The votive plaques in the Oxus and Mir Zakah treasures: what they tell about the cults
Abstract
54
Full text
The votive plaques in the Oxus and Mir Zakah treasures: what they tell about the cults Frantz Grenet ABSTRACT This paper proposes acomparison between the votive plaques which were part of the Oxus and Mir Zakah treasures, both dating for the main to the Achaemenid period. In afirst section the questions of provenance and authenticity are discussed. Arguments adduced against the authenticity of some plaques are dismissed in the light of comparative material, some of which was recently discovered in safe contexts. Asecond section addresses the religious significance of these offerings. The Oxus treasure collection is consistent with known chracteristics of the god Wakhsh (Oxus), e.g. its association with the horse and its possible assimilation to Tishtrya. The Mir Zakah collection is more associated with farming and it also comprises asignificant proportion of medical ex -votos (or propitiatory offerings), completely absent from the Oxus treasure. Such arepertoire could suggest that this part of the Mir Zakah treasure originally belonged to atemple of Anāhitā. KEYWORDS Achaemenid; Anāhitā; Bactria; forgeries; medicine; Mir Zakah; Takht -iSangin; votive plaques; Zoroastrianism. We have only few means to access the religious beliefs and practices of non -elite people in ancient Central Asia. Funerary archaeology gives us quite aprecise idea of ritual practices and their frequent conformity to the prescriptions of the Vidēvdāt, the ‘exorcism’ book of the Avesta, but the eschatological beliefs are documented only by adozen or so figured Sogdian ossuaries, extremely interesting but which probably were not intended for the common folk. 1 The very abundant series of terracotta figurines offers better perspectives. They often depict donors bringing flowers or offerings, or seated goddesses. The production of anthropomorphic terracottas had almost ceased in Central Asia between the late Bronze Age and the Hellenistic period. Even then, at Ai Khanoum, the main site, this material is neither particularly abundant nor imaginative. The last studies, in particular by Jangar Ilyasov, Tigran Mkrtychev, Kazim Abdullaev, and Nigora Dvurechenskaya, 2 tend to indicate that adiversified production started only in the post -Hellenistic period, initially with some influence from Parthian art. In Sogdiana this material remained quite stereotyped until the 6th century, after which it gave way to some individualized depictions of deities.3 Fortunately, for the Achaemenid and, to some extent, the Hellenistic period, we have aremarkable and focused documentation with multiple ex ‑votos or rather propitiatory offerings, drawn or incised on gold leaves. They come from two sources: the Oxus Treasure found in 1 The up -to -date collection of the evidence is in Shenkar 2014, see figs. 34, 52–53, 112, 133, 169–181, and pl. 6. 2 Ilyasov– Mkrtychev 1991/1992; Abdullaev 2003; Abdullaev 2004; Dvurechenskaya 2016. 3 Meshkeris 1989; Kidd 2007. See also the very informative entries in Zeymaľ ed. 1985, especially no. 486–509 by Boris Marshak. STUDIA HERCYNIA XXVII/1, 54–64
55FRANTZ GRENET 1877 near the Afghan border in present Tajikistan, now in the British Museum,4 and the Mir Zakah Treasure found in 1992 in the mountains to the southeast of Kabul, now in the Miho Museum in Japan.5 The Oxus series comprises 51 plaques, the Mir Zakah series more than three hundred. Similar plaques were found in the temple of Takht -e Soleiman (the Ādur Gushnasp, one of the three major Fire temples of the Sasanian empire), but they are only afew.6 In both the Oxus and Mir Zakah treasures the golden plaques were only part of the votive offerings, and by far not the richest one, but they lend themselves better to astudy of religious attitudes because of their very special character: uniform by the function, but individual by the choice of images expressing specific intentions. In both cases