Archa¨Ologische Mitteilungen Aus Iran Und Turan

DeutschesArcha¨ologisches Institut Eurasien-Abteilung Außenstelle Teheran

Sonderdruck aus:

Archa¨ologische Mitteilungen aus Iran und Turan

BAND 46 * 2014

DIETRICH REIMER VERLAG * BERLIN Deutsches Archa¨ologisches Institut Eurasien-Abteilung Außenstelle Teheran

Archa¨ologische Mitteilungen aus Iran und Turan

Band 46 * 2014

DIETRICH REIMER VERLAG * BERLIN IV + 342 Seiten mit 201 Abbildungen, 39 Tabellen und 5 Karten

Herausgeber: Svend Hansen und Mayke Wagner Redaktion: Judith Thomalsky und Nikolaus Boroffka WissenschaftlicherBeirat: Abbas Alizadeh (Chicago) David Braund (Exeter) Henri-PaulFrancfort (Nanterre) Ernie Haerinck(Ghent) StefanR.Hauser(Konstanz) Lorenz Korn (Bamberg) Stephan Kroll (Mu¨nchen) Michael Pfrommer (Trier) Susan Pollock (Berlin) Robert Rollinger (Innsbruck) Miroslav Salvini (Roma)

Mitglieder des Deutschen Archa¨ologischen Instituts und Studenten der Altertumswissenschaften ko¨nnen die Archa¨ologischen Mitteilungenaus Iran und Turan zum Vorzugspreis von 53,20 a zuzu¨glich Versandkosten abonnieren. Bestellungen sind an die Schriftleitung zu richten.Studenten werden um Vorlage einer Studienbescheinigung gebeten. Die Beendigung des Studiumsist unverzu¨glich mitzuteilen.

Redaktionsschluss ist der 31.M¨arz fu¨rden im folgendenJahr erscheinenden Band. Bei der Abfassungder Manuskripte sind die ,,Richtlinienfu¨rVero¨ffentlichungen der Außenstelle Teheran der Eurasien-Abteilung des Deutschen Archa¨ologischen Instituts‘‘ zu beachten, die unter http://www.dainst.org/index.php?id¼7490 in ihrer jeweils aktuellgu¨ltigen Formaufgerufen werden ko¨nnen.

Die Redaktion beha¨lt sich vor, Manuskripte zu ku¨rzen.

° 2016 by DeutschesArcha¨ologisches Institut, Eurasien-Abteilung ISSN 1434-2758 Redaktion: Deutsches Archa¨ologisches Institut, Eurasien-Abteilung, Im Dol 2–6,D-14195 Berlin Satz,Druck undBindung:Beltz BadLangensalza GmbH,Neusta¨dter Straße 1–4,D-99947 BadLangensalza Kommissionsvertrieb: Dietrich Reimer Verlag GmbH, Berliner Straße 53,D-10713 Berlin Inhalt III

Inhalt

Aufsa¨tze

HamedVahdati Nasab und GeoffreyA.Clark,The UpperPaleolithic of the Iranian Central Desert:the Delazian Site – aCase Study ...... 1 OmranGharasian,LeilaPapolis und Hamide Fakhr-e Ghaemi,Qaleh Khan aSite in Northern Khorassan and the Neolithic of North Eastern Iranian Plateau ...... 21 MohammadEsmaeilEsmaeili Jelodar und SaeedZolghad,Central Zagros, Highland Fars, and Lowland, Susiana Sphere of Interactioninthe 5th Millennium BC.: Evidence from salvage excavation at Haji Jalil 2,Kuhrang, Iran ...... 51 Massimo Vidale,DeboraVendemi undEdoardo Loliva, Uncertainty and errors in the Painted Buff Ware of Shahr-e Sukhte (Sistan,Iran) ...... 71 KaiKaniuth, Die Bestattungen des Hu¨gels 4 von Dz˘arkutan...... 95 Mutin-Razzakov, Cultural contacts across the Hindu Kush in the early BronzeAge. Additional insights from Sarazm – Soundings 11–11A(Tajikistan) ...... 123 Ruben Badalyan, Adam T. Smith, IanLindsay,Armine Harutyunyan,Alan Greene, Maureen Marshall, Belinda Monahan und RomanHovsepyan,mit Beitra¨gen von Kha- chatur Meliksetian, Ernst Pernicka und Samuel Haroutunian, APreliminary Report on the 2008, 2010,and 2011 Investigations of Project ArAGATS on the Tsaghkahovit Plain, Republic of Armenia . 149 AlirezaAskari Chaverdi,Pierfrancesco Callieriund Emad Matin, Tol-eAjori: a Monumental Gate of the Early Achaemenian period in the Persepolis Area. The 2014 excavation season of the Iranian-Italian project ‘‘From Palace to Town’’, with an appendixbyGian Pietro Basel- lo, Afragment of another inscribedglazed brick from Tol-e Ajori...... 223 MichaelFedorov, Money circulation in Central Asia under T¯ımu¯rand T¯ımu¯rids ...... 255 AlexandreTourovetz, The search for abetter organisation of the space and its use: apossible cause to explain the outbreak of the Achaemenian Architecture ...... 281 Mernoush Soroush,The Misrof‘Askar Mokram: Preliminary Report and Framework for Future Research ...... ˙ ...... 299 Yaghesh Kazemi, Solar Considerations in Planning of Three Circular Cities in Ancient Persia ... 321

Buchbesprechungen:

W. Roberts/C. P. Thornton (eds.), Archaeometallurgy in global perspective. Methods and syntheses (Springer, Heidelberg, New York 2014)(B. Helwing) ...... 331

Nachruf

Yuriy Fedorovich Buryakov. 1934–2015 (A. E. Berdimurodov) ...... 335 Vadim Nikolaevich Yagodin. 1932–2015 (A. D. Iskanderova) ...... 339 IV Inhalt

Table of content

Articles

HamedVahdati Nasaband GeoffreyA.Clark,The UpperPaleolithic of the Iranian Central Desert: the Delazian Site – aCase Study ...... 1 OmranGharasian,Leila Papolis and Hamide Fakhr-e Ghaemi, Qaleh Khan aSite in Northern Khorassanand the Neolithic of North Eastern Iranian Plateau ...... 21 MohammadEsmaeil Esmaeili Jelodarand SaeedZolghad,CentralZagros, Highland Fars, and Lowland, Susiana Sphere of Interaction in the 5th Millennium BC.: Evidence from salvage excavation at Haji Jalil 2,Kuhrang, Iran ...... 51 Massimo Vidale,DeboraVendemi and Edoardo Loliva,Uncertainty and errors in the Painted Buff Ware of Shahr-eSukhte (Sistan, Iran) ...... 71 KaiKaniuth, Die Bestattungen des Hu¨gels 4 von Dz˘arkutan ...... 95 Mutin-Razzakov, Cultural contacts across the Hindu Kush in the early Bronze Age. Additional insights from Sarazm – Soundings 11–11A(Tajikistan) ...... 123 Ruben Badalyan,AdamT.Smith,Ian Lindsay,Armine Harutyunyan, Alan Greene, MaureenMarshall, BelindaMonahanund RomanHovsepyan,with contributions by Khachatur Meliksetian, Ernst Pernicka and Samuel Haroutunian, APreliminary Report on the 2008, 2010,and 2011 InvestigationsofProject ArAGATSonthe Tsaghkahovit Plain, Republic of Armenia . 149 AlirezaAskariChaverdi,PierfrancescoCallieri andEmadMatin,Tol-e Ajori: a Monumental Gate of the Early Achaemenian periodinthe Persepolis Area. The 2014 excavation season of the Iranian-Italian project ‘‘From Palace to Town’’, with an appendix by Gian Pietro Basel- lo, Afragment of another inscribed glazed brick from Tol-e Ajori...... 223 MichaelFedorov, Money circulationinCentral Asia underT¯ımu¯rand T¯ımu¯rids ...... 255 AlexandreTourovetz, The search for abetter organisation of the space and its use: apossible cause to explain theoutbreakofthe Achaemenian Architecture ...... 281 MernoushSoroush,The Misrof‘Askar Mokram: Preliminary Report and Framework for Future Research ...... ˙ ...... 299 Yaghesh Kazemi, SolarConsiderations in Planning of Three Circular CitiesinAncient Persia... 321

Book reviews

W. Roberts/C. P. Thornton (eds.), Archaeometallurgy in global perspective. Methods and syntheses (Springer, Heidelberg, New York 2014)(B. Helwing)...... 331

Orbituary

Yuriy FedorovichBuryakov. 1934–2015 (A. E. Berdimurodov) ...... 335 Vadim Nikolaevich Yagodin. 1932–2015 (A. D. Iskanderova) ...... 339 Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS on the Tsaghkahovit Plain, Republic of Armenia

By Ruben Badalyan, Adam T. Smith,Ian Lindsay, Armine Harutyunyan, Alan Greene,Maureen Marshall, BelindaMonahan, Roman Hovsepyan

With contributions by Khachatur Meliksetian,Ernst Pernicka and Samuel Haroutunian

Keywords: Armenia, Tsaghkahovit Plain, Early Bronze Age, Kura-Araxes, Late Bronze Age, Gegharot, Tsaghkahovit, Aragatsi Berd ˚ºþ÷åâßå æºîâà: ÀðìåíŁ', ÖàıŒàîâŁòæŒà'ðàâíŁíà, ðàííŁØ ÆðîíçîâßØ âåŒ, Œóðî-àðàŒæ, ïîçäíŁØ ÆðîíçîâßØ âåŒ, ˆåıàðîò, ÖàıŒàîâŁò, ÀðàªàöŁ Æåðä

Fig. 1 Archaeologicalsites of the Tsaghkahovit Plain and surrounding areas

In 1998 the joint Armenian-American Project for the detail the processes that drove major historical Archaeology and Geography of Ancient Transcauca- transformationsover the longue dure´e.2 sian Societies(Project ArAGATS) began asystematic The Tsaghkahovit Plain (2000 ma.s.l.) is a investigation of Bronze and Iron Age sites in the 10–12 km wide upland basin bounded by the Tsaghkahovit Plain of central Armenia.1 Our goal northern slope of Mt. Aragats (4090 m), the south- was, and remains, to provide an encompassing account of the rangeofsocial, economic, and political practices that characterizedlife in the region and 2 See the Project ArAGATS website (http://aragats.arts.cornell.edu) for overviews of each season and all of ouranalyticaland field research.This contributionwas sent to AMIT during the summer of 2012.Since that time, Project ArAGATS has continuedits in- 1 Avetisyan et al. 2000;Badalyan et al. 2008;Smith et al. 2004; vestigations with field seasonsin2013, 2014, 2015,and 2016. Smith et al. 2009. The resultsofthat work will be reported separately. 150 Ruben Badalyan et al.

western slopes of the Pambak range, and the east 2008 and 2011 have provided critical new informa- flank of Mt. Kolgat. The region is crossed by akey tion on both of these central concerns.5 north-south transit route that connects the Ararat Plain with points north and hosts the headwaters of the Kasakh river, akey tributary of the Araxes. Excavations at Gegharot Over the course of three field seasons in 2008, 2010,and 2011,Project ArAGATS completed Excavations were conducted at the settlement of anew phase in its investigations.3 Building upon Gegharot (Fig. 2) in three general areas: the west continuing work at the sites of Gegharot and Tsagh- terrace (operations T-2E, T-19,T-31), on the west kahovit, we conductedadditional excavations at citadel(operations T-21,T-23,T-24,T-26,T-28,T- the BronzeAge settlement at Aragatsi Berd, aLate 30)and on the east citadel (operations T-20,T-22, Bronze Age Cemetery at Tsaghkahovit (Burial Clus- T-27,T-32,T-33,T-34). On the western terrace, we ter 12), and the medieval ijevanatun (caravanserai) completed the excavation of the T-2Esanctuary at Arai.4 In addition, topogeodetic work was con- (henceforth, sanctuary 1)described in our previous ducted at the large Bronze and Iron Age settlement preliminary report,6 uncovering asmall 12.5 m2 at AparaniBerd and the Iron Age settlement at Ni- area in the northeast corner of the building. In ad- gavan as afoundationfor future plannedwork in dition,wecompleted the excavation of T-19,a the AparanPlain (Fig. 1).Taken together, our inves- 110 m2 operation near the center of the terrace, tigations providenew insights into adeep, but dis- and initiated the excavation of T-31,a50 m2 continuous, historical sequence of human occupa- trench, on the northern end of the terrace.7 On the tions: the EarlyBronze Age (Gegharot, Aragatsi west citadel, we excavatedtwo adjacent trenches Berd), the Late BronzeAge (Gegharot, Tsaghkaho- (T-21,T-23)with atotal area of 104.2 m2,and three vit, Aragatsi Berd), the Iron 3 (Achaemenid) period 5 Â 10 moperations T-26,T-28,and T-30 which (Tsaghkahovit), and the Middle Ages (Arai). continued aline of excavationsalong the western In this article, we provide an extensive preli- rim initiated in 2002.Inaddition, operation T-24 minary report on the work related to the Bronze was excavated as asingle 5 Â 5 moperation near Ageoccupations of the regionatGegharot, Aragatsi the northern edge of the summit. On the east cita- Berd, and Tsaghkahovit. The Iron Age occupation of del, we completed the excavation of T-20,anop- Tsaghkahovit and excavationsatthe Medieval site eration initiated in 2006 and extended investiga- of Arai are described in separate communications tions in the area with five additional operations (T- (Franklin 2014;Khatchadourian 2014). Our investi- 22,T-27,T-32–34), covering atotal area of gations into the Bronze Age Tsaghkahovit Plain 271.2 m2.Lastly, in concert with aremote sensing have focused on two key anthropological problems: project at the site,8 we opened asmall 2 Â 3 mex- the genesis, growth, and abandonment of the Early ploratory sounding (T-29)north of the main hill. Bronze Age Kura-Araxes village and the formation, The stratigraphy in all excavation units was expansion,and destructionofthe region’s earliest generally similar, with evidence of two primary cul- complex societies during the Late BronzeAge. The tural levels dating to discrete episodes of settle- data from the field seasonsconducted between ment during the Early and Late Bronze Ages. Both of these levels were in turn composed of two construction horizons.However,the preservation and 3 Funding for Project ArAGATS was providedbythe NationalSci- robustnessofeach layer varies considerably in dif- ence Foundation(grant nos. BCS-0964154 and BCS-1147577), ferent parts of the settlement, due to asteep slope, the University of ChicagoDepartment of AnthropologyLichtstern the geological characteristics of the outcrop (a Fund (2008), and the IAE NAS RA (2011). folded granodiorite and quartz diorite exposed to 4 In addition to the authors of this article, the members of the 2008–2011 expeditions included:architects L.Ter-Minasyan severe erosion), and the repeated episodes of de- (2008, 2010–2011,IAE NAS)and H. Sargsyan (2008, 2010,Yere- struction and reconstruction that took placeduring van State University), D. Narimanishvili (2010,Tbilisi State Univer- the life of the settlement. sity, Georgia), V. Vardazaryan (2011,Yerevan State University), E. Fagan (2008, 2010,University of Chicago), K. Franklin (2008, 2011,University of Chicago), H. Chazin(2010–2011,University of Chicago), K. Kearns (2010–2011,Cornell University), J. Leon (2010–2011,Cornell University), C. Wiktorowicz (2010–2011,Pur- 5 due University), J. Nabel (2011,Cornell University). Geodetic sur- The Project ArAGATSexcavation database is available online at veys at AparaniBerd and Nigavan were conducted by Smbat and http://aragats.arts.cornell.edu.Asandbox read-only version, Vahe Davtyan. Restoration of ceramic materials describedhere available to the public, contains extensive descriptionsofeach was conducted by L. Manukyan and A. Ayvazyan (IAE NAS). Con- locus and find. 6 servation of the metal artifactswas completed by L. Atoyants Badalyan et al. 2008. (HMA). Drawings of ArAGATS materialswere provided by H. Sarg- 7 Because the excavation of T31 was not completed duringthe syan (Fig. 3, 5, 9, 13, 14, 15, 22–1,25, 32,1–6), N. Mkhitaryan (Fig. 2011 field season, we will leave areport on it for afuture publi- 6, 7)(IAE NAS), and S.Haroutunian (Fig. 35a, b, c) while artifact cation. photos were done by V. Hakobyan (IAE NAS). 8 Lindsay et al. 2010. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 151

Fig. 2 Map of Gegharot 152 Ruben Badalyan et al.

The Early Bronze Age (Kura-Araxes) tools includingaquern, acylindrical pestle, adis- settlement coid mortar, and abifaced hand axe that perhaps served as an adze or hoe (Fig. 3).Further to the north, we encounteredanirregular triangular stone As our prior excavationshad revealed, the Early construction (W1905, 1.6 Â 2.0 m) composed on a Bronze Age (EB) layers at Gegharot are divided into single course of stones (locus 123)whose function two horizons, each containingahomogenous and remains unclear. distinct Kura-Araxesceramic assemblage. The lower Just to the north of the locus 112 floor was a horizon consists of ceramics of the ‘‘Elar-Aragats’’ stonemasonry wall fragment(W1906, 1 mlong) type while the upper horizon is defined by ‘‘Karnut- and an associatedsection of an EB floor (loci 509– Shengavit’’ type wares. Aseries of calibrated radio- 511)just over 1 m2,which was articulated with the carbon determinations suggests that the lowerhori- wall by an adjoining two part hearth with burnt zon dates to ca. 3350–2900 BC while the upper edges(0.5mindiameter), reinforced around the dates to ca. 2900–2500 BC.9 perimeter by occasionalstones. The center of the The specific manifestationsofthis basic strati- hearth was filled with ablack,organic-rich soil that graphic schema are distinct in different parts of the contained arelatively large collection of botanical settlement. AlthoughEBmaterials are presentto material, including barley, wheat, and undefined varying degrees in all Gegharot operations, the pre- grains. Within and under the hearth was an in situ servation of EB layersonthe citadel is primarily fragmented vessel. due to the scale andcharacter of Late Bronze Age All of the occupation surfaces defined in op- (LB) construction. Because LB buildings were eration T-19 can be assigned to Gegharot’s upper erected at some distance (approximately 3.5– ‘‘Karnut-Shengavit’’ horizon. Within the west ter- 5.5 m) from the circumferential walls that defined race, anumber of excavated contexts are stratigra- the citadeland terrace,EBlevels in that buffer zone phically related to this occupationlevel, including remained undisturbed, preserving earlierEBcon- an area in T-16 (bounded by walls W205 and structions; underneath LB occupation areas, EB re- W1601), and the upper EB layer uncoveredinT-2E mains have survived only as small, fragmentary (W206–207). It is possible that T-19’s upper EB lenses. horizon covers an earlierEBIlayer, as was the case On the west terrace, excavations in T-19 un- in T-16 and T-2E. This possibility is indicated by covered an intactEBlayer in the western portion of both afew fragments of pottery decorated with the operation. Here, alarge LB sub-terracing wall dimple ornaments (‘‘Elar-Aragats’’ type) such as (locus 119/126;see below) dividedthe excavated thoseknown from T-31 and possibly by the radio- area into two parts: an area east of the wall repre- carbon determination from T-19 locus 112 AA- sented by LB cultural deposits and an area to the 92844 (see discussion below). west with an undisturbed EB occupationlayer. This While the 2008–2011 excavations on Gegha- layer was represented by aseries of stratified levels rot’s west terrace uncovered only the upperlayer of of diagnostic materials.The upper level was repre- EB occupation at the site, excavations on the west sented by aconcentration of EB ceramic fragments citadel provided acomplete stratigraphic picture of (loci 108, 504, 507), overlapping wall fragments, EB cultural layers (Fig. 4).11 On the western edge of and limitedareas of apreserved floor. The first wall the citadel, the preservation of EB levels was pri- (W1904 locus 122), composed of two rowsof marily due to the buffer area that later LB inhabi- stones (1.0 mwide, 3.4 mlong) was oriented per- tants maintained between theperimeter wall and pendicular to the slope; only one course of stones their primaryoccupation areas.Inthis intervening was preserved on the lowerreaches of the wall space (the western portions of operations T-23,T- (18 cm in height) while threecourses remained in 26,T-28,and T-30), two EB horizons were left rela- the upper (45 cm in height). tively undisturbed. To the north of W1904 was apreserved sec- In the easternportionofoperations T-21 and tion of an associated floor (locus 112, 2.2 Â 3.5 m) T-26,EBlevels survived only as lenses of fragment- which contained in situ materials, most notably a ed ceramic vesselslying directly under LB depos- fragmented ceramic vessel of the ‘‘Karnut-Shenga- its. In the eastern half of T-21,thin (10–16 cm) vit’’ type,ahorseshoe-shaped andironfragment charredblack and orange lenses of EB cultural with sheepprotome (similar to the discovery of pit layers were preserved (locus 35: 2.50 Â 0.50 m 6.2 of T-2Elocus 624),10 and acollectionofstone and locus 39: 1.60 Â 0.90 m). In the northern sec-

9 Badalyan et al. 2008;Badalyan et al. 2010; `àäàº'í 2011. 11 The lower ‘‘Elar-Aragats’’ horizon was documented on the west 10 `àäàº'í/ÑìŁò 2008,table IV, 9;Badalyan et al. 2008, 56, terrace during our earlier investigationsinoperation T-2E(Ba- Fig.13. dalyan et al. 2008). Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 153

Fig. 3 Gegharot.Materials from the Upper EB level of T-19. 1–7 ceramic; 8–9, 11 stone; 10 fragment of ceramic andiron tion of the operation, the EB layer extended under- specially noted. It is avessel in the form of the neath aLBwall (W2104). This layer contained a classic ‘‘Karnut-Shengavit’’ cup, the lowerbody of large number of richly ornamented pottery of the which,however, has an intricate profile that may ‘‘Karnut-Shengavit’’ type, comprising at least 13 have led to an open base (Fig. 5,5);afragment vessels (locus 35) (Fig. 5),one of which must be with asimilarprofile was discovered at the settle- 154 Ruben Badalyan et al.

Fig. 4 ment of Karnutinthe adjacent Shirak Plain.12 Per- that had been restored during the EBA with bitu- Gegharot. Plan of west haps these vessels are fragments of vases, censers, men (Figs. 7,8).16 citadel operations T-21, T-23,T-26,T-28,and or so-calleddouble-vessels – jars with arim The lower(western) half of T-21 was greatly T-30.EBconstructions shaped like abowl or goblet similartothose from eroded by the encroaching slope.NoEBlevel was shaded in gray Serkertepe13 in north-east Azerbaijan and Torpakh- preserved, only ahomogenous assemblage of EB kala14 in Dagestan (in the latter example, however, ceramics within deposits of fine sand formed by the the upper cup was not connected to the lower destruction of the granite substrate. Within this opening). Aradiocarbon date (Bln-5374; 4255 Æ 15–20 cm thick sandy deposit (locus 36), EB cera- 39 BP) from Torpakh-kala points to adate contem- mic fragments covered an area of 2.2 Â 4.3 m. No porary with the ‘‘Karnut-Shengavit’’ (upper EB) constructions or floors were found in association layers at Gegharot.15 An equally rich lens of the with this assemblage. However, its stratigraphicpo- upper EB floor was found in locus 39 (2.5 mnorth- sition corresponds to the floor in operation T-18 west of locus 35). Amongst the dense assemblage (locus 29,excavated in 2006), which included two of ‘‘Karnut-Shengavit’’ pottery (Fig. 6) was an ar- large in situ EB vessels, set into the floor. senical bronzedouble-pointed awl (length = On the lower, western,portions of T-26 and T- 2.3 cm, see discussionofArAGATS metals below). 28,weuncovered two EB occupation levels.The The next fragment of the ‘‘Karnut-Shengavit’’ upper horizon (elev. 2292.21–2291.88 ma.s.l.), horizon was uncoveredinoperation T-26,separated which corresponds stratigraphically to the floors dis- from the previous floors by the large rock outcrop cussed above, included arectangular room in the which occupied much of T-23 (locus 7, 6.5 mnorth- western portion of T-26 bounded by stone masonry northwest of T-21 locus 39). Here, aburnt orange walls (W2601 and W2602). The north wall (W2602,lo- clay lens 0.6 Â 1.5 minarea was preserved under the LB floor along the southern edge of the excava- 16 Analysis of the bitumen at the Institute of Geological Sciences tion unit in ashallow (30 cm), possibly artificial, de- of NAS RA (P.Tozalakyan, R.Gazumyan) confirmedthat the sub- pression in the bedrock. Among the ceramic inven- stance was ametamorphic bitumenofthe asphalt class. The tory (Fig. 7) it is important to note aminiature cup use of bitumen for the restoration of small vessels (filling cracks, repairing edges) and joining separatelymanufactured parts (e.g., the neck and body of medium-sized jars) has been repeatedly observedonGegharot materials (e.g., T-2clocus 10/T-2Elocus 110;T-2E, loci 538 and 555;T-20 loci 9–10). A number of examples (e.g., Gegharot T-19 locus 501;Aragatsi 12 Berd AB4 locus 7)show that the edge of the body and necks Storage of the Museum of the History of Armeniano. 3017/31. were corrugatedinorder to create abetter surface for joining. 13 Ìóæàåâ 2006, 44,Table XXIII,2. The use of bitumenhas also been recordedinthe Kura-Araxes 14 ˆàäæŁåâ/Ìàªîìåäîâ 2008, 280 Fig. 4,7–8. ceramics of settlements at Mets Sepasar, Tsaghkasar,Teghout I 15 ˆàäæŁåâ/Ìàªîìåäîâ 2008, 281. (Dzori Gex), and the tomb at Ardasubani. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 155

Fig. 5 Gegharot. EB ceramics from operation T-21 locus 35 cus 31)was 2.6 mlong and 0.2 mwide, preserved in courses (0.35–0.55 m). The southern continuation 1–3 courses of stones to aheight of 0.30–0.37 m. of W2601, 2 minlength, was highly disturbed, per- The eastern wall of the room (W2601’s north- haps due to an episodeofcollapse and partial re- ern half, locus 30)was constructed of regular stone construction. Interestingly, 2.2 meters south of the masonry, 0.26 mwide in 1–2 lines of stones and corner of W2601 and W2602 we uncovered asmall 2.2 minlength, preserved to aheight of 3–5 perpendicular line of stones, 1,45 minlength, sug- 156 Ruben Badalyan et al.

Fig. 6 gestive of apartition. The subsequent discovery un- low pit, encircled by small stones and encircled by Gegharot. EB ceramics derneath these stones of ahearth (locus 8)sug- aclay ring. The floor of the room (loci 2, 5)in- from T-21 locus 39 gests this rudimentary construction was associated cluded EB ceramics, aflint sickle blade, and frag- with arebuildingepisode. The hearth was made ments of obsidian. The general lack of in situ mate- from afragment of alarge EB vessel set in ashal- rials suggests that the building was abandoned. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 157

Near the center of the adjacent operation, T- littered with an unusually large number of animal Fig. 7 28,weencountered the remnants of the same bones. Asection of the floor immediately adjacent Gegharot.EBceramics from T-26 locus 7 upper EB layer with clay floors, bounded to the to wall W2806 (locus 12)was very densely packed east and southeast by wall W2806 (locus 29). This and contained only very small pottery sherds and curved stone masonry construction, 3.2 mlong pieces of highly fragmented, calcified bone. Some and 0.5 mwide, was built of fairlylarge stones17 areas included multiple floor layers that were most and was preserved in 2 courses to aheight of visible on the southernedge of the operation. 0.57 m. Ayellow clay floor (locus 12, 21) 3.5 Â Near the center of the floor, covered by the last 3.9 mand 5–12 cm thick was recorded at an ab- episode of re-flooring, was asmall shallow hearth solute elevation of 2292.15 ma.s.l. In addition to (locus 22) 35 Â 26 cm, surrounded by acircle of ceramic fragments and stone tools, the floor was scorched earth.Below this upperfloor was asan- dy erosionaldeposit containing EB ceramics and a large number of animal bones(locus 20). At 17 Averagestone dimensions: 25/40/55 Â 35/50/75 Â 25/30 cm. 0.5 meters parallel to the western end of W2810, 158 Ruben Badalyan et al.

