From Goddess to Prophet: 2000 Years of Continuity on the Mountain of Aaron near Petra, Jordan1 ANTTI LAHELMA & ZBIGNIEW T. FIEMA University of Helsinki The Lord said to Moses and Aaron at Mount Hor, on the border of the land of Edom [...] ‘Take Aaron and Eleazar his son and bring them up to Mount Hor; and strip Aaron of his garments, and put them upon Eleazar his son; and Aaron shall be gathered to his people, and shall die there.’ Moses did as the Lord commanded; and they went up Mount Hor in the sight of all the congregation [...] and Aaron died there on top of the mountain. (Num. 20: 23–28.) Abstract The Mountain of Aaron (Jabal Haroun) near the ancient city of Petra, Jordan, is the traditional burial place of the Old Testament prophet and a site considered sacred by the three world religions of Christianity, Judaism and Islam. Since 1997, a Finnish archaeological project has been investigating the mountain through the excavations of a Byzan- tine pilgrimage complex on its high plateau and an intensive survey of its surroundings. In the course of the excavations, it has become clear that the Byzantine structures were preceded by a monumental building, probably a temple of the Nabataean-Roman period. More- over, already in the pre-Christian period a pilgrim route probably led from central Petra to Jabal Haroun. The article explores the history and archaeology of Jabal Haroun, which shows a remarkable degree of continuity and opens up the prospect that the local folk tradition may preserve elements of Nabataean religion. Using contemporary rituals and beliefs associated with the mountain as a reference point, we suggest that the pre-Christian ‘deity of Jabal Haroun’ can be identi- fied as the Nabataean goddess al-‘Uzza. Keywords: Petra, Jordan, Nabataean, Byzantine, religion, al-‘Uzza, Aaron 1  The research on the subject of this article has been carried out within the framework of the Research Centre ‘Ancient and Medieval Greek Documents, Archives and Libraries’, directed by Prof. Jaakko Frösén, which is part of the ‘Centres of Excellence in Research’ program of the Academy of Finland. We are grateful to Stephan Schmid (University of Basel), Jaakko Frösén and Päivi Miettunen (both of the University of Helsinki) for their critical comments. All errors of omission and interpretation are ours. © The Finnish Society for the Study of Religion Temenos Vol. 44 No. 2 (2008), 191–222 192 ANTTI LAHELMA & ZBIGNIEW T. FIEMA Figure 1. Location of Petra (left) and sites discussed in the text. Map: H. Junnilainen, K. Koistinen, P. Kouki & A. Lahelma. Ever since its rediscovery in 1812 by the Swiss explorer Johann Ludwig Burckhardt (1784–1817), the ruins of the ancient Nabataean city of Petra, located in the Jordanian desert ca. 200 km south of the capital Amman (Figure 1), have captured the Western imagination.2 The Nabataeans were a Semitic people, originally nomadic but later increasingly involved in trading and agriculture; they first appear in Greek historical sources in the fourth century bce. For centuries, the kings of Nabataea successfully resisted Greek and Roman domination. Petra, the capital of the Nabataean Kingdom, con- veniently located at the crossroads of ancient caravan routes between East and West, experienced conspicuous prosperity during the Late Hellenistic and Roman periods, as exemplified by the monumental edifices, such as rock-cut royal tombs and temples. Even after eventual Roman annexation in 106 ce, the city continued to flourish. In the course of the Late Roman and Byzantine periods, however, a shift in trade routes led to Petra’s gradual decline (Fiema 2003, 50). The city was 2  The most recent and comprehensive treatment of Petra and the Nabataean culture is by Schmid 2001. For an extensive bibliography on the subject, see Crawford 2003. The ancient sources concerning the Nabataeans are presented and discussed in Hackl et al. 2003. FROM GODDESS TO PROPHET 193 Figure 2. The cenotaph of Aaron (Haroun) inside the weli bears evidence of the long history of reli- gious observance related to Aaron. Built partly of reused marble chancel screen posts from the Early Byzantine shrine, it is covered with Hebrew inscriptions left by Jewish pilgrims. The Arabic inscription in the central panel in the front records the building of the grave monument by Emir Seif-ed-Din in the fourteenth century ce. Photo: Anna Erving/FJHP. struck, moreover, by a disastrous earthquake in 363 ce, which left many of the temples and other monumental structures permanently in ruins. The city was long thought to have been abandoned soon after the earthquake. The general scarcity in Byzantine sources of historical references to Petra has further added to the romantic mystery surrounding the Nabataeans. Recent archaeological excavations, however, have