these finds were associated with coin hoards which do not match chronologically with them and had most probably been accumulated separately. Both series of plaques are entirely accessible in published catalogues, and both have been submitted to detailed and sometimes polemical studies. In January 2006 Ihad direct access to the Miho collection thanks to the help of the curator Professor Inagaki.7 These two series cannot be treated as reliable source material without further ado. Several questions have to be addressed, and in fact have been addressed. QUESTIONS OF PROVENANCE AND AUTHENTICITY The first question is the provenance. Contemporary information consistently associates the Oxus Treasure with afortified site named Takht -iQobad/Takht -iKuwad, just on the northern bank of the Amu Darya. In 1976, one century after its discovery, excavations at Takht -iSangin, five kilometres upstream, began to bring to light amonumental temple from the Hellenistic and Kushan periods, with dedicatory inscriptions all mentioning the god Oxus, in Bactrian Wakhshu; the place is in fact mentioned by Ptolemy under the name Oxeiana.8 The excavator Igor Pichikyan then proposed to recognize Takht -iSangin as the real place of discovery of the Oxus Treasure.9 Amajor objection is the time discrepancy between the composition of this treasure, entirely Achaemenid except perhaps for one or two objects and the coins (probably hoarded separately), and the foundation of the Takht -iSangin temple in the early Seleucid period. Ishall not dwell on this point as Iconsider it was convincingly solved by Evgenii Zeymaľ,10 then by Paul Bernard:11 they demonstrated that the Oxus Treasure came from adestroyed or still undiscovered Achaemenid predecessor of the Takht -iSangin temple, and was concealed at Takht -iQobad at the time of Alexander’sinvasion or in asubsequent period of troubles. 4 Dalton 1964; Zeymaľ 1979; Curtis– Searight 2003. 5 Miho Museum 2002. 6 Huff 2011, 86–88 with fig. 4a–c. 7 Photographs from the Miho Museum exhibition catalogue are reproduced with permission from Pr. Inagaki. From here onwards Isubstantially revise and update the second half of an article published in Italian some years ago (Grenet 2010). 8 The most elaborate excavations report is Livinskij– Pičikjan 2002. The identification with Oxeiana was first proposed in Grenet– Rapin 1998, 85. 9 He expressed this idea in many publications, see especially Pitschikjan 1992. 10 Zeymaľ 1979, 10–15. 11 Bernard 1994.
56 STUDIA HERCYNIA XXVII/1 Like the Oxus Treasure, the Mir Zakah treasure does not come from controlled excavations, but, contrary to what is sometimes stated, precise information was recovered on the spot in the following years by members of the Afghan Institute of Archaeology, at great risk to themselves (Fig. 1). 12 The golden plaques were part of ahuge hoard which also comprised four tons of coins and many prestigious objects in precious metal. It appears that it was all put in bags and thrown simultaneously into basins dug around aspring and lined with wooden logs, which had been already evidenced in 1948 during alimited rescue excavation (Fig. 2).13 The coin series stops with the Kushan king Vasudeva Iwhose reign ended in about 230 AD, after which the Kushans lost Bactria to the Sasanians and, after some time, retreated to the east of the Khyber Pass.14 Aplausible scenario is that these treasures were part of the holding of the Kushan state and were evacuated at the time of the Sasanian invasion or other concomitant 12 Bopearachchi– Flandrin 2005. See below for more detail. 13 Curiel– Schlumberger 1953. 14 Bopearachchi– Flandrin 2005, 158, 245. Fig. 1: Osmund Bopearachchi and members of the Afghan Institute of Archaeology at Mir Zakah, 2 March 2005, checking the Miho catalogue with local people’srecollections. ©Osmund Bopea‑ rachchi. Fig. 2: The only scientific excavations at Mir Zakah, 1948: basins lined with wooden logs; right, wooden drain. Curiel– Schlumberger 1953, pl. VII: 1–2.
