(PDF) Banquets Scenes from Hatra

ARAM, 17 (2005) 61-82 L. DIRVEN 61 BANQUET SCENES FROM HATRA Dr. LUCINDA DIRVEN (University of Amsterdam) INTRODUCTION This article discusses several representations of human and divine banquet- ing figures from Hatra, a pre-Islamic city located in the North-Mesopotamian steppe.1 Images of male figures who recline on their left side and hold a drink- ing cup in their hand are very common in the Syrian-Mesopotamian region during the first three centuries of the Common Era.2 Best known are numerous representations of this kind from Palmyra. Compared to Palmyra, reclining fig- ures are rare in Hatra. However, the Hatrene specimens are extraordinary in spite of their small number and common appearance. Contrary to the majority of contemporary images, the Hatrene monuments are found in a cultic instead of a funerary context. 3 The present contribution seeks to demonstrate that they relate to one of the activities that took place in Hatrene sanctuaries, the cel- 1  The idea for this paper was born during a conversation with Professor Michal Gawlikowski, that took place at a conference in Wiener Neustadt in the spring of 2003. Although Professor Gawlikowski ought to be cited as source of inspiration, responsibility for the opinions and mis- takes are the writer’s alone. I also owe thanks to Dr. Ted Kaizer for his useful comments on a draft of this article. 2  The motif has a long history in the Near East. It is first attested on a relief of King Assur- banipal from Nineveh dated to the seventh century BCE. The origin of the motif in the iconogra- phy of the Orient is still debated. Cf. J.M. Denzer, Le motif du banquet couché dans le proche- orient et le monde grec du VIIe au IVe siècle avant J.C. (Paris 1982). For a list of banqueting figures dated to the Parthian and early Sasanian period, see J.M. Dentzer, “L’iconographie iranienne de souverain couché et le motif du banquet”, AAAS 21 (1971), 39-50. Last but not least, there are many instances of reclining terracotta figurines from Hellenisic and Parthian Mesopotamia. The clothed, semi-draped or naked reclining female was particularly popular in Babylon, Seleucia and Uruk. Like the men, the women recline on their left side and lean on a cushion. Frequently, they have a small cup in their left hand. A few male reclining figures are also known from Mesopotamia, but they are typologically younger than the female figurines: K. Karvonen-Kannas, The Seleucid and Parthian Terracotta Figurines from Babylon. In the Iraq Museum, the British Museum and the Louvre (Florence 1995), 59-64. For the motif in the Roman world, see K. M. Dunbabin, The Roman Banquet. Images of Conviviality (Cambrige 2003), esp. 103-140. 3  In addition to a funerary context, representations of banqueting figures have been found in a domestic context in the Roman Near East. Well known are the paintings found in one of the houses in Dura-Europos: L. Dirven, The Palmyrenes of Dura-Europos. A study of Religious In- teraction in Roman Syria (Leiden 1999), 291-293, fig. 10 and pl. XI, with references to previous publications. In addition to being found in graves, reclining terracotta figurines were found in temples and residential areas (in Seleucia, most were found in the rooms of Block B). 62 BANQUET SCENES FROM HATRA ebration of ritual meals.4 Subsequently, it will be argued that some of these meals were celebrated in memory of the ancestors of the participants. It will be shown that funerary meals celebrated in temples or private homes are a well known feature in West Semitic, Israelite and Mesopotamian family religion. An in-depth analysis of the monuments shows that these millenia-old tradi- tions still prevailed in Parthian Hatra. THE MONUMENTS Before we turn to an analysis of the monuments, they deserve to be de- scribed in detail. In total, Hatra so far yielded six instances in stone, one crude painting and several graffiti. Of the representations in stone, four picture hu- man figures and two represent a reclining Heracles-figure. The majority of the stelae come from the fourteen small temples that have so far been discovered in the domestic area that surrounds the enormous temenos in the centre of the city. None of the stelae is dated by an inscription. Like the majority of the sculpture from Hatra they can roughly be dated between 150 and 240 of the Common Era.5 The first and best known specimen of a banqueting figure in relief embel- lishes a limestone lintel that originally decorated the main entrance of the so- called fifth temple, a small shrine located north of the great temenos (fig. 1).6 In the centre of the relief a bearded man reclines on his left side. He is identi- fied by the inscription to the right of his head as Nasru.7 Nasru wears the usual Parthian tunic and trousers. On his head is a large bulb decorated with wavy incisions, probably a knot of hair.8 In his left hand he holds a flat cup or patera, decorated with vertical incisions.9 In his right hand is a small object, 4  When speaking of ritual meals or banquets, I do not distinguish between eating and drinking parties. Strictly speaking, the latter should be called symposia. However, the available sources do not enable us to make this distinction in Hatra. 5  On the date of Hatrene sculpture see H. E. Mathiesen, Sculpture in the Parthian Empire. A Study in Chronology, Vol. I (Aarhus 1992), 73-77. 6  IM 56751 (2/Hatra/101). The lintel is 0.58 m. high and 2.67 m wide. It was on display in the National Museum in Baghdad, where I was able to study it in March 2002. Originally, it was encased in the façade, in the centre of the arch above the central entrance (cf. the reconstruction in F. Safar and M. A. Mustafa, Hatra. The City of the Sun God (Baghdad 1974), fig. 15 (in Ara- bic). Cf. also Mathiesen, Sculpture in the Parthian Empire, Vol. II, 213-214, no. 212, fig. 81, with references to earlier publications. 7  Hatra no. 33 (Henceforth abbreviated H33 etc.). Unless stated otherwise, the readings in the present article follow B. Aggoula, Inventaire des inscriptions hatréennes (Paris 1991). 8  This may be inferred from a statue of a priest with the same hairdo that was found at the East Gate of the city: W. al-Salihi, “A Statue of the Chief Priest from Hatra”, BAI 5 (1991), 35-40, figs. 1-2. Before this find, it was thought that Nasru was wearing a crown of some sort. 9  This flat cup deviates from the kantharos, the two-handled bowl normally carried by reclin- ing figures in Hatra. Parallels for this flat bowl can be found in Palmyrene funerary banquet scenes from the beginning of the third century onwards: M. A. R. Colledge, The Art of Palmyra L. DIRVEN 63 presumably a money bag. On the ground next to the upper end of the mattress a small figure stands on a base. It is badly weathered, but polos and base indi- cate this is the statue of a deity. To the right of Nasru a clean shaven man in Parthian costume approaches him. He is identified by the inscription to the left of his head as Vologash. Both Nasru and his oldest son Vologash appear in the ruling dynasty of Hatra in the second century CE. Nasru probably ruled in the thirties of the second century CE, whereas his eldest son ruled in the forties.10 With his right hand Vologash presses a round, unidentifiable object against his chest. In his lifted left hand he holds a leaf. Winged Nikai, clad in the tradi- tional peplos, approach the scene from either side. Both hold out garlands in their extended hands. A second young man stands on the extreme right of the lintel.11 A second stele that represents a reclining man in relief was found in the pronaos of shrine 13, a temple located to the east of the great temenos (fig. 2).12 Originally, this limestone stele was probably inserted into the front of the podium of the cella. The relief is now badly weathered. The stele shows a fig- ure reclining on his left side, supporting himself on his left elbow. The right (London 1976), 155. A flat bowl or phiale (usually made of metal) decorated with lobed or fluted patterns is common in Persian art: J. Boardman, Persia and the West (London 2000), 184-194. A Parthian silver bowl dated to the second – third century of the Common Era, now in the British Museum, has the same vertical flutes. The medallion in the centre pictures a reclining man who holds the very same shallow cup in his hand: D. Collon, Ancient Near Eastern Art (British Mu- seum Press, London 1995), 197, fig. 163. In all likelihood the king on the lintel from temple 5 holds a drinking cup as well. However, it cannot be excluded that it is actually a libation bowl. Phialai are frequently used as libation bowls: RE, Suppl. 