The " Figure" at Hatra and : Problems of Interpretation Author(s): Ted Kaizer Reviewed work(s): Source: Iraq, Vol. 62 (2000), pp. 219-232 Published by: British Institute for the Study of Iraq Stable URL: http://www.jstor.org/stable/4200491 . Accessed: 27/01/2012 17:08

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

British Institute for the Study of Iraq is collaborating with JSTOR to digitize, preserve and extend access to Iraq.

http://www.jstor.org 219

THE "HERACLES FIGURE" AT HATRA AND PALMYRA: PROBLEMS OF INTERPRETATION

By TED KAIZER*

Introduction In this paper I intend to use the Heraclesfigure, a primeexample of the applicationof Graeco- Roman imageryto the ever-varyingprocess of expressingNear Easternforms of religionin the Romanperiod, to illustrateand guidea briefdiscourse on the methodologicalproblems concerning the approachto a religiousworld which was more heterogeneousthan is sometimesthought.' Iconographicrepresentations of a male figurewith club and lion's skin as his main attributesare widespreadin the Near East, and indeed far beyond, in the Roman period.2In what follows I will concentrateon the place and functioningof this so-calledHeracles figure within the context of the religiouslife of Palmyraand Hatra, two desert cities which, each in their own distinctive way, present examples of a complex religioussystem in which differentelements coexisted and might have influencedeach other. Evidencefrom elsewherein the Near East that may contribute to our perceptionof the variety of values which a Heraclesfigure could embody for different groups of worshipperswill also be taken into account.3 With regardto the places on which I am focusing, I preferto referto the figurewith club and lion's skin not as "Heracles"but as the "Heraclesfigure", because neither at Palmyranor at Hatrais that figureever called by his Greekname in accompanyinginscriptions. What the evidence does reveal,though, is that the Heraclesfigure enjoyed great popularity(especially at Hatra) and was clearlyconceived of as deeplyrooted in the divineworld of both places.As we will see below, it is possible, althoughnot "proven",that both in Palmyraand in Hatra the Heraclesfigure was identifiedwith , a deity with certain chthonian aspects from the Mesopotamiandivine world.4Nevertheless, the identificationsare problematic,and one ought to attend to the names and epithetsactually given by worshippers. It should be stated at the outset that the nature of our sources makes them very difficultto interpret.The evidencefor religiouslife in the Near East in the Roman period mainly consists of sculpturaland epigraphicmaterial, and we know close to nothing with regardto mythologyand ritual life. In the recent past the Heraclesfigure has been the subject of a growing number of studieswhich are mainlypreoccupied with hypothesesconcerning his originand travelsthroughout the Near East and beyond, and with theoriesconcerning the variousinterpretations and assimila- tions which he underwent.5In what follows I will leave those problemsaside as much as possible. 2 *Brasenose College, Oxford. See e.g. J. Boardman, The of Classical Art in 1 Diffusion Earlier versions of this paper were given in September Antiquity (1994), ch. 4: "The East after Alexander the 1998, at conferences in Cardiff at the University of Wales Great", pp. 75-153. and, a few days later, in Baghdad. I am grateful to the 3It is worth mentioning separately the material from Iraqi National Commission for Education, Science and Seleucia, collected by W. van Ingen, Figurinesfrom Seleucia Culture for making it possible for me to attend the confer- on the Tigris (1939), pp. 106-9, and especially the famous ence in Baghdad. In addition, I owe many thanks to bronze statue of a Heracles figure which the Parthian ruler Professors Fergus Millar and Robert Parker, and to Dr Vologaeses IV took from Mesene, with a bilingual inscrip- Andreas Bendlin, for their comments on several drafts of tion in Greek and Parthian (the latter written in Aramaic the text and for being very generous with their time, and characters) that identifies the figure in Greek as Heracles, to Margherita Facella for providing me with references and in Parthian as Verethraghna (wrtrgn). See A. Invernizzi, helpful suggestions on numerous occasions. Abbreviations: "Heracles a Seleucie du Tigre" in Revue Archeologique ANRW H. Temporini and W. Haase (eds.), Aufstieg (1989), fascicule 1, pp. 65-113, with further references, and und Niedergang der romischen Welt (1972-). also H. Stierlin, Stadte in der Waste (1987), p. 177 Abb. H 1, 2, etc. Inscriptions from Hatra. Same numbering 158. On the text see especially F. A. Pennacchietti, adopted by the main collections: F. Vattioni, "L'iscrizione bilingue greco-partica dell'Eracle di Seleucia" Le iscrizioni di Hatra (1981); B. Aggoula, in 22 (1987), pp. 169-85. Inventaire des inscriptions hatreennes (1991); 4 On this deity see E. von Weiher, Der babylonische Gott K. Beyer, Die aramaischen Inschriften Nergal (1971), and W. G. Lambert, "Studies in Nergal" in aus , Hatra und dem ubrigen Bibliotheca Orientalis 30 (1973), pp. 355-63. See also S. Ostmesopotamien(1998). Dalley, "Near Eastern patron deities of mining and smelting PAT D. Hillers and E. Cussini, Palmyrene Aramaic in the Late Bronze and Early Iron Ages" in Report of the Texts (1996). Department of Antiquities, (1987), pp. 61-6. 5 RTP H. Ingholt, H. Seyrig and J. Starcky (eds.), On the possible typological and historical antecedents Recueil des tesseres de Palmyre (1955). of the Heracles figure in the more , see

IraqLXII (2000) 220 TED KAIZER

Ratherthan implyinghis Graeco-Romanness,or his Orientalbackground for that matter,I wish to approachthe Heraclesfigure at Palmyraand Hatra at face value, as we encounterhim in the first three centuriesAD. Since the evidenceis well known and has been studied thoroughlyby others, another overviewof the materialwould not, at present,help to furtherour knowledge. Instead,I wish to touch upon some generalproblems of approachfirst, and only then to see how far a study of the Palmyreneand HatreneHeracles figure, and of the contexts in which he was worshipped,can help to advanceour understandingof the natureof religiouslife in these places. There were of course many similaritiesbetween the religionsof the severaldifferent towns and local areas in the Near East, but above all the variousplaces were very differentfrom each other. Both Palmyraand Hatra - each in their own terms- had a culturewhich cannot be described unambiguously,and when one attempts to label the various facets of their respectivecultures either as "western"or as "eastern",those generalisationsturn out to be very problematic.6 While the archaeologicalremains of Palmyra,situated in the Syriansteppe, show a city which was largely Hellenised in its outward appearance,its distinctiveand still partly enigmaticart points to a wide rangeof influences,7and while both Greekand the local dialectof Aramaicwere used alongsideeach other as writtenlanguages inth the religionof Palmyra is usually describedas completelySemitic. The view expressedin the concludingremarks of the main work on Palmyrenereligion by Gawlikowskiis that an indigenoussubstratum was trans- formedby Babylonianand Canaaniteinfluences in pre-Hellenistictimes, followedby the penetra- tion of so-called Arab and Syrian deities by the first centuryBC. With the urban development from the early Roman empireonwards came theologicalsystems and astrologicaldoctrines which may have had their origin in HellenisticBabylonia.8 A numberof Graeco-Romandivine names and images also appear,but identificationof the Palmyrenedeity with the Greek one is usually perceivedas secondary,and believedto have also been regardedas such in Roman times. Hatra, the enigmaticcity in the desertof North Mesopotamiawhich probablyoriginated from a combinationof strategicand religiousfactors, gives a differentimpression from Palmyraat first glance,with the circularplan of its city walls (at least by the last stage)and the enormousvaulted structuresin its central temple complex. Its sculpturesreveal many resemblancesto those from Palmyra(see above), and the written language was a local dialect of Aramaic,with no Greek inscriptionsbeing found, and only three ones from the brief period of Roman occupation before the city was capturedby the Sasanians.Due to the state of researchat present,knowledge of Hatrenereligion can only be provisional,but the religiouslife of the city in whichthe Heracles figure enjoyed such popularityis believed, in accordancewith the epigraphicevidence, to have been almost completelySemitic.9