57FRANTZ GRENET troubles. The closure of the two rooms containing the Begram treasure (more probably part of apalace treasury than amerchant’scache) might be related to this stampede. The fact that the largest hoard of Kushan gold coins ever discovered, at Debre Damo in Ethiopia, near the mouth of the Red Sea, also ends with issues of Vasudeva is perhaps not just acoincidence but could also be in relation to the disruption of the Kushan Empire, rather than with trade.15 As for the cargo to Peshawar, for some unknown reason it was diverted to aside valley, where the load was hastily hidden by people who certainly intended to recover it but were never able to come back. These State holdings were obviously the product of along and heterogeneous accumulation, including afar older stock of temple offerings; in fact, very few plaques from Mir Zakah indicate the Hellenistic period, especially two with Athena (Pl. 2/1), and none belongs clearly to the Kushan period; the Buddhist repertoire is totally absent.16 Another issue is authenticity. In 2003 Oscar White Muscarella took an extreme position.17 According to him both series of plaques were, at least largely, the product of modern forgers working one century apart from each other: ‘not one single gold plaque can be readily accepted as an ancient artefact.’ As far as the Oxus Treasure is concerned, this contention was refuted with sound arguments by John Curtis. 18 The Mir Zakah Treasure presents amore complex situation, for forgers working in the late twentieth century could benefit from the well spread knowledge of Achaemenid art. In an article published in the Archäologische Mitteilungen aus Iran und Turan in 2011, Dietrich Huff took amiddle position, proposing to eliminate several objects including the most beautiful ones, one from the Oxus Treasure and five from Mir Zakah, all showing worshippers with mouth -covers (padām) and the bundle of ritual twigs (barsom) (Pl. 2/2).19 Imust admit that Iam not convinced by his arguments, very carefully presented as they are: 1) The Oxus Treasure plaque is supposed to be mistaken in its depiction of the suspension of the akinakes (Scythian and Persian dagger), the lower belt being too thin. But if we compare the recently discovered image of Srōsh at Akchakhan -kala in Chorasmia, this belt looks hardly thicker (Pl. 2/3).20 2) The bashlyk as reproduced on the Mir Zakah plaques is supposedly copied from published coins of the Frataraka rulers of Persis and misunderstood. In fact, it is almost identical on agenuine coin from the Oxus Treasure, issued in Bactria as shown by the legend Wakhshu on the reverse (Pl. 2/4).21 3) Huff’smain objection to the genuineness of these images is the double string attaching the padām, which he considers inspired by modern medical masks. Adouble string, however, appears on the figure of apriest in a6th -century Sogdian funerary relief from China (Fig. 3).22 15 Mordini 1960. Trade is not the most likely explanation for Roman gold flew towards the Kushan Empire and not vice versa. Adiplomatic gift carried to the king of Axum and diverted on the way is also apossibility, see Berzina 1984. 16 Same conclusions in Bopearachchi– Flandrin 2005, 166. 17 Muscarella 2003. 18 Curtis 2004. 19 Huff 2011, 88–100. 20 Bettsetal. 2016, 133–134, fig. 18, pl. 7. 21 Zeymaľ 1979, no. 0454; Alram 1986, no. 383. All that can be read of the obverse legend is w.gw (The second and fifth letters can be either d, k, or r), certainly not whšwwr (Wakhshuwar, Greek Oxyartes), contrary to no 382 which bears no portrait of alocal ruler. 22 Couch in the Miho Museum: Wertmann 2015, 305, fig. 86, left.