7 (1940), 1026, “ƒialj”, (Luschey). In the Greek world, the sponde, the pouring of a liquid by a hand-held jug or bowl, is an integral part of Greek symposia. When poured to the ground instead of on an altar, libations are destined for the dead and for the gods who dwell in the earth. This explains why banqueting deities are frequently represented with a phiale instead of an ordinary drinking cup in the Greek world (with the exception of Dionysos, who invariably carries the kantharos). 10  Nasru mry’ is attested 34 times in the inscriptions from Hatra. Three inscriptions are dated between 128/9 and 137/38 CE (H272, 338, 346). According to H67, inscribed on the lintel of the cella of shrine 10, Nasru was the high-priest of the god: K. Dijkstra, Life and Loyalty. A Study in the Socio-Religious Culture of Syria and Mesopotamia in the Graeco-Roman Period based on Epigraphical Evidence (Leiden 1995), 235. Nasru certainly died before 176/77 CE, for in this year his son Sanatruq I was king of Hatra (H82). A Lord Vologash, son of Nasru is mentioned in H348. A Vologash mry’ is mentioned in three inscriptions, a Vologash mlk’ in another three. These inscriptions are not dated, but it is likely that Vologash succeeded his father. In turn, Vologash was probably succeeded by Ma’nu (H288 a-c), who was followed by Sanatruq I, brother of Vologash and son of Nasru. Cf. also M. Sommer, “Hatra – imperiale und religionale Herrschaft and der Steppengrenze”, KLIO 85 (2003), 384-398. 11  According to Safar, there is another inscription next to the head of the figure on the ex- treme right. It reads ‘Abdullah'. 12  Field number and Museum number are unknown. I saw the piece in the upper storeroom of the National Museum in Baghdad during a visit in March 2002. W. al-Salihi, “Inscriptions of Hatra”, Sumer 44 (1985-86), 104-105 (in Arabic), fig. 13 (drawing); W. al-Salihi, “The Excava- tion of Shrine XIII at Hatra”, Mesopotamia 25 (1990), 33, fig. 24. The stele is 0.81 m. high and 1.15 m. wide. A rectangular angle is cut from the left corner, from which follows that it was de- signed for a specific place. 64 BANQUET SCENES FROM HATRA arm extends along his right thigh, holding a small bag of coins. In his left hand the man holds a two-handled cup or kantharos. He wears a knee-length tunic with longs sleeves and Parthian trousers. The bearded face is badly damaged. The hair of this bare headed figure is rendered in snail-curls. Some of the curls were broken off in antiquity and an attempt was undertaken to repair them with iron pegs. On his right forearm stands an eagle with folded wings. Its body is represented in three-quarter view; the head is now missing.13 In antiq- uity an attempt was made to repair the bird’s head with an iron pin. The figure is identified by an inscription to the right of his leg as Hyusha, the son of Asalmy, servant of (the god). In a second line to the right of the eagle is an inscription in which his son Ryat asks to be remembered.14 A third instance of a reclining figure is an unpublished relief of unknown provenance that was kept in one of the storerooms at Hatra. The man is not accompanied by an eagle, but in other respects it is very similar to the relief from shrine 13. The last example of a reclining human figure in stone comes from the rim of a great crater made of Mosul marble that was found in situ in the pronaos of temple 10.15 The handles and the rim are decorated with different patterns of vine scrolls. At the height of the handles, two small figures can be seen reclin- ing on the rim. Unfortunately they are badly damaged and hardly visible on the photograph that was published by Safar. However, personal examination of the object confirmed the figures are represented in the common reclining pose and wear the habitual Parthian dress. The first stele that represents a reclining Heracles was found on the platform in the northern iwan in the great temenos (fig. 3).16 The rectangular stele is made of Mosul marble and represents the naked hero reclining on his left side on a narrow ledge. The body is frontal. The extended left leg rests on the floor and is slightly bent. The right leg, now largely missing, was originally bent and placed in front of the left leg. The right arm, most of which is now miss- ing, was originally bent and held the club, of which traces can still be seen. The left arm leans on a big lion’s head that in turn rests on paws that stand on 13  According to al-Salihi, “Excavation of Shrine XIII”, 33, there are remains of several neck- laces around the bird’s neck. I am unable to see these on the photograph. 14  H412. Cf. the reading proposed by F. Vattioni, Hatra, supplemento n. 81 di Annali vol. 54 (1994) fasc. 4. 15  IM 5815 (4/H/146). Safar, Hatra, fig. 317. The crater is 0.43 m. high, with a diameter of 0.59 m. 16  IM 73088 (11/Hatra/343). W. al-Salihi, The Sculptures of Divinities from Hatra, (Unpub- lished dissertation Princeton University 1969), 74-76; Safar, Hatra, fig. 93; Aggoula, Inventaire, 109; S. Downey, The Heracles Sculpture (The Excavation at Dura-Europos conducted by Yale University and the French Academy of Inscriptions and Letters. Final Report III.I, New Haven 1969), 86-88, pl. XXIII, 3. The stele is 0.36 m. high and 0.64 m. wide. The neck and head of the hero are missing, as well as the right leg from above the knee, the right arm from above the el- bow onwards and the head of the eagle. L. DIRVEN 65 the ground. The forearm and hand are broken off and with them any possible attribute. To his left, against the edge of the slab, is a tree trunk around which a snail curls up. The head of the snake is directed towards the eagle with out- spread wings that stands on top of the trunk. To the right of the reclining Heracles is an Aramaic inscription that reads ‘may be remembered Gedyhab the majordomo and Astet the scribe’. The inscription on the plinth informs us the stele was made by Shamashahab and Habyb.17 A second stele with Heracles Cubans was found in situ in a room on the southern side of the court of temple 5.18 Unfortunately the relief is badly weathered and many details are no longer visible. Heracles is shown reclining on his left side on the base of the relief, now chipped off. Face and torso are represented frontally, whereas the legs are rendered in profile. The front left leg is stretched; the right leg bent and placed in front of it. Heracles’ right arm rests on the bent leg. In his right hand is a large club, the top of which points towards the ground. The god is bearded and has short hair. In all likelihood, he is naked, but the relief is so damaged that this can no longer be established with certainty. In addition to these representations in stone, several drawings of reclining figures were found in Hatra. A crude painting on the right-hand side of the so- called ‘eagle niche’ at the East Gate of the city, pictures a male figure on a couch.19 The man is bare headed and dressed in the usual Hatrene costume. Of note is a religious standard that stands to the left of the couch. The standard was the object of cultic worship in Hatra. Hence this painting recalls the divine statue placed next to the reclining figure of Lord Nasru on the lintel from tem- ple 5. Roberta Venco-Ricchiardi published several graffiti of reclining figures from a domestic context.20 In all likelihood, most represent dining as part of daily life and hence ought to be distinguished from the representations found in the temples. However, a graffito from room no. 61 in the so-called ‘house of Manu’, is of particular interest for the present discussion (fig. 4).21 It pictures a group of figures reclining under a large arched structure. The two tiers and stepped crenellation recall the religious architecture of the large iwans in the great temenos, on the other side of the street of Beit Manu where the 17  H221. Cf. Aggoula, Inventaire, 109. 18  2/Hatra/38. Downey, Heracles Sculpture, 86, pl. XX, 1; al-Salihi, Sculptures of Divinities, 76-77, fig. 28; Safar, Hatra, fig. 229. The rectangular limestone stele is 0.50 m. high (according to al-Salihi 0.38 m.) and 0.79 m. wide. The upper corner on the left hand side and the lower part of the god’s left arm are missing. The base and legs of the god are partly chipped off. 19  J. K. Ibrahim, Pre-Islamic Settlement in Jazirah (Baghdad 1986), 161, no. 29, pl. 174. 20  R. Venco-Ricciardi, “Pictorial Graffiti in the City of Hatra”, Ancient Iran and the Mediter- ranean World. Proceedings of an International Conference in Honour of Professor Józef Wolski held at the Jagiellonian University Cracow, in September 1996 (ed. E. Dabrowa), Electrum 2 (1998), 187-205. 21  Venco-Ricciardi, “Pictorial Graffiti”, 198-199, fig. 11. 