7 W. Burkert, Structure and History in and Like those from Hatra, the reliefs and from Ritual ( 1979), ch. IV "Heracles and the Master of Animals", Palmyra have been characterised as "Parthian art", a term pp. 78-98, and idem, "Oriental and Greek mythology: the used for the art of East and North Mesopotamia on meeting of parallels" in J. Bremmer (ed.), Interpretationsof the basis of resemblances in style, of which the consistent Greek Mythology (1987), pp. 10-40. On the various equi- frontality is the most important characteristic. On the vocal processes of assimilation see C. Bonnet, "Heracles en problematic use of this term and for a recent overview of Orient: interpr6tations et syncr6tismes" in idem and C. the scholarly debate, see H. J. W. Drijvers, "The Syrian Jourdain-Annequin (eds.), Heracles. D'une rive a l'autre de cult relief" in Visible Religion 7 (1990), pp. 69-82. See also la Mediterranee. Bilan et perspectives (1992), pp. 165-98. Millar, The Roman Near East, pp. 329-30. 6 For a survey of the history and culture of both cities 8M. Gawlikowski, "Les dieux de Palmyre" in ANR W II see H. J. W. Drijvers, "Hatra, Palmyra und Edessa. Die 18.4 (1990), pp. 2605-58, esp. 2652-3. 9 Stadte der syrisch-mesopotamischen Waste in politischer, While still awaiting the first full-scale study of Hatrene kulturgeschichtlicher und religionsgeschichtlicher Beleuch- religion, see H. J. W. Drijvers, "Monotheismus und tung" in ANRW II 8 (1977), pp. 799-863. In general see Polytheismus in der hatraischen Religion" in Studies in the F. Millar, The Roman Near East (1993), throughout, History of Religions, Suppl. Numen 31 (1975), pp. 240-9; and especially the epilogue, "East and West", pp. 489-532; idem, "Mithra at Hatra? Some remarks on the problem of L. van Rompay, "Palmyra, Emesa en Edessa: Semitische the Irano-Mesopotamian syncretism" in Acta Iranica 17 steden in het gehelleniseerde Nabije Oosten" in Phoenix 36 (1978), pp. 151-86; K. Dijkstra, "Aramese votiefteksten en (1990), pp. 73-84; M. Goodman, The Roman World, 44 votiefgeschenken uit Hatra" in Phoenix 40 (1994), BC-AD 180 (1997), pp. 242-50, "The Northern Levant and pp. 184-94; idem, Life and Loyalty. A Study in the Socio- Mesopotamia". See now also K. S. Freyberger, Diefriihkai- religious Culture of Syria and Mesopotamia in the Graeco- serzeitlichen Heiligtumer der Karawanenstationenim hellenis- Roman Period Based on Epigraphical Evidence (1995), ierten Osten (1998). pp. 171-244. THE "HERACLESFIGURE" AT HATRA AND PALMYRA 221

The problem of "Hellenism" and theformation of religious life inthe NearEast the in Romanperiod Generalisations usually turn out to be very problematic when applied to religious life in the Near East in the Roman period. This becomesvery clear when one tries to analyseand describe religiousligiouinres tfeatur oes in teraction between an indigenous oriental substratum and structures which were introduced from the Graeco-Roman world. The various aspecs of Greek culture were transmitted into the Near Eastern world in such a way as not only to become a medium by which local culture could find renewedexprenewed expression, but also to open up a completely new world of historical and cultural knowledge to he Near East. In neither way should Greek culture, "Hellenism", be seen automatically as opposed to the local traditions.10 Furthermore, if Hellenism really, to quote Bowersock, "provided the means for a more articulate and a more universally comprehensible expression of local traditions,"1" this should not so much be seen in terms of quality but rather as conditioned by continuously changing historical and cultural circumstances. It also depends from which side the matter is theviewed: for indigenous population it may have been, consciously or not, a vigorous new way to make their cultural traditions intelligible. But in his definition of , Sartre explains the phenomenon in terms of Greeks and Romans recognising their own gods in the appearance of the local deities of the East.12 The nature of our evidence makes the problem even worse, in that it represents three centuries of snapshots rather than revealing any functioning entity. The more or less sudden ending which came to the civilisations of Hatra and Palmyra, when these cities were captured by the Sasanians in the early 240s and by the Roman emperor Aurelian in the early 270s respectively, makes it of course understandableto view the cultural and other processesin these places as representing their culmination shortly before their disappearance.13 But it is not correct to view the religious life of those placesas a perennialsystem which, althoughstill open to new influences,nevertheless reached its final and conclusive stage in the three centuries from which our evidence dates, thus treating its development in the preceding formative periods as a cultural prelude to the new synthesis.This trap of making teleologicalassumptions is not new, but there cannot be enough warning against it. On the other hand, what we now call "the religion" of any place in the Near East in the first three centuries AD not only becomes an entity because we choose to study it as such, but also becausethe evidencedates from exactly that period. Whetherone ought to think of the local cultures of Palmyra and Hatra as mysteriously emerging as they are known to us, or whether one should rather understand the sudden appearance of their material cultures in terms of a so-called epigraphic and sculptural habit, are questions which cannot be answered. But any glance at any item which can be connected to the religious world of the Near East in the Roman period reveals the outcome at a particular moment of various historical, social and religious processes. That said, one should of course acknowledgethat the population of a Near Easterncity or region, in this case Palmyraor Hatra, could still experienceits religious life as a unified and integrated whole. To quote Stewart, "when viewed at a certain moment in time continuing processesdo take on the appearanceof discretestates and not only do social scientistsfrequently perceive social and cultural forms to be establishedand enduring, but the social actors they describeoften work with this premiseas well."'4 It is, admittedly,only naturalthat, in a world which is not only dynamic but also very traditional,many elements are recognisableover the whole of our period.Connected with this is the notion of authenticitywhich people could ascribe to their religious traditions. The question whether such traditions are historically"true" and original or recently invented is not of any importance. For example, the label araTpcaoS,"ancestral", is regularlyattached to Palmyrenedeities in Greek inscriptions,not so much to stress that the deity is literallyinherited from one's father,but ratherto asserta specialclaim on the deity who

10G. Bowersock, Hellenism in Late Antiquity (1990), p. 7; 130Or,for that matter, the destruction of Dura-Europos Millar, The Roman Near East, pp. 326 and 523. by the Sasanians in AD 256/257. It should be noted that in ' Bowersock, Hellenism, p. 9. the case of Palmyra the end of its local civilisation was less 12M. Sartre, L'Orientromain (1991), p. 491. He continues sudden or immediately definitive; see Millar, The Roman by calling it, p. 496, nothing more than a superficial veneer, Near East, pp. 335-6. applied only to give a Graeco-Roman aspect to the local 14C. Stewart, "Relocating syncretism in social science deities, and argues that their "nature" or cult celebrations discourse" in G. Aijmer (ed.), Syncretism and the Commerce are not really affected. of Symbols (1995), p. 30. 222 TED KAIZER in this way becomes tutelary. 5 The fact that in the religious life of a city such as Palmyra or Hatra variousstrata can be tracedwhich all point to differentspheres of influencedoes not affect the possibilityof claims to authentictraditions either.16 The problemgets even more complicated when one takes into consideration the fact that "new" traditions, once introduced alongside the "native traditions", in a way become "native" as well: elements can cease to be regarded as importedwhile simultaneouslybeing subjectto a continuousperception of importation.Equally, elementswhich are among the most traditionalones can still functionas "comingfrom abroad". Finally,with regardto the stabilisingeffects of writtentexts on differentindigenous traditions, as proposed by Stewart,it has to be concludedthat we have no evidencefor the role of liturgical directions,ritual manuals,mythological narratives and other sacredtexts in creatingany kind of "standard"form of religionin the Near East in the Roman period.17

The hiddennature of Near Eastern deities: identificationof the Heraclesfigure with Nergal? Towards the middle of the second century AD a Graeco-Roman cella was built over the existing chapel of the goddess Allat at Palmyra, and a marble statue representinga classical Parthenos was imported and up in the new temple.'8 Since the way in which Allat was represented in her sanctuary had by then developed from her being seated on a throne in between two lions to standing with helmet, spear, round shield and aegis, it could be argued that this particularAthena statue was called Allat by the Aramaic-speakingworshippers in her Palmyrene temple. But if the same statue had been found in an apparentlymore HellenisedNear Eastern context,e.g. in one of the cities of the so-calledDecapolis, the answermight have been different.'9 "As in other cases it is futile to ask what such a deity really 'was'," writes Millar,20and that should also serve as a warningagainst hasty and improperidentifications of the Heraclesfigure in Palmyraand Hatra,where he is neveridentified unambiguously by accompanyinginscriptions.2