58 STUDIA HERCYNIA XXVII/1 For my part Ido not see any decisive reason to eliminate these objects. Concerning specifically Mir Zakah, it is necessary to return to some facts linked with the discovery. As mentioned above, in 1993, when the pillage was still going on, members of the Afghan Institute of Archaeology managed to inspect the place; among them was Nader Rasooli, atrained archaeologist who later on became director of the Institute. Ten years later, after the Miho Museum exhibition catalogue had appeared, he recognized in it about twenty objects or series of objects he had seen, including most of those illustrated in the present article. 23 Even taking into account some possible confusion, Muscarella’sopinion cannot be reasonably maintained in view of this testimony. Certainly, Icannot exclude that some plaques of childish execution are modern forgeries,24 though Iam not even sure of that (some plaques in the Persepolis Museum do not look better). Bad taste, inaccuracy in the reproduction of realia, awkward copying of good specimens by less talented hands, all phenomena considered as betraying modern forgeries, could just as well occur in temple workshops where craftsmen with various training worked side by side. Considering the huge mass of objects found at Mir Zakah and simultaneously poured on the antique market, one wonders why the dealers would have wasted money to commission forgeries whose contribution to the gross price was utterly negligible.25 23 Pl. 2/2: Miho Museum 2002, no. 65a–d (priests?, tentatively dismissed by D. Huff), Pl. 2/5: Miho Mu‑ seum 2002, no. 70c–d (plaques with twisted tongues), Pl. 2/6: Miho Museum 2002, no. 86 (ploughing) and Miho Museum 2002, no. 87a–l (domestic animals). 24 See especially Miho Museum 2002, n o 71–72 (indeed not included among the objects recognized by N. Rasooli). I owe to Henri-Paul Francfort the important remark that several objects from Ai Khanoum would have been rejected as fakes if they had not been found in controlled excavations. 25 The find was never dispersed in various shops: as awhole (including the coins and the heavy precious objects) it was transported to Peshawar by Mangal tribesmen under the control of the brigand Jan Baz; there the London art dealer Nigel Markham arranged air transportation to London, and Fig. 3: Sogdian funerary couch in Miho Museum. Lamentation scene, detail of priest. Wertmann 2015, fig. 86 left.
59FRANTZ GRENET THE REPERTOIRES AND THEIR POSSIBLE RELIGIOUS INTERPRETATION Having tried in my own way to clear the field, Ican now come to real issues of religious history. Judging from the diversity of their repertoire, especially at Mir Zakah, and in both places major differences in quality and consequently in price, it is clear that these plaques were purchased individually by pilgrims, either as ready -made or executed on the spot. In the shape they came down to us, quite often folded or crumpled, they had probably been recycled as raw material, except for the best specimens. How were they used before? Some had been nailed to asupport; others had long tongues cut in the same metal (Pl. 2/5). In his contribution to the Miho Museum exhibition catalogue, Anthony Green has supposed that such tongues were wrapped around aparcel which constituted the real offering.26 Paul Bernard proposed an alternative idea: the tongues could have been used to suspend the plaques to tree branches, like the ribbons and pieces of cloth which in amodest way continue such offerings in modern Zoroastrian and Islamic sacred places.27 Itook the photo reproduced here (Fig. 4) in 2002 in ashrine built around aspring at Shaartuz, not far from Takht -iSangin, and which somehow perpetuates the devotion to Waters which had been the main function of the ancient temple. Perhaps in this context one should compare the mention by Herodotus (VII, 31) of Xerxes suspending ‘golden ornaments’ to aplane tree in Lydia. subsequently to the free trade zone in Basel where it was ‘legalized’ and acquired by Noriyoshi Horiuchi, another art dealer, on behalf of the Miho Museum. All this went on quickly: there is no doubt that the objects (except for the coins) were in the Miho Museum in 1997 at the latest, when Igor Pichikyan published the first study of the hoard which (misinformed about the real provenance) he assumed to come from the same source as the Oxus Treasure, an opinion which has never gained much approval (Pichikyan 1997). Pichikyan’sstudy did not include the votive plaques but he mentions there were several hundreds. For all this see Bopearachchi– Flandrin 2005, 126–149, ascrupulous inquiry based on multiple interviews. This history excludes any substantial addition of modern forgeries. 26 Green 2002, 225. Approved by Huff 2011, 99–100. Indeed, these tongues are another indication that the pieces are genuine: they were not copied from the plaques of the Oxus Treasure where, on rare specimens, they are much shorter; why would forgers have invented them, wasting gold and work? 27 Personal communication. Fig. 4: Ribbons attached to trees in asanctuary near Takht ‑iSangin. Photo by the author.
