Settings

The Anatolian Cult of Sabazios

¶1 In 1989 I published an article on examples of Attic vase painting which illustrate non-Greek divinities and cult rituals.1 Among the pieces I discussed was an Attic red figured krater from Spina, now in Ferrara, with a scene of two seated divinities, one male and one female, both making offerings from a phiale.2 The female divinity can be identified with some certainty as the Mother of the Gods, the Greek Kybele, but the male figure presents more of a puzzle. Following a suggestion of Erika Simon,3 I offered an identification as Sabazios. Shortly after this article appeared in print, I received a letter from Eugene Lane telling me that I was mistaken; whoever this divinity was, it was not Sabazios. As was so often the case in matters dealing with eastern divinities in Greek and Roman cult, he was right and I was wrong. This was my first contact with Professor Lane, but it proved to be the start of a rewarding association from which I have benefitted enormously. With this in mind, I would like to offer the following comments on Sabazios, with the hope that the scholar who has contributed so much to our understanding of this divinity in ancient Mediterranean cult may find a few original points in it to stimulate his interest.

¶2 Sabazios is a deity who seems to exist on the margins of Greek and Roman cult practice, attested everywhere yet fully at home nowhere. His origins, his place in ancient Mediterranean cult practice, and his character have all been widely discussed, with little consensus achieved.4 Much of the interest in Sabazios has focused on questions of syncretism between his cult with that of other deities. Some of the more far-fetched attributions of syncretism, such as the placement of Sabazios’s origins in Thrace5 or the erroneous connection between Sabazios and Judaism,6 have largely been discredited. Modern discussions of Sabazios which place him in a Hellenic context and focus on the literary evidence from Greek and Roman sources seem more sensibly grounded, since these sources give direct information on the position of Sabazios in Greek and Roman cult. Moreover, we have references to Sabazios in the Greek world at an earlier date than in Anatolia, a fact that appears to lend greater authority to the comments of Greek authors. This approach, however, has the effect of limiting our view of the cult of Sabazios to a Greek and Roman perspective, thus separating the god from his place in Anatolian cult. As an example, Sabazios is routinely considered a Phrygian divinity, in large part because Aristophanes, who mentions the god in four of his plays, ascribes his origins to Phrygia.7 Yet from an Anatolian perspective, this is a highly problematic assumption, since there is no evidence for an equivalent deity in Phrygia in the late fifth century BCE, the time when Aristophanes attests to his presence in Athens. Therefore my goal in this paper is to analyze the information from Anatolia related to the cult of Sabazios and focus on the god’s place in Anatolian religious practice. From this we will see that Sabazios was very much at home in Anatolia where he was one manifestation of the principal male divinity, whom the Greeks and Romans identified as Zeus. While his cult is only rarely attested in Phrygia, it appears more frequently in other regions of western Anatolia, especially Lydia and Lycia, where the god’s worship and his status in Anatolian cult seem quite different from what is suggested by the Greek evidence. Let us review the sources on the cult of Sabazios from Anatolia and try to form an assessment of the god’s place in the religious practices of his homeland. Then we can turn to the Greek view of Sabazios and evaluate it more carefully. Such an analysis will help clarify the Hellenic filter through which this deity is viewed and provide further insight into his position in both Anatolian and Greek cult.

¶3 Evidence from Anatolia on the cult of Sabazios is provided by a series of inscriptions recording dedications to the god or regulating his cult; these are supplemented by votive reliefs and statuettes depicting the god. Since the inscriptions provide a longer and more complete record of the god’s cult, they will be considered first. We may start with the earliest known example, an interesting text from Sardis.8 Inscribed in the first half of the second century CE, the text records a decree originally promulgated in the mid-fourth century BCE, probably during the reign of Artaxerxes II.9 The decree opens with a dedication to Zeus Bardates by the Persian satrap stationed in Sardis. The body of the text is a cult regulation which explicitly forbids participation in the mysteries of the deities Sabazios, Angdistis, and Ma by the neokoroi, the temple attendants, particularly those responsible for crowning the god and carrying the implements of fire and incense burning.10 Zeus Bardates, Zeus the Law Giver in Persian, may be identified with the chief Persian male deity Ahura Mazda, and the intent of the text was apparently to keep the worship of Ahura Mazda from being contaminated by local Anatolian cults.11 Thus by grouping Sabazios together with the cults of indigenous Anatolian deities, particularly Angdistis, another name for Matar, the Phrygian Mother goddess, and Ma, an Anatolian deity originally at home in Cappadocia, the text emphasizes Sabazios’s Anatolian origins.

¶4 The text suggests other issues as well. The Sardis text presents Sabazios as a distinct entity, separate from Zeus Bardates. The goal of the decree may have been to distinguish a major Anatolian male deity from the main Persian male deity, and thereby keep the Iranian identity of Zeus Bardates intact. This will have further implications as we review subsequent evidence on Sabazios, since, as we will see, the god’s name normally appears as an epithet of Zeus. Could the Persian governor of Sardis have been concerned to keep Zeus Bardates from being assimilated by the local people with an Anatolian counterpart, a male deity who might also identified with Zeus?12

¶5 This suggestion is supported by another inscription from Sardis that provides evidence of the worship of Sabazios there. The text is a dedication by a priest of Sabazios named Menophilos, of the tribe of Eumeneis, to king Eumenes, and it seems likely that this refers to Eumenes II of Pergamon, 197-159 BCE.13 The name of the god is spelled Sauazios, a form that we will meet in other Anatolian texts, and one which may well reflect the local pronunciation of the name. On the stone, a gap of uncertain dimensions before the word Sauazios exists, and it has been plausibly suggested that this should be restored as Zeus Sa<b>azios. If this is correct, it will be the earliest evidence for the identification of Sabazios with Zeus, an identification which was prominent in Pergamon and was to become virtually standard in later Sabazios texts.