5 See R. MacMullen, Paganism in the hamana au naos. Le temple palmyr6nien hellenise" in Topoi (1981), pp. 3-4, on the possible uses of 7rarpq,oS (and 7 (1997), p. 841. For the statue, now in the Museum of patrius). The word can both be attached to a deity "for Palmyra, see K. Tanabe (ed.), Sculptures of Palmyra I, remembrance's sake and from a sense of being a stranger Memoirs of the Ancient Orient Museum 1 (1986), Pls. abroad" and to "him who watches over us here at home". 161-2. An equivalent in Palmyrene Aramaic can be found in PAT '9 Example given by E. Will, "Les aspect de l'int6gration 0324, where Shamash is called "the god of the house of des divinites orientales dans la civilisation gr6co-romaine: their fathers" ('lh byt 'bwhn). In the bilingual inscription langage conventionnel et langage clair" in L. Kahil and PAT 0273, the phrase -rp4oLL O6ots is rendered into C. Aug6 (eds.), Mythologie greco-romaine, mythologies per- Palmyrene by 'Ihy 'tby', "the good gods". ipheriques. Etudes d'iconographie (1981), pp. 159-60. See 16See also the introduction in C. Stewart and R. Shaw also H. J. W. Drijvers, "De matre inter leones sedente. (eds.), Syncretism/Anti-Syncretism. The Politics of Religious Iconography and character of the Arab goddess Allft" in Synthesis (1994), p. 7: "both putatively pure and putatively M. B. de Boer and T. A. Edridge (eds.), Hommages d syncretic traditions can be 'authentic' if people claim that Maarten J. VermaserenI (1979), pp. 331-51, and J. Starcky, these traditions are unique, and uniquely their (historical) "Allath, Athena et la deesse syrienne" in Kahil and Aug6 possession." Compare the introduction in G. Aijmer (ed.), (eds.), Mythologie greco-romaine,mythologies peripheriques, Syncretism and the Commerce of Symbols (1995), p. 12, pp. 119-30. 20 where the author describes the cultural process of what he Millar, The Roman Near East, p. 326. calls "generating faked traditions out of a constantly ongo- 21 It should be stressed that even the identification of a ing bricolage". For further references to recent approaches Heracles figure by an accompanying inscription still does to the so-called "invention of tradition", see Stewart and not mean that we know what the deity really "was" for the Shaw, op. cit., p. 1. local worshippers. One of the famous dexi6sis reliefs that 17Stewart, "Relocating syncretism", p. 31, where the Antiochus I erected throughout the kingdom of Commagene author describes non-literate religions, as opposed to reli- in the first century BC labels the naked figure with club and gions of the Book, as "basically more receptive to the lion's skin who shakes hands with the king as "Artagnes- incorporation of diverse, exogenous deities into its repertoire Heracles", while "Ares" is joined to this name in another of worship". Recently, the hypothesis has been put forward inscription. But the remnants of this royal dynastic cult that since it has been proved that cuneiform continued into proceed from the religious and political programme of the first century AD, temples in which Babylonian deities Antiochus himself and do not tell us anything about the were worshipped preserved their liturgy and regulations "in local culture of the area. See J. Wagner, "Dynastie und cuneiform script as it had been for millennia", see Herrscherkult in Kommagene" in Istanbuler Mitteilungen M. J. Geller, "The last wedge" in Zeitschriftfir Assyriologie 33 (1983), p. 186, and Millar, The Roman Near East, und VorderasiatischeArchdologie 87 (1997), especially p. 47. pp. 452-4. See now also U. Huttner, Die politische Rolle Whether this can or should be connected to the der Heraklesgestalt im griechischen Herrschertum (1997), "Babylonian" cults which are attested at Palmyra and pp. 198-210, who stresses that Antiochus's interest in the Hatra remains more doubtful. Heracles figure resulted from the influence of Hellenistic 8 See M. Gawlikowski, "Le temple d'Allat f Palmyre" ruler ideology on the royal house of Commagene from in Revue Archeologique (1977), p. 269, and idem, "Du Alexander onwards. THE "HERACLESFIGURE" AT HATRA AND PALMYRA 223

In the each deity has its own well-defined sphere of influence and it may be tempting to think that the Near Eastern deity to whom the Graeco-Roman one is assimilated always performedsimilar functions. But in most cases it remainsuncertain, when iconographic featuresof a Classicaldeity are taken over, whetherhis or her powers are equally transmitted, and if so, whether they remain unchanged or become reinterpreted.22 As I have brieflystated above, it is generallyaccepted that both in Palmyraand in Hatra the Greek Heracleswas identifiedwith the MesopotamianNergal.23 It ought to be stressed,though, that in both places the identificationis not only uncertain,in the sense of "not entirelyproven", but also not complete, in as far as it took place at all. We will go through some of the other evidence below. Here I only mention the material usually referred to in this context. The Palmyrene identification of Heracles with Nergal, as stressed by Seyrig, rests mainly on two tesserae, small tokens which are usually interpretedas entrancetickets to religiousmeals.24 The first one (RTP 227) has a club, the main attributeof the Graeco-RomanHeracles, on the obverse, while the dialectnamename of Nergal the is(nrg)is written of ininon local the reverse. The obverseof the second tessera (RTP 231) shows the same club, while on the reversea figure with a double axe is depicted. Although Seyrig has argued, with regard to RTP 227, that on none of the obverse connected with the attributeof another on the reverse,there are plenty of exampleswhere either the names or the images of two or three deities are impressedon the two sides of the tessera. In any case, even if there seems to have been some connection between the two kinds of figures, it remains doubtful whether representations of the god with double axe ought to be taken into account as well when discussingwhat we call the "Heraclesfigure". A group of tesserae (RTP 233-6) on which the deity with double axe, eitherholding a lion or with the animalleaping at him, is depictedtogether with an unspecifiedgoddess, is describedby Seyrigas representingthe Graeco-Orientalimage of Nergal and his female companion.25But whetherthe cult statue of Nergal at Palmyratook the form of the deity with double axe, while the Graeco-RomanHeracles figure could equally be applied to represent him on certain reliefs and tesserae, as Seyrig argued, must remain a hypothesis.26 Both the Heraclesfigure with club and lion's skin (eithernaked or in Hatrenedress) and a deity with a double axe seem also to have been very popularat Hatra. But in contrastto Palmyra,the evidencefrom Hatra consists of a great numberof statues,ranging from man-sizedto minuscule, and found in the various smallershrines spread throughout the city.27The Hatreneidentification of Heracleswith Nergal is believedto have been provedby an inscriptionwhich reads nrgwl klb' ("Nergal the dog"?) underneatha statue base which shows bare feet and the lower part of a club.28The identificationof the two deities, although not completely certain, may have been

22 To make matters even more complicated, one ought to 26Seyrig, "Heracles-Nergal", p. 67. A. Bounni, on the take into account that the sphere of influence of a Classical other hand, seems to distinguish between the Greek and deity was not actually static. Graeco-Roman gods often the oriental god, arguing that the naked and bearded figure had more than one "field of interest", which could gradually must be Heracles, while the figure holding a double axe alter its direction and should be seen within the context of represents Nergal, even when he carries an attribute of a wider divine world. See M. Beard, J. North and S. Price, Heracles in addition. See "Iconographie d'Heracles en Religions of 1: A History (1998), p. 16. Syrie" in L. Kahil, C. Aug6 and P. Linant de Bellefonds 23 Bonnet, "Heracles en Orient", pp. 180-3. (eds.), Iconographie classique et identites regionales (1986), 24H. Seyrig, "Heracles-Nergal" in Syria 24 (1944/5), p. 385. Note that Bounni stresses military presence as being pp. 62-80 (=idem, Antiquites syriennes IV (1953), decisive for Heracles' popularity throughout the Near pp. 1- 19). East, p. 387. 25Ibid., pp. 66-7, Nos. 12-15. It is worth referring here 27 See the evidence listed by W. I. al-Salihi, "The to a relief found in the so-called Camp of , Sculptures of Divinities from Hatra" (PhD Princeton 1969, showing a figure in Palmyrene dress holding a club in his available on microfilm from UMI), pp. 63-100, and by S. right hand and another object in his left hand, standing B. Downey, The Heracles , The Excavations at next to a small altar and an even smaller dog; see Drijvers, Dura-Europos, Final Reports III, Part 1, Fascicle 1 (1969), The Religion of Palmyra, PI. XV, and Tanabe, Sculptures of pp. 83-96. See also D. Homes-Fredericq, Hatra et ses Palmyra, PI. 134. Following the interpretation of M. sculpturesparthes (1963). Gawlikowski, "Un nouveau type d'Heracles a Palmyre" in 28 See W. I. al-Salihi, "-Nergal at Hatra" in Iraq Etudes et Travaux3 (1966), pp. 141-9, most scholars accept 33 (1971), P1. 34d and p. 114, where he states that the that the relief is the result of the juxtaposition of Greek various pieces of evidence "clearly illustrate the assimilation and Babylonian traditions, and thus represents Heracles- of Hercules to Nergal" in Shrine X. The inscription is H71, Nergal. Unfortunately, the accompanying inscription (PAT see B. Aggoula, Inventairedes inscriptionshatreennes ( 1991), 1933) is damaged and does not help us any further. and now also K. Beyer, Die aramaischen Inschriften aus 224 TED KAIZER facilitatedby a resemblancein the sense that both could be associatedwith a lion. It has also been arguedthat "the main reasonfor the fusion of Heraclesand Nergalshould be tracedin their chthoniannature," from which it has been concludedthat the so-called "Cerberusrelief" is an "artisticrapprochement" between the Greek and the Babyloniangod and also "expressesan actual fusion of religiousideas".29 Unfortunately, since the figuresdepicted on the relief are not identified,it is pure speculation,encapsulated in a circularargument, that the three-headeddog on a leash is ,and any possible interpretationof the evocative relief must remain a hypothesis.30Since the writtenlanguage of Hatra is a local dialectof Aramaic,it is not surprising that, as at Palmyra, the Greek name Heracles is not attested.31But in one of the few Latin inscriptionsfound at Hatra, on the base of anotherstatue of a Heraclesfigure, a Roman officer who was encampednear the city around AD 240 seems to identify the Genius of his cohort, to which the statue was dedicated,as Hercules sanctus.32 Whetherthe epithet sanctus points to an oriental origin of the deity under concern is another matter.33The popularitywhich Hercules enjoyedamong Roman soldiersdoes not make orientalinfluences in this case necessary.But it is of course possible that the iconographicaldetails of the statue, especiallythe amulet, have been influencedby local artisticmotifs.