shown that Petra survived the disaster of 363, and was in fact able to recover to the extent that in the sixth century the city boasted several churches adorned with mosaics and marble decorations (e.g. Fiema et al. 2001; Fiema 2002, 2008). Following Burckhardt’s visit, the ruins of Petra were subsequently de- scribed by numerous Western scholars and travellers. Today they form a major tourist attraction as well as a UNESCO World Heritage site. In local tradition, however, the holy mountain of Jabal Haroun (the Mountain of Aaron), located ca. 5 km south-west of the centre of the city, has been far more significant than the ruins of a long-disappeared civilization. Indeed, Burck- hardt himself travelled to Petra disguised as a Muslim pilgrim on his way to pay homage to the Prophet Haroun (Aaron), who according to tradition lies buried in a subterranean chamber under the fourteenth-century Muslim shrine (weli)3 built on the highest peak of the mountain (Figure 2). Until recently, archaeological activity in Petra has been concentrated in the centre of the ancient city, where several large-scale excavation projects have taken place (e.g. Hammond 1996; Bignasca et al. 1996; Joukowsky 1998; Fiema et al. 2001; Schmid & Kolb 2000; Ruben 2003). Apart from a 3  The word weli is a dialectal form of the Arab word wali, which means ‘friend’ and refers to both the saint and, in informal language, his shrine. The word maqām, meaning ‘sacred site’ or ‘tomb of a saint’, is also used by the locals. 194 ANTTI LAHELMA & ZBIGNIEW T. FIEMA few brief reports concerning ruined structures on Jabal Haroun (Wiegand 1920, 136–145; Peterman & Schick 1996), no substantial archaeological re- search had been conducted there until 1997, when the Finnish Jabal Haroun Project (FJHP) began its investigations. Headed by Professor Jaakko Frösén and funded by the Academy of Finland and the University of Helsinki, an international team of archaeologists, conservators, cartographers and other specialists has spent eight fieldwork seasons (between 1997 and 2007) exca- vating a ruined architectural complex (Figure 3) located on a large plateau below the summit of Jabal Haroun, and surveying its environs. The project is now nearing completion, and the first volume of its three-volume final publication has just been published (Fiema & Frösén 2008). The volume contains major presentations on the figure of Aaron in ancient sources, as well as on the history of Jabal Haroun in past centuries (Frösén & Miet- tunen 2008; Miettunen 2008). However, the data are still being analysed, and some of the observations made in this paper – especially with regard to the pre-Byzantine history and archaeology of Jabal Haroun – are therefore preliminary and tentative. The purpose of this paper is to explore one significant aspect of the site, i.e. the literary, archaeological and ethnographic evidence for the long-term continuity of sacred significance of Jabal Haroun. The FJHP excavations as Figure 3. A view of the FJHP excavation site (in 2007) from the summit of the mountain. Photo: Matti Mustonen/FJHP. FROM GODDESS TO PROPHET 195 well as the intensive survey (Lavento et al. 2006) of its surrounding area, indicate that some aspects of religious tradition pertaining to the mountain extend from the Nabataean period to the present. Our findings may thus shed some new light on the Nabataean cult and religion, which due to the scarcity of ancient written sources remains poorly understood. The most recent study on the subject – John Healey’s The Religion of the Nabataeans: A Conspectus (2001) – begins by stating the need for ‘a certain creativity of thinking, [without which] all that we will be left with will be a recital of epigraphic and archaeological details’ (Healey 2001, 2). While we ac- knowledge the speculative nature of the subject and the necessity of crea- tive imagination in reconstructing Nabataean religion, we emphasise the unexplored potential of the archaeology of Nabataean religion, which can and does reach beyond ‘details’ that merely serve to illustrate the written records. We also wish to explore one particular strand of evidence so far largely ignored: the beliefs and practices of the contemporary populations of the Petra region, which – perhaps unexpectedly – appear to shed new light on the ancient past. A ‘Direct Historical Approach’ to the Cult of Jabal Haroun The investigations of the FJHP made it increasingly clear that the roots of the cult of Jabal Haroun lie in the Nabataean religion. It is significant that even though the city of Petra gradually became impoverished and was ultimately abandoned in the Early Islamic period, the mountain of Jabal Haroun remained in use as a holy site. Indeed, aside from the modern town of Wadi Mūsā (formerly known as the village