Fig. 8 was aline of large stones (locus 33).18 The wall lel rows of stones. The northern boundary of the Gegharot. Stratigraphic blocks extending perpendicular to wall W2812’s room was defined by wall W3009 (2 mlong, 0.45 section from operation T-28,south baulk lower course were separated from it by athin mwide). In the northeast corner of the room was layer of soil. an open bell-shaped pit (locus 28), 0.50 mindiam- The excavationofT-30 revealed asingle large eter at the mouth. The pit was dug into the bed- EB room in the western half of the trenchthat was rock and the opening was defined by arim of clay stratigraphically and planigraphically correlated with 10–12 cm wide. The walls of the pit were lined with the T-28 structures discussed above. The room is small stones in some places. The pit was excavated defined on its western edge by acomplex stone to adepth of 2.8 mbelow the EB surface, but the masonry wall, W3006 (length 3.8 m). The central bottom was not reached prior to the end of the portion of the wall (0.3–0.4 mwide) is composed 2011 field season. At adepth of 2.4 mweencoun- of anumber of largestones, while the northern tered aburnt layer which included piecesofchar- (0.5–0.6 mwide) and southern (0.5 mwide) sec- coal.Itisnot clear at present whether this layer tions are composed of small rounded cobbles in was deposited as part of the original operation of two or three courses. The wall delimits aclay floor the pit or if it was the resultofasubsequent de- that was litteredwith EB ceramics (loci 22, 23,elev. positional event. 2292.20–2291.94 ma.s.l.). Setinto the floor, Half ameter west of the pit was ahearth-fur- 20 cm east of W3006 was asmall (0.45 mwide) nace (locus 29;elev. 2292.01 ma.s.l.). It is possi- cobble ‘‘platform’’ (W3008)built of two-three paral- ble that this oval (0.5 Â 0.9 m) featurehad an arched roof, as evidenced by the collapsedremains of baked clay within the chamber found lying with 18 Average stone size: 25/45/60 Â 35/55/60 cm. theirinner sides face down. The collapsed pieces of Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 159

Fig. 9 Gegharot.EBmaterials from operation T-20 locus 104. – 1–7 ceramic; 8 stone 160 Ruben Badalyan et al.

Fig. 10 Gegharot. Photos of EB unique finds. – 1 EB plaque from T-22 locus 115 (dimensions: length 44 cm bottom, 49.2 cm top; height 8.5 cm middle, 10.2 cm sides; – 2 EB ‘‘table’’ from T-30 locus 22.a profile; bdetail of ornamentation (without scale); – 3 EB ceramic vessel from T-22 locus 9–10 made in two parts andjoined with bitumen; – 4 EB ceramic vessel from T-22 locus 9–10

baked clay were marked with traces of finger im- the construction were 2 cm thick. Near the opening pressions, as was the trough-shaped bottom of the was found acylindrical tripod cup, decorated with furnace. The northern end of the furnace narrows at geometric patterns. The cup is morphologically simi- the neck to adiameter of 18 cm. The clay walls of lar to an unornamentedvessel found in the ‘‘Kar- Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 161 nut-Shengavit’’ levels of T-2E.19 2.15 meters south the northern and southern baulks and under wall of the hearth-oven we uncovered acircular shallow W2810.The rooms were rectilinear, oriented along depression20 cut into the bedrock (30 cm in diam- anorth-south axis. The extantportions of the north- eter, 12 cm deep), lined and rimmed with clay ern room define a 2–2.5 Â 2.3 mspace; the surviv- (the clay rim was 0.6 mthick) (locus 32;elev. ing portion of the south room was 1–1.25 Â 2.4 m 2292.11 ma.s.l.). This hearth was highly burned. in area. The walls were defined by one or two Traces of asimilar feature were discovered 1 mto courses of large and small stones set directly on the northeast (locus 32)extendingunderneath the the bedrock. In the southernarea (locus 46,elev. LB wall W3001.Along the southern edge of the 2291.55 ma.s.l.) we recordedafloor on which lay room, we excavated anumber of stones that may rectangular blocksofclay, flat stones, and ceramic comprise the south wall of the room. This construc- sherds. The central floor area was bordered by a tion was disturbed by aLBpit (locus 30)which cut thin strip (2 cm) of yellow clay which ran along the through the EB floor. Another stone wallorplatform base of the walls, perhaps indicatingthe eroded re- (W3011)was excavated paralleltothe LB wall mains of clay plastering. The primary material exca- W3004;W3011 encircles the eastern edge of the vated from these rooms was ceramic sherds, as bell-shaped pit (locus 28). It appears that the con- well as several fragments of andirons, includinga struction of the LB wall W3004 destroyed the east- piece of arectangular stand with an embossed spir- ern wall of the EB room. In any case,within the al ornament. Fragments of four- and/or three-legged area delimited by LB walls in the eastern half of T- potstands with similar patterns are knownfrom the 30,weencountered alarge concentration of EB ma- excavationsatShirakavan and burial 9 at Keti.21 terial (locus 13). The entire floor area was covered Constructions uncovered in the western third with dense clusters of EB ceramics, most notably a of T-26 correspond stratigraphically and planigra- ‘‘table’’ on four legs with ornamentation on one phically with the lower horizon rooms locatedinT- side (Fig. 10,2).The only non-ceramic artifacts from 28.The lowerEBlevel in T-26 is represented by a the room were aflint sickle blade, atuffdisc, and a trapezoidal room with aclay hearth. This room (lo- pumice tool. cus 27)issmaller than the upper horizon room The western portions of T-26 and T-28,the (1.3–1.9 mN-S Â 1.85 mE-W). It is delimited by upper EB leveldescribed above,contained adi- walls W2609,W2610,and W2611 (loci 38, 39, 40) verse assemblage of ‘‘Karnut-Shengavit’’ pottery, and extends under the western baulk of the excava- and covered an earlier EB horizon. The most com- tion unit. The walls were built of flat stone blocks plete stratigraphic picture of EB cultural deposits preserved in 1–2 courses of masonry. Asmall was provided by the excavation of T-28 (Fig. 8). (70 Â 35 cm) semicircular construction of pebbles Here, the floor of the upperhorizon (loci 12 and 21 was set against the eastern wall (W2610), from bounded by wall W2806)sealed the sand deposit which we recoveredafragmentofaquern. At a which contained re-deposited EB ceramics. The depth of 2291.26 ma.s.l. we encounteredlargely shape and texture of the intervening sand layer in- sterile granitic sand that marked thebase of our dicated arelatively long period of erosional accu- excavations. mulation.Atthe same time, the sand layer’s flat Asmall line of stones built against the east- horizontal upper limit suggests that the surface was ern (external) fac¸ade of W2610 divides the deposits graded in ordered to providealevel surface for the into two discrete portions.Tothe southweencoun- construction of the upperEBoccupation.Wall tered largelysterile granitic sand while to the north W2810,discussed above, served as aterrace wall deposits were composed of brown clay with high for the upper horizon, delimiting thecitadelem- artifact densities. Between walls W2602 and W2609 bankment. Underneath the layer of sand lay a 55– at adepth of 2291.45 ma.s.l. we uncovered acon- 65 cm thick horizon, featuring aseries of thin 2– structionofhighly fired clay. The poor preservation 3 cm black burnt layers angled down slope (loci 20, of the feature precludes arestorationofthe shape 34, 41)and separated by intermittentlenses of of the construction, but it was most likely around granitic sand. Underneaththese layers,weuncov- hearth 0.65 mindiameter. Within the feature we ered two rooms, portions of which extend under recovered only fragments of burnt clay and fragments of an obsidian arrowhead. Asimilarpoint was found under the EB wall W2601.Ingeneral, 19 Badalyan et al. 2008,Fig. 12b. the artifactual remains from the floor layer con- 20 The presence in the same room of standardround hearths and sisted primarily of ceramic sherds. To put the T-26/ avaulted oven trains our attention on the specific function of the latter feature. Thereare no characteristic elements that al- 28 lower horizon constructionsinawider context, low us to identifythe specific function of the furnace. However, its shape and size are similar to the smelting furnaces excavated on the secondhill of Babadervish (see Ìàıìóäîâ et al. 21 Torosyan et al. 2002, 14,Table II,6; ˇåòðîæ'í 1989, 45 Table 1968). 34,Fig. 4. 162 Ruben Badalyan et al.

tery and ahigh density deposit of ceramic fragments (locus 108)set directly on the bedrock. This deposit lacked aclear context due to both the erosion of the slope and later LB construction. The only associated architectural element was asingle row of stones that may represent asurviving fragment of awall which defined the eastern limit of the occupation.The material assemblages from locus 108 include black burnished pottery fragments with ornamentation in high relief in the form of single and double(?) spirals combined with lozenges bear- ing (possibly zoomorphic) figures, abird with atriangular body, and large dimple ornaments (diameter 1.5–2.0 cm). Thesematerials can be typologically assigned to the ‘‘Elar-Aragats’’ early EB complex. Several notable finds came from apartially disturbed context and thus cannot be securely assigned to aspecific EB horizon. Among the materials excavated in T-22,those from locus 115 deserve specialnote. First, we recovered fragments of arectangular ‘‘plaque’’ with smooth raised sides and a flat bottomthat was taperedonboth ends Fig. 11 they are similar in layout and orientation to the (Fig. 10,1).The internal space of the uppersurface Gegharot.EBbronze rooms uncoveredinT-17/18 and define asingle is divided into nine fields by aseries of alternating artifacts. – 1 Pendant from T-29 locus 4 (di- stratigraphic and planigraphic unit. triangles formedbypaired lines in relief. Inside the mensions: 3.3 cm long, In the east citadel operations (Fig. 16),EBoc- four upside down trianglesare vertical rods with 1.85 cm maximum cupation levels were discovered in the central and double volutes; four of the five triangular fields width, 0.1–0.2 cm thick, 3.1 g.); – 2 Be- southern areas of operation T-20.Here we uncov- (one was not preserved)contained ornithomorphic ads from T-22 locus ered a 3.7 mlong retaining wall oriented on a images. Aseries of similar finds comes from Gyuze- 115 (dimensions: north-south axis (locus 105)which created asmall lovy22 and fragmentary comparanda are also known 0.8–0.9 cm in diameter, 23 0.45–0.55 cm thick, terrace 100–120 cm high. Atop the terrace we un- from Sos Ho¨yu¨k. Second, the excavation of T-22 1.0–1.1 g) covered fragments of aclay floor (locus 104, 6– locus 115 revealed twocopperbiconical beads si- 10 cm thick)strewnwith an inventory of in situ arti- milar to those found in sealed contexts from the facts, including intact (locus 108)and fragmented Kura-Araxesburials 22, 29, 35 at Aradetis Orgora,24 ‘‘Karnut – Shengavit’’ceramic vessels and stone burial 8 at Kvatskhelebi, and burials 2, 3,and 4 at tools (Fig. 9);the southern portion of the same Tulepia (Tvlepias Tskaro).25 floor, excavated in 2006,contained fragments of Although our extramural sounding,operation ceramics, complete vessels (Fig 10,3–4) and a T-29,did not locate apreserved occupation layer, hearth (loci 9–10). Near the retaining wall, intrud- we did recover (locus 4)abronze triangular pen- ing into the embankment was apit (locus 109) 30– dant with two symmetrical volutes at the top and 35 cm in diameter and 40 cm deep, containing only lateral loop for hanging (Fig. 11,1).The pendant afew fragments of pottery.The floor area, which has no direct analogues, but the motif of paired seems from the artifactual remains to have hosted spirals closely resembles the anchor-shaped pen- domestic activities,appears to have been an exter- dants on the EB necklace from operation T-2dat ior space as it was not planigraphically associated Gegharot (Fig. 32,1).26 with any extant wall constructions. However, to the To summarize our 2008–2011 investigations north and south of the floor, we uncoveredcorner of the EB at Gegharot, we should first note the si- fragments of contemporary buildings and their cor- milar stratigraphic pattern on the west terrace(T- responding interiors, defined by walls of one or two 2E) and west citadel (T-28). Both stratigraphic col- courses of small stones (15–20 cm). Inside the sur-

viving area of the northern building(Locus 107)we 22 recoveredalarge assemblage of EB ceramic frag- Kos¸ay/Vary 1967, 42,Lev. XII, XXXVII, G.131,G.178,G.206. 23 Sagona et al. 1995,Fig. 8:8;Sagona et al. 1996,Fig. 11:5, 6; ments (approximately 1,500 sherds, of which only Sagona et al. 1997,Fig.7:5. 24 belonged to the LB), while lithic and bone den- 24 Koridze/Palumbi 2008, 128 Fig. 5.4; 131 Fig. 16.3; 133 Fig. 19.3; ღამბაშიძე et al. 2010,Table XI. sities were only moderate. 25 In operation T-22,EBlayers were represented Glonti et al. 2008; ღამბაშიძე et al. 2010,Table VIII. 26 Smith et al. 2004,Fig. 15;Hayrapetyan 2005;Meliksetian et al. by both sporadic findsof‘‘Karnut-Shengavit’’ pot- 2009. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 163

Site Operation Material Lab 14C Calibrated Date Number Date (BP) (BC) 95.4%Probability

Gegharot, 2008 Ò-20,l.104 charcoalAA-92623 4283 Æ 40 3020–2763

Gegharot, 2008 Ò-19,l.112 charcoalAA-92844 4620 Æ 120 3646–3017

Gegharot, 2010 Ò-22,l.105 charcoalAA-92622 4128 Æ 41 2872–2580

Gegharot, 2010 Ò-28,l.21 charcoalAA-92621 4104 Æ 40 2872–2500 Tab. 1 Gegharot, 2011 Ò-28,l.38 charcoalAA-95617 4119 Æ 42 2873–2576 Radiocarbondetermina- Gegharot, 2011 Ò-30,l.22 charcoalAA-95615 4204 Æ 52 2906–2630 tions related to Gegha- rot EBA contexts from Gegharot, 2011 Ò-28,l.41 charcoalAA-95616 4374 Æ 42 3263–2899 the 2008–2011 field seasons (calibrations Gegharot, 2011 Ò-28,l.34 charcoalAA-95618 4391 Æ 49 3325–2902 by OxCalv.4.1.7Bronk Ramsay 2010 r. 5) umns containeddiscrete EB layers comprised of mains, and both have birds of varying sizes, but two isolated horizons, eachofwhich was character- these non-mammalian remains are rare (see ized by adistinct ceramic complex, ‘‘Karnut-Shenga- Tab. 2).Among the mammalsidentified to the level vit’’ and ‘‘Elar-Aragats’’.The hiatus between the of genus, sheep and goats form the majority of two primary phases of EB occupation was marked both samples, but the Elar-Aragats sample has a by apronounced wedge-shaped layer of sandy col- significantly higher proportion of cattle than does luvial soil, which was, accordingtothe extantradio- the Karnut-Shengavit sample, which more closely carbon dates, relatively brief. resembles other Tsaghkahovit Plain assemblages. Of the eight radiocarbon determinations sub- Equidsare presentinboth samples; however, only mitted from the EB layers excavated at Gegharot in the Karnut-Shengavit sample could they be iden- from 2008–2011 (Tab. 1),six (AA-92623,AA-92844, tified more specifically, with one domesticated AA-92622,AA-92621,AA-95617,AA-95615)belong horse (T-19.6)and one onager present. Pigs are to the upper ‘‘Karnut-Shengavit’’ horizon. Most of present in the Karnut-Shengavit sample (from T- the dates fit comfortablywithin the date range of 19.109,T-26.05,T-28.11), but absent from the Elar- 2900–2500 BC defined for this occupation by prior Aragats sample. determinations,although the results of AA-92623 Given the small size of the Elar-Aragats sam- only partially correspondtothis range. The sample ple, no further work could be done to examine from T-19 locus 112 (AA-92844)was stratigraphi- herdingpatterns within this horizon.Survivorship cally linked to the upper ‘‘Karnut-Shengavit’’ hori- was calculated for sheep and goats in the Karnut- zon, so its surprisingly early date may indicate dis- Shengavit sample, but even this suffered from sam- turbance of an earlier horizon by later construction. ple size issues. Only Stage Ahad asufficiently large In any case, the magnitude of the statistical range samplesize to be considered reliable; survivorship for this sample makes it aless useful determina- in this stage is comparable to that in other Tsagh- tion. Lastly,the dating of the lower ‘‘Elar-Aragats’’ kahovit Plainsamples (see Tab. 3).Survivorship in horizon, defined by samples AA-95616 and AA- the other two stages, especially Stage C, is extre- 95618,indicates arange of 3350–2900 BC, sup- mely low and would indicate aherd that was not porting previousdeterminations for the age of the viable,but the sample size for both stagesmakes lower horizon of EB settlement at Gegharot. these figures suspect. The division of the EB levels at Gegharot into It has been suggested that settlementinthe two distinct horizons makes possible an examina- EB period became more nomadic over time,28 and tion of change over time in the fauna at the site, the higher proportion of cattle in the earlierhorizon albeit on alimited scale. As with all of the other at Gegharot might support this idea. However, the faunal samples from the Tsaghkahovit Plain, both presence of pigs, albeit in small quantities, in the of the EB faunal sample sets from Gegharot are Karnut-Shengavit sample suggests that the popula- composed primarily of mammals,27 the Karnut- tion was not entirely nomadic. Nevertheless, the Shengavit sample also contains snake and toad re- differences in the two samples indicate an intri- guingdivergence in pastoral practices between the two horizons of the EBA. 29 27 Only the faunafrom contexts whichcould be identified definitively as Elar-AragatsorKarnut-Shengavit are included in this analysis. The total analyzed Early Bronze Age fauna sample in- cludes additional contexts that have not been assigned to a specific sub-horizon. 28 e.g. Kushnareva 1997. 164 Ruben Badalyan et al.

GegharotEarly Bronze Elar-Aragats Gegharot Early BronzeKarnut-Shengavit NISP %NISP %NISP NISP %NISP %NISP to genus to genus Bufonidae Toads 00.00% 23 0.81% Squamata Snakes 00.00% 10.04% Fish 00.00% 30.11% Large bird 20.19% 00.00% Medium bird 10.09% 00.00% Small bird 00.00% 10.04% Indeterminate 656 61.71% 810 28.64% Large mammal 85 8.00% 419 14.82% Medium mammal 126 11.85% 571 20.19% Small mammal 30.28% 42 1.49% Artiodactyl 00.00% 00.00% Large artiodactyl 00.00% 00.00% Medium artiodactyl 00.00% 00.00% Bovid 30.28% 10.04% Bos Cattle 85 8.00% 49.95% 207 7.32% 22.28% Capra Goat 40.38% 2.16% 37 1.31% 3.98% Ovis Sheep 60.56% 3.24% 120 4.24% 12.92% Gazella Gazelle 10.09% 0.54% 20.07% 22.00% Ovis-Capra-Gazella 10.09% 00.00% Ovis-Capra 83 7.81% 44.86% 442 15.63% 47.58% Cervid 00.00% 23 0.81% Cervus Red deer 10.09% 0.54% 10.04% 0.11% Dama Fallow deer 00.00% 0.00% 10.04% 0.11% Equus 20.19% 40.14% 0.43% Equus caballus Horse 00.00% 0.00% 10.04% 0.11% Equus asinus Donkey 00.00% 0.00% 00.00% 0.00% Equus hemionus Onager 00.00% 0.00% 10.04% 0.11% Equus asinus/hemionus 00.00% 0.00% 00.00% 0.00% Sus Pig 00.00% 0.00% 30.11% 32.00% Carnivore 00.00% 00.00% Canid 00.00% 20.07% Large canid 10.09% 20.07% Canis familiaris Domesticated dog 30.28% 1.62% 30.11% 0.32% Canis lupus Wolf 00.00% 0.00% 10.04% 0.11% Vulpes Fox 00.00% 0.00% 00.00% 0.00% Meles Badger 00.00% 0.00% 00.00% 0.00% Lepus Hare 00.00% 0.00% 11 0.39% Tab. 2 Rodentia 00.00% 10.04% Faunal assemblage for Gegharot EB contexts Microtus Vole 00.00% 0.00% 15 0.53% 1.61% (2007–2011 field Spermophilous Souslik 00.00% 0.00% 80 2.83% 8.61% seasons) accordingto Number of Identified Total 1063 2828 Specimens. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 165

Stage AStage BStage C Gegharot Early Bronze Karnut-Shengavit 75.00% 47.62% 25.00% Tab. 3 Sheep-goat survivor- GegharotLate Bronze 69.75% 55.29% 33.42% ship percentages for Aragatsi Berd Late Bronze 65.22% 33.33% 42.86% Project ArAGATS Bronze Age contexts TsaghkahovitLowerTown Late Bronze 82.91% 82.69% 62.50% (2007–2011 field seasons)29

Fig. 12 Gegharot.Photo of LB Shrine 2,Operation T-21

Paleobotanical remains from the EB occupation to hulled varieties,and apart of this belonged to at Gegharot provide considerable materialfor under- multi-rowed subspecies Hordeum vulgare ssp. vul- standingboth the local environment and developing gare convar. vulgare), breadwheat (Triticum aesti- plant economies. Gegharot’s EB layers have to date vum amongst which was common Triticumaestivum yielded the greatest quantity of archaeobotanical ssp. vulgare)and club wheat(Triticum aestivum ssp. material of the Tsaghkahovit Plain sites (approx. compactum). The barley to wheat ratio is 95%to5% 31,200 units; Tab. 4) thanks in large part to the EB in EB samples from Gegharot. Species of Galium,fa- occupation uncoveredinoperations T17 and T18.30 milies of Poaceae and Polygonaceae are prevailing The following taxa of cultivatedplants prevail in the amongstEBA samplesofGegharot. EB samples from Gegharot (Tab. 4):cultivated barley EB agriculture on the Tsaghkahovit Plain, as re- (Hordeum vulgare – the majority of which belonged presented by paleobotanical evidence from Gegha- rot, appearstohave been based on cereal cultiva- tion, particularly hulled barley and bread wheat. 29 Stage Aiscomprised of the following elements: distalhume- rus, innominate, proximal radius, scapula,first phalanx, and se- This pattern is in contrast with earlier Neolithic and cond phalanx. For sheep and goats it represents roughly ages Chalcolithic agricultural traditions in the region 6–13 months; for cattle, 10–24 months. Stage Biscomprised which were based on an almost equal representa- of distal tibia anddistal metapodials. For sheep and goatsit representsroughly 15–24 months, for cattle, 24–32 months. tion of cereals, pulses, and oil plants in the surviv- Stage Ciscomprised of distal femur, proximal femur, proximal ing archaeological record.31 humerus,distal radius,proximal tibia, and proximal ulna. For sheep and goats it represents roughly 36–42 month,for cattle, 42–48 months. Ages from Schmid 1972 and Silver 1969. 30 Hovsepyan 2008. 31 Hovsepyan/Willcox 2008. 166 Ruben Badalyan et al.

Gegharot, EBA Gegharot, LBA Aragatsi Berd, Tsaghkahovit LBA (SLT), LBA Number of samples 36 114 13 71 Volume of processed sediments (liter) 237 1,728 37 945 Concentrationoftotal carpological material (unit/liter) 131.75.63.33.4 Concentrationofcultivated plants carpological material (unit/liter) 127.51.11.21.5 Taxa number 179 95 19 62 Plant Taxa 31,207 Preservation* 26,931 9,680 121 3,216 Cultivated cereals 96.8% 96.3% 19.1% 36.4% 43.5% Unidentifiedgrains(fragments) ratio inside cereals 2.9% 26.9% 43.2% 61.4% Triticeae grains C 861 453 19 858 gen. spp. unidentifiable M 37 fragments rachis internodesC 1 glumes bases C 6 awns M+ Barley –wheatratio 94.6%74.8%56.0%74.5% Hordeum vulgare unidentifiable grains C 264 603 13 311 M 25 rachis fragments C 22 triplet middle grains C 22 Hordeum vulgare (hulled) triplet position C 25,770 225 172 unidentifiable hulledgrains M 63 Hordeum vulgare triplet middle hulled grains C 356 32 10 (hulled) Hordeum vulgare triplet left hulled grains C 721 20 3 subsp. vulgare convar. vulgare triplet righthulled grains C 644 8 Wheat -barley ratio 5.4% 25.2% 44.0% 25.5% Triticum spp. grains C 154 100 664 glume base C 11 1 Triticum sp. (hulled) spikeletfork C 3 Triticum aestivum/ grains C 628 51 115 turgidum Triticum cf. grains C 25 18 4 aestivum Triticum aestivum internode C 21 Triticum cf. aestivum grains C 91 5 (small grained) Triticum cf. aestivum ssp. grains C 527 16 28 vulgare Triticum cf. aestivum ssp. grains C 128 10 18 compactum Triticum cf. aestivum ssp. grains C 9113 sphaerococcum Triticum cf. aestivum ssp. spelta grains C 21 1 Triticum durum/dicoccum grains C 12 9 Triticum cf. durum grains C 24 1 Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 167

Gegharot, EBA Gegharot, LBA Aragatsi Berd, Tsaghkahovit LBA (SLT), LBA Triticum cf. dicoccum grains C 8404 M 2 Triticum dicoccum grains C 12916 Triticum cf. dicoccum glume base C 1 Triticum dicoccum spikelet C 1 Triticum cf. dicoccum spikelet ‘‘forks’’ C 1 Triticum dicoccum spikelet ‘‘forks’’ C 2231 Triticum dicoccum internode C 2 Triticum cf. monococcum grains C 14 Triticum monococcum grains C 1 Triticum cf. monococcum spikelet ‘‘forks’’ C 32 Triticum dicoccum/Secale sp. grains C 22 Other cereals (possibly cultivated) 0.0%2.8%0.0%1.5% cf. Secale sp. grains C 3 Secale sp. grains C 317 cf. Panicum miliaceum hulled grains M 71 naked grains C 7 Panicummiliaceum naked grains C 11 Grape (Vitaceae) Vitis vinifera pip C 2 Weeds 3.2% 80.9% 63.6% 56.1% Poaceae Poaceaegen. spp. grains C 60 104 93 Paniceae gen. spp. grains C/M 15 11 Bromus sp. grains C 21 15 3 Lolium sp. grains C 89 11 Hordeum sp. (wild) grains C 3 cf. Aegilops sp. grains C 71 cf. Avena sp. grains C/M 91 Poa bulbosa bulb-grains C 1 Fabaceae Viceae gen. spp. seeds C 23 17 Fabaceae gen. spp. 2 seeds C 511 (small seeded) Rubiaceae Galium sp. mericarps C/M 25 40 130 Galium cf. spurium mericarps C/M 435 974 22 352 Galium cf. aparine mericarps C/M 36 38 Asperula sp. mericarps C 13 3 Boraginaceae Buglossoides arvensis erems&seeds M/C 16 3920 32 179 Anchusa arvensis erems&seeds M/C 55 3 Echium sp. erems M 49 1 168 Ruben Badalyan et al.