66 BANQUET SCENES FROM HATRA graffito was found. Hence the graffito in all probability pictures a sacred ban- quet. BANQUETING FIGURES AND BANQUETING ROOMS The fact that most bas-reliefs were set up in sanctuaries is bound to have implications for their meaning. Reclining figures have seldom been found in Syrian-Mesopotamian sanctuaries dated to the Seleucid and Parthian periods. Best known are Palmyrene tesserae that picture banqueting priests. As is well known, these small clay tokens were invitations for banquets celebrated in Palmyrene sanctuaries by a religious association in honour of one or several deities.22 The finding place of the Hatrene representations as well as the simi- larities with the Palmyrene tesserae, suggest the Hatrene reliefs testify to an actual ritual practice, viz. that cultic associations of some sort celebrated ritual banquets in Hatrene sanctuaries. This hypothesis is substantiated by textual and archaeological evidence from Hatra. Admittedly, the material from Hatra is not abundant and confined primarily to non-textual evidence. However, compared to contemporary mate- rial from other places in the Syrian-Mesopotamian realm, our scattered sources become revealing. Banqueting rooms in or attached to sanctuaries are a well- known feature throughout the contemporary Semitic world. Both inscriptions and archaeological material testify to the practice of cultic meals in Delos, Nabatea, Palmyra, the Palmyrène and Dura-Europos.23 In Hatra, only one inscription was found so far that unequivocally testifies to the celebration of banquets in temples. It comes from temple 13, a small shrine located east of the great temenos. Temple 13 consists of a sanctuary at the back of a courtyard that is lined with a number of small rooms. On the lin- tel that was once placed above the door of the room in the south west corner of the court H408 is inscribed. It commemorates the dedication of an iwan (kpt’) that is built by a certain Oqe son of Barnai son of Oqe son of Nabuketab, to the Great Gad of Ramgu in 235 CE.24 Oqe expresses the wish that he and his sons may recline (gn’) and be administered in the iwan for ever. The word kpt’, is hardly ever used for a banqueting room.25 In contemporary inscriptions from Palmyra and Nabatea we frequently encounter smk’, with the Greek equivalent 22  H. Seyrig, “Les tessères palmyréniennes et le banquet rituel”, Mémorial Lagrange (Paris 1940), 51-58. 23  For the practice in Palmyra and Nabatea and elsewhere in the Roman Near East, see the material assembled by J. Healey, The Religion of the Nabateans. A Conspectus (Leiden 2001), 165-169; T. Kaizer, The Religious Life of Palmyra. A Study of the Social Patterns of Worship in the Roman Period (Stuttgard 2002), 213-234. 24  H408. Cf. Dijkstra, Life and Loyalty, 205-207. 25  In addition the Hatrene inscription, one bilingual inscription from Palmyra, CIS 3912, uses the word in this sense. L. DIRVEN 67 sumpósion,26 whereas Greek inscriptions from Dura-Europos and Delos speak of ândrón, náov, o¤kov or êzédra.27 However, the word gn’, ‘to re- cline, to lie down’, leaves no doubt on the function of this kpt’ in temple 13. Subsidiary rooms that are used for ritual dining are a well known feature in the contemporary Semitic world. We know of similar situations in various temples in the Palmyrène.28 In Dura-Europos, many temples have subsidiary rooms that were used for ritual meals.29 In Hatra, subsidiary rooms are to be found along the courts of shrine 5 and 8, 9, 10 and 12.30 By analogy with the evidence listed above, it is likely that at least several of these were used for ritual dining.31 Archaeological evidence from Hatra suggests ritual banquets were not only celebrated in subsidiary rooms around the court, but also in the sanctuary proper. Most of the small sanctuaries in Hatra consist of a long narrow room with an entrance on one of the long sides and a cult chamber or niche in the rear wall opposite the door.32 In many cases, the long room or pronaos has a shallow platform that is about 10 centimetres high and one metre wide along its walls. 33 Similar low benches have been found in sanctuary units that were used for dining in the Palmyrène and on Delos.34 Presumably, the platforms 26  TEAD VII-VIII 156, note 20 (F. Brown). For an enumeration of the Greek words used in the Semitic world, see J. Starcky, “Autour d’une dédicace palmyrénienne à Sadrafa et à Du’anat”, Syria 26 (1949), 64-65. 27  DNWSI smk2; J. T. Milik, Dédicaces faites par des dieux (Palmyre, Hatra, Tyr) et des thiases sémitiques à l’époque romaine (Paris 1972), 149. The term both means couch, banquet and banqueting hall. PAT 0177 and PAT 1571 in all likelihood refer to a building. Perhaps the word also occurs in an inscription in archaic script that was recently published by A. Schmidt- Colinet and Kh. As’ad, “Zur Urbanistik des hellenistischen Palmyra. Ein Vorbericht”, Damaszener Mitteilungen 12 (2000), 71. In an inscription from the temple of Artemis in Dura- Europos, smk’ is transcribed in Greek sommako: F. Cumont, Fouilles de Doura-Europos (1922- 1923) (Paris 1926), 192-193, no. 50. Cf. E. Lipinski, Studies in Aramaic Inscriptions and Onomastics I, Orientalia Lovaniensia Analecta 1 (Leuven 1975), 69-70, who interprets the word as a reference to the participants of the banquet. In Nabatea, the word occurs in the inscription from the Turkmaniyyeh tomb in Petra: J. Healey, The Nabatean Tomb Inscriptions of Mada’in Salih (Oxford 1993), 23-42. 28  Instances are listed by D. Schlumberger, La palmyrène du nord-ouest (Paris 1951), 101-105. 29  S. B. Downey, Mesopotamian Religious Architecture. Alexander through the Parthians (Princeton 1988), 127-128. 30  Downey, Mesopotamian Religious Architecture, figs. 80, 84, 85 and 86. Unfortunately, ex- cavation of the courts of the temples was often not completed. Temples 6 and 11 are situated at the back of a courtyard, but it is not clear whether there were any subsidiary rooms around this court. Temple no. 14, still largely unpublished in September 2004, is situated at the back of a courtyard without subsidiary rooms. 31  The remainder of the small shrines (temple 1 in its first stage, and the temples 2, 3 and 4), opened directly onto the street. 32  As suggested already by D. Schlumberger, l’Orient Hellénisé (Paris 1970), 136. Downey, Mesopotamian Religious Architecture, 172, denied this possibility, arguing that the rooms are too big. 33  The low platforms are frequently made of marble, whereas the remainder of the floor of the pronaos is of plaster. 34  For the Palmyrène, see Schlumberger, Palmyrène, 104-105. For Delos, E. Will, “Banquets et salles de banquet dans les cultes de la Grèce et de l’Empire romain”, Mélanges d’histoire an- cienne et d’archéologie offerts à Paul Collart, Cahiers d’Archéologie Romande 5 (1976), 353-362. 68 BANQUET SCENES FROM HATRA were covered with mattresses or – perhaps more likely- with wooden beds on which the participants lay down for dinner.35 The functioning of the pronaoi of the small temples as ritual dining rooms is confirmed by craters that have been found here. Some craters were obviously still in situ at the moment of their discovery.36 Similar objects have been dis- covered in banquet halls elsewhere in the region, notably in the Palmyrène and Palmyra.37 As is well known, craters were used for the mixing of water and wine. That people in Hatra drank wine is confirmed by an inscription from temple 5 dated to 238 CE, that mentions a certain Qaimi, the daughter of Abdsimia, the wine seller.38 Some of the craters from sanctuaries in Palmyra and the Palmyrène bear inscriptions that dedicate the object to a particular de- ity; undoubtedly the god in whose honour the drinking occurred.39 Wine had of old been the appropriate drink during religious banquets and it may safely be assumed that the Hatrene banquets were no exception to this rule.40 In fact, 35  In a room in a temple in Delos, provided with a similar plinth, an inscription was found that mentions the dedication of six beds: G. Roux, Études déliennes, BCH Bulletin de correspon- dance hellenique, suppl. I (Paris 1973), 548, fig. 16. The plinth is also to be found elsewhere. According to Roux, op. cit. 552, its function is to protect the wooden legs of the bed against wa- ter that is used for cleaning. The option of the wooden klinè is preferable for Hatra, in view of a graffito from room 61 in Beit Ma’nu that pictures a bed on legs: Venco, “Pictorial Graffiti”, fig. 10. 36  Basins were found in shrines 3, 7, 8, 10 and 11: Safar, Hatra, 354, 359, 364, 366. In addi- tion, a drinking vessel (?) made of marble was found in the pronaos of chapel 7 (0.22 high, Safar, Hatra, fig. 259; 3/H/20). A drinking vessel made of alabaster was found in the pronaos of chapel 8a (Safar, Hatra, fig. 291. Diam. 0.15 m., IM 57780; 3/H/30). 