From Pella to

Hybridisation and Identity in the Art and Architecture of the Hellenistic East

Edited by

Anna Kouremenos, Sujatha Chandrasekaran and Roberto Rossi

with a foreword by

Sir John Boardman.

BAR International Series 2221

2011

Published by

Archaeopress Publishers of British Archaeological Reports Gordon House 276 Banbury Road Oxford OX2 7ED England [email protected] www.archaeopress.com

BAR S2221

From Pella to Gandhara: Hybridisation and Identity in the Art and Architecture of the Hellenistic East

© Archaeopress and the individual authors 2011

ISBN 978 1 4073 0779 4

Cover image: ‘Cybele Plate’ (silver and gold, d. 25 cm, c. 3rd B.C.) from Ai Khanum, the Temple with Indented Niches. National Museum, Kabul. Mus. No: 04.42.7. After F. Hiebert and P. Cambon (eds.), Afghanistan, Hidden Treasures from National Museum, Kabul, cover image/Pl. 11. Washington. (ISBN 978-1-4262-0295-7).

Printed in England by Blenheim Colour Ltd

All BAR titles are available from:

Hadrian Books Ltd 122 Banbury Road Oxford OX2 7BP England www.hadrianbooks.co.uk

The current BAR catalogue with details of all titles in print, prices and means of payment is available free from Books or may be downloaded from www.archaeopress.com

TEMPLE ARCHITECTURE IN THE IRANIAN WORLD IN THE * Michael Shenkar Hebrew University of Jerusalem

The  period following the Macedonian conquest of the The Avesta and the Rig Veda do not mention any temples is traditionally considered one of the or cultic structures.5 According to these sources, the ‘dark ages’ in the history of ancient .1 Very little is ancient Indo-Iranians practiced their religion under the known about Iranian religions and cult practices during open sky. However, besides open air terraces, the time when Alexander’s successors ruled most of the archaeological excavations have revealed that a second Iranian world.2 We possess no Iranian written sources type of cultic structure, ‘closed temples’, existed in the from this period and there are only a few brief mentions Iranian world before the Macedonian conquest.6 It is in Classical authors regarding Iranian religion or places instructive, that even during the reign of the Achaemenid of worship, leaving us with incomplete and inconsistent kings, who united most of the Iranian world under their evidence.3 Some indications (Herodotus, Histories 1.131- authority, no single architectural temple canon was 132) date from the period before the Macedonian developed and it seems that cultic structures reflected a conquest, while others (Plutarch, De hide et Osiride 46- variety of heterogeneous local cults, which existed among 47; Strabo, Geography 15.3.13-15) are from the later the Iranians. The Achaemenid kings probably worshipped period and refer mostly to Western Iran. The historians of under the open sky, since no closed temples have been Alexander’s campaigns do not compensate for this found in their royal capitals of Western Iran.7 On the deficiency.4 other hand, a number of buildings that had cultic and sacral significance were excavated in the Iranian world.8  I would like to thank Professors Joseph Patrich, Institute of Archaeo- In general, it should be noted that many more temples logy, Hebrew University of Jerusalem, Shaul Shaked, Department of were uncovered in Eastern than in Western Iran. Comparative Religion and Department of Islamic and Middle Eastern Studies, Hebrew University of Jerusalem, Frantz Grenet, École pratique des hautes études, Hannah M. Cotton, Department of Classics, Hebrew Fire was, and still is, one of the common components and University of Jerusalem Dr. Claude Rapin, École Supérieur, centre d’Archéologie, and Dr. Julia Rubanovich, Department of Islamic and supplementary elements of cultic activities in many Middle Eastern Studies, Hebrew University of Jerusalem, for their diverse cultures around the world. In Iranian cults, fire comments and invaluable help in improving this paper. and ashes had a major place and probably also a special 1 The term Iran as well as the Iranian world/region as used in this paper sacral significance, so prominent that the place of reaches beyond the borders of the modern Islamic Republic and refers to the region from the Hindu-Kush mountains in the East to the Zagros worship of modern Zoroastrians is usually called ridge in the West and from Transoxania in the North to the Persian Gulf ātaškada, (‘house of fire’) where the ever-burning fire is in the South, which in the 1st millennium BCE and the 1st millennium kept.9 Indeed, when one thinks about Iranian places of CE was inhabited by Iranian-speaking tribes and possessed common worship, the term ‘fire-temple’ is the one immediately cultural and religious ancestry. 2 This situation has even led to attempts to discuss the religious situation 5 of the Hellenistic period based on the Sasanian evidence: Hjerrild 1990. The Rig Veda and the oldest parts of the Avesta (which are ascribed to See a useful survey of the period, in: Frye 1984, 137-191; Boyce and Zoroaster) are generally dated, mainly on the linguistic grounds, around nd Grenet, 1991, 3-34. the second half of the 2 millennium BCE and c. 1000 BCE 3 These fragments were collected and published by de Jong 1997 with respectively. The dating of the Avesta is, of course, bound to the extensive commentary. See also P’yankov 1997. question of ‘Zoroaster’s time’, see no. 12. 6 4 Thus, Diodorus Siculus (Bibliotheca historica 17.114.4) mentions that On temples in the Iranian world before the Hellenistic period, see: Alexander followed a Persian custom of quenching the ‘sacred fire’ Shenkar 2007. 7 after the death of Hephaistion. Curtius Rufus (Historiae Alexandri Boucharlat 2005, 281. Magni 6.7.5) reports that the plotters against Alexander took oaths in 8 See: Shenkar 2007, 175-176. On recent important discoveries of the temple that was located in Drangiana in Seistan. See Boyce and sanctuaries in eastern Iran, see Rapin 2007, 39-42. Grenet 1991, 3-17. 9 See: Boyce 1989.

117 FROM PELLA TO GANDHĀRA

Fig. 1: The Iranian World in the Hellenistic Period. Drawing by Mitia Frumin invoked. However, the reality is that Avesta, our oldest the place of his doctrine in the broad context of other written source for the Iranian religion and the royal Iranian cults? What were the stages in the development inscriptions of the Achaemenian kings, do not mention and transformation of Zoroastrianism into the dominant such a term. It seems that the earliest certain evidence for religion in the and to what extent did the the existence of ‘fire-temples’ from which the ātaškada religious practices and beliefs of the Sasanian period and dar-e Mehr (‘house of Mithra’)10 of modern (recorded in Pahlavi literature) reflect those of earlier Zoroastrians directly derives, is only from the Sasanian periods? These extremely important questions for the period.11 It thus appears methodologically unsound to history of Zoroastrianism are not unanimously answered. apply this anachronistic term to Iranian temples before the rise of the Sasanian dynasty. It seems more appropriate to use the broad term ‘Iranian temples’ to refer to temples found in the Iranian world in The same caution is probably justified regarding the term the periods before the rise of the Sasanian dynasty, and to ‘Zoroastrian’ itself. Despite indefatigable scholarly suppose that they are connected to some various ‘Iranian efforts, the pre-Sasanian history of the Zoroastrian faith is cults’ whose exact nature eludes us, but of which still barely known. Of the beliefs and religious practices Zoroastrianism as it is known from the Sasanian period recorded in the Avesta, we do not know for sure what onward was obviously a part. peoples, periods and places are reflected, nor do we know when Zarathustra, the great prophet of the ancient Iranian religion, lived and preached his teaching.12 What was the WESTERN IRAN exact nature of his teachings and reforms and what was Persepolis 10 This name has been used for the fire-temple since the Islamic conquest: Boyce 1989. However see also Boyce 1993, where she claims Five temples may be attributed with relative certainty to that ‘dar-e mehr should have its origin in the Achaemenid period, the Hellenistic period: four in Eastern and one in Western despite its late attestation’. 11 On fire-temples in the Sasanian period see: Keall 1971; Boucharlat Iran (Fig. 13). The temple in Western Iran called the 13 1999b. ‘Frataraka Temple’ was excavated by Ernst Herzfeld in 12 The literature on ‘Zoroaster’s time and homeland’ is enormous. See for example: Gnoli 1980; P’yankov 1996; Gnoli 2000; Kellens 2001; 13 ‘Leader’s’ or ‘Governor’s’ temple, named after the title (frataraka) Shahbazi 2002; Stausberg 2002, 21-62; Dandamaev and Lukonin 2004, used by the dynasts of Pars during the Seleucid and Parthian period. On 320-329; Shaked 2005, 183-187. Most scholars date his activity around frataraka, see: Frye 1984, 158-162; Boyce and Grenet 1991, 110-116; the end of the 2nd ― beginning of the 1st millennium BCE and place his Wiesehöfer 1994, 101-136; Wiesehöfer 2001; Panaino 2002; homeland in Eastern Iran. Wiesehöfer 2007; Potts 2007, 272; Callieri 2007, 115-146.

118 M. SHENKAR: TEMPLE ARCHITECTURE IN THE IRANIAN WORLD IN THE HELLENISTIC PERIOD

Fig. 2: Frataraka Temple and the adjacent structures (after: Stronach 1985, fig. 4)

1923 and is located some 300 meters north-west of the one square and stepped, the other a typical Achaemenian Persepolis platform (Figs. 2-3).14 torus column, probably reused from the platform destroyed by Alexander.16 Square column bases are This mud-brick edifice consisted of rooms and corridors unknown in Achaemenian architecture.17 These, and the and an ayvān (portico) with eight columns. In the central, fact that the temple was not part of the royal complex on four-columned hall, a stone base for a statue was placed the Persepolis platform, strongly suggest a post- close to the rear wall.15 The columns were of two types: Achaemenian date.18

14 Herzfeld 1976, 275; Schmidt 1953, 56; Schippmann 1971, 177-185. For recent discussion, see: Litvinskiy and Pichikyan 2000, 226-240 and 16 Schmidt 1953, 56. especially: Callieri 2007, 51-68. 17 Stronach 1985, 615-616; Callieri 2007, 53. 15 See detailed discussion in: Callieri 2007, 56-62. 18 Callieri 2007, 64.

119 FROM PELLA TO GANDHĀRA

EASTERN IRAN

The Oxus Temple

In Eastern Iran, the remains of four temples were found. The first to consider was excavated during the 1970s and 1980s by Soviet archaeologists at Takht-i Sangin in , at the confluence of the Vakhsh and Panj rivers.24 The excavators date the erection of the temple to the end of the 4th, beginning of the 3rd century BCE.25 It was a major religious centre during the rule of the Seleucid and Greco-Bactrian kings, until it was devastated by nomad forces in the middle of the 2nd century BCE. However, it was subsequently restored and continued to serve as a temple till the 4th century CE.26