¶6 The identification of Sabazios with Zeus is made explicit by the fuller documentation on the cult of Sabazios in Pergamon.14 Two letters by Attalos III, both written in October 135 BCE, report that the cult of Zeus Sabazios was brought to Pergamon in 188 BCE by Stratonike, wife of Attalos II and mother of Attalos III.15 Stratonike was originally from Cappadocia; assuming that she brought the deity with her from her homeland, this strengthens the argument for Sabazios’s Anatolian roots by placing him in another region altogether. The principal official of the cult of Zeus Sabazios was a member of the royal family, Athenaios, a cousin of Attalos III, who was further honored with the priesthood of Dionysos Kathegemon, another cult of importance to the royal family.16 Because of the royal patronage, the cult of Zeus Sabazios enjoyed high standing in Pergamon. In 135 BCE the god’s cult was placed on the acropolis of Pergamon in the temple of the city’s most important deity, Athena Nikephoros, and he was addressed as the most honored deity, the deity who stood by the city in times of danger.17 He was honored with sacrifices and processions and mysteries conducted for him on behalf of the city, πρὸ πόλεως.18 Zeus Sabazios was only one form of Zeus, distinct from the Zeus worshipped at the Pergamon Altar, yet was clearly an important presence in the city. Sabazios was the recipient of private cult in Pergamon also, as is evident from a small inscribed votive column dedicated to the god by Philotera, a private citizen.19 The god’s status seems to have been most closely related to the patronage of the Attalid dynasty, however, and thus the end of the dynasty, only two years after the letters were written, marks the end of our information about Sabazios in Pergamon.

¶7 Only slightly later than the documents from Pergamon is a text from Tlos, in Lycia, in which the name Sabazios appears alone, not as an epithet of an Olympian deity.20 Dating to the end of the second century BCE, the text records honors to an individual (whose name is not preserved) for his benefactions to the people of the city and to the race of Lycians, in military, civic, and political duties. One of these was his service as priest for life of Sabazios πρὸ πόλεως on behalf of the city. This language, noted above in the letter of Attalos III of Pergamon, defines the priesthood as a civic office and the deity as an official god of the city.21 The text from Tlos makes clear that the Sabazios cult in Anatolia was not limited to circumstances specific to the royal family of Pergamon, but could function as a cult of a πόλις as well. Nor was it limited to the highly Hellenized parts of western Anatolia, but was also a presence in more remote areas such as Lycia.

¶8 I have treated the Anatolian testimonia on Sabazios from the fourth through second centuries BCE in some detail because these texts introduce the god to us and give a picture of his rituals and social status. During the Roman era, dedications to Sabazios, ranging in date from the late first through early third centuries CE, become more abundant; yet at the same time they fall into predictable patterns and so can be treated more briefly. In the majority of the examples the god is addressed as Zeus Sabazios.22 The name of the deity appears in variant spellings, the most common of which is Saouazios, implying that regional pronounciation often used a soft semi-vowel, like the English w, for the internal b.23 The majority of Sabazios inscriptions were found in western Anatolia, in Bithynia, Ionia, Lydia, and Lycia.24 Despite Aristophanes’ comment that Sabazios was a Phrygian, evidence for the god’s cult in Phrygia is much rarer; to date only two texts are known from Phrygia,25 although a well preserved statuette of the god, found in Çavdarli, near Afyon, discussed below, adds further evidence to the god’s presence in this region.26

¶9 Dedications to Sabazios exhibit the range of subjects typical of religious dedications from the Roman era in Asia Minor. The majority record dedications in fulfillment of a vow to the god.27 Rarely is the reason for the vow specified, although one touching case records a vow from a freedman who prays for the safety of his father, presumably still a slave.28 One vow is offered for good crops by the inhabitants of a village who dwell on sacred land, presumably a temple estate of the god.29 A few are confessions of wrongdoing and atonement, usually confessions from individuals who stole or damaged property belonging to the god.30 There are also records of new sanctuaries dedicated to Sabazios. Texts from Maionia in Lydia and Sakcilar in Bithynia report the establishment of private sanctuaries, while in a text from Koloe (modern Kula), the whole village joins in the establishment of a shrine of Zeus Sabazios; the event was commemorated by an elaborate relief depicting the worshippers processing to the altar of the god.31 One is a funerary stele, set up by followers of the god, the Sabaziastai, for a woman, Euboula.32 In some dedications Sabazios is honored jointly with other aspects of Zeus. A certain Ploution, from Philadelphia, dedicates a statue of Zeus Sabazios to Zeus Koryphaios, while another individual in Phrygian Epiktetos makes a joint dedication to Zeus Bronton and Sabazios.33 Occasionally there are indications that the god could be an important presence in a given community. An example is furnished by an inscription on a large base of the third century CE from an imperial estate in Ormeleis, in Lycia.34 The text, inscribed by the participants in the god’s mysteries, attests both to the prestige and to financial support of the cult in this region. But overall we are left with the general impression that Sabazios played only a minor part in the religious life of Anatolia. The number of texts is fairly small, fewer than thirty, and forms only a small part of the corpus of religious inscriptions from Anatolia during the first three centuries of the Roman era.

¶10 Let us place Sabazios within the overall picture of local cults in Anatolia. His origins and his place in the Anatolian pantheon remain problematic. Despite the attribution of a Phrygian origin in Greek sources, Sabazios is most at home in western Anatolia. The centers of his worship are fairly diffused, with the strongest presence in Lydia and Lycia. Testimonia to his cult are, however, widely enough distributed to show that the cult was not limited to one specific area, indicating that the god was more than a purely regional presence. The background of his cult is also problematic. We have no mention of him in Anatolia before the mid-fourth century BCE, nor is there any visual image of him before the early Roman period. Before the fourth century BCE, the most conspicuous feature of Anatolian cult was the Mother Goddess, originally a Phrygian deity but one whose cult spread widely, particularly in western Anatolia.35 In the third century BCE and later, however, male divinities had come to dominate the Anatolian pantheon. Of these, the most prominent is Zeus, worshipped with various epithets and in various guises.36 In fact, Zeus appears in so many regional versions and with so many different epithets that it seems likely that we see, not the Greek Zeus, but rather the principal male divinity in Anatolian cult practice whom the Greeks would have equated with Zeus.37 Just as Matar, the Mother, the principal female divinity in Anatolian cult practice, was addressed with a large number of topographical and descriptive epithets, so was the major Anatolian male deity.38 These epithets are so variable and numerous that it is not fully clear to what extent these male deities, the forms of Zeus in the various regional cults, were considered to be the same figure, or whether we are seeing a large number of local cult figures, between whom only minimal connections would have existed in the minds of their worshippers. As we have seen, Sabazios is often identified with Zeus, and the theonym Sabazios was used as an epithet of Zeus in the majority of the texts. Could Sabazios then be the Greek perception of the principal male Anatolian cult figure, one aspect of this Anatolian god whom the Greeks would later identify with Zeus?