Gegharot, EBA Gegharot, LBA Aragatsi Berd, Tsaghkahovit LBA (SLT),LBA Asperugoprocumbens erems M/C 10 10 5 Rochelia sp. erems M/C 5 Heliotropium sp. erem C 1 Lapula sp. erem M 1 Polygonaceae Polygonaceae gen. spp. nutlets C/M 22410 Rumex sp. nutlets C/M 19 31 16 Polygonum spp. nutlets C/M 12 359 1 47 Polygonum cf. aviculare nutlets C/M 33958 Polygonumconvolvulus nutlets C/M 14 234 143 Polygonaceae/Cyperaceae spp. nutlets, seeds C/M 24 208 145 Cyperaceae Cyperaceae gen. spp. nutlets C/M 13 109 41 Scirpus sp. nutlets C 30 19 Carex sp. nutlets C 340144 Eleocharis sp. nutlets C 63 Brassicaceae Brassicaceae gen. sp. seeds C/M 7213 Thlaspi spp. seeds C/M 66359 Neslia sp. capsules C/M 77248 Alyssum sp. seeds C/M 16 cf. Camelina sp. seeds C/M 3 cf. Bunias sp. capsule C/M 4 Convolvulaceae Convolvulus sp. nutlets C/M 5372 Chenopodiaceae Chenopodium sp. seeds C/M 13 236 152 Solanaceae Hyoscyamus sp. seeds C/M 27 26 79 Ranunculaceae Adonis sp. seeds C 2311 Ranunculus spp. seeds C 42 Scrophulariaceae Veronica sp. fruit (capsule) C 1 Asteraceae Asteraceae gen. spp. seeds C/M 36 3 Apiaceae Chaerophyllum sp. mericarpC/M 19 Apiaceae gen. sp.2 seeds C/M 25 cf. Prangos sp. seeds M 1 Lamiaceae Lamiaceaegen. spp. nutlets C/M 4506 Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 169

Gegharot, EBA Gegharot, LBA Aragatsi Berd, Tsaghkahovit LBA (SLT), LBA

Caryophyllaceae

Caryophyllaceae gen. sp. seeds C/M 31 5

Vaccaria sp. seeds C/M 48 2

Scleranthus sp. fruit C/M 37 11

Papaveraceae

Papaveraceae gen. spp. seeds C 11

Malvaceae

Malvaceae gen. sp. seeds C 1 cf. Althaea sp. seeds C 11

Violaceae cf. Viola sp. seeds C 61

Urticaceae cf. Urtica sp. seeds C/M 145

Rosaceae

Rubus sp. nutstonesC/M 1

Rosa sp. nutstonesC 33415

Cupressaceae

Juniperus sp. twigs C/M 63

UnidentifiedSpecies 0.6% 6.8% 3.3% 12.2%

seeds C/M 186 657 4392

Granary weevils &other C/M 22 insects

The Late Bronze Age (Lchashen-Metsamor) since 2002 date the two phases of occupation Tab. 4 broadlytothe late 15th–late 11th cent. BC (Tab. 5).32 Plants identifiedinthe settlement macrobotanical samples Gegharot during theLBA constitutes aparticu- from the primary occu- larly unique site due to the presence of three shrine pation contexts exami- LB levels succeed the EB occupations in all excava- complexes. The first of these shrines, uncovered on ned by Project ArAGATS (2008–2011 field sea- tion units within Gegharot’s citadel and west terrace the west terrace (T-2E), was discussedinour pre- sons) and in the extramural areas explored by operations vious report.33 In 2008,asecondshrine complex T-25 and T-29.While the EB occupationsatGegha- was discovered in operation T-21 on the west cita- rot indicate alargely egalitarianagricultural village del (Figs. 4; 12).This LB layer was represented in withoutclear evidence of social differentiation, the the central and eastern portions of the excavation LB occupation is defined by greater social stratifica- which contained two occupation levels. The lower tion and political centralization. Moreover, the exca- level, 0.63 mthick (loci 3, 6)beginning at adepth vations of LB Gegharot havedocumented anumber of 2293.14 ma.s.l., covered an area of 16–20 m2. of dramatic episodes that took place in the Tsagh- kahovit Plain during the second half of the 2nd mill. BC which resulted in the total destruction of the 32 In collaboration with Stuart Manning and the Malcolm and Caro- lyn Weiner Laboratory for Aegean and NearEastern Dendrochro- site, the partial reconstructionofseveral complexes, nology at Cornell University, we are currently working to refine and the final immolation and abandonment of the this chronology in order to both definethe temporal limits of each settlement. occupation phase and establish the dates for the two destruction episodes. The results of this work will be reported in aseparate Radiocarbon determinationsconducted on communication when the results are available. samples recovered from Gegharot’s LB horizon 33 `àäàº'í et al. 2005;Badalyan et al. 2008. 170 Ruben Badalyan et al.

Site Operation Material Lab Number 14CDate (BP) Calibrated Date (BC) 95.4%Probability Gegharot, 2010 T21,l.35 charcoal AA-92624 2891 Æ 39 1250-939

Gegharot, 2010 Ò-21,l.6 charcoal AA-92625 2971 Æ 39 1370-1053

Gegharot, 2010 Ò-21,l.2 charcoal AA-92626 2909 Æ 42 1261-980

Gegharot, 2010 Ò-21,l.12 charcoal AA-92627 2943 Æ 39 1291-1017 Tab. 5 Radiocarbon determina- Gegharot, 2011 Ò-31,l.41 charcoal AA-95613 3094 Æ 39 1439-1265 tions related to Gegha- rot LBA contexts from Gegharot, 2011 Ò-31,l.7 charcoal AA-95614 3039 Æ 40 1413-1133 the 2008–2011 field Gegharot, 2011 Ò-32,l.54 charcoal AA-95611 2960 Æ 43 1368-1026 seasons(calibrations by OxCal v. 4.1.7 Bronk Gegharot, 2011 Ò-32,l.20 charcoal AA-95980 3015 Æ 40 1392-1129 Ramsay 2010 r. 5)

The eastern portion of the deposit (locus 3, 35cm by several pits (loci 37, 54, 55)except in afew thick), consisted of acharcoal rich soil with high areaswhere it overlay portions of an earlier EB densities of LB pottery and burnt bones, including floor (see above). Against the eastern limit of the adiscoid pendant with acentral hole made from a shrine near the southeast corner, was akidney- catfish (Silurus glanis)vertebra (Fig. 15,21),aswell shaped ‘‘altar’’ (locus 22, 1.5 Â 1.25 m), set on a as acollection of stone tools (pumice polishers, dis- bedrock ledge (locus 41). The perimeter walls of coid objects,sharpening stones, etc.). The western the baked clay ‘‘altar’’ were 5 cm wide clay bottom portion of the lower level (locus 6)was part of a that may have been lined with small stones. Inside rectangular area (approximately 4.7 Â 4.2 m), or- the ‘‘altar’’ were seven vessels, includingalarge iented to the west, with awell-preserved in situ as- storage jar (locus 12), bowls, acup with knobs on semblage. The northern and eastern walls of the the perimeter of the body, and asimilarly deco- room joined with natural outcrops of the substrate rated small jar (locus 23,inside this vessel was a bedrock.The north wall (W2104)was 5.6 mlong, bronze needle) (Fig. 14,1–7).Inthe northern part defined by aseries of stone masonry blocks; the of the ‘‘altar’’ was astationary ceramic hearth (lo- survivingfragment of the east wall (W2106), 2.2 m cus 4, 0.56 Â 0.6 m). This vertical installation stood in length and consistingof2courses of stone ma- on acircle of small stones, and was covered with sonry (0.25–0.60 m) (locus 43), was built on arock clay and baked throughout. At the bottom was set outcropfrom large and smallunworked boulders. aflat round stone. Inside the chamber was a man- The south wall of the room (W2103), which sur- ghal.Asecond, larger manghal was located in the vived only to alength of 4.4 m, was set into an southern part of the ‘‘altar,’’ to the right of the sta- erodedsection of the naturalrock. The wall is composed of one line of medium sized (35 Â 55 cm) stones and preserved in severalcourses of masonry (0.15 m). However, it is not entirely clear that South Caucasus derive from Middle Bronze to Iron 1 period W2103 was part of the described complex. It could settlements (i.e. Aragatsi Berd, Gegharot, Kuchak, Metsamor, be associated instead with abuilding complex exca- Shirakavan, Tsaghkahovit,Uch-Tepe,Uzerlik-Tepe) and they are typically associatedwith either stationaryovens or concen- vated in operation T18 in 2006 that included acur- trationsofcharcoal and ash. Some, though not all of the ex- vilinear wall (W1807), as well as, perhaps, W1806. tant manghals show evidence of severe burning. Three factors At adepth of 2292.79 ma.s.l. we uncovered have ledtoawidely dispersed interpretation of manghalsas ritual paraphernalia: their repeatedassociation with cultic com- afloor (locus 6)littered with in situ intact and frag- plexes (e.g., sanctuary two at Metsamor and all three shrines mented vessels,including aceramic form knownas at Gegharot), the discovery of small manghal models (a model a‘‘manghal’’, as well as stoneand bone tools from Kuchak was just 11 cm wide and 5.6 cm tall), and percei- ved morphological similaritiesto‘‘censers’’/‘‘cult boxes’’. How- 34 (Fig. 13). The floor was laid atop bedrock and cut ever, there areatleasttwo reasons to be wary of the traditio- nal view. First, the large number of manghals foundat Gegharot,where at least six examples have been documented 34 In the archaeologicalliterature of the South Caucasus,the term (T-2Elocus 218,T-27 locus 56,T-32 locus 16,T-21 locus 4, ‘‘manghal’’ (Arabic: mankal)isused to describe the portable T-21 locus 21;T-21 locus 22)suggests they were not exceptio- hearths or braziers. However, notonlyisthe specific function nal elements of the ceramic repertoire. Second, while censers of these vessels unclear but even the exact orientation of their from Gegharotand elsewhere (e.g., Lchashen; Petrosyan 2002) use is unknown. The manghal found outsidethe altar in the tend to be rather lavishly decorated and well made, manghals T-21 shrinewas found installed verticallywith the open side are quite ordinary in appearance, madeofpoorly levigated clay up. In contrast, the manghal found near the altar in the T-32 with high densities of fine sand. Formally, then, manghalsap- shrinewas lying horizontally, although it is not clear whether pear to havebeen regarded as rather utilitarianobjects, even this was its original position or if it had fallen during the de- if their prosaic function (perhapsbaking bread?)was related,in struction event. Almost all of the manghalsexcavated in the some cases, to sacred institutions and practices. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 171

Fig. 13 Gegharot.T-21 LB materials from floor of Shrine 2 172 Ruben Badalyan et al.

tionary hearth. Around this manghal was an assem- of pottery as well as bone and stone tools. The blage of ceramic sherds, aclay model wheel, a stone tool inventory was relatively rich, including large retouched obsidian blade, atriangular obsi- small mortars and pestles as well as fragmentsof dian arrowhead with anotched base, fragments of grinding stones and pumice polishers. obsidian flakes, as well as bones both charredand The ceramic assemblageofthe lower LB hori- unfired. zon which included both the floor and the ‘‘altar’’ To the south-west of the ‘‘altar’’, running un- was generally homogeneous, representingwares of derneath wall W2102,was apit constructed in the LB II phase dating to the end of the 14th three sections. The largeupper chamber divided through the first half of the 13th cent. BC. into two distinct deeper chambers. The pits were The lower (initial) layer of LB construction de- dug into the bedrock to varying depths. The two tected in T-21 was also documented in aseries of deep chambers (chamber A, locus 37 to the east pits uncovered in the eastern half of operation T-26. and chamber B, locus 54 to the west) have adepth The pits were cut into the granite bedrock substrate, of 1 m, while the shallower large upper chamber (C, tapering at the bottom. Aportion of pit 1 (locus 20, loci 37, 54, 55)was 0.3 mdeep. Chamber A(di- 1.6 Â 1.4 m, 0.97 mdeep), extendsunder the edge mensions 1.2 Â 0.95 m) was filled with ablack of the northern baulk and wall W2603,aconstruction burned soil rich in ash, charcoal, and small bones. of the second LB phase. The pit was sealed by aclay The walls were plastered with clay, beneath which floor and burnt fragments of charredbeams (identi- we uncovered two large boulders on opposite fied as Betula/birch;36 loci 16, 19). Withinthe pit fill, sides. At adepth of 2291.81 ma.s.l. we encoun- we recoveredsizable assemblages of fragmented LB tered alarge LB storage jar (locus 45), asmall bowl ceramics, faunal remains, obsidian pieces, broken (locus 47)and fragments of three or four additional stone tools, and adiscoid stone bead. Pit 2 (locus vessels, includingtwo pots with ornamentation un- 22, 1.65/1.9 Â 1.1 m, 0.96 mdeep) was located to der the rim and alarge bowl with vertical handles the west of pit 1 and also extended under the north and aflat cut rim (Fig. 14,8–12). baulk. The pit was filled with charcoal and organic Asecond two-part pit (2 Â 1 m) was found to matter along with alarge collection of LB pottery the northeast of pit 1,demarcated by adouble row fragments. Pit 3 (locus 23, 1.95 mindiameter, 0.9 m of flat stones on its western edge. The pit was filled deep) was sealed by the construction of walls with charcoal,afew fragments of ceramics and a W2606 and W2608 during the 2nd phase of LB con- small faunal assemblage. 1.10 mwest of pit 2 was struction, as well as aclay feature (locus 17). The athird large pit which may date to the first LB con- upper layer of the pit contained large pieces of obsi- struction phase. LB ceramic fragmentsfound in the dian, as well as two or three fragments of asingle pit were identical to those of the vessel recovered ceramic vessel.Within the deeper fill we recovered from pit 2 (locus 54)and the LB ceramics fromun- fragments of LB pottery, high densities of faunal re- der the floor of the upper LB phase. Inside the pit mains, obsidian pieces, afragmentofahandstone, we recovered arelatively large faunal assemblage. pumice polishers, amodel wheel of unbaked clay, a Found in situ on the edge of the pit, amongst the spindle whorl with punctate ornamentation, aplum- rocks of W2105 (locus 38), we recovered acollec- met,acup with pressed nail decoration on thesur- tion of 7 bone awls, 2 bone spindle whorls, 2 clay face, and alarge number of bone implements. The model wheels, and aretouched dacite tool,35 all of smallest of T-26’s LB pits (0.60 mdiameter, 0.35 m which appear to belong to the lower LB horizon deep), pit 5 (locus 29), containedanalmost com- (Fig. 15). plete LB vessel set on asemicircle of tuff stones. The western portion of the T-21 shrine was The early LB occupation was also documented disturbedbyconstructionofthe upper LB occupa- in operation T-28 where we uncovered architectural tion. Directly under topsoil we uncovered afrag- constructions, aseries of pits, and fragments of ment of acurvilinear wall (W2102)which represents large storage jars. Pit 1 (locus 8; 1.75 Â 1.1 m, this second LB phase. WallW2102 is composed of 0.75 mdeep) extendsunder wall W2803.The cera- asingle line of stones set in 2–4 courses preserved mic remains from the pit were exclusively LB. The to aheight of 35–70 cm. The extant arc of the wall pit was sealed by asurface littered with charred is 4.70 mlong. In association with the wall, we ex- beams. Pit 2 (locus 9, 0.8–0.9 Â 1.0 m, 0.55 m cavated stone floor slabs covered with aclay coat- deep) lay 30–45 cm south of pit 1.The bottom of ing at adepth of 2292.36 ma.s.l., which survived only in afew places. The material assemblages from this upper LB occupation includedfragments 36 Identification of woodspecies conducted by the Cornell Den- drochronology laboratory as part of an in-progress collaborative program of research, the full results of which will be detai- 35 Morphologically, the artifact more closelyresembles aPaleoli- led in aseparate publication. Our thanks to Sturt Manning and thic tool than an implement of the BronzeAge. Kate Seuferfor the data describedhere. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 173

Fig. 14 Gegharot.T-21 LB materials from Shrine 2 altar (1–7)and pit (8–12) 174 Ruben Badalyan et al.

Fig. 15 Gegharot T-21 LB materials from Shrine 2. – 1–2 ceramic; 3–14, 18, 20 bone; 15 obsidian; 16 antler; 17 flint; 19 dacite; 21 Silurus glanis (catfish) vertebrae

the pit was filled with small packed stones set atop stones separated by layers of yellow clay, atop athin layer of soil. The pit artifact inventory in- which sat a‘‘stone box’’, that perhaps served as a cluded afew fragments of LB pottery, two frag- post support. The northwestern perimeter of the pit ments of obsidian and arelatively large collection was paved with small stones. On the south side of of faunal remains. T-28 Pit 3 (locus 24, 2 mdiam- the pit was a 2–3 course stonewall (30 cm high) eter) was filled with three layers of tightly packed built of flat stones, including severalreused hand- Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 175

Fig. 16 Gegharot.Plan of east citadel operationsT-20, T-22,T-27,T-32,T-33, T-34 176 Ruben Badalyan et al.

Fig. 17 Gegharot.Photos of shrine 3. – 1 T-27 locus 5 and 11 from the north; – 2.T-32 platform and altar from the north

stones. Amongthe finds from the pit were frag- tended into the south baulk, and large fragmentsof ments of stone implements, used as fill, as well as burnt beams, covering a 3–5 cm layer of yellow complete tools,including pestles,graters, and a clay. Pit 2 (locus 30, 1.25 Â 0.6 mindiameter, mortar. Pit 4 (locus 40, 0.90 Â 0.70 m, 30 cm deep) 0.63 cm deep) contained only small fragments of containedamodel wheel, fragments of tuff weights, pottery,obsidian flakes, and bone. Pit 3 (locus 31, and asmall tuff ball. Pit 5 (locus 43, 1.35 mindia- 1.0 Â 0.75 mindiameter, 0.72 mdeep)located meter, 50 cm deep) was filled with ahigh density near the northeast corner of the operation, con- of large animal bones, including astragali. In addi- tained alarge collection of faunal remains, afrag- tion, the pit containedasmall LB ceramic vessel ment of abone awl, and aceramic collection that and several stone tools, includingapumicepol- contained anumber of EB sherds alongsideLBfrag- isher, atuff disc, and flat basalt tools. ments. The initial LB occupation was thus repre- The granite paving stones in the northeast sented by excavationsunderneath walls W2803, sector of T-30 were set directly on bedrock. Atop W2804,W2807,and W2808 and within the inter- the paving was an overturned LB bowl (locus 20). vening spaces of the later constructions, where we The paved surface was sealed by a 10 cm layer of uncovered fragments of 2–3 large storage jars. The burnt earth containingalarge amount of charcoal blackened soil matrix for this deposit was rich in and fragments of charred beams which separates charcoal, ash, burnt bones and organic remains as the paved floor from the bottom course of masonry well as large fragments of burned beams. This early in wall W3002.The lower LB horizon recorded in LB layer extends west to the EB wall W2809 and the excavations of T-26,T-28,and T-30 was almost south to wall W2805 (locus 19)where alight yellow alwayssealed by alayer of burnt earth with frag- clay LB floor directly overlies the EB surface. ments of charcoaland charredbeams. Operation T-30 also revealed evidenceofthe The upperLBoccupationatGegharot was re- initial LB occupation level at Gegharot, including presented in T-26 by constructions W2603,W2604, several pits. Pit 1 (locus 24, 1.35 Â 0.9 mindiam- W2606,W2607,W2608,ahearth, and ayellow clay eter, 1.12 mdeep) contained two complete LB floor which sealed pits 1–3,and 5.This partially vessels (loci 25, 26), including abowl which ex- preserved yellowclay floor was rich in LB ceramics, Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 177

Fig. 18 Gegharot.Photos of shrine 3; – 1 T-32 altar and quern installation from the east; – 2 T-32 locus 18,overview of altar including fragments of a manghal and crucible, as ishers, etc.). In the northeast corner, parallel to the well as faunal remains, large pieces of obsidian, north baulk, we uncovered awall fragment – and stonetools (mortars, grinding stones, pol- W2603 (locus 32, 2.25 mlong, 0.35–0.45 mwide, 178 Ruben Badalyan et al.

Fig. 19 Gegharot.Shrine 3 ceramics. Pots Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 179

Fig. 20 Gegharot.Shrine 3 ceramics. Cups, bowls, small jars, and otherforms 180 Ruben Badalyan et al.

Fig. 21 Gegharot. Shrine 3 ceramics. – 1–2 manghals: – 3–6 storage jars

0.3–0.55 mhigh), which intersected at an acute an- ashallow depression in the centersuggesting it gle with an adjacent construction that consisted of was either areusedmortar and/or perhaps apost two stone ‘‘boxes’’ set within adistinctly irregular support. The architectural configuration of the later linear feature made of small stones arranged in 2– LB constructions in T-26 is difficult to understand.It 3 layers (W2604,locus 33). The bottom of the is possible that walls W2603 and W2607 were por- western box was constructed of alarge stone with tions of the same construction.However, these Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 181

Fig. 22 Gegharot.Shrine 3 and related materials. – 1 Mittanian commonstyle cylinder seal from T-20; – 2 carnelian beads from shrine 3; – 3 paste beads from shrine 3; – 4–5 bronze arrowheads from shrine 3; – 6–7 stone pendants from shrine 3; – 8 ceramic stamp from T-30; – 9 ceramic stampfrom T-32 shrine 3 182 Ruben Badalyan et al.

fragments do not clearly articulate with other walls 2.2 mlong, 0.45 mwide) built against bedrock.It which might havecomposed aroom. was preserved in 4–6 courses of stone masonry Another wall of the upperLBhorizon – (1 m). In front of W3002 and alongthe northern W2606 (locus 35, 2.2 mlong, 0.4 mwide) – was baulkwas adense packed clay floor (size located along the southern baulk. Perpendicular to 0.9 Â 2.5, 15–20 cm thick). In the area adjacent to the W206 was an attached wall (W2808,locus 37, the small recess we uncovered astone ‘‘box’’, simi- 1.10 mlong, 0.35 mwide), preserved as 1–2 lar to thosefrom T-26,which consisted of apit courses of masonry (0.5 m).Inthe small recess (30–35 cm in diameter, 60 cm deep)partially lined formedbetween W2606,W2607,and W2608,we with stones. The entire floor area bounded by walls excavated arectangularclay feature (locus 17)that W3001,W3002 and W3004 (loci 3, 7)was filled resembled ahearth, set into the center of the floor; with amatrix of blackened soil, rich in charcoal. however, the lack of evidence of fire in the feature The excavated material from T-30’s upper LB level suggests that it served adifferent function. In the was quite varied, with LB ceramics, arange of clay southeast corner of the T-26 we opened afeature objects (plummets, disks), faunal materials, and more clearly identifiable as ahearth (locus 4, stonetools. Of particular interestwas around cera- 0.7 Â 0.4 m), consisting of arectangular stone laid mic stamp (diameter 4.5 cm, height 4 cm), with a flat and perpendicular to acluster of small stones symmetricalcross in relief (Fig. 22,8). that was covered with clay. Inside the hearth was The third shrine at Gegharot, excavated in an ash matrix which contained the bottom of a 2010 and 2011,was located on the easterncitadel, pan, or perhaps manghal,impressedwith the im- primarily within the boundaries of operations T-27 print of awoven mat as well as burned bones. In and T-32 (Fig. 17–18).The shrine was arectangular and aroundthe hearth we also found charred stonebuilding set on anorthwest-southeast axis pieces of twigs and charcoal. that followed the contour of the outcrop. The east- Excavationsalong the northern edge of T-28 ern wall of the room was not preserved due to the revealed agroup of structuresassigned to the later natural erosion of the slope. The surviving western phase of LB occupation. Taken together, walls (W3201)and northern (W3202)walls of the room W2803,W2804 (loci 27, 28, 29), and W2807 form a circumscribed an interior space of approximately narrow rectilinear structure (0.60 Â 1.70 m) com- 37.6 m2.The masonry blocks of W3201 had been posed of small stones (18–30 Â 10–20 cm, rarely set directly into sculptedbedrock. Both W3202 and 35 Â 50 cm) with 2–3 surviving courses of ma- W3201 were constructed of large granite blocks (up sonry.37 In addition, asmall curvilinear line of to 1 mlong Â 0.5 mwide). The main interior feature stones (W2808) – better described as apartition of shrine 3 was astone platform, covered with clay, than awall, extended W2804 to the west curving atop which was set asymmetrically placedcircular into the northern baulk. As awhole, the construc- ‘altar’ (locus 18), 1,4 Â 0,95 mbuilt againstW3201. tions in the area were rather poorly made – and, The floor of the shrine was constructed of perhaps, poorly preserved – which restricts our packed clay and occupation surfaces throughout ability to provide acomplete account of their layout the room were overlain by athick terminal destruc- or the functionofthe area they circumscribed. tion layer of ash, charcoal, and burned beams that Walls W3001,W3002 and W3004,may com- sealed theunderlying deposits. The beam remains prise the walls of asingle upper phase LB room. included three primary tree specimens: Populus Wall W3004 (locus 18,asexcavated 2.8 mlong, (poplar, 9 samples), Pinus (pine, 5 samples), and 0.45 wide, 3.2 mhigh) was built of fairly regular Betula (birch, 1 sample).39 One sample included a courses of stone masonry.W3005 abuts the east joint with atapered birch fragment wedged into a face of W3004 defining asmallspace between it larger beam of pine. The floor of the shrine was lit- and W3001.38 The enclosed recess (locus 8)was tered with in situ artifactsand features related to a distinguished by the presenceofayellow clay floor. host of practices tied to LBA rituals of divination In the northeast corner of the operation we uncov- and devotion(Smith and Leon 2015). Several fea- ered asubstantial wall fragment (W3002,locus 16, tures appear to be linked to production activities. Near the southern wall of the building (W3301)was alarge (50 Â 45 cm) basalt quern, set into astone 37 Similar trapezoidal constructionswere uncovered elsewhereat pedestal covered with clay. Asmall clay basin be- Gegharot in operations T-2d, T-18,and T-34. low the eastern wall of the pedestalcontained a 38 Asmallirregular semicircle of stones connected the terminus large handstone, as if fallen from the quern. Adja- of W3005 with W3001,however, this line was so poorlypreser- ved that it remains unclear whether the area delimited by W3001,W3004,and W3005 was arecess or fully enclosed room.The very different floor matricesfrom within this recess 39 Wood identifications thanks to Kathryn Seufer and Sturt Man- and the adjacent paved area suggests it was at the very least ning of the Malcolm and Carolyn Wiener Laboratoryfor Aegean adistinctspace. and Near Eastern Dendrochronology at Cornell University. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 183 cent to the north wall of the pedestal were two sels, like censers. The centralspace of the room, complete large jars and fragments of athird vessel between W3202 and alarge pit near the eastern (loci 50, 51, 52). Between W3401 and the basin was edge of the preserved living area, yielded approxi- astone paved conical depression (0.6 Â 0.5 mand mately 45 discrete vessels (Fig. 19–21) includinga 0.55 cm deep), with acircular indentationonthe large churn (Fig. 21,6) and a manghal (Fig. 21,1). bottom that may have once served as astationary North of the altar, atop aplatform was alarge sto- mortar. rage vessel or karas;just south of the altar was a The ceramic repertoire from the shrine was manghal.Inside the altar were twelve complete large and remarkably diverse,ranging from large vessels, including small pots, miniature jars, acen- storage jars to small cups and special purpose ves- ser (Fig. 20,12),two ceramic potstandsor‘idols’

Fig. 23 Gegharot.Shrine 3 materials. 1.Bone objects; 2.Pebbles from altar; 3.Idols from altar 184 Ruben Badalyan et al.