of Elji), Jabal Haroun appears to be the only significant archaeological site in the entire Petra region where an apparent continuity from Nabataean to modern times is attested. This great historical trajectory, related to a site considered sacred by Judaism, Christianity and Islam, is remarkable in itself and worth recounting. On a different level, it enables a ‘direct historical approach’ to the archaeology of Jabal Haroun, based on aspects of local folklore and ritual in addition to historical sources. The direct historical approach – a form of ethnographic analogy – was first elaborated in early twentieth-century North American archaeology (e.g. Steward 1942), where an abundance of ethnographic research combined with a lack of ancient historical sources made it natural to apply ethnographic material to the interpretation of archaeological finds. The theoretical basis of the direct historical approach lies in cultural continuity and transmis- 196 ANTTI LAHELMA & ZBIGNIEW T. FIEMA sion. Quite simply, it is a method by which one works ‘from the known to the unknown’ (Steward 1942, 337) – from the ethnographic present towards prehistory. By emphasising the cultural and geographical links between past and present, the approach provides a rather straightforward solution to the problem of choosing the most relevant analogies for interpretation. Exactly how far back in time the approach can be extended is a difficult question, to which there is no clear-cut answer; but dramatic changes or breaks in the archaeological record may also indicate breaks in cultural transmis- sion. There is also a need to exercise caution in the use of ethnography in interpreting archaeological remains, as the meanings of ritual practices and religious symbols may have undergone radical transformations in the course of time (Trigger 1995). Great care should thus be accorded to establishing (rather than assuming) the continuity of cultural forms. This approach has enjoyed somewhat varying fortunes in different countries and subdisciplines of archaeology. A search for cult continuity has been persistently popular in Classical and Near Eastern archaeology, but in Central and North European archaeology the prehistoric past has often been viewed as completely disconnected from the present.4 In the United States, where the use of the direct historical approach had seemed so natural in the early twentieth century, the method fell into general disrepute with the New Archaeology of the 1960s (Lyman & O’Brien 2001). This development is related to what may be seen as the ‘downside’ of the direct historical ap- proach: its deceptive simplicity and obvious interpretative force can easily lead to misguided use. As Ann Stahl (1994, 181–2) points out, archaeolo- gists who use the approach have tended to disregard change and stress the similarities between past and present, using the method rather freely to illustrate the less accessible aspects of past lifeways. This has caused some archaeologists (e.g. Fahlander 2004) to view the approach as essentially patronizing, branding indigenous cultures as ‘cold’ and stagnant, without social progress or flexibility. The critique of the direct historical approach is related to a wider scepticism concerning the use of analogy in archaeol- ogy (see Orme 1974; Wylie 1985), which has sometimes been erroneously perceived as somehow optional (e.g. Binford 1967). It is obvious, however, that if the goal of archaeology is to understand the past rather than merely describe it, the use of analogy is unavoidable. Critiques of the uses of 4  The main exception to this is the influential work of the early twentieth-century German archaeologist Gustaf Kossinna, whose notion of Siedlungsarchäologie is essentially a form of direct historical approach (Trigger 1989, 165). FROM GODDESS TO PROPHET 197 analogy in archaeology (including those pertaining to the direct historical approach) should thus be seen as critiques of flawed uses of analogy rather than of analogy as such. Today, the direct historical approach is regaining popularity in both American and European archaeology (Lyman & O’Brien 2001; Cunningham 2003), apparently because the New Archaeology failed to deliver on its promise of universal laws of human behaviour. Another possible reason is the growing interest in the archaeology of religion; as Bruce Trigger writes, the direct historical approach is the only technique available to archaeolo- gists that allows detailed insight into culturally specific aspects of prehis- toric ideologies and rituals (Trigger 1989, 342). True, direct continuation presupposes a causal order, which may be problematic; the development of religion usually involves multiple interacting causes. Ideally, relational analogies should be brought in to support direct