¶11 Visual representations of the god support this hypothesis by providing strong links to the Greek Zeus. There are several sculptural images of Sabazios from Anatolia, all of the first and second centuries CE. Four of the inscribed dedications to Sabazios are surmounted by a relief depicting the god, and one statuette of him, found near Afyon, is known. One stele, from Balat, in Bithynia, shows a mature bearded male seated on a throne, wearing a long gown and a headdress; he holds a phiale in his right hand and a spear in his left.39 The piece is now lost, however, and only imprecise older drawings of it survive, making an analysis of the visual iconography tentative at best. Clearer information can be derived from reliefs on three stelai from Lydia, from Maionia, Koloe, and Philadelphia.40 On each piece the god is depicted as a mature bearded male, standing and pouring a libation from a phiale. On two pieces he pours the libation onto a small altar, and on the third into a krater with a tree behind it. In each example the god is shown wearing the costume of a typical Greek male, a long robe drawn into a roll around his waist and pulled up over his left shoulder. On the stele from Maionia the god wears a distinctive headdress, a cap with a pointed tip, traditionally labeled Phrygian (although as we shall see below, this label is incorrect).41 On this work also a snake appears above the altar. Apart from this example of cap and snake, none of the reliefs has attributes that identify the god specifically as Sabazios.42 Indeed, apart from the cap, the general form and costume of the mature bearded male are quite similar to Greek representations of Zeus, further suggesting the connection of Sabazios with Zeus.

¶12 The statuette of the god from Afyon represents a different visual tradition.43 The piece depicts a mature male figure, standing with the weight on the left leg and the right leg bent and raised as if striding forward, as he steps onto a ram’s head. Both arms are outstretched, the right lowered and the left raised. Unfortunately both hands are missing, and so we cannot tell if the right hand was depicted with the tradition three-finger gesture of blessing characteristic of Sabazios votive hands.44 The costume is distinctive: the god wears a belted knee-length tunic with leggings and soft boots which reach to his ankles; he also wears a high cap with a point tipping forward and ear flaps, the conventional Phrygian cap. This representation of the deity follows the standard Sabazios iconography known from numerous reliefs and statuettes of Sabazios and from figurines attached to votive hands of the god, all of which, apart from this piece, come from the western Roman Empire.45 This costume might strike Greek and Roman eyes as Anatolian, but it has no parallels among the visual representations of the god from any region of Anatolia. It is instead a Greek creation, originally based on an Achaemenian prototype. It first appears in Greek vase painting during the first half of the fifth century BCE in depictions of Achaemenian Persian figures, who are shown with a very similar belted tunic, leggings, soft boots, and pointed cap with ear flaps. The use of the costume was also extended to represent a variety of mythological figures believed by the Greeks to be of Eastern origin, such as the Amazons and the Trojan Paris.46 A comparison with Greek representations of Attis is particularly telling, since Attis, a deity to whom the Greeks imputed Phrygian origin, wears a costume that is virtually identical.47 Here the same costume has become part of the regular iconography of Sabazios, in which this so-called Phrygian costume is worn by a mature male with a full beard. The image is that of an Oriental Zeus, in which the Greek body and face of the god are clothed in Eastern dress. The first application of this costume to Sabazios probably did not take place in Anatolia, but in the western Empire, perhaps in Rome itself, from where it spread widely throughout the western Empire.48 It seems likely that this iconographic type was imported into Anatolia during the early Roman period, a suggestion reinforced by the fact that the other Anatolian votive reliefs depicting the god do not follow this type.

¶13 In the case of Attis, not only the costume, but indeed the god himself can be shown to be a creation of Greek iconography and cult practice, first apparent in the fourth century BCE.49 Sabazios, however, is unlikely to have been a creation of Greek cult, as Attis was, despite the fact that pictorial representations of him were created under strong Greek influence. The evidence for his presence in Anatolia and his position as a deity of honor there is too widespread to deny his Anatolian roots. The text from Sardis sets him firmly in the company of other gods, Angdistis and Ma, of clear Anatolian origin. Thus one may wonder why representations of Sabazios drew on a visual scheme originally based on a Greek model. Why was an Anatolian image of the deity not adopted, as was done in the case of the Phrygian Matar Kubeliya, who became the Greek Meter? The answer seems to be that there was no Anatolian visual prototype of Sabazios to draw on because the god had never been represented in Anatolian art. When the cult of Sabazios spread to the west, a visual iconography was created for the god based on the widely disseminated schema used to represent another deity associated with Anatolia, Attis.

¶14 These observations give us some tools with which to address the questions posed at the beginning of this essay about the origins of Sabazios and the specifically Anatolian character of his worship. Both his name in Anatolia, Zeus Sabazios, and his physical appearance indicate that Sabazios was one aspect of the principal male divinity in Anatolian cult whom the Greeks identified as Zeus, so abundantly attested within the Hellenistic and Roman periods in Anatolia. Given this situation, one may wonder at the lack of evidence for Sabazios before the fourth century BCE. There is no cult image of any mature male figure, or indeed of any male figure at all, in central or western Anatolia that might be identified as Sabazios. Nor does his name appear in any Paleo-Phrygian text.50 One might claim that Sabazios did exist in indigenous Anatolian cult practice but that no visible evidence survives of him, but this seems unlikely when one contrasts Sabazios’s absence from texts and cult monuments with the abundant testimonia, both written and visual, for the principal Phrygian divinity Matar, the Mother. Indeed, the Phrygian Mother’s absolute prominence in Phrygian cult and her lack of a consort are among her most distinguishing features. The dearth of iconographic representations of Sabazios in Anatolia is particularly puzzling, and we may wonder if the high status of the Anatolian male deity, the prototype of Zeus Sabazios, led to a prohibition against visual depictions of this deity. This would explain the discrepancy between the infrequency of evidence for the cult of an important male divinity in Anatolia before the Hellenistic period, and the regular appearance of such a deity during the Hellenistic and Roman eras. These many Zeus figures were addressed with a plethora of local epithets in different regions of Anatolia, and the name Sabazios may originally have been one such regional or descriptive epithet of this Anatolian Zeus. The texts from Philadelphia and Phrygia Epiktetos which record joint dedications to Zeus Sabazios and another aspect of Zeus further suggest that Sabazios was one of many Zeus figures.51 We cannot be sure of the chronological progression of the early Sabazios cult, but the references to Sabazios in Greek literature some two generations before the Sardis text strongly suggest that his cult was a factor in Anatolian religious life before the earliest Anatolian texts in the Greek language make his presence visible to the modern viewer.