with aflat base, four tapered sides, and corners censers,stone pendants, and aMittanian common pulled into ‘horns’ (Fig. 23,3).The altar also con- style cylinder seal43.Insum, excavations at the LB tained an unusual collection of eighteen smoothed settlementatGegharot have to date exposed three pebbles (Fig. 23,2).40 Two stonependants of flat buildings which,based on their built features and oval pebbles (Fig. 22,6,7) and afragment of acop- artifact inventories, appear to have served as per bracelet were also recovered from within the al- shrines. tar. In addition, paste and carnelian beads (Fig. Giventhe considerable social, political, and 22,2,3) were found both within the altar and in the economic differences between the Early and Late immediately surroundingarea. Amongst the shrine Bronze occupations at Gegharot, it is somewhat assemblage were also asizable collection of stone surprising that the faunal assemblages are virtually and bone tools (includingaweavingcomb) (Fig. identical in terms of the number of identified speci- 23,1) and two bronze arrowheads (Fig. 22,4,5).One mens (Tab. 6).Asinthe EB settlement at the site, of the most strikingfinds from the room was a non-mammalian species are rare from LB levels; ceramic stamp in shape of equilateralcross with a non-mammalian species represented in the LB fau- swastikadesign (Fig. 22,9).Weshould also note nal sample include toads, fish and birds of varying the concentration of more than 20 cattle astragali sizes. Altogether, these specimensmake up less recovered from around the quern pedestal and than 0.1%ofthe faunal assemblage.Among the within the adjacent vessels of loci 50–52.41 Lastly, specimensidentified to genus, sheep and goatsare the original inventory of the buildinglikely included most common, makingupmore than 50%ofthe the Mittanian common style cylinder seal found in assemblage, with cattle making up just under 40%. operation T-20 (locus 4)during the 2006 field sea- The only other type of animal that makesupmore son.42 It is possiblethat the seal was redeposited than 1%isequids, both domesticated horses and as post-depositional formation process destabilized wild onagers. Small mammals (voles, souslik, and portions of the southern wall. hares) are more frequent than in other samples, The ceramic assemblage from the east citadel but many of these are likely to be intrusive and can shrine is typologically assigned to the periodfrom be explained by the fact that the LB levels at Ge- the end of the 14th through the first half of the 13th gharot are closertothe surface than the EB depos- cent. BC. Radiocarbon determinations largely con- its. Comparison of NISP between shrine contexts firm this dating of the final destruction event, which and between shrine and non-shrine loci suggests also razed the west terrace sanctuary in operation that there is no difference in the proportion of ani- T-2Eand the workshop in operation T-15. mals found within shrines and outside of them.Nor Asignificant number of finds from the east ci- are there any recognizable differences between fau- tadel shrine echoes the assemblage recovered from nal assemblages from the west terrace, east citadel, the first shrine discoveredonthe west terrace, in- and west citadel. cluding ceramic potstands/‘‘idols’’, ceramic stamps, Past examinations of survivorship at Gegharot indicated that less than 25%ofthe herd was surviving until physical maturity.44 Thesepatterns were 40 The exclusive concentrationofthese pebbles within the altar, their absence in the culturallayer and their heterogeneous interpreted as anon-viable herding strategy, indi- compositionofandesite,basalt, basaltic andesite, siliceous catingthat the inhabitants of Gegharot were being mudstone,jasper, sandstone, dacite, quartz,limestone(ana- supplied with meat. An updated survivorship analy- lysis conducted by the Institute of Geological Sciences of Arme- nia) – indicates that they do notbelong to the local geological sis, however, is noticeably higher with over 33%of substrate and were thus collected and intentionallyrelocated the herd surviving until maturity (see Tab. 3),apat- to Gegharot.Nine pebblesare undoubtedly polished artificially. tern that still appears to indicate anon-viable herd- 41 All three shrines contained alarge numberofcattle astragali, manyofwhich were striatedonone side andoccasionally bur- ing strategy. Thus, it seems that the LB inhabitants ned. Smithand Leon (2015)notethat in both shrines 1 and 3, of Gegharot were being supplied with meat in some aratio of approximately 64%left astragali to 36%right astra- form. Sheep/goat survivorship also appears to differ gali indicates aroughly 2 : 1 preferencefor curating astragali from the left side of the animal.Asimilar propensity for left between shrines and areas outside of the shrines. astragali was foundinshrine 2.The emphasis on left astragali In Stage C, the age at which animals reach maturity, from thesecontexts is all the more striking, when taken along- survivorship in the three shrines combined45 is side the cattle astragali from non-shrine contexts where the ratio of left to right astragali was roughly 1 : 1.Clearly cattle 22.45%, while in the non-shrine contexts survivor- astragali, and especially an asymmetrical ratio of left and right, playedarole in the ritual activities associated with the shrines. Garth Gilmour (1997)and others (Koerper/Whitney-Desautels 43 Badalyan et al. 2008,Fig. 2,f,g,h; fig. 24,m,n;fig. 25,b;Fig. 1999, 74–75;Bejenaru et al. 2010)have notedthat astragali 23,i; Fig. 24-k,i. in Near Easternand Aegean contexts are often associated with 44 see Monahan 2012:Badalyan et al. 2008. ritual, where the rolling of the bones as dicewould have 45 Neither of the shrinesonthe terrace had asufficiently large functioned as adivinatory activity. Indeed divination appears sample to reliably calculate survivorship, but the presumed si- to have been acritical aspect of devotional ritual at Gegharot. milarity of theirfunction seemed to suggestthat they couldbe 42 Badalyan et al. 2008, 72–73,Fig. 25a. constructively compared to areas outsideofthe shrines. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 185 ship is noticeably higher at 29.33%. It seems, then, that the shrineswere more likely to have received Gegharot Late Bronze juvenile animals, which may reflectnot only their NISP %NISP %NISP to genus greater social and economicvalue, but also their augmented cultural importance as sites of religious Bufonidae Toads 40.01% and divinatory rituals. Body-part representation, how- Squamata Snakes 00.00% ever, withinthe shrines and outside of theshrines Fish 20.00% is the same, so theshrines were not getting differ- Large bird 13 0.02% ent cuts of meat than the non-shrine areas. Cattle were living longer than sheep and Mediumbird 20.00% goats during the LB, with nearly 70%surviving until Small bird 20.00% physical maturity. It is likely that cattle were more Indeterminate 414 0.68% frequently used for their secondary products, such Large mammal 2954 4.87% as milk and traction,than were sheep and goats, Medium mammal 3081 5.08% which were primarily ameat source. Differences also appear between shrine and non-shrine con- Small mammal 277 0.46% texts, but they are the reverse of what is seen Artiodactyl 10.00% among sheep and goats. In shrine contexts 75.00% Largeartiodactyl 50.01% of the animals are surviving past Stage C, or into Medium artiodactyl 20.00% full physical maturity (Tab. 7).Innon-shrinecon- texts, just under 70%ofthe animals are surviving Bovid 42 0.07% to the same age. As with sheep and goats, body Bos Cattle 5167 8.52% 37.51% part representation is roughly similar between Capra Goat 275 0.45% 2.00% shrine and non-shrineareas at the site, although Ovis Sheep 1378 2.27% 10.00% there are more astragali in shrines as discussed Gazella Gazelle 16 0.03% 0.12% above. Thus the picture presented by the LB fauna at Gegharot is complicated. The types of animals Ovis-Capra-Gazella 20 0.03% presentand the body parts of these animals are Ovis-Capra 6528 10.76% 47.39% more or less the same in the shrines and outside of Cervid 66 0.11% the shrines, but the ages of the animals differmark- Cervus Red deer 27 0.04% 0.20% edly in the two different areas. Dama Fallow deer 20.00% 0.01% The plantregimes of LB Gegharot appear to be broadly similar to those documented in the EB Equus 157 0.26% 1.14% levels at the site (Tab. 4).The maincrops for LB Equus caballus Horse 28 0.05% 0.20% Gegharot were: cultivated barley (Hordeum vul- Equus asinus Donkey 00.00% 0.00% gare), common bread wheat(Triticumaestivum Equus hemionus Onager 10.00% 0.01% ssp. vulgare), and club wheat (Triticum aestivum ssp. compactum)aswell as emmer (Triticumdicoc- Equus asinus/hemionus 16 0.03% 0.12% cum)and rye (Secale sp.), two cereals that were Sus Pig 73 0.12% 0.53% less numerous in EB samples. The barleytowheat Carnivore 10.00% ratio for the LB samples is 75%to25%. Two spe- Canid 10.00% cies well represented in LB samplesbut, to date, Large canid 10.00% absent from EB layers are broomcorn millet (Pani- cum miliaceum)and grape (Vitis vinifera). It is im- Canis familiaris Domesticated dog 20 0.03% 0.15% portant to note that both of these speciesprefer Canis lupus Wolf 30.00% 0.02% warmer temperatures (additionally, millet thrives in Vulpes Fox 17 0.03% 0.12% more humid climates) and thus cannot grow in high Meles Badger 20.00% 0.01% mountainous areas such as the Tsaghkahovit Plain. Their appearance at Gegharot thus in all probability Lepus Hare 30.00% 0.02% is evidence of exchange relationswith the Ararat Rodentia 12 0.02% Plain to the south. According to an earlier recon- Microtus Vole 15 0.02% 0.11% 46 struction, millet appeared in the territory of Arme- Spermophilous Souslik 35 0.06% 0.25% nia during the Early Iron Age and spread widely Total 13914

Tab. 6 Faunal assemblage for Gegharot LB contexts (2007–2011 field seasons) according to Number 46 Hovsepyan 2009. of Identified Specimens. 186 Ruben Badalyan et al.

1.6 ha citadel,48 although significant spoil heaps Stage AStage BStage C precludeacomplete understandingofthe damage. Gegharot Late Bronze 87.00% 80.77% 69.84% Surface collections during the ArAGATSpedestrian TsaghkahovitLowerTown Late Bronze 95.00% 89.90% 63.49% survey included ceramics from the Early and Late Bronze Age.49 Excavations at four separateopera- tions on the site’s northern and eastern terraces in Tab. 7 only during the Urartian era. Thus the recentar- 2006 and 2008 revealed significant Early and Late Cattle survivorship per- chaeobotanical investigations at Gegharot and BronzeAge occupations at the site. centages for Project ArAGATS Bronze Age Tsaghkahovit demonstrate that broomcornmillet contexts (2007–2011 was knowntolocal populationsofthe SouthCau- field seasons) casus at least as early as the LBA, even if it was The Early Bronze AgeatAragatsi Berd not ahigh volume staple. Certain weedspecies, such as representatives of the Boraginaceae, Poly- While no in situ EB levelswere exposed at Aragatsi gonaceae, Rubiaceae and Cyperaceae are well-at- Berd in 2006 or 2008,significant amounts of EB tested in the LB layersofGegharot. potterywere recoveredinlayers of colluvial over- burden in all excavation operations. On the whole, these materials were typologically homogeneous, Excavations at Aragatsi Berd assignable to the ‘‘Karnut-Shengavit’’ group of the later EB. They included standardcups with tripartite The Aragatsi Berd excavationsfollow the site’s initi- bodies, hemispherical bowls, pots, jugs, and sto- al documentation by researchers from the Institute rage jars. The dominantornamental style utilizes of Archaeology and Ethnography and Project ArA- horizontal circumferential belts filled with geometric GATS.47 Set atop atall (80–100 mhigh, 2175.6 m. motifs.Two samples are particularlynotable: (1)a a.s.l.) conical outcrop, Aragatsi Berd is located body fragment of alarge jar (AB4,locus 7), which 1.5 km northofthe village of Alagyaz and 5.86 km held traces of bitumen on an upper horizontal, cor- southeast of Gegharot (the two sites are not inter- rugatededge (see the discussion of bitumen from visible, obscured by the western cliffs of the Sipan Gegharot above), and (2)abowl fragmentwith a Canyon) along the Yerevan – Vanadzor road. It also vertical handle that starts at the rim. The bowl sits several kilometers south of the southern en- sherd features ageometric pattern that extends trance to the SpitakPass that leads to the Spitak – along the entire handle. An analogoushandle, fea- Pambak Valley. In geological terms, theAragatsi turing similarform and decoration, was recovered Berd hill is composed of foldedpalaeotypalande- from Sos Ho¨yu¨kIVa.50 site basalts with abase diameter of about 500 m Asingle EB radiocarbon determination from and atotal surface area of approximately 25 ha. operation AB1 (locus 6), the initial 4 Â 4 msound- The slopes of the hill are steeply graded (from 15– ing at the site carried out in 2006,indicated adate 25%), while the crown has been cut by Bronze Age range of 3100–2870 BC (2 sigma) (Tab. 8).Based terraces and features excellent sight lines for sev- on the predominance of LBA artifacts in the opera- eral kilometers south along the Kazakh River, north tion however (see below), this date is most likely through the Sipan Canyon, and southwesttoward the result of EB deposits having been pulled into Tsaghkahovit Fortress and Mt. Aragats. On the an LB space during terrace construction and use. north side of the hill, five terraces, each approxi- Also of note for the EB investigations at Ara- mately 10 mwide, are visible. On the western, gatsi Berd were theexcavations in the eastern half southern, and eastern sides of the hill two major of operation AB4,a3Â5munit located on the terraces are visible, averagingbetween 10 and eastern portion of the second terrace (Fig. 24,d). 20 minwidth. The second highest terrace on the Loci 7, 11,and 14 in this sector of the operation crown circumscribes the entire hill, largely uninter- encountered dense depositsoffallen wall cobbles rupted, and constitutes the first visible terrace on and silty soil that permitted deeper excavation than every side save the northern. It is this ‘‘second ter- in the western AB4 precincts, which all met bedrock race,’’onwhich all of the excavation units de- (see below). The collected materials also featured scribed below were located. higher frequencies of EB pottery than the western During the late Soviet Period atelevision re- half of the operation and eventually gave way to lay antenna was installedontop of the site, which undifferentiated EB culturalfill. Ahorn fragment destroyed much of the architecture and cultural deposits visible in earlier aerial photographsofthe 48 Smith et al. 2009,Plate 42. 49 Avetisyan et al. 2000:30;Badalyan/Avetisyan 2007, 62–65; 47 Avetisyan et al. 2000, 30;Badalyan/Avetisyan 2007, 62–65; Smith et al. 2009, 281–286. Smith et al. 2009, 281–286. 50 Sagona 2000,Fig. 17.7. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 187

Site Operation Material Lab 14CDate (BP) Calibrated Date (BC) Number 95.4%Probability Aragatsi Berd, 2006 AB1,l.6 charcoal AA-72050 4323 Æ 51 3091-2878 Aragatsi Berd, 2008 AB2,l.13 charcoal AA-827913078 Æ 35 1424-1265 Aragatsi Berd, 2008 AB3,l.12 charcoal AA-827923040 Æ 35 1410-1209 Aragatsi Berd, 2008 AB3,l.16 charcoal AA-827933005 Æ 36 1383-1128 Aragatsi Berd, 2008 AB3,l.19 charcoal AA-82790 3037 Æ 50 1416-1129 Aragatsi Berd, 2008 AB3,l.22 charcoalAA-82794 2946 Æ 35 1291-1027 Aragatsi Berd, 2008 AB4,l.3 charcoalAA-82795 3060 Æ 40 1426-1213 Tab. 8 Radiocarbon determina- Aragatsi Berd, 2008 AB4,l.6 charcoal AA-827882774 Æ 66 1113-807 tions related to Aragat- si Berd Early and Late Aragatsi Berd, 2008 AB4,l.6 charcoal AA-86868 2958 Æ 46 1369-1018 Bronze Agecontexts Aragatsi Berd, 2008 AB4,l.6 charcoal AA-868692877 Æ 46 1212-922 from the 2006–2008 field seasons (calibra- Aragatsi Berd, 2008 AB4,l.8 harcoal AA-827892949 Æ 51 1370-1008 tions by OxCal v. 4.1.7 Bronk Ramsay 2010) from azoomorphic EB andiron, similartoothers sers, and manghals were also collected.Other LB collected from surface and colluvial strataatAragat- materials of note include stone tools (basalt hand- si Berd, was recovered as well. An LB foundation stones, avariety of hammerstones, four partialob- wall (W401,locus 4)situated at ahigher elevation sidian projectile points, and one unifacialLBjewelry in the operation’s western sector, did not extend mold fragment), bone awls and punches, ceramic into this downslope area of the excavation, likely disks, and numerous incised and drilled astragali of having eroded away. Thus, the depositional context cattle, in addition to large collections of paleozoolo- closely resembles EB spaces uncovered on the Ge- gical and paleobotanical material. gharot western terrace and citadel, which have The materials from operation AB3,a6Â3m been exposedimmediately downslope and beneath excavation unit opened in aflat area on the northern eroded LB spaces.51 Due to time constraints, the extent of the second terrace, provide the most re- excavation of the eastern sector of AB4 was not vealing and substantial informationabout the site’s completed during the period of excavations de- LB occupation. Surface architecture visible within scribed in this report. the operation was confirmed to constitute several superimposed wall segments (W301 and W302,locus 4), forming two triangular architectural spaces positionedagainst the ‘‘outermost’’ builtfeature of The Late Bronze Age at Aragatsi Berd the terrace (Fig. 24,c).These terracing features were faced on the citadelside, where theexcavation Excavations at all four operationsatAragatsiBerd matrix consistedofsilty ash with high frequencies of revealed in situ architectural and material remains carbon particles. To the south of this tangle of built attributable to the Late Bronze Age and collected a contexts, two LB features were uncovered. significant amount of ceramic materials (n > 600). The first was a 1.5 mmeter deep pit positioned LBA pottery at the site is dominantly affiliated with immediately south of the architecture (Pit 301,loci the LB II and III periods. Only four potsherds, all 18–24)and overlain by flat paving stones, burned from operations AB1 and 2,can be assigned to the beams, charred and darkenedearth, aspiral bronze LB Iphase. The assemblage consists of short and ring, acollectionofstriated and unstriated cow as- tall-necked jars (often with high shoulders), globu- tragali, and abroken-in-place manghal (Fig. 25,1).52 lar pots,shallow bowls, basins, and larger storage Double-ended,ovoid ceramic objects, manghals are vessels characteristic of the LB. The majority of the open on both (long) ends, like the censersrecov- assemblage is decorated with incised, circumferen- ered from the LB shrinesonthe Gegharot western tial motifs consisting of lines, belts, bands, and ob- terrace.53 But whileone orifice is wide and follows lique hatching and is dominated by black and gray continuously from the vessel walls, the other end of reduced-fired wares.Significantproportions of buff the vessel features anarrower, slotted orifice.Also and red pottery are also included, however. Sherds from special vesselssuch as mini jars, cups, cen- 52 The manghal from Aragatsi Berd is 41 cm wide at the base, 54.5 cm wideatthe top, and 32.5–34.5 cm tall. The basal aperture is 30 cm long and 5 cm wide. 51 `àäàº'í/ÑìŁò 2008;Badalyan et al. 2008. 53 Badalyan et al. 2008, 66 Fig. 21. 188 Ruben Badalyan et al.

Fig. 24 Aragatsi Berd. Plan and section views of operations AB1–4

characteristicofthese slab-built objects, but not al- 2 cm in diameter 15 cm below each ‘‘rim’’, the Ara- ways present, is aperforation halfway up each of gatsi Berd example features five pairs of impressed the sides, so that air could fuel pyrotechnical activ- grooves on opposing sides. Manghals often feature ity or acord or rod be insertedthrough the vessel additional evidence of pyrotechnical activity in the lengthwise to support it, hang it, or suspend some- form of carbon smudges and cloudsonboth interior thing inside of it. In addition to lateral perforations and exterior surfaces (see the two examples from Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 189

Fig. 25 Aragatsi Berd. 1.LB manghal; 2.unifacial jewelry mold recovered from Pit AB3–01 190 Ruben Badalyan et al.

operation T-32 at Gegharot above, which feature al- of LB ceramics were recovered in its immediate vici- most identical smudging patternstothe Aragatsi nity. The floor itself formed acontinuous surface, Berd example). reaching to the lip of the AB3 pit (301). Multiple The ‘‘paved’’ pit underlying this collection of lenses of black carbon and ash were distributed materials, composed of inward-sloping bedrock over the entire length of the floor, terminating in faces, is afeatureencounteredinLBcontexts exca- the burned beams sitting atop the Pit 301 pave- vated on the Gegharot western terrace and cita- ment. Analysis of radiocarbon samples collected del.54 Towardsthe center of the feature, the paving from both the floor and pit features in AB3 date the slabs (locus 17)slumped lower,having sunken complex to the fourteenth century BC (Tab. 8). from their original positionatthe lip over the mil- Operation AB4,a5Â3mexcavation unit on lennia (Fig. 24,c).The pit’s silty ash matrix included the eastern portion of the second terrace, revealed moderate frequencies of bone, including additional additional LB built features and activity areas cow astragali, the mandible and teeth of asingle (Fig. 24,d).Atthe southern edge of the operation, dog,horse vertebrae, asingle pig tooth, obsidian an LB foundation wall (W401,locus 4)and paved debitage, LB ceramics, and aunifacial jewelry mold floor (locus 6)were uncovered merely 50 cm below (Fig. 25,2).55 The mold was discovered with its the surface. The wall location corresponded exactly non-engraved side up, approximately 10 cm away to the placement of architecture suggested by sur- from acomplete cow horn, and featured aperfora- face rockiness and mounding and was composed tion in one corner, likely used to either attach a primarily of rectangular, single-faced blocks,aver- cover or face plate to the mold or to positionitdur- aging 40 cm in length (the position of W401 against ing casting. Ashallow divot in the oppositecorner the southern operation baulk made it difficult to as- would have been used to more securely articulate sess the presenceofsingle-ordouble-faced the mold with aplate. In its design forms and size blocks). the Aragatsi Berd mold most closely resembles the The paved floor was found to abut the foun- mold recovered from operation T-15 in the Gegha- dation wall to the north, covered by a 2 cm layer of rot citadel,56 although it appears to be made of ar- pasty, yellow clay. It didnot host asignificant gillite. The designs, includingbars, spoked wheel amount of materials on its surface, but all of the shapes,a‘‘tree’’ of circularbeads,zigzag decorated diagnostic ceramic materials collected were asso- cylinders, abi-conical decoration, among others, ciated with the LB. Analysis of carbon samples col- find analogies in both the Gegharot T-15 mold and lected from the floor date the LB complex to the the Gegharot T-2Emold discovered in the western thirteenth century BC (Tab. 8).The paving stones terrace excavations.57 The Aragatsi Berd mold was were subsequently removed in order to determine discovered at an elevation where the sloping pit whether the floor represented merely the most re- walls stopped their inward taper and began anear cent level of site occupation, or the only (LB) occu- vertical cylinder. This cut continued down to the pation level in the western half of the operation. pit’s base, between 20 and 30 cm in total cylindri- Two pits were exposed beneath thepaving cal depth. stones: one againstthe western operation cut (Pit The second significant AB3 feature consisted 401,locus 8 western sector),running into the of alevel bedrock floor surface (locus 6)covered baulk, and asecond(Pit 402,locus 8 eastern sec- lightly by several millimeters of yellow clay, upon tor) on the eastern side of aprojectionofcrumbly, which sat acarved, freestanding basalt basin. red bedrock, expanding eastward towards the cen- Placed at the eastern edge of the floor, the basin ter of the operation. Pit 401 contained silty, ashy measured 50 cm tall and 40 cm in diameter. Such loam and no artifacts of note; most of the feature installations, which could serve any number of appeared to lie within the operation’s western working, grinding, or washing purposes, are com- baulk. The larger Pit 402 produced moderate mon at LB fortresses on the Tsaghkahovit Plain and amounts of ceramics and high frequencies of stone elsewhere throughout Southern Caucasia. Similar tools: seven in total, including hammer-stones, flat examples have been recoveredfrom contexts at anvil-like slabs, and two basalt handstone frag- both Gegharot and Tsaghkahovit fortresses.58 Three ments. Alarge, perforated, triangular bone object, groundstone tools, as well as moderate quantities perhapsafitting or an irregular spindle whorl, was also recovered there. Most interesting, however, was the discovery of amini LB jar in two halves, 54 See Badalyan et al. 2008,Figs. 5 and 9. just north of W401.The bottom half was exposed 55 The mold measures 7.0 Â 5.5–6.0 Â 0.7 cm and weighs 45.2 approximately 20 cm southeast of the top portion. g. Such ‘‘mini vessels’’have been recovered from con- 56 Badalyan et al. 2008:64,Fig. 19. 57 Badalyan et al. 2008:71,Fig. 24. texts on the Tsaghkahovit and Gegharot terraces, 58 `àäàº'í/ÑìŁò 2008. as well as the Gegharot citadel(see above). Like Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 191 the paved surface overlaying it, radiocarbondeter- poseful deposition of the cow suggests aceremo- minations associate Pit 402 with the thirteenth cen- nial closure of the pit, perhaps associated with the tury BC (Tab. 8).Due to time constraints, the exca- abandonment of the room or its turnover to new vation of this pit was not completed during the uses. fieldwork describedinthis report. The botanical remainsfrom LB Aragatsi Berd Finally, operationsAB1and 2,adjacent 4 Â 4 m also proved illuminating, although the total carpolo- excavation units on the northeastern side of the gical sample recovered(162 units) is small for accu- second terrace separated by a 1 mbaulk, provided rate statistical analysis. Therefore,the ratios of uni- glimpsesofthe LB retaining walls and terracing ar- dentified cereals, wheat,barley, and weeds chitecture employed at the site, as well as artifact presented here remain provisional. The highest con- caches purposefully deposited when this sector of centrationofarchaeocarpological material was re- the fortress went out of use (Fig. 24,1.2).Excava- corded for Pits 201 and 301.The cultigens-to- tions in AB1 exposed aten-course retainingwall weeds ratio for LB Aragatsi Berd was 35%to (W101,locus 3)set on and againstacarved bench 65%, and the wheat-to-barleyratio was 45%to of hillside bedrock and directlyassociated with a 55%. Thus, preliminary data suggest that barley floor surface built of leveled bedrock.The curvi- predominated over wheat during the site’s LB occu- linear wall turned northeast at the operation’s east- pation. Four cultivated cereals (Poaceae) are pre- ern edge, articulating with ashorter wall segment sent in the Aragatsi Berd dataset: common bread (W102,locus 11)orientedroughly west-east as it wheat (Triticum cf. aestivum subsp. vulgare), club passedbetween the northern and eastern operation wheat (Triticum cf. aestivum subsp. compactum), baulks, just at the terrace edge. The trajectory of round grained wheat (Triticum cf. aestivum subsp. these walls off the terrace suggested that the rest sphaerococcum), and cultivated barley(Hordeum of the room had eroded off the side of the hill. The vulgare), at least aportion of which are hulled.Cer- bedrock floor in AB1 (loci 6 and 14)featureda eals are therefore the only field crops found at Ara- small in situ collection of artifacts purposefully gatsi Berd, which is not uncommon among archae- placed beneath aboulder 50 cm in length, includ- obotanical investigations of Bronze Age sites in the ing groundstone tools, manghal sherds, acache of region. astragali, ceramic disks,aspindle whorl, and a In addition to field crops, the composition of bone spatulate tool. weedy flora recoveredatAragatsi Berd is generally Excavations in AB2 confirmed indications from common for Bronze and Iron Age Armenia. The only AB1 that the terrace architecture of Aragatsi Berd exception is the presence of Asperugo cf. procum- consisted of aconfiguration of curvilinear, s-shaped bens –– aboraginaceous ruderal and sometimes se- retainingwalls built against hillside bedrock that is getal weed, which is common for locations of hu- similar to the known LB terrace preparations at Ge- man activity. Its presenceatLBAragatsi Berd gharot and Tsaghkahovit.59 Two wall segments marks the first instance of its documentation in the were exposed (W201 and W202,loci 4 and 5), once archaeoflora of Armenia. Cyperaceae family species again abuttingthe rear of the terrace (Fig. 24,b). (Scirpus and Carex)are also present at Aragatsi They sat on aprepared bedrock surface much like Berd and indicate the existence of nearby wetlands the floor in AB1.These wall sectionslikely con- ecosystems/humid conditions during the LBA. Lim- nected to each other within the AB2 southernbaulk ited quantities of weed seeds belong to Chenopo- and to the AB1 retaining wall across the operation’s dium,Brassicaceae, Lamiaceae,and Poaceae fa- eastern baulk. milies, suggesting possible dietary supplements in At the northernmost AB2 margin the bedrock addition to cereals. floor gave way to alower progression of cultural Like the paleobotanical collection, the Aragat- fill, an area subsequently identified as apit ap- si Berd faunal sample is one of the smallestana- proximately 1.75 mindiameter(Pit 201,loci 12 lyzed samples from the Tsaghkahovit Plain. In and 13). The bell-shaped feature contained articu- terms of the taxa represented, it is consistent with lated portionsofacow spineand legs approxi- the other LB faunal samplesfrom the Tsaghkahovit mately thirty centimeters below the lip. The pit fill Plain, but in kill-off patterns, it seems to differfrom includedceramic disks and most of the diagnostic them in important ways. pottery was associated with EBA. However, asingle Among the specimensidentified to the level radiocarbon date from the pit (Tab. 8) has affiliated of genus, the Aragatsi Berd sample is composed the matrix to the LBA, suggestingitwas filled with primarily of sheep and goats (Tab. 9),aswith all EB debris, but last used in the LBA. The rather pur- the other samples from the LBA Tsaghkahovit Plain. The proportionofsheep/goatsisslightly higher than at other LB sites and the proportion of cattle 59 Badalyan et al. 2008,Fig. 9;Smith et al. 2004,Fig. 6. is slightly smaller, but these differencesare slight. 192 Ruben Badalyan et al.