60 STUDIA HERCYNIA XXVII/1 Both series of plaques contain amajority of images of worshippers, including afew women. Those who carry barsoms are not necessarily priests, for there is textual evidence that still in the Sasanian period barsoms were also used by the laity when uttering prayers or thanksgiving.28 Those wearing apadām, however, are likely to be priests, for the padām is associated with the recitation of the Avestan prayers in front of the sacred fire, aduty of priests, and there is no record of their use by others except by servants in the presence of their master.29 One worshipper with padām in the Oxus Treasure carries an akinakes (Pl. 2/3), which does not necessarily preclude his function as apriest.30 If we move now onto other categories of the repertoire, some significant differences can be noticed between both sites. In the Oxus Treasure the only subject depicted except for worshippers are individual figures of animals: one camel and three unsaddled horses (Fig. 5), to which one can add two figures of horses cut around. It is plausible that they were substitutes for sacrificed animals. At Mir Zakah the majority of animals depicted are bulls or zebus. There are also camels, horses, donkeys, sheep, and goats. Some compositions are evocative of farming, like ascene of ploughing and transportation of the harvest in acart, which suggests awish for agricultural abundance (Pl. 2/6). Another difference is the mass presence at Mir Zakah of miniature images of weapons, mainly akinakes and bows (Pl. 2/7), and their total absence from the Oxus Treasure. There is also an image of alancer charging (Pl. 2/8). The Takht -iSangin votive deposits, however, include ahuge quantity of real weapons, mainly arrows.31 28 Boyce– Kotwal 1971, 298–302. 29 Information kindly supplied by Almut Hintze. 30 Still recently in India weapons were kept in some fire temples and carried in processions (information Almut Hintze). In Sogdian painting some characters dressed in white and supposed to be priests carry adagger or asword at their belt: Grenet– Azarnouche 2007, 168–169 with figs. 12–13. 31 Litvinsky 2001. Fig. 5: Oxus Treasure, votive plaque with horse, Zeymaľ 1979, no. 99.
61FRANTZ GRENET The main contrast concerns images of body parts, totally absent in the Oxus Treasure but relatively frequent at Mir Zakah where they represent 8% of the total. They include eyes, arms, legs, hands, and feet (Pl. 2/9). Clearly these offerings reflect requests for cures. APersian Zoroastrian text, the Saddar Bundahishn, actually mentions the offering of a‘golden eye’ to the Ādur Gushnasp temple in order to recover eyesight.32 Scratched figures of babies might express requests for fertility (Pl. 2/10), but contrary to what is found in Greco -Roman curative sanctuaries no object alludes to sexual functions, in contrast to the repertoire of Sasanian magical seals which include some images of intercourse and uteri.33 The only figures showing genitalia are clearly of Indian inspiration (Pl. 2/11). Does the absence reflect the extremely reserved attitude of ancient Iranians towards nudity and bodily functions, which surprised Greek writers so much?34 While remaining cautious in the religious interpretations, one can venture some explanation for the contrasts between the repertoires at both sites. Several peculiarities of the Oxus Treasure are consistent with what we know of the cult to the river -god Oxus. Horses were in general sacrificed to rivers, and in the case of the god Oxus there are, moreover, indications of an assimilation to Tishtrya, originally the rain god, who comes from the waters every New Year in the shape of awhite horse fighting the black horse which embodies the demon of drought (Avesta, Yasht VIII, 20–23). AChinese account from the time of the Arab conquest contains an echo of this myth, associated with the description of atemple which is most probably Takht -iSangin or what subsisted of it at that time.35 An assimilation of Oxus and Tishtrya is also possibly reflected by apersonal name in the Aramaic archive of Balkh: Wakhsu ‑abra ‑dāta‑, ‘given by the cloud of Oxus’.36 More accretions are suspected in the Hellenistic period, such as the Lydian river -god Marsyas (depicted by avotive