¶15 The texts from Anatolia also provide us with some insight about the rites celebrated for Sabazios and his status in Anatolian cult. One type of ritual celebrated for Sabazios was mystery cult. The texts from Sardis, Pergamon, and Ormeleis all refer to the mysteries for Sabazios, and imply that his worship was limited to initiates. The god was also worshipped in civic rites open to the whole community. Texts from Pergamon and Tlos attribute a high status to the god and state that the cult of Sabazios was important to the πόλις. Emphasis was placed on sacrifices and processions in addition to mystery rites. Civic cult is also implied by the foundation of a sanctuary for Sabazios by a village community.52 Yet the numerous private vows, dedications, comments on personal cult foundations and a private association of Sabaziastai suggest the power of the god to influence private individual cult as well. The wide appeal of his cult is also implied by the Sardis text; otherwise the Persian governor would not have needed to prohibit the worship of Sabazios to those engaged in Persian cult practice. Throughout we see a deity who was integrated into many facets of Anatolian religious life, public and private. Nowhere is there any hint of ecstatic cult ritual, no illustrations of tympana, nor any signs of connection with rituals celebrated for Dionysos. In all these aspects the cult of Sabazios in Anatolia is fairly typical of the broad range of activities found in many Anatolian cults of the Hellenistic and early Roman period.

¶16 Several of these conclusions might seem surprising, since they contrast sharply with the standard Greek and Roman picture of Sabazios. In Greek and Roman sources, Sabazios was often connected with Dionysos and his cult was noted primarily for its ecstatic rites, viewed as typical of foreign deities. Therefore let us turn now to a consideration of the Greek view of Sabazios and examine the factors that shaped the Greek view of the god.

¶17 A key factor of Sabazios’s place in Greek cult was his non-Greek origin.53 In most modern literature, discussions of non-Greek deities rest on the assumption that the character of the foreign deity and the rites held for him/her in the Greek world replicated the cult practices of the deity’s homeland. Greek reactions to foreign deities, however, mask a complicated situation, and the variable status of Anatolian deities in Greek cult well illustrates this. Some Anatolian deities whom we meet through Greek sources were the Hellenic counterparts of deities worshipped within their country of origin, while others were substantially altered by Greek cult practice and bore little relationship to an equivalent figure in their supposed ethnic homeland. An example of the former is the Greek Mother of the Gods, or Kybele. In ascribing a Phrygian origin to her, the Greeks’ assumptions were correct. A clear line of transmission and development can be followed from the Phrygian Matar Kubeliya to the Greek Meter Theon, the Mother of the Gods.54 On the other hand, the deity Attis, known to the Greeks as the consort of the Mother of the Gods, was also widely considered to be a Phrygian god, but here the Greeks’ attribution of origin appears to be wrong. No deity equivalent to Attis can be found in Phrygian cult monuments and inscriptions from central Anatolia never use the word Attis as a theonym until the Hellenistic when the deity appears as an import from Greece.55 Our task will be to understand where Sabazios fits into this spectrum: which Greek comments about his cult record genuine information, and which reflect the alterations experienced by the Sabazios cult in the Greek world.

¶18 One point of confusion, the question of the god’s origins, can be addressed fairly easily. Sabazios was regularly called a Phrygian by the Greeks, despite the fact that his cult was most abundantly attested in Lydia and Lycia and only rarely in Phrygia. Comment on Sabazios’s origin can be found in the earliest Greek source on the god, the comedies of Aristophanes, where Sabazios is identified as a Phrygian god in a fragment from the Horae, an attribution reinforced by a scholiast on a passage in the Birds.56 It is unlikely, however, that Aristophanes was recording precise ethnic information about Anatolian cult practices. Rather, his intent was to suggest the god’s Eastern origin and, perhaps, his low social status, at a time when labeling someone a Phrygian was tantamount to calling that individual a slave.57 Other comments by fifth century Athenian authors indicate that to contemporary Athenians, little distinction existed between Anatolian ethnic groups such as Lydians and Phrygians.58 Because of the significant presence of Phrygians in the slave population of Athens, they would be the dominant Anatolian group that Athenians came into contact with and so all Anatolian figures would likely be called Phrygian.

¶19 The rites celebrated for Sabazios offer another key to his status in the Greek world. Since Sabazios was a deity of non-Greek origin, we would expect his worship to differ from rites for the established deities of a Greek πόλις, which emphasized publicly conducted sacrifices and shared meals. The limited information we have about the rites for Sabazios in Greece, pertaining almost entirely to Attica, reinforces this expectation. Aristophanes implies that Sabazios’s rites emphasized ecstatic ritual designed to facilitate direct communication with the deity, either through the use of a trance-like state (Wasps 9-10) or through an emotionally agitated state induced by cries and drumming on the tympanum (Lysistrate 387-388). In this, the rites of Sabazios were similar to those held for other deities of non-Greek origin, including Adonis and Meter, the Mother of the gods.59 The primary social function of ecstatic rites was not to promote communal bonds, as was the case in a πόλις cult, but to encourage individual religious expression. As a result the individualistic tendencies of such an ecstatic cult could have a divisive effect, which may be one reason why they were looked on with disfavor.

¶20 Ecstatic rites, however, were not limited to foreign deities, but were also characteristic of some Greek divinities, most notably Dionysos.60 This circumstance may lie behind the frequent association of Sabazios with Dionysos, one which is stated explicitly in several Greek and Roman literary sources although never attested in any document concerning the cult of Sabazios from Anatolia.61 There is no reason to assume that the cult of Sabazios had any connection with Dionysos. No Dionysiac epithets were ever applied to Sabazios, and the god’s visual image, as has been discussed above, draws heavily on Greek images of Zeus, but shows no affinity to images of Dionysos. The equation between the two seems to arise from the common use of mystery cult and ecstatic rites, rather than from any commonality of cult identity.

