was a well-maintained settlement in the early middle ages, possibly a religious center. One of the first single-nave basilica churches in Armenia is the 4th century St. Astvatsatsin church of Agarak Manucharyan 2017: 35-37, 45. Asatryan 2004: 72-81; Grigoryan 2012: 231-239. 7 In a straight line to the north-west of the complex at a distance of 2.4 km. 5 6

Nora Yengibaryan and Lilit Ter-Minasyan: The Early Medieval Complex of Agarak with its unique planning and spatial solution.8 It is not clear to which ministerial house belonged the village of Agarak. Aghdzk, located in the vicinity of Agarak, belonged to the Mamikonians,9 thus Agarak could have belonged to that ministerial house.

is evidenced by the fact that the position of the walls relative to each other remained unchanged, the right angles of the walls were not disturbed. The southeastern corner of the complex was also destroyed during the construction of the road. The north-eastern part of the building is best preserved, the walls here stand at a height of 1.3-1.5 m.

General Composition of the Complex The Agarak complex is a rectangular structure. The walls of the compex are precisely oriented to the cardinal points. The structure occupies a total area of​​ 428 m2 and has east-west and north-south oriented outer walls (24.410 m and 17.5 m respectively, see Figure 19). The ratio of longitudinal to lateral sides is 1:1.4 m. The complex is divided into two almost equal blocks by a wall placed along the longitudinal axis of the structure. On the east side of the building, there is an annex with a width of 5.2 m, of which the north-east walls are partially preserved (Figure 34). Eastern block has a 10.5 m × 15 m inner size and is divided into two parts by an east-west oriented wall with a ratio of 3:2. To the south of the wall, the larger part is occupied by the columned hall (room no. 1),11 and rooms no. 2 and no. 3 are located to the north. Western block has a 10.2 × 15 m inner size and is divided into two equal parts by an east-west oriented wall. To the north of the wall are the smallest room no. 4 and an almost square room no. 5. To the south of the wall is the second ‘hall’ (room no. 6). The floor marker of the western block is higher by 0.55 m than the floor marker of the eastern block. That is why a double stone step was built in the openings connecting the blocks. The preserved entrance to the building is on the south wall. The main role in organizing the space in the composition of the complex is played by the columned hall. It is the most spacious, richly organized room in the building. The Agarak complex has a persistent layout, where the principles of symmetry and volume balance have been applied. Current state of the building. The building has beed damaged significantly due to later burials and more recent road construction. The western block of the complex is the most damaged. The south-north walls of the block are destoryed almost entierly and the western wall and the partition dividing the block into two parts, deviate from the axes of the structure in the opposite direction of the clock by 40 (Figure 20). The deviation most likely occurred as a result of explosions during road construction. The bedrock has moved and the walls built on it have deviated from the axes. This

Columned hall (Room no. 1) occupies 10.5 m × 8.6 m space (Figure 35). The south-east walls are external. The ratio of the sides of the room is 1:1.2. The maximum preserved height of the walls is 1.5 m. The interior design is symmetrical to the longitudinal axis. In the right center of the south wall there is a 1.2 m wide doorway through which the complex was accessed. On the longitudinal axis of the hall, with a deviation of 0.5 m from the center to the north, there is a tuff fireplace (1.3 m × 1.1 m).12 The stove is made of one-piece stone and has a depth of 0.15 m (Figure 29a). The hall has four tuff bases. The bases are placed so that they form a square in the center of the hall. The distance between them is 4.8 m. A single row of the stone bench (0.74 m × 4.7 m) is preserved parallel to the western wall in the hall. The surface of the rock in that section is leveled to a depth of about 0.1 m. The floor of the room was paved, some stones were preserved. Approximately 0.3 m × 0.7 min size, the stones have irregular sides and are adapted to each other. The southern corners of the columned hall are destroyed. Only the 1.95 m long north edge of the western wall stands. In the central part of the wall, a stone of the outer face has been preserved, due to which the direction of the wall is determined. On the west wall, next to the north wall, there is a 0.82 m wide door opening through which the columned hall communicates with the western block. The central part of the south wall with a length of 7.4 m is preserved in 1-2 rows. The eastern wall is preserved in 1-4 rows. In the central part of the wall, a 2.4 m long section has collapsed (Figure 27a). The north wall is the best-preserved with 3-5 rows (Figures 23a, 37). The 0.92 m wide door connecting the hall with room no. 2 is located at a distance of 1.14 m from the western edge of the wall. The eastern edge of the wall has collapsed. Based on the composition of the columned hall, on the large space occupied by the room (90.3 m2), the presence of a fireplace and a bench, it can be concluded that the room was used for crowded receptions and celebrations. Room no. 2 occupies the north-eastern corner of the complex (5.8 m × 5.1 m, Figure 38). The north-east walls of the room are external. Room no. 2 is connected to room no. 3 by a 0.98 m-wide door on the western wall.

700 m in a straight line from the complex, east. Manucharyan 2017: 42. 10 The dimensions mentioned in the text are in meters, unless there are other explanations. 11 See room numbering in Figure 19.

Similar fireplaces were found in Dvin (1.5 m × 1.5 m × 0.3 m) room no. 4 (1.0 m × 0.95 m) of Zakari Berd fortress and room no. 1 (1.02 m × 1.05 m) of complex no. 14. The fireplaces are located in the central parts of the rooms. See Asatryan 2005: 28-29, 31-32, Figures 16-17; Ghafadaryan and Kalantaryan 2002: 67.

8

12

9

455

Systemizing the Past The western wall has been preserved in 5-6 rows. The north-east corner of the room is the most damaged. The northern wall is preserved in 1-5 rows (Figure 23b), the eastern wall in 1-4 rows. There is a niche on the southern wall of the room (1.0 m × 0.85 m). It is located 1.9 m from the eastern edge of the wall. In the north-western and north-eastern corners of the room, about 0.5 m from the walls, there are two tuff bases. The ash layer in the central part of the room indicates the presence of a hearth.

attached to the north wall of the room. The bench is located along the longitudinal axis of the room, which excludes the presence of a partition in the area, which supports the idea of a ‘hall.’ The western and southern walls were external. The western wall, the western part of the northern wall, and the south-western corner have been preserved, all in a single row. Annex (room no. 7) is also very poorly preserved (Figure 42). The north wall originates from the central part of the east wall of the complex but deviates from the main axis of the walls in a clockwise direction by 80. The building had a width of 6.6 m (external size), but we can only speculate about its longitudinal size, only the 3.6 m long section of the eastern wall has been preserved. If this building was built to preserve the southern façade of the complex, it was 8.6 m in length (Figure 21). The right angle between the walls of the building has been preserved. The width of the walls is the same as in the main structure, the eastern wall is made with ‘fir’ lining. This technique is also used in other parts of the complex.

Room no. 3 is located in the north-west corner of the eastern block (3.7 m × 5.1 m) (Figure 39). The north wall of the room is external. The room was used as a corridor. Through it room no. 2 is connected to the western block of the columned hall. The walls are 1.3-1.4 m high, except for the south-western corner of the room, which is preserved only to 0.5-0.7 m. Lime mortar is partially preserved on the walls. On the west wall, next to the south wall, there is a 1.2 m wide door opening (Figure 24a). In the south-west corner of the room, in the eastwest direction, there is a large stone (1.2 m × 0.7 m × 0.4 m). The plane of the stone front is adapted to the western plain of the southern opening. The stone may have had a religious significance (Figure 29b).

Artifacts. Mainly fragments of pottery were found in the complex. For the purposes of typology and classification, the pottery was compared to contemporanious materials found in Oshakan, Zakari Berd, Aghdzk, as well as in Dvin and Sotk in the same region.

Room no. 4 is located in the eastern corner of the western block. It is the smallest room in the complex (2.6 m × 3.1 m, Figure 40). The north wall of the room is external. The thickness of the south wall is 0.85 m. On the wall next to the east wall there is a 0.77 m wide door opening.

Jars (karas). From these large vessels, pink, brown, black, and brick-colored fragments of rims, body sherds, andbases, all made of sandy clay were found. The jars were used to store dry and liquid food. The distinction between these large vessels was made by the structure of the rim, the body, and the base. The bodies of the jars are lined with a dotted ornament made on wet clay (Figure 4/1, 6). The bases are flat (Figure 5/4) or convex (Figure 5/5), the bodies extend from the bottom or widen upwards. The jars with a convex bottom are inserted into the ground at the bottom. The rims of the jars are either attached to the waist (Figures 1/4; 2/1-2, 5; 3/4, 6) or short neck (Figures 1/1-3, 5-7; 2/34, 6-7; 3/1-3, 5). The crowns are mostly flat at the top, the rim is straight (Figures 1/6; 2/2, 5; 3/3-5), sloping (Figures 1/1-2, 5; 2/4; 3/6), inverted (Figures 2/6-7; 3/1), protruding (Figures 1/3, 7; 3/4). Other options are inward-pointing (Figure 1/4), wide-striped (Figure 2/2) rims with a rounded top (Figures 2/3; 3/5-6). In some examples, the part joined to the back of the folded rim is decorated with a rope ornament (Figure 1/1-3), a grain-ornamented line passes along the upper part of the body (Figure 2/7), parallel lines (Figure 2/3), and a comb ornament (Figures 2/3; 1/4). A cross is inverted on wet clay on one of the rims (Figure 3/1), the other is painted (Figure 1/5). Signs on pottery are not common during this period, but cross-shaped or intersecting

Room no. 5 is located on the northwest corner of the western block. The room is square (6.7 m × 6.8 m, Figure 41). The north-western walls of the room are external. The north wall and the north edge of the west wall have completely collapsed. The south-east corner of the room is also destroyed. The eastern wall is preserved in 3-5 rows and the eastern-southern walls in a single row. As we have already mentioned, the western block, except for the north-western corner, deviates from the main axes of the structure. In room no. 5, the eastern wall and two tuff anchors remained in place. The anchors are located at a distance of 1.1 m from the east wall, the distance between them is 3.7 m. No entrance to the room is clearly preserved. A door opening on the south edge of the east wall seems more likely. The second ‘hall’ (room no. 6) is the most poorly preserved room in the building (10.2 × 6.7 m). It occupies the southern half of the western block. We can not exclude that this section also had other divisions that have not preserved. But we believe that the area was more likely to be an undivided space. For a room with a wide axis, it is logical to have partitions in the longitudinal direction, but there is a stone bench (1.41 m × 0.54 m) 456

Nora Yengibaryan and Lilit Ter-Minasyan: The Early Medieval Complex of Agarak

Figure 1. Jar rim fragments.

Figure 2. Jar rim fragments.

Figure 3. Jar rim fragments.

Figure 4. Jar body fragments.

457

Systemizing the Past

Figure 5. 1-3 griddles, 4-5 jar bases, 6 stone crusher, 7 grindstone, 8 ‘game piece’.

Figure 6. Jug rim fragments.

Figure 7. Pot rim fragments.

Figure 8. Bowl rim fragments.

458

Nora Yengibaryan and Lilit Ter-Minasyan: The Early Medieval Complex of Agarak

Figure 9. Pottery fragments.