statuette) and his rival Apollo, which could explain the presence in the material from that time of many arrows and musical instruments, especially flutes of pure Greek manufacture, 37 but they are not relevant for the Achaemenid period documented by the Oxus Treasure. One function of river deities, however, was probably not assumed by the god Oxus: Ihave in mind the healing function. The river Weh Dāytī (today the Daryā -iPanj) which joins the Oxus (today the Wakhsh) just opposite Takht -iSangin is considered in the Bundahishn (XIa, 8) as infested with noxious creatures (xrafstar).38 Perhaps this mention concerns mainly silurids, large sludge fish unfit for consumption and which can attack domestic animals drinking in the river. Fish were apparently not caught or at least not consumed at Takht -iSangin, according to the collection of bone remains found in the excavations.39 Adeep staircase dug in the temple courtyard, cleaned by the archaeologists to adepth of twelve meters without reaching the bottom, had no other purpose than giving access to the water table (Fig. 17);40 it shows that the river water, though flowing at avery short distance, was not considered suitable for the cult and possibly not even for drinking, no more than it is today. Iwould suggest that the total absence of medicine in the functions reflected by the offerings in the Oxus Treasure could reflect this situation. 32 Transl. Dhabhar 1932, 536. 33 Gyselen 1995, 50–51 with figs. 50–56. 34 Dio Chrysostomus, Oration 13.24: being naked in public is one the worst thing to do for Persians; Procopius, De bello Persico I, 7.18: it is not allowed for women to be seen naked by men. 35 Drège– Grenet 1987; Bernard 2015, 59–61, 68. 36 Naveh– Shaked 2004, 59, 211; Tavernier 2017, 118, prefers to read Wakhsu ‑abda ‑dāta‑ ‘given by the wonderful Oxus’, an alternative already considered by Naveh– Shaked 2012. 37 Bernard 1987; Bernard 2015, 55–58. 38 Cereti 2004, 32–33. 39 Angelina Drujinina, personal communication. 40 Drujinina– Lindström 2013; Bernard 2015.
62 STUDIA HERCYNIA XXVII/1 As for the Mir Zakah material, the combined allusions to war, agricultural abundance, and medical cure could hint at atemple of Anāhitā, for these are the three main benefits her worshippers expected from her. The wish for good health is expressed by the worshipper Tusa in the Ābān Yasht (Yasht V, 53), and is consistent with her function as protectress of running water, as all pure waters are considered as healing. The bull which according to Plutarch (Lucullus, 24.6–7) was sacrificed to Anāhitā, and still was in the Zoroastrian shrine of Pir -e Bānu Pārs until the 19 th century, 41 is well represented among animal images at Mir Zakah. Alady holding avase and wearing an indented crown (Pl. 2/12) could be Anāhitā, or perhaps rather aqueen emulating Anāhitā, for in the Ābān Yasht (128) the goddess is described with acrown with ‘eight towers’ (ašta.kaožda).42 The history of the formation of the hoard as we can try to reconstruct it does not, unfortunately, give any clues to the place where this Achaemenid and Hellenistic temple could have stood. Balkh, one of the seven cities mentioned by Berossos (apud Clement of Alexandria, Protrepticus V, 65.3) as having received atemple of Anāhitā under Artaxerxes II, is just apossibility. It does not seem proper to push the analysis further, and possibly Ihave tried to make these modest objects tell more than they could. Ihope, however, to have shed some light on the expectations of these anonymous pilgrims, otherwise utterly lost to history. 41 Boyce 1977, 248–255. 42 Panaino 2000, 38 note 16. This proposal has not gained wide acceptance, most translators preferring to understand ‘eight parts’ or ‘eight rays’. It is, however, etymologically attractive (kaoždaka- > Middle Persian kōšk = pavilion, palace, kiosk). The towered crown could be aheritage of Ishtar. On the famous Achaemenian seal from Gorgippia showing adignitary praying in front of Anāhitā depicted as Ishtar on her lion, the goddess has acrown with four indentations, presumably eight ‘towers’ in profile view (Shenkar 2014, fig. 18). On the Mir Zakah plaque five indentations are visible. Fig. 6: Takht ‑iSangin, staircase. Drujinina– Lindström 2013, fig. 3.