¶21 These factors should be kept in mind as we assess one of the most problematical passages concerning the Sabazios cult in Greece, Demosthenes, On the Crown 259-260. In it Demosthenes attacks his opponent Aischines by accusing Aischines’ mother of participating in ecstatic rites involving processions, wearing special crowns, handling snakes, and the use of the cry Εὐοῖ Σαβοῖ. The language of Demosthenes, while vivid and compelling, may not be a literal description of the rites of Sabazios, since Demosthenes is clearly exaggerating for rhetorical effect.62 Moreover, the god’s name is not directly mentioned in this passage, and the exact meaning of the cry Εὐοῖ Σαβοῖ is unclear. It seems likely, however, that these words preserve a ritual cry for Sabazios, since they are coupled with another ritual cry, ὑῆς ἄττης, which may be the earliest reference to another deity of supposed Anatolian origin, Attis.63 It is likely that because of his Anatolian origins, Sabazios was coupled in Greek minds with Attis. Demosthenes’ comments confirm the image found in the Aristophanic comedies, that Sabazios’s cult in Athens had a distinctively exotic quality that kept it on the fringes of Athenian religious practice.64 They do not, however, justify the conclusion that Sabazios was connected with Dionysos.

¶22 There is, moreover, a strong sense of low social status attached to the cult of Sabazios in Demosthenes’ description, and this contempt is further suggested by Cicero, who cites a (now lost) comedy of Aristophanes in which Sabazios and certain other foreign gods were put on trial and expelled from the city.65 Such scorn seems to be present in the comments of Theophrastos also; he mocks one of his characters, the Late Learner, who gussies himself up for the festival of Sabazios, while the Superstitious Man calls on Sabazios’s name whenever he sees a snake.66 We cannot discard the information from Greek authors about the Sabazios cult altogether, for there seems to be some knowledge of Anatolian cult practice in the use of the snake, which appears in the passages of both Demosthenes and Theophrastos and in several Anatolian votive reliefs dedicated to the god.67 The common thread in the information gained from Greek authors, though, is that Sabazios’s foreign origin and the unusual character of his rites resulted in a marginal position for the god in Greek religious practice. This surely accounts for the distorted reaction in ancient literary sources to Sabazios, since it is based on the Greek experience with his cult and not on the god’s character and status in his Anatolian homeland.

¶23 Thus we can see that Sabazios’s place in Greek religious practice was quite different from his position in Anatolia. The cult of Sabazios spread to the Greek world, presumably through Anatolian immigrants, including slaves, since Anatolians are well attested in the Greek slave population, particularly in Athens. The date when the cult first appeared in Greece is uncertain, but must be during or earlier than the mid-fifth century BCE, since by the 420′s Sabazios was a well enough known figure in Athens that Aristophanes could use him as the butt of jokes which the local audience would be expected to understand. The supercilious quality of the observations by Aristophanes, Demosthenes, and others implies a lack of social respectability among the god’s followers, and this was enhanced by the use of ecstatic, emotionally charged ritual, presented as typical of foreign ritual. Yet the cult of Sabazios clearly struck a chord with many individuals in the Greek world, for we find it attested in several centers on the Greek mainland and the Aegean islands.68 As a non-Greek deity, Sabazios’s cult would have been outside the regular cult associations of a πόλις, and so we would expect to see some special organizations established to regulate his cult, as was done for the cults of other non-Greek divinities. This we do see, both in the Piraeus, an inscription recording the names of the ἐρανίσται of Sabazios,69 and in Rhodes, a text recording honors to an individual, Ariston of Syracuse, for his services to the κοίνον of the Sabaziastai.70

¶24 Taken together, the sources give us some insight into the identity of Sabazios and his place in Anatolian and Greek cult. In Sabazios we see one of the earliest manifestations of the principal male deity in Anatolian cult, later identified by the Greeks and Hellenized Anatolians with Zeus and worshipped under a large number of epithets. The name Sabazios may be one regional epithet of the Anatolian Zeus, or an epithet indicating a particular area of concern to the divinity or his worshippers. What we cannot tell is why people turned to the cult of Sabazios and what attracted worshippers to him.71 His presence in mystery and ecstatic rites suggests that his cult must have filled some real personal need to his worshippers, and this suggestion is further supported by the wide distribution of the Hellenized definition of Sabazios throughout the Roman Empire. The great majority of Sabazios dedications, both texts and images, come from areas far from his Anatolian homeland, attesting to the power of this cult and its ability to transcend cultural boundaries. Because of his identity with Zeus, Sabazios became a figure in Orphic cult as well, attested in several of the Orphic hymns as the Zeus figure who was responsible for the birth of Dionysos.72 In sum, the cult of Sabazios illustrates how a local cult that started with only a small regional following came to influence areas far beyond its original setting. It furnishes one small part in the larger picture of the impact of the native peoples of Anatolia on areas outside their homeland.

1 Roller 1989.

2 Ferrara, Museum no. T 128, ARV2 1052, 25. The identity of the two divinities on the vase has been much discussed, and no generally agreed upon identification of the male divinity has yet been reached. For further discussions of the vase, see Naumann 1983, 171-175; Bérard et al. 1989, 24-29; Matheson 1995, 278-79 (who assigns the vase to the Curti Painter). I have amended my own earlier opinion, see Roller 1999, 151-155.

3 Simon 1953, 79-87.

4 No one can address the topic of the cult of Sabazios without acknowledging the enormous debt owed to Eugene Lane, whose numerous publications on this subject, particularly CCIS II and III, correct many errors of earlier scholars and form the basis of any future study. Some earlier works such as Eisele 1909-1915 and Schaefer 1920, remain useful as compilations of ancient written sources on Sabazios; earlier attempts at synthesis, including Picard 1961, Fellman 1981, and Johnson 1972-1987, have been superseded by Lane’s work.

5 Sabazios as a Thracian deity, Picard 1961; Johnson 1972-87, 1506-07. As Eugene Lane points out, there is no evidence for the cult of Sabazios in Thrace before the first century BCE (see CCIS III, 1, 40). The evidence discussed by Fol 1998, concerning the connection of Sabazios and Orphism, with its presumed Thracian connections, is a product of the highly Hellenized cult of Sabazios known from Asia Minor in the second century CE (see below, n. 72); it does not prove the existence of a Sabazios cult in Thrace before this time.

6 On this, see Lane 1979; Johnson 1972-87, 1602-03.

7 Wasps 9-10, Birds 876, Lysistrate 387-390, Horae, frag. 566 (PCG 296); in the Horae fragment, Sabazios is called a Phrygian deity.