37  Best known is the crater Daniel Schlumberger found in situ in the sanctuary of Abgal in Khirbet es-Semrine: Palmyrène, 60-61, pl. XXXV-XXXVI. It was this find that led Schlum- berger to identify many rooms in the Palmyrène as banquet-halls. In addition, Schlumberger found fragments of craters in Khirbet Chteib (op.cit, 31 and 70 no. 5 (not readable); close to the cella), Khirbet Madaba (op.cit, 34 and 75, no. 6, inscr. no. 49; in a room with benches along the walls), Marzouga (op.cit., 42 and 85 no. 10, inscr. No. 62, found in the cella) and Ouéchel (op. cit., 48 and 85, no. 4, inscr.no. 63; found in a room with benches along the walls). In Palmyra, a crater that was dedicated to Aglibol and Malakbel in 29 CE was found reused in the road between the theatre and the agora: A. Bounni and X. Teixidor, Inventoire des inscriptions de Palmyre XII (Damas 1975), 36-37, pl. IX, no. 39. Cf. also Kh. Al-As’ad and M. Gawlikowski, The Inscriptions in the Museum of Palmyra. A Catalogue (Warsaw 1997), 71, no. 109. Recently, a krater, probably from the Palmyrene, was published by F. Briquel Chattonet, “Un cratère palmyrénien inscrit: nouveau document sur la vie religieuse des Palmyréniens”, ARAM 7 (1995), 153-163. 38  H35. It follows from the dress of this statue that Qaimi was a priestess: Safar, Hatra, fig. 243. For the inscription, Dijkstra, Life and Loyalty, 225-226, with references to earlier literature. 39  For example the crater from Palmyra dedicated to Aglibol and Malakbel (above, note 37). 40  For the role of wine in the marzeah in the early period (Ugarit 13th century BCE and Judea in the 8th-7th century BCE), see D. Tarrier, “Banquets rituels en palmyrène et en nabatène”, ARAM 7 (1995), 170-171. For the role of wine in Palmyrene thiasoi, see Milik, Dédicaces, 188- 192. The word hmr, ‘wine’, is frequently attested on Palmyrene tesserae. Cf. the index of H. Ingholt e.a., Recueil des tessères de Palmyre (Paris 1955). A Palmyrene inscription dated to 243 CE (PAT 2743), records that a symposiarch bought wine for the priests for a whole year. In case wine was not served in the cult of a particular deity, this was a peculiarity worth mentioning in Palmyra: PAT 0319 (CE 132), a Palmyrene dedication by a Nabatean horseman to the god Shai‘al-Qaum, ‘who does not drink wine’. In Dura-Europos, a wine cellar was dedicated to Adonis: TEAD VII-VIII, no. 875. L. DIRVEN 69 craters played such a prominent role in religious banquets, that the word for crater, ’gn is used on Palmyrene tesserae for the banquet itself.41 The decora- tion of the crater from temple 10 described above, nicely illustrates the func- tion of the object in Hatra. Not only are the handles and rims embellished by elaborate vine scrolls, but two figures with drinking cups in their hand recline on the rim.42 In Hatra, the celebration of ritual banquets was not confined to the small shrines. In all likelihood, these festivities were also held in buildings in the great temenos in the centre of the city. Most illustrative is the graffito from Beit Ma’nu described above, that shows people who recline in a structure that recalls the large iwans in the main temenos in Hatra. Like the small shrines, the interior of the large iwans is surrounded by a shallow plinth; in likelihood the basement for portable beds. The amount of people that could participate in the festivities celebrated in the great temenos must have been very substantial. It may be compared to the large company that could assemble in the banquet hall in the temenos of the temple of Bel in Palmyra.43 Like some of the small shrines in the city, the court of the great temenos is lined with small rooms. Unfortunately, these rooms have only been partly ex- cavated and the results were insufficiently published. In one of the rooms along the west wall of the temenos two graffiti were found that mention a pÌr’, an assembly or symposion.44 The idea that worshippers also assembled in rooms around the court of Hatra’s main temple, accords well with Ernest Will’s suggestion that in Palmyra during the great festivals the majority of worshippers were seated outside the dining hall, in the portico of the temple of Bel.45 THE SOCIAL ORGANISATION OF RELIGIOUS BANQUETS Now that we have established that religious banquets at which wine was drunk were celebrated in the small temples in Hatra as well as in the sanctuary in the great temenos, we turn to the question of the character of these ban- quets; who participated in these events and for what reason? This is an ex- tremely difficult question. First, because we hardly have any information from Hatra to this respect. Second, because the scattered references that we do have, compared with similar material from the Semitic world, strongly suggests 41  Milik, Dédicaces, 108-119. 42  Compare the crater decorated with Dionysiac imagery that presumably originates from the Palmyrène published by Briquel-Chatonnet, “Cratère palmyrénien”, 155. 43  E. Will, “Les salles de banquets de Palmyre et d’autres lieux”, TOPOI 7 (1997), 873-887. 44  H282 and H283. Cf. DNWSI II, 907. In all likelihood, the word is attested twice on Palmyrene tesserae: Ingholt, Tessères, nos 23 and 304; Cf. M. Gawlikowski, Le temple Palmy- rénien, Étude d’épigraphie et de topographie historique (Palmyra VI, Warsaw 1973), 31, note 15. 45  Will, “Salles de banquets”, 878. 70 BANQUET SCENES FROM HATRA there is no unequivocal answer to this question. On the basis of the available material, it may be assumed that various cultic organisations existed side by side at the same time in the same place. The first thing that catches the eye when one looks at the religious organisa- tion in Hatra, is the dichotomy between the enormous banquet halls in the great temenos and the small banquet halls of the small shrines in the city around the temenos. Clearly, the dining rooms within these two types of build- ings were frequented by social organisations that differed in size. Probably, they differed in social composition as well. Probably, for unfortunately the in- formation on Hatra’s cultic and social organisation is scattered and insuffi- ciently studied. At a religious level, it is important to note that the deities wor- shipped in the temenos (the triad Maren, Marten, Barmaren, the goddess Allat and several other deities), also occur in inscriptions and representations from the small shrines. In contrast, a number of deities that figure prominently in the texts and other finds from the small shrines do not appear in the great temenos. On a social level, the rulers of Hatra play a prominent role in the construction and cult of the central sanctuary.46 The small shrines were appar- ently built and used by small social groups. In all likelihood many of these groups were based on kinship. The inscription from temple 13 mentioned above, in which an iwan is built for the use of a specific family, is an illustra- tive example of the family-based organisation of the small shrines.47 In view of Hatra’s location on the fringe of the desert it is feasible that these groups partly consisted of desert-dwellers, whereas others had settled in the city.48 Hypo- thetically, we may opine that the central temenos was the place of a set of cen- tralized cults, supported by the royal house and important for the city as a whole, whereas the small shrines were frequented by social groups of a small scale.49 The fact that these groups worshipped their own individual deities in 46  In Hatra, the political rulers simultaneously functioned as high-priest of the god (in all like- lihood Shamsh, the main deity of the city). This may be inferred from H67, in which Lord Nasru is identified as ’pkl’ rb’ d’lh’, high-priest of the god. Cf. Dijkstra, Life and Loyalty, 235-236, for a discussion of the idea that the political rulers yielded supreme religious power in Hatra. A lintel from the temple of Allat, pictures King Sanatruq, identified as such by an inscription, in his func- tion as priest: W. al-Salihi, “Allat-Nemesis. Iconographical Analysis of two Religious Reliefs from Hatra”, Mesopotamia 20 (1985), 131-132, 137 and fig. 43. 47  Above, p. 66. Of course this does not imply all cultic associations in Hatra were family based. Familial and professional organisation may very well have co-existed side by side. A pro- tective deity of a professional association is perhaps to be found in H58, found in front of the façade of temple 8. According to the reading of Aggoula, it mentions ‘the gad of the fullers’: Aggoula, Inventaire des inscriptions, 45. Followed by Kaizer, Religious Life of Palmyra, 220. In Dura-Europos, a cultic organisation composed of individuals from different families from Anat on the Euphrates assembled in the temple of Aphlad: TEAD V, 114, no. 418. Cf. Dijkstra, Life and Loyalty, 266. In the Syrian sanctuary at Delos, we know of a cult group that has a profes- sional basis. On the professional composition of the marzeah, compare below, note 56. 