Local religions in the Near East: notions of continuousrenegotiation In an importantarticle on the relation between Hellenismand the divine world of Palmyra, Gawlikowskidescribes what he calls an amalgamof religiousnotions, correspondingto a civilis- ation that is orientalin characterbut "marqueedu sceau indelebilede l'hellenisme".It is argued that the Greek or Graeco-Romanappearance of the sanctuariesonly disguisesthe fact that the deities worshippedinside are oriental,and that a "doublenature" can only be ascribedto a few gods, such as Nebo- or Nergal-Heracles,originating in HellenisedBabylonia. Generally speaking,Gawlikowski calls the adoption of Hellenisticmodels relativelysuperficial, and charac- tenrisesit as havingbeen mostly of a decorativecharacter.34 But concentratingon the little evidence we have from the pre-Roman,, Gawlikowskistresses that what we encounter thereis not the "hellenisationd'une entite deja formee"but a "syncretismeou l'apporthellenique participeen solution."35He then argues that an approachwhich isolates the various elements which constitutethe "religionof Palmyra"does not take sufficientaccount of the melange,since "tout syncretismeresulte precisement en une entite homogene."This would be especiallyclear in the case of Palmyra,with the substantialreduction of its indigenoussubstratum and the develop- ment of the Palmyrenecivilisation "tous azimuts".36Gawlikowski's remarks are very thought- provoking and deserve some extended discussion. In what follows I will use his authoritative study, despite some minor criticism,as a base for an interpretationof local Near Easternreligion

Assur, Hatra und dem ubrigen Ostmesopotamien(1998). See transliteration:Nemesis (nmsys ) is mentionedin an inscrip- also Bonnet,"Heracles en Orient",p. 182. tion fromWadi 'Arafa in the Palmyrenefrom AD 153(PAT 29Thus V. Christides,"Heracles-Nergal in Hatra" in 1568) and in a bilingualone from Dura-Europosfrom AD Berytus30 (1982), pp. 105-15, quotationson pp. 106-7. 244 (PAT1078). See Gawlikowski,"Les dieux de Palmyre", The "Cerberusrelief", found in the first shrineexcavated p. 2642. It is worth mentioninghere that the name of in Hatraoutside the centraltemple complex, owes its name Heracleshimself appears in Syriactransliteration (hrqls) in to the three-headeddog on a leash next to the main figure the enigmatic Oration of Meliton the Philosopher, see W. on the relief.See Stierlin,Stddte in der Waste,p. 203 Abb. Cureton (ed.), Spicilegium Syriacum (1855), p. 24, 11. 17 188. The main figureis a beardeddeity with horns,topped and 26. by an eagle, surroundedby and scorpions, and 32 For the inscription see Annee Epigraphique 1958, holding an axe in his right hand. Next to him stands a No. 240. Publishedby D. Oates, "A note on three Latin so-called "divine standard", and in the backgrounda inscriptionsfrom Hatra"in Sumer11 (1955), No. 81. For goddess is seated betweentwo lions. See also Freyberger, the statue, see Downey, The Heracles Sculpture, PI. XIX, 2. Die friihkaiserzeitlichen Heiligtumer, p. 102. 33CompareBonnet, "Heracles en Orient",p. 182. 30Surprisingly, Christides seems to have missed the 34M. Gawlikowski, "L'Hellenismeet les dieux de inscriptionon which al-Salihi(see above) based his identi- Palmyre" in 0 EAAHNIZMOE ETHN ANATOAH, ficationof the two deitiesat Hatra,which makes even more Internationalmeeting of historyand archaeology, speculativehis own discussionof "how his (= Heracles') 6-9 November 1986 (1991), especially pp. 246-7, 251 and chthoniannature was spreadand understoodin the remote the concludingremarks on p. 256. town of Hatra" 115). 35Ibid., p. 245. (p. 36 31With regard to PalmyreneAramaic it ought to be Ibid., p. 246. stressedthat Greek divine namescould appearin Semitic THE "HERACLESFIGURE" AT HATRA AND PALMYRA 225 in the Roman period which emphasisesthe notion of continuous renegotiationof old and new elements more than is generally done. The pre-Roman evidence from Palmyra, usually referred to by the term "archaic", mainly comes from the temple of and seems to make clear that before the beginningof our era his cult was accompaniedby those of variousother deities. Some of them probablyoriginated in the Babylonian religious world, while others have a more enigmatic background or might be labelled as indigenous.37 Amongst the oldest Palmyrene sculptures is a small relief which depicts a naked Heracles figure with club and lion's skin alongside a radiate goddess and two radiate gods, one of them with a crescent on top.38 While the other three deities all wear common Palmyrene dress, the nuditythe of Heracles figure is exceptional at Palmyra. If anything, the relief seems to show that, no matter when orwh from ere he arrived at the oasis, the Heracles figure had been fully integrated into the divine world of Palmyra by the time that the evidence became more abundant. His nudity might seem different to us when compared with the way in which other deities were represented, and might also have struck the ancients asadivergent, but that does not mean that he cannot have been regarded as as much part of the religious life as other gods and goddesses were.39 I have already argued that religious life in the Nearin the East s three centuries from which our evidence dates should not be regarded as an entity that formed the final stage of an evolutionary process. The notion of continuing change and development should not be overlooked, and it is only naturalthatural that the various elements which may alreadyahave coexisted for ages, or were introducedmore recently,were constantly renegotiated and kept on influencingeach other. Indeed, there is no reasonwhy this should be automaticallyseen as "progressive"or "logical".If we thus ask ourselves how far cultures of multiple origin represent a cultural , and how far they were being continuously reintegrated, we are probably asking questions to which there are no unambiguousanswers. With every new question and from every new viewpointthe scene seems to change as easily and quickly as a kaleidoscope. Generalisations are difficult to avoid, but either give no explanation at all or do not do justice to the complexity of the material. Efforts have been made recently to reinstate the most notorious term of them all, "syncretism", long avoided because of its implication of a melting pot.40 Although all religions are "syncretic" in the sense that the coming together of at least two different sources is involved, Stewart has argued that the term can still be valid in referringto that particularaspect of culturewhich stressesthe borrowingand reinterpretationof divergentelements as part of its very nature.41But Drijvers'argument, with regardto the multi-interpretablecultures of the Near East in the Roman period, that "a culture assimilatesother elementsto its own traditionand pattern,but does not mingleor mix everything