The temple stood in the middle of a massive temenos Fig. 3: Frataraka Temple (after: Litvinskiy protected by four towers. Its symmetrical plan consisted and Pichikyan 2000, fig. 57) of a tetrastyle main hall surrounded by L-shaped corridors and an eight-columned ayvān between the two wings (Fig. 4). These wings concealed altar rooms in In 1923 Herzfeld uncovered two reliefs in the area of the which layers of ashes were found.27 Outside, two temple depicting male and female figures in the position monumental altars of dressed limestone, of purely Greek of adoration.19 He thought that they were connected to the style and technique stood in front of the wings.28 temple cult; however, the reliefs in fact belonged to the Innumerable bones of sheep and goats were found separate building adjacent to the temple and should associated with them – a practice wholly alien to probably be dated to the later period.20 Another find of Zoroastrianism.29 cultic significance thought to be associated with the temple, but whose exact find spot is unknown, is a group More than 5000 artifacts of various kinds – votives, of Greek dedicatory inscriptions, to Zeus Megistos, weapons, and objects of art, dating from the pre- Apollo, Helios, Artemis and Athena.21 The excavators, Achaemenian period to the third century CE were Herzfeld and Schmidt, proposed that these are uncovered in the temple, scattered all over its floors or interpretatio Graeca of the Iranian deities Ahura Mazdā, dumped into many pits.30 Mithra and Anāhitā.22 This reflects a desire among part of the scholars to equate Greek deities found in Iran as The most important cultic find, dated to the Hellenistic representing Iranian divinities with similar characteristics period (first half of the second century BCE) was the and attributes. However, in most cases, when there are no statuette of Marsyas playing the flute, set on top of a further indications, one should be restrained from arriving votive altar with a Greek inscription: Εὐχὴν ἀνέθηκεν at such identifications. For instance, without additional Ὰτροσωκης Ὸξωι (‘Atrosokes dedicated [his] vow to evidence, preferably epigraphic, we simply cannot know Oxus’).31 In the same pit (no. 4, at the northern end of the whether the Iranians considered Greek imagery to corridor no. 2) a small female silver statuette (height 7.6 represent their own gods. The same holds true for the inscriptions mentioning Greek divinities. 24 The first two volumes of the final report so far published are dedicated to the architecture and the religious life: Litvinskiy and In the case of Persepolis, there is no evidence to support Pichikyan 2000; and to arms and armour: Litvinskiy 2001. Another the identification of the Greek deities with the Iranian important publication which discusses the finds from the Oxus temple in the broad context of Achaemenid and Hellenistic Bactria is: Pichikyan ones and the question of possible Greco-Iranian 1991. For good summaries in French, English and German respectively syncretism in the ‘Frataraka temple’ remains open. Its see: Litvinsky and Pichikyan 1981; Litvinsky and Pichikian 1994a; plan, lacking Greek architectural characteristics, suggests Litvinskij and Pičikjan 2002. The excavations at Takht-i Sangin were an Oriental cult, but the pedestal found in the cella is renewed in the late 1990s, see: Drujinina, 2001, Drujinina 2008. 25 Litvinskiy and Pichikyan 2000, 183; Litvinskiy 2001, 10. typical for Greek and Hellenistic marble statues, and 26 Litvinskiy 2001, 10. Long after the temple was destroyed the place fragments of such a statue have apparently been continued to be venerated by the local population: Litvinskiy and uncovered by Iranian archaeologists in recent Pichikyan 2000, 363-367. See also: Grenet and Drège 1987. excavations.23 Therefore it is possible that this temple 27 Litvinskiy and Pichikyan 2000, 98. housed a cultic statue of a deity whose identity is 28 Litvinskiy and Pichikyan 2000, 169-175. 29 Litvinskiy and Pichikyan 2000, 172. unfortunately not known to us. 30 Descriptions of many of these objects are found in: Pichikyan 1991, 19 some are the subjects of special publications: Litvinskii and Pichikian Herzfeld 1976, 286, Pl. LXXXVI. 1994a; Litvinsky and Pichikiyan 1994b; Litvinskij and Pičikian 1995a; 20 Boyce and Grenet 1991, 117; Koch 1993, 178. Litvinskij and Pičikian 1995b; Rapin 1995; Litvinsky 2000a; Litvinskiy 21 Herzfeld 1976, 275; Callieri 2007, 56. 2000b. See also Wood, this volume. 22 Herzfeld 1976, 275; Schmidt 1953, 56. They are joined by 31 Litvinskiy, Vinogradov, Pichikjan 1985; Bernard 1987; Pichikyan Wiesehöfer 2007, 39. 1991, 160-172. It is an interesting fact that many flutes were found in 23 Callieri 2007, 61-62. the temple: Litvinskij and Pičikian 1995a, 149.

120 M. SHENKAR: TEMPLE ARCHITECTURE IN THE IRANIAN WORLD IN THE HELLENISTIC PERIOD

Fig. 4: The Oxus Temple (after: Litvinsky and Pichikian 2000, fig. 50) mm, width 4.6 mm) was uncovered (Fig. 5).32 She is the ayvān.36 Of special interest are three heads made of dressed in chiton and himation and her bended right hand unbaked clay which were found in the eastern corridor. is holding an object which has been interpreted as a Two of them, wearing a royal diadem, are apparently pomegranate by Uzyanov.33 Only a small fragment of the portraits of both a young and an elderly Hellenistic ruler37 headdress (probably a kalathos) has survived, suggesting while the third shows a bearded man with a typical that it could be a goddess. Iranian headdress (kyrbasia).38

Also noteworthy is a clay statue (height 35 cm, width 16 The excavators proposed that the plan of the Oxus cm) of a youth with a quiver, resembling Apollo (Fig. Temple is derived from the monumental Bronze Age 34 temples of and whose main features 6), four korai who probably belong to the early 39 35 are axial symmetry and tower-like façades. Despite the Hellenistic style and fragments of a colossal bronze statue which probably stood on a plinth at the entrance to 36 Judging by the preserved fragments of feet (65 cm.) the statue could have been as tall as 5 m. Based on the material, which is highly unusual 32 The statuette is discussed in: Uzyanov 1987. However, in the first for the local artistic tradition and on the stance (left foot brought volume of the final report Litvinskiy and Pichikyan do not list this forward) the excavators think that the style of the statue was Greek: object among the finds of pit 4 in corridor 2: Litvinskiy and Pichikyan Litvinskiy and Pichikyan 2000, 123. The finds from recent excavations 2000, 86-87, nor the article of Uzyanov is mentioned in their indicate that the statue was probably four-footed: Drujinina and Inagaki bibliography. The statuette was, however, published by Pičikjan in 2008, 103-104. Zeymal’ 1985, n. 223. It is dated by Uzyanov 1987, 294, ‘not later than 37 They are identified as Seleucid kings by Pichikyan 1991, 189-194 and nd 2 century BCE’ on stylistic grounds and according to the excavators, as Greco-Bactrian kings or governors of Bactria, kings’ philoi by the latest object in this pit is the votive altar of Oxus with a Greek Litvinskij 2003, 51, 55. dedication: Litvinskiy and Pichikyan 2000, 87. But in the previous 38 st Pichikyan 1991, 194-196, calls it ‘a head of priest’ but there are publication of the statuette, Pichikyan, attributed it to the 1 centuries absolutely no indications for this identification, as nothing seems to CE: Zeymal’ 1985, n. 223. suggest a religious connection. Perhaps it should rather be considered as 33 Uzyanov 1987, 291. a representation of a donor. 34 Pichikyan 1991, 181-182. B. Litvinsky argues that this statue 39 Litvinskiy and Pichikyan 2000, 242-247. This is supported in a recent rd – nd represents a boy or Eros, and dates it on stylistic grounds to the 3 2 study by Korovchinskiy 2007, 18-19. Victor Sarianidi 1996, 325 argued centuries BCE: Litvinsky 2004, 69. that some architectural features of the Oxus Temple are already found in 35 Pichikyan 1991, 182-183; Litvinskii and Pichikian 1994a, 59-60. the temples of Margiana dated around 1000 BCE.

121 FROM PELLA TO GANDHĀRA

Hasanlu in Western Iran. They are also frequent in Achaemenian architecture at the royal seats of Western Iran such as Persepolis and .42 In recent years two Achaemenid sanctuaries incorporating tetrastyle halls were excavated at Sangyr-tepe and Bandikhan in Eastern Iran.43

It is worth noting that four-columned halls formed a core for a number of religious and secular buildings in the Iranian world in subsequent periods, such as the Parthian edifices at Old Nisa,44 Mansur-depe,45 Nippur and Assur, the so-called āyadanā in Susa,46 the temples of Bard-e Neshandeh47 and Kuh-i Khwaja, which most probably should also be attributed to the Parthian period (247 BCE – 224 CE), the Kushan royal temple at (second century CE),48 the sacred complex of Dedoplis Mindori in Georgia (2nd-1st centuries BCE),49 the rd th 50 Chorasmian palace at Topraq-Qala (3 -6 centuries CE) and the Sogdian temples of Panjikent (5th-8th centuries Fig. 5: Silver statuette of a goddess wearing a kalathos CE).51 A tetrastyle core also features in the temples of (?), The Oxus Temple (after: Uzyanov 1987, fig. 2) Sur, Sahr and Si‘a in Hauran, Syria (1st century BCE-1st century CE).52

Therefore it seems that the four-columned hall as an architectural element has a long history in the Iranian world and it appears not only in temples, as was sometimes claimed in the past,53 but also in secular and residential buildings.54 It is evident that in temples it was one of the central spaces of the edifice, but its exact function and the nature of the rituals performed there are unknown. In secular buildings the four-columned halls seem to have various functions, such as gates, assembly halls and probably even store rooms.

The altar rooms filled with ashes in the Oxus Temple indicate that activities and rituals connected to fire

42 Hopkins even thought that the tetrastyle hall was an Achaemenid development: Hopkins 1942, 16-17. Colledge also assumed that this ‘type of “centralized square” hall was perhaps contributed by Achaemenid Iran’: Colledge 1986, 10. For the most recent discussion of the origin and significance of the columned halls, see Gopnik 2010. 43 See: Rapin, Isamiddinov and Khasanov 2010, 21-22, and Borofka 2009. 44 Pilipko 1996. 45 Košelenko, Lašhin, Novikov 1989; Košelenko, Lašhin, Novikov 2000. Fig. 6: Statue of Apollo (?), The Oxus Temple 46 th (after: Zejmal’ 1985, fig. 204) This building excavated, in the 19 century, was for a long time considered a ‘fire-temple’ of the Achaemenid period: Schippmann 1971, 266-274; Litvinskiy and Pichikyan 2000, 209-216. However, it should probably be dated to the later Hellenistic-Parthian period and its close similarity between the layouts (especially the tower- cultic interpretation is very doubtful: Boucharlat 1997, 62-63; Boucharlat 2005, 242. like façade), it is difficult to bridge the chronological and 47 geographical gap. Ghirshman 1976, 5-55; Schippmann 1971, 251-258. 48 Schlumberger 1961; Schlumberger 1975; Fussman 1983; Schlumberger, Le Berre, Fussman 1983. Tracing the evolution of the central, four-columned hall is 49 Gagošidze 1983; Gagošidze 1992. of great importance in understanding the origin of both 50 Rapoport 1994; See also: Grenet 1986. the Oxus and the Frataraka temples.40 An open-air altar 51 Škoda 1987; Shkoda 1996; Shkoda 1998 and especially: Shkoda 2009. surrounded by four pillars was part of the edifice of 52 41 On these temples see: Netzer 2003, 103-109. On Nabatean temple Jarkutan in Northern Bactria between 1400-1000 BCE. architecture, see also: Tholbecq 1997. Such halls already appeared between 1000-800 BCE in 53 Most recently: Sarianidi 1996, 321. 54 Pugachenkova 1973. This is also supported by the latest 40 See detailed discussion in: Litvinskiy and Pichikyan 2000, 259-264. investigations at Takht-i Sangin conducted by A. Drujinina, who 41 Askarov and Širinov 1991; Askarov and Shirinov 1994; Görsdorf and supposes that one of the altar rooms (5) was used as ātaŝhgāh already Huff 2001. from the earliest period, see Druzhinina 2010, 230.