48  For the tribal factor in Hatrene society and administration see Dijkstra, Life and Loyalty, 185-188. 49  Also suggested by Kaizer, “Some Remarks about the Religious Life of Hatra”, TOPOI 10 (2000), 231. L. DIRVEN 71 combination with the gods of the city, illustrates the uniting role of religion in this city on the fringe of the desert.50 The division into a central cult and a family based religious organisation re- calls the situation in contemporary Palmyra.51 Information on Palmyra may therefore help to illuminate the situation in Hatra. Whereas in Hatra we are best informed about the banquet halls in the small temples, information from Palmyra mainly concerns the central cult. The banquet halls that have been discovered in Palmyra are not numerous, but they are big.52 Best known is the enormous banquet hall in the temenos of the temple of Bel, in which was space for about 150 people. Within this hall, the marzeah of the priests of Bel played a crucial role. Following J. Milik, E. Will rightly pointed out the civic character of both the banquet hall and the marzeah of Bel.53 In Palmyra and Nabatea the main Aramaic term to denote the cult groups is marzeah, rendered in Greek as sumpósion. The institution is already attested in texts from Ugarit dated to the14th century BCE.54 A hotly debated issue with respect to the character of the marzeah is its presumed funerary character. Recently, the funerary character of this institution was denied by Alavoine.55 Given the scattered references, long history and occurrence over a large geo- graphical area, it is difficult to arrive at a general conclusion with respect to the character of the marzeah. In most cases, however, a funerary role is indeed not very likely. The main argument to this respect is not provided by obscure 50  If we are right in suggesting that the small shrines were visited by citizens as well as by their kinsmen from the desert, they play an important role in the construction of a family-identity. In turn, the deities from the central cult and the royal house tie these families to the city and its ruling class. 51  Dirven, The Palmyrenes of Dura-Europos, 28-29, with further references. 52  In total, four banquet halls were found in the city. E. Will, “Salles de banquets de Palmyre”, 873-887. By analogy with the situation in the Palmyrène and in Hatra, it may be pre- sumed there were many shrines in which the religion of the family figures prominently. Unfortu- nately, however, none of these shrines was found so far. A family-based cult may also be inferred from the tesserae that mention the bny… PN. Dijkstra, Life and Loyalty, 90, suggest these are children of the same father, who jointly bore the expenses of the feast. In all likelihood, these feasts are celebrated in honour of the deity, mentioned on the tesserae, who functioned as the god of the family. In addition to family-based groups, there were professionally based groups in Palmyra as well: Kaizer, Religious Life, 220. 53  Will, “Salles de banquet de Palmyre”, 877, following Milik, Dédicaces, 109-110 and 280- 281. 54  Several texts from Ugarit inform us that the marzeah owned a house (bt, byt cf. KTU3.9; RS15.088; RS15.70). Unfortunately the texts do not specify the location of these houses. KTU3.9 speaks of a marzeah in the house of Shamumanu, which suggests we are dealing with a secular building. A very illuminating passage in which no mention is made of a marzeah, but that does refer to dining inside temples, can be found in the Epic of Aqhat from Ugarit, dated to the 14th century BCE. A passage summarizes the responsibilities of a paterfamilias. One of these is.. to eat his (i.e. his father’s) emmer in the temple of Baal [and] his part in the temple of El. KTU 1.17 I 32-33. Cf. the translation and interpretation with K. van der Toorn, Family Religion in Babylonia, Syria and Israel. Contnuity and Change in the Forms of Religious Life (Leiden 1996), 154-155. This passage moreover suggests the place of the father was taken by the son. 55  V. Alavoine, “Le Mrzh est-il un banquet funéraire? Étude des sources épigraphiques et bibliques (Am. 6.7 et Ier. 16, 5)”, Le Museon 113 (2000), 1-23. 72 BANQUET SCENES FROM HATRA and multi-interpretable references in literary and epigraphic sources, but by the fact that the marzeah is organised on a professional and not on a family basis.56 Throughout the history of the ancient Near East, the funerary cult or cult of the ancestors, is based on kinship ties.57 This implies that the character of the marzeah cannot be used as an explanatory key for all representations of reclin- ing figures. The institution is particularly unfit to interpret representations that belong to a funerary context, such as the banquet scenes from Palmyra.58 It 56  This seems to have been the case in Ugarit: Van der Toorn, Family Religion, 169. In Palmyra, the best known marzeah, that of the priests of Bel, is best defined as a professional or- ganisation. Other marzeah known from Palmyrene inscriptions have a professional basis as well. Cf. Kaizer, Religious Life, 215-218, for the relevant texts. Healey, Religion of the Nabateans, 166, reports of a series of marzeah inscriptions, at that time still unpublished, from el-Madras at Petra. These make clear that the thiasoi would often consist of a professional or trade group. 57  In Palmyra, the organisation of the tombs proves this still to be the case in the Roman pe- riod. Cf. J.-B. Yon, Les notables de Palmyre (Beyrouth 2002), 197-221. However, concomitantly, we have to reckon with the possibility that the marzeah of the priests of Bel also played a role in celebrating the memory of its members. This is suggested by the funerary symbols on several (definitely not all!) tesserae. For the funerary symbolism of tesserae, compare below, note 58. In the Roman world, professional associations played a prominent role in the interment of its mem- bers. Possibly, this Roman custom influenced the character of the marzeah in Roman Palmyra. O. M. van Nijf, The Civic World of Professional Associations in the Roman East (Amsterdam 1997), 31-72. 58  Most recently on banqueting scenes in Palmyrene tombs: K. Parlasca, “Palmyrenische Sarkophage mit Totenmahlreliefs- Forschungsstand und Ikonographische Probleme”, Akten des Symposiums ‘125 Jahre Sarkophag-Corpus’, Marburg, 4.-7. Oktober 1995. Sarkophag-Studien I (Mainz am Rhein 1998, ed. G. Koch), 311-317. The non-funerary character of the marzeah has become an important argument in favour of the supposed purely profane character of banqueting figures in Palmyrene tombs. The first to advocate a profane meaning of the funerary banquet was Henry Seyrig, “Le repas des morts et le “banquet funèbre à Palmyre”, AAAS 1 (1955), 32-40. Seyrig does identify certain funerary busts from loculi as a reference to funerary meals, but he (rightly) notes that these funerary rituals ought to be distinguished from commemorative ban- quets (the remains of meals or table ware that have been found inside the tombs are in all likeli- hood connected to these funerary rituals, and not to the commemorative meals). Seyrig argues the reclining banqueting figures from the tombs have a purely profane meaning and illustrate the so- cial prestige of the deceased. Seyrig’s interpretation is followed by many scholars; E. Will, “Le Mrzh est-il un banquet funéraire? Étude des sources épigraphiques et bibliques (Am. 6.7 et Ier. 16, 5)”, “Le relief de la tour de Kithot et le banquet funéraire à Palmyre”, Syria 28 (1951), 70- 100; Tarrier, “banquets rituels”, 177; Alavoine, “Le Mrzh est-il un banquet funéraire?”, 13. The profane interpretation of the funerary banquet scenes is directed against the older view advocated by F. Cumont, Recherches sur le symbolisme funéraire des romains (Paris 1942), 421, who as- sumed a relationship between the ritual banquet (as represented on the tesserae) and the funerary banquet in the tombs. A similar interpretation is advanced by Mesnil du Buisson. Recently again: A. Sadurska, “L’art et la société. Recherches iconologiques sur l’art funeraire de Palmyre”, Archeologia (Warsawa) 45 (1994), 14; A. Sadurska, A. Bounni (en colloboration avec K. Al- Ass’ad and K. Makowski), Les sculptures funeraires de Palmyre (RdA, suppl. 