37 Most of the "archaic"material comes from a founda- buildingstructures to Herta, Nanai and Relef, the gods tion wall in the court of the templeof Bel whichis ascribed (Ihrt'wlnnywlr?p 'lhy'), see PAT 2766, after stating that "ein to the sanctuarythat precededthe presentone. For a list aus dem Osten eingewanderterGott ist auch Nergal, ein of the evidencesee M. Gawlikowski,Le templepalmyrenien Unterweltsgott,der mit dem kanaanaischenGott Reschef (1973), pp. 56-60. Some of the materialis fragmentary, identifiziertwurde. In Reliefstragt er Zilge des ." undatedand only ascribedto the so-calledHellenistic temple Althoughit is indeedpossible to go one step furtherand to becauseit is believedto be "archaic".It is worthmentioning arguethat the Resef from the inscriptionfrom 6 BC should that two Greekinscriptions, unfortunately too fragmentary be connected with the Heraclesfigure on the Hellenistic to give any information,are amongthe materialwhich can relief,this must remaina hypothesis. certainlybe ascribedto the older structure.The following 39Which is a good reasonwhy a referenceto the reliefas necessarilydeals with Palmyra only, for we haveno evidence representing"H6racles et des divinit6sde Palmyre"in the of any culticlife in Hatrain the last centuriesBC. In a way, above-mentionedstudy of Hellenism by Gawlikowski, the suddenappearance of Hatrenecivilisation as it presents p. 248, Fig. 2, ought to be avoided.The iconographyof the itself to us raises problemseven more serious than those Heraclesfigure may be Greek,but the style of all the deities concerningPalmyra. that are depictedis still very local. Compare 38 Boardman, Drijvers,The Religionof Palmyra,PI. XIV and p. 12. The Diffusion of Classical Art, p. 317, who writes that Drijversidentifies the other three gods as ,Aglibol althoughthe use of nudity"could have made sense only to and Yarhibol,which is possiblebut not certain.It is possible a Greek",the nakednessof Heracleswas acceptedthrough- that originallymore deities were depicted.Note that the out the East. figure on the right is very similar to the one which is 40 M. Pye, Syncretism versusSynthesis, BritishAssociation depictedon a stele found at the templeof Nebu, see ibid., for the Study of Religions, OccasionalPapers 8 (1993); PI. L, 2 and p. 19. W. Pratscher,"Das Pantheon von Stewart and Shaw (eds.), Syncretism/Anti-Syncretism; Palmyra"in P. W. Haider,M. Hutterand S. Kreuzer(eds.), Aijmer (ed.), Syncretism and the Commerce Symbols. 41 of Religionsgeschichte Syriens (1996), p. 224, refers to an Stewart, "RelocatingSyncretism", pp. 26-7. See also inscriptionfrom the templeof Bel which is dated to 6 BC Pye, Syncretism versus Synthesis, p. 7. and in whichthe priestsof the goddessHerta offered various 226 TED KAIZER together," may be more satisfactory, for it seems to take into account the fact that once so-called syncretistic forms are established, they cease to be syncretistic and become original.42Gawlikowski, therefore,is right to dismissthe idea that Hellenismat Palmyrafunctioned as nothingmore than a veneer.43That would not only imply too static a situation but also ignore the role which Hellenism was due to play from its first appearance onwards as an original ingredient. Of course, Gawlikowski is right when he argues that an approach which isolates the various elements that constitute the "religion of Palmyra" does not take sufficient account of the melange. But I do not agree when he further states that all syncretism results in precisely a homogeneous entity.4 In a recent sociological study Aijmer has argued against this view of homogeneous cultural traditions, stating that "there is no a priori reason to assume that customs and conventions should fit into one and the same logical format. Rather, our present knowledge points in the direction that a cultural tradition embraces a number of 'modalities',culturally constructed universes, each with its own given presuppositions."45 With regard to the religious world of Palmyra or Hatra one could argue, to make matters even more complex, that the different groups of people who came from different backgrounds could stillshare similar conceptions of how to visualise the divine world.46 It is doubtful, though, how far the notions of "dominance" or "competition", e.g. as used by sociologists in the sense of applying symbolism to the integration of various social groups under the acceptance of one source of dominance,47 are applicable to our case. In ancient polytheism neither common concepts of visualisation nor a multiplicity of forms automatically imply the creation of potential conflicts.48 And although we come across a variety of social groups who may have arrived at either Palmyra or Hatra at various points in history, the evidence does not seem to reveal any degree of antagonism on the part of one (religious) group towards another. What it does reveal is, if anything, significant cases of religious worlds which were distinctively local. And, needless to say, a local religion is made up of various aspects which do not necessarilyhave to be local. Admittedly,it ought to be said that the model which I have been sketchingis very dynamic, while the limited nature of the evidence does not help to express those notions of continuous renegotiation we are looking for. Besides, the religious world of the Roman Near East may have been more dependent on tradition than the model seems to take into account. But, although the model cannot be "proven"with our evidence(and necessarilywill neverbe proven),it might still be put forwardagainst (or ratherin additionto) the currenttheory of an accumulationof religious layers, of which the Hellenisticone is believed to have had no real impact. I do not deny, of course, that differentspheres of influencecan be traced, and I support the idea that the divine world of a place like Palmyrain the Roman period still reveals two main strata, which may originallyhave been distinguishedby ethnicity.But in order to explainthe developingnature of a culture,and the way in which developmentand renegotiationwith regardto a systemof beliefs wereperceived by groupsof worshippers,the model of an "additiveextension of an open system", as proposedby Bendlinwith regardto religiouscommunication in the Roman empire,might be more helpful.49

42... and become as such part of the culture's tradition 43Gawlikowski, "L'Hellenisme", p. 246. 44 and pattern to which further elements are assimilated. For In an early article though, Gawlikowski referred to the the quotation see H. J. W. Drijvers, Cults and Beliefs at "heterogenese de la culture, et surtout de la religion Edessa (1980), p. 17. See also Aijmer's introduction in idem palmyrenienne": see "Un nouveau type d'Heraclbsa (ed.), Syncretism, p. 12. I do not believe that the distinction Palmyre", p. 149. between "syncretism" and "synthesis", as made by Pye, 45Aijmer'sintroduction in idem, Syncretism, p. 5. Syncretism versus Synthesis, especially on p. 6, is very helpful 46Or in Aijmer's words again, ibid., pp. 5-6: "People do in this matter. Pye distinguishes "synthesis", which "implies not live in one society only, but simultaneously in several that, out of multiple possibilities, a new conclusion has societies, which exist in parallel and apart." 47 been reached," from other resolutions such as "assimila- See ibid., pp. 3 and 6. Compare the introduction in tion" (here in the sense of "the outright dominance of one Stewart and Shaw (eds.), Syncretism/Anti-Syncretism, strand of meaning by another") or "dissolution". Those pp. 19-22. 48 resolutions supposedly come out of the dynamically open Contra A. Feldtkeller, "Synkretismus und Pluralismus and syncretistic situation in which "the potential claims of am Beispiel von Palmyra" in Zeitschriftfur Religions- und the constitutive elements are still alive." Especially the idea Geistesgeschichte48 (1996), especially pp. 27-8, who makes of "synthesis" as the "coherent mixture" which "represents an unnecessary plea for reconsiderationof the old theory of the conclusion to a process which is thereby completed" is rivalry between the two so-called supreme gods at Palmyra. much too artificial and static to be applied to any reli- 49 A. Bendlin, "Peripheral centres- central peripheries: gious world. religious communication in the Roman empire" in H. THE "HERACLESFIGURE" AT HATRA AND PALMYRA 227

Graeco-Romanart as mere "decoration"in the East? In additionto the sculpturesproduced in the differentlocal styles of the Near East, a number of divine and heroic representationsin Greekor Graeco-Romanstyle appearin the variouscities of the area. For example, two sculpturedblocks from Palmyra, one representingthe Heracles figureand the other a Hermesfigure holding a purseand probablya caduceus,are very divergent in style from the typicallyPalmyrene reliefs.50 The so-called"pure" Greek forms in the East are often playeddown by explainingthem as having been of a decorativecharacter.5' But one should not, I believe, try to make sense of the complex religiousworld of the region by restrictingas many ambiguitiesas possible.Two found in houses east of the templeof Bel at Palmyra are good examplesof those approachestowards obvious ambiguities.The mosaics, which have been judged contemporaryon stylisticgrounds, are both in Classicalstyle, and the main figures are labelled by the use of mosaic inscriptionsin Greek.52The first one represents,among other figures, Cassiopeia and . But in contrast with the "standard"Greek myth, in which Cassiopeia'sbragging about her looks arousedPoseidon's anger and in which she lost the beauty contest against the Nereids, the Cassiopeia from the Palmyrenemosaic is triumphant,with Poseidon in the role of judge in the contest.53Although it was originallyattributed to the years between AD 245 and 273, it was later believedto have originatedin the time that Palmyrahad become part of the chain of the Strata Diocletiana.54But the fact that two other depictionsof this Near Easternversion of the myth, discoveredat Apamea in Syria and on Cyprus,both date from a later period is no guaranteeof a later date for the Palmyreneone.55 The second mosaic from Palmyra,believed to be contemporarywith the one just described,shows the discoveryof Achillesamong the daughtersof Lycomedeson Scyrusby Odysseusand Diomedes.This Classical themewas also reproducedon a muralpainting in the so-called"hypogeum of the threebrothers", a tomb which was founded in the first half of the second century AD.56In the same way the downplayingof a sculptureof Dionysos in Greek style becomesless acceptablewhen considering anotherthird-century fresco in a hypogeumon whichthe samedeity is represented"au banquet".57 Whateverform belief in an afterlife,if therewas any, may have taken at Palmyra,we should not expect the decorationin the necropolisto have been deprivedof all significance.58When Graeco- Roman materialin the Middle East is classifiedas "decorative",it still means that with the new "iconographicrepertoire" other religiousexperiences originally represented by this repertoirein the Graeco-Romanworld could also enter the Near Easterndivine world.59 That we do not really