Figure 10. Pottery fragments.

lines are found on the rim of a jar found in Oshakan13 and on the body of the karas and on the bottom of ajug from Zakari Berd.14 The parallels of the presented jars can be found in the materials of Zakari Berd,15 Oshakan,16 Dvin,17 and Hatsavan.18

(Figure 7/10; 8/10), inverted or everted straight upper part (Figures 7/1-2, 4; 8/2), with a rounded upper part, inwardly inclined (Figure 7/3). Pots with similar rims were found in Zakari Berd19 and Oshakan.20 The pots are decorated with grooved lines (Figures 10/1-2; 11/2), wavy and parallel lines made by compression (Figures 10/3-4; 11/4), wavy lines made by compression and cutting (Figure 10/5-6, 8), grain-line decoration (Figure 10/9), ribbon-strip with oblique lines (Figure 10/7) and inverted ovals (Figure 11/3). On one of the fragments, a quadrangular appendage perpendicular to the waist has been preserved (Figure 10/10). The surface of a group of fragments is represented by images of an irregular parallel ridge left by a grinder or a ‘comb’. As the above-mentioned pottery is very fragmentary, some fragments of the body may belong to jars or bowls. The gray fragment stands out, with a series of deep oblique grooves passing along the bottom (Figure 11/6). We do not know of such a decoration on medieval vessels.

Pots. Numerous fragments of brown, gray, ash-brown, brown-reddish rims, body sherds, and bases were found. They are mainly made of clay containing fine sand. The pots are both with and without handles, taps, the body sherdsare both decorated and plain, some of them are ash-covered. The surface of the fragments is uneven, sometimes with crumbling grains of sand or well-groomed. The pots were classified according to the form of the rims: rounded top and a deep groove inside (Figures 7/7; 8/3-6, 8-9; 9/8), cut obliquely (Figures 7/5-6, 8-9; 8/1, 7), inverted in the center Yesayan and Kalantaryan 1978: Plate IV/9. Asatryan 2005: 56, Plates XXV/12, XXIX/3. 15 Asatryan 2005: Plates XVI/4-7, XVII/6, 10, 13-14, XVIII/1, 4-15, XX/3, 9. 16 Yesayan and Kalantaryan 1978: Plate IV/4, 8. 17 Kalantaryan 1970: 8, 11, Plate XXV/1, 3. 18 Tiratsyan 1963: 109. 13 14

19 20

459

Asatryan 2005: 56, Plates XXII/1, 8-10, XXVII/ 5, 14, 19, 22. Yesayan and Kalantaryan: 1978, Plate IV/1, 5, 7, 10-11.

Systemizing the Past

Figure 11. 1-4 pottery fragments, 6-7, 9 base fragments, 8 lamp fragment.

Figure 12. Pottery handles.

The components of the jars are cylindrical taps with a rounded edge. One of the taps belongs to a large vessel made of gray clay with large inclusions and an uneven surface (Figure 13/7). The second belongs to abrickgray sample of a small jar (Figure 13/8). A portion of the rim was preserved on a fragment of the tap of light pink, well-designed wide-jar (Figure 13/9). The parallels of these taps are known from Zakari Berd fortress-settlement.21

neck and straight rim with a rounded edge (Figure 6/38). The parallels of the last two types can also be found in the materials of Oshakan, Zakari Berd, and Aghdzk.23 There are many fragments of yellow jugs decorated with wavy lines or ‘comb’ ornament (Figure 11/1). These jugs have a flat base (Figure 11/9). Handles. There is a variety of handles: light yellow (yellow-green), red, light pink, red-brown, brown, and black. Mostly not with a flat bottom, round or oval in shape. These handles are connected to the waist of the vessel from the rim or the upper part of the neck, except for one example. The latter probably belonged to a churn (Figure 12/9). These handles are undecorated (Figures 12/6-8; 13/1) or decorated with an inverted wedge pattern (Figure 12/3, 5), a frame ornament (Figure 12/4), holes (Figure 12/1-2), and with spherical protrusions (Figure 13/2), with weak longitudinal grooves (Figure 13/3-6). There are parallel lines on two examples (Figures 12/1; 13/5). The handles

Jugs. The yellowish, brown, gray, gray-brown, and dark pink fragments found belong to several types of jugs. The first are jugs with a narrow neck and round or three-pointed rims, dominated by the fragments with a yellowish surface (Figure 6/2). Such decoration of rim is known from the Urartian period, which is also attested in the Middle Ages. The parallels can be found in the materials of Oshakan and Zakari Berd.22 Next are the jugs with a narrow neck and a straight rim at the top (Figure 6/1). The third variant are jugs with a narrow Asatryan 2005: Plate XXX/1, 2, 8-11, 13-14, 18-19. Asatryan 2005: Plate XXI/5-6; Kalantaryan and Simonyan 2005: Plate 1/5; Yesayan and Kalantaryan 1978: Plate V/16.

Asatryan 2005: Plates XXI/1-4, XXII/1-3, 16, XXVII/18; Kalantaryan and Simonyan 2005: Plate 1/4; Yesayan and Kalantaryan 1978: Plate V/4, 14.

21

23

22

460

Nora Yengibaryan and Lilit Ter-Minasyan: The Early Medieval Complex of Agarak 9/7, 9), also known from Oshakan29 and Zakari Berd.30 The third typeis vessels with deep, almost straight walls and rounded rims (Figure 9/5-6, 8, 10). The parallels are known from Zakari Berd.31 The round bottoms (Figure 11/5, 7) belonged to bowls and have their parallels in the materials of Oshakan, Dvin, and Aghdzk.32 Griddles. They are brown or brown-brick, made of sandy clay. They are mostly flat, rough-processed, with thick, straight, rounded (Figure 5/1-2), or flat-cut (Figure 5/3) walls. These were used for baking bread. We have contemporaneous parallels of these items from the Zakari Berd fortress.33A fragment of a flat bowl with a rounded taop (Figure 11/8) has its parallel in the materials of Zakari Berd. It was used as a lamp.34 A fragment of a brown bowl with rounded sides was also found. Similar items are considered to be ‘game piece’ (Figure 5/8). A ‘game piece’ was also found in Sotk.35 Two stone tools have been found: a basalt grinder (Figure 5/6) and a sharpener (Figure 5/7). Construction Elements and Building Technique The presence of bases in the three rooms of the building indicates that a column-beam system was used parallel to the load-bearing walls.36 The roof rested on both the walls and the columns. The bases are all made from tuff and are differ in size. The bases in the hall are square (Figure 29c). Room no. 2 has bases on a rectangular slab (plinth) with a diameter of 0.28 m, shelves with cylindrical protrusions 0.03 m in height, on which the columns sit (Figure 29d). It is interesting to use the shelves directly on the plinth, as they are typical of more complex base designs (i.e., Achaemenid bases, toroidal form bases dating back to the Classical Age).37 These bases occupy an intermediate position between the complex compositions and simple geometric shapes. In room no. 5, the bases have a form of a rectangular prism and are heavily weathered (Figure 29e). All bases are polished. With the simplicity of the forms and size these bases have parallels in the examples of Zakari Berd, but the examples from the latter site are often found with a hole made for the column38 which indicates that a greater ‘load’ was placed on the columns. There is an interesting pattern to the

Figure 13. 1-6 pottery handles, 7-9 taps.

belonged both to jars and pots. The yellowish handles with shallow, wide grooves belonged mainly to combornamented jars. A similar example of a fully preserved handle (Figure 12/1) is found in the materials of Zakari Berd fortress.24 These handle types have their parallels in Oshakan25 and Sotk.26 At Zakari Berd there are parallels of most of the handles presented here.27 Parallels of handles with the above mentioned wedge or circular inversionare not known. Bowls. These are brown, gray-brown, gray, black, pink, and red-colored, with uneven, well-worked surfaces. Fragments made of clay containing fine sand constitute a large group. There is a single example of a hemispherical bowl with a straight cut rim at the top (Figure 9/1-5). Similar examples are present in the materials of Oshakan28 and Zakari Berd. The next group is bowls with inverted or everted rims (Figure

Yesayan and Kalantaryan 1978: Plate V/ 7, 9. Asatryan, 2005: Plates XXXII/3, XXXIII/19, 21. 31 Asatryan 2005: Plate XXXIII/1, 13. 32 Kalantaryan 1970: Plate XXVIII/1-3; Kalantaryan and Simonyan 2005: Plate 1/7; Yesayan and Kalantaryan 1978: Plate V/17, 20. 33 Asatryan 2005: 54. 34 Asatryan 2005: 56, Plate XXXII/10. 35 Mirijanyan and Grigoryan 2018: Plate 2/11. 36 The column-beam system has been used in Armenia since the 3rd millennium BC: Bayburtyan 2011: 33. 37 Kanetsyan 2020: 17. 38 Asatryan 2005: 28-29, 33, room no. 4, complex no. 14, room no. 3. 29 30

Asatryan 2005: Plates XXIV/2, XXV/5, 17, XXIV/15-16, XXV/10. Yesayan and Kalantaryan 1978: Plate V/15, 19. 26 Mirijanyan and Grigoryan 2018: Plate 2, Figures 7, 10. 27 Asatryan 2005: XXIV/1-2, XXV/5, 10. XXVI/4, 8, 11. 28 Yesayan and Kalantaryan 1978: Plate V/1, 3, 5-6, 10, 12. 24 25

461

Systemizing the Past distance between the size of the bases and the distance between the columns. Bases make up about 1/10 of the flight between columns. The flight of the columned hall is 5.4 m (from axis to axis), the bases measure 0.57 m × 0.57 m × 0.54 m. The flight in room no. 2 is 3.9 m and the bases measure 0.38 m × 0.38 m × 0.15 m. The flight in room no. 5 is 4.1 m, and the bases measure 0.41 m × 0.35 m × 0.35 m. Naturally, the size of the columns changed depending on the size of the bases. Due to the small size of the bases, wooden columns were placed on them.

If we add the outer plaster, we get to the wall thickness mentioned above. The external faces of the first rows of the eastern block of the complex consist of large stones (0.7 m × 0.6 m). The stones of the external rows measure 0.3 m × 0.4 m. A similar principle is applied in the ‘first hall’ of the early medieval complex of Aruch,43 where the external face of the walls in the classical midis is made of larger stones.44 The openings are mainly located at the intersections of the walls, due to which it was possible to avoid additional angles and complex volumes. The door openings are about 1.2 m wide, with the exception of the doors of room no. 4, which connect the hall with the columns to the western block, which measure 0.8 m. No window openings were preserved. Due to the poor preservation of the walls, it is difficult to say with certainty whether there were any windows at all. We can note that there are no traces of openings on the relatively well-preserved north wall. It is possible that the windows were ‘sitting’ on the last row of stone, within the layer of unfired mudbricks. Researchers investigating Armenian architecture, starting with T. Toramanyan, have emphasized the lighting of secular structures through garrets.45 However, it is difficult to imagine that all the rooms in this building had garrets, so we consider the existence of windows to be possible.