13, Rome 1994), 191-2. A compromise of the two positions was recently advanced by Dominique Tarrier in a pa- per presented at the Aram conference in 1995. Tarrier follows Seyrig’s profane interpretation of the funerary banquet, but insists on a possible relationship between banquets and commemorative rituals for the deceased: “banquets rituels”, 178. According to Tarrier, this is due to the fact that the non-funerary, Semitic marzeah assimilated with the Greek Dionysiac thiasos in the Hellenis- tic period. Consequently, the marzeah changed from a festive meeting into an organisation that was intended for the salvation, or at least the commemoration, of its members. I agree with Tarrier that the character of the marzeah changed in Roman Palmyra, but not for the same rea- sons (cf. the preceding note). Eating and drinking in honour of the dead is not a typical western L. DIRVEN 73 falls beyond the scope of the subject of this paper to re-open the discussion on the meaning of funerary banquets in Palmyra.59 However, the following argu- ments have important implications for this discussion as well. It is perhaps not without importance that banquet halls abound in Hatra, whereas the term marzeah is missing in the epigraphic record from the city. The only identifiable group that celebrated religious meals together is organ- ised on a family basis. In the tradition of the ancient Near East, there are many occasions on which ritual meals were celebrated in temples.60 One of these occasions is the commemoration of the death. In his evocative study on family religion in Babylonia, Syria and Israel, Karel van der Toorn has shown that cultic meals in honour of the dead played a prominent role in these three Semitic cultures. In Babylonia the cult of the ancestors or kispum, included food offerings to the ancestors. On a daily basis, small offerings were made, whereas the ancestors were presented a full meal at the end of the month. Once a year the Babylonians celebrated a kind of all souls. A similar situation existed in ancient Israel. The account in 1 Samuel 20, suggests all members of a family were supposed to be take part in this yearly event.61 In Babylonia, the meal of the ancestors was ordinarily celebrated in the house of the family. In Israel and the West Semitic milieu, however, the dead were addressed in the temple of the family god.62 This results from the Syro-Phoenician conception that the dead are presumed to dwell with the god of the family. Post-funerary offerings first have to be accepted by the family god before they can reach the ancestors. Consequently, the temple of the fam- ily god sometimes functions as the place where the dead were addressed. Of course the old texts referred to above cannot be cited as proof for the celebration of funerary banquets in the small temples in Hatra or elsewhere in Syria and Mesopotamia during the first centuries of the Common Era. In the Palmyrène, however, banquet halls in temples are frequently intended for the well-being and memory of their founders and their family. From this Françoise Briquel Chattonet inferred that the memory of the deceased was related to the phenomenon. In fact, it is custom with a long tradition in the Near East. Hence there is absolutely no need to use the Dionysiac ideology to explain the banquet in memory of the deceased. 59  I plan to devote a study to this subject in the near future. In spite of the great quantity of archaeological and epigraphic material, Palmyrene funerary beliefs are neglected in recent stud- ies on Palmyrene religion such as Ted Kaizer’s The Religious Life of Palmyra. One of the few studies devoted to this subject is H. J. W. Drijvers, “After Life and Funerary Symbolism in Palmyrene Religion”, La soteriologia dei culti orientali nell’ Impero Romano. Atti del Colloquio Internazionale su La soteriologia dei culti orientali nell’ Impero Romano, Roma 24-28 Settembre 1979, ed. U. Bianchi and M. J. Vermaseren (Leiden 1982), 709-733. 60  Nor was the commemoration of the death the only occasion at which meals were celebrated in temples. For example, from Mari we know of the pagra’um offerings; that can be compared with Hebrew Zebah selamim, the sacrificial banquet at which visitors enjoyed consecrated meat; Archives royals de Mari: 14: 217 and ARM 21: 160-161. 61  Van der Toorn, Family Religion, 212. 62  Van der Toorn, Family Religion, Chapter 7: Family Religion in the West: Persisting Pat- terns. 74 BANQUET SCENES FROM HATRA ritual of the sacred banquet.63 As for Hatra, the representations of reclining human figures and the reclining Heracles-figures indicate that meals in memory of the ancestors were celebrated in Hatrene sanctuaries. INTERPRETING RELIGIOUS BANQUETS IN HATRA The first indication to this respect is provided by the stelae from temple 13, that pictures a reclining man with an eagle on his arm. The accompanying in- scription identifies him as Hyusha, the son of Asalmy, servant of the god.64 The same individual asks to be remembered in a text inscribed on the architrave of the cella, that is the place where the relief was originally set up. Hence we may presume Hyusha was involved in the foundation of this part of the building.65 This is confirmed by the small moneybag he carries on the relief. At first Hyusha seems to be an ordinary human being. However, the eagle provides the scene with a supra-natural flavour. To establish the meaning of this bird is, however, not an easy matter. The eagle occurs in Hatra in many different constellation and has various connotations.66 A major clue to its sense in the present context is the text inscribed to its right, in which Hyusha’s son Hyat asks to be remembered.67 This is unusual and seems to imply that the stele was not set up by Hyusha, but by his son in memory of his father. On contemporary funerary monuments from the Hauran, the living commissioner of the monuments frequently asks to be remembered.68 Possibly, therefore, the eagle signifies that Hyusha passed away. Numerous examples of eagles in a funerary context are known from Edessa, Zeugma, Palmyra and the remainder of Syria.69 A striking parallel for the Hatrene relief comes from a funerary cave in Kara Keupru, a site within walking distance of ancient Edessa. A rock relief that presumably dates to the third cen- tury CE, pictures the reclining figure of the deceased and the bust of a woman. 63  Briquel-Chatonnet, “Cratère palmyrénien”, 163. 64  H412. 65  Cf. H407, the priest Hyusha (and all his family?) asks to be remembered before the great Gad (i.e. the Gad of Ramgu). Cf. the inscription on the architrave of the main entrance of the pronaos or antecella, H406, in which two other individuals ask to be remembered. 66  The closest parallel is provided by the eagle on the tree next to the reclining Heracles-fig- ure. Unfortunately, its significance is equally unclear. For example, the eagle is related to Maren, the head of Hatra's ‘pantheon', and as such a symbol for the sacred precinct. In some texts Maren is labelled the eagle (nsr’), from which it may be inferred that the eagle was some sort of mani- festation of this god: H88 and H155. In addition, an eagle stands on the diadem of Hatrene kings and several Hatrene gods, and perches on the top of Hatrene cultic standards. The bird is also represented on coins minted in the city. 67  The name Hyat is also attested in H284, but this must be another individual. For the mean- ing of the Arabic name, see S. Abbadi, Die Personennamen der Inschriften aus Hatra (Texte und Studien zur Orientalistik, Hildesheim 1983), 163. 68  A. Sartre-Fauriat, Des tombeaux et des morts. Monuments, iconographie et société en Syrie du Sud Ier s. av.-VIIIe apr. J.C., vol. 2 (BAH 158, Beyrouth 2001), 222. 69  Cumont, “L’aigle funéraire”, Études syriennes (Paris 1917), lists the material 38-56. For an alternative interpretation of the funeray eagle: J. Wagner, Seleukia am Euphrat/Zeugma (Wiesbaden 1976), 157-158. L. DIRVEN 75 Above the man’s right arm is an eagle with outstretched wings.70 A very similar representation comes from a funerary cave of a veteran of the 8th legion, in Qâtoûrâ in Djebel-Halaka, west of Aleppo.71 Several Palmyrene tesserae pic- ture a reclining priest together with an eagle with outspread wings.72 The funerary symbolism of the tesserae is a hotly debated issue. However, the funerary meaning of these particular tesserae is likely in view of the popularity of the eagle on funerary monuments. In the tomb of Hairan and other tombs, the eagle is frequently depicted above a series of cubiculae.73 On a funerary relief of unknown provenance that is now in the local museum in Palmyra, an eagle perches above a man’s head (fig. 5). A second palmyrene relief that is now in the museum in Istanbul pictures a pair of eagles that hold up a wreath.74 We may conclude that the relief of Hyusha was erected by his son in his memory. In all likelihood the family of Hyusha celebrated banquets in his memory in this sanctuary. Apart from the present relief, we know of at least one other representation of a deceased person that was set up in one of the small shrines. I am referring to the famous statue of Abu, daughter of Gabilu, which was erected for her in the fourth shrine by her husband Asa, son of Shamshaqab after she died, only 18 years old.75 It is noteworthy that no 70  M. Pognon, Inscriptions sémitiques (1907), 182, pl. IX. J. Leroy, “Mosaïques funéraires d’Edessa”, Syria 1957, 329-330; H.J.W. Drijvers and J. Healey, The Old Syriac Inscriptions of Edessa and Oshroene. Texts, Translations and Commentary (Leiden 1999), pl. 10. 