122 M. SHENKAR: TEMPLE ARCHITECTURE IN THE IRANIAN WORLD IN THE HELLENISTIC PERIOD

probably occupied an important place in the temple cult.55 The god Oxus-Wakhš was apparently the chief deity of According to P. Bernard the evidence for the extensive the temple.65 However, the cultic and constructional use of fire in the altar rooms (layers of ashes) dates only duality of the Oxus Temple (a pair of altars in the altar to the Kushan (1st-2nd centuries CE) and later periods, and rooms in the wings, a pair of monumental altars in the the original function of these rooms is uncertain.56 How- courtyard, a pair of cylinder bases in the central hall) may ever, the excavators claim that it could be safely estab- suggest that another deity beside Oxus was worshipped in lished that the altar, in at least one room, is contemporary the temple. If the silver statuette, mentioned above, with the wall, which was erected in the Hellenistic indeed dates to the Hellenistic period, one should not period.57 Water, the element associated with the river god exclude the possibility that it depicts a goddess Oxus, also seems to have had a significant part in the worshipped alongside Oxus, though any attempt to temple cult.58 This is not surprising, since fire and water identity her is speculative at best.66 Stressing the played a major role in ancient Iranian religious practice.59 importance of water in the cult of the temple, the excavators propose that the second deity of the Oxus Although the Greek altars, a statue of Apollo (?) and Temple was the Iranian fertility and water goddess numerous donations of objects of Hellenic art could Anāhitā.67 However, her cult is not attested in Eastern Iran indicate that a Greek cult occupied a prominent place at until the Kushano-Sasanian period. the Temple of Oxus, the temple was undoubtedly also attended by the local Bactrians. Greco-Iranian cultic The numerous finds of purely Greek art may be hybridization comes to fruition in the Dedication of considered as evidence for the cult of a Greek deity, Atrosokes mentioned above. An individual with an perhaps Apollo, who is possibly represented by the statue Iranian name made an offering to a god called Oxus, as found in the temple. However, examples of non-Greek art attested by the Greek dedication inscribed on the altar. In (such as the clay head of a bearded man with a kybrasia), all probability, ‘Oxus’ refers to the Iranian god Wakhš, an the dedication of Atrosokes and the very architecture of ancient Iranian name of the Amu-Darja River.60 The the building, indicates that the temple was undoubtedly a visual representation on top of the altar of the Greek major religious centre for the local Bactrians. If we Marsyas, also a river deity from Asia Minor (Phrygia).61 are to try to suggest interpretatio iranica for the son of However, this representation did not became a standard Zeus and Leto, Mithra and Tīr are the attested iconography of Oxus, since on the coin of the Kushan identifications.68 It is perhaps significant that the king Huviška (150/51-191/92 CE), Wakhš (OAΧϷO) is representation of Tīr (identified by the inscription presented as a bearded man with a nimbus surrounding ‘TEIPO’) – on the unique coin of the Kushan king his head and holding a staff (or perhaps a trident or spear) Huviška – shows him as a feminine youth holding a bow in his right hand and a fish in his left.62 It seems that by in one hand and pulling out an arrow from the quiver with the Kushan period, the iconographic type of Marsyas the other.69 playing the flute had been abandoned and forgotten (or, perhaps, the statuette from the Temple of Oxus is an Vakhsh and Panj rivers: Grenet and Drège 1987, this might possibly reflect a survival of local traditions of sanctity related to the long example of the unique type which did not gain wide destroyed Temple of Oxus: Litvinskiy and Pichikyan 2000, 315-316. popularity), and the representation of Oxus was probably For the discussion and interpretations of the motif of the horse in the being modeled on the image of Poseidon as he appears on Sogdian art see: Riboud 2003. the coins of the Indo-Scythian king Maues (3rd quarter of 65 Bernard 1994a, 97-98; This is confirmed by a recently found Greek the 2nd century BC).63 Moreover, in the Sogdian period, dedicatory inscription: ‘Seiromois from Molrparles, the son of Nemiskos, presented this 7 talents bronze vessel to the newly revived the probable depiction of the worship of Oxus in form of Oxus’: Drujinina 2008. It is plausible that Oxus-Wakhš was one of the a horse is found on the Sogdian funerary coach, known as most important divinities in Bactria. As late as 7th and 8th c. CE, several ‘The Miho coach’.64 Bactrian documents name him ‘The king of gods’: Sims-Williams 2001; Sims-Williams 2003. He was a popular deity not only in Bactria, but 55 also in other regions of Eastern Iran, as is indicated by many personal The excavators decisively identified these rooms as ātašgāh (‘fire names containing the name of this god. The inscription on a coin of place’), repository for a sacred fire and ashes in a Zoroastrian fire- Andragoras, the ruler of , demonstrates that Oxus was probably temple: Litvinskiy and Pichikyan 2000, 98, 206; Grenet 2004, 378. 56 worshipped in Parthia and Hyrcania: Lerner 1999, 25. Wakhš was also Bernard 1994a, 86-90. As F. Grenet notes, ‘the facts were not very prominent deity in Sogdiana where many anthroponyms contain properly registered and are now beyond recovery’ Grenet 2004, 378. his name: Livshits 2004, 190; Grenets 2004, 377. Korovchinskiy at first 57 Litvinskiy and Pichikyan 2000, 105. However, much of the thought that the cult celebrated in the Oxus Temple was connected to information concerning these rooms is ambiguous: Grenet (forthcoming Mithra: Korovchinskiy 2001. However, in his dissertation he changed b) n. 98. his mind and suggested that the Oxus Temple was dedicated to Oxus 58 See: Litvinskiy and Pichikyan 2000, 308-312. and the goddess Nana: Korovchinskiy 2007, 18-22. See also: Boyce and 59 Bernard 1994a, 109. Grenet 1991, 179-181. On Nana and her cult, see: D’Yakanova and 60 Litvinskiy and Pichikyan 2000, 313. Smirnova 1967; Azarpay 1976; Tanabe 1995; Grenet and Marshak 61 Litvinskiy and Pichikyan 2000, 304; Boyce and Grenet 1991, 177, 1998; Potts 2001; Ambos 2003; Ghose 2006. 180. Göbl 1984, 241/1. 66 For instance, Uzyanov 1987, 294 identifies her as syncretistic Magna 62 Rosenfeld 1967, 92, pl. VIII, 155. Mater, combining the iconographic traits of ‘local goddess of fertility, 63 Rosenfeld 1967, 92. Greek Demeter, Aphrodite Urania and Anāhitā’. 67 64 Grenet 2007, 410. The couch is kept in the Miho Museum in Japan. Litvinskiy and Pichikyan 2000, 353. 68 The discoveries of additional tombs of Sogdian merchants in northern See the comprehensive discussion and in-depth analysis in: Bernard in recent years suggest that the Miho couch comes from Anyang 1990b, 52-62. region and should be dated to the second half of the 6th century CE. See: 69 Rosenfeld 1967, 101, pl. X, 204; Bernard 1990b, 55, fig. 23. In Marshak 2001, 233-244; Grenet 2007, 409-410. It is noteworthy, that Sogdian religious iconography Tīr was probably represented as a four- one 8th century CE Chinese source even speaks of a temple with a handed deity holding an arrow: Grenet and Marshak 1998, 10- remarkable statue of a horse which was located at the confluence of the 15.

123 FROM PELLA TO GANDHĀRA

Fig. 7: The ‘Dioscuri Temple’, earliest stage (after: Kruglikova 1977, fig. 2)

Dilberjin associated with the wall, the fresco was dated to the end of the Greco-Bactrian period.74 The site of Dilberjin is located in the desert of northern Afghanistan, some 40 km from the city of Balkh. The Several attempts have been made to reconsider this dating remains of a large, fortified city dating to the Hellenistic and to attribute the mural and thus also the erection of the and Kushan periods were uncovered at the site by an temple to the Kushan or even later periods.75 On the other Afghan-Soviet expedition during the 1970s.70 According hand, if the paintings were indeed of the Kushan period, to the excavators, it was founded by the end of the Greco- one would expect to recognize in their iconography the Bactrian era (c. 150 BCE), and existed till the 5th century strong admixture of Iranian and Indian art which CE.71 characterizes the period, while the Dioscuri from Dilberjin are ‘of pure Greek form and in a broadly Greek 76 The temple was erected in the middle of a temenos in the style’. Despite the fact that only one Greco-Bactrian 77 northeastern corner of the site, close to the city wall. Its coin was found under the earliest floor in the temple , the plan consists of three discrete units: a cella surrounded by Greek style of the Dioscuri painting, devoid of any 78 three corridors and additional rooms arranged on either visible Oriental influence, and the ceramics, many types side of the entrance, which was on the eastern side (Fig. of which fit in well with the Hellenistic assemblages of 79 7). Benches were set up close to the walls of the side Ai Khanoum, indicate that the construction of the rooms and along the outer wall at the back of the central 74 Kruglikova 1976; Kruglikova 1986, 106. hall.72 75 Fitzsimmons 1996; Bernard 1996, 236; Lo Muzio 1999. 76 Boardman 1993, 103. According to Kossolapov and Marshak, the The excavations of the earliest stage of the temple did not paintings are ‘probably Hellenistic or, at least, preserving the Hellenistic yield any significant finds except for wall paintings of tradition’see Kossolapv and Marshak 1999, 24. Boyce and Grenet 1991, 173 n. 91, also think that ‘their type is purely Greek’. It is interesting to standing Dioscuri which decorated the western wall of compare them with the image of Dioscuri, as identified by Frantz the ayvān, flanking the entrance (height 2.15 m., width Grenet, from a 2nd century CE wall painting from Dalverzintepe: Grenet 1.60 m) (Fig. 8).73 The divine twins are depicted naked, (forthcoming a). The ‘Oriental’ style of the Dalverzintepe mural differs significantly from that of the Dilberjin Dioscuri. with two white horses, and wearing their distinctive pylos 77 hats. Based on its artistic style and on ceramic material The bronze coin of king Euthydemos I probably dated between 235- 200 BCE: Kruglikova 1986, 18. 78 The excavator, I. Kruglikova, emphasizes that the style and the 70 Kruglikova 1974; Kruglikova and Pugachenkova 1977; Kruglikova technique of the painting is wholly different from the later period 1977, 407-427; Kruglikova 1986; Bernard 1996. murals uncovered in Dilberjin: Kruglikova 1976, 91. On the technical 71 Kruglikova 1977, 407. Some architectural remains and ceramics with peculiarities of the Dioscuri painting, see: Buryj 1976, 111-113. parallels in the material of the Achaemenid period were also found at 79 Kruglikova 1986, 21-23. According to Bernard the ceramics collected Dilberjin: Dolgorukov 1984, 75-77. from the earliest stages of the temple allow its dating to the Hellenistic 72 The most detailed publication of the temple is: Kruglikova 1986, 6- period. However, B. Lyonett specifies that this material ‘is not 34. incompatible’ with a slightly later date, after the fall of the Greco- 73 Kruglikova 1974, 22-24; Kruglikova 1976, 88-93. Bactrian kingdom: Bernard 1990a, 55.

124 M. SHENKAR: TEMPLE ARCHITECTURE IN THE IRANIAN WORLD IN THE HELLENISTIC PERIOD

Fig. 8: Mural painting showing the Dioscuri, The ‘Dioscuri Temple’ (after: Kruglikova 1979, fig. 52)

Dioscuri temple should rather be dated to the Hellenistic goddess Nana, the most important female deity in Bactria period. and Sogdiana whose connection to the moon is well- known. It seems unlikely that the temple at Dilberjin was dedicated to the Dioscuri. This was a very widespread It has been proposed to link the plan of the Dioscuri motif in Hellenistic and Greco-Roman art.80 In the Greek temple to the plan of the Ai Khanum temples, which I world, the divine twins were frequently described as ‘gate shall discuss shortly, or to look for possible prototypes in keepers’ and the location of the painting suggests that this Bactrian secular architecture.86 It seems, however, that the could also be their function at Dilberjin.81 They are often layout of the Dilberdjin temple is in fact very close to the depicted standing on either side of a deity; in the West, Oxus Temple at Takht-i Sangin.87 The two temples are this would be a male supreme god, for example Zeus, composed of three main units,88 a central hall surrounded while in the East the Dioscuri are more commonly shown by corridors and side rooms which open to the courtyard. flanking a female deity, probably connected to the The main difference between them seems to be the moon.82 absence of columns in the Dioscuri temple, which, together with a dearth of valuable finds and offerings, Possible evidence for the cult of a goddess was found on indicates that it was a relatively modest building wall paintings in two rooms excavated not far from the compared with the monumentality and extraordinary temple.83 The mural depicts a female head wearing a wealth of the Oxus Temple. helmet, an image widely interpreted as Athena-Anāhitā84 85 or Athena-Arštāt. But there is no direct connection The temple of Mohra Maliaran at Taxila, dated to the 1st between these rooms and the Dioscuri temple and they century BCE, could be attributed to the same type and are also dated to the later Kushan period. Thus the perhaps regarded as an ‘intermediate’ variant between the identity of the principal deity of the temple remains Oxus and the Dioscuri temples.89 It consists of a obscure. Although we can speculate that she could be the rectangular cella with two columns and four additional

80 columns set in an ayvān. Another interesting example of For a survey of iconography of the Dioscuri, see: Azarpay 1988; Lo this architectural type is the Kushan palace/temple at Muzio 1999, 46-49. 81 Boyce and Grenet 1991, 173. However see Bernard 1996 234, who which has two columns in the cella like that thinks that the temple was dedicated to the Greek Dioscuri and Kruglikova 1976, 88 who proposes the identification with the Indian 86 Bernard 1988, 55; Boyce and Grenet 1991, 172; Rapin 1994, 136. Ašvins. 87 Bernard 1994a, 96-97. 82 Lo Muzio 1999, 46-49. 88 This layout is known also in other examples of Bactrian architecture, 83 Kruglikova 1976, 96-100; Kruglikova 1986, 109-110. but it is not unique to Bactria as claimed by Rapin 1994, 136. 84 Kruglikova 1986, 110; Lo Muzio 1999, 58. 89 Rapin 1995, 284-287. Although a non Iranian cult, perhaps a 85 Grenet 1987. Buddhist one, was celebrated in this temple.