The walls are laid on a rock without any preparatory work. There are numerous buildings of medieval Armenian architecture built directly on the rock outcrops that have preserved. The walls of the complex are about 1.3 m thick, except for a small partition of room no. 4, which is 0.85 m thick. The walls are three-layered, consisting of rows of stones placed on both sides, filled with clay and small stones. The stones that make up the walls are slightly leveled and adjusted to each other. This type of three-layer wall is typical of ancient and early medieval secular architecture. It precedes the formation of the Armenian midis technique. From the Early Middle Ages39 the midis, which was widely used in Armenian architecture, was made of lime beton, and the walls were polished. In the Agarak complex, lime is used only in wall plaster. The walls are preserved in a maximum of 5-6 rows. In well-preserved areas, the walls end at about 1.45 m above the floor. The intersection of the walls in the same plane suggests that they were lined above that mark with unfired mudbricks (Figure 29g). Such wall construction was used in the Urartian period,40 as well as in Classical period structures of Draskhanakert, Shirakavan, and Yervandashat, dated between the 5th century BC and 3rd century AD.41 To this day, thre are 18th-19th century buildings in Yerevan where the lower part of the walls is made of stone and the upper part is made of unfired mudbricks. Unfired mudbricks of different sizes were also used in the early medieval layer of Dvin.42 One of the supportive evidencefor using unfired mudbricks at Agarak are the fragments of these (0.4 m × 0.25 m × 0.1 m) found from the eastern wall of the columned hall. They were used during the modification of the door between the columned hall and room no. 2. Unfired mudbricks (0.4 m × 0.4 m × 0.1 m) were also used in the niche. The opening here is 1 m long and after placing two rows of bricks, the remaining 0.2 m area was filled with clay and small stones (Figures 29f; 43). From this we can assume that the raw bricks were made in the molds, otherwise, they would habe been adjusted to the size of the opening. The thickness of the stone walls is 1.3 m, then three rows of bricks were applied, the plaster between the bricks is 1.5 cm.

Roof constructions. It is possible to make judgments on the roof constructions based on the composition of the building, the preserved bases, and the walls. No wooden roof constructions have been preserved. During the excavations, only fragments of tiles were found, which indicate the existence of a sloping roof. Since the bases in the columned hall are placed in a way that they form a square in the central part of the room, it seems probable that the roof was made according to the principle of hazarashen,46 known from traditional folk architecture, and ended with a garret. (Figure 31). The fact that the tuff hearth of the hall is placed with a slight deviation from the center, so as not to appear directly under the garret, lends support to the existence of a garret. Fragments of tiles were found mainly in the columned hall. In the Early Middle Ages, tile was used mainly in religious buildings,47 the Agarak complex being an exception in this respect. About a dozen fragments of flat tiles (solen) were found (Figure 14/1-5, 7) along Matevosyan 2020: later proved that the building is a basilica of the 4th-5th centuries. 44 Harutyunyan 2002: 69. 45 Toramanyan 1942: 354. 46 In this type of square roof layout, the beams of the previous layer serve as a support for the beams of each subsequent layer, gradually reducing the flight to a tent-shape volume. 47 Mirijanyan 2010: 75-76 . 43

39 Basilicas of Kasakh, Agarak (4th-5th centuries), Parpi, Yereruyk (5th century). 40 Hovhannisyan 1996: 135. 41 Ter-Martirosov and Mikaelyan 2007: 23. 42 Kalantaryan 1970: 20.

462

Nora Yengibaryan and Lilit Ter-Minasyan: The Early Medieval Complex of Agarak

Figure 14. Tile ( 1-5, 7) and calipter fragments (6).

Figure 16. 1 jar rim, 2-4 pottery body, 5-6 jug rims, 7 bowl, 8 pot.

with one semicircular (calipter) tile fragment (Figure 14/6). The fact that the tile is fragmented does not allow us to talk about its size. The fragments have the characteristics of early medieval tiles: the lateral sections of flat tiles are about 5 cm, the sides of the tiles are narrowed at the bottom to be placed in the upper part of the inner tile. The lower parts begin to narrow at a distance of 6.5 cm from the edge of the tile. The fragment of the calipmer has a collar tubercle which serves as a support for the next tile. The thickness of the tiles is about 2 cm. Some fragments are painted in red. Plan Analysis Figure 15. 1-4 Jar rim fragments, 5-7 tile fragments.

Precise systems of proportions were used in the ancient and early Christian architecture of Armenia. A. Sahinyan48 referred to the detailed analysis of the 48

463

Sahinyan 1955: 141-189.

Systemizing the Past

Figure 18. 1-3, 5, 7 body fragment, 4 lamp, 6, 8, tap.

block. The construction module is also kept in separate rooms. Room no. 2 is 2 modules, no. 3 – 1.5, no. 4 – 1, no. 5 – 2.5, no. 6 – 4. The wall thickness is 0.5 modules. The proportions are not perfectly preserved in all cases, as the building has undergone degradation over time.

Figure 17. Pottery handles.

regularities used in religious architecture. The Agarak complex is built with more modest construction than other contemporary religious buildings, but here we present some intersting patterns as well.

In early medieval Armenian architecture, the kangun (about 0.5 m), was used as a unit of length. Different sizes of kanguns are obtained based on the module used for different monuments.50 Taking into account the size of the module of the Agarak complex, a kangun with a length of 0.53 m was used here.

Examining the layout of the complex, it becomes clear that there are systematizations between different parts of the structure and general sizes. The first thing that stands out is that the columned hall is divided into 4 equal parts and a construction module was used during the construction (Figure 22). Each part or module is 2.65 m, from the bases to the walls is 1 module, the distance between the bases is 2 modules.49 There are 2 modules in the northern part of the eastern block, 2.5 in the northern-southern parts of the western

Comparison with Other Similar Sites There are not many secular structures that have reached us from early medieval Armenian architecture. Of these monuments, the palace complexes are better studied in terms of architecture: Dvin, Aruch, Zvartnots, and Avan. An important group of secular monuments constitute the fortress-settlements (Zakari

49 This has parallels with the early Christian three-nave basilicas, where the size of small naves is 1 module, the main nave is 2-3, see Sahinyan 1955: 174.

50

464

Sahinyan 1955: 175.

Nora Yengibaryan and Lilit Ter-Minasyan: The Early Medieval Complex of Agarak

Figure 19. Measurement.

Figure 20. The plan in the first stage of operation.

465

Systemizing the Past

Figure 21. The plan in the second stage of operation.

the similar structures with a four-column composition is from room no. 2 of 2D complex at the site of Horom52 dated between the 7th-5th centuries BC (Figure 28a) and the hall of the Teishebaini multi-section residential house, dated to the 7th century BC (Figure 28 b).53 Areshian and Ghafadaryan note that ‘The centralstructured buildings were known as far back as the Urartian period, and the tower-shaped temples with a square plan of Altıntepe later appear in Achaemenid architecture’.54 Folk houses with similar compositions are known in Georgia, as well as in residential and religious buildings in Afghanistan, India, and Central Asia.55 Armenian architecture has always been influenced by the traditions of folk construction techniques. The glkhatun had an influence on the early medieval secular monumental architecture. Part of the palace buildings were the halls with four or three pairs of columns, which served as a throne room, a dining room for ceremonial receptions.56 According to the accepted view,57 each of the four columns of the palace halls was covered with a hazarashen type ‘dome’, which had a four-sided pyramid cut out from the outside and ending with a garret. We have two pairs of columns in

Berd fortress, four fortresses of Ashnak, Kakavadzor, Davtashen) and castle-fortresses (Kosh, Hatsavan, Ditak).51 Of the mentioned monuments, only Zakari Berd has been fully excavated, which, according to a preliminary plan, is a fortress-settlement and dates back to the 3rd-5th centuries AD. The Agarak complex has parallels with both Zakari Berd and the palaces of Dvin and Aruch. The similarities with Zakari Berd refer mainly to the building technique and construction elements. Both monuments have the same three-layer walls, with the difference that the walls in Zakari Berd were made of stone at full height. Due to the general construction technique, the size of the rooms and the distances between the columns in Zakari Berd and the Agarak complex are also similar to each other. However, in terms of composition, Agarak has more parallels with the early medieval palace structures, as they are based on the principle of ​​a glkhatun (lantern roof). The glkhatun is one of the most popular types of traditional residential houses in Armenia. In the plan, it is square, sometimes rectangular. The glkhatun is covered with a hazarashen ending in a garret. The roof rests on wooden columns that stand freely or are attached to the walls. The hearth or tonir is one of the obligatory elements of the house. The glkhatunis attested at least from the first millennium BC. One of 51

Badalyan et al. 1994: Figure 8. See also Genç and Konyar in this volume. Areshyan and Ghafadaryan 1996: 20. 55 Ilyina 1946: 22-29. 56 Areshyan and Ghafadaryan 1996: 75. 57 Hovhannisyan 1996: 75. 52 53 54

Mirijanyan 2013: 8-11.

466

Nora Yengibaryan and Lilit Ter-Minasyan: The Early Medieval Complex of Agarak

Figure 22. Graphic analysis of the plan.

Agarak. There are three pairs of columns in the first Catholicos’s palace of Dvin dating to the 5th century (Figure 28d) and governmental palace of Aruch dated to the 7th century58 (Figure 28e) and four pairs of columns in the 7th century Dvin Catholicos’s palace (Figure 28f).59 Each central set of forty columns in the halls is turned into a unique volume unit by being covered separately. If we arrange it according to the growth, it turns out that in Agarak (Figure 28c) we have one such unit, in the first Dvin palace and Aruch palace – two, in the second Dvin palace – three. Comparing these structures, we see that the Agarak complex is the closest to the traditional folk residential house, as the columned hall fully corresponds to the composition of the glkhatun. But with its size, the complexity of the design, and the construction techniques used, it is much more than an ‘ordinary’ residential house. Thus, the Agarak complex occupies an intermediate position

58 59

between the ‘palace structures’ and traditional folk residential houses. Changes in the Use of the Complex The Agarak complex has undergone several changes during its existence. There is reason to believe that the structure was completed in two stages. The eastern block was built in the first phase and the western block in the second. The assumption has several arguments. First, the north-western corner of the eastern block is lined with large stones (0.82 m × 0.93 m), only the outer corners of the building have a similar solution, in contrast to the corners formed by the outer walls and partitions (Figure 20). The next argument is that smaller stones are used in the first rows of the western block walls compared to the first rows in the eastern block. The difference in rows is visible on the north façade of the building (Figure 26a). Second, the door between the columned hall and room no. 2 was closed and turned into a niche. The closed

Khalpakhchyan 1971: 92. Khalpakhchyan 1971: 95.

467

Systemizing the Past

Figure 23. Transverse sections.

door is clearly expressed in the north wall of the hall (Figure 26a). The opening in the columned hall was closed with a layer of stone and in room no. 2 the opening of the door was closed with 4 rows of unfired mudbricks sitting on 2 slabs (Figures 29f; 43). The depth of the niche is 0.5 m. It sits 0.75 m above the floor of the room, and inside was plastered with lime mortar. Due to this change, room no.2 was cut off from direct communication with the columned hall.

reconstruction, and the period of its existence suggest that it may have been adapted for the Zoroastrian temple. An example of this assumption is the only known fire temple in Armenia, about which there is both literary evidence and archaeological data. Armenian and Georgian historians (Hovhannes Draskhanakerttsi, Tovma Artsruni, Sebeos, Kirakos Gandzaketsi, Arsen Saparatsi, and others) mentioned the Zoroastrian temple built in Dvin.60 The Dvin fire temple was built next to the western wall of the columned hall of the Catholicos’s palace. Organically it was included in the layout scheme, at the same time it is noticeable that it belongs to the second construction period of the structure. The bases of this four-column building (one has been preserved on the site, the second was found in a wall structure of a later period) are slightly different in detail from the bases in the columned hall. The floor of the building differs in its height (0.7 m) and composition. In the central part, a square made of three large tuff stones with a surface cracked from heat, a grooved center (1.5 m × 1.5 m × 0.3 m), and a jar full of ashes (Figure 28d) were uncovered. A large

Third, on the south wall of the columned hall, the doorway on the axis was closed with a row of stones. It may have been moved to the western edge of the wall, as it ends in a smooth cut surface. The structure in that area is so poorly preserved that it is difficult to come to a definitive conclusion. Fourth, the construction of the building on the east side of the complex. The changes made may be due to changes in the function of the complex. Reasonable grounds were needed to transform the building into a concise, welldesigned plan. The location of the Agarak complex, its

60

468

Ghafadaryan and Kalantaryan 2002: 67; Hakobyan 2011: 196-197.