71  In a niche in the vestibule of the tomb is a representation of the deceased reclining in relief, with above him an eagle with outspread wings. M. Prentice, American Expedition to Syria III, Greek Inscriptions (1908), 127, no. 111. A good reproduction of the monument in M. Jullien, Sinaï et Syrie, 66. 72  Already mentioned by Cumont, “Aigle funéraire”, 53, according to whom the tesserae are invitations for a funerary banquet (my italics). Cf. Ingholt, Tessères, nos. 774, 700 (pl. XXXIII). Under the coach is an inscription that identifies this priest as lqnys brs. The same individual is represented reclining on no. 613. Here, he is approached by a Victory instead of an eagle. Per- haps busts with an eagle hovering above their heads are likewise representations of deceased per- sons (nos. 464, 578, 1119, 1127). Cf. also Tubach, Schatten des Sonnengottes, 432. 73  Drijvers, “After life in Palmyrene Religion”, 718. 74  Colledge, Art of Palmyra, 158, with note 618. Funerary eagle: Palmyra museum 28 and 424. Pair: Istanbul 3801. 75  IM 56730 (2/Hatra/60). D. Homès-Fredericq, Hatra et ses sculptures parthes. Étude stylistique et iconographique (Istanbul 1963), no. 27, pl. VI, 4; Safar, Hatra, fig. 211. H30, in- scribed on the front of the pedestal reads: ≥lmt’ dy ’bw brt gblw dy ’qym lh ’s’ br sms‘qb b‘lh dy mytt brt snt 18 bgn mrn wmrtn wbrmryn wb‘lsmn w’tr ‘t ‘l mn dy …….. whdy lh w‘l ns’ dy mly [ ] [ ] Translation: This is the statue of Abu, daughter of Gabilu, which has erected for her Asa, son of Shamshaqab, her husband, as she died at the age of 18. Mercy of Maren and Marten and Barmaren and Baalshamin and Atargatis for whoever …… 76 BANQUET SCENES FROM HATRA funerary monuments with figural representations and inscriptions were found in the graves at Hatra. The statue of Abu and the relief of Hyusha suggest monuments of this kind were erected in the temples instead.76 Hatra conveys the impression that the living and the dead were very close. In Hatra, ritual banquets were not only celebrated in memory of the ances- tors but also in memory of the deceased ruler, especially if the latter had played a role in the foundation of the shrine. This can be inferred from the re- lief that pictures the reclining Lord Nasru from temple 5. In the past, several scholars interpreted the scene as a victory banquet, on which Lord Nasru cel- ebrates his victory over Trajan in 116 CE.77 This interpretation does not carry conviction, if only because Nasru did not rule Hatra at this early date.78 The idea of a military victory is based primarily upon the two Nike figures who flank the scene.79 However, these two divine figures carry a garland, instead of a wreath, the usual attribute of the goddess of victory. The fact that we are dealing with two instead of one figure, also suggests we are no longer dealing with the personification of Victory proper. Instead, the two figures must have a more abstract meaning. Like similar figures in contemporary monuments from Syria and Mesopotamia, they probably refer to divine peace and prosperity. Seen in this light, the lintel of shrine 5 becomes a remarkable parallel to the inscription on the lintel originally placed above the entrance of the cella of shrine 10, that reads: ‘with good prospects for the life of Lord Nasru, the high- priest of the god’. 80 In the discussion of this inscription, Klaas Dijkstra sug- gested Nasru may very well have instigated the foundation of temple 10.81 The money bag that Nasru carries on the lintel from temple 5 indeed suggests he contributed to its construction. It does not follow from Nasru’s presence on the lintel of temple 5 that the ruler personally attended the banquets celebrated here. By analogy with in- scriptions that are erected in honour and in memory of the founder, we may assume Lord Nasru was remembered during the banquets in shrine 5. This may already have been the case during his lifetime. We may, however, safely assume that it was his intention to be remembered for his generosity after his death as well. This may be inferred from the presence of prince Vologash and 76  As already suggested by F. Dorna Metzger, “Hatra: edifici funerari”, TOPOI 10 (2000), 197-215. 77  Denzter, “iconographie iranienne”, 41-42; al-Salihi, “Excavation of Shrine XIII”, 33. 78  Sommer, “Imperiale und religionale Herrschaft”, 390. 79  At the basis of this hypothesis stands the reading of A. Caquot, “Nouvelles inscriptions araméens de Hatra”, Syria 30 (1953), 238-239: nsr ’ blgs, ‘victory of Vologash’. This reading is followed by Homès-Fredericq, Hatra et ses sculptures parthes, 58. Some way or another the in- terpretation of victory prevailed, even after Nasru was read as a personal name. 80  P. Linant de Bellfonds, “Nike (in peripheria orientali)”, LIMC VIII.1 (1997), 882, argued that Tyche-figures in the Roman Near East assimilated certain features of Gad, the Semitic per- sonification of good fortune. 81  H67. Cf. the reading and commentary with Dijkstra, Life and Loyalty, 235. L. DIRVEN 77 the Nike-figures. The position taken by prince Vologash reminds one of the usual position of servants on Palmyrene funerary reliefs.82 In an actual ban- quet, it is most unlikely that the prince fulfilled such a humble role. Alterna- tively, we may assume he brings funerary offering to his deceased father. These offering will reach his father via the god who is represented standing next to him. The idea that at one level the scene may be interpreted as an offer- ing represented to the deceased king and his god fits well with the garlands carried by the Nike-figures.83 On representations from Hatra, Palmyra and Dura-Europos, garlands are brought as offerings to the deity.84 The hypothesis that people in Hatra celebrated funerary banquets in honour of their ancestors and deceased kings in the small shrines is confirmed by the stelae representing a reclining Heracles in relief. In the iconography of Hatra, Heracles is the only divine figure who is represented in a reclining pose. Ap- parently, Heracles was closely associated with drinking wine in Hatra. Apart from his reclining pose, he is frequently represented standing holding a drink- ing cup in his hand. When we rely on religious representations, Heracles may rightfully be called the most popular deity worshipped in the small shrines. Out of the 107 statues of divinities that were unearthed here, 29 are representa- tions of the Greek god. Not only do about a quarter of all divine images repre- sent Heracles, the god is also attested in 9 out of the 13 sanctuaries.85 His re- clining position and drinking cup suggest this popularity is connected with the religious banquets that were celebrated in these shrines. In turn, Heracles’ role in these banquets is related to his chthonic character in Hatra. The reclining Heracles or Heracles Cubans is well known in the Graeco- Roman world. The type already occurs on Greek vases and is attested in sculp- 82  Especially popular on the small third-century banquet reliefs: Colledge, Art of Palmyra, 131-132 and pl. 109. 83  Instances of Nikai or Victoriae carrying garlands are rare. Cf. “Nike”, LIMC VI (1992) (A. Moustaka, A. Goulaki-Voutira, U. Grote), where the attribute is missing altogether. In the Roman period, there are somes instances, but few: “Victoria”, LIMC VIII (1997) (R. Voll- kommer), no. 158 and nos. 207-208 and 238-239. Victory-figures carrying garlands are, how- ever, a common motif on the so-called ‘girlande sarcophagus’, produced in Asia Minor between 140-170 CE. Cf. Koch and Sichtermann, Römische Sarkophage, figs. 480-2, 501, 509, 512, 516, 518, 533, 535, 539, 540, 562, 572-3, 575. Contrary to the Nikai in Hatra, these Victoriae carry large garlands over their shoulders, that are used as a decorative device in their own right. 84  Parallels in Hatra of garlands being offered to the statue of a deity: a relief from the North Gate: Sumer, 1980, 173, fig. 26. Safar, Hatra, fig. 177: relief representing three women (god- desses?), two of whom carry a garland. On Palmyrene sarcophagi with sacrificial scenery, serv- ants carry garlands towards an altar: Kaizer, Religious Life of Palmyra, 180-81. A garland can be recognised in the object carried by the figure of Conon on the far left of the so-called ‘Conon’ fresco from the temple of Zeus (formerly the Palmyrene gods) in Dura-Europos: Cumont, Fouilles, 44-46, pls. XXXI-XXXII; Colledge, The Parthians, fig. 70. In Palmyra, altars are em- bellished with garlands in relief: Colledge, Art of Palmyra, 52. In a passage of De Dea Syria (par. 46) we read that in Hierapolis there is an altar that is always garlanded. To fulfil a vow, people carry garlands to the altar. 