Cancik and J. Rilpke (eds.), Romische Reichsreligion und PI. II-III and especially XIII, and Colledge, The Art of Provinzialreligion (1997), pp. 35-68, especially 52-4. In Palmyra, PI. 115. The inscription which gives the date of contrast I remain sceptical about any deliberate foundation of the tomb is PAT 0066 (AD 133). 57 "Romanisation" as recently argued for by M. A. R. For the statue see Gawlikowski, "L'Hellenisme", p. 251 Colledge, "Roman influence in the art of Palmyra" in n. 11, and idem, "Un banquet dionysiaque a Palmyre" in AAAS 42 (1996), pp. 363-70. Ltudes et Travaux 15 (1990), pp. 157-61. For the fresco, 5°These blocks were reused in the medieval citadel into now inaccessible, see J. Starcky and M. Gawlikowski, which the propylaea of the temple of Bel were converted. Palmyre (1985), Fig. 13, and Colledge, The Art of Palmyra, See Seyrig, "Heracles-Nergal", PI. IV and p. 75. He sug- P1. 118. gested that the two reliefs were originally used in a gymna- 58 Just as an example, it is worth mentioning the unpub- sion or a palaistra. lished graffiti of an altar and scorpions (the animal displayed 51E.g. Gawlikowski, "L'Hellenisme", p. 251. Compare in a prominent position on some reliefs from both Palmyra M. Colledge, The Art of Palmyra (1976), pp. 104-5 and 242. and Hatra) in tomb tower No. 12, and the painting of a 52Both were published by H. Stem, Les mosalques des satyr head in tomb tower No. 19. The graffiti are mentioned maisons d'Achille et de Cassiopee a Palmyre (1977), with by M. Gawlikowski, Monuments funeraires de Palmyre detailed photographs. (1970), 77 n. 41. The is mentioned K. 53 p. painting by See J. Ch. Balty, "Une version orientale meconnue du Michalowski, Palmyre 1961 (1963), p. 198, with Fig. 242, mythe de Cassiop6e" in Kahil and Auge (eds.), Mythologie and sketched by Colledge, The Art of Palmyra, p. 84 Fig. 47. greco-romaine, mythologies peripheriques,pp. 95-106. I am grateful to Agnes Henning for drawing my attention S4 Stem, Les mosaiques, p. 42. Contra J. Balty, Mosalques to this material and for providing me with some excellent antiques de Syrie (1977), pp. 32-4. photographs. See also, K. "R6mische 5 briefly, Parlasca, For the mosaic from Apamea see Balty, Mosafqes, Elemente in der Grabkunst Palmyras" in F. Zayadine, Petra pp. 82-7, and J. Ch. Balty, Guide d'Apamee (1981), and the Caravan Cities (1990), pp. 191-6, and on the pp. 212-15. For the mosaic from Cyprus see Bowersock, question of meaning in general see H. J. W. Drijvers, "After Hellenism, pp. 50-1 and PI. 3. life and funerary symbolism in Palmyrene religion", in U. 56 For the mosaic see Stem, Les mosafques, pp. 5-26, and Bianchi and M. J. Vermaseren (eds.), La soteriologia dei Balty, Mosaiques, pp. 30-1. For the fresco see C. H. culti orientali nell'Impero Romano (1982), pp. 709-33. Kraeling, "Color photographs of the paintings in the tomb 59See also Beard, North and Price, Religions of Rome of the three brothers at Palmyra" in AAS 11-12 (1961/2), 1, p. 164. 228 TED KAIZER know how far these experiencesbecame newly interpretedin the East, how far they maintained their Classicalcharacter, and how they were receivedby the various sections of the population, goes without saying. The questionwhether there ever was a so-calledoriginal core in Palmyrenereligion cannot be pursuedhere, but if anythingwas recognisablein the Roman period to which one would now refer as the "indigenoussubstratum", it can only be known to us in a context which must have been at least partiallydifferent from the world in which it originallydeveloped or from wherever it emerged.When and how the Heraclesfigure made its entranceto the divine worlds of the variousplaces in the Near East is of course not detectable.The exampleof Tyre, situatedon the southernPhoenician coast, shows that identificationof Heracleswith an indigenousouslddeity go back to at least the mid-fifthcentury BC, but in generalone could say that a new dimension was added to the appearancesof the Greekhero in the East afterthe campaignsof Alexanderthe Great.6 As we have seen above, by theetime from whichthe our sourcesdate the Heraclesfigure had become firmly rooted in the religious worlds of the various localities in the Near East. And althoughthe three centuriesof snapshotswith which our evidenceprovides us necessarilyleave a lot of the questionswhich we would like to ask open, it is still possible to get an impressionof the possible ways in which the Heraclesfigure could functionand have its own place within the local divine worlds. In this regardit will be necessaryto approachthe materialin various and sometimesseemingly contradictory ways.

The Heracles figure on Near Eastern coins Consideringthe role that coinage may have fulfilled in familiarisingthe Near East with a Graeco-Romanstandard form of representingthe Heracles figure, it is surprisinghow little evidencethere is. The abundanceof the collection of coins from Tyre which depict the laureate head of Heracles-,the lion's skin knotted around his neck, is exceptionalin this regard and ought to be seen in the context of the particularrole which the cities on the Phoeniciancoast played within the Roman world.61But the coins issued in Palmyra62or Hatra,63or found at Dura-Europos,64do not help us any further,and it is only the coinageof the cities of the so-called Decapolis, in particularGadara and Philadelphia,that providesus with some examplesof ste- reotypicalGraeco-Roman representations of the Heraclesfigure: seated on a rock or restingon his club (even holding the golden applesof the Hesperidesin one case), fightingthe Nemeanlion, or his bust with the lion's skin knottedaround his neck.65But here as well the local contextcould

60Bonnet, "Heracles en Orient", pp. 167-72. See also di Vaballato" in Numismatica (Rome) n.s. 2 (1961, Boardman, The Diffusion of Classical Art, p. 321. On Tyre pp. 123-8. Despite the appearance of various other local see C. Bonnet, Melqart. Cultes et mythes de l'Heracles tyrien deities on the more typically Palmyrene coins, there is no en Mediterranee, Studia VIII (1988). On the evidence yet for the depiction of the Heracles figure on any particularities of the process of hellenisation in Phoenicia of them. See A. Krsyzanowska, "Le monnayage de see F. Millar, "The Phoenician cities: a case-study of Palmyre" in Proceedings of the 9th International Congress hellenisation" in Proceedings of the Cambridge Philological of Numismatics, Berne, September 1979 I (1982), Society 209, n.s. 29 (1983), pp. 55-71. pp. 445-57, and, in general, W. Szaivert, "Die Manzen von 61A. Burnett, M. Amandry and Pere P. Ripolles, Roman Palmyra" in E. Ruprechtsberger(ed.), Palmyra. Geschichte, Provincial Coinage I (1992), Nos. 4619-706 (silver) and Kunst und Kulturder syrischen Oasenstadt(1987), pp. 244-8, 4707-19 (bronze). See also above, n. 60, and on the region with further references. in general, Millar, The Roman Near East, pp. 264-95. 63J. Walker, "The coins of Hatra" in The Numismatic 62 There are a few coins on which Zenobia's son Wahballat Chronicle, 6th series, 18 (1958), pp. 167-72; J. J. Slocum, (or Vaballathus) is associated with a Heracles figure. See "Another look at the coins of Hatra" in The American G. Mazzini, Monete imperiali romane IV (1957), p. 171 with Numismatic Society. Museum Notes 22 (1977), pp. 37-47. 64 Tav. L, M/4v (RIC V,2, p. 585, No. 4). Compare RIC V,2, A. R. Bellinger, The Coins. The excavations at Dura- p. 585, No. 7. See also E. Equini Schneider, Septimia Europos. Final Report VI (1949). 65 Zenobia Sebaste (1993), p. 87, n. 2, for further references. A. Spijkerman, The Coins of the Decapolis and Provincia But any coinage of the brief period that the actual position Arabia, ed. M. Piccirillo (1978), and Y. Meshorer, City- of Roman emperor was claimed by Wahballat ought to be Coins of Eretz-lsrael and the Decapolis in the Roman Period interpreted in an "imperial" context. In associating himself (1985). In addition to Gadara and Philadelphia there are with Heracles/Hercules, the Palmyrene usurper would have examples from Abila, Adraa and Pella. Golden apples continued an imperial tradition in order to substantiate his (damaged): Spijkerman, Pella No. 10; : claim to the throne. Compare V. Picozzi, "Le monete Meshorer, Abila No. 213. THE "HERACLESFIGURE" AT HATRA AND PALMYRA 229 be veryimportant, and the seriesof coins of Philadelphiadepicting a chariotwith canopysupported by four pillars,drawn by four horses, which is identifiedin Greek as the "chariotof Heracles", is usuallyinterpreted by scholarsas pointing to an underlyingindigenous deity.66 A coin from Ptolemais-Aco,on the Phoeniciancoast again, shows Heraclesbeing given a herb by the local river to heal the wounds he had receivedin his fight with the Hydra.67In this case one of the laboursof Heraclesis used and reinterpretedin a local context, and it raises the questionwhether we should imaginethe existencein the Near East of a fixed cycle of his exploits as known from the Classicalworld. Graeco-Roman influence on the representationof the Heracles figure at Hatra and Dura-Europoscan be detected in the use of motifs like the apples of the Hesperidesand the battle with the Nemean lion, but whether the whole Greek corpus of the Heracles myth was known or accepted there and elsewherein the Near East, either for the ClassicalHeracles who was newly introducedor for any deity to whom he became assimilated, remainsvery doubtful.68Since we know close to nothing with regardto Near Easternmythology in the Roman period, correspondencebetween myth and iconographicrepresentation, as known from the Classical world, is thereforelacking. Whether that situation will change with future discoveries,or whetherit reflectshistorical reality, remains of course uncertain.