AragatsiBerd Late Bronze Tsaghkahovit SLT Late Bronze NISP %NISP %NISP NISP %NISP %NISP to genus to genus Bufonidae Toads 00.00% 00.00% Squamata Snakes 00.00% 00.00% Fish 00.00% 00.00% Large bird 20.16% 00.00% Medium bird 00.00% 30.06% Small bird 00.00% 00.00% Indeterminate 602 46.81% 145 2.69% Large mammal 130 10.11% 1229 22.81% Medium mammal 307 23.87% 732 13.59% Small mammal 20 1.56% 55 1.02% Artiodactyl 00.00% 00.00% Large artiodactyl 00.00% 00.00% Medium artiodactyl 00.00% 00.00% Bovid 20.16% 26 0.48% Bos Cattle 49 3.81% 23.44% 1479 27.45% 46.52% Capra Goat 50.39% 2.39% 100 1.86% 3.15% Ovis Sheep 18 1.40% 8.61% 334 6.20% 10.51% Gazella Gazelle 00.00% 0.00% 10.02% 0.03% Ovis-Capra-Gazella 50.39% 70.13% Ovis-Capra 122 9.49% 58.37% 1099 20.40% 34.57% Cervid 90.70% 50.09% Cervus Red deer 00.00% 0.00% 70.13% 0.22% Dama Fallow deer 00.00% 0.00% 10.02% 0.03% Equus 30.23% 1.44% 73 1.36% 2.30% Equuscaballus Horse 00.00% 0.00% 15 0.28% 0.47% Equus asinus Donkey 00.00% 0.00% 40.07% 0.13% Equus hemionus Onager 00.00% 0.00% 00.00% 0.00% Equus asinus/hemionus 00.00% 0.00% 10.02% 0.03% Sus Pig 20.16% 0.96% 17 0.32% 0.53% Carnivore 00.00% 00.00% Canid 00.00% 10.02% Large canid 00.00% 00.00% Canis familiaris Domestica- 90.70% 4.31% 00.00% 0.00% ted dog Canis lupus Wolf 00.00% 0.00% 00.00% 0.00% Vulpes Fox 00.00% 0.00% 00.00% 0.00% Meles Badger 10.08% 0.48% 90.17% 0.28% Lepus Hare 00.00% 0.00% 00.00% 0.00% Tab. 9 Rodentia 00.00% 30.06% NumberofIdentified Specimens for Aragatsi Microtus Vole 00.00% 0.00% 31 0.58% 0.98% Berd andTsaghkahovit Spermophilous Souslik 00.00% 0.00% 10 0.19% 0.31% SLT LB contexts (2007–2011 field Total 1286 3223 seasons) Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 193

However, among those specimens so identified, the suggest that the occupation of AragatsiBerd in that sheep to goat ratio at Aragatsi Berd is 1.6:1,con- period covered the entire upperportion of the hill siderably lower than the other LB sites, which range and that the citadel and terraces were intenselyoc- from over 5:1 and never drop below 3:1.Without cupied. The later datingofLBdeposits in AB4 indi- further information aboutthe differences in the use cate that the eastern terraces, as well as immedi- of sheep and goats in the LB plain, it is not possi- ately adjacent sub-divisions running alongthe first ble to definitively interpret this pattern. Outside of southern terrace, may have been constructed and Aragatsi Berd, the lowest sheep to goat ratio is at occupied in the later LB, while the second northern the Gegharot citadel. terrace appears to have been occupied earlier. The Survivorship of sheep and goats is also simi- destroyed fourteenth century room in AB3 exhibited lar to that at other sites on the Tsaghkahovit Plain. many of the features and materials familiar from in- Just over 40%ofthe herd reaches adulthood, a stitutionalritualand storage contexts exposed at pattern which, at EB Gegharot has been interpreted Gegharot and Tsaghkahovit, as wellasmore distant as an example of localized productionfor localized fortresses such as Metsamor in the Ararat Valley. consumption. Thus, it seems that the pastoralists at Ongoing research with the materials from Aragatsi Aragatsi Berd were not pursuingaspecialized pro- Berd is intended to delineate exactly how the for- duction strategy, but rather practicing ageneralized tress was incorporated into the broader political- herdingstrategy for localized consumption. economic rhythms of LB social life on the ancient There is anoticeable difference between the plain. Aragatsi Berd sample kill-offcurve and those of the Tsaghkahovit citadel and Gegharot western terrace, which are the only two assemblages with reliable samples. For both of the latter two contexts,more Excavations of the Late Bronze Age than half the herd was killedoff by Stage D, which Settlement at Tsaghkahovit is the pointatwhich animals reach maturity. In these samples, kill-off after this stage is very low. In the AragatsiBerd sample, kill-off at stage Dis50 The monumentality and vertical relief of LBA for- percent, leaving agreater proportionofthe herd to tresses such as Gegharot, with well-demarcated live to maturity. To alimited extent, this pattern is spacesofritual practicesand metal production, also true for kill-off curvesgeneratedusing epiphy- stand in marked contrast to LBA residential settle- seal fusion. In the first stage, at about one year of ments.Excavations in 2010 and 2011 at the resi- age, kill-off is noticeably less than in any of the dential complex at the base of the Tsaghkahovit other four samples. In the final stage, aboutthree fortress, designated the Tsaghkahovit South Lower years of age, kill-off is slightly greater than the Town (SLT), advanced our overall goal of under- Tsaghkahovit Residential Complex, but lowerthan standing the impact of new fortress-centered politi- the other three samples. The Residential Complex cal institutionsongrassroots populations. Addition- kill-off pattern was originally interpreted as the sin- ally, the settlement’s site structureuncovered gle example of aself-sustaining herding practice. during our 2003/2005 excavations60 raisedimpor- The low kill-off of young animals in the Aragatsi tant questions about the potential for seasonal resi- Berd sample suggests that this might also be an in- dential mobility of the occupants of the LBA settle- stance of aself-sustaining herd. ment, which we have begun to resolvethrough a Despite these intriguing initial data, the Ara- combination of amagnetic survey61 and further in- gatsi Berd sample size remains too smalltodraw tensive excavations. Overall, our results thus far de- reliableinferences at present and all of the above monstrate asubstantial energetic investmentinset- trends should be noted as tentative and prelimin- tlementarchitecture concurrent with repeated ary. ephemeral (likely seasonal) uses of the settlement In summary, excavations to date at Aragatsi space. The evidencefrom site-structure and materi- Berd have confirmedthe EB and LB occupations of al culture resulting from our 2010/2011 investiga- the site, identified theirrelative chronological se- tions at SLT support an interpretation of acommu- quence, and provided adetailed evaluation of the nity engaged in amixed farming and herding damagetothe citadel. EB investigations, while re- economy whospent portions of the year on the strictedtolevels of colluvium and cultural fill, have move but returned to the site repeatedly. In this confirmed the existence of achain of small EB set- sense, the TsaghkahovitSouth Lower Town repre- tlements along the southern Pambak slopes that sents an important case study in the complex set- extended from Jarjaris to Gegharot. Their close placement and intense occupations indicate the high 60 Badalyan et al. 2008. densityofthe plain’s EB settlement. LB exposures 61 Lindsay et al. 2010. 194 Ruben Badalyan et al.

tlement and subsistence patterns of mobile pastor- likely servicing as foundations for an organic super- alists.62 structure such as felt tenting or wattle-and-daub. Our 2010 and 2011 excavations were espe- Examples of the short, single- and double-course cially focused on tracing the settlement architecture walls are visible around an oblongcentral space in uncovered in 2005 to broaden our scope of the operation SLT 14,another ovoid room spanning the economic activities represented at the settlement southwest corner of SLT 9 into SLT 10 (walls la- and to trace the stratigraphicrelationship between beled W1401,W0901,and W0101 in Fig. 26),as the primary LBA occupation and the subsequent well as rooms exposed in operations SLT 1 and Iron 3 occupation noted in portions of the settle- 5.64 The type of roofing material used to shelter ment during prior field seasons. We therefore these irregularly-shaped rooms is still unclear, but a opened five large units to the south of our 2005 series of small pits (20–50 cm in diameter) dug into exposures;in2010 we opened operations SLT 9 the centralfloor space of SLT 9,including one lined (7 Â 7 m), SLT 10 (7 Â 7 m), SLT 11 (5 Â 7 m), and with aring of recessedcobbles, suggest they may in 2011 we initiated operations SLT 13 (6 Â 7 m), have once supported structural columns or tent- and SLT 14 (6 Â 7 m) (Fig. 26).Two other ancillary posts. operationswere conducted in the vicinity of the lar- In addition to the loose structure of the site ger trenches; SLT 8 was a 2 Â 2 mtest trench exca- plan, evidence for rearrangement of wallsand stra- vated during the course of the 2008 magnetic sur- tified floor levels signal episodes of reoccupation vey,63 and SLT 15 was arobbed Iron 3 tomb within the LBA. In operation SLT 14,many of the exposedwhile cleaning the baulk near the north- basalt wall stones of W1401 are faced on one side, west corner of SLT 13. but the faces are jumbled,suggesting the walls The architecture uncovered in the five primary stones were recycledfrom prior occupational operations (SLT 9–11, 13, 14)offer further evi- phases. In addition, walls associated with different dence of the LBA construction style of cutting into floor levels abut and overlap each other, reinforcing bedrock in preparation for building and using ba- the impression of remodeling prior residential salt outcrops in wall construction and for stationary spaces to fit the changing needs of the group in mortars. Indeed, the perspective of four seasons of subsequent occupations. In operations SLT 9, 10 excavations suggests that LBA builders went to and 14,stratifiedfloor levels were visible in profile, great effort in leveling the natural west-east grade alternatingwith thin layers of floor prep; in SLT 9, of the settlement area into at least three distinct the floors thicknessesranged from 2–5 cm sepa- vertical elevations. Illustrating these descending le- rated by floor prep, while in SLT 14,the strata be- vels of prepared surfaces, the prepared bedrock tween floors were separated by larger intervals be- surface in the western-mostoperation of SLT 11 (lo- tween 10 and 40 cm apart, perhaps indicatinga cus 4)is2142.1 m, the primary prepared surface in longer interval between occupations. However, neighboring SLT 9 (locus 15)is2141.0 m, and in three radiocarbon determinations (AA-95623,AA- SLT 13 (locus 10)is2138.8 m; even with some un- 95624,AA-95626)from charcoal samplescollected evenness in the bedrock surfaces, we can observe from threevisible floor levels in SLT 14 all returned at leasta1mdrop between the prepared bedrock dates in the LB II phase between the last quarter of surfacesinSLT 11 and 9 to the east, and at least a the 15th century and the mid-14th century BC (Tab. 2 mdrop between SLT 9 and SLT 13. 10),which suggests that the time between thefloor After this initial expenditure of energy prepar- layers can be measured in generations rather than ing the space for construction, however, compara- distinct horizons. If this assessment is correct, it tively little formal planningappearstohave been further supports our hypothesis that portions of the put into much of the architecture that might give domestic complex were occupied intermittently by one asense of asustained occupation. Indeed, one seasonally mobile groups. of the most strikingaspects of the layout of the Whenthe settlement was in use, the features complex is the informality of the room plans. and artifacts recorded reinforce the domestic scale Rooms in the settlementare generally ovoid in of production and consumption, while providing shape, constructed of variably-sized, poorly worked further insight into the rangeofsubsistence activ- basalt stones comprising walls that stand to a ities represented in different rooms within settle- height of up to four courses or approximately 1.5 m ment. For example, in operation SLT 9 features on (typically builtagainst cut bedrock). In most cases, the living surface include hearth features – com- however, the walls are only one or two courses, mon to many of the exposed rooms – but the most defining characteristic of this space is aseries of

62 sensu Barnard/Wendrich 2008;Hanks/Linduff 2009;Porter 2012. 63 see Lindsay et al. 2010 for details. 64 Badalyan et al. 2008,Fig. 28. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 195

Fig. 26 Tsaghkahovit. Plan of South Lower Town residential complex, including operations SLT 9, 10, 11, 13,and 14,the primary units under discussion 196 Ruben Badalyan et al.

Site Operation Material Lab Number 14CDate (BP) Calibrated Date (BC) 95.4%Probability Tsaghkahovit SLT, 2008 SLT8,l.9 charcoal AA-82785 2993 Æ 44 1390-1058 Tsaghkahovit SLT, 2008 SS1,l.6 charcoal AA-82786 3231 Æ 39 1608-1430 Tsaghkahovit SLT, 2008 SLT8,l.12 charcoal AA-82787 2883 Æ 66 1290-900 Tsaghkahovit SLT, 2010 SLT9,l.12bcharred pea AA-92628 2647 Æ 60 972-571 seed Tsaghkahovit SLT, 2010 SLT9,l.13 charcoal AA-92629 2651 Æ 40 900-782 Tsaghkahovit SLT, 2010 SLT9,l.21 charcoal AA-92630 3080 Æ 39 1434-1260 Tsaghkahovit SLT, 2010 SLT9,l.35 charcoal AA-92631 3067 Æ 38 1426-1218 Tsaghkahovit SLT, 2010 SLT10,l.14 charcoal AA-92632 2702 Æ 38 919-801 Tsaghkahovit SLT, 2010 SLT10,l.14 charcoal AA-92633 3111 Æ 39 1491-1270 Tsaghkahovit SLT, 2010 SLT10,l.16 charcoal AA-92634 3102 Æ 39 1448-1267 Tsaghkahovit SLT, 2010 SLT10,l.19 charcoal AA-92635 3110 Æ 38 1490-1270 Tsaghkahovit SLT, 2011 SLT13,l.13 charcoal AA-95619 2353 Æ 39 726-366 Tsaghkahovit SLT, 2011 SLT13,l.15 charcoal AA-95620 2885 Æ 39 1211-934 Tsaghkahovit SLT, 2011 SLT13,l.19 charcoal AA-95621 2495 Æ 39 787-417 Tab. 10 Tsaghkahovit SLT, 2011 SLT13,l.22 charcoal AA-95622 3185 Æ 87 1682-1262 Radiocarbon determinations related to Tsaghkahovit SLT, 2011 SLT14,l.5 charcoal AA-95623 3080 Æ 51 1488-1210 Tsaghkahovit South Lower Town from the Tsaghkahovit SLT, 2011 SLT14,l.9 charcoal AA-95624 3182 Æ 41 1530-1386 2008–2011 field sea- Tsaghkahovit SLT, 2011 SLT14,l.12 charcoal AA-95625 3117 Æ 40 1494-1294 sons (calibrations by OxCalv.4.1.7Bronk Tsaghkahovit SLT, 2011 SLT14,l.14 charcoal AA-95626 3148 Æ 40 1504-1316 Ramsay 2010 r. 5)

variably sized pits dotting the floor, includingfive to the east. This operation uncoveredalarge ovoid very large pits on the order of 1–1.5 mindiameter semi-subterranean room enclosed by amulti-course by 1 mdeep (Fig. 27).Most of these pits contained stonewall W1301 (Fig. 28).The interior features of grains of domesticated wheat, barley, and pulse re- the room includedalarge stone basin on the bed- covered by flotation, fromwhich we conclude that rock floor measuring 1.5 mindiameter. The basin the occupants of the lower town engaged in grain was propped up on its eastern side by cobbles production during the short summer growing sea- where thebedrock floor was uneven; this attempt son to augment theirpastoral economy (see be- to level such aheavy basin may indicate it was low). These pits were in addition to the smaller used to hold liquid (perhaps water or oil) for scale pits noted above that may have had arole in household use. The basin’s function remains some- structural support. thing of amystery, but it is nearly identical to a The primary use of this space as astorage fa- basin discovered on an LBA floor in operation SLT 5 cility – encompassing operation SLT 9 and portions excavated in 2005.66 Othermaterials in SLT 13 of SLT 10 and 11 – is supportedbyanotable lack included astationary stone mortarcarved into the of evidence for food processingand consumption; bedrock outcropatthe western end of the room indeed, only two pieces of groundstone were recov- and seven pieces of groundstone, indicating apos- ered from operationsSLT 9, 10,and 11.This is in sible focus on food processing in this space. contrast to findings of similarly sized operations Although storage was not as dominant acharacter- just to the north – SLT 1, 4,and 5 – where we istic as in neighboring SLT 9,wedid encounter collected 15 piecesoflargegroundstone and two three substantial storage pits along the eastern in situ cooking vessels,artifacts associated more di- baulk radiocarbon dated to the LBA (samples AA- rectly with preparing and cooking food, in rooms 95620 and AA-95622). Immediately overlaying the that we interpreted as living spaces.65 large LB pits was aclay floor with several burned The storage features that characterize SLT 9 areas radiocarbon dated to the mid-to-late 1st mill. also contrast with materials documented in SLT 13 BC (Iron 3 period, samples AA-95619 and AA-

65 Badalyan et al. 2008. 66 Badalyan et al. 2008,Fig. 28. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 197

Fig. 27 Tsaghkahovit. Opera- tion SLT 9,storage facility in the South Lower Town complex

Fig. 28 Tsaghkahovit. Operati- on SLT 13,semi-subterranean room in South Lower Town complex, looking west toward W1301 198 Ruben Badalyan et al.

95621). The presence of LB pits belowanIron 3 the lower town. This is consistentwith the sugges- context mirrors the stratigraphic situation in operation that residents of the area may have been at tion SLT 1 where an LB pit was filled in and paved least semi-mobile, since transhumant pastoralists over with an Iron 3 flagstone floor. tend to rely on these secondary products. Survivor- With larger ceramic samples from new expo- ship among cattle strongly resembles that from sures and acurrent total of 28 radiocarbon determi- other sites in the Tsaghkahovit Plain, withover 60% nations analyzed between 2003–2011,wecan con- of the herd living until physical maturity (Tab. 7). clude there was an initial occupation of the South While the macrobotanical assemblages of crop Lower Town during the II phase of the LBA, fol- and weed remains are broadly similar across the lowed by asecondmajor reoccupation during the Bronze Age sites of the TsaghkahovitPlain, the mid-1st mill. BC (Iron 3 period); the latter occupa- plant profile for the LB occupation at Tsaghkahovit tion is well-attested by three radiocarbon samples is nearly identical to that of Gegharot (Tab. 4).The from SLT 1 and SLT 13,and afourth from SLT 6 (a Tsaghkahovit assemblage includeshulled barley test trench locatedabout 30 mtothe south of op- (Hordeum vulgare,part of which belongs to six-ro- erations discussed here), in addition to the ceramic wed subspecies), common breadwheat (Triticum sequencesevident from the settlement. aestivum ssp. vulgare)and club wheat(Triticum An intriguingdevelopment related to the oc- aestivum ssp. compactum), emmer (Triticum dicoc- cupationalsequence of the Tsaghkahovit South cum), rye (Secale sp.), and millet (Panicum milia- Lower Town is the mounting radiocarbon evidence ceum).The barley to wheat ratio is the same as for for continued sporadic use of the settlement during LB Gegharot: 75:25%. Prevailing weeds include the Iron 1 period; so far, six C14 determinations speciesfrom Polygonace, Rubiaceae, Boraginaceae have returned dates from the turn of the 1st millen- and Cyperaceae families. nium BC, complemented by asmall handfulof sherds with diagnostic Iron 1 stylistic attributes. Though we lack aclear occupational level attribu- Excavations of the Late Bronze Age ted to the Iron 1 period, it now appears that the Cemetery at Tsaghkahovit (Ts. BC 12) SLT settlement area continued to be revisited – perhaps for the lush summer pastures on Mt. Ara- gats – long after the Gegharotand Tsaghkahovit In its 1998–2000 survey, ProjectArAGATS took the fortresses weredestroyed and abandoned. first steps toward asystematicspatial investigation The faunal remains from the SLT operations of aLate BronzeAge mortuary landscape, recording at Tsaghkahovitsupport the interpretation of the the location of burial clusters, estimating the num- area as acommunity engaged in amixed farming ber of tombs within burial clusters, identifying types and herding economy (Tab. 9).Aswith all of the of spatial relationships (or sub-clusters) between other assemblages in the TsaghkahovitPlain, the cromlechs67 (e.g. lines, ribbons, densely packed, majority of the specimensidentified to the level of etc.), and noting morphological typeswhen visible genus were sheep and goats,although cattle in this (e.g. standard, paved, spiral, etc.).68 In order to sample are aclose second.Equids (domesticated further investigate these spatial relationships and horses outnumber donkeys where they could be mortuary practices,excavations were conductedin distinguished)are the only other taxon that makes 2006 and 2008 at Tsaghkahovit Burial Cluster 12 up more than 1%ofthe sample. As usual, pigs are (Ts BC 12). Ts BC 12 has one of the largest number present in extremely low proportions, which does of tombs per burial cluster (n = 160)inthe Tsagh- not rule out the possibility that at least part of the population was highly mobile. Survivorship patterns also indicate production for the localized area rather than pursuit of aspe- 67 Cromlechs refer to circular mound architecture constructed from cialized economy or production for exchange. Survi- stonesplaced in various arrangements; like ‘‘kurgans’’they ha- vorship of sheep and goats in Stage Aisnoticeably ve an outermost ring that is also termed ‘‘cromlech,’’ but unli- higher than in all other Late Bronze occupations on ke kurgans they do not have an earthen mound and are generally smaller than 10 mindiameter.Bayern (1882, 17) the plainand it remains as high in Stage B (Tab. 3). introduced the term ‘‘cromlech’’ to the Caucasusinorder to Over 60 %ofthe herd reached physical maturity;a distinguishthese tombs from the ‘‘dolmen,’’ that he had iden- proportion that is more similartocattle survivor- tified in Circassia. These dolmen are megalithic tombs with three or more uprightwalls supporting alarge flat horizontal ship from other LB Tsaghkahovit Plain sites. This capstone and often with a‘‘portal’’ in one wall. He noted that indicatesnot only apattern of local production for the cromlechs of the South Caucasuswere not exactly the sa- local consumption, but may also suggest that the me as those in Europe, but that the term was appropriate since they were partofacultic tradition dating to the Bronze-Iron secondary products of sheep and goats, including transition. wool and milk, were important to the inhabitants of 68 see Smith et al. 2009. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 199 kahovit Plain,69 which suggestedtousthat this particular burial cluster was apopular place for in- terring the deceasedand may have been used either by alarge number of people (or groups)over acomparatively long duration of time. Ts BC 12 is located 350 msoutheast of TsaghkahovitFortress on aspur of Mt. Aragats overlooking the southernedge of the Tsaghkahovit Plain. The burial cluster covers an area of approximately 5.5 hectares, encompassing two facing slopes of asmall valley oriented northeast-southwest. Based on the map we generated in 2008 (Fig. 29),each slope has approximately the same number of tombs (76 and 84), but the tombs on the northern slope are more densely packed in a smaller area. In particular, sub-group70 1 is the most densely packed with approximately 56 cromlechs in 274 square meters and with 80%ofthese cromlechs overlapping each other. In contrast, only 35%ofthe cromlechs overlapinsub-group 2 and 23%overlap in sub-group 3.Instead of overlapping, several of the cromlechs in sub-group 3 appear to be evenly spaced out alonghorizontal lines. Thus, there appear to be distinctions in surface ar- chitecturalconstructionand placement of tombs that distinguish areas withinthe burial cluster. Ex- cavations were thus undertaken in order to investigate whether there is acorrelation between the spatial distribution of tombs within the burial cluster (i.e. sub-clusters) and any of the following ele- Three of the tombs on the southern slope Fig. 29 ments: surface tomb construction, sub-surface tomb (B02,B03,and B04)were dated to the LBA and Tsaghkahovit BC 12. Topographicmap of Ts construction, chronology, artifact inventories, post- shared several architectural and intermentfea- BC 12 with cromlech mortem treatment of the deceased,numberofindi- tures.74 All three of these tombs consisted of pri- tombs shaded in gray viduals present (MNI),age, sex, or any other biolo- mary fully-articulated intermentsinwhich the de- gical information.71 ceased were placed within earthen pits that were In 2006,weselected four cromlechs(B01, sealed with large basalt capstonesand covered B02,B03,and B04)for excavation that appeared to over with basalt stones in a‘‘standard’’ cromlech be in ahorizontal ‘‘line’’ on the southern slope. In design (see Tab. 11 for architectural details). In par- 2008,72 we randomly selected cromlechs from the ticular, B02 and B03 had very similar interment more densely packed cromlechs from subgroup 1 styles; within these tombs, asingle individual was (B09 and B10)and subgroup 2 (B06 and B07;B11 placed in aflexed position on the right side with was excavated with B07)onthe northern slope.73 ceramic vessels surrounding or in articulationwith the deceased’s body (Fig. 30).75 Both individuals were adult males and 35–49 years old at the time

69 An estimated 91 tombs were counted during the survey, but the mapping project revealed ahigher estimate of 161. 70 Three sub-groups were identified within Ts BC 12 based on to- 74 B01 was distinct from the other excavated tombs on the South- pography as well as differences between elements of form and ern slope in terms of its surface architecture – aspiral design construction(see Smith et al. 2009,190). – and the lack of interred materials. It is thus likely that B 01 71 Due to the limitations of space, this report briefly describesthe was acenotaph, but withoutany associated materials the con- results of the excavations. Amore detailed discussion and ana- struction cannot be securely dated to the LBA. lysis of these features,includingbioarchaeological analyses, is 75 However, on the floorofB02’s pit were three sheep (Ovis): an available in Marshall (2014). articulated male 6–12 months old and the partial remainsofa 72 The 2008 excavations were supportedbyaFulbright Hays female and male each 2–6 months old. This placement is simi- DDRA grant. lar to the Western Chamber of GegharotKurgan 1;however, 73 Visible cromlechsineach subgroup were assigned numbers the ceramics suggested atransitional late Middle Bronze Age and then local excavation team membersrandomly selected a assemblage and no human remains were recovered from the tomb from each subgroup. Two cromlechs (see Avetisyan et al. chamber(see Badalyan et al. 2008: 59–61). All faunal remains 2000)had already beenexcavated from subgroup 2 in 1998. were analyzed by Dr. Belinda Monahan. 200 Ruben Badalyan et al.