historical ones. However, since religion is often culturally conservative and resistant to change, study- ing it may be particularly well suited to the direct historical approach. Its potential certainly needs to be explored in the study of Nabataean religion, where epigraphers and historians of religion are desperately short of use- ful data. Jabal Haroun in Written Sources In the Jewish, Christian and Muslim traditions, Mount Hor is specified as the place where Moses’ brother Aaron died and was buried according to the Old Testament (Num. 20: 23–28, supra).5 The identification of Mount Hor is by no means clear, and Jabal Haroun, while a very strong candidate for this location, is only one of several potential places. It is noteworthy, however, that the tradition that places it near Petra is a very old one, going back at least to the writings of the first-century Jewish historian Flavius Josephus. As Josephus tells us in ‘Jewish Antiquities’ (Antiquitates Judaicae, written ca. 94 ce): [Moses] led his forces away through the desert and came to a place in Arabia which the Arabs have deemed their metropolis, formerly called Arce, to-day named Petra. There Aaron ascended a lofty mountain range that encloses the spot, Moses having revealed to him that he was about to die, and, in the sight of the whole army – for the ground was steep – he divested himself of his 5  For a detailed description of the sources see Frösén & Miettunen 2008. 198 ANTTI LAHELMA & ZBIGNIEW T. FIEMA high priestly robes and, after delivering them to Eleazar his son, upon whom by right of age the high priesthood descended, he died with the eyes of the multitude upon him. (Jewish Antiquities IV, 82–83; Josephus 1991, 516–7.) Notably, Josephus feels confident in identifying the site as located near Petra, even though the location only barely fits the itinerary of the Israel- ites as described in the Old Testament (cf. Frösén & Miettunen 2008, 5–6). Approximately two centuries later, Eusebius refers to Mount Hor in his Onomasticon (ca. 295), as ‘the mountain on which Aaron died near the city of Petra, on which even until today is shown the rock (from which water) flowed by Moses’ (Eusebius of Caesarea 2005, 165). In the following century, St. Jerome confirms ‘Mount Or’ as the mountain ‘on which Aaron died, near the city of Petra, where even until today the stone is shown which Moses struck and gave water to the people’ (ibid; translation according to Frösén & Miettunen 2008, 11). It is worth noting that while these writers do not mention a tomb or sanctuary associated with the burial, they indicate that by this time tradition placed not only Aaron’s death but also Moses’ miraculous spring on the same mountain near Petra, and that the latter was being shown by the locals to pious visitors. This forms an interesting discrepancy with the Biblical account (Num. 20), certainly well known to both Eusebius and Jerome, in which it is very clear that these events took place in two different locations, Kadesh and Mount Hor. In the sixth century we find the first references to a monastery near Petra. In 536 ce, two members of a ‘Monastery of Aaron’ are mentioned as attending the Church Councils in Constantinople and Jerusalem (Frösén & Miettunen 2008, 12). The exact location of the monastery is not specified, but it was probably in Petra, as one of the Petra Papyri (inv. 6a, dated to 573) being studied by a Finnish team of papyrologists refers to ‘the House of our Lord the Saint High-Priest Aaron’ outside of the city of Petra (Frösén & Fiema 1994; Gagos & Frösén 1998, 477; Frösén 2001, 491; 2004, 142).6 Despite the Islamic conquest of the region in the early seventh century, a Christian presence seems to persist for several centuries at Jabal Haroun. The Vita of St. Stephen the Sabaite mentions a ‘Mār Hārūn’ – probably Jabal Haroun – as a holy site visited by Christian monks of the mid-eighth cen- tury (Leontios of Damascus 16.2, 1991, 96–7). The site is mentioned again 6  With two volumes published so far (Frösén et al. 2002; Arjava et al. 2007), the scientific publication of the papyri is still underway. Papyrus inv. 6a will be published in a forthcoming volume of the Petra Papyri series. FROM GODDESS TO PROPHET 199 around 955/956 by the historian Al-Mas’ūdī, the ‘Herodotus of the Arabs’, who writes of Jabal Haroun as a Christian holy mountain in the possession of the Melkites (al-Mas’ūdī 1894, 143–4). Two centuries later, the mountain reappears in the accounts of the Crusades. The Crusader leader Baldwin, just before he became king of the Latin Kingdom, visited Jabal Haroun in 1100. His chaplain, Fulcher of Chartres (b. 1059), describes the visit: ‘Fur- thermore we found at the top of the mountain the Monastery of St Aaron where Moses and Aaron were wont to speak with God. We rejoiced very much to behold a place so holy and to us unknown.’ (Fulcher of Chartres 1913, 381; translation according to Frösén & Miettunen 2008, 13.) This account proves that the monastery, in whatever form (Fiema 2008), was still in existence in the twelfth century. Magister Thetmarus, who visited Petra in 1217 during a truce between the Christians and the Arabs following the battle of Hattīn, mentioned Jabal Haroun and wrote that a church on its summit was inhabited only by two Greek Christian monks (Magister Thetmar 1983, 272; cf. Frösén & Miettunen 2008, 14). However, he did not visit the summit but merely reported what he had heard. Quite soon after his account, the Christian history of Jabal Haroun seems to have come to an end. Islamic records of Jabal Haroun are few in number, suggesting that the site was mostly of local importance. In 1276, the Mameluke Sultan Baybars traveled through Petra and ascended the slopes of the mountain on which was ‘the tomb of Aaron, Prophet of God, the Brother of Moses, Son of ‘Um- ran’ (Zayadine 1985, 173). The account of Baybars’ journey does not mention any specific religious structures on the site. The major event in the Islamic history of the mountain was the construction of the Islamic shrine (weli) in the fourteenth century. The site is mentioned, once more, by the Arab geographer Yaqūt in the 15th century (al-Salameen & al-Falahat 2007, 259), but after this all mentions of Jabal Haroun become exceedingly scarce. A few Jewish writers, including Rabbi Jacob (writing between 1238–1244), and an anonymous author who visited the site in 1537, refer to Jabal Haroun as a site of pilgrimage (Peterman & Schick 1996, 478). Thereafter Jabal Haroun more or less disappears from the historical sources until its ‘rediscovery’, for the West, by Johann Burckhardt in 1812. In addition to these written sources, some inscriptions found near the site clearly indicate that the site was frequented in the Islamic period (al- Salameen & al-Falahat 2007). The oldest inscription, a short plea to Allah, is fragmentary and contains no date, but is according to Zeyad al-Salameen and Hani al-Falahat paleographically datable to the eighth or ninth century. A second, longer inscription refers to the mountain as a ‘blessed and hon- 200 ANTTI LAHELMA & ZBIGNIEW T. FIEMA oured place’ and asks for Allah’s forgiveness for all Muslims. It includes the date 690 AH / 1291 ce, thus predating the construction of the weli. Other inscriptions, probably dated to the Ottoman period (including one with the date 1019 AH / 1610 ce), complete the evidence for the sanctity of the site throughout the Islamic period. Three major events associated with the history of Jabal Haroun took place in the twentieth century. The German explorations at the summit of the mountain revealed the remains of a Byzantine Christian church beneath the Islamic weli, probably also associated with the cult of Aaron (Wiegand 1920, 136–45). A large quadrangular ruin, located on the high plateau of Jabal Haroun, ca. 70 m below the peak with the Islamic shrine, was visited in 1991 by Peterman and Schick (1996), who concluded that these ruins should be identified with a Byzantine monastery of St. Aaron mentioned in Byzantine sources. Finally, the information derived from the Petra Papyri (supra), combined with the religious tradition associated with Jabal Haroun and with the results of the early explorations in the area, would strongly suggest that the architectural remains on the high plateau of the mountain, which are the focus of the FJHP excavations, should indeed be identified as the Byzantine Monastery of Saint Aaron.7 From the Western Building to... the Western Building – The Archaeologi- cal Sequence at Jabal Haroun Like the written sources, the archaeological record of Jabal Haroun pro- vides evidence for a religious tradition which, although ‘reinvented’ over the course of centuries, has continued for at least the past two thousand years. The focus of the FJHP investigations is a ruined building complex on the high plateau, just below the summit, measuring ca. 62.9 m N-S x 46.7 m E-W.8 Although most of the currently visible structures in their extant form belong to the Byzantine monastery dedicated to St. Aaron which existed roughly between the later fifth and the eighth/ninth century, the site contains more Nabataean remains than initially assumed (Figure 4). A brief description of the site is therefore followed by an account of the spatial and temporal changes which occurred from the Nabataean until the Early Islamic periods. 7  See also Lindner 2003, 177–204 for a recent description of Jabal Haroun and the way leading to the mountain from Petra. 8  For references to all reports as of now published concerning the FJHP see Fiema & Frösén 2008.