125 FROM PELLA TO GANDHĀRA

Fig. 9: The ‘Temple with Indented Niches’, Aï Khanum, stage III (after: Bernard 1970, fig. 16) of Mohra Maliaran and one row of four columns in probably Eucratadea, was founded by Seleucus I around contrast to two such rows in the Frataraka and Oxus 300 BCE and plundered and devastated by the temples.90 nomads invading from the north some 155 years later.92 It had a triangular shape (1600 x 1800 m.) and was divided Aï Khanum into two main parts, an acropolis (60 m. height) and a lower town (400 x 1800 m.) where most of the public The remaining two temples dated to the Hellenistic buildings were located. The city’s necropolis was located period were uncovered during the French excavations at outside the massive city ramparts built of mud-bricks. the Greco-Bactrian city of Ai Khanum, which occupied a The architecture of Ai Khanum displays a blend of Greek 93 strategic position at the confluence of the two rivers, and Oriental elements. Along with the gymnasium, 94 95 Amu-Darya (Oxus) and its tributary Kokcha in northern theatre, fountain and a heroon of the city’s oikist 96 Bactria.91 The city, whose name in the later period was (founder), – institutions characteristic of a typical Greek

92 90 Pugachenkova 1966; Pugachenkova 1971. For the most recent survey, Bernard 2008, 87, 104. 93 see: Bernard 2008 and especially Bernard 2009, which summarizes the Bernard 1975, 189-193; Bernard 1976a, 293-303; Bernard 1978, research done by the DAFA at the site 421-429; Veuve 1987; Bernard 2008, 89-91; Bernard 2009, 43- 91 Several final reports have already been published so far in Mémoires 44. 94 de la Délégation archéologique française en Afghanistan (vols. 21; 26- Bernard 1976a, 314-322; Bernard 1978, 429-441; Bernard 2008, 89- 32) along with numerous preliminary reports written by the principal 91; Bernard 2009, 44-45. excavator of Ai Khanum Paul Bernard: Bernard 1966; Bernard 1967a; 95 Bernard 1976a, 307-314; Bernard 2009, 45; Pichikyan 1991, 224-225. Bernard 1968; Bernard 1969; Bernard 1970; Bernard 1971; Bernard 96 Bernard, Le Berre, Stucki 1973. Two additional mausoleums were 1973; Bernard 1974; Bernard 1975; Bernard 1976a; Bernard 1978. The uncovered at the site: Bernard 1975, 180-189; Bernard 2009, 47-49; pre-Hellenistic name of the site was probably *Oskobora: Rapin 2005. Pichikyan 1991, 232-239.

126 M. SHENKAR: TEMPLE ARCHITECTURE IN THE IRANIAN WORLD IN THE HELLENISTIC PERIOD

city – a huge palace complex,97 residential mansions, and temples followed Oriental architectural traditions. 98

What What seems to be the principal temple of Ai Khanum was an almost square building which stood on a three-stepped podium within a wide temenos on the main street of the city, close to the palace.99 The exterior of its massive mud-brick walls was decorated with indented niches which gave the edifice its modern name – ‘Temple with Indented Niches’ (Fig. 9).100 The temple had a flat roof with a decorative cornice. A vestibule led to a cella divided into three rooms, a central one flanked by two smallersacristies. Marble fragments of a foot (Fig. 10) and a hand were all that was left of a huge acrolithic statue, three times a man’s size, which once stood in the cultic niche in the back wall of the central hall.101 Fig. 10: Fragment of a foot of a huge acrolithic statue, Three mud-brick bases were found on both sides of a ‘Temple with Indented Niches’, Aï Khanum pronaos, close to the walls. It seems that mud and stucco (after: Bernard 1969, fig. 16) statues once stood on them.102 French archaeologists have succeeded in reconstructing two ‘masks’ from the fragments, male and female, which may be attributed to the Greek sculptural tradition.103

Another significant find at the temple was a gilded silver plate probably originally nailed to a staff and which served as a cultic standard (Fig. 11).104 A goddess that is generally interpreted as Cybele, is shown standing in a chariot driven by Nike and drawn by two lions across a mountainous landscape towards a priest in a Syrian-style outfit who is standing on a stepped altar, and is placing an incense burner on it. Another priest marches behind the chariot, holding a parasol above the goddesses. A star, the crescent moon and the sun-Helios are depicted above this procession.

Paul Bernard was of the opinion that the plate was imported from Syria, since the scene is typical for northern Syria and the priest’s garments resemble those

97 Bernard and Le Berre 1973; Bernard 1976a, 288-293; Bernard 1978, 444-461; Bernard 1994b, 110; Bernard 2008, 88-89; Bernard 2009, 41- Fig. 11: The ‘Cybele plate’, ‘Temple with Niches’, 43. Aï Khanum (after: Bernard 2008, fig. 23) 98 Pichikyan 1991, 114; Bernard 2008, 91-92; Bernard 2009, 41-43. 99 The first temple was constructed under Antiochus I and the second Greco-Bactrian temple was built during the reign of Diodotus II: Martinez-Sève 2010. of priests of Atargatis.105 Pugachenkova speculated that 100 Also known as temple à redans or temple à niches indentées. Only this scene could depict a true historical event – the the small finds from the temple have been fully published so far: transportation of a monumental statue of Cybele to Francfort 1984. For preliminary reports and discussion of architecture 106 and finds see: Bernard 1970, 321-348; Bernard 1971, 414-432; Bernard Bactria. Although the plate was found in one of the 1989; Bernard 1990a, 51-53; Downey 1988, 65-73; Hannestad and Potts smaller rooms of the cella, it remains difficult to decide if 1990, 93-95; Pichikyan 1991, 239-243; Boyce and Grenet 1991, 165- it originally belonged to the temple or perhaps was 171; Litvinskiy and Pichikyan 2000, 283-289; Bernard 2008, 92-93; brought there later, after the temple was devastated and Bernard 2009, 49-51, and especially the recent and very important 107 article by Martinez-Sève, who is currently preparing the final report on turned into a storage place On the other hand, the ‘The Temple with Indented Niches’, to be published in the Mémoirs de remains of a plate showing a lion harnessed to a chariot DAFA. were discovered in the Oxus Temple at Takht-i Sangin.108 101 Bernard 1969, 338-341; Pichikyan 1991, 245-246. If the plate from the Oxus Temple depicted a very similar 102 Bernard 1969, 344; Pichikyan 1991, 249. 103 scene, the relevance of the ‘Cybele plate’ to the cult of Boyce and Grenet 1991, 169 proposed that these could be the images of Greek donors. According to Korovchinskiy 2007, 12 there are 105 absolutely no physiological indications that one of the ‘masks’ is Bernard 1970, 345; Francfort 1984, 98, 102-103. female, the faces are possibly the images of male twins – the Dioscuri 106 Pugachenkova 1974, 125. who are connected to Zeus. 107 Bernard 2008, 104. 104 Bernard 1970, 339-347; Francfort 1984, 93-105; Pichikyan 1991, 108 Pichikyan 1991, 103-104. It is noteworthy that a plaque depicting 256-259. Helios was also found in the Oxus temple, see Litvinsky 2000b.

127 FROM PELLA TO GANDHĀRA

the ‘Temple with Indented Niches’ seems to be statue that stood in the central hall should be interpreted reasonable. Moreover, it prompts important questions as the remains of the chief deity venerated in the temple. regarding the possible connection between the cults of the The pure Greek naturalistic style of the fingers and a two temples. winged thunderbolt which decorates the sandal, suggest that they were once part of a statue of Zeus.117 However, It was noted by the scholars who discussed the ‘Temple some scholars have been skeptical about the possibility with Indented Niches’ that its layout follows an Oriental that a Greek god was worshipped in a purely Oriental or more specifically, Mesopotamian architectural building and proposed that a syncretistic Greco-Oriental tradition.109 Interesting attempts have been made to draw deity was the lord of the ‘Temple with Indented parallels with northern Syrian temples of the 3rd and 2nd Niches’.118 In his early publications, the excavator of Ai millennia BCE110 but the chronological gap is too big to Khanum, Paul Bernard, cautiously proposed that it should be effectively bridged. There is also a theoretical be identified with Ahura Mazdā, as he is known in the possibility that the layout of the ‘Temple with Indented royal cult of Antiochus I at Commagene.119 Niches’ could be the Hellenistic continuation of a Bactrian architecture of the Achaemenian period, but Vessels buried in the temple podium probably indicate examples of such architecture are still to be found.111 that libations were part of the cultic practice. Based on that evidence and on the fact that on coins of the Greco- The prototype for the plan of the ‘Temple with Indented Indian king Heliocles I, Zeus is depicted with rays of Niches’ could be the Neo-Babylonian ‘Temple A’ at light around his head, and during the reign of Hermaeus Ashur.112 This distinctive rectangular, nearly square, plan and Amynthas he even wears a ‘Phrygian cap’, Franz of a tripartite cella, entered via a vestibule, appears in Grenet suggested that in Eastern Iran Zeus was identified Mesopotamia also in a later, Parthian period in the with Mithra, a solar deity who had some chthonic temples of Dura-Europos (especially the Temples of characteristics, and not with Ahura Mazdā, and thus Artemis and of Zeus Megistos).113 This indicates the Zeus-Mithra could be the god of the ‘Temple with continuity of this type in Mesopotamia. Probably under Indented Niches’.120 Further suggestions seek to identify Parthian influence, a very similar layout also appears in the Zeus of Ai Khanum with the Syrian Hadad121 or three late Nabatean temples (Qasr Bint Firaun in Petra, propose that a Seleucid king was worshipped in the Qasr Rabba, Dibon).114 temple as Zeus.122

The high-stepped podium on which the Ai Khanum A few more words should be said about the significance temple stood is apparently part of the Iranian cultic of Hadad. This Semitic storm-god was introduced in Iran tradition and is frequently found in Iranian architecture, at the beginning of the 2nd millennia BCE and became a the most famous examples being the mausoleum of Cyrus part of the Elamite pantheon.123 He is mentioned in the 115 the Great at Pasargadae. The initial prototype behind Persepolis Fortification Tablets and was probably these podiums could be the ziggurats of Mesopotamia considered a native god by the inhabitants of Achaemenid 116 and Elam, but it seems that during the Achaemenid Fars and its rulers,124 an example of the Iranian and period it was fully absorbed into the eclectic imperial Elamite religious acculturation, which was actually the architecture. Persian Persepolis pantheon.125 Therefore it would not be surprising to find Hadad worshipped in Hellenistic Despite the rich finds made in the temple, it is not easy to Bactria, but we do not possess any decisive evidence for establish what deity or deities were worshipped in the this. ‘Temple with Indented Niches’. Fragments of a colossal