Burial #Period Location Sub Group Surface Diameter (m) CapstonesCapstone Architecture Length (m) 1 N/A South Slope 3 Spiral 5.92 11.5 2 LB ISouth Slope 3 Standard 6.98 11.75 3 LB ISouth Slope 3 Standard 4.932.5 4 LB South Slope 3 Standard 6.15 42.05 6 LB II North Slope 2 Paved 4.83 21.75 BA LB II North Slope 2 n/a n/a 11.55 BB LB II North Slope 2 Paved 9.321.7 7 LB II North Slope 2 Stepped 10.131.75 9 LB II North Slope 1 Standard 3.710.85 10 LB II North Slope 1 Standard 4.522.05 11 LB II North Slope 2 Paved/Bud 2.96 21.38

Burial# Subsurface Length (m) Width(m) Height (m) Orientation Preparation Courses Architecture 1 Earthen pit 1.81.51.58 NS n/a n/a 2 Earthen pit 1.84 1.44 1.52 EW n/a n/a 3 Earthen pit 1.64 1.63 0.95 NESW n/a n/a 4 Earthenpit 2.93 1.80.96 EW n/a n/a 6 Stone Cist 2.78 1.35 1.49 NWSE Shaped 2 to 3 BA Stone Cist 1.70.91.15 NESW Worked 2 to 3 BB Stone Cist 1.75 1.05 1.1 NWSE Worked 2 7 StoneCist 2.42 1.46 1.65 EW Worked 2 Tab. 11 Tsaghkahovit BC 12. – 9 Stone Cist 1.05 0.80.57 EW Worked 1 surface architecture of 10 Stone Cist 1.31.15 1.15 NWSE Worked 2 excavated tombs; – subsurface architecture 11 StoneCist 0.58 0.52 0.56 EW Worked 1 to 2 of excavated tombs

of death.76 The ceramic vessels that surrounded In contrast to the tombs on the southern these individuals were mostly complete undeco- slope, the five tombs that we excavatedonthe rated bowls and jars.77 The vessels from these two northern slope all proved to be stone-walled cist tombs suggest aLBIperiod assemblage.While chambers, containing only decorated ceramic ves- B04 was similartoB02 and B03 in terms of archi- sels, and, based on the ceramic assemblages, at- tectural features, two younger adult (30–45 and tributable to the LB II period. However, they also 20–35 years old) males were placed in the tomb demonstrated agreaterdiversity in surface architec- fully-articulated in flexed positions on their right ture morphology (cromlech type), post-mortem sides, facing in opposite directions, and with their treatmentofthe deceased, placement of vessels, feet and legs overlapping. No ceramic vessels were and construction techniques for the cist walls. For placed in the tomb; however, basedonabone tog- example, B06 was apaved oval shaped cromlech gle pin found near the ribs of one of the indivi- with interlocking capstones (Fig. 30) that covered a duals, B03 can be attributed generally to the LBA. cist chamber incorporating basalt bedrock to make up the southwestern wall. Here, the fully articulated human remains78 were placed in aflexed position on the right side with the arms extended down to 76 Skeletal age and sex were determined according to Buikstra and Ubelaker et al. 1994.The human remainsofthe individual the legs and five vessels (four bowls and one jar) interred in B03 were not well preserved and could have been olderthan the 35–49 estimated. 77 B02 contained seven whole vessels and onelarge karas frag- 78 The human remains were that of an adult, but were not well ment, while B03 contained four whole vessels, one of which preserved due to the fact that they were covered in aclay ma- (B03.09)had decorative elements. trix that filled the chamber. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 201

Fig. 30 Tsaghkahovit BC 12. Plan andsection views of cromlech tombs; – a–cBurial 02; d–gBurial 06 202 Ruben Badalyan et al.

were stacked in the southern corner.79 In contrast, top of each other.84 The chamber contained two in- the surface architecture of B07 was alarge stepped terments,one apartially articulatedadult male 22– cromlech with three largebasalt capstonescovering 32 years old and one infant (birth Æ 2 months).The alarge cist chamber in which acomplete bowl was adult male was placed in the tomb in avery tightly placed in one corner80 and the remains of the de- flexed position on the right side, perhaps originally ceased were disarticulated and scattered on the leaning against the southeastern wall. While the ma- floor. B11,asmall ‘‘bud’’cromlech that overlapped jority of bones were present, this individual was B07,contained no human remains at all. One bowl missing the skull and may have been ‘‘bundled’’ and four jars were placed on the floor of the cham- after death and secondarily interred. The infant was ber; three of these vessels contained faunal (med- placed between the adult and the northeastern wall. ium mammal/sheep/goat) and/orbotanical remains While the excavations at Ts BC 12 in 2006 (cerealgrains, one pulse seed, and weeds).81 B09 and 2008 were small scale,they have contributed and B10 both had standard cromlech architecture82 to our understanding of the history of mortuary and actually shared some of their surface stones. practices in the Tsaghkahovit Plain. In particular, Yet they differedinterms of subsurface architecture they have suggested parallel distinctions between and interment styles. B09 was constructed on aba- location, surface architecture, subsurfacearchitec- salt bedrock floor and one individual, 12 years old ture, ceramic forms and decorative elements, and Æ 30 months, was placed on this floor fully articu- placement of vessels. For example,onthe southern lated in aflexed position on the right side. The de- slope the excavated tombs were either standard or ceased was wearing abronze bracelet aroundthe spiral types and had earthenpits sealed with cap- left arm and abronze earring was found near the stones that were embedded or covered with addi- skull. The individual’s legs, feet, and part of the tional stones. Three of these tombs contained pri- pelvis were covered by alarge black jar that con- mary interments of fully articulated adult tained cultigens83 including an emmer grain that individuals and two of the tombs included ceramic was boiled prior to charring, anumber of medium vessels placed aroundthe body.The majority of mammal ribsand sheep/goat remains that may these vessels were undecorated (80%) bowls (70%) have beenfrom abutchery unit, and afragmentof and datedtothe LB Iperiod. In contrast, the exca- obsidian. Atwo tanged gray obsidian arrowhead vated tombs on the northern slope demonstrated with retouching on both faceswas also deposited in greater variationincromlech architecture and were the tomb. In contrast, B10 contained comparatively all stone-lined pit chambers covered by several cap- few material objects; two bowls, one of which con- stones that formed part of the surfacearchitecture. tained medium mammal faunal remains, were Greater variationinpost-mortemtreatment was placed in the southern corner of acist chamber. also observed on the northern slope with the inter- B10,however, was distinct from B09 and the other ment of fully articulated primary interments, apar- excavated tombs in terms of the construction of its tially articulated body without the skull, and the in- walls: the northwestern wall was constructed from terment of select disarticulated or excarnated body several small basalt stones that were stacked on parts. The deceased also included an infant and a subadult as well as adults.Vessels tended to be placed in corners or along walls, but there was at 79 Two of the bowls containedfaunal remains (sheep and cattle) that may haveformed units of meat (e.g. the shoulder of a least one example of placing vessels around and sheep). on top of the deceased’s body. The majority of the 80 Agap in thesecapstonesaswell as materials recovered from vessels in this ceramic assemblagewere decorated the fill of the chamberindicate that Burial 07 was disturbed and looted, although the eastern partofthe chamber may not (92%) jars (60%) and dated to the LB II period. have been disturbed as there was still acapstone abovethis While these associations could be influenced area and human remains. The human remains included: aright by the small sample size of excavated tombs, the tibia and fibula that were in correct anatomical relation, aright ulna, athoracic vertebra,and aleft fibula. The epiphyses, or present evidence suggests two possible interpreta- articular surfaces, had been removed from the long bones. tions. The first is that these overlapping distinctions 81 Paleobotanicalanalysis in 2008 was conducted by Dr. Roman reflect temporal changes in mortuary style and prac- Hovsepyan. tices that may also be associatedwith achanging 82 Although, B09 had one small capstone and smallgap or empty spaceinthe center of the cromlech; asecond capstone could socio-political context. In this interpretation, the have originally covered this area,oritcould represent aburial form that has no central capstone. 83 The large black jar, locus 5,contained bread wheat, emmer, hulledcultivated barley, as well as unidentified cultivatedcere- 84 This chamber construction is very similar to thatdescribed by als, unidentified wheat, tetr- or hexaploid wheat, and cultivated Bayern (1882)for the majority of tombs at Redkin Lager. Bay- barley. The jar also containedseveralweedsand wild rose. ern hypothesized that these walls were originallythe entrance, Threeother jars and onebowl were placed around, or in arti- and that once the body and accompanying goodshad been culation with,the deceased; one of the jars (locus 6)alsocon- lowered into the chamberthroughadromos, the entrance was tained macrobotanical remains (wheat, barley, and weeds). sealedoff with smallerrocks Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 203 tombs would have first been constructed on the the Tsaghkahovit Plain, including Gegharot, Aragatsi southern slope, followed by ashift to the northern Berd,and Tsaghkahovit. It will focus on the chemi- slope during the LB II and possibly culminating in cal composition and some lead isotope ratios of EB the construction of terrace walls that marked off and LB copper based artifacts. Apossible relation- the densely packed sub-group 1.Itisdifficult to ship of the artifacts to Armenian and other regional compare this construction history with construction copper ores in the region is also discussed. Two at TsaghkahovitFortress due to its depositional his- electron microprobe analyses (EMPA)ofIron 1 peri- tory, but it is worth pointing out that the LB Ito od metal objects from Mantash, published earlier,86 LB II transitionatGegharot was marked by asuper- are also used for discussion and comparison. imposed rebuilding phase.85 The second interpretation attributes the distinctions to different mortuary Samples and analytical techniques traditions. The different slopes or subgroups could have beenused by different social (kin-based, poli- Altogether, 40 EB and LB samples were analysed tical, economic, or religious) groupswithin LBA so- for chemicaland trace element composition, and a ciety. In this case, different ways of placing the selection of 17 were analysed to determine their deceased and materialobjects, of constructing sub- lead isotope abundance ratios. Energy dispersiveX- surface chambers, and of arranging cromlechs may ray fluorescence analysis (EDXRF) was used to de- have been part of different traditions of practice terminethe major and trace element composition and may have served to distinguish group member- of the artifacts, using methods described in detail ship and assert socio-political claims related to by Lutz and Pernicka.87 identity, status, or resources. Isotopic analyses were performed with inductively coupled plasma-mass spectrometry with aVG Elemental AXIOM, adouble focussingmagnetic sec- Conclusions tor-based multiple collector inductively coupled plasma mass spectrometer (MC-ICP-MS) using the In sum, the ongoing investigation of Project ArA- protocol defined by Niederschlag et al.88 All ana- GATS in the Tsaghkahovit Plain of central Armenia lyses were continuously compared with certified in- are shedding new light on the social practices and ternationalanalytical reference materials. The re- historical processes that shaped communities in the sults of the EDXRF analyses are given in Tab. 12, South Caucasus during the Bronze Age. The combi- while lead isotope abundance ratios are available nation of systematic survey data and detailed large- in Tab. 13.All analyses were performed at the Curt- scale excavationshave established the region as a Engelhorn-Centerfor Archaeometry in Mannheim, critical location for defining material sequences,es- Germany. tablishing regional chronologies, defining paleo- economies and trackingkey shifts in socio-political EB metal objects dynamics. As we continueour investigations in the coming years, our emphasis will be on expanding Altogether 16 EB metalobjects from Gegharot were into adjacent regions, such as the Aparan Valley, in analyzed for their chemical composition. Some of order to better define variation within Bronze Age them were also examined by lead isotope analysis communities and fill in critical temporal gaps in and some exceptionally interesting high arsenic al- Tsaghkahovit Plain occupation, including the Middle loys were also subjected to metallographic analysis Bronze Age and the Iron 1 period. by optical microscopy and scanning electron microscopy (SEM). Based on the compositions determined by EDXRF, analysis of the metalartifacts Appendix 1.Chemical composition and Lead were classified into four chemical groups: pure cop- Isotope Analysis of metal objects, per (3 objects), copper-arsenic alloys (10 objects), 2002–2010. copper-arsenic-lead alloys (3 objects) and asingle tin bronze dated to the late ‘‘Karnut-Shengavit’’ By K. Meliksetian and E. Pernicka phase of the EB. These groupsare listed in Tab. 12 and an overview of their relative abundances is gi- This contribution presents asummary of archaeo- ven in Fig. 31. metallurgical investigations of metal artifacts recovered by the investigations of Project ArAGATSin

86 Tedesco 2006. 87 Lutz/Pernicka 1996. 85 Badalyan et al. 2008;Smith et al. 2009. 88 Niederschlag et al. 2003. 204 Site Provenance Description XRF sample Chemical Cu %FeCoNiZnPbBiSnAsSbSeTeAuAgTotal group

EBA

1 Gegharot 2002 T-2d/c-5 spiral bead FG-030653 Cu 99 <50 <50 140 <1000 400 <50 <50 10300 350 <50 <80 200120 100

2 Gegharot 2002 T-2d/c-5 double-voluted pendant FG-030654 Cu+As 95 <50 <50 <100 <1000 200 <50 <50 46000 100 140 <80 <100 110 100 (cooper colored-A)

3 Gegharot 2002 T-2d/c-5 double-voluted pendant FG-030655 Cu+As 94 <50 <50 <100 <1000 <100 <50 <50 61000 <50 130 <80 <100 400 100 (silver colored-B)

4 Gegharot 2002 T-2d/c-5 cilindircal bead with relief FG-030656 Cu+As 84 <50 <50 <100 <1000 300 <50 <50 158000 750 <50 <80 <100 130 100

5 Gegharot 2002 T-2d/c-5 barrel shaped bead FG-030657 Cu+As 81 1200 <50 <100 <1000 <100 <50 <50 186000 250 <50 <80 <100 210 100

6 Gegharot 2002 T-2d/c-5 bead with notches FG-030658 Cu+As 80 <50 <50 <100 <1000 600 <50 <50 194000 <50 60 <80 <100 200 100

7 Gegharot 2002 T-2d/c-5 teardrop-shaped biconical FG-030659 Cu+Pb+As 85 <50 <50 <100 <1000 91000 <50 <50 32000 250 270 <80 <100 120 97 bead

8 Gegharot 2002 T-2d/c-5 teardrop-shaped spherical FG-030660 Cu+Pb+As 87 800 <50 120 <1000 37000 <50 <50 82000 450 80 <80 100250 99 bead

9 Gegharot 2006 T-17-104 spear head – a MA-071538 Cu+As 98 200 <100 <100 <2000 <100 <100 20 17200 30 <50 <50 <100 320 100

10 Gegharot 2006 T-17-104 spear head – b MA-071539 Cu+As 98 200 <100 <100 <2000 <100 <100 50 17200 <50 <50 <50 <100 289 100

11 Gegharot 2006 T-20-4 awl MA-071540 Cu+Sn 96 <200 <100 550 <2000 2340 <100 26800 9500 750 <50 <50 <100 230 100

12 Gegharot 2006 T-2E-656 dart blade (?) MA-071541 Cu+As 96 <200 <100 180 <2000 <100 <100 80 38000 50 <50 <50 <100 220 100

13 Gegharot 2008 T-21-39 awl MA-082086 Cu+As 95 <200 <100 100 <2000 180 <100 <50 52000 260 <50 <50 <100 1090 100

14 Gegharot 2008 T-20-103 awl MA-082087 Cu+As 97 <200 <100 240 <2000 <100 <100 60 34000 <50 <50 <50 <100 140 100

15 Gegharot 2010 T-29.4 pendant MA-110580 Cu+Pb+As 85 <200 <100 99 <2000 18800 <100 <50 130000 181 95 <50 <100 340 100

16 Gegharot 2010 T-22.115.M02 bead MA-110581 Cu 100 340 <100 215 <2000 730 <100 3702470 125 117 52 <100 124 100

17 Gegharot 2010 T-22.115.M01 bead MA-110582 Cu 100 200 <100 126 <2000 680 <100 56 2410 203 108 <50 <100 102 100

Tab. 12 Chemical composition of the analyzed artifacts determinedwith EDXRF. Cu concentrations of all samples are giveninweight percents, allothers in ppm Ruben Badalyan et al. Ap

Site Provenance Description XRF sample Chemical Cu %FeCoNiZnPbBiSnAsSbSeTeAuAgTotal reliminary group

LBA

18 Gegharot 2003 T-2E-6 pin FG-050605 Cu+Sn 89 500 70 1300 <1000 6300 <50 100000 4400 490 80 <80 190 240100 report

19 Gegharot 2003 T-2E-9 bracelet FG-050606 Cu+Sn 96 1100 130 800 <1000 11300 70 21400 1960 430 <50 <80 100 380100

20 Gegharot 2005 K1-9a ‘‘dagger’’ MA-071542 Cu+Sn 95 390 490 768 <2000 140 <100 45000 6800 208 97 59 <100 65 100 on

21 Gegharot 2005 K1-9b arrowhead MA-071543 Cu+As 97 <200 <100 250 <2000 160 100 260 26000 900 <50 <50 <100 160100 the

22 Gegharot 2006 T-16-117 awl MA-071548 Cu+Sn 99 <200 <100 770 <2000 3600 100 9200 100 250 <50 <50 <100 30 100 2008

23 Gegharot 2006 T-18-10 bracelet MA-071544 Cu 99 <200 <100 540 <2000 2020 100 2710 660 570 <50 <50 300 370100 , 2010 24 Gegharot 2006 T-19-7 arrowhead MA-071545 Cu 99 <200 <100 540 <2000 2080 100 3500 740 850 <50 <50 180 250100 ,a 25 Gegharot 2006 T-20-19 bead MA-071546 Cu 99 200 <100 <100 2189 960 <100 700 <100 7300 <50 64 338 101100 nd

26 Gegharot 2006 T-20-2 button MA-071547 Cu 99 200 <100 <100 <2000 890 <100 1320 225 1850 <50 <50 732 264100 2011

27 Gegharot 2006 T-20-5 fragment MA-071549 Cu+Sb 81 4200 <100 <100 <2000 1300 <100 51 1060 18000 <50 <50 374 340100 investiga 28 Gegharot 2008 T-21-49 pin (?) MA-082088 Cu+Sn 96 <200 <100 830 <2000 8600 100 28900 1330 170 <50 <50 <100 80 100

29 Gegharot 2008 T-21-23 pin/needle MA-082090 Cu 99 1900 210 920 <2000 4300 150 2260 3600 180 <50 <50 <100 500100 tions 30 Gegharot 2010 T-19.506 button MA-110587 Cu+Sn 82 430 195 600 <2000 740 <100 173000 7800 420 51 <50 <100 81 100

31 Gegharot 2010 T-27.56.M02 fragment MA-110586 Cu+Sn 87 550 <100 690 <2000 3800 <100 118000 3200 420 69 <50 <100 350100 of

32 Gegharot 2010 T-27.56.M01 arrow head MA-110588 Cu+Sn 97 <200 143 1450 <2000 2900 <100 23200 <100 177 50 <50 <100 263100 Project

33 Gegharot 2010 T-27.11.M01 arrow head MA-110589 Cu+Sn 97 1410 180 710 <2000 1640 <100 23000 740 150 50 <50 <100 310100 ArAGATS 34 Gegharot 2011 T 30.03.M.01 awl MA-120958 Cu+Sn 94 2300 100 600 <2000 7500 200 42000 1200 200 <100 100 <100 400100

35 Gegharot 2011 T 30.03.M.02 awl MA-120959 Cu 99 1900 400 900 <2000 2500 400 <100900 <100 <100 <100 <100 400100

36 Gegharot 2011 T 31.33.M.01 fragment MA-120960 Cu 99 100 <100 500 <2000 400300 1030 200 2300 <100 <100 <100 400100

37 Gegharot 2011 T 52.20.M.01 bracelet MA-120961 Cu 99 700 <100 600 <2000 1400 100 <1002500 1200 <100 <100 <100 300100

38 Aragatsi-Berd AB-2-1 hook (?) MA-082091 Cu 100 <200 <100 580 <2000 350100 80 100 50 <50 <50 280 150100 2008

39 Aragatsi-Berd AB-3-13 spiral ring MA-082092 Cu 99 <200 <100 520 <2000 670100 2560 7700 1160 60 <50 450 240100 2008

40 Tsakhkahovit TS-B-09-07 bracelet MA-082093 Cu 99 <200 <100 550 <2000 310100 2790 3500 1380 <50 <50 140 230100 2008

41 Tsakhkahovit TS-B-09-11 earring (?) MA-082094 Cu 97 <200 <100 1770 8500 1700 100 8000 1560 80 340 <50 5200 6700 100 2008

Iron 1

42 Mantash 2000 tomb no.3 dagger Cu+Sn 94 1500 – 780 – 900 – 50000 ––––– 310100

43 Mantash 2000 tomb no.3 mace head Cu+Sb 97 10 – 100 – 1080 – 3300 2860 21600 ––– 680100

Tab. 12 continued 205 206 Ruben Badalyan et al.

Sample Orig sample No Site Description Alloy type Dating 208Pb/206Pb 207Pb/206Pb 208Pb/204Pb 207Pb/204Pb 206Pb/204Pb

FG-030653 GE-T-2d/c-5-Gegh-1 Gegharot necklace beads Cu EBA 2,07940 0,84045 38,5890 15,5970 18,5580

FG-030654 GE-T-2d/c-5-Gegh-2 Gegharot necklace beadsCu+AsEBA 2,07160 0,83517 38,6860 15,5970 18,6740

FG-030655 GE-T-2d/c-5-Gegh-3 Gegharot necklace beads Cu+AsEBA 2,06920 0,8350738,7240 15,6280 18,7140

FG-030656 GE-T-2d/c-5-Gegh-4 Gegharot necklacebeads Cu+AsEBA 2,072900,8366938,7200 15,628018,6780

FG-030657 GE-T-2d/c-5-Gegh-5 Gegharot necklace beads Cu+AsEBA 2,07520 0,83752 38,7090 15,6230 18,6530

FG-030658 GE-T-2d/c-5-Gegh-6 Gegharot necklace beads Cu+AsEBA 2,07060 0,83399 38,7370 15,6030 18,7080

FG-030659 GE-T-2d/c-5-Gegh-7 Gegharot necklace beadsCu+As+Pb EBA 2,09520 0,85139 38,2890 15,5590 18,2750

FG-030660 GE-T-2d/c-5-Gegh-8 Gegharot necklace beads Cu+As+Pb EBA 2,09080 0,8483638,3630 15,5670 18,3490

MA-071540 GE-T20-4 Gegharot awl Cu+SnLate EBA 2,08300 0,84428 38,593015,6430 18,5280

MA-071547 GE-T20-2 Gegharot button Cu LBA 2,08720 0,84697 38,5340 15,6360 18,4610

MA-071548 GE-T16-117 Gegharot awl Cu LBA 2,06060 0,83095 38,9100 15,6910 18,8830

MA-071545 GE-T-19-7 Gegharot arrowheadCuLBA 2,08490 0,84576 38,5600 15,6420 18,4940

MA-071546 GE-T-20-19 Gegharot bead Cu LBA 2,074600,83866 38,7200 15,652018,6630

MA-071542 GE-K1-9aGegharot tanged blade Cu+SnLBA 2,05880 0,82906 38,9460 15,6830 18,9160

MA-071543 GE-K1-9bGegharot arrowheadCu+AsLBA 2,06900 0,82885 39,2200 15,7110 18,9560

MA-071544 GE-T-18-10 Gegharot braceletCuLBA 2,076000,83890 38,6470 15,617018,6160

MA-071549 GE-T-20-5 Gegharot fragmentCu+SbLBA 2,08570 0,84495 38,6230 15,6470 18,5180

Tab. 13 Lead isotope Aspearhead from Gegharot was sampled at 144.5 g), 88 chalcedony and 217 talc beads.89 In abundance ratios of two different locations(MA-071538 and MA-071539), the metal beads of the necklace three types of al- analyzed artifacts the shaft (a) and the blade (b) respectively, but as loys were identified:90 doublevolute beads consist expected the compositionofboth parts is identical, of arsenical copper with 4.6–6.1%As; conical and with only minordifferences. spherical teardrop-shaped beads consist of leaded Of particular interest is the necklace of Ge- arsenical copper; and, finally, cylindrical and barrel- gharot (Fig. 32) consisting of 99 metal (totalweight shaped beads are made of acopper alloy with extremely high As contents ranging between 15.8% and 19.4%. These beads are mainly characterized by agray ’’silvery’’ color, sometimeswith ayellow- ish ‘‘bronze’’ tint. We assume that the ancient craftsman used alloys with different colors to give the necklace an extraordinary, ‘‘precious’’ appearance. Lead isotope analyses of objects from the necklace of Gegharot show aconsiderablevariation in the different alloys.Inparticular, theCu+As alloys are clearly different from Cu +As+Pbones (Fig. 34).Therefore, we assumethat more than one ore source was used for the production of these alloys.91 The SEM examination of one of the beads of the Gegharot necklacewith high arsenic content revealed an eutectic microstructure, which is responsible for the silvery gray color of the beads.92 It should be mentioned that it is difficult to produce Fig. 31 Gegharot.Statistical distribution of the al- 89 loys used for EBA arti- Smith et al. 2004;Hayrapetyan 2005. 90 facts. Most of the ob- Meliksetian et al. 2009. jects consist of Cu+As 91 Meliksetian et al. 2009. alloys and unalloyed Cu 92 Meliksetian et al. 2011b. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 207

Fig. 32 Gegharot,EBnecklace. Photo (left) and drawings of individual bronze elements (right). – 1 Double-volutedbeads (FG-030654, FG-030655); – 2 Cylin- drically shaped bead with raised transversal rims (FG-030656); – 3 Cylindrically shaped bead with parallel oblique cuttings (FG-030658); – 4 Bar- rel-shaped bead (FG-030657); – 5–6 Two types of teardrop- shaped beads: conical and spherical (FG-030659, FG-030660) such copper–arsenicalloys, becausearsenic is in the vapour state at the melting temperature of copper (1085 C). But the SEM analyses revealed that some eutectic microstructuresare present with a much lower melting point such as the mixture of 79%Cu+21%Asthat represents the copper-arsenic eutectic point at 685 C. This is still higher than the sublimation point of arsenic (615 C) but obviously it is much easier to achieve in asmall crucible by combinedmelting of copper with an arsenic-rich material, possibly native arsenic. Amore detailedpaperdiscussingthe chemical, lead isotope, optical microscopy and SEM analyses of high arsenic (15–28%) alloys from Gegharot and from Lori Berd is now in preparation. Fig. 33 It is known that the addition of arsenic makes Statisticaldistribution of the alloys used for copperbrittle but changes the color of the metalto LBA artifacts from Geg- asilvery and dark-silvery color with aunique shine. harot, Aragatsi Berd, Considering the fact that ahigh arsenic content Tsakhkahovitand Man- (7%and more) is related primarily to EB decorative tash. Most of the objects consist of Cu+Sn objectsand jewelry in Armenia, rather than tools alloys and unalloyed Cu 208 Ruben Badalyan et al.

Fig. 34 Diagram of 207Pb/ 206Pb vs. 208Pb/ 206Pb in EBA and LBA artifacts from Gegharot combined with the iso- topecompositional fieldsofArmenian ores: ‘‘Radiogenic’’ lead, ‘‘ordinary’’ lead and ‘‘old’’ lead. Dotted lines show the lead isotope fields for ores from Anatolia, Oman as well as from Feinan, Jordan

and weapons,93 we consider that an attempt to As noted above, three objects from Gegharot achievethis distinctive appearance (brightshine, represent unalloyed copper (about 99%Cu), with silvery color) for decorative objects and jewelry was other impurities comprising about 1%orless. Such the major reason for the intentional production of acomposition is widely distributed during the EB in copperalloys with high arsenic contents in prehis- Anatolia,the Middle East96 and in the South Cauca- toric times. sus.97 Anotherinteresting type of alloy used at EB Asingle tin bronze is representedbyanawl Gegharot was acopper-arsenic-lead alloy (3 ob- dated to the ‘‘Karnut-Shengavit’’ phase of the EB jects, FG-030659,FG-030660,MA-110580). In Arme- and marks the periodwhen tin began to replace nian Early Bronze Age contexts, such tertiary alloys arsenic as the most important alloying component. are known also from Harich.94 Lead is afrequent It should also be mentioned that this object con- trace element in copper based artifacts, as it is a tains about 1%arsenic and also high Ni (550 ppm) common minorcomponentincopper ores. How- compared with other objects. But, in general, none ever, because the concentration of lead in the Ge- of the artifacts can be classified into ahigh nickel- gharot artifactsranged from 3.7 to 9.1%, we as- silver group, such as that revealed for some EB arti- sume an intentional addition of lead. facts from Gyumri and Talin.98 During the EB, the alloyingofcopperwith As for the silver contents, only asingle object lead is uncommon but has been supposedfor (MA-082086), an awl with 5.2%arsenic, was rela- some artifacts in the Aegean and other regions of tively high in this metal, containing 1090 ppm Ag; Mediterranean.95 However, Aegean copper-lead al- all other EB objects from Gegharot contained just loys are not high in arsenic. 100–400 ppm.