85  No representations of Heracles were found in the temples 2 and 6. 78 BANQUET SCENES FROM HATRA ture and other media from the Hellenistic period onwards.86 The scene is not connected with one particular event from Heracles’ life and its meaning varies according to period and context.87 Heracles Cubans is already attested in the East in the Seleucid period. In Behistun, in present day Iran, a Greek named Hyacinthus son of Pantauchus, dedicated a colossal relief of Heracles Cubans for the safety of the governor Cleomenos in 148 BC.88 The Greek inscription dedicates the monument to Heracles Kallinikos, an epithet that refers to Heracles’ victorious deeds and that is often connected to his last labour, his de- feat of Cerberus and descend into the underworld.89 Later, the reclining He- racles appears on the Palmyrene tesserae, and two reliefs from Dura-Euro- pos.90 In the Roman West, Heracles Cubans sometimes occurs in a funerary con- text. The reclining position connects him with the god Dionysos and therewith refers to a happy afterlife.91 The divinised hero becomes a prototype for the deceased who is sometime represented as Heracles Cubans on Roman sar- cophagi.92 Of course it cannot be inferred from the situation in the Roman West that Heracles fulfilled the same function in the eastern provinces. How- ever, circumstantial evidence does suggest that the same association was made in Palmyra. Apart from Heracles, the only god who is pictured in a reclining pose on Palmyrene tesserae is Dionysos.93 Whereas further information on Heracles is missing, there can be no doubt that Dionysos had funerary conno- tations in Palmyra. A reclining Dionysos was painted in a third-century tomb.94 The chthonic character of Heracles Cubans in Hatra is best illustrated by the relief that pictures him next to a tree with a snake and an eagle. Though the reclining hero belongs to the standard Greco-Roman type of Hercules Cubans, there are no western parallels for the tree with the snake and eagle.95 In all  For the motif on Greek vases see S. Ruth Wolf, Herakles beim Gelage: eine Motiv- und 86 Bedeutungsgeschichtliche Untersuchung des Bildes in der archaisch-frühklassischen Vasen- malerei (Köln 1993). For other monuments dated to the Hellenistic and Roman periods, see H. Scharmer, Der gelagerte Herakles (Berlin 1971). M. Bieber, “Archaeological Contributions to Roman Religion”, Hesperia 14 (1945), 272-277. 87  In general, it can be said to refer to two different events. First, it may represent an unspeci- fied moment from the hero’s life, picturing him at rest after one of his labours. This may turn out negatively, viz. the drunken hero. Second, it may picture the hero after his death and divination, participating in a divine banquet. 88  Downey, Heracles Sculpture, 1, pl. I.1 89  “Herakles”, PW III (1918), 1002 (Gruppe) 90  Ingholt, Tessères, no. 245. Downey, Heracles Sculpture, 33-34, nos. 40-41, pl. XVII. 91  On the mysteries of Dionysos and hopes for the afterlife, see W. Burkert, Greek Religion. Archaic and Classical (Oxford 1985), 293-295. 92  F. Cumont, Recherches sur le symbolisme funéraire des romains (Paris 1942), 415-416, pl. XLIII.1. 93  RTP nos. 14 and 202. 94  Colledge, Art of Palmyra, 87, pl. 118. 95  In western art, a tree often appears in the background of Hercules Cubans, but generally only bow and quiver are hanging from it. The tree with the snake around it recalls the tree of the Hesperides. However, Heracles is standing next to this tree in western art. L. DIRVEN 79 likelihood, these motifs are a local invention. Though it is at present impossi- ble to give a coherent interpretation of these motifs, the snake is best explained as a reference to the god’s chthonic character.96 The most adequate parallel for the snake can be found on the so-called ‘Cerberus relief', a stele from temple 1.97 Snakes are omnipresent on this stele that represents a bearded deity with three dogs. They spring from the god’s forehead, shoulders and belt and are painted and carved in great quantity in the background. Elsewhere in Hatra, dogs are related to Nergal, the Babylonian god of the underworld. 98 Conse- quently, the god on the stele from temple 1 can surely be identified as such. The snakes are in total accord with the chthonic character of this Babylonian deity. In spite of a recent article on the identity of Heracles at Hatra, a strong case can be made for Nergal’s identification with Heracles in Hatra.99 The fact that on the relief under discussion Heracles Cubans is represented with one of the animals of Nergal, substantiates this identification. In the tradition of the ancient Near East contact with the deceased is nor- mally made with gods of the underworld, or with gods that regularly visit the underworld, such as Shamash. This contributes to the explanation of Heracles- Nergal’s popularity in the small shrines. After all, one of the functions of these rooms was the celebration of commemorative banquets for the ancestors. It was through the family god and through Heracles-Nergal, the god of the un- derworld, that contact was being made with the deceased. At these occasions, the living and death constituted one big family. It is only by way of analogy 96  Already suggested by Downey, Heracles Sculpture, 53 and 87. 97  Safar, Hatra, fig. 183. H. Stierlin, Städte in der Wüste. Petra, Palmyra und Hatra- Handelszentren am Karawanenweg (Stuttgart 1987), pl. 188 (in colour). In addition, snakes fig- ure prominently on two other stelae that in all likelihood also picture Nergal: on an altar from temple 2, a bearded god is represented with a double axe in his lifted right and two snakes in his lifted left hand (Safar, Hatra, fig. 191; Stierlin, Städte in der Wüste, fig. 183). On a relief from temple 5, a snake is represented behind a bearded, winged god who holds a goat (Safar, Hatra, fig. 230), and in the hand of a bearded deity who stands next to Athena-Allat (Safar, Hatra, fig. 225) Finally, a snake is represented next to a beardless military deity who has rays around his head and a crescent rising from behind his shoulders: Safar, Hatra, fig.75. Cf. Tubach, Schatten des Sonnengottes, fig. 5, who identifies this deity as Barmaren. 98  Most evidence to this respect comes from temple 10. H70, inscribed on a plain plate, men- tions Nergal the dog. H72 is inscribed on a statuette of a dog and refers to the dog or dogs of Nergal (Safar, Hatra, fig. 313). H73 that reads ‘made for Nergal’, is inscribed on the cover of an offering box representing a dog in relief (Safar, Hatra, fig. 318). H71 is dedicated to ‘Nergal the dog’ and is inscribed on the base of a statuette of Heracles from whom only the bare feet, lion skin and club remain (4/Hatra/55). The upper fragment of a small incense altar pictures a naked (?) god with an eagle on his head on the front, and seated dogs on the sides of the altar (IM 58115 (4/Hatra/132): S. Downey, “A New Relief of Nergal from Hatra”, Sumer 26 (1970), 227-229, fig. 1; al-Salihi, Sculptures of Divinities, 113-5, pl. XXXIII-V); Safar, Hatra, fig. 312. 99  Contra. T. Kaizer, “The ‘Heracles Figure’ at Hatra and Palmyra: Problems of Interpreta- tion”, Iraq 62 (2000), 219-231. V. Christides, “Heracles-Nergal in Hatra”, Berytus 30 (1982), 105-115, already advocated that both deities were identified in Hatra due to their shared chthonic character. Admittedly, Christides did not advance persuasive arguments in favour of Heracles’ chthonic character. The arguments advanced above may be cited to substantiate Christides' hy- pothesis. 80 BANQUET SCENES FROM HATRA with Babylonian and Syrian traditions that we can speculate about the fre- quency of these meals. As stated above, in Babylonian and west-Semitic tradi- tion, food was offered to the death on a daily, monthly and yearly basis. The latter occasion was a kind of all souls, an event in which all family-members took part. Although this must remain mere speculation, it is not hard to imag- ine how in Hatra members of a family, settled as well as nomadic, assembled once a year in the sanctuary of their clan to commemorate their ancestors. Fig. 1: lintel from shrine 5. After H. E. Mathiesen, Sculpture in the Parthian Empire, Vol. II, fig. 81. Fig. 2: Stele from shrine 13. After W. al-Salihi, “The Excavation of Shrine XIII at Hatra”, Mesopotamia 25 (1990), 33, fig. 24 L. DIRVEN 81 Fig. 3: Stele of Heracles from the great northern iwan. After Safar, Hatra, fig. 93 Fig. 4: Graffito from room 61 in the house of Manu. After Venco-Ricciardi, “Pictorial Graffiti”, 198-199, fig. 11. 82 BANQUET SCENES FROM HATRA Fig. 5: Funerary relief of a man with an eagle perched on his head from the depot of the Palmyra museum. Photo: Kim van Straten.