Place andfunction of the Heraclesfigure It is thereforeof the uttermostimportance that the Heraclesfigure appearsamong a row of deities on one of the very few monumentsfrom Palmyrathat gives us some insight into the local mythology,and as such above all servesas a reminderof how much we do not know. The relief on one of the beams from the peristyle of the temple of Bel depicts an encounter between Palmyrenedeities and a female figure with snakes instead of legs. It is usually interpretedas representingthe victory of Bel over , known from EnumaElish, the Babylonianepic of creationwhich used to be recitedduring the Akitu festival.69The deitieslined up on the right side of the relief are difficult,if not impossible,to identify unambiguously,but at least they seem to point to a local Palmyreneadaptation of the myth of the slaughterof Tiamat.70Very recently Dirven has arguedthat the most importantfigure on the relief is not Bel, ridinga charioton the left fragmentwhich is now missing,but Nebu ( in ,where he is the son of - Bel), riding a horse and occupyingthe most central place on the relief, and that the six deities standingin a row were connectedwith the temple of Nebu,71who seems to have been identified

66Spijkerman, The Coins of the Decapolis, Philadelphia ibid., Nos. 28-32. One Palmyrene tessera (RTP 1032) is Nos. 21-2, 29, 35, 40, 43 and 46 (from Marcus Aurelius to believed to show Heracles hurling the Ceryneian hind to Elagabal). Meshorer, City-Coins, p. 96, puts forward the the ground, but this must remain a hypothesis. The mosaic hypothesis that the canopy covers the sacred stone which from Philippopolis on which a drunken Heracles (identified was worshipped in the cult of Heracles, and that the Greek as such in Greek) attends the wedding of Dionysos and deity was identified with both Melqart and the Ammonite Ariadne ought to be interpreted in the context of the rapid deity Milkom. Bonnet, Melqart, pp. 146-7, although writing changes by which the village of Shahba was transformed in with regard to the depictions of Heracles on the coins of a Graeco-Roman city worthy of being the birthplace of the the Decapolis that "rien cependant ne le distingue du type "Arab" emperor Philip. For the mosaic see Balty, Mosalques grec traditionnel" (p. 147 n. 9), connects the coins dep- antiques de Syrie, pp. 50-7. icting Heracles'chariot with the title of eyEpaue[in?vTOO] 69 See Gawlikowski, "Les dieux de Palmyre", 'HpaKA0ov([] ("the one who raises Heracles"), given to a pp. 2614-15, with further references. For the relief see citizen from Philadelphia who is further identified as gymna- Tanabe, Sculptures of Palmyra, Pls. 32-4, and Drijvers, The siarch, senator and president of the council and assembly, Religion of Palmyra, PI. IV, 2. For an English translation and argues that "l'Heracles d'Amman n'est pas simplement with introduction of Enuma Elish see S. Dalley, Myths from le heros grec." Mesopotamia (1989), pp. 228-77. For a new interpretation 67Meshorer, City-coins, No. 13. Stephan of Byzantium, see now L. Dirven, "The exaltation of Nabil. A revision of s.v. AKt,, tells the story how Heracles had founded the city the relief depicting the battle against Tiamat from the in gratitude, the Semitic name Aco being interpreted as temple of Bel in Palmyra" in Die Welt des Orients 28 AK?,, the Greek word for "healing". (1997), pp. 96-116. 8Compare Boardman, The Diffusion of Classical Art, 70See Dirven, "The exaltation of Nabu", pp. 106-8. p. 328 n. 74. For the Heracles figure holding one or more Whether the Akitu festival as such was really celebrated in apples at Hatra, see al-Salihi, "The Sculptures of Divinities Palmyra, and - if it was - whether it was also subject to from Hatra", pp. 65-6 and Nos. 16-19. Equally, what local influences and full of Palmyrene peculiarities, we myth, if any, is represented on a cult bank from Hatra that cannot know for sure. 71 shows three scenes of a Heracles figure, possibly supported Ibid., p. 115. It has indeed been recognised for a long by an Athena figure, fighting a , remains unknown. time that any attempt to explain these attendant figures by See Downey, The Heracles Sculpture, p. 88 and PI. XXI. referring to the cult of Bel causes a lot of problems, see For the fight with the Nemean lion at Dura-Europos see Gawlikowski, "Les dieux de Palmyre", p. 2615. Equally, a 230 TED KAIZER at Palmyrawith Apollo.72For two of them this can be shown with some plausibility.First, a goddess with a bow probably representsthe BabylonianNanaia, identifiedat Palmyra with Artemis,Apollo's twin sister.73Secondly, the Heraclesfigure holding the club appearstwice on tesseraealongside this deity with a lyre, and he is also depictedon a relieffrom the templewhich is believedto have been "of Nebu".74Nevertheless, one will have to wait for the final publication of this temple to see whetheror how the Heraclesfigure really fitted in with a group of deities who have been identifiedtentatively as Shadrafa, Nanaia or Atargatis, the son of Atargatis, and .75 Another deity with whom the Heraclesfigure could be associatedat Palmyrais Yarhibol,the protectorof the Efqa springand hence one of the oldest Palmyrenedeities. They appearon the obverseand thereverse of thessera,same Yarhibolidentified by his name, the Heraclesfigure withoutany accompanyinginscription.76 A recliningHeracles figure is also depictedon the obverse of a tessera which shows two unidentifiedradiate deities, one of them with a crescenton his shoulders.77The Heraclesfigure also appears,sculptured in the local basalt stone, in the Hauran area in southernSyria,78 but, at present,at Palmyrahe does not seem to have been attachedto the group of deities centredaround Allat and -Shamin,who were among the most popular deities in the Hauran(whence they may actuallyhave reachedPalmyra). In contrast,the evidence from Hatra reveals at least some connection between the Heraclesfigure and Allat: one could refer to a relief from the so-called temple of Allat, on which a Heraclesfigure and the goddess Allat arebelieved to have beendepicted, and a furtherconnection might be foundin the appearance of Allat on a stele from Shrine XIII, whose main deity seems to have been a Heraclesfigure labelledGad (see below).79Thus, if the presentstate of the evidencefrom Palmyraseems to show that the Heraclesfigure was mainly associatedwith those deities who are either to be labelled "indigenous"or "Babylonian",one should not automaticallyexclude other possiblerelations nor jump to conclusionswith regardto his "originaldivine surroundings".The evidenceis equivocal, often unclear,and in generalinsufficient to trace his originsin Near Easternreligion.80 Throughoutthis paper I have stressedthat the Heraclesfigure at Palmyraand Hatra is never identified as "Heracles",and that the so-called "syncretistic"appearance of Heracles-Nergal remainsunclear. It is thereforestriking that when the figurewith club and lion's skin actuallyis identifiedby an accompanyinginscription, he is labelled(but only at Hatra), the personifica- tion of "Good Fortune",and a termwhich could also be used to addressa distinctdeity.81 Shrine leading role on the part of Nebu on the relief is not 77 RTP 245 (and 246?). incompatible with Nabu's rise to power within the religious 78See e.g. M. Dunand, Le musee de Soueida (1934), world of Babylon, which by the late Babylonian period had No. 48, and the catalogue of the National Museum of made him reach a status equal to Marduk-Bel and which Damascus, p. 112 (C5183). D. Sourdel, Les cultes du Hauran is believed to have been reflected in the Akitu festival; see a l'epoque romaine (1952), p. 35, put forward the hypothesis Dirven, "The exaltation of NabCi", pp. 111-13. that the Hauran functioned as a passage for Phoenician Nevertheless, the cult of Nebu at Palmyra is, to say the influences to Trans-Jordan, although she stressed the local least, quite difficult to reconstruct. aspect of the Heracles figure in the Hauran. Bonnet, 72See A. Bounni, "Les representations d'Apollon en Melqart, p. 145, argues against the old viewpoint and sees Palmyrene et dans le milieu syrien" in Kahil and Auge the Greek model standing behind the Heracles figure in the (eds.), Mythologie greco-romaine, mythologies peripheriques, Hauran instead. Compare Bounni, "Iconographie pp. 107-12. A deity identified as Nebu and holding a lyre d'Heracles en Syrie", p. 383. often appears on the tesserae. Note that Bounni, op. cit., 79For the relief see W. I. al-Salihi, "Further notes on pp. 108-10, interprets one of the deities lined up on the Hercules-Gnd' at Hatra" in 38 (1982), Fig. 1. For right side of the relief as Apollo instead. the stele see idem, "Palmyrene sculptures found at Hatra" 73Note that the identification of the Greek twins with in Iraq 49 (1987), P1. XIIIa. the Babylonian couple is also made by , Geogr. XI 80For example, at Dura the Heracles figure appears on a 1, 7. fresco from the temple either known as "of the Palmyrene 74 For the tesserae see RTP 168 and 237. For the relief gods" or "of Bel", on the tableau immediately left of the see Drijvers, The Religion of Palmyra, PI. XIII, 2. Tyche of the Palmyrenes and the Tyche of Dura. The 75See Dirven, "The exaltation of Nabu", pp. 107-8, for context remains unclear. See F. Cumont, Fouilles de Doura- a full description of the deities lined up on the right side of Europos (1926), p. 118 and PI. XLIX. On the problems the battle relief. For some preliminary reports on the temple with regard to the labelling of this and other temples see "of Nebu" see A. Bounni and N. Saliby, "Six nouveaux F. Millar, "Dura-Europos under Parthian rule" in emplacements fouill6s a Palmyre" in AAS 15, 2 (1965), J. Wiesehofer (ed.), Das Partherreich und seine Zeugnisse, pp. 121-38, and A. Bounni, "Le sanctuaire de Nabu a Historia-Einzelschrift 122 (1998), pp. 473-92. Palmyre" in F. Zayadine (ed.), Petra and the Caravan Cities 81 See in general T. Kaizer, "De Dea Syria et aliis diis (1990), pp. 157-67. deabusque. A study of the variety of appearances of Gad in 76RTP 247. Aramaic inscriptions and on sculptures from the Near East THE "HERACLESFIGURE" AT HATRA AND PALMYRA 231