93 Meliksetian et al. 2009;Meliksetian/Pernicka 2010;Meliksetian et al. 2011b. 96 Tylecote 1976. 94 Meliksetian/Pernicka 2010. 97 ˆåâîðŒ'í 1980;Meliksetian et al. 2010. 95 Pernicka et al. 1990. 98 Meliksetian/Pernicka 2010. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 209

LBA and Iron 1 metal Objects metal rather than in the slag and metallurgical processes do not affect the lead isotope signature. 20 metal objects excavated from the LB Tsaghkaho- Table 13 contains lead isotope analyses of se- vit Plain were analyzed by EDXRF, including 16 from lected EB and LB objects from Gegharot. They are Gegharot, 2 from Aragatsi Berd, and 2 from Tsagh- quite variable in the beads from the EB Gegharot kahovit. Analyses of two objects from Mantash necklace as discussed above.101 The analytical re- were takenfrom the PhD dissertation of Laura Te- sults for both Cu +Asand Cu +As+Pb alloys desco. Òhe chemical compositions and groups of (Fig. 34) lead us to conclude that more than one these LB artifactsare given in Tab. 12 and do not ore source was used in the production of the neck- reveal any surprises. Most of the objectswere com- lace.Lead isotope ratios of four Cu +Asbeads from posed of unalloyed copper and tin bronze; one the necklace, asingle EB, and two LB unalloyed Cu copper-antimony and one copper-arsenic object objects match the ‘‘ordinary’’ lead compositional were also present. An overview of the relative field in Armenia,which is partially overlappedby abundances of the chemicalgroups of LB objects the Anatolian field. These data suggest that the raw from Gegharot is provided in Fig. 33. material for these artifacts was derived from alocal These chemical groups are consistent with source while artifacts which exhibit wider variations, other ArmenianLBartifacts analyzed duringthe last matching ‘‘old’’ lead compositional field and Anato- six years.99 The chemical composition of LB objects lian compositional fields, relate to more distant par- indicates that in the Tsaghkahovit Plain, tinwas ent ores. widely in use. But one can also assume that the As for the tin bronzes from Gegharot, none abundant presence of unalloyed copper, some anti- exhibit an unusual lead isotope signature,suggest- mony-copper alloy, and arsenical copper likely indi- ing that all can be matched to local ore sources. cates that tin was expensive and rare. That unusual pattern is typical for most of EBA tin It is noteworthy that an arrowhead (MA- bronzes from Talin,102 from Troy,103 and sites in 071543), the single copper-arsenic alloy (2.6%As), the Aegean,104 the Persian Gulf 105 and Dage- could easily be anatural alloy since arsenic is a stan.106 Thus, we can assumethat imported tin was common minor element in copper ores of Armenia, mixed with local copper to produce the artifacts or alternatively, the composition is inherited from from Gegharot,since lead isotopes indicate the recycled earlier metal. copper source rather than the tin source. Most of the tin bronzesare also relatively high in nickel (600–1500 ppm). Ahigh silver con- Conclusions tent is typical only for the earring from Tsaghkaho- vit (6700 ppm). This object also exhibits an unu- In this contribution, abrief summary of archaeome- suallyhigh gold content, about(5200 ppm), as well tallurgical investigations of metal artifactsrecovered as zinc (8500 ppm) and tin (8000 ppm). Such a by the investigations of Project ArAGATS in the compositionmay indicate the use of polymetallic Tsaghkahovit Plain, including Gegharot, Aragatsi sulphide ores. Berd and Tsaghkahovit was discussed.InEBA artifacts four chemical groups of metal can be distin- Lead Isotope Analysis guished: pure copper (3 objects), copper-arsenic alloys (10 objects), copper-arsenic-lead alloys(3 Lead isotope geochemistry, introduced in archaeo- objects) and asingle tin bronze dated to the late metry,100 is an important fingerprintingmethod for ‘‘Karnut-Shangavit’’ phase of EBA. Some of the relating artifactstotheir ore sources.Natural lead beads of the Gegharot necklacehave extremely is usually found in copper ores, occurringasamix- high As contents between 15.8%and 19.4%They ture of 204Pb and three radiogenic isotopes 206Pb, were subjected to special metallographic investiga- 207Pb, 208Pb, which are products of the radioactive tions that revealed an eutectic microstructure, decay of 238U, 235Uand 232Th respectively. So lead isotope abundance ratios, such as 208Pb/206Pb, 101 Abrief discussion of Bronze Age isotopic data from artifacts 207 206 208 204 207 204 206 204 Pb/ Pb, Pb/ Pb, Pb/ Pb, Pb/ Pb found in Armenia generally, includingthose of Gegharot, is are important geochemical fingerprints of ore de- available in Meliksetian et al. 2009 and Meliksetian et al. posits of various geologicalages and metallogenic 2011a. In addition,acomparison of lead isotope ratios for EB artifacts from Armenia analyzed prior to 2006 with local and formations. It is important to note that during regional ores is availableinMeliksetian/Pernicka 2010. smelting,lead has atendency to concentrate in the 102 Meliksetian/Pernicka 2010. 103 Pernicka et al. 1990. 104 Begemann et al. 1992;Pernicka et al. 1990;Stos-Gale et al. 1984. 99 Meliksetian et al. 2011a. 105 Weeks 1999. 100 Gale/Stos-Gale 1982. 106 Weeks 2002. 210 Ruben Badalyan et al.

which is responsible for the silvery gray color of the western slope and the base of western slope of the beads. Lead isotope analysis of the beads from Ge- hill, extending to the present village of Gegharot. gharot suggest that more than one ore sourcewas Among the EBA pottery discovered at Ge- used for the production of these alloys. gharot,asample of 1417 sherds was selected for LBA and EIA metal objects mostlyconsist of closer examination,107 and incorporated into adigi- unalloyedcopper (12 objects), tin bronze (11 ob- tal database of EBA pottery.Priority was givento jects), as well as copper-antimony (2 objects) and the material from stratified levels with diagnostic copper-arsenic alloys (1 object). The chemical com- interest (recognizable profile, presence of decora- position of LB objectsindicates that in the Tsaghka- tion, tracks of fabrication techniques, etc.). Also, hovit Plain, tin was widely used as alloyingmetal. the goal is to include in the corpus aselection of But one can also assume that the presenceofunal- pottery which are as much as possible representa- loyed copper, some antimony-copper alloy, and ar- tive of the totality of EBA potteryfrom Gegharot. senical copper likely indicates that tin was expen- Approximately 45%ofthe studiedcorpus has an sive and rare. identifiable profile and recognizable morphological Lead isotope ratios of the analyzed LBA arti- type. This corpus also contains 27 complete ves- facts match Armenian and Anatolian isotope com- sels discovered in situ.The form and the morphol- positional fields, and we can assume alocal or re- ogy of all the Early Bronze potteryhas been thor- gional origin of the copperinthese artifacts. oughly examined to establish achrono-typological classification.

Appendix 2.The Early Bronze Age pottery Productiontechnique and surface treatment from Gegharot The Early Bronze Age pottery from Gegharot was By S. Haroutunian manufactured by techniques characteristic of Kura- Araxes culture. Visualinspection of the fragments The most abundant archaeological artifact in Ge- indicates that they were mostly hand-made, using gharot, pottery, is amaterial of first importance for the coiling technique,streaks of which are clearly the study of the Kura-Araxes Culture. Pottery has visible to the naked eye. Petrographic analysesof long played acritical role in investigations of the Gegharot pottery108 have revealed inclusions of dif- Kura-Araxes, serving as achronological marker and ferent rocks (basalt, granite), minerals (quartz, pla- an indexofcultural identity within settlements and gioclase, biotite, apatite, epidote, chalcedony) and across regions. volcanic glass (obsidian). The size of the inclusions During the last few decades, our knowledge fluctuatesbetween 0.2 mm and 1.5 mm, with few of the Early BronzeAge in the South Caucasus has exceptions where theinclusions are very irregular increasedconsiderably thanks to new research pro- and exceed 2 mm. However, the dominant size of jects that allow us to defineanew understanding the inclusions is between 0,2 and 0,5 mm, and al- of Kura-Araxes society and its chronology. most the half of them are regular and well sorted. The excavations at Gegharot have brought to The vessels are generally burnished or smoothed. light asignificant corpus of EBA pottery. The study They have generally uniform colors, with some ex- of this material allows us to better understand the ceptions where dark colors are mixedwith stainsof modalities of occupation of the site, to clarify the light colors. The colors of the interior or exterior chronologicaldivisions of the settlement’s history, surfaces were obtained by controllingthe variations and to define the evolution of Kura-Araxes Culture of the firing atmosphere109 (reducing for dark colors within the Tsaghkahovit Plain. It also allows us to and oxidizing for light colors). examine the relationships between Gegharot and other sites in the region.

Pottery inventory

The six field seasons, conducted between 2003 and 107 This studyispart of my Ph.D.thesis that Iampreparing in 2011 at Gegharot by Project ArAGATS, recovered Paris-SorbonneUniversity (Paris IV), under the joint supervisi- 26,919 sherds of EBApottery, comprising 53%of on of professor Jean-Yves Monchambert (Paris-Sorbonne Uni- versity)and professor RubenS.Badalyan (Institute of Archae- the total ceramic corpus from the site, excluding in- ology and Ethnography, Academy of Sciences, Republic of determinate ceramics whose period could not be Armenia). identified with certainty. The size of the EB ceramic 108 Hayrapetyan 2008.For comparative data on petrographicana- lyses of Kura-Araxes pottery, see for example Mason/Cooper assemblage indicatesadense occupation during 1999;Batiuk 2000;Iserlis et al. 2010;Kibarogˇlu et al. 2011. the EB period, especially on the western and south- 109 Yon 1981, 61;Balfet et al. 1998, 66. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 211

Decoration the typology of vessels, we can distinguish seven main categoriesofvessels, which are subdivided Elar-Aragats group pottery from Gegharot is largely into several types and variants. Each group, deter- undecorated, except for afew schematic parallel or mined by its profile, corresponds to acategory of perpendicular lines and afew dimples. In contrast, use. Karnut-Shengavit group vessels were highly decorated, typicallywith schematic geometric forms Jars. There are severaltype of jars from Gegharot. (spirals,parallel or perpendicularlines, triangles, There are small high-necked jars, triple-handled, rhombi). Typically, bands of geometric forms were with elliptic or globular body (Fig. 35,1–4).This made on the vessel neck, while large spirals and type of EB II jar is decorated (Fig. 35,3–4).Another other complex patterns were set on the body and type of jar is the small elliptic jar (Fig. 35,jars 2) shoulder. We can differentiate four different decora- with awide biconic (Fig. 35,jars 3a) or elliptic (Fig. tion techniques.110 35,jars 3b) body. Some of these jars are two- handled (Fig. 35,10.12) or have aledge handle Incised decoration: Decoration created by removing (Fig. 35,6.9.11).The third type of jar is the large the vessel surface clay using asharp tool. The dis- storage jar, which can be subdivided into two cate- tal end of the incising tool is maybe circular or line- gories: large jars with biconic bodies (Fig. 35,11.12) ar. This type of decoration is generally found on and large jars with ovoid bodies (Fig. 35,13.14). the vessel neck.Itisalso used to hatch geometric designs. The line thickness is typically between Bowls.These are open vessels of varying dimensions. 0.5–1.5 mm. We can distinguish large bowls with amaximum diameter of more than 30 cm (Fig. 36,15–18),medium Micro-incised decoration: Partly similar to the pre- size bowls, with adiameter between 20–30 cm vious technique, this decoration was created by a (Fig. 36,19–25) and small bowlswithamaximum diam- very sharp tool able to cut lines less than 0.5 mm eter of less than 20 cm (Fig. 36,27–38).The bowls thick. This decoration is sometimes not visible with can have flaring (Fig. 36,15.18) or hemisphere-trun- the naked eye from acertaindistance. cated (Fig. 36,19–25) bodies, with an exterior bead (Fig. 36,19–25,35–38) or asquare (Fig. 36,17– Relief decoration: These decorations were madeby 18.31–34) rim. One of the small bowls has atripod the addition of modeled material to the surface of foot (Fig. 36,26). the vessel. This techniquewas especially used to model large spirals or other geometric forms on the Dishes.These are large vessels with aflattened body of the vessel. base. The maximum diameterisatleast three times more than the height (Fig. 36,39–41). Plastic decoration: These forms were directly modeled on the surface of the vessel, withoutthe addi- Goblets: Alarge number of goblets have been tion of additional material. The forms were created found at Gegharot. With amaximum diameter of using atool or fingers. less than 20 cm, these vessels are characterized by biconic or spheric bodies and high flaring rims Typology (Fig. 37,43–53).This type is widespread at Gegha- rot and represents 35%ofthe totality of vessels in In order to complete the study of the Gegharot EBA the sample. It can also haveahandleonone side pottery, it was essential to establish chrono-typolo- (Fig. 37,42–44),orlugs (Fig. 37,52).The neck may gical classification, based on the criteria of form, have an incised decoration composed of alternately color, decoration, and handle type, taking as are- hatched geometric motifs. Sometimes, the body is ference the pottery classification schemedescribed decorated with complex motifs (double spirals, by Jean-Claude Gardin.111 The typology of Kura-Ara- polygons with angular edges,etc.). xes pottery has already involved many scholars since the late 1960swho createdcomplete or par- Cups: This is auniqueform (Fig. 37,54),which re- tial typologies of Kura-Araxes pottery.112 Studying presents avessel whose height is at least twice as large as the mouth diameter.Ithas ahandle in one side (broken). 110 For detailed descriptions of pottery decoration techniques, see Yon 1981, 23, 85;Balfet et al. 1998, 85, 111;Lorenzi et al. 2000, 18. Mugs: These are small containers with vertical walls 111 Gardin 1976. and flattened base (Fig. 37,55–58).The height is 112 For example, Khanzadyan 1967;Kushnareva/Chubinishvili greater than the maximum diameter.One of the 1970;Sagona 1984;Palumbi 2008;Marro 1997;Smith et al. 2009;Gopnik/Rothman 2011. vessels (Fig. 37,56),isequipped with ahandle 212 Ruben Badalyan et al.

Fig. 35 Gegharot. Brief chrono- typologyofEBA pottery: jars. Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 213

Fig. 36 Gegharot.Brief chrono- typology of EBA pottery: bowls and dishes 214 Ruben Badalyan et al.

Fig. 37 Gegharot. Brief chrono- typologyofEBA pottery: goblets, cup, mugs, lugs Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 215

Tab. 14 List of Gegharot EB pottery integrated in the brief typology 216 Ruben Badalyan et al.

(broken).AnEBIIcup has atripod foot (broken) New decoration techniques and motifs: One of the and two faced holes situated 1 cm below the lip particularities that we can remark is the presence of (Fig. 37,58). new decorations during the EB II. These new modes are visible not only in decoration techniques, but Lids: Lids of various diameters were discoveredin also in decoration motifs, which becomemore and Gegharot (Fig. 37,59–63).These covers are mostly more complex. equipped with ahandle in the center.The exterior surface of some of them is decorated with incised Change of surface colors: The dominantcolors dur- lines. ing EB Iwere generally light colors (7.5YR 6/4, 5YR 5/6, 7YR 5/2,etc.), while those during the EB II are Indeterminate groups: Some of the forms, which darker (GLEY12.5N, 5YR 3/1, 7.5YR 3/2 etc.). This are recognizable, cannot be included in any of the evolution could be aresult of improvements of fir- previousgroups. These groups are composed of ing techniques during the EB II, which gave the bases, which can be flat, where the entire surface possibility to control the firing atmosphere. touchesthe support or the basesare concave, The Early Bronze Age potteryfrom Gegharot where the central part is not in contact with the reveals certain homogeneity of forms and manufac- ground,and coarse pottery, which is composed of turing techniques.The recent excavations in Ge- very irregular and coarsely hand-made pottery. gharot broughtforth an occupation stratigraphy of Among the vessels studiedinclassification two phases of the Kura-Araxes culture and allows corpus, approximately 13%are decorated. to draw up acontinuous chronological sequence of the Early Bronze pottery and to build up its evolu- Conclusion: comparative analysis tion in the plain of Tsaghkahovit. The analysis of typology and the comparisonof Also, the EBA pottery from Gegharot provides two horizons of Kura-Araxes culture highlights some astarting pointfor alarger typology of the cultural specific points. area which can clarify the regional interactions. However, the typological approach of the Kura-Ara- Evolution of existing forms: One of the problems of xes pottery has its limits and for abetter under- Kura-Araxes culture in the region and particularly in standing the interactions we need also to confront the Tsaghkahovit Plainisits evolution and the rela- the typology of EB pottery with acomplete techni- tionships between two cultural horizons of the Early cal analysis of chaıˆne ope´ratoire.Future research Bronze Age. In fact, in spite of alocal hiatus be- will certainly help to fill knowledge gaps about tween these two periods of occupation at Gegharot, many problems of the Kura-Araxes culture. we see an evolution of forms which is clearly visible if we compare the typology of the two horizons. Catalogue of ceramics The three-handled jars, very characteristic to Kura- Araxes pottery, became larger in EB II, with incised Fig. 31ext. black, Gley 12.5/N; int. very dark decoration on their shoulders. The handles, coarse gray, Gley 13/N; – 2 ext. very dark gray, Gley 13/ and thin, are developed to alarger, regular and N; int. dark gray 2.5Y 4/1,very dark gray, Gley 13/ symmetric handle. N(rim); – 3 ext. gray, Gley 15/N; int. light brownish gray 2.5Y 6/2,dark gray, 2.5Y 4/1 (rim); – 4 ext. Continuance of some forms: However, some particu- black, Gley 12.5/N; int. dark gray, 2.5Y 4/1; – 5 ext. lar forms are used in both EB Iand EB II. One of very dark gray, Gley 13/N, pale brown, 10YR 6/3 these forms is the biconicbody which concerns (rim); int. polishlines – very dark grayish brown, both EB I (Fig. 35,5.11) and EB II vessels 10YR 3/2,reddish yellow, 5YR 6/6; – 6 ext. gray, (Fig. 35,12),but also special features like the tripod Gley 15/N; int. grayishbrown, 10YR 5/2; – 7 ext. foot (Fig. 36,26;Fig. 37,58),which is used during yellowish red, 5YR 4/6, 5/6,black, Gley 12.5/N; int. both EB Iand EB II. yellowish red, 5YR 5/6,very dark gray, Gley 13/N. Fig. 51ext. black, 2.5Y 2.5/1,light brownish Change of forms: We can observe the extinction of gray, 2.5Y 6/2;int. grayish brown, 2.5Y 5/2; – 2 ext. some type of pottery which existed in EB I. The black, Gley 12.5/N, dark gray, Gley 14/N; int. cups with vertical walls and the mug are used espe- grayish brown, 2.5Y 5/2; – 3 ext. black,Gley 12.5/ cially during the EB Iperiod. In contrast, in EB II we N, dark gray, Gley 14/N; int. dark gray, 2.5Y 4/1, assist to the development of large (Fig. 36,15–18) grayishbrown 2.5Y 5/2 (neck); – 4 ext. black Gley and small (Fig. 36,27–38) bowls with opened shape 12.5/N dark gray, Gley 14/N (neck); int. dark gray- and high flaring rims which does not exist during ish brown, 10YR 4/2; – 5 ext. black, Gley 12.5/N; EB I. int. very dark gray, Gley 13/N, very dark brown 10YR 2/2 (rim); – 6 ext. black, Gley 12.5/N, yellow- Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 217 ish red 5YR 4/6 (neck); int. dark gray, 7.5YR 4/1, 15 ext. dark gray, Gley 14/N; int. dark gray, gley 1 strong brown 7.5YR 4/6; – 7 ext. blackGley 12.5/ 4/N, yellowish brown, 10YR 5/4; – 16 ext. gray, N; int. dark grayishbrown, 10YR 4/2,very dark Gley 15/N, dark gray, 2.5Y 4/1 – dark grayish grayishbrown, 10YR 3/2 (rim); – 8 ext. black,Gley brown, 2.5Y 4/2;int. black, Gley 12.5/N, dark gray 12.5/N; int. very dark grayish brown, 10YR 3/2; – 2.5Y 4/1. 9 ext. black, Gley 12.5/N; int. black, 2.5Y 2.5/1, Fig. 91ext. black, Gley 12.5/N; int. gray, brown, 7.5YR 4/3 (spotonthe neck); – 10 ext. 7.5YR 5/1; – 2 ext. spots: brown, 7.5YR 4/3,black, black, Gley 12.5/N; int. dark gray, 7.5YR 4/1,brown Gley 12.5/N; int. brown, 7.5YR 5/3; – 3 ext. black, 7.54/2;–11 ext. black, Gley 12.5/N; int. very dark Gley 12.5/N; int. red, 2.5YR 4/8,brown, 7.5YR 5/2; gray Gley 13/N, dark brown, 7.5YR 3/2 (rim); – 12 ext. – 4 ext. black, Gley 12.5/N; int. grayish brown, black, Gley 12.5/N; int. neck (lines): grayish brown, 10YR 5/2,dark grayish brown 10YR 4/2 (polish rim 10YR 5/2,very dark gray, 10YR 3/1;body (spots): and neck), dark gray Gley 14/N (bottom); – 5 ext. very dark gray Gley 13/N, brown 10YR 5/3; 13 ext. black, Gley 12.5/N; int. grayish brown, 10YR 5/2, black, Gley 12.5/N; int. very dark gray, Gley 13/N, dark grayish brown 10YR 4/2 (polish rim); – 6 ext. brown, 7.5YR 4/3 (rim) black, Gley 12.5/N; int. brown, 10YR 5/3,grayish Fig. 61ext. black, Gley 12.5/N; int. brown brown, 10YR 5/2; – 7 ext. dark gray Gley 14/N, gray 7.5YR 4/3, 4/2,darkgray 10YR 4/1; – 2 ext. black, Gley 15/N (neck); int. dark grayish brown 10YR 4/2 Gley 12.5/N; int. reddish yellow, 5YR 4/4; – 3 ext. (polishrim and neck), gray 10YR 5/1. – 8 Stone. black, Gley 12.5/N, brown, 7.5YR 4/4;int.dark gray, 10YR 4/1,yellowish brown, 10YR 5/4 (spot); – 4 ext. dark reddish brown, 5YR 2.5/2 (neck), very Bibliography dark brown, 10YR 2/2 (body); int. dark reddish brown, 5YR 3/3,grayish brown, 10YR 5/2 (body); – Avetisyan et al. 2000 5 ext. black, Gley 12.5/N; spot on the rim: brown, P. Avetisyan/R. Badalyan/A. T. Smith, Preliminary report 7,5YR 4/3;int. reddish brown, 5YR 4/3 (neck), dark on the 1998 archaeological investigations of project gray, 2.5Y 4/1 (body); – 6 ext. black, Gley12.5/N; ArAGATS in the Tsakahovit plain, Armenia. Studi Micenei int. dark grayish brown, 10YR 4/2,dark gray, 10YR ed Egeo-Anatolici 42, 2000,V.1,19–59. Badalyan/Avetisyan 2007 4/1; – 7 ext. black, 10YR 2/1,black, Gley 12.5/N; R. S. Badalyan/P. S. Avetisyan, Bronze and Early Iron Age int. reddish brown, 5YR 4/4,weak red, 10YR 5/4; archaeological sites in Armenia I. Mt. Aragats and its sur- dark gray, Gley 14/N; – 8 ext. black- Gley 12.5/N, rounding region. British archaeological report 1697 (Ox- pale brown, 10YR 6/3;int. grayish brown, 10YR 5/2, ford 2007). dark gray, 10YR 4/1; – 9 ext. black, Gley 12.5/N; Badalyan et al. 2008 int. dark gray, 10YR 4/1,brown, 7.5YR 5/3, 4/2; – R. Badalyan/A. T. Smith/I. Lindsay/L. Khatchadourian/P. 10 ext. black, Gley 12.5/N; int. grayish brown, Avetisyan, Village, fortress, and town in Bronze and Iron 10YR 5/2,dark gray, 10YR 4/1. Age Southern Caucasia: Apreliminary report on the Fig. 71ext. gray, 2.5Y 5/1;int. gray, 2.5Y 5/ 2003–2006 investigations of project ArAGATS on the Tsaghkahovit Plain, Republic of Armenia. Archa¨ologische 1; – 2 ext. brown, 7.5YR 4/4,dark gray, 7.5YR 4/1; Mitteilungen aus Iran und Turan 40, 2008, 45–105. int. strongbrown, 7.5YR 5/6 (neck), black, Gley 1 Badalyan et al. 2010 2.5/N; – 3 ext. dark grayishbrown, 10YR 4/2,dark R. S. Badalyan/A. T. Smith/L. Khatchadourian, Project gray, 10YR 4/1;int. strongbrown, 7.5YR 5/6 (neck ArAGATS: 10 years of investigations into Bronze and &rim), dark gray, 10YR 4/1; – 4 ext. very dark Iron Age sites in the Tsaghkahovit Plain, Republic of Ar- gray, Gley 14/N, brown, 10YR 5/3 (rim); int. grayish menia. Tu¨rkiye Bilimler Akademisi Arkeoloji Dergisi 13, brown, 10YR 5/2; – 5 ext. dark gray, 10YR 4/1;int. 2010, 263–276. light yellowish brown, 10YR 6/4; – 6 ext. dark gray, Balfet et al. 1998 Gley 14/1;int.dark grayishbrown, 2.5Y 4/2; – 7 ext. H. Balfet/M.-F.Fauvet-Berthelot/S. Monzon, Lexique et brown 7.5YR 5/4,very dark gray, 7.5YR 3/1;int. typologie des poteries. CNRS Editions (Paris 1998). Barnard/Wendrich 2008 brown, 7.5YR 3/1; – 8 ext. black, Gley 12.5/N; int. H. Barnard/Willeke Wendrich (ed.), The archaeology of very dark grayish brown, 10YR 3/2,dark grayish mobility: Old World and New World nomadism (Los An- brown, 10YR 4/2; – 9 ext. gray, Gley 15/N; int. geles 2008). grayish brown, 2.5Y 5/2; – 10 ext. dark gray, Gley 1 Bayern 1882 4/N; int. grayish brown, 2.5Y 5/2; – 11 ext. black, F. Bayern. ContributionaL’Arche´ologie du Caucase (Ly- Gley 12.5/N; int. brown, 10YR 5/3; – 12 ext. black, on 1882). Gley 12.5/N, very dark gray, Gley 13/N; int. black, Batiuk 2000 Gley 12.5/N; – 13 ext. brown, 7.5YR 4/3,dark S. Batiuk, Petrographic analysis of ETC ceramics from brown 7.5YR 3/2;int. brown, 7.5YR /4; – 14 ext. the Bayburt region, northeastern Anatolia: an explorato- black, Gley 12.5/N – dark grayish brown 10YR 4/2; ry study. Ancient Near Eastern Studies, 37, 2000, 153–163. int. very dark gray, 10 YR 3/1,brown, 7.5YR 5/4; – 218 Ruben Badalyan et al.