Nora Yengibaryan and Lilit Ter-Minasyan: The Early Medieval Complex of Agarak

Figure 24. Longitudinal sections.

Armenians, which in 572 AD turned into a rebellion. The rebels led by (Junior) Vardan Mamikonyan killed the governor and destroyed the fire temple.61 The fire temple survived for a very short time. The layout of the hall with the columns of Agarak complex has a lot in common with Dvin temple. It is also four-column, made of one-piece tuff, with a hollow hearth inside full of ashes, and the hall has been modified. Given the strong foundations of Christian religion in the region, namely the abundance of churches, monuments, the

red lid of a water bowl was also found. According to the analysis of the material, the building was not built especially for the fire temple, but one of the rooms of the palace was adapted for that purpose. Suren Tchiovr, a Persian nobleman who in 564 AD was appointed as a governor of Armenia by Khosrow I (531-579 AD), the Sasanian King of Kings, pursued a harsh and heavy tax policy and religious intolerance towards the locals. The Persians stripped the Catholicos’s palace, making it the governor’s residence, and built a Zoroastrian temple in the western part of the palace. The construction of a fire temple created a dissatisfaction among the

61

469

Ghafadaryan and Kalantaryan 2002: 69-70; Hakobyan 2011: 198-199.

Systemizing the Past

Figure 25. Eastern facade.

presence of the tomb of Aghdzk, and the unique church of Agarak, the Persians turned the building into a fire temple to strengthen their positions. Both the Dvin temple and the Agarak columned hall are close in design to the most popular ‘chahārtāhg’ type temples in Sasanian Iran, which translates to ‘four arches’. Chahārtāhg was a domed hall with a square plan, with arches resting on four columns, where the sacred fire was lit. Such a building presupposed the existence of a second closed service area, where the sacred fire was kept. In addition, the presence of water nearby – a river, a stream or a lake – was a factor.62 The entrance to the temple was mainly from the east, sometimes from the west and the south. The Agarak complex meets almost all these requirements. Both Dvin and Agarak temples existed for a very short period of time. If we accept the hypothesis of turning the complex into a fire temple, the building built in the east could serve as an auxiliary area for the preservation of the sacred fire. Naturally, a 62

door had to be opened in the east wall of the hall with columns to communicate with the service area. There are two burials in that part of the wall, and there are no burials directly on the wall in any other part of the main structure. In other words, the wall in this part was already collapsed due to the opening, due to which it was possible to lower the burials into the wall (Figure 19). Room no. 5 of the Agarak complex is the only one that was later used as a living space. Evidence of this are the remains of the walls and the hearth. Griddle fragments were found in the hearth. The room is the only one where no burials were found (Figure 19). Auxiliary structures serving the house were not preserved. They were carried out with more modest construction means, usually, and only the foundation rows preserved. As the walls in Agarak were placed directly on the rock, the traces of those buildings did

Voronina 1970: 341.

470

Nora Yengibaryan and Lilit Ter-Minasyan: The Early Medieval Complex of Agarak

Figure 26. Northern facade.

Figure 27. Southern facade.

471

Systemizing the Past

Figure 28. Comparison with similar buildings.

472

Nora Yengibaryan and Lilit Ter-Minasyan: The Early Medieval Complex of Agarak

Figure 29. Constructive elements.

473

Figure 30. 3D restoration of the building.

Systemizing the Past

474

Figure 31. Partial 3D restoration of the roof.

Nora Yengibaryan and Lilit Ter-Minasyan: The Early Medieval Complex of Agarak

475

Figure 32. 3D restoration of the building in the second phase of operation.

Systemizing the Past

476

Nora Yengibaryan and Lilit Ter-Minasyan: The Early Medieval Complex of Agarak

Figure 33. 3D restoration of the interior of the columned hall.

Figure 34. General view of the complex from the air.

477

Systemizing the Past

Figure 35 View from the west.

Figure 36. View from the east.

478

Nora Yengibaryan and Lilit Ter-Minasyan: The Early Medieval Complex of Agarak

Figure 37. View from the north.

Figure 38. Room no. 2.

479

Systemizing the Past

Figure 39. Room no. 3.

Figure 40. Room no. 4.

480

Nora Yengibaryan and Lilit Ter-Minasyan: The Early Medieval Complex of Agarak

Figure 41. Room no. 5.

Figure 42. Annex.

481

Systemizing the Past

Figure 43. Niche of room no. 2.

not reach us. Remains of walls and rock-cut mounds have been preserved.

there were columns in room no. 5 with the same flight and a smaller flight in room no. 2, then there should have been columns here as well.

Three-Dimensional Restoration of the Site

In the profiles, the walls on the fronts were filled with stone up to 1.45 m, and above that with unfired mudbrick (Figures 23c; 24c; 25b; 26b; 27b). Of course, we do not know exactly how high the walls were, based on the layout we can assume 3-4 m heights. As mentioned, a construction module of 2.65 was used in the layout of the complex. In early medieval Armenian architecture, the structural module was expressed in the facades of the building. It is possible that it was similar to that in the Agarak complex. The maximum height of the structure, that is, from the floor of the columned hall to the hazarashen garret, we conventionally accepted two modules: 5, 3. We built the hazarashen, which was 2.1 m high. In the flat part, the height of the room is 2.9 m, the height of the walls on the facades is 3.4 m. Dimensions expressing height are conventional and were used to visualize the original appearance of the monument. The rest of the rooms of the building are supposed to be covered with a traditional flat or a slightly sloping roof, except for room no. 5, where it is possible that a hazarashen roof was also used. The basis for such an assumption is the square layout of the room. There were several complexes of dwellings with glkhatun in

A three-dimensional computer model was built to illustrate the original appearance of the monument (Figures 30-33). 3D modeling has become one of the most important tools in the study of monuments in recent decades. The digital model can be used to represent the different stages of the construction of a monument. In the case of a multi-layered monument, it is possible to view each construction horizon separately, all at the same time. In the case of the Agarak complex, when most of the walls are preserved at the level of one row, and some are not preserved at all, we can get an idea of t​​ he spatial solutions of the building through a three-dimensional model. The simulation was performed using AutoCAD software. To get the three-dimensional model, we first tried to reconstruct the structure in two-dimensional drawings. The collapsed walls were filled in the plan (Figure 20). The missing bases in rooms no. 2 and no. 5 have been filled in. As both rooms are almost square, the presence of columns in one direction seems illogical. We also added bases in room no. 6, because if 482

Nora Yengibaryan and Lilit Ter-Minasyan: The Early Medieval Complex of Agarak Archaeology, Ethnology and Folklore of the Caucasus: 52–57. Etchmiadzin, Press of Holy Etchmiadzin (in Russian). Badalyan, R., Kohl, Ph. L., Stronach, D. and Tonikjan, A. V. 1994. Preliminary report on the 1993 excavations at Horom, Armenia. Iran 32: 1–29. Bayburtyan, E. 2011. The Sequence of the Oldest Cultures of Armenia based on Archaeological Material. Yerevan, History Museum of Armenia (in Russian). Ghafadaryan, K. and Kalantaryan, A. 2002. Dvin II. The City Dvin and Its Excavations (1973-1980). Yerevan, Publishing House ‘Gitutyun’, NAS RA (in Armenian). Grigoryan, G. 2012. Early Medieval Four-Sided Stelae in Armenia. Yerevan, History Museum of Armenia (in Armenian). Тoramanyan, Т. 1942. Materials on the History of Armenian Architecture. Yerevan, Armfan (in Armenian). Hаkоbyan, N. 2011. Hovsep Orbeli and the Sasanian culture. Museum Annual Magazine: 194–199 (in Armenian). Harutyunyan, V. 2002. Secular architecture. In M. Hasratyan (ed.), History of Armenian Architecture, volume 2: 52–79. Yerevan, Publishing House ‘Gitutyun’, NAS RA (in Armenian). Hovhannisyan, K. 1996. Urartian architecture. In G. Tiratsyan (ed.), History of Armenian Architecture, volume I: 87–181. Yerevan, Publishing House ‘Gitutyun’, NAS RA (in Armenian). Ilyina, M. 1946. The oldest types of housing in the Caucasus. Architecture Institute Communications 5: 1–48 (in Russian). Khalpakhchyan, O. 1971. Civic Architecture of Armenia. Moscow, Construction Literature Publishing House (in Russian). Kalantaryan, A. 1970. The Material Culture of Dvin in the IV-VIII Centuries (Archaeological Monuments of Armenia 5). Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Armenian). Kalantaryan, A. and Simonyan, H. 2005. Archaeological value of Aghtsk (results of the 1974-1975 excavations). Hushardzan 3: 140–151 (in Armenian). Kanetsyan, A. 2020. Bases of columns of Ancient Armenia. In Historical and Cultural Heritage of Shirak: Contemporary issues of Armenology 10th International Scientific Conference (Shirak Centre Armenological Studies 2): 15–35. Gyumri, Shirak Armenological Centre (in Armenian). Matevosyan, K. 2020. Historical-Archaeological Observations. Yerevan, Mughni (in Armenian). Manucharyan, A. 2017. The Princely Homes in Aragatsotn in the IV-VII centuries. Banber Erevani hamalsarani (Bulletin of Yerevan University) 3: 35–46 (in Armenian). Mirijanyan, D. 2010. On one type of constructional ceramics used in the Early Middle Age. Etchmiadzin 11: 74–79 (in Armenian). Mirijanyan, D. 2013. Early Medieval Material Culture of Aragatsotn, PhD thesis, Yerevam, Institute

the Armenian traditional folk architecture, which were documented in Zangezur and other places.63 In the three-dimensional model of the monument, in addition to the structural elements, interior details were restored: benches, hearth, and paving stones (Figure 33). Conclusion The complex was built with a preliminary plan. The application of the principles of structural module and symmetry indicates that the construction was carried out by professional masons. The size of the building testifies to the high social position of the customer and financialpossibilities. In the first stage of occupation, it was a private house. In the second stage, the function of the complex changed possibly serving as a Zoroastrian temple for a very short time. The composition of the glkhatun of a traditional folk house is used in the Agarak complex. It is much more than an ‘ordinary’ house in terms of size, dersign, quality, and construction materials used. The types of household and construction fragments found in the complex with their preparation techniques, peculiarities of decoration, and quality correspond to the materials of Zakari Berd fortress (3rd-5th centuries), Oshakan castle-fortress (4th-7th centuries), the early medieval layer of Dvin (4th-6th centuries) and are dated to the end of the 4th – 6th centuries. In all probability, it was destroyed in the seventies of the 6th century and was turned into a cemetery. The Agarak complex is a unique secular structure that was preserved from the early middle ages. It occupies an intermediate position and is a connecting link between the traditional folk houses and the early medieval palace complexes. It fills a gap in a series of few studied secular monuments. Bibliography Areshyan, G. and Ghafadaryan, K. 1996. The architecture of the Armenian Highlands in the NeolithicEneolithic era. In G. Tiratsyan (ed.), History of Armenian Architecture, volume I: 19–32. Yerevan, Publishing House ‘Gitutyun’, NAS RA (in Armenian). Asatryan, E. 2004. The Monuments of Talin Region. Yerevan, Hushardzan Press (in Armenian). Asatryan, E. 2005. Zakari Berd (Archaeological Excavations in Armenia 19). Yerevan, Publishing House ‘Gitutyun’, NAS RA (in Russian). Avetisyan, P. 2003. Preliminary results of excavations of the Agarak site. In A. Kalantaryan, L. Abrahamyan, R. Badalyan, L. Bardanyan and H. Xaratyan (eds), 63

Khalpakhchyan 1971: 69.