XIII at Hatra has revealedfive inscriptionsthat refer to "the (great) Gad of Ramgu", while sculpturesof the sameprovenance show that it was the Heraclesfigure that servedas the protecting Good Fortuneof those frequentingthe sanctuary.82From the North gate of the same city comes a similar statuette of a Heraclesfigure in Hatrene dress, identifiedby the Aramaic inscription underneathas Gad (qdmgnd').83 As I have argued elsewhere,the anonymous Gad from one of the city's gates could be compared with the so-called Gad of the gate, who appears twice at Hatra and seems to have fulfilled a public function as divine city protector.84And although the interpreta- tion of the evidenceremains a hypothesis,the Heraclesfigure may have played a similarrole as guardianat the borderof the city territoryof both Hatra and Palmyra.An Aramaicinscription found 25 km east of Hatra refersto a sacredplace "of Nergal", and that site has been identified with a place which was called, by Ptolemy, 'HpaKASovus O/lOL,and, on the Peutingeriana Tabula, ad Herculem.5 Equally, a Heracles figure on a column was found at a distance of c. 10 km from Palmyra.86In addition, it is worth mentioning that in Dura-Europos he is often attested in private houses.87 Whether these apotropaic characteristics of the Heracles figure should be explained in terms of Greek influence,where Heracleswas the pre-eminentprotector who could ward off evil in any period of distress, or by referring to the similar role which Nergal could play in ancient Mesopotamia,88 remains an open question.

Conclusion This leads us to my concludingremarks. We cannot know what the varietyof representations of the Heraclesfigure meant to his worshippersin Palmyra,Hatra and elsewherein the Near East. Typologicalantecedents in the ancient Near East are of course possible, but any Semitic substratumwhich may have lain beneaththe evidencefrom the Roman period remainsunknown. The Greek and Graeco-Roman Heracles fitted in well in various contexts due to his susceptibility to change and interpretation.89 But, as I have tried to argue throughout this paper, an ongoing process of Hellenisation is not sufficient to explain the developing nature of the local cultures in which the Heraclesfigure was deeply rooted in the first three centuriesAD. In the local contexts in which I have tried to place the evidence,the appearanceof "Nergal"is not necessarilyto be perceivedas "Babylonisation",and neitheris the appearanceof the Heraclesfigure to be seen as "Hellenisation"or "Graeco-Romanisation".The gradualadaptation of new materialmakes more sense when this materialis interpretedas additionalto an open local culturein whichnew material becomesauthentic and in whichthe variouscultural elements were constantly reinterpreted.90 And "reinterpretation",quoting Boardman,"is a respectablecreative exercise."91

in the first three centuries AD", Part 1 in Orientalia discussed by W. I. al-Salihi, "A note on a statuette from Lovaniensia Periodica 28 (1997), pp. 14-66, Part 2 in OLP Hatra" in Sumer 29 (1973), pp. 99-100. For photographs 29 (in press). see the Arabic part of the same volume, pp. 151-6 with 82For the temple and its evidence see W. I. al-Salihi, Figs. 1-5. "The excavation of shrine XIII at Hatra" in Mesopotamia 84H 297, gd' dy' bwl' (North gate); J. K. Ibrahim, Pre- 25 (1990), pp. 27-35. See Fig. 18, a decapitated statue of Islamic Settlement in Jazirah (1986), No. XIV (H 1031 in the Heracles figure in Hatrene dress wearing an amulet, and Beyer, Die aramaischen Inschriften), nys' dgnd' d'bwl' (East especially Fig. 21, a statue of the Heracles figure, identified gate). See Kaizer, "De Dea Syria et aliis diis deabusque", as the Gad of Ramgu (H413 II), standing in a small niche Part 2, for full references. behind the dedicator. On the former see also idem, "Two 85For the Aramaic text see now Beyer, Die aramaischen cult statues from Hatra" in Iraq 58 (1996), pp. 105-8. The Inschriften, p. 116 (S 1). For Ptolemy see Geographia (ed. inscriptions from this shrine are published by B. Aggoula, C. Nobbe), V 18, 1 and VI 3, 4. See also Bonnet, Melqart, "Remarques sur les inscriptions hatr6ennes XVIII" in Syria p. 151, and idem, "Heracles en Orient", p. 183. 67 (1990), pp. 405-19. The Gad inscriptions from this 86See Bonnet, "Heracles en Orient", p. 180. Professor temple are listed with translation in Kaizer, "De Dea Syria Michal Gawlikowski kindly informed me that the column et aliis diis deabusque", Part 2. See now also Beyer, Die was found at the source of an aqueduct leading there. 87 aramadischenInschriften. The fact that the forms gadda and See Downey, The Heracles Sculpture, p. 82. 88 ganda appeared alongside each other in these texts can be Compare Drijvers, "Hatra, Palmyra und Edessa", explained as a regular case of "dissimilatorische p. 834. Geminatenauflosung", see Kaizer, op. cit., Part 1, 89G. K. Galinsky, The Herakles Theme. The Adaptations pp. 151-2. of the in Literature from Homer to the Twentieth 83 For the inscription see Aggoula, Inventaire, H 296. See Century (1972), p. 2. also Kaizer, "De Dea Syria et aliis diis deabusque", Part 2, 90See above, n. 49. 91 and Beyer, Die aramaischen Inschriften. The statuette is Boardman, The Diffusion of Classical Art, p. 320. 232 TED KAIZER

Addendum After submittingthis articleI was shownthe proofsof a relevantforthcoming paper by Professor Michal Gawlikowski, "Heracles au desert. Autel dedie a Heracles-Nergal", to be published in Bulletin de correspondancehellenique, Suppl. 38, pp. 155-60. I am also grateful to him for some minor correctionson my final text. For a more recent and slightly alteredversion of the article by L. Dirven in Die Welt des Orients 28 (1997), pp. 96-116, see now her book The Palmyrenes of Dura-Europos. A Study of Religious Interaction in (1999), pp. 147-55. Of some interestto the topic is also an older articleby J. C. Greenfield,"Nergol dh§pt'" in Hommages et Opera Minora, vol. XII, A Green Leaf Papers in Honour of Professor Jes P. Asmussen (1988), pp. 135-43.