Begemann et al. 1992 Republic of Armenia. Archa¨ologische Mitteilungen aus F.Begemann/S. Schmitt-Strecker/E. Pernicka, The metal Iran und Turan 40, 2008, 45–105. finds from Thermi III–V: achemical and lead isotope- Hovsepyan 2009 study. Studia Troica 2, 1992, 219–239. R. A. Hovsepyan,Dashtayin mshakabuysereyev tarats- Bejenaru et al. 2010 vats molakhotere Hayastanitaratsqum nor qari – erkati L. Bejenaru/D. Monah/G.Bodi, Adeposit of astragali at darashrjannerum). Ph. D. thesis in biology by specializa- the Copper Age tell of Poduri-Dealul Ghindra, Romania. tion of ‘‘Agriculture’’. Scientificcenter of Agriculture and Antiquity: Project Gallery 84 (323), 2010. Plant protection (Yerevan 2009). http://www.antiquity.ac.uk/projgall/bejenaru323/ Hovsepyan/Willcox 2008 Buikstra/Ubelaker 1994 R. Hovsepyan,G.Willcox, The earliest finds of cultivated J. E. Buikstra/D. H. Ubelaker (eds), Standards for data plants in Armenia: evidence from charred remains and collection from human skeletal remains. Proceedings of crop processing residues in pise´ from the Neolithic sett- aseminar at the Field Museum of Natural History. Ar- lements of Aratashen and Aknashen. Vegetation History kansas archeological report research series 44 (Fayette- and Archaeobotany 17, 2008,supplement 1, 63–71. ville 1994). Iserlis et al. 2010 Franklin 2014 M. Iserlis/R. Greenberg/R. Badalyan/Y.Goren, Bet Yerah, K. J. Franklin. ‘‘This World is an Inn’’: Cosmopolitanism Aparan III and Karnut: Preliminary observations on Kura- and Caravan Trade in Late Medieval Armenia. Ph.D. dis- Araxes homeland and diaspora ceramic technologies. sertation, UniversityofChicago. Tu¨rkiye Bilimler Akademisi Arkeoloji Dergisi 13, 2010, Gale/Stos-Gale 1982 245–262. N. H. Gale/Z. A. Stos-Gale, Bronze Age copper sources in Khachatryan 1979 the Mediterranean. Science 216, 1982, 11–19. T. S. Khachatryan, Artikski Nekropol (Yerevan 1979). Gardin 1976 Khanzadyan 1967 J.-C. Gardin. Code pour l’analyse des formes de poteries. E. Khanzadyan,Haykakan lernashkharhi mshakuyte CNRS Editions (Paris 1976). m.t.a. III hazaramyakum (Yerevan 1967). Gilmour 1997 Khatchadourian 2014 G. Gilmour, The nature and function of astragalus bones L. Khatchadourian Empire in the Everyday: APreliminary from archaeological contexts in the Levant and Eastern Report on the 2008–2011 Excavations at Tsaghkahovit, Mediterranean. Oxford Journal of Archaeology 16,H.2, Armenia. American Journal of Archaeology 118(1): 137– 1997, 167–175. 169. Glonti et al. 2008 Kibarogˇlu et al. 2011 L. Glonti/M. Ketskhoveli/G. Palumbi, The cemetery at M. Kibarogˇlu/A. Sagona/M. Satir, Petrographicand geo- Kvatskhelebi. In: A. Sagona/M. Abramishvili (ed.), Ar- chemical investigations of the late prehistoricceramics chaeology in Southern Caucasus: perspectives from from Sos Ho¨yu¨k, Erzrum (Eastern Anatolia). Journal of Georgia. Ancient Near Eastern Studies, supplement 19 Archaeological Science 38, 2011, 3072–3084. (Leuven, Paris, Dudley, MA 2008) 153–184. Koerper/Whitney-Desautels 1999 Gopnik/Rotman 2011 H. C. Koerper/N. A. Whitney-Desautels, Astragalus bo- H. Gopnik/M. S. Rotman, On the high road. The history nes: artifacts or ecofacts? Pacific Coast Archaeological of Godin Tepe, Iran (Ontario 2011). Society Quarterly 35,H.2–3,1999, 69–80. Hanks/Linduff 2009 Koridze/Palumbi 2008 B. K. Hanks/K. M. Linduff (eds), Social complexity in pre- I. Koridze/G.Palumbi, Kura-Araxes tombs at Aradetis Go- historic Eurasia: monuments, metals, and mobility (New ra. In: A. Sagona/M.Abramishvili (ed.), Archaeologyin York 2009). Southern Caucasus: perspectives from Georgia. Ancient Hayrapetyan 2005 Near Eastern Studies, supplement 19,(Leuven, Paris, A. Hayrapetyan, Vagh bronzedaryan vznots Gegharotits. Dudley, MA 2008) 125–152. In: A. A. Kalantaryan/R. S. Badalyan/P. S. Avetisyan Kos¸ay/Vary 1967 (ed.), Hin Hayastani mshakuyte XIII. Nyuter hanrapeta- H. Z. Kos¸ay/H. Vary, Gu¨zelova Kazisi. Ausgrabungen von kan gitakan nstashrjani (Erevan 2005) 78–85. Gu¨zelova (Ankara 1967). Hayrapetyan 2008 Kushnareva 1997 A. Hayrapetyan, Some technical aspects of the pottery K. K. Kushnareva, The Southern Caucasus in prehistory of the Early Bronze Age site of Gegharot (Armenia). In: (Philadelphia 1997). K. S. Rubinson/A. Sagona (ed.), Ceramics in transitions. Leon/Smith 2012 Chalcolithicthrough Iron Age in the highlands of the J. Leon/A. T. Smith, Devotion and divination: the temple Southern Caucasus and Anatolia. Ancient Near Eastern fortress at Gegharot and rituality in the Ancient Cauca- Studies, supplement 27 (Leuven, Paris, Dudley, MA sus and Near East. Paper delivered at the 77th Annual 2008) 71–86. Meeting of the Society for American Archaeology (Mem- Hovsepyan 2008 phis TN) 2012,Pages?? R. Hovsepyan,Appendix 2.The palaeobotanical remains Lindsay et al. 2010 from Early Bronze Age Gegharot. In: R. Badalyan R/A. T. I. Lindsay/A.T.Smith/R. Badalyan, Magnetic survey in Smith/I. Lindsay/L. Khatchadourian/P. Avetisyan, Village, the investigation of sociopolitical change at aLate Bron- fortress, and town in Bronze and Iron Age Southern ze Age fortress settlement in Northwestern Armenia. Ar- Caucasia: apreliminary report on the 2003–2006 inves- chaeologicalProspection 17, 2010, 15–27. tigations of project ArAGATS on the Tsaghkahovit plain, Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 219

Lorenzi et al. 2000 and their possible relation with ore deposits of the Erz- F. Lorenzi/A. Nonza/H. Paolini-Saez, Terminologie pour gebirge. Archaeometry 45, 2003, 61–100. l’e´tude typologique de la ce´ramique (Ajaccio 2000). Palumbi 2008 Lutz/Pernicka 1996 G. Palumbi, The Red and Black. Social and cultural in- J. Lutz/E. Pernicka, Energy dispersiveX-ray fluorescence teraction betweenthe Upper Euphrates and Southern analysis of ancient copper alloys: empirical values for Caucasus communities in the fourth and third millenni- precision and accuracy. Archaeometry 38,H.2,1996, um BC (Roma 2008). 313–323. Pernicka et al. 1990 Marro 1997 E. Pernicka/F. Begemann/S.Schmitt-Strecker/A. P. Grima- C. Marro, La culture du haut-Euphrate au bronze ancien: nis, On the composition and provenance of metal ob- essai d’interpre´tation a` partir de la ce´ramique peinte de jects from Poliochni and Lemnos. Oxfrod Journal of Ar- Keban, Turquie (Istanbul et Paris 1997). chaeology 9,H.3,1990, 263–298. Marshall 2012 Petrosyan 2002 M. E. Marshall, Processes and practices of death: to- L. Petrosyan, Hayastani tsisakan arkghere. In: A. A. Ka- wards abioarchaeology of dynamic societies.In: C. lantaryan, S.Harutyunyan (ed.), Hayastani hnaguyn Hartley/B.Yazgicolu/A. T. Smith (ed.), The archaeologyof mshakuyte, 2.E.Khanzadyanihobelyanin nvirvats gitaz- power and politics across Eurasia: regimes and revoluti- hoghovi nyuter (Erevan 2002), 31–36. ons (Cambridge 2012) 157–172. Porter 2012 Marshall 2014 A. Porter, Mobile pastoralism and the formation of Near M. E. Marshall, Becoming bioarchaeology? Traditions of Eastern civilizations (Cambridge 2012). physical anthropology and archaeology in Armenia. In: Sagona 1984 B. O’Donnabhain, M. C. Lozada Cerna (ed.), Archaeologi- A. Sagona, The Transcaucasian region in the Early Bron- cal human remains: Global perspective. Cham 2014, ze Age. British Archaeological Reports, International se- 29–40. ries 214 (Oxford 1984). Mason/Cooper 1999 Sagona 2000 R. Mason/L. Cooper, Grog, petrology and Early Transcau- A. Sagona, Sos Hoÿu¨kand the Erzurum region in late casians at Godin Tepe. Iran, 37, 1999, 25–31. prehistory:aprovisional chronology for Northeast Ana- Meliksetian/Pernicka 2010 tolia. In: C. Marro/H. Hauptmann(ed.), Chronologies des Kh. Meliksetian/E. Pernicka, Geochemicalcharacterisati- Pays du Caucase et de l’Euphrate aux IVe-IIIe millenai- on of Armenian Early Bronze Age metal artefacts and res (Paris 2000) 329–373. their relation to copper ores. In: S. Hansen/A. Haupt- Sagona et al. 1995 mann/I. Motzenbacker/E.Pernicka(Hrsg.), Von Majkop A. Sagona/C. Sagona/H. O¨zkorucuklu, Excavations at bis Trialeti. Gewinnungund Verbreitung von Metallen Sos Hoÿu¨k 1994.First preliminary report. Anatolian Stu- und Obsidian in Kaukasien im 4., 2.Jt. v. Chr. (Bonn dies XLV, 1995, 193–218. 2010) 41–58. Sagona et al. 1996 Meliksetian et al. 2009 A. Sagona/M.Erkmen/C. Sagona/I.Thomas, Excavations Kh. Meliksetian/E. Pernicka/R. Badalyan, Compositions at Sos Hoÿu¨k 1995.Second preliminary report. Anatoli- and some considerations on the provenance of Armeni- an Studies, XLVI, 1996, 27–52. an Early Bronze Age copper artifacts. 2nd International Sagona et al. 1997 conference ‘‘Archaeometallurgy in Europe 2007’’. Se- A. Sagona/M. Erkmen/C. Sagona/S. Howells, Excavations lected papers (Milano 2009) 125–134. at Sos Hoÿu¨k, 1996.Third preliminary report. Anatolica Meliksetian et al. 2011a XXIII, 1997, 181–226. Kh. Meliksetian/S. Kraus/E. Pernicka, Metallurgy of pre- Schmid 1972 historic Armenia. Proceedings of the InternationalSym- E. Schmid, Atlas of animal bones (Amsterdam 1972). posium ,,Anatolian Metal V‘‘. Early mining and metallur- Silver 1969 gy in Anatolia and neighboring areas 13–15 November I. A. Silver, The ageing of domestic animals. In: D. 2008 (Bochum 2011) 201–210. Brothwell/E. Higgs (ed.), Science in archaeology (London Meliksetian et al. 2011b 1969) 283–302. Kh. Meliksetian/R. Schwab/S.Kraus/E. Pernicka/M. Smith and Leon 2015 Brauns, Chemical, lead isotope and metallographic ana- A. T. Smith/J. Leon, Divination and Sovereignty: The lysis of extraordinary arsenic-rich alloys used for jewelry Late Bronze Age Shrines at Gegharot, Armenia. Ameri- in Bronze Age Armenia. Archaeometallurgy in Europe III, can Journal of Archaeology. 118: 549–563. Bergbau Museum (Bochum 2011) 211–212. Smith et al. 2004 Monahan 2012 A. T. Smith/R. Badalyan/P. Avetisyan/M. Zardaryan, Early B.Monahan,Beastly goods: pastoral production in the complex societies in Southern Caucasia: Apreliminary Late Bronze Age Tsaghkahovit plain. In: Ch. Hartley/G. report on the 2002 investigations by project ArAGATS Bike Yazicioglu/A. T. Smith (ed.), The archaeology of on the Tsakahovit Plain, Republic of Armenia. American power and politics in Eurasia: regimes and revolutions Journal of Archaeology 108, 2004, 1–41. (Cambridge 2012) 337–347. Smith et al. 2009 Niederschlag et al. 2003 A. T. Smith/R. S. Badalyan/P. Avetisyan,The archaeology E. Niederschlag/E. Pernicka/Th. Seifert/M. Bartelheim, and geography of ancient Transcaucasian societies. Vol- Determination of lead isotope ratios by Multiple Col- ume 1.The foundationsofresearch and regional survey lector ICP-MS: acase study of Early Bronze Age artifacts in the Tsaghkahovit Plain, Armenia (Chicago 2009). 220 Ruben Badalyan et al.

Stos-Gale et al. 1984 `àäàº'í et al. 2005 Z. A. Stos-Gale/N. H. Gale/G. R. Gilmore, Early Bronze —. `àäàº'í/ˇ. ÀâåòŁæ'í/À. ÑìŁò, Ñâ'òŁºŁøå ïî Age Trojan metal sources and Anatolians in the Cycla- çäíåªî Æðîíçîâîªî âåŒà ˆåıàðîòà.In: A. A. Kalantary- des. Oxford Journal of Archaeology 11,H.2,1984 an/R. S. Badalyan/P. S. Avetisyan (ed.), Hin Hayastani 155–77. mshakuyte XIII. Nyuter hanrapetakan gitakan nstashrjani Tedesco 2006 (Erevan 2005) 109–115. L. A. Tedesco, Redefining technology in Bronze Age `àäàº'í/ÑìŁò 2008 Transcaucasia: copper-alloy metallurgy in Armenia in the —. `àäàº'í/À.ÑìŁò, ˇîæåºåíŁå ˆåıàðîò: îæíîâíßå ðå 3rd–mid 2nd millennium BC. PhD dissertation New York çóºüòàòß ðàæŒîïîŒ 2005–2006 ªª.In: P. S. Avetisyan/A. University 2006. A. Kalantaryan/R. S. Badalyan (ed.), Hin Hayastani msha- Torosyan et al. 2002 kuyte XIV. Nyuter hanrapetakan gitakan nstashrjani (Yere- R. M. Torosyan/O. S. Khnkikyan/L. A. Petrosyan, Hin Shi- van 2008) 45–68. rakavan (1977–1981 tt. peghumneri ardyunqnere) (Ere- ˆàäæŁåâ/Ìàªîìåäîâ 2008 van 2002). Ì. Ñ. ˆàäæŁåâ/—. ˆ. Ìàªîìåäîâ, Òîðïàı-Œàºà – Œóðî- Tylecote 1976 àðàŒæŒîå ïîæåºåíŁå ŁæàæàíŁäæŒîå ªîðîäŁøå â R. F. Tylecote, Ahistory of metallurgy (London 1976). þæíîì ˜àªåæòàíå.In: ˝. ß. Ìåðïåðò/Ñ. ˝. ˚îðåíåâæ- Weeks 1999 ŒŁØ (ed.), ÀðıåîºîªŁ'˚àâŒàçà Ł`ºŁæíåªî ´îæòîŒà. L. Weeks, United Arab Emirates:new data regarding the ÑÆîðíŁŒ Œ 80-ºåòŁþ ÷ºåíà-Œîððåæïîíäåíòà —À˝, ïð ‘tin problem’inWestern Asia. Antiquity 1999, 73, îôåææîðà —. Ì. Ìóí÷àåâà (ÌîæŒâà 2008) 276–297. 49–64. ˆåâîðŒ'í 1980 Weeks 2002 À. Ö. ˆåâîðŒ'í, ¨ç ŁæòîðŁŁäðåâíåØłåØ ìåòàººóðªŁŁ L. Weeks, SummaryofVelikent compositional and lead Àðì'íæŒîªî íàªîðü' (¯ðåâàí 1980). isotope analyses. Appendix: In: P. L. Kohl, Bronze pro- ˚ółíàðåâà/×óÆŁíŁłâŁºŁ 1970 duction and utilization in south-eastern Dagestan, Rus- ˚. Õ. ˚ółíàðåâà/Ò. ˝. ×óÆŁíŁłâŁºŁ, ˜ðåâíŁå Œóºüòó sia: ca. 3600–1900BC. In: M. Bartelheim/E. Pernicka/R. ðß Þæíîªî ˚àâŒàçà (¸åíŁíªðàä 1970). Krause (eds), The beginnings of metallurgyinthe Old Ìàıìóäîâ et al. 1968 World (Freiberg 2002) 179–183. Ô. À. Ìàıìóäîâ/—. Ì. Ìóí÷àåâ/¨. ˆ. ˝àðŁìàíîâ, ˛äð Yon 1981 åâíåØłåØ ìåòàººóðªŁŁ ˚àâŒàçà. ÑîâåòæŒà' M. Yon, Dictionnaire illustre´ multilingue de la ce´ramique àðıåîºîªŁ' 4, 1968, 16–26. du Proche-Orientancien. Maison de l’Orient et de la Me´- Ìóæàåâ 2006 diterrane´e(Lyon 1981). ˜. Ìóæàåâ, ÑåðŒåðòåïå – ïîæåºåíŁåýïîıŁ ðàííåØ Æðîíçß (`àŒó 2006). `àäàº'í 2011 ˇåòðîæ'í 1989 —. `àäàº'í, ˚âîïðîæóîïåðŁîäŁçàöŁŁ ŁıðîíîºîªŁŁ ¸. À. ˇåòðîæ'í, —àæŒîïŒŁïàì'òíŁŒîâ ˚åòŁ Ł´îæ- ˚óðî-ÀðàŒææŒîØ Œóºüòóðß ÀðìåíŁŁ.In: Ìå ŒåàæŒà (III – I òßæ. äî í.ý.) (¯ðåâàí 1989). æäóíàðîäíà' íàó÷íà' ŒîíôåðåíöŁ' «ÀðıåîºîªŁ', ღამბაშიძე et al. 2010 ýòíîºîªŁ', ôîºüŒºîðŁæòŁŒà ˚àâŒàçà». ÑÆîðíŁŒŒðàòŒ ი. ღამბაშიძე/გ. მინდიაშვილი/გ. გოგოჭური/კ. Łı æîäåðæàíŁØ äîŒºàäîâ (ÒÆŁºŁæŁ 2011) 63–68. კახიანი/ი. ჯაფარიძე, უძველესი მეტალურგია და სამთო საქმე საქართველოში ძვ. წ. VI–III ათასწლეულებში (თბილისი 2010).

Ruben Badalyan ArmineHarutyunyan Institute of Archaeologyand Ethnography Institute of Archaeologyand Ethnography of the National Academy of Sciences of the National Academy of Sciences of the Republic of Armenia of the Republic of Armenia 15,Charentsistr., Yerevan, 0025 15,Charentsi str., Yerevan, 0025 E-mail: [email protected] E-mail: [email protected]

AdamT.Smith Alan Greene Department of Anthropology Department of Anthropology Cornell University Stanford Archaeology Center 261 McGraw Hall Building 500, 488 Escondido MallMC Ithaca NY 14853 2170 Stanford, CA 94305 E-mail: [email protected] E-mail: [email protected]

Ian Lindsay Maureen Marshall Department of Anthropology Russian, East European, and Eurasian Center Purdue University UniversityofIllinois at Urbana-Champaign 700 W. State St., Suite 219 104 InternationalStudies Building West LafayetteIN47907 910 S. Fifth Street E-mail: [email protected] Champaign, IL 61820 E-mail: [email protected] Apreliminary report on the 2008, 2010,and 2011 investigations of Project ArAGATS 221

Belinda Monahan Khachatur Meliksetian 7430 Ridge Blvd. Institute of Geological Sciences Chicago IL 60645 of the National Academy of Sciences E-mail: [email protected] of the Republic of Armenia 24A, Marshal Baghramyan Ave., Yerevan, 0019 RomanHovsepyan E-mail: [email protected] Institute of Archaeologyand Ethnography of the National Academy of Sciences ErnstPernicka of the Republic of Armenia Institut fu¨rUr- und Fru¨hgeschichte und Archa¨ologie des 15,Charentsistr., Yerevan, 0025 Mittelalters der Universita¨tTu¨bingen, Germany; E-mail: [email protected] Curt-Engelhorn-Zentrum Archa¨ometrie, Mannheim,Germany E-mail: [email protected]

SamuelHaroutunian Paris – Sorbonne University (Paris IV) 1,rue Victor Cousin, 75005 Paris E-mail: [email protected]

Summary rot, Aragatsi Berd und Tsaghkahovit, Tsaghkahovit-Ebene (Republik Armenien) realisiert worden sind. This report presents the results of the collaborative ar- Auf den Siedlungen Gegharot und Aragatsi Berd chaeologicalfield investigations undertakenbetween 2008 wurden Schichten der Fru¨hen Bronzezeit (Kura-Araxes Kul- and 2011 under the auspices of the joint Armenian-Amer- tur) und Spa¨tbronzezeit, in Tsaghkahovit – eine Siedlung ican Project for the Archaeologyand Geography of Ancient und ein Gra¨berfeld der Spa¨tbronzezeit – untersucht.Die Transcaucasian Societies (Project ArAGATS). Here we focus stratigraphische Reihe der fru¨hbronzezeitlichen Schicht our discussions on investigations into the Bronze Age von Gegharot hat diskreten Charakter der Kura-Araxes communities of the TsaghkahovitPlain of central Armenia. Schicht, welche durch isolierte Horizonte repra¨sentiert ist; In particular, we detail here the results of excavationsat jeder davon charakterisiert einen spezifischen keramischen the Bronze Age complexes at the sites of Gegharot, Ara- Komplex. gatsi Berd, and Tsaghkahovit. At the settlements of Ge- Die fru¨hbronzezeitliche Schicht von Gegharot stellt gharot and Aragatsi Berd, investigations uncovered well- eine landwirtschaftliche Siedlung dar, dessen Konstruktio- preserved occupations of the Early Bronze Age (Kura-Ara- nen spiegeln die egalita¨re und funktionell schwach diffe- xes culture) and the Late Bronze Age. At Tsaghkahovit, renzierte Lebensweise der Bevo¨lkerung wieder. where excavations explored both settlement and mortuary Die spa¨tbronzezeitliche Schicht der Siedlung Gegha- contexts, our research was focused on the remains of the rot ist vor allem durch spezifische Komplexe gekennzeich- Late Bronze Age community. Study of the Early Bronze net, und zwar durch drei synchrone (Ende 14. – erste Age levels at Gegharot revealed stratified layers of occu- Ha¨lfte 13.Jh. v. Chr.) Heiligtu¨mer mit, im Allgemeinen, ana- pation representing discrete phases of the Kura-Araxes logem Innenbereich und Inventar, die eine einheitliche horizon defined by unique ceramic complexes. The Early Struktur des Fundortes zeigen. Bronze Age occupationsatGegharot indicate an agro-pas- Durch die Ausgrabungen der Siedlung ist eine Reihe toral village organized around agenerally egalitarian and von Episoden dramatischer Art dokumentiert, die die Sied- minimally differentiated social order. Late Bronze levels at lung in der Tsghkahovit Ebene wa¨hrend der zweiten Ha¨lfte Gegharot have uncovered three contemporary shrines, dat- des 2.Jt. v. Chr. gepra¨gt haben: sie zeugen von einer tota- ing to the late 14th through first half of the 13th centuries len Zersto¨rung dieser Komplexeund teilweise Rekonstruk- B.C. In contrast, investigations at the base of Tsaghkaho- tion von manchen davon. vit indicate aresidential complex,perhaps representing seasonal occupations around the fortress. Excavations at sites across the region have documented episodes of dra- —åçþìå matic conflagrationduring the second half of the 2nd millennium B.C. Taken together, the collective research of ˝àæòî'øŁØ îò÷åò ïðåäæòàâº'åò îæíîâíßå ðåçóºüòàòß Project ArAGATS reported here provides detailed new un- ðàÆîò, ïðîâåäåííßı àðì'íî-àìåðŁŒàíæŒîØ ýŒæïåäŁöŁåØ derstandings of life and death during the Bronze Age in âðàìŒàıˇðîåŒòà ArAGATS â 2008, 2010, 2011 ªª. íà the South Caucasus. ìíîªîæºîØíßı ïàì'òíŁŒàı Æðîíçîâîªî âåŒà ˆåıàðîò, ÀðàªàöŁ Æåðä ŁÖàıŒàîâŁò âÖàıŒàîâŁòæŒîØ ðàâíŁíå (—åæïóÆºŁŒà ÀðìåíŁ'). ˝à ïîæåºåíŁ'ı ˆåıàðîò Ł Zusammenfassung ÀðàªàöŁ Æåðä ŁææºåäîâàºŁæü æºîŁ ðàííåªî Æðîíçîâîªî (Œóðî-àðàŒææŒà'Œóºüòóðà) Łïîçäíåªî Æðîíçîâîªî âåŒîâ, Vorliegender Bericht fast die Hauptergebnisse der Arbeiten âÖàıŒàîâŁòå – ïîæåºåíŁå ŁìîªŁºüíŁŒ ˇ`´. ˙àôŁŒæŁ- zusammen,die durch die armenisch-amerikanische Expedi- ðîâàííà' ðàæŒîïŒàìŁ —` æºî' ˆåıàðîòà æòðàòŁªðàôŁ- tion im Rahmen des Projektes ArAGATS 2008, 2010, 2011 ÷åæŒà' ŒîºîíŒà îòðàæàåò äŁæŒðåòíßØ ıàðàŒòåð Œóðî- in den mehrschichtigen Fundstellen der Bronzezeit Gegha- àðàŒææŒîªî æºî', ïðåäæòàâºåííîªîäâóì' ïî æóøåæòâó 222 Ruben Badalyan et al.

ŁçîºŁðîâàííßìŁ ªîðŁçîíòàìŁ, ŒàæäßØ Łç Œîòîðßı ıàðàŒ- – ïåðâà' ïîºîâŁíà XIII ââ. äî í.ý.) æâ'òŁºŁøàìŁ æ òåðŁçóåòæ'æïåöŁôŁ÷íßì ŒåðàìŁ÷åæŒŁì ŒîìïºåŒæîì. àíàºîªŁ÷íßì âöåºîì Łíòåðüåðîì ŁíàÆîðîì Łíâåíòàð', —àííåÆðîíçîâßØ æºîØ ˆåıàðîòà ïðåäæòàâº'åò æåºüæŒî- îïðåäåº'þøŁìŁ óíŁŒàºüíßØ îÆºŁŒ ïàì'òíŁŒà. —àæŒîï- ıîç'Øæòâåííîå âæâîåØ îæíîâå ïîæåºåíŁå, ïîæòðîØŒŁ ŒàìŁ ïîæåºåíŁ' äîŒóìåíòŁðîâàí ð'ä äðàìàòŁ÷åæŒŁı Œîòîðîªîîòðàæàþò äîæòàòî÷íî ýªàºŁòàðíßØ ŁôóíŒ- ýïŁçîäîâ, ŁìåâłŁıìåæòîâÖàıŒàîâŁòæŒîØ ðàâíŁíå íà öŁîíàºüíî æºàÆî äŁôôåðåíöŁðîâàííßØ îÆðàç æŁçíŁ ïðîò'æåíŁŁâòîðîØ ïîºîâŁíß II òßæ. äî í.ý. ŁâßðàçŁâ- íàæåºåíŁ'. ˇîçäíåÆðîíçîâßØ æºîØ ïîæåºåíŁ' ˆåıàðîò łŁıæ'âòîòàºüíîì ðàçðółåíŁŁýòŁı ŒîìïºåŒæîâ Ł÷àæ- ïðåäæòàâºåí ïðåæäå âæåªî ŒîìïºåŒæàìŁ âåæüìà æïå- òŁ÷íîØ ðåŒîíæòðóŒöŁŁ íåŒîòîðßı Łç íŁı. öŁôŁ÷íîªîıàðàŒòåðà – òðåì' æŁíıðîííßìŁ (Œîíåö XIV