483

Systemizing the Past of Archaeology and Ethnography, NAS RA (in Armenian). Mirijanyan, D. and Grigoryan, A. 2018. Medival Sotk under the light of the newest archaeological investigations. Etchmiadzin 3: 87–102 (in Armenian). Sahinyan, A. 1955. The Architecture of Kasakh Basilica. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Armenian). Ter-Martirosov, F. and Mikaelyan, L. 2007. The old roots of construction techniques. Architecture and Construction 5: 22–24 (in Armenian).

Tiratsyan, G. 1963. The excavations of Hatsavan fortress in 1962. Lraber hasarakakan gitut‘yunneri (Herald of Social Sciences) 2: 103-110 (in Armenian). Voronina, V. 1970. Architecture of the Sassanid period. In O. Khalpakhchyan, E. Kvitnitskoy, B. Pavlova and A. Pribtkovoy (eds), Architecture of the Ancient World, volume 1: 328–347. Moscow, Strojizdat (in Russian). Yesayan, S. and Kalantaryan, A. 1978. Early medieval settlement of Oshakan. Lraber hasarakakan gitut‘yunneri (Herald of Social Sciences) 3: 79–90 (in Armenian).

484

Women in Urartian Rituals Nora Yengibaryan

Institute of Archaeology and Ethnography, NAS RA, Yerevan, Armenia Abstract: This article reviews the available evidence on the role of participants, and more specifically women, in the various rituals dedicated to the gods of the Urartian pantheon. Although written sources are rare, these can be augmented by iconographic evidence found ritual objects in order to attempt a reconstruction of the place of women in ritual performance. Keywords: Urartu, women, priestesses, ritual, sacrifice, belt, medallion, votive plate, ritual utensils

Introduction There is scarce information in the written sources on the process, number, and the ‘role’ of participants in rituals dedicated to the gods of the Urartian pantheon.1 The main source of information is the ‘Door of Mher’ inscription, which lists the names of the gods and the type and number of sacrifices that were made for them in the month of the Sun god.2 There is also some written information about sacrifices performed at the end of the construction of cultic buildings.3 Rituals accompanied by sacrifices in the presence of the king are probably described in three incomplete inscriptions from Armavir and Karmir Blur.4 These inscriptions mention groups of people engaged in certain rituals, who probably belonged to the priestly class or served in the temple (LÚurbikani (sacrificer), kurni,5 LÚšeluini, LÚ purunurda, LÚiriunili, LÚšiarkira, LÚaṷeiteni, LÚpalagi).6 Undoubtedly, there were also priestesses in the Urartian temples, but there is no written information about them. This gap of information, in some extent, is filled by the iconography on ritual objects including bronze belts, votive plaques, and medallions. Belts, Votive Plaques, and Medallions On the depictions of these ritual objects (especially on belts), it is mainly women who are making sacrifices and offerings to the gods, pray, and participate in the ritual procession. The depictions of worship scenes on the votive plaques can be divided into two types: 1) the worshiping woman 1 I would like to thank Y. Grekyan for the choice of the theme, the literature, and valuable comments. 2 Harutyunyan 2001: no. 38; Salvini 2008: A 3-1. The month of Sun or Shivini perhaps corresponds to Armenian month of Areg (Karagyozyan 1981: no. 6, n. 1). 3 Harutyunyan 2001: nos 122, 123, 415, 418, 421; Salvini 2008: A 5-87, A 125-89, A 12-41, A 12-6, A 12-8. 4 Harutyunyan 2001: nos 193, 270, 424, 425; Salvini 2008: A 8-14, A 12-3, A 12-2. 5 Cf. the Armenian word k‘urm ‘priest’ (Harutyunyan 2001: no. 193, n. 24). 6 Harutyunyan 2001: nos 193, 270, 418, 424; Salvini 2008: A 8-14, A 12-3, A 12-6, A 12-2.

Systemizing the Past (Archaeopress 2023): 485–496

is depicted alone7 or 2) standing in front of a goddess8 or a god.9 Sometimes a sacrificial goat is also depicted. Worshipers are depicted veiled or bareheaded, with their hands in an orant pose or holding a standard, spike, bucket, paternoster consisting of series of beads and more often holding two of these ritual objects. The same objects are depicted in the hands of the gods/ goddesses. The scenes on several votive plaques are different from the more commonly found scenes in terms of iconography. On two of these, the worshiping woman is shown appealing to goddess/queen, who is sitting on the throne with folded legs and a raised bowl in front of the chancel table with crescent-shaped bread.10 Alternatively, the goddess is sitting on a throne under the winged sun disk with a worshiping woman and sacrificial goat in front of her.11 In another version, the table is missing and the throne of the goddess/ queen is situated in front of fortress or temple doors.12 These are unique examples in Urartian iconography depicting scenes of offering. Most ritual plaques depicts plain worship scenes. The depictions on the votive plaques is repeated on four medallions showing scenes of worship. Noblewomen in an orant pose are making offerings to the goddess and to the god seated on a throne or standing on the back of a bull. In two examples they are depicted with a sacrificial goat/kid.13 Depictions of rituals on narrow Urartian bronze belts are different. This type of belt is believed to have been worn by women. The main scene depicted on these belts is the ritual performed at the chancel table. The iconography of these objects of great artistic value can be divided into two types. The images of the belts of the first type are surrounded by a chain of buds, while the

Caner 1998: 184-286. Caner 1998: 289-345; Hori et al. 1982: nos 72-73, 75. 9 Caner 1998: 346-377; Hori et al. 1982: no. 77. 10 Caner 1998: 412, 423. 11 Caner 1998: 425. 12 Caner 1998: 424. 13 Yengibaryan 2012: Figure 2/3-5, 9. 7 8

Systemizing the Past scenes are separated by rows of dots14 (in one instance, by rosettes15). In these isolated parts are depicted birds, fish, sheep (in one case also goats), priestesses, or groups of temple servers who are carrying offerings or ritual utensils. In the central part is the depiction of a queen or a priestess, raising a bowl and seated on a throne in front of the chancel table with a crescentshaped and round bread and ritual ware. On the other side of the table stands the worshiper or server. The latter are absent in the depictions of two belts. Behind the throne of a queen or high-ranking priestess stand servants with a fan, sometimes with a bowl, a ritual bucket, a libation vessel, or a paternoster. The animals and people are depicted in motion and pointing towards the chancel table. On the belts of this group there are images of a temple or a fortress. These two belts are also distinguished by iconography. One depicts only women playing musical instruments, griffins, rows of eagles, and sheep.16 The second depicts griffins, holding a goat with a beak and/or standing near the tree of life, shedus, a musician, a libation ritual, priestesses/ servants with bowls and jugs. Thus far, this is the only image of a tree of life depicted on a narrow belt known to us. Only bread is placed on the chancel tables of the mentioned belts. The absence of buildings suggests that the festival was performed outdoors.

shaped and round bread, an askos, and two large bowls. The participants of the ritual procession carry buckets, paternosters, bowls, towels, shoes, and conical items, probably for ornamentation (Figure 2). Scenes of ritual libation, female musicians, and an acrobat are depicted here.21 The upper register of the second belt separated the figures into groups by three rosettes depicted between them. The figures of procession are repeated (Figure 1).22 The iconography of the second belt at first sight seems to represent household episodes: bakers near a tandoor, bed decorators, carpet or rug weavers, and a woman walking under a sunshade. The image of the latter is repeated also on an incomplete belt depicting a wedding ritual (Figure 4/2).23 This belt, decorated with buds and dot ornament depicts a two-wheeled chariot with two oxen and a decorated carriage, the coachman is wearing a in knee-length tunic. The participants of the procession in front of the chariot include music-playing women, an acrobat, and a richly dressed queen or priestess, with a piece of clothes in her hand and a veil reaching her knees. She is walking under a sunshade held by two servants. In front of them two women carring a bowl, a high-heeled cup, a standard, and sunshades. Among them, there is a noblewoman wearing a long veil, holding her hands in the orant pose.

A single partially preserved belt depicting a building, perhaps, a temple, is the only exception (Figure 4/1).17 The preserved segments of the belt show images of eagles, a temple, sheep in two rows, musicians, and probably a dancer with a bud attached to her head jewelry, standing between them.18 The goddess/ queen, sits on a throne in front of the chancel table, with her feet on the footrest, holding a bowl in her right hand, and a paternoster in the left hand. A large bowl, crescent-shaped and round bread are placed on the table. Probably the king or a high-ranking official is standing in front of the chancel table with a raised right hand and holding an undetermined object in his left hand. He wears knee-lenght decorated clothes. A jeweled ornament on a buckle is attached to the front part of his headpiece.19 This is the only depiction of a woshiping man on the mentioned belts.

Part of the central segment of the belt is missing. Only the chancel table with crescent-shaped and round bread, two bowls, and an undetermined object are preserved. Under the wall of the temple depicted at the right end of the belt sits a woman in simple clothes, with her hair gathered around her neck. Women carrying bowls, buckets, a single-handled jar, and a chancel table with crescent-shaped bread, and three small bowls are depicted leaving the temple. There is also depiction of veiled women, holding two bowls and two other figures taking wine from a vessel. A fragment of the central part and some other fragments have been preserved from the next belt with the depictions of the ritual procession.24 On the chancel table are two crescent-shaped and round bread and large bowls. On the ground there is a single-handled jar. The women participating in the ritual procession carry the same objects. The entire process of wool processing, including the looms, is depicted on this belt.25 Libation vessels with bull-headed handles placed on high fourfoot basements are depicted on this belt.26 There is an image of a similar vessel on the fragment of another belt (Figure 4/3). In both cases, the libation is done by similar single-handled jars. Two fragments of this

The iconography of the second type is represented on two complete belts, which depict a ritual procession starting from a temple or fortress gates (Figure 1/2).20 In one case, on the chancel table there is a crescent-shaped bread, while in the second case there is a crescent14 Kellner 1991a: nos 261, 266, 279; 1991b: 17; Seidl 2004: Falttafel A 1-2; Ziffer 2002: Figure 3. 15 Hori et al. 1982: no. 28; Kellner 1991a: no. 251; Ziffer 2002: Figure 2. 16 Ziffer 2002: Figure 2. 17 Ziffer 2002: Figure 3. 18 Seidl 2012: 166. 19 Ziffer 2002: Figure 4. 20 Kellner 1991a: 282; Seidl 2004: Falttafel A 4.