Together with all of these attempts at interpretation, it 109 For example: Pugachenkova 1982, 34-35; Colledge 1986, 10; Downey 1988, 73; Litvinskiy and Pichikyan 2000, 287; Bernard 2008, should be noted that the official pantheon of the Greco- 93. Bactrian kings, as demonstrated by their coinage, mostly 110 Litvinskiy and Pichikyan 2000, 296-299. Supported by: represents deities of a purely Greek iconography,126 and Korovchinskiy 2007, 9. until a relatively late period we do not find iconographic 111 Bernard 1990a, 52. Mairs (forthcoming), tries to find additional arguments for this. 117 Bernard 1990a, 53. 112 Downey 1988, 67. For the description of this temple, see: Downey 118 For example: Downey 1988, 73. 1988, 149-151. 119 113 Bernard 1974, 298. On Commagene and its royal cult see: Bernard 1990a, 51-52. On the temples of Dura-Europos, see: Downey 1988, 76-131. Waldmann 1991; Boyce and Grenet 1991, 309-352. Additional evidence for the identification of Zeus with Ahura Mazdā is citied in: Boyce and 114 Netzer 2003, 68-72, 99-102. Grenet 1991, 205, 261. 115 Stronach 1978, 24-44. As well as Achaemenid enigmatic tower-like 120 Grenet 1991; Boyce and Grenet 1991, 162-165. This suggestion was structures, Ka’aba-i Zardusht at Naqsh-i Rustam and Zindan-i Suleyman also adopted by Paul Bernard in his recent publications: Bernard 2008, at Pasargadae: Bernard 1990a, 53. On the function of these structures 93; Bernard 2009, 51. with a summary of previous scholarship, see: Potts 2007, 278- 121 Francfort 1984, 124-125; Korovchinskiy 2007, 4-5. 295. 122 116 Abdullaev 1994, 243. Zournatzi 1993. However, see Ghirshman 1964, 222 who thinks that 123 ‘The Iranian terrace derives from Urartu rather than from Mesopotamia’ Henkelman 2008, 311. 124 and Boardman 2000, 53-61 who points out Lydian prototypes and Henkelman 2008, 305. Lydo-Ionian building techniques and features. For the in-depth 125 Henkelman 2008, 58-59, 323. discussion and evaluation of the architectural antecedents for the Tomb 126 Among their main gods are: Zeus, Heracles, Apollo, Athena, of Cyrus, see Stronach 1978, 39-44. Artemis, Nike, Dionysus, Poseidon and Tyche.

128 M. SHENKAR: TEMPLE ARCHITECTURE IN THE IRANIAN WORLD IN THE HELLENISTIC PERIOD

evidence for syncretism and hybridization with Oriental considered an interpretatio semitica of Ahura Mazdā.138 gods and goddesses.127 What gives this prospect even more weight is the substitution of Ahura Mazdā with Bel in a copy of According to some scholars, the presence of a small Behistun inscription found in ,139 which provides chapel in the temenos, various objects depicting females, important evidence that the syncretism between these two and the ‘Cybele plate’ suggest that possibly another deity, deities might go back to the Achaemenian period. perhaps a goddess, was worshipped in the sacred precinct of the ‘Temple with Indented Niches’.128 The available It is very tempting, therefore, to connect the Bel from the evidence does not permit us to safely establish whether Aramaic documents of the Achaemenian period, the this was Anāhitā,129 Nana,130 Cybele, Atargatis131 or some ‘Zeus’ of the ‘Temple with Indented Niches’ and the composite syncretistic female deity who combined the Kushan ŌOROMOZDO into a continuous, uninterrupted attributes and qualities of these Oriental goddesses.132 The tradition of worship of a hybrid Zeus-Belos-Ahura Mazdā situation is very similar to the statuette from the Oxus in Bactria, spanning from the Achaemenian to the Kushan temple who, like the goddess on the ‘Cybele plate’, also periods, but the current evidence might still appear too wears kalathos. fragmentary and insufficient for such an ambitious hypothesis. However it is should be noted that the Recently, new textual evidence of extraordinary recording of libation (zwtr’), as part of the cult of Bel in significance has become available, which may shed light the Aramaic document cited above, may indicate that the on the mysteries of the ‘Temple with Indented Niches’ at vessels found buried in the temple podium were used in a Ai Khanum. This consists of 4th century BCE Aramaic cult of Zeus-Belos. documents from Bactria, which have been translated by Shaul Shaked and Joseph Naveh. One of these documents One can speculate that a local Bactrian population (or mentions a ‘Libation for the temple, to Bel’.133 The perhaps even a population of Mesopotamian descent existence of a temple (bagina) of Bel in Bactria as early living in Bactria before the Hellenistic period), settled in as the Achaemenian period indicates that a Ai Khanum by the Seleucid king, brought with it a Mesopotamian cult and perhaps also Mesopotamian Mesopotamian cult and took an active part in the temple architecture were part of the Bactrian cultural construction of temples according to the Mesopotamian landscape even before the Macedonian conquest, and architectural tradition. could therefore provide an explanation for the architectural layout of the Ai Khanum temples and for the The second temple at Ai Khanum was uncovered some identity of the principal deity of the ‘Temple with 100 meters to the north of the city’s fortifications.140 This Indented Niches’. ‘Extramural Temple’ had a layout very similar to that of the ‘Temple with Indented Niches’, and shows that the The syncretism between Zeus and Bel (under the name of Mesopotamian plan chosen for the main temple was not Zeus-Belos) is a well known phenomenon in accidental (Fig. 12). Like the ‘Temple with Indented Mesopotamia, Elymais and Syria.134 Franz Grenet has Niches’ it stood on a three-stepped podium and the pointed out similarities between the appearance of Zeus exterior walls were decorated with indented niches. on the coins of Seleucia on the minted during the However, instead of a roofed vestibule, the ‘Extramural reign of the Parthian king Phraates II (129-128 BCE) and Temple’ was entered via an open courtyard. Three the deity ŌOROMOZDO (‘Ahura Mazdā’, on two types separate staircases led to three juxtaposed rooms. This abbreviated to ŌROM), on the coins of Huviška.135 Two layout is remarkably similar to the temple of Zeus deities are shown wearing a polos and if the ‘Zeus’ on a Megistos at Dura-Europos.141 coin from Seleucia is indeed Zeus-Belos, as argued by Grenet, and not, for instance, Sarapis, whose cult is also The ‘Extramural Temple’ was found devoid of any finds. attested in the Hellenistic East,136 the identification of It is thus impossible to establish who was worshipped in Zeus-Belos with Ahura Mazdā may be considered a this monumental structure. It has been proposed that the possibility. One Aramaic inscription from Arebsun cella, divided into three chambers, each entered via a (Cappadocia) names Bel as the brother and the husband separate staircase, might indicate a triad of deities, the of ‘Mazdayasnian religion’.137 Bel here is usually possible contenders being Ahura Mazdā, Mithra and Anāhitā.142 127 Boyce and Grenet 1991, 160-165. 128 Korovchinskiy 2007, 9. 129 Downey 1988, 73. In addition, a third religious monument was uncovered on 130 Korovchinskiy 2007, 4-5. the acropolis – a monumental stepped platform placed in 131 Francfort 1984, 124-125. the wide temenos, which probably served as an open-air 132 Korovchinskiy 2002 and 2007, 4, 9-15 thinks that she combined the characteristics of all these deities. 138 Boyce 1982, 275; De Jong 1997, 246, no. 35. .Naveh and Shaked 139 Siedl 1999, 109-110; Calmeyer-Siedl 1999, 229 :(זותרא על בגנא לביל) zwtr’ ‘l bgn’ lbyl 133 (forthcoming) C1:37. 140 Bernard 1974, 287-289; Bernard 1976a, 303-307; Bernard 1976b, 134 See for example, Downey 2004. 272-273; Downey 1988, 73-75; Bernard 1990a, 53-54; Litvinskiy and 135 Grenet 1991, 148. Pichikyan 2000, 289-290. 136 Boyce and Grenet 1991, 25. 141 Bernard 1990a, 54. 137 See Boyce 1982, 274-275; Lemaire 2003. 142 Bernard 1990a, 75; Boyce and Grenet 1991, 172.

129 FROM PELLA TO GANDHĀRA

Fig. 12: The ‘Extramural Temple’, Aï Khanum (after: Bernard 1976, fig. 11) shrine.143 It obviously belonged to the Iranian tradition of dating of the most interesting complex at Shami which worship in high open-air sanctuaries known from earlier has yielded significant finds of is completely periods.144 Two cultic platforms were also excavated in unclear.151 The earliest stage of a temple on the mount of Masjid-i Sulaymān145 and Bard-e Neshandeh146 in Kuh-i Khwaja, for which even Achaemenid foundations Elymais, but their dating to the Hellenistic period remains were claimed,152 was later dated to the late Parthian153 or uncertain.147 Sasanian periods.154

I should also mention a number of monuments once To conclude: Iranian temples of the Hellenistic period considered Hellenistic temples, but recent research and (four in Eastern Iran and one in Western Iran) can be excavations of these sites have demonstrated that they divided into two types: should rather be dated to the later periods, or that they 1. The ‘Iranian type’ (Fig. 14), to which the temples of had no religious function, or both. The ruins of Kangavar Persepolis, Takht-i Sangin and Dilberjin could be and Khurha, located in Western Iran, were frequently 148 attributed, probably developed in Western Iran and cited as examples of Greek temple architecture in Iran. originated in Achaemenian royal architecture. This The first is now thought to be a late Sasanian palace149 150 type is characterized by a rectangular tetrastyle cella and the latter a Parthian estate. The interpretation and surrounded by corridors and flanked by wings entered

143 from the exterior. Not all these elements are always Bernard 1976a, 306-307; Bernard 1990a, 54. The most detailed description of this monument is: Boyce and Grenet 1991, 181-182. For present in each temple of this type; there are the photograph of this platform see: Bernard 2009, fig. 19. variations. The layout of the ‘Iranian type’ was never 144 See: Boyce and Grenet 1991, 182-184; Shenkar 2007, 177. On canonized, but its variants became widespread in the terraces in Iranian architecture see also: Kleiss 1998. 145 Iranian world and beyond (for example in Syria and Ghirshman 1976, 55-149. Ghandhara) in the Parthian, Kushan and even in later 146 Schippmann 1971, 251-258; Ghrishman 1976, 5-55. 147 For the chronological discussion, see: Downey 1988, 131; Bernard periods. 1991, 52 no. 7; Boyce and Grenet 1991, 44; Hannestad and Potts 1990, 115; Boucharlat, 2005, 238. The temples whose remains were excavated 151 See: Schippmann 1971, 227-233; Sherwin-White 1984; Boyce and on these platforms seem to date to the Parthian period: Boucharlat Grenet 1991, 42-43. 1999a, 34. 152 Gullini 1964, 253-273. 148 For example: Ghirshman 1962, 24; Schippmann (1971), 307-308; 153 Schippmann 1971, 57-70; Boyce and Grenet 1991, 150; Mousavi Frye 1984, 162. 1999, 81-85; Ghanimati 2000, 146. 149 Lukonin 1977; Azarnoush 1981, 82-94; Hannestad and Potts 1990, 154 Kawami 1987, 24. There are however still scholars who date it to the 111-112; Azarnoush 1999; Kleiss 2005; Azarnoush 2009. Achaemenid period: Litvinskiy and Pichikyan 2000, 219-226 and the 150 Hakemi 1990; Rahbar 1999. Hellenistic period: Bivar 2003, 2-3.