8 Robert 1975 publishes the text with extensive commentary. See also Sokolowski 1979 for some corrections on Robert’s translation and commentary.

9 Robert 1975, 310. While it seems likely that the text is a faithful copy of the original fourth century BCE document, some of the language may well have been altered to suit second century CE understanding. Note the comments of Robert 1975, 313, on the lack of iota subscript in the dative, typical of the second century CE. Robert also suggests that the last sentence, in which an individual named Dorates is specifically forbidden to take part in the mysteries, may be a later addition to the text.

10 Sokolowski 1979, 68.

11 Robert 1975, 314-317, 325.

12 On the evidence for Zeus in Sardis, see Hanfmann 1983, 228-229; Hanfmann distinguishes at least three other Zeus figures.

13 Johnson 1968; CCIS II, 30. On the variant spelling of the god’s name, see the examples listed below, n. 23.

14 On the cult of Sabazios in Pergamon, see Welles 1934, 264-273; Ohlemutz 1940, 269-272; Allen 1971, 8-9; Allen 1983, 130-131, 174-175.

15 Welles 1934, 267, letter 67 = CCIS II, 27. See Welles, 264-273, for a discussion of three texts, letters 64-67, which give information on the cult of Zeus Sabazios in Pergamon and his priest Athenaios. On the chronology of Stratonike, see Allen 1983, 203.

16 Welles 1934, 264-269, letters 65 and 66; Allen 1983, 130-131.

17 Τιμιωτατήν,Welles 1934, 266, letter 66 line 9; ἐμ πολλοῖς κινδύνοις παραστάτηγ, Welles 1934, 266, letter 67 lines 7-8.

18 Welles 1934, 267, translates this phrase as “before the city,” apparently meaning rites held in front of the city gates. The occurrence of the same phrase in an inscription from Tlos, CCIS II, no. 44a, discussed below, shows, however, that the term refers not to the place where Sabazios’s rituals were held but to his status as an official city god.

19 CCIS II, no. 26.

20 CCIS II, no. 44a.

21 J. and L. Robert 1983, 171-75; CCIS II, 21, commentary on no. 44a.

22 CCIS II, nos. 21, 22, 24, 25, 27, 29, 30 (Zeus is restored), 32, 33, 34, 36, 37, 38, 39, 40, 41, 42, 43, 45, nineteen of Lane’s twenty-eight Sabazios inscriptions. To the list in CCIS II, nos. 21-45, should be added two additional texts published by Frei 1988, nos. 9N and 10N. 10N is dedicated to Zeus Sabazios, while 9N is dedicated Διὶ Βροντῶντι κὲ Σαουαδίω; assuming that the theonym Zeus is to be taken with both epithets (see Frei, 23), this would make a total of twenty-one out of thirty Sabazios dedications.

23 CCIS II, nos. 24 and 25, from Bithynia; nos. 30 and 42, from Lydia; nos. 43 and 44, from Lycia; Frei 1988, nos. 9N and 10N, from Phrygia Epiktetos. On variant spellings of the name, see Johnson 1972-87, 1585.

24 In Bithynia, CCIS II, nos. 21-25; Ionia, nos. 28-29; Lydia, nos. 30, 32-42; Lycia, nos. 43-44.

25 Frei 1988, nos. 9N and 10N.

26 Drew-Bear and Naour 1990, 1947 n. 131. For the statuette, see Asgari et al. 1983, 125, no. B 346l; CCIS II, no. 77.

27 My data on the Sabazios dedications are drawn almost entirely from the inscriptions published by Eugene Lane, CCIS II, 9-22, nos. 21-45. To his list should be added two new inscriptions published by Frei 1988, above n. 25. For vows to the god, see CCIS II, nos. 21, 23, 24, 25, 29, 32, 35, 37, 37a, 37b, 38, 42; Frei, 9N and 10N.

28 CCIS II, no. 32.

29 Frei 1988, 22-23, no. 10N.

30 CCIS II, no. 33, “I cut down the trees of the god’s sacred grove”; no. 34, “the god found me lacking,” this from a sacred slave; no. 36, “[unknown] caused Trophimos, sacred slave of Meter Hipta and Zeus Sabazios, to be dragged off”; no. 40, “I took the sacred doves of the gods” (also a dedication to Zeus Sabazios and Meter Hipta), for a discussion of this last inscription, see Buckler 1914-1916, 169-172.

31 CCIS II, no. 22, from Sakcilar; no. 39, from Maionia; no. 41, from Koloe; see also Robert 1948, 111-113 and Lane 1975, 105-108.

32 CCIS II, 28.

33 CCIS II, no. 42, Philadelphia text; Frei 1988, 19, no. 9N. On the latter, see above, n. 22.

34 CCIS II, 43; cf. Mitchell 1993, I, 240.

35 For the evidence on the cult of Meter, the Mother Goddess, in Lydia and Ionia, see Roller 1999, 126-132.

36 Mitchell 1993, II, 22-24.

37 Hanfmann 1983, 228, suggests that the Lydian Zeus may have been a descendant of the older Hittite weather and storm god, a hypothesis that may well be valid for Zeus in other areas of Anatolia as well.

38 Note the many different Zeus cults, signified by different epithets, discussed by Drew-Bear and Naour 1990 and Mitchell 1993, II, 22-24.

39 CCIS II, no. 25.

40 Stele from Maionia, CCIS II, 39; from Koloe, CCIS II, 41; and from Philadelphia, CCIS II, 42. One additional relief of Sabazios, CCIS II, no. 45, depicts the god on horseback, riding towards a small altar with a krater at its base; behind the altar is a tree with an eagle on one branch holding a snake in its mouth. The god is depicted as a mature male figure who holds a group of lightning bolts in one hand. Both the appearance and the attribute held by the figure are standard markers of Zeus, leading further credence to Sabazios’ identification with Zeus, but the Anatolian provenience of the piece is not secure, so it is not included in this discussion.

41 The term “Phrygian cap” appears in Roman literature of the first century CE and later; see Vergil, Aeneid 4.215, 9.616, and Juvenal 6.516. On the origin of the cap and the reasons for its attribution to Phrygia, see Roller 1994, 249-250.