Seidl 2004: Falttafel A 4. Kellner 1991a: 282. Ziffer 2002: Figure 4. 24 Çavuşoğlu 2014: Drawing 1a, Figure 1. 25 Çavuşoğlu 2014: Figures 6-9. 26 Çavuşoğlu 2014: Drawing 1d-e. 21 22 23

486

Nora Yengibaryan: Women in Urartian Rituals

Figure 1. Belt with the depictions of the ritual procession and a ritual structure.

487

Systemizing the Past

Figure 2. Belt with the depictions of ritual structure, ritual procession, chancel table, libation scene, bread baking, weaving, bed making, musicians.

belt with depictions of chariots, offering bearers, and musicians have been preserved.27

legs of the queen/high-ranking priestess are placed on the chair or hung, but mostly sitting with the knees bent and the feet resting on the seat. All of them wear belts and richly decorated headdresses. Judging by the images, some of them wear diadems, in the central part ornamented with a ball30 or a protrusion (Figure 3/12,4,8).31 Headdresses were decorated also with metal jewelry, as evidenced by excavated materials. In the burial jar found at Argishtikhinili, small pieces of silver jewelry sewn on the headdress were placed around the skull and neck of the deceased.32 The headdress covered the woman’s head reaching the waist or heels.

The Participants of the Ritual28 Queen/high-ranking priestess They are depicted sitting on a throne, mostly holding only a bowl (s) or a jar with a paternoster and a richly decorated headdress (Figure 3/1-8). In both cases there is a depiction of a branch together with a bowl.29 The Seidl 2004: Abb. 99. See Grekyan 2008; Hmayakyan 2001. See also Krzysztof Jakubiak, in this volume. 29 Kellner 1991a: 263; Yengibaryan 2012: Figure 2/3. 27

Seidl 2004: Abb. 102. Seidl 2004: Falttafel A4. 32 Karapetyan et al. 1996: 36. 30

28

31

488

Nora Yengibaryan: Women in Urartian Rituals

Figure 3. 1-8. Depictions of chancel tables on the belts, 9. Depiction of a clergyman with flabellum.

489

Systemizing the Past

Figure 4. 1. Belt with the depictions of animals, building, chancel table, musician and a dancer, 2. Belt with the depictions of the chariot, chancel table, libation scene, ritual wedding, musicians, acrobat, building. 3. Belt with the depictions of chariot, musicians, libation scene.

490

Nora Yengibaryan: Women in Urartian Rituals the above mentioned instruments.38 The flabellum was used in church rites and produces sound by shaking.

The headdresses at the same time emphasized the social status of the wearer. Interestingly, in addition to the three other women in the procession, those sitting on the ground near the loom are also wearing headdresses.33 The spinning of thread and weaving was also the occupation of high-ranking women and queens. A depiction of a Babylonian queen spinning with a spindle has been preserved, who was served by the servant holding a sunshade.34 The image of a weaving queen is known from many cultures of the Near East and the Mediterranean, and perhaps most familiar is that of the Queen of Ithaca, Penelope, whom Homer often depicted as weaving on a loom.35

Furniture There are chancel tables, thrones/armchairs, and chairs in the depictions. The chancel tables are fourlegged, though there is a single example pf a threelegged one.39 They are covered with cloths and bread and libation vessels, which are mainly bowls of different sizes, askos, and rhyton are placed on these. Thrones are often covered with decorated blankets. The armrests are straight or curved. In one case, a footrest was placed near the throne. The depicted furniture is framed with various metal accessories and the legs are shaped like an animal’s paw or hoof. The furniture has its parallels in the materials of Urartian tombs.

Offering bearers These women in similar clothes could be priestesses, temple servants, or maids (Figures 1-2). They wear ornamented dresses made of decorated textile, and sometimes wear skirts with rounded hems. All are depicted wearing belts, with shoulder-lenght hair mostly without any headsresses. The similarity of their clothing and hairstyles may be an indicator of status. The offering bearers depicted on one of the belts stand out with their ornamental cloth headbands, the entire part of which covers the top of the head and descends to the neck like a veil (Figure 4/2). There is a second version of the headband on the same belt: a semicircular hat with the long ends of the ribbon hanging down to the shoulders.36 These women may have had a different status than others. A separate group in the depictions are the wool makers, sitting on chairs with folded legs, who are probably temple workers.

Buildings The depicted buildings have one or two rows of windows, arched doors, buttresses ending with towers, the middle parts of which are decorated with dentil ornamentation. Only the buildings on one of the belts have two or three doors (two with a rectangular frame). The entrances of the fortress and the fortification wall may be depicted here (although the fortress doors are not installed consequently as in the depiction). It is not possible to identify the depicted buildings with any of the cultic structures mentioned in Urartian written sources.40 The most common of the Urartian ritual structures are the Haldean gates and the susi temples. Both are mainly associated with the worship of Haldi.41 Rituals and sacrifices were performed for Warubane in the Haldi temple. The first Urartian cult structures are the open-air stelae, which are not present in the discussed depictions. This leads us to assume that these crowded ritual processions took place in the squares in front of the gates and susi temples. And only those who had a direct relationship with the deity could enter the temple: the king, queen, high priest, high priestess, or special people who serve a certain god/goddess.

Musicians All of them are women, dressed similar to the participants of the festival. They are playing string, percussion, and brass instruments.37 String instruments are types of the lyre. The instruments with rectangular or square frames and parallel running wire resemble lyre. Of the brass instruments, only the double-barreled flute is depicted. The percussion instruments are the bass and bronze bowls. The latter were sounded by striking. In the depictions they appear together. The last instrument is a long-tailed, oval-shaped instrument with buds attached to the edges, played by shaking. This may be a prototype of the flabellum used during Armenian church ceremonies (Figure 3/9). The flabellum is a circular object with a long wooden tail, and metal sheet, with buds attached to the edges. In one of the miniatures, it is depicted in the same position as

Ritual Ware and Accessories Libation vessels There are two types of libation vessels in depictions under the discussion. One is a large clay vessel with a narrowing body in the center. Slightly different types were found at Patnots-Giriktepe (the height of the central registers made of triangles: 41 cm), Karmir Blur (decorated with three bull heads in the narrowing part

Çavuşoğlu 2014: Figure 1a; Kellner 1991a: 282; Ziffer 2002: Figure 4. Vardiman 1990: 218. 35 Vardiman 1990: 217. 36 Ziffer 2002: Figure 4. 37 For musical instruments see Seidl 2012. 33

Gevorkyan 1973: Plate XLVIII/1. Kellner 1991a: 282. 40 For more details see Grekyan 2018: 76-109. 41 Grekyan 2018: 80-81.

34

38 39

491

Systemizing the Past of the body, height: 50 cm, diameter: 51 cm), and temple area of Ayants (height: 107 cm, diameter: 83 cm):42 Miniatures of these vessels are also known from Karmir Blur and Noratus.43

Single-handed jars This type of pottery, dark or red polished, with a spherical body, three-pointed (oinochoe) or round mouth, and a grooved handle is the most common in Urartian sites. Only one of the rooms of the Karmir Blur citadel contained 1,036 single-handed jars of different sizes.52 The jars are depicted in the hands of pourers and ones taking wine from a large vessel. On the belt depictions these are placed near the chancel tables. Bronze and silver examples of single-handed jars are also known.53

Cauldron In the depictions, the libation is made in bronze cauldrons with bull-headed handles, placed on fourlegged basements, which end in bull’s hoofs. A similar ritual cauldron with bull-headed handles and an iron basement was found in the royal tomb of Altıntepe.44 There are also Urartian cauldrons with bull-headed handles of unknown provenance.45 Clay examples of cauldrons are also known.46

Bowls There are bowls in the hands of goddesses/queens, worshipers, as well as in the hands of participants in the ritual procession. From the images, it can be concluded that bowls were of different sizes. Most likely, clay and metal samples were used. Both versions have been recorded at numerous Urartian sites.54 A large number of bronze bowls were found at Karmir Blur.55 Examples of deep metal bowls are kept in various museums.56

Bucket In the depictions, the participants hold one or two buckets in their hands. The presence of this ritual object in Urartian iconography is everywhere. On the wall paintings and objects, there is a depiction of a bucket with sacred fruits in the hands of the priests standing next to the tree of life.47 Silver and bronze examples are also known.48

Askos These less common vessels have a slightly elongated oval body, a spout in the central part, and a handle joining the upper part of the body from the lip. They were of ritual significance. Examples of clay57 and iron58 askoses are also known.

Goblets Participants hold high-stemed goblets that could be made of metal or clay. Clay goblets with red, engobe surface and a spherical body are known from Kayali Dere, Karmir Blur, Oshakan, Erebuni, and Shikahogh. There are also examples of unknown provenance.49 A silver goblet with a narrow, long body is also known.50

Rhyton The next ritual item on the chancel table are the rhytons in the form of a ram or bull’s head. Such rhytons made of clay is known from Bastam.59

Single-handed goblets The participants of the ritual make pouring or take the wine with the use of these goblets. These have a flat or accentuated round bottom with a gradually widening neck. The handle has various ornamentation. This version of the goblet is not widely used.51

Paternoster Both the worshipers and the goddesses carry in their hands one or two strands of beads, each consisting of large, single-size beads. Large, round beads made of sard and transparent green glass are known from famous Urartian settlements and tombs (Karmir Blur,

42 Belli 2004: 137; Kozbe et al. 2001: Plate XVI/18. It has hieroglyphic signs under the lips in the form of circles and stylized jars; see Grigoryan 2018: 252 (HMA, pr. 2010-20). 43 Grigoryan 2018: 254 (HMA, pr. 2205-12, 2010-95); Piotrovsky 1969: 55; Yengibaryan 2019: Figure 9. 44 Merhav 1991a: Figure 7. 45 Hori et al. 1982: Plates I-II; Merhav 1991a: Figures 30a, 39a. 46 Kroll 1979: Table 2; Vanden Berghe and De Meyer 1983: no. 194 (height 36.4 cm); Yesayan et al. 1995: Plate VI, Figure 2 (height 30.5 cm, diameter 40.5 cm). 47 Hovhannissian 1973: Figures 2, 7, 49; Piotrovskij 1962: Plate XVI-XIX. 48 Hori et al. 1982: Plate III; Merhav 1991b: Figures 19, 20a, 21-22; 1991c: Figures 25-26; Yesayan et al. 1995: Plate X/2. 49 Burney 1966: Figure 15; Grigoryan 2018: 255 (HMA, pr. 2682-40); Yesayan and Kalantaryan 1988: Plate LVII/3; Belli 2004: 141. 50 Merhav 1991c: Figure 24. 51 Avetisyan 2001: Plate LXXVI/2, 7-8; Belli 2004: 140, 142; Grigoryan 2018: 236 (HMA, pr. 1814b, 2052/40); Martirosyan 1974: Figure 69.