130 M. SHENKAR: TEMPLE ARCHITECTURE IN THE IRANIAN WORLD IN THE HELLENISTIC PERIOD

Fig. 13. Iranian temples of the Hellenistic period. 1. “Frataraka Temple”, Persepolis; 2. “Oxus Temple”, Takht-i Sangin; 3. “Temple with Niches”, Ai Khanum; 4. “Extramural Temple”, Ai Khanum; 5. “Dioscuri Temple”, Dilberjin

2. The ‘Mesopotamian type’ (Fig. 15) includes the two architecture or cult in their kingdoms. It seems they temples of Ai Khanoum and is characterized by thick simply followed the religious policies of their mud-brick walls frequently decorated with niches, a predecessors, the Achaemenians. As in the earlier roofed or open pronaos and cella divided into three periods, there are more temples in Eastern than in chambers. The origin of this type is most probably in Western Iran, suggesting that throughout antiquity the Mesopotamia where it can be found also in the later peoples of Western Iran (Medians, Persians etc.) probably Parthian period (Temple of Artemis and Temple of tended to practice their cults in high places under the Zeus Megistos at Dura Europos). The layouts of open sky. The reasons for this difference between the cult several Nabatean temples (Qasr Bint Firaun at Petra, practices of the peoples of Eastern and Western Iran still Dibon, Qasr Rabba) also demonstrate close similarity remain to be discovered. to this ‘Mesopotamian type’ and it seems that they were influenced by it. The Greco-Macedonian conquest of Iran had an interesting and controversial impact on temple The layout of the temples shows that the Hellenistic architecture. The influence of Greek architectural rulers did not attempt to introduce uniform ‘state’ temple tradition on Iranian temples was apparently limited to a

131 FROM PELLA TO GANDHĀRA

Fig. 14. The ‘Iranian type’. 1. The “Frataraka Temple”, Persepolis; 2. The “Oxus Temple”, Takht-i Sangin; 3. Mohra Maliaran, Taxila; 4. Khalchayan; 5. The “Dioscuri Temple”, Dilberjin; 6. Sur; 7. Sahr; 8. “Temple of Ba‘al Shamin”, Si‘a; 9. Surkh Kotal; 10. Kuh-i Khwaja; 11. Susa; 12. Panjikent; 13. Dedoplis Mindori number of decorative elements. It is significant that the whose plan is entirely based on Mesopotamian and plans of all five temples discussed here were taken from Iranian architectural models. They (at least the rulers of the repertoire of Oriental architectural types, and that they Ai Khanoum) were certainly acquainted with Greek all were built using local construction techniques which temple architecture, as a heroon of Kineas demonstrates, were characterized by thick walls made of mud-brick.155 but for some reason we do not know, they chose Oriental models.156 One possible explanation is that the architects One can only guess what caused the Hellenistic rulers (if who planned these temples were of non-Greek origin, they actively participated in the projects and personally versed in the Mesopotamian building tradition and chose the plans of the buildings), whose culture, language and customs were predominantly Greek, to erect temples 156 It is worth noting that further to the East, a Jandial temple near Taxila did include a number of elements borrowed from Greek temple 155 Pichikyan and Shelov-Kovedyaev 1989, 53. architecture: Rapin 1995, 287-291.

132 M. SHENKAR: TEMPLE ARCHITECTURE IN THE IRANIAN WORLD IN THE HELLENISTIC PERIOD

Fig. 15. The ‘Mesopotamian type’. 1. The “Temple A”, Ashur; 2. The “Temple with Niches”, Aï Khanum; 3. The “Extramural Temple”, Aï Khanum; 4. The “Temple of Artemis”, Dura Europos; 5. The “Temple of Zeus Megistos”, Dura Europos; 6. “Qasr Bint Firaun”, Petra; 7. Dibon; 8. Qasr Rabba incorporated some elements of Achaemenian royal ‘Greek’ objects, whose exact nature is still unclear, poses architecture. It is puzzling that the Mesopotamian interesting and important questions regarding the Greco- influences seem to have passed over Western Iran and Iranian syncretism and acculturation. In the present state that they appear only in Bactria. Perhaps some cultural of research it is clear that the Bactrian material in and historical factors still unknown to us, and not only particular (as four out of five temples are from Bactria) the state of archaeological research in Western Iran, are reflects the coexistence, and perhaps also a merging, of a responsible for our lack of clarity. number of different cultural traditions – Greek, Iranian and Mesopotamian – and that these temples met the needs Despite the fact that the architecture of the temples is of the heterogeneous, multicultural population of entirely Mesopotamian and Iranian, the majority of finds Hellenistic Bactria.157 in them could be labeled ‘Greek’ or ‘Hellenistic’. The relationship between the ‘Oriental’ architecture and the 157 Lindström 2009, 131.

133 FROM PELLA TO GANDHĀRA

Bibliography BERNARD, P. 1971. La campagne de fouilles de 1970 a Ai Khanoum (Afghanistan). Comptes rendus des ABDULLAEV, K. 1994. Recenziya na: History of séances de l’Académie des Inscriptions et Belles- Zoroastrianism. Vol. III. Zoroastrianism under lettres, 385-455. Macedonian and Roman Rule, by Mary Boyce and BERNARD, P. 1974. Fouilles d’Ai Khanoum Frantz Grenet, with contribution by Roger Beck, (Afghanistan) campagnes de 1972 et 1973. Comptes Leiden, 1991. In Vestnik Drevnej Istorii 3, 239- rendus des séances de l’Académie des Inscriptions et 245. Belles-lettres, 280-309. AMBOS, C. 2003. Nanaja – eine ikonographische Studie BERNARD, P. 1975. Campagne de fouilles 1974 à Ai zur Darstellung einer altorientalischen Göttin in Khanoum (Afghanistan). Comptes rendus des séances hellenistisch-parthischer Zeit. Zeitschrift für Assyrio- de l’Académie des Inscriptions et Belles-lettres, 167- logie, 93/2, 231-272. 197. ASKAROV, A. and ŠIRINOV, T. 1994. The Palace, BERNARD, P. 1976a. Campagne de fouilles 1975 à Ai Temple and Necropolis of Jarkutan. Bulletin of the Khanoum (Afghanistan). Comptes rendus des séances Asia Institute 8, 13-26. de l’Académie des Inscriptions et Belles-lettres, 287- ASKAROV, A. and ŠIRINOV, T. 1991. Le temple du feu 322. de Džarkutan, le plus ancien centre cultuel de la BERNARD, P. 1976b. Les traditions orientales dans Bactriane septentrionale, in P. Bernard, F. Grenet l’architecture gréco-bactrienne. Journal Asiatique (eds), Histoire et cultes de l’Asie centrale préislami- 267, 245-275. que, 129-137. Paris. BERNARD, P. 1978. Campagne de fouilles 1976-1977 à AZARNOUSH, M. 1981. Excavations at Kangavar. Ai Khanoum (Afghanistan). Comptes rendus des Archäologische Mitteilungen aus Iran 14, 69- séances de l’Académie des Inscriptions et Belles- 95. lettres, 421-463. AZARNOUSH, M. 1999. Kangavar: un temple séleucide BERNARD, P. 1989. Āy Kānom. Encyclopedia Iranica d’Anahita devient un monument Sassanide. Dossiers III, 124-126. d’Archéologie 243, 52-54. BERNARD, P. 1990a. L’Architecture religieuse de l’Asie AZARNOUSH, M. 2009. New Evidence on the centrale à l’époque hellénistique. In Akten des XIII Chronology of the ‘Anahita Temple’. Iranica Antiqua internationalen Kongresses fur klassische Archao- 44, 393-402. logie, 51-59. Berlin. AZARPAY, G. 1976. Nanā, the Sumero-Akkadian BERNARD, P. 1990b. Vicissitudes au gré de l’histoire Goddess of Transoxania. Journal of the American d’une statue en bronze d’Héraclès entre Séleucie du Oriental Society 96/4, 536-542. Tigre et la Mésène. Journal des Savants, 3-68. AZARPAY, G. 1988. The Roman Twins in Near Eastern BERNARD, P. 1994a. Le temple du dieu Oxus à Takht-i Art. Iranica Antiqua 23, 349-360. Sangin en Bactriane: temple du feu ou pas?. Studia BERNARD, P. 1966. Première campagne de fouilles Iranica 23, 81-121. d’Ai Khanoum. Comptes rendus des séances de BERNARD, P. 1994b. The Greek States in ”. l’Académie des Inscriptions et Belles-lettres, 127- In J. Harmatta (ed), History of Civilizations of Central 133. Asia, vol. II, 99-131. Paris. BERNARD, P. 1967a. Deuxième campagne de foulles BERNARD, P. 1996. Delbardjin. Encyclopedia Iranica d’Ai Khanoum en Bactriane. Comptes rendus des VII, 234-237. séances de l’Académie des Inscriptions et Belles- BERNARD, P. 2008. The Greek Colony at Ai Khanoum lettres, 306-324. and Hellenism in Central Asia. In F. Hiebert and P. BERNARD, P. 1967b. Ai Khanum on the Oxus: a Cambon (eds.), Afghanistan, Hidden Treasures from Hellenistic City in Central Asia. Proceedings of the National Museum, Kabul, 81-106. Washington. British Academy 53. BERNARD, P. 2009. La découverte et la fouille du site BERNARD, P. 1968. Troisième campagne de fouilles hellénistique d’Aï Khanoum en Afghanistan: d’Ai Khanoum en Bactriane. Comptes rendus des comment elles se sont faites. Parthica 11, 33-57. séances de l’Académie des Inscriptions et Belles- BERNARD, P. and Le BERRE, M. 1973. Architecture. lettres, 263-457. Le quartier administratif: l’ensemble nord. Fouilles BERNARD, P. 1969. Quatrième campagne de fouilles d’Ai Khanoum I, Mémoires de la Délégation d’Ai Khanoum (Bactriane). Comptes rendus des Archéologique Française en Afghanistan 21. Paris. séances de l’Académie des Inscriptions et Belles- BERNARD, P., Le BERRE, M. and STUCKI, R. 1973. lettres, 315-355. Architecture. Le Téménos de Kinéas. Fouilles d’Ai BERNARD, P. 1970. Campagne de fouilles 1969 à Ai Khanoum I: Mémoires de la Délégation Archéologi- Khanoum en Afghanistan. Comptes rendus des que Française en Afghanistan 21. Paris. séances de l’Académie des Inscriptions et Belles- BIVAR, A.D.H. 2003. Cosmopolitan Deities and lettres, 301-349. Hellenistic Traces at Kuh-e Khwaja in . In C.