42 For a discussion of the various attributes of Sabazios, see CCIS III, 23-37.

43 Asgari et al. 1983, 125, no. B 346; CCIS II, no. 77. The piece was reportedly found in a sanctuary, but not in controlled archaeological excavation. It has not yet been fully published.

44 The Sabazios votive hands have been collected in CCIS I.

45 The costume and pose of the god can be seen in reliefs and statuettes from Italy (CCIS II, nos. 80-82 from Rome; no. 83, from Volsinii illustrates a bust of the god wearing the “Phrygian” cap, Spain (CCIS II, no. 85), Gaul (CCIS II, no. 86), Dalmatia (CCIS II, no. 87), Moesia (CCIS II, no. 78), and Pannonia (CCIS II, no. 88). Figurines and busts of Sabazios with the same costume and pose also appear on numerous bronze hands dedicated to the divinity, from Italy (CCIS I, no. 8, 10, 12-15, 19, 27, 29), Sicily (CCIS I>, nos. 20-25), Sardinia (CCIS I, nos. 33, 36, 37), Spain (CCIS I, no. 39), Gaul (CCIS I, no. 43), Helvetia (CCIS I, no. 47), Germany (CCIS I, no. 50), and several others of unknown provenience (CCIS I, nos. 65, 68, 71, 79, 80, 83, 86, 87, 92).

46 On the history of the costume and its use in Greek art, see Miller 1997, 171, and Roller 1994, 249-251, with earlier bibliography.

47 For the application of this costume to Attis, see Roller 1994, 251-253.

48 For examples, see note 45. This discussion of the creation and dissemination of the Sabazios visual iconography is inevitably tentative, since few of the Sabazios figurines can be securely dated and many do not have a secure provenience.

49 Roller 1994.

50 One possible occurrence may be a Paleo-Phrygian text from Midas City, which preserves the word Sabas carved into one of the rock faces of the upper citadel, Brixhe and Lejeune 1984, no. M-08. The reading of the word is quite uncertain, however, and the five letters seem to form the end of a longer text; thus it is unclear if this is a distinct name or part of an unrelated word.

51 Keil and Premerstein 1911, 94, no. 2; CCIS II, no. 41; Frei 1988, 19-23, no. 9N. See above, notes 22, 33.

52 CCIS II, no. 41, from Koloe.

53 I am using the word “foreign” here in the sense of non-Greek. The remarks of Parker 1996, 152-198, on foreign cults in Athenian religious practice offer a valuable overview on the variable status of foreign divinities in Athens.

54 See the discussion in Roller 1999, 121-141.

55 On the origin of Attis, see Roller 1994.

56 The citations of Sabazios in Aristophanes’ comedies are given above, n. 7. Sabazios as a Phrygian, Horae, frag. 566 (PCG 296); Σ Aristophanes, Birds 874.

57 On the frequency of Phrygian slaves in Athens, see Hermippos, Phormophoroi, line 18 (PCG frag. 63); Aristophanes, Wasps 433, Lysistrate 1211, Frogs 1345; Euripides, Alcestis 675; cf. Strabo 7.3.12. In general, see Long 1986, 108-110 and Bäbler 1998, 156-163.

58 According to Pindar, Ol. 1.24, Pelops was a Lydian, while Herodotos, 7.11.4, called him a Phrygian. See also Euripides, Alcestis 675, a Lydian or Phrygian slave mentioned as equivalents; and Timotheus, Persians 115-150, a Phrygian slave who comes from Sardis, near Mt. Tmolos (both in Lydia).

59 Aristophanes, in Lysistrate 387-388, mentions rites for Adonis and Sabazios together, strongly implying that to contemporary Athenians, the two seemed much the same. On ecstatic rites in the cults of non-Greek divinities, see Versnel 1990, 102-130, esp. 114-118 on Sabazios.

60 As attested most vividly in Euripides, Bacchae, lines 64-168.

61 On the conflation of rites for Dionysos and Sabazios, see Aristophanes, Σ Birds 873, ΣWasps 9; Alexander Polyhistor, FGrHist 273 F 103; Diodoros 4.4; Cicero, De Natura Deorum 3.23; Strabo 10.3.15; Harpocration, s. v. Σαβοί; Souidas, s. v. Σαβάζιος; Etymologicum Magnum, s. v. Σαβάζιος. Sabazios is considered the equivalent of Dionysos by some modern scholars as well, e.g. Nilsson 1955, 566; Versnel 1990, 114.

62 As Eugene Lane has correctly pointed out, CCIS II, 52, testimonium D1; CCIS III, 2-3.

63 See the comments in Roller 1994, 255 on the cry ὑῆς ἄττης in the same passage, and its relationship to the pseudo-Anatolian deity Attis.

64 Parker 1996, 159.

65 Cicero, De Legibus 2.37; on this passage, see Parker 1996, 158 n. 20.

66 Theophrastos, Characters 27.8, the Late Learner; 16.4, the Superstitious Man.

67 A snake can be seen on the relief from Balat, CCIS II, no. 25; from Maionia, CCIS II, no. 39; and from Koloe, CCIS II, no. 41. The snake also appears on a number of the Sabazios hands (although none from Anatolia), see CCIS III, 23.

68 The evidence is collected in CCIS I, no. 5, and CCIS II, nos. 46-55.

69 IG ii2 1325 = CCIS II, no. 51.

70 Kontorini 1983, 71-79 = CCIS II, no. 46.

71 For a discussion of this issue, see Lane’s comments in CCIS III, 39-48.

72 Quandt 1973, 35, Hymn 48; see West 1983, 96-97, 106-107

Abbreviations

ARV2 = J. Beazley, Attic Red-Figure Vase Painters, 2nd ed. (Oxford 1963).

CCIS I = M. J. Vermaseren, Corpus Cultus Iovis Sabazii vol. I. Ètudes Préliminaires aux Religions Orientales dans l’Empire Romain 100 (Leiden 1983).

CCIS II = E. Lane, Corpus Cultus Iovis Sabazii vol. II. Ètudes Préliminaires aux Religions Orientales dans l’Empire Romain 100 (Leiden 1985).

CCIS III = E. Lane, Corpus Cultus Iovis Sabazii vol. III. Ètudes Préliminaires aux Religions Orientales dans l’Empire Romain 100 (Leiden 1989).

FGrHist = F. Jacoby, Fragmente der griechischen Historiker (Leiden 1926-58).