Grigoryan 2018: 239, 241-245 (HMA, pr. 2010-93, 2000-5, 2199-5, 2201-5, 2203-30, 2204-68, 2452-199-5, 2580-129); Kleiss 1971: Abb. 21; Kroll 1976: Typ 46: Piotrovskij 1970: Figure 15; 2011: nos 613-614. 53 Grigoryan 2018: 159 (HMA, pr. 2010-65); Kleiss 1971: Abb. 23; Merhav 1991b. 18; 1991d; Piotrovskij 1970: Figure 61; Vanden Berghe and De Meyer 1983: nos 170, 172. 54 Grigoryan 2018: 261-265 (HMA, pr. 2206-4, 2203-8, 2050-78, 2203-1), 185 (HMA, pr. 2165-32; 2165-32). 55 Grigoryan 2018: 176-182, 184 (HMA, pr. 2010-7, 19, 22, 38, 47, 2010166); Piotrovskij 2011: 953-967. 56 Merhav 1991d: Figures 1/1-3, 2/3-4. 57 Belli 2004: 140; Khachatryan 1996: Figure 175; Kroll 1976: Typ 79; Martirosyan 1964: Plate XXIII/4; Piotrovskij 1959: 175-176; 1962: 109. 58 Grigoryan 2018: 246 (HMA, pr. 2010-162 28); Kellner 1979: Abb. 4-5. 59 Calmeyer 1979: Abb. 1, Tafel 46. 52

492

Nora Yengibaryan: Women in Urartian Rituals the heart and liver of the animal were often offered to the gods. The blood of the victim was poured into a vessel, on bread, or onto the ground.63

Metsamor, Argishtikhinili). Two agate beads bearing the inscriptions of Argishti I and Sarduri II are also known. According to the inscription, Argishti brought it from the land of Eriahe, both beads were dedicated to the goddess Varubani.60 It is worth mentioning that the neck of the statue of Warubane is decorated with beads. Beads are worn by different participants of the ritual procession.

One or more bread loavs were sacrificed during all the rituals. There are many types of ritual bread, which were of different colour (black, white, red), sizes, forms, and condition (warm, fresh, soft, sour). For some rituals, anthropomorphic and zoomorphic (also representing parts of the body: teeth, tongue, eyes) bread were made. The sacrifice of bread was conducted by breaking it into pieces.

The above-mentioned ritual vessels, with a few exceptions, are made of metal (silver, iron, bronze) and clay.

The Hittites used wine, beer, marnuwa (a type of beer), milk, honey, etc. for the ritual libation. According to the Hittite ritual, the first wheat, the first baked bread, a newborn lamb, or a bull were taken for sacrifice. The culmination of the ritual was the libation or drinking of wine or beer from the animal-shaped vessel (mainly in the form of the head) made of a precious metal BIBRU. The animal symbolized the original representation of the deity. Wine or other beverages replace the blood of that animal and bread represents the meat of the victim.64

Hittite sources Let’s try to investigate the material presented above from the chronologically earlier data of Hittite written sources. Numerous written sources have been preserved about Hittite rituals. The Hittite religious practice ncluded rituals of sacrifice, prayers, witchcraft, divination, and festivals (EZEN). A total of 170 rituals and festivals are mentioned in various texts, which were performed in a regular manner, suggesting the existence of a ritual calendar. The names of the rituals are known from the inscriptions, but not all of them are described.61 The main rituals were performed in spring and autumn connected to the cycles of nature and agricultural work. The main purpose of the rituals was to assure abundant crops. The most important festivals were the 37-day long EZENAN.TAḪ.ŠUMSAR and 43-day long EZEN nuntarriyašḫaš autumn and spring festivals. Some of the festivals include the New Year EZENpurulli/wurulli/ purulliyaš festival of the earth, EZENKI.LAM (translated as the gates/market squares festival), EZENGIŠ.GEŠTIN (tuḫšuwaš), the grape harvest festival, as well as rain (EZENḫewaš/ZU-UN-Nİ), spring (EZENḫamešḫanda/DIŠU/ TEŠİ), and winter (EZENgimmant) rituals. Numerous rituals and festivals dedicated to gods, sacrificing and libation to the gods, farming, harvesting, and agricultural tools are also known.62 Ritual festivals took place in the capital and other cities, where the cult centers of the gods were located, with the participation and guidance of the king, queen, and more rarely the heir. During the festivals, the king and queen visited important cities, sanctuaries, mountains and performed rituals near the stelae, sacred springs, and inside of the temples. There were also many local Hittite festivals and rituals performed without the royal couple, but rather by men having a right to perform such rituals. Animals were sacrificed during all the rituals: bull, cow, ram, sheep, lamb, goat, kid, and sometimes also pig, dog, and rarely horse. In addition to the head, shoulders, chest, and legs of animals, which were considered ritually pure,

A chancel table or tables were used for the sacrifice. The sacrifice to one or more gods could either be placed on one table or set on separate tables one for each god. Sacrifices were performed both in closed areas (temples, sanctuaries etc.) and outside (groves, mountains, in the vicinity of stelae and springs, and even on rooftops).65 The sacrifices were made to feed the gods or were eaten by participants. There was a list of participants of the ritual: society, craftsmen, servants, by which the sacrifice was distributed throughout the celebration.66 The king and the queen performed most rituals together. The role of the Hittite queen was important and was associated with worship. There were some rituals that the queen performed without the king, especially during large festivals such as nuntarriyašḫaš, the autumn festival. During the spring festival EZENAN. TAḪ.ŠUMSAR the queen sacrificed an antahsum plant, beer, bread with oil, grain, and lambs, whose liver and heart were grilled.67 Hittite sources mention a high-ranking priestess called ‘sister of the god.’ In the texts, she is mentioned after the king and the queen and performs the same rituals. In some cases she leads the rituals, and in particular, the ritual dedicated to the god Tetešapi. The latter was performed with a group of priestesses called ‘zintuḫi’. Ardzinba 2015a: 51, 175. Ardzinba 2015a: 167, 170. 65 Ardzinba 2015a: 169-171. 66 Ardzinba 2015a: 110. 67 Ardzinba 2015b: 95. 63

Salvini 2012: 253, D 8-1, 253. (diameter 2,2-2,4 cm, diameter 2,8-3,0 cm). 61 Demirel 2016: 21. 62 Demirel 2016: 23-27.

64

60

493

Systemizing the Past At the beginning of the ritual, the head of the girls was praying, singing, exclaiming, while others were just praying. The participants of the ritual, tel by the ‘sister of the God’, then walk to the city of Tawinia. The priestess carries with her the statue of the god Teteshapi. During the journey, the ‘zintuḫi’ girls sing and pray. In the forest, the statue is placed in a tent that served as a temple. The ‘sister of the god’ and ‘mother of the God’ priestesses enter the tent taking with them a goat. At the end of the ritual, the statue is returned to the temple, where the ‘sister of the god’ makes libation for the other gods both sitting or standing. Afterwards the bread is given, which she cuts and returns to the priestess, it is distributed among the participants. After she enters the ritual room of the temple, puts the symbols of power there, changes her clothes, and comes out. The ‘zintuḫi’ girls sing and exclaim all the time. In the end, the ‘sister of the god’ cleans the room and washes her hands. In this ritual, she acts as the main sacrificer. During other rituals, she performs various dances, both alone or with other participants.68

percussion instruments. In addition to music, dances were performed during the festivals, and rarely during sacrifices. These were mainly performed by ‘ḫapi people’, with the participation of ‘wolf people’ and ‘zintuḫi’ girls.72 Priests, doorkeepers and weavers were privileged in the most important cult centers of the Hittite state. Their profession was considered sacred. During one of the Hittite rituals, the king and queen are spinning. Two courtiers standing before the queen and the head of the weavers help them.73 Let’s try to explain the scenes of the above-mentioned Urartian belts with the help of the presented Hittite material. Bread and drinks are placed on the chancel tables. There are two types of depicted bread: roundshaped and crescent-shaped. The Urartians made libation during the rituals, about which there is written evidence. For that purpose they mainly used wine.74 The presence of different vessels in the depictions suggests the use of different drinks at the same time. Sargon II, king of Assyria (722-705 BC), mentions a huge metal container taken from the temple of Muṣaṣir, which was used to make libations in honour of the god Haldi, the head of the Urartian pantheon. Another large cauldron with capacity of 1500 liters is known from Karmir Blur. The Urartians could use other drinks besides wine and beer for making libations.75

In Mesopotamia, the ‘sister of the god’ acts as a highranking priestess who performs the ritual of ‘sacred marriage’ with the king.69 In Babylonia, the ‘bride of the god’ is established by an oracle. The chosen girl must be from the royal dynasty or from the elite. Even kings wanted to see their daughters in this role. According to researchers this ritual was performed also by Hittites. Through the ritual marriage, the king or the god ensures the welfare of the country, the fertility of the land, and the birth of animals.70

Urartian written sources mention agricultural festivals, which are mainly associated with grapes (prooning, proping up, harvest) accompanied by sacrifices.76 In one of the inscriptions of Güzak the construction of the Haldean gate and a fortress, and the establishment of a vineyard of king Minua are mentioned. Sacrifices for the harvest must be made in front of the Haldean gates and stela. When the grapes are gathered, a festival for Haldi and Warubani, his consort, should be conducted in front of the stela and the Haldean gates, during which new wine must be poured. According to Y. Grekyan, it is the juice squeezed from fresh grapes and not wine as it was previously belived.77

In the descriptions of festivals several priestesses or temple workers are mentioned: ‘mother of the god’, ‘wife of the Thunder god’, ‘wife of the god of wheat’, ‘priestess-wife’ of the goddess Hebat, ‘priestess-mother’ of the goddess Titiutti. The exact function of these priestesses is not described in the texts. The ‘zintuḫi’ priestesses are more often mentioned, who take part in ritual processions, ceremonies, festivals, pray, sing, dance, and exclaim.71 Among the women participating in Hittite rituals are musicians and weavers. Hittite musicians were temple servants and participated in all the rituals. The music was always played if the ritual was performed by the king, queen, or the heir to the throne. During the ritual processions, they accompanied the king and queen, and they also participated in the New Year festivals. The Hittites had stringed and percussion instruments. According to written sources, the female musicians played only the cymbal and tambourine

Having no cultic texts mentioning the participation of Urartian women in the rituals, let us interpret the Urartian iconographic evidences based on Hittite sources. In the depictions, the participants are in motion. If we fasten the belt, the movement becomes circular. Among the depicted noblewoman, one can conventionally separate a goddess, a queen, or a priestess. Only a living person can sit with knees bent Ardzinba 2015a: 135, 285. Ardzinba 2015a: 35, 295. Hmayakyan 1990: 76-77. 75 Piotrovskij 1950: 28. 76 Harutyunyan 2001: no. 3830-31. 77 Grekyan 2017: 41-50․ 72 73

Ardzinba 2015a: 292. Ardzinba 2015a: 292. 70 Vardiman 1990: 52-53. 71 Ardzinba 2015a: 288. 68