134 M. SHENKAR: TEMPLE ARCHITECTURE IN THE IRANIAN WORLD IN THE HELLENISTIC PERIOD

Cereti (ed.), Religious Themes and Texts of Pre- DOWNEY, S. 1988. Mesopotamian Religious Islamic Iran and Central Asia, 1-6. Wiesbaden. Architecture: Alexander through the Parthians. BOARDMAN, J. 1994. The Diffusion of Classical Art in Princeton. Antiquity. Princeton. DOWNEY, S.B. 2004. Zeus the Greatest in Syria. BOARDMAN, J. 2000. Persia and the West: An Parthica 6, 117-129. Archaeological Investigation of the Genesis of DRUJININA A. 2001. Die Ausgrabungen in Taxt-i Achaemenid Persian Art. London. Sangīn im Oxos-Tempelbereich (Süd-Tadžikistan). BOROFFKA, N.G.O. 2009. Bandichan zwischen Vorbericht der Kampagnen 1998-1999. Archäolo- Spätbronzezeit und Frühmittelalter, In S. Hansen, A. gische Mitteilungen aus Iran 33, 278-281. Wieczorek, and M. Tellenbach (eds.), Alexander der DRUJININA, A. 2008. Gussform mit griechishcer Große und die Öffnung der Welt Asiens Kulturen im Inschrift aus dem Oxos-Tempel. Archäologische Wandel, 134-144. Mannheim. Mitteilungen aus Iran 40, 121-137. BOUCHARLAT, R. 1997. Susa under the Achaemenian DRUZHININA, A. 2004. Predvaritel’nye rezul’taty Rule. In J. Curtis (ed.), Mesopotamia and Iran in the issledovaniya gorodishcha Takht-i Sangin i Persian Period: Conquest and Imperialism 539-331 opredelenie granits goroda ellinisticheskogo vremeni. BC, 54-68. London. Arkheologicheskie raboty v Tadzhikistane 29, 224- BOUCHARLAT, R. 1999a. Villes, palais et temples. 237. Dossiers d’Archéologie 243, 30-35. DRUZHININA, A. and INAGAKI, Kh. 2008. Obshchie BOUCHARLAT, R. 1999b. Temples du feu sassanides. rezul’taty arkheologicheskikh issledovanij na goro- Dossiers d’Archéologie 243, 68-71. dishche Takht-i Sangin v 2007 g. Arkheologicheskie raboty v Tadzhikistane 33, 101-137. BOUCHARLAT, R. 2005 “Iran”. In R. Boucharlat and P. Briant (eds.), L’Archéologie de l’empire FITZSIMMONS, T. 1996. Chronological Problems at the achéménide: nouvelles recherches, 221-283. Temple of the Dioscuri, Dil’berdzin Tepe. East and Paris. West 46, 3-4, 271-298. BOYCE, M. 1989. Ātaškada. Encyclopedia Iranica III, 9- FRANCFORT H.P. 1984. Le sanctuaire du temple à 11. niches indentées. 2. Les trouvailles. Fouilles d`Ai Khanoum 3, Mémoires de la Délégation Archéolo- BOYCE, M. 1993. Dar-e mehr. Encyclopedia Iranica VI, 669-670. gique Française en Afghanistan 27. Paris. BOYCE, M. and GRENET, F. 1991. A History of Zoro- FRYE, R.N. 1984. The History of Ancient Iran. Munich. astrianism III: Zoroastrianism under Macedonian and FRYE, R.N. 1984. Religion in Fars under the Roman Rule. Leiden. Achaemenians. Acta Iranica 23, 171-179. BURYY, V.P. 1976. Tekhnika zhivopisi. Drevnjaja FUSSMAN, G. 1983. Surkh Kotal. Tempel der Kuschan- Baktrija: materialy Sovetsko-Afganskoj ekspedicii Zeit in Baktrien. Munich. 1969 – 1973 gg, t. 1, 111-124. GAGOŠIDZE, J. 1983. Le temples de l’époque CALLIERI, P. 2007. L’archéologie du Fars à l’époque préchrétienne en Géorgie. In IV Symposium achéménide: quatre leçons au Collège de France, 8, International sur l’Art Géorgien, 1-12. Tbilisi. 15, 22 et 29 mars 2007. Persika 11. Paris. GAGOSIDZE, J. 1992. The Temples at Dedoplis COLLEDGE, M. 1986. The Parthian Period. Leiden. Mindori. East and West 42, 27-48. CALMEYER-SIEDL, U. 1999. Eine Triumphstele GHANIMATI, S. 2000. New Perspectives in the Darius’ I. aus Babylon. In J. Renger (ed.), Babylon: Chronological and Functional Horizons of Kuh-i Focus mesopotamischer Geschichte, Wiege früher Khwaja in Sistan. Iran 38, 137-150. Gelehrsamkeit, Mythos in der Moderne, 297-306. Saarbrücken. GHIRSHMAN, R. 1976. Terrasses sacrées de Bard-é Néchandeh et Masjid-i Soleiman Mémoires de la DANDAMAEV, M.A. and LUKONIN. V.G. 2004. The Délégation Archéologique Française en Afghanistan Culture and Social Institutions of Ancient Iran. 45. Paris. Cambridge. GHIRSHMAN, R. 1962. Iran. Parthians and Sassanians. De JONG, A. 1997 Traditions of the Magi: London. Zoroastrianism in Greek and Literature. GHIRSHMAN, R. 1964. Persia from the Origins to Leiden. . London. D’YAKONOVA, N.V. and SMIRNOVA, O.I. 1967. K voprosu o kul’te Nany (Anakhity) v Sogdiane. GHOSE, M. 2006. Nana, the ‘Original’ Goddess on the Sovetskaya Arheologiya 1, 139-175. Lion. Journal of Inner Asian Art and Archaeology 1, 97-112. DOLGORUKOV, V.S. 1984. Oboronitel’nye sooruzhe- niya Dil’berdzhina In I.T. Kruglikova (ed.), Drevnjaja GNOLI, G. 1980. Zoroaster’s Time and Homeland. Baktrija: materialy Sovetsko-Afganskoj ekspedicii Naples. 1969 – 1973 gg, t. 3, 58-92. GNOLI, G. 2000. Zoroaster in History. New York.

135 FROM PELLA TO GANDHĀRA

GÖBL, R. 1984. System und chronologie der HJERRILD, B. 1990. The Survival and Modification of Münzprägung des Kušanreiches. Vienna. Zoroastrianism in Seleucid Times, in P. Bilde, T. GOPNIK, H. 2010. Why Columned Halls? In J. Curtis Engberg-Pedersen, L. Hannestad and J. Zahle (eds.), and J. Simpson (eds.), The World of Achaemenid Religion and Religious Practice in the Seleucid Persia. History, Art and Society in Iran and in Kingdom, 141-150. Aarhus. Ancient Near East, 195-207. London. HOPKINS, C. 1942. The Parthian Temple. 7/1, GÖRSDORF, V.G. and HUFF, D. 2001. C-Datierungen 1-18. von Materialien aus der Grabung Džarkutan, KAWAMI, T.S. 1987. Kuh-e Khwaja, Iran, and Its . Archäologische Mitteilungen aus Iran 33, Wall Paintings. Metropolitan Museum Journal 22, 13- 75-87. 52. GRENET, F. 1986. Palais ou palais-temple? Remarques KEALL, E. 1971. Archaeology and the Fire Temple. In sur la publication du monument de Toprak-kala. C.J., Adams (ed.), Iranian Civilization and Culture, Studia Iranica 15, 123-135. 15-22. Montreal. GRENET, F. 1987. L’Athéna de Dil’berdzin. In P. KELLENS, J. 2001. Zoroastre dans l’histoire ou dans le Bernard and F. Grenet (eds.), Cultes et monuments mythe? À propos du dernier livre de Gherardo Gnoli. religieux dans l’Asie Centrale préislamique, 41-45. Journal Asiatique 289/2, 171-184. Paris. KLEISS, W. 1998. Terrassenanlagen in der iranischen GRENET, F. 1991. Mithra au temple principal d’Ai Architektur. Archäologische Mitteilungen aus Iran Khanoum? In, P. Bernard, and F. Grenet (eds.), 30, 227-268. Histoire et cultes de l’Asie centrale préislamique, KOCH, H. 1993. Feuertempel oder Verwaltungszentrale? 147-151. Paris. Überlegungen zu den Grabungen in Takht-e Sangin GRENET, F. 2004. The Cult of Oxus: A Reconside- am Oxos. Archäologische Mitteilungen aus Iran 26, ration, Tsentral’naya Aziya ot Akhemenidov do 175-218. Timuridov, 377-379. St-Petersburg. KOROVCHINSKIY, I.N. 2001. Kul’ty Takht-i Sanginskogo khrama (Severnaya Baktriya) v III-II vv. GRENET, F. 2007. Religious Diversity among Sogdian Merchants in Sixth-Century China: Zoroastrianism, do n.e. In Tezisy dokladov Mezhdunarodnoy Buddhism, , and Hinduism. Compara- konferentsii studentov i aspirantov, ‘Lomonosov 2001’. Moscow. tive Studies of South Asia, Africa and the Middle East 27/2, 403-418. KOROVCHINSKIY, I.N. 2002. Disk iz Ay Khanum i kul’ty zhenskikh bozhestv pozdneantichnoy Tsent- GRENET, F. (forthcoming a) Between Written Texts, ral’noy Azii. In Tezisy dokladov Mezhdunarodnoy Oral Performance and Mural Paintings: Illustrated konferencii studentov i aspirantov, ‘Lomonosov Scrolls in Pre-Islamic Central Asia. In J. Rubanovich 2002’. Moscow. (ed.), Orality and Textuality in the Iranian World: Patterns of Interaction Across the Centuries. KOROVCHINSKIY, I.N. 2007. Kul’ty ellinisticheskoy Jerusalem. Baktrii (na materialakh “khrama s ustupchatymi nishami” v Ay Khanum i khrama Oksa v Takht-i GRENET, F. (forthcoming b). Zoroastrianism among the Sangine). Avtoreferat dissertatsii. Moscow. Kushans. In M. Boyce, and A. de Jong, (eds.), A History of Zoroastrianism, vol. IV: Parthian KOŠELENKO, G.S., LAPŠIN, A.G., and NOVIKOV, Zoroastrianism. Leiden. S.V. 1989. Mansur-depe Excavations. Bulletin of the Asia Institute 3, 45-52. GRENET, F. and DRÈGE, J-P. 1987. Un temple de l’Oxus près de Takht-i Sangin, d’après un témoig- KOŠELENKO, G.S., LAŠHIN, A.G. and NOVIKOV, nage chinois du VIIIe s. Studia Iranica 16/1, 117- S.V. 2000. The Mansur-depe Excavations of 1986- 121. 1987. Parthica 2, 87-125. GRENET, F. and MARSHAK, B. 1998. Le mythe de KOSSOLAPOV, A.J and MARSHAK, B.I. 1999. Nana dans l’art de la Sogdiane” Arts Asiatiques 53, 5- Stennaya zhivopis’ Sredney i Tsentral’noj Azii. St- 18. Petersburg. GULLINI, G. 1964. Architettura Iranica dagli KRUGLIKOVA, I.T. 1974. Dil’berdzhin: Raskopki achemenidi ai Sasanidi. Turin. 1970-1972. Moscow. HAKEMI, A. 1990. The Excavation of Khurha East and KRUGLIKOVA, I.T. 1976. Nastennye rospisi West 40, 11-41. Dil’berdzhina. Drevnjaja Baktrija: materialy HANNESTAD, L. and POTTS, D. 1990. Temple Sovetsko-Afganskoj ekspedicii 1969 – 1973 gg, vol. 1 Architecture in the Seleucid Kingdom, in P. Bilde, T. 87-111. Engberg-Pedersen, L. Hannestad and J. Zahle (eds.), KRUGLIKOVA, I.T. 1977. Les fouilles de la Mission Religion and Religious Practice in the Seleucid archéologique soviéto-afghane sur le site gréco- Kingdom, 91-124. Aarhus. kushan de Dilberdjin en Bactriane (Afghanistan). HERZFELD, E. 1976. Iran in the Ancient East. Comptes rendus des séances de l’Académie des Oxford. Inscriptions et Belles-lettres, 407-427. Paris.