PCG = Kassel, R. and C. Austin. edd., Poetae Comici Graecae (Berlin 1983--).

Bibliography

Allen, R. E. 1971. “Attalos I and Aigina,” Annual of the British School at Athens 66, 1-12.

Allen, R.E. 1983. The Attalid Kingdom, a Constitutional History (Oxford).

Asgari, N., B. Madra, and M. Soysal. 1983. Anadolu Medeniyetleri II (Istanbul).

Bäbler, B. 1998. Fleissige Thrakerinnen und wehrhafte Skythen. Nichtgriechen im klassischen Athen und ihre archäologische Hinterlassenschaft (Stuttgart and Leipzig).

Bérard, C., C. Bron, J.-L. Durand, F. Frontisi-Ducroux, F. Lissarague, A. Schnapp, and J.-P. Vernant. 1989. A City of Images: Iconography and Society in Ancient Greece. Translated by D. Lyons. (Princeton). Originally published as Une cité des images (Lausanne 1987).

Brixhe, C. and M. Lejeune. 1984. Corpus des inscriptions paléo-phrygiennes (Paris).

Buckler, W. H. 1914-1916. “Some Lydian Propitiatory Inscriptions,” Annual of the British School at Athens 21, 169-183.

Burkert, W. 1985. Greek Religion, translated by J. Raffan (Cambridge, MA).

Drew-Bear, T. and C. Naour. 1990. “Divinités de Phrygie,” Aufstieg und Niedergang der römischen Welt II.18.3 (Berlin and New York), 1907-2004.

Eisele, 1909-1915. Ausführliches Lexikon der griechischen und römischen Mythologie, s. v. Sabazios (Leipzig, repr. Hildesheim 1965), 232-263.

Fellman, R. 1981. “Der Sabazios-Kult,” in Die Orientalischen Religionen im Römerreich, ed. M. J. Vermaseren (Leiden), 316-340.

Fol, A. 1998. “Pontic Interactions: the Cult of Sabazios,” in The Greek Colonisation of the Black Sea Area, ed. G. R. Tsetskhladze (Stuttgart), 79-84.

Frei, P. 1988. “Phrygische Toponyme,” Epigraphica Anatolica 11, 9-33.

Hanfmann, G. M. A. 1983. “On the Gods of Lydian Sardis,” in Beiträge zur Altertumskunde Kleinasiens, ed. R. M Boehmer and H. Hauptmann (Mainz), 219-231.

Johnson, S. E. 1968. A Sabazios Inscription from Sardis, Religions in Antiquity, ed. J. Neusner (Leiden).

Johnson, S.E. 1972-1987. “The Present State of Sabazios Research,” Aufstieg und Niedergang der römischen Welt (Berlin) II..17.3, 1583-1613.

Keil, J. and A. von Premerstein. 1911. Bericht über eine zweite Reise in Lydien. Denkschriften der kaiserlichen Akademie der Wissenschaften in Wien, Phil-hist. Klasse 542 (Vienna).

Kontorini, V. 1983. Inscriptions inédites relatives à l’histoire et aux cultes de Rhodes au IIe et au Ier s. av. J.-C. Rhodiaka I (Louvain).

Lane, E. 1975. “Two Notes on Lydian Topography,” Anatolian Studies 25, 105-110.

Lane, E. 1979. “Sabazius and the Jews in Valerius Maximus: a Re-examination,” Journal of Roman Studies 69, 35-38.

Lane, E. 1980. “Toward a Definition of the Iconography of Sabazius,” Numen 27, 9-33.

Long, T. 1986. Barbarians in Greek Comedy (Carbondale and Edwardsville).

Matheson, S. B. 1995. Polygnotos and Vase Painting in Classical Athens (Madison, Wis.).

Miller, M. C. 1997. Athens and Persia in the Fifth Century BC (Cambridge).

Mitchell, S. 1993. Anatolia. Land, Men, and Gods in Asia Minor (Oxford) I-II.

Naumann, F. 1983. Die Ikonographie der Kybele in der phrygischen und der griechischen Kunst. Istanbuler Mitteilungen, suppl. 28 (Tübingen).

Nilsson, M. P. 1955. Geschichte der griechischen Religion I, 3rd ed. (Munich).

Ohlemutz, E. 1940. Die Kulte und Heiligtümer der Götter in Pergamon (Würzburg), repr. Darmstadt 1968.

Quandt, W. 1973. Orphei Hymni, reprint of 1941 edition (Berlin).

Parker, R. 1996. Athenian Religion, a History (Oxford).

Picard, C. 1961. “Sabazios, Dieu thraco-phrygien: Expansion et Aspects nouveaux de son Culte,” Revue Archéologique, 129-176.

Robert, L. 1948. “Inscriptions grecques de Lydie,” Hellenica 6 (Paris).

Robert, L. 1975. Une nouvelle inscription grecque de Sardes: Règlement de l’autorité perse relatif à un culte de Zeus, Comptes Rendus de l’Académie des Inscriptions et Belle-Lettres, 306-330.

Robert, J. and L. 1983. Fouilles d’Amyzon en Carie. Explorations, histoire, monnaies et inscriptions (Paris).

Roller, L. E. 1989. “Foreign Cults in Greek Vase Painting,” Proceedings of the Third Symposium on Ancient Greek and Related Pottery, ed. J. Christiansen and T. Melander (Copenhagen) 506-515.

Roller, L. E. 1994. “Attis on Greek Votive Monuments: Greek God or Phrygian?” Hesperia 63, 245-262.

Roller, L. E. 1999. In Search of God the Mother: the Cult of Anatolian Kybele (Berkeley and Los Angeles).

Schaefer, 1920. Realencyclopädie der classischen Altertumswissenschaft I A2, s. v. Sabazios, 1540-1551.

Simon, E. 1953. Opfernde Götter (Berlin).

Sokolowski, F. 1979. “Τὰ ἔνπυρα: On the Mysteries in the Lydian and Phrygian Cults,” Zeitschrift für Papyrologie und Epigraphik 34, 65-69.

Versnel, H. S. 1990. Ter Unus. Isis, Dionysos, Hermes. Three Studies in Henotheism. Inconsistencies in Greek and Roman Religion I (Leiden).

Welles, C. B. 1934. Royal Correspondence in the Hellenistic: A Study in Greek Epigraphy (New Haven).

West, M. L. 1983. The Orphic Poems (Oxford).