74

69

494

Nora Yengibaryan: Women in Urartian Rituals Demirel, S. 2017. An essay on Hittite cultic calendar based upon the festivals. Athens Journal of History 3/1: 21–32. Grekyan, Y. 2008. The woman in Urartu, the queens of Urartu. In A. Petrosyan A. (ed.), Gift From Heaven: Myth, Ritual and History, Studies in Honor of Sargis Harutyunyan on His 80th Birthday: 292-309. Yerevan, Publishing House ‘Gitutyun’, NAS RA (in Armenian). Grekyan, Y. 2017. The Urartian word for ‘wine’. Etchmiadzin 1: 41–50 (in Armenian). Grekyan, Y. 2018. Cultic structures in Biainili-Urartu. In Y. Grekyan, M. Badalyan, N. Tiratsyan and A. Petrosyan, Biainili-Urartu: Gods, Temples, Cults: 76–109. Yerevan, Institute of Archaeology and Ethnography Press (in Armenian). Gevorkyan, A. 1973. The Crafts and Mode of Life in Armenian Miniatures. Yerevan, Publishing House ‘Hayastan’ (in Armenian). Grigoryan, G. (ed.) 2018. Urartu, the Kingdom of Van. Finds from the Urartian Sites in Armenia (860-585 BC), from the Collections of the History Museum of Armenia. Yerevan, History Museum of Armenia Press (in Armenian). Harutyunyan (Arutjunjan), N. 2001. Corpus of Urartian Cuneiform Inscriptions. Yerevan, Publishing House ‘Gitutyun’, NAS RA (in Russian). Hmayakyan, S. 1990. The State Religion of the Van Kingdom. Yerevan, Press of the Academy of Sciences of Armenia (in Armenian) Hmayakyan, S. 2001. The women in Urartu. In D. Vardumyan and L. Vardanyan (eds), Problems of the Study of Family. Materials of the Republican Conference in Memory of Emma Karapetyan: 59–66. Yerevan, Publishing House ‘Gitutyun’, NAS RA (in Armenian). Hori, A., Miyashita, S. and Ishida, K. 1982. Studies in the Urartian objects from Japanese collections (1). Catalogue. Bulletin of the Ancient Orient Museum 4: 68–79. Hovhannissian, K. 1973. The Wall-Paintings of Erebooni. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Armenian, Russian and English). Karagyozyan, H. 1981. Urartian sources on the Armenian Highland and neighbouring states. In P. Hovhannisyan and G. Abrahamyan (eds), Christomatia of the History of the Armenian People: 37–93. Yerevan, Yerevan University Press (in Armenian). Karapetyan, I., Hmayakyan, S., Hakobyan, H., Vardanyan, R., Tumanyan, G., Avetisyan, P. and Tiratsyan, N. 1996. The funeral of a young Urartian woman. In Abstracts of the Reports of the 10th Scientific Session Dedicated to the Results of 1993-1995 Archeological Researches: 36–37. Yerevan, Institute of Archaeology and Ethnography Press (in Armenian). Kellner, H-J. 1979. Eisen in Urartu. Archäologische Mittailungen aus Iran 6: 151–156. Kellner, H-J. 1991a. Gϋrtelbleche aus Urartu (Prähistorische Bronzenefunde, Abteilung XII, Band 3). München, Fronz Stiner Verlag Stuttgart.

over the throne and her feet on the seat, so it cannot be a depiction of a goddess sitting on a throne.78 In the Hittite ritual, a high-ranking priestess was equated with the queen in the right to lead the ritual, and then there are priestesses among those sitting women. In our opinion, the woman walking under a sunshade is a high-ranking priestess, whom we see in the depictions of the ‘sacred marriage’ scene on the two belts. The depictions of the ritual bed stand out on one of the belts. In the scene with the chariot the high-ranking priestess walking under a sunshade is accompanied by women wearing headdresses, which by their status may be similar to the Hittite ‘mother of the God’, ‘priestess mother’ or ‘wife of the Thunder god’. The woman who differs from the other participants in the procession by wearing a headband and takes a small chancel table with bowls and bread, undoubtedly has a different status. The bearers of offerings, buckets, towels, shoes and other utensils for the rituals are the priestesses or temple servants who had a special status. These ritual participants: musicians, dancers, and weavers have their archetypes in Hittite written sources. By sacrificing bread made from the first crop, ritual animals, and freshly squeezed grape juice, the Urartians wanted to ensure prosperity and abundance for the country for which the Urartian priestesses served. Bibliography Ardzinba, V. 2015a. Rituals and Myths of Ancient Anatolia. In V. Chirikba and A. Vyatkin (eds), Civilaztion of Ancient Asia Minor, History and Culture, vol. 1: 48–266 Moscow and Sukhum, Publisher MBA (in Russian). Ardzinba, V. 2015b. Civilizacii Drevnej Maloj Azii (Civilaztions of Ancient Asia Minor). In V. Chirikba and A. Vyatkin (eds), Civilaztion of Ancient Asia Minor, History and Culture, vol. 1: 27–47. Moscow and Sukhum, Publisher MBA (in Russian). Avetisyan, H. 2001. Aragats. Yerevan, Yerevan University Press (in Armenian). Belli, O. (ed.) 2004. Urartu: Savaş ve Estetik /Urartu, War and Aesthetice. İstanbul, Yapı Kredi Yayınları. Calmeyer, P. 1979. Zum Tongefäß in Form eines Gazellenkorfes. In W. Kleiss (ed.), Bastam I, Ausgrabungen in den Urartäischen Anlagen 1972-75: 195–201. Berlin, Gebr. Mann Verlag. Caner, E. 1998. Bronzene Votivbleche von Giyimli (Archäologie in Iran und Turan 2). Rahden/Westf., Verlag Marie Leidorf GmbH. Çavuşoğlu, R. 2014. A unique female belt from the Van Museum of Archaeology. In A. Çilingiroğlu, K. Köroğlu, Z. Çulha and G. Öncü (eds), Urartians, A Civilization in the Eastern Anatolia, The Proceedings of the 1st International Symposium, 13-15 October, İstanbul: 171–182. İstanbul, Printing MAS Matbaacılık A.Ş. 78 Cf. with the statue of the Warubane goddess found in Darabey, or with the depiction of the gold medallion from Toprakkale.

495

Systemizing the Past Kellner, H-J. 1991b. Groping and dating bronz belts. In R. Merhav (ed.), Urartu, A Metalworking Centre in the First Millennium B.C.E.: 142–161. Jerusalem, Sabinsky Press. Khachatryan, H. 1996. Arménie, Trèsors de l`Arménie ancienne, Catalogie. Paris, Somogy éditions d’Art. Kleiss, W. 1971. Bericht über Erkundungsfahrten in Nordwest-Iran im Jahre 1970. Archäologische Mittailungen aus Iran 4: 51–111. Kroll, S. 1976. Keramik Urartäischer Festungen in Iran, Typologie der Keramik (Archäologische Mittailungen aus Iran, Ergänzungsband 2): 107–184. Berlin, Dietrich Reimar Verlag in Berlin. Kroll, S. 1979. Urartäische Keramik. In H.-J. Kellner (ed.) 1979. Urartu. Ein wiederentdeckter Rivale Assyriens: 62– 63. München, Ausstelung in Krefeld. Kozbe, G., Sağlamtemir H. and Čevik Ö. 2001. Pottery. In A. Çilingiroğlu and M. Salvini (eds), Ayanis I. Ten Year’s Excavations at Rusaḫinili Eiduru-kai 1989-1998 (Documenta Asiana 6): 85–153. Roma, CNR - Istituto per gli studi Micenei ed Egeo-Anatolici. Martirosyan, H. 1964. Armenia in the Era of Bronze and Early Iron Ages. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Russian). Martirosyan, H. 1974. Argishtikhinili (Archaeological Monuments of Armenia 8. Urartian Monuments I). Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Russian). Merhav, R. 1991a. Cauldrons and their stands. In R. Merhav (ed.), Urartu, A Metalworking Centre in the First Millennium B.C.E.: 226–243. Jerusalem, Sabinsky Press. Merhav, R. 1991b. Ceremonial backets. In R. Merhav (ed.), Urartu, A Metalworking Centre in the First Millennium B.C.E.: 214–220. Jerusalem, Sabinsky Press. Merhav, R. 1991c. Comments on silver vessels. In R. Merhav (ed.), Urartu, A Metalworking Centre in the First Millennium B.C.E.: 221–225. Jerusalem, Sabinsky Press. Merhav, R. 1991d. Bowls, gobblets, jugs and lamps. In R. Merhav (ed.), Urartu, A Metalworking Centre in the First Millennium B.C.E.: 200–213. Jerusalem, Sabinsky Press. Piotrovskij, B. 1950. Karmir Blur I. The Results of the Excavations of 1939-1949. Yerevan, Press of the Avademy of Sciences of the Armenian SSR (in Russian). Piotrovskij, B. 1959. The city of the god Teisheba (Karmir-Bloor). Soviet Archaeology 2: 169–186 (in Russian).

Piotrovskij, B. 1962. The Art of Urartu in the 8th-6th centuries BC. Leningrad, State Hermitage Publishing House (in Russian). Piotrovskij (Piotrovsky), B. 1969. Urartu. Geneva, Nagel. Piotrovskij, B. 1970. Karmir Blur (Аlbum). Leningrad, Avrora (in Russian). Piotrovskij, B. 2011. The History and Culture of Urartu. Catalog. St. Petersburg, Art of Russia (in Russian). Seidl, U. 2004. Bronzekunst Urartus. Mainz am Rhein, Verlag Philipp von Zabern. Seidl, U. 2012. Rattling and clapping Urartian girls: Idiophones in Urartu. In P. Avetisyan and A. Bobokhyan (eds), Archaeology of Armenia in Regional Context: 163– 169. Yerevan, Publishing House ‘Gitutyun’, NAS RA. Salvini, M. 2008. Corpus dei testi urartei, vol. I. Le iscrizioni su pietra e roccia: I Testi (Documenta Asiana 8/1). Roma, CNR Istituto di studi sulle civiltà dell’Egeo e del Vicino Oriente. Salvini, M. 2012. Corpus dei testi urartei, vol. IV. Iscrizioni su bronzi, argilla e altri supporti, nuove iscrizioni su pietra, paleografia generale (Documenta Asiana 8/4). Roma, CNR Istituto di studi sulle civiltà dell’Egeo e del Vicino Oriente. Vanden Berghe L. and De Meyer, L. 1982. Urartu: een vergeten cultuur uit het bergland Armenië (Katalog der Ausstellung Gent 27). Gent, Centrum voor Kunst en Cultuur Sint-Pieterabdij. Vardiman, E. 1990. Woman in the Ancient World. Moscow, Nauka. Yengibaryan, N. 2012. An Urartian medalion from Sodk. Aramazd: Armenian Journal of Near Eastern Studies 7/1: 80–85. Yengibaryan, N. 2019. Middle Iron Age tombs from Noratus (Sevan region, Armenia). Aramazd: Armenian Journal of Near Eastern Studies 13/1: 1–43. Yesayan, S. and Kalantaryan, A. 1988. Oshakan I. The Main Results of the Excavations of 1971-1983. Yerevan, Press of the Academy of Sciences of the Armenian SSR (in Russian). Yesayan, S., Biyagov, L., Hmayakyan, S. and Kanetzyan, A. 1995. The Urartian Tomb of Yerevan, 2 (Archaeological Monuments of Armenia, Urartian Monuments III): 55-79. Yerevan, ‘Armenian Army’ Military Science Magazine Press (in Armenian). Ziffer, I. 2002. Four new belts from the Land of Ararat and the feast of the women in Esther 1:9. In S. Parpola and R. M. Whiting (eds), Sex and Gender in the Ancient Near East: Proceedings of the 47th Rencontre Assyriologique Internationale, Helsinki, July 2-6, 2001: 645–656. Helsinki, The Neo-Assyrian Text Corpus Project.

496