136 M. SHENKAR: TEMPLE ARCHITECTURE IN THE IRANIAN WORLD IN THE HELLENISTIC PERIOD

KRUGLIKOVA, I.T. 1979. Nastennye rospisi v Oksa v Severnoy Baktrii. Vestnik Drevnej Istorii 4, pomeshchenii 16 severo-vostochnogo kul’tovogo 84-110. kompleksa Dil’berdzhina. Drevnjaja Baktrija: LIVSHITS, V.A. 2004. Nadpisi i dokumenty. In B.I. materialy Sovetsko-Afganskoj ekspedicii 1969 – 1973 Vainberg (ed.), Kalaly-Gyr II. Kul’tovyy tsentr v gg, vol. 2, 120-145. drevnem Khorezme IV-II vv. do n.e., 188-213. KRUGLIKOVA, I.T. 1984. Dil’berdzhin. Khram Moscow. dioskurov. Moscow. Lo MUZIO, C. 1999. The Dioscuri at Dilberdjin LEMAIRE, A. 2003. Les pierres et inscriptions (Northern Afghanistan): Reviewing Their Chronology araméennes d’Arebsun, nouvel examen. In S. Shaked and Significance. Studia Iranica 28, 41-71. and A. Netzer (eds.), Irano-Judaica 5, 138-165. LUKONIN, V.G. 1977. Khram Anaity v Kangavare. LERNER, J.D. 1999. Impact of Seleucid Decline on the Vestnik Drevnej Istorii 2, 105-111. Eastern Iranian Plateau. Stuttgart. MAIRS, R.R. (forthcoming). The ‘Temple with Indented LINDSTRÖM, G. 2009. Heiligtümer und Kulte im Niches’ at Ai Khanoum: Ethnic and Civic Identity in hellenistischen Baktrien und Babylonien – ein Hellenistic Bactria. In R. Alston and O. van Nijf Vergleich” In S. Hansen, A. Wieczorek, and M. (eds.), Cults, Creeds and Contests in the Post- Tellenbach (eds.), Alexander der Große und die Classical City. Leuven. Öffnung der Welt Asiens Kulturen im Wandel, 127- MARSHAK, B.I. 2001. La thématique sogdienne dans 133. Mannheim. l’art de la Chine de la seconde moitié du VIe siècle. LITVINSKY, B.A. 2000a. A Finial from the Temple of Comptes rendus des séances de l’Académie des Oxus in Bactria. Parthica 2, 131-143. Inscriptions et Belles-lettres, 227-264. Paris. LITVINSKIY, B.A. 2000b. Gelios v khrame Oksa. MARTINEZ-SEVE, L. 2010. À propos du temple aux Vestnik Drevnej Istorii 1, 57-64. niches indentées d’Aï Khanoum: quelques LITVINSKIY, B.A. 2001. Khram Oksa v Baktrii, tom 2, observations, In P. Carlier and Ch. Lerouge-Cohen Baktriyskoe vooruzhenie v drevnevostochnom i (eds.), Paysage et religion en Grèce antique, 195-208. grecheskom kontekste. Moscow. Paris. LITVINSKIJ, B.A. 2003 Hellenistic Clay Portraits from MOUSAVI, M. 1999. Kuh-e Khadjeh: un complexe the Temple of the Oxus. Parthica 5, 37-63. religieux de l’est iranien. Dossiers d’Archéologie 243, 81-85. LITVINSKIJ, B.A. 2004. Ellenisticheskoe iskusstvo v Khrame Oksa. In Transoksiana: istoriya i kul’tura, NAVEH, Y. and SHAKED, S. (forthcoming), Ancient 67-71. Tashkent. Aramaic Documents from Bactria (Fourth Century B.C.E.). London. LITVINSKY, B.A. and PICHIKYAN, I.R. 1981. Découvertes dans un sanctuaire du die Oxus de la NETZER, E. 2003. Nabatäische Architektur. Mainz am Rhein. Bactriane septentrionale. Revue archéologique 2, 195- 216. PANAINO, A. 2002. The ‘bagān’ of the Fratarakas: Gods or ‘divine’ Kings? In C. Cereti, M. Maggi and E. LITVINSKII, B.A. and PICHIKIAN, I.R. 1994a. The Hellenistic Architecture and Art of the Temple of the Provasi (eds.), Religious Themes and Texts of Pre- Oxus. Bulletin of the Asia Institute 8, 47-66. Islamic Iran and Central Asia, 283-306. Wiesbaden. LITVINSKY, B.A. and PICHIKIYAN, I.R. 1994b. A PICHIKYAN, I.R. 1991. Kul’tura Baktrii. Ahemenidskiy Rhyton from Takhti Sangin. Ancient Civilizations i ellinisticheskiy periody. Moscow. from Scythia to Siberia 3, 355-364. PICHIKYAN, I.R. and SHELOV-KOVEDYAEV, F.V. LITVINSKIJ, B.A. and PIČIKIAN, I.R. 1995a. River 1989. Grecheskie bozhestva v ellinisticheskoy Deities of Salute the God of the River Oxus- epigrafike i izobrazitel’nom iskusstve zapadnogo i Vakhsh. In A. Invernizzi (ed.), In the Land of the vostochnogo Irana. Istoriko-Filologicheskiy zhurnal Gryphons, 129-150. Florence. 126, 48-58. LITVINSKY, B.A. and PICHIKYAN, I.R. 1995b. Gold PILIPKO, V.N. 1996. Staraya Nisa. Zdanie s kvadratnym Plaques from the Oxus Temple (Northern Bactria). zalom. Moscow. Ancient Civilizations from Scythia to Siberia 2, 196- P’YANKOV, I.V. 1996. Zoroastr v istorii Sredney Azii: 221. problema mesta i vremeni. Vestnik Drevnej Istorii 3, LITVINSKIY, B.A. and PICHIKYAN, I.R. 2000. 3-23. Ellinisticheskiy khram Oksa v Baktrii vol. I. P’YANKOV, I.V. 1997. Srednyaya Aziya v antichnoy Moscow. geograficheskoy traditsii. Istochnikovedecheskiy LITVINSKIJ, B.A. and PIČIKJAN, I. R. 2002. Taxt-i analiz. Moscow. Sangīn. Der Oxus-Tempel. Grabungsbefund, POTTS, D.T. 2001. Nana in Bactria. Art and Stratigraphie und Architektur. Mainz. Archaeology 7, 23-37. LITVINSKIY, B.A., VINOGRADOV, Yu.G. and POTTS D.T. 2007. Foundation Houses, Fire Altars and PICHIKYAN, I.R. 1985. Votiv Atrosoka iz khrama the Frataraka: Interpreting the Iconography of Some

137 FROM PELLA TO GANDHĀRA

Post-Achaemenian Persian Coins. Iranica Antiqua 42, SCHLUMBERGER, D. 1975. Sur la nature des temples 271-301. de Surkh Kotal. Central’naja Aziya v Kushanskuyu PUGACHENKOVA, G.A. 1966. Khalchayan. Tashkent. epokhu 2. Moscow. PUGACHENKOVA, G.A. 1971. Skul’ptura Khalcha- SCHLUMBERGER, D., Le BERRE, M. and FUSSMAN, yana. Moscow. G. 1983. Surkh Kotal en Bactriane. Vol. I. Les temples: architecture, sculpture, inscriptions. PUGACHENKOVA, G.A. 1973. K arkhitekturnoy Mémoires de la Délégation Archéologique Française tipologii v zodchestve Baktrii i Parfii. Vestnik Drevnej en Afghanistan 25. Paris. Istorii 1, 121-131. SCHMIDT, E.F. 1953. Persepolis I: Structures, Reliefs, PUGACHENKOVA, G.A. 1974. O kul’takh Baktrii v Inscriptions. Chicago. svete arkheologii. Vestnik Drevnej Istorii 3, 124-135. SHAHBAZI, A.S. 2002. Recent Speculations on the PUGACHENKOVA, G.A. 1982. K tipologii monumen- ‘Traditional Date of Zoroaster’. Studia Iranica 31/1, tal’nogo zodchestva drevnikh stran sredneaziatskogo 7-45. regiona. Iranica Antiqua 17, 21-42. SHAKED, S. 2005. Zoroastrian Origins: Indian and RAHBAR, M. 1999. Khorheh, une résidence parthe Iranian Connections. In J.P. Arnason, S.N. Eisenstadt sur le plateau iranien. Dossiers d’Archéologie 243, and B. Wittrock (eds.), Axial Civilizations and World 44-47. History, 183-201. Leiden. RAPIN, C. 1994. Svyatilishcha Sredney Azii v epokhu SHENKAR, M. 2007. Temple Architecture in the Iranian ellinizma. Vestnik Drevnej Istorii 4, 128-140. World before the Macedonian Conquest. Iran and the RAPIN, C. 1995. Indo- and Vishnuism: on an Caucasus 11/2, 169-194. Indian Object from the Sanctuary of the Oxus and ŠKODA, V.G. 1987. Le culte du feu dans les sanctuairies Two Temples in Taxila in A. Invernizzi (ed.), In the de Pendžikent. In P. Bernard and F. Grenet (eds.), Land of the Gryphons, 275-291. Florence. Cultes et monuments religieux dans l’Asie Centrale RAPIN, C. 2004. L’Afghanistan et l’Asie centrale dans la préislamique, 63-72. Paris. géographie mythique des historiens d’Alexandre et SHKODA, V.G. 1996. The Sogdian Temple: Structure dans la toponymie des géographes gréco-romains. and Rituals. Bulletin of the Asia Institute 10, 195- Notes sur la route d’Herat à Begram, In O. 206. Bopearachchi and M.-Fr. Boussac (eds.), Afghanistan. Ancien carrefour entre l’Est et l’Ouest, 143-172. SHKODA V.G. 1998. Iranian Traditions in Sogdian Bruxelles. Temple Architecture. In V.S., Curtis, R. Hillenbrand and J.M. Rogers (eds.), The Art and Archaeology of RAPIN, C. 2007. Nomads and the Shaping of Central Ancient Persia. New Light on the Parthian and Asia, In J. Cribb and G. Herrmann (eds.), After Sasanian Empires, 122-132. London. Alexander: Central Asia before Islam. Proceedings of the British Academy 133, 29-72. Oxford. SHKODA, V.G. 2009. Pyandzhikenstkie khramy i problemy religii Sogda (V-VIII vv). St. Petersburg. RAPIN, C., ISAMIDDINOV, M. and KHASANOV, M. 2010. Koktepe, ville jumelle de Maracanda. Dossiers SIEDL, U. 1999. Ein Monument Darius I. aus Babylon. d’Archéologie 341, 20-24. Zeitschrift für Assyriologie und vorderasiatische Archäologie 89, 101-114. RAPOPORT, Yu.A. 1994. The Palaces of Topraq-Qal‘a. Bulletin of the Asia Institute 8, 161-186. SIMS-WILLIAMS, N. 2001. Bactrian Legal Documents from 7th- and 8th-Century Guzgan. Bulletin of the Asia RIBOUD, P. 2003. Le cheval sans cavalier dans l’art Institute 15, 9-31. funéraire sogdien en Chine: À la recherche des sources d’un thème composite. Ars Asiatique 58, 148- SIMS-WILLIAMS, N. 2003. A Bactrian Quarrel. Bulletin 161. of the Asia Institute 17, 9-17. ROSENFIELD, J.M. 1967. The Dynastic Art of the STAUSBERG, M. 2002. Die Religion Zarathushtras, Kushans. Los Angeles. Geschichte – Gegenwart – Rituale, vol I. Stuttgart. SARIANIDI V.I. 1996. Proiskhozhdenie iranskikh STRONACH, D. 1978. Pasargadae: A Report of the khramov ognya. Convegno internazionale sul tema: Excavations Conducted by the British Institute of La Persia e l’Asia Centrale da Alessandro al X Persian Studies from 1961 to 1963. Oxford. secolo, in collaborazione con l’Istituto Italiano per il STRONACH, D. 1985. On the Evolution of the Early Medio ed Estremo Oriente (Roma, 9-12 novembre Iranian Fire Temple. Papers in Honor of Prof. Mary 1994), 319-329. Rome. Boyce. Acta Iranica 25, 605-627. SCHIPPMANN, K. 1971. Die iranischen TANABE, K. 1995. Nana on Lion. East and West in Feuerheiligtümer. Berlin – New York. Sogdian Art. Orient 30/31, 365-381. SCHLUMBERGER, D. 1961. The Excavations at Surkh THOLBECQ, L. 1997. Les sanctuairies des Nabatéens. Kotal and the Problem of Hellenism in Bactria and État de la question à la lumière de recherches India. Proceedings of the British Academy 47. archéologiques récentes. Topoi Orient-Occident 7/2, London. 1069-1095.

138 M. SHENKAR: TEMPLE ARCHITECTURE IN THE IRANIAN WORLD IN THE HELLENISTIC PERIOD

UZYANOV, S.A. 1987. O serebryanoy statuetke, WIESEHÖFER, J. 1994. Die “dunklen Jahrhunderte” naydennoy na Takhti-Sangine. In B.B. Piotrovskij, der Persis. Munich. and G.M. Bongard-Levin (eds.), Tsentral’naya Aziya, WIESEHÖFER, J. 2001. Frataraka. Encyclopedia Iranica novye pamyatniki pis’mennosti i iskusstva, 290-297. X, 195. Moscow. WIESEHÖFER, J. 2007 Fars under Seleucid and Parthian VEUVE, S. 1987. Le gymnase. Architecture, céramique, Rule. In V.S. Curtis, and S. Stewart (eds.), The Age of sculpture. Fouilles d’Ai Khanoum 6. Mémoires de la Parthians, The Idea of Iran, vol. II, 37-50. London. Délégation Archéologique Française en Afghanistan 30. Paris. ZEYMAL’, E.V. 1985. Drevnosti Tadzhikistana. Katalog vystavki. Dushanbe. WALDMANN, H. 1991. Der kommagenische Mazdais- mus. Tübingen. ZOURNATZI, A. 1993. Cyrus. vi. The Tomb of Cyrus. Encyclopedia Iranica VI, 522-524.

139