A Religious Association at Sardes

A RELIGIOUS ASSOCIATION AT SARDES Abstract: An inscription at Sardes, carved in the second century AD, preserves a dedication of a statue by an Achaemenid governor (fifth or fourth century BC) and two unusual injunctions banning the therapeutai who tend the statue from participating in mysteries of Sabazius, Agdistis, and Ma. The paper argues that the two injunctions date from Roman Imperial times, the privileged therapeutai reacting against cults that they regarded as novel and exotic. An inscription at Sardes, discovered in 1974 on a building block that was not in situ, offers a text on sacred matters, engraved by a single hand in the second century AD.1 Of its three sentences, the first is a dedication of a statue, dating originally from Achaemenid times; the second and third bar some cult officials from participation in certain rites: 4 8 12 ἐτέων τριήκοντα ἐννέα Ἀρταξέρξεω βασιλεύοντος, τὸν ἀνδριάντα Δροαφέρνης Βαρ<ά>κεω Λυδίης ὕπαρχος Βαραδατεω Διί. ❦ προστάσσει τοῖς εἰσπορευομένοις εἰς τὸ ἄδυτον νεωκόροις θεραπευταῖς αὐτοῦ καὶ στεφανοῦσι τὸν θεὸν μὴ μετέχειν μυστηρίων Σαβαζίου τῶν τὰ ἔνπυρα βασταζόντων καὶ Ἀνγδίστεως καὶ Μᾶς. προστάσσουσι δὲ Δορατῃ τῷ νεωκόρῳ τούτων τῶν μυστηρίων ἀπέχεσθαι. ❦ 4 ΒΑΡΛΚΕΩ lapis In year 39 of Artaxerxes as king, Droaphernes son of Barakes, governor of Lydia, (dedicated) the statue to Baradateo Zeus. He orders that his temple-warden devotees who enter the inner sanctum and crown the god are not to share in mysteries of Sabazius of the censerbearers(?),2 or of Agdistis, or of Ma. They order Dorates the templewarden to abstain from these mysteries. 1 Robert (1975), with photograph of squeeze [SEG XXIX 1205; G.H.R. Horsley, New DocumentsIllustratingEarlyChristianityI, no. 3; E. Lane, CorpuscultusIovisSabazii, no. 31; M.J. Vermaseren, CorpuscultusCybelaeAttidisque, no. 456; Paz de Hoz (1999) 130, no. 2.1; F. Canali de Rossi, I.Estremo oriente 235]; Briant (1998), from autopsy, with photographs. 2 Or ‘carriers of burnt offerings’. For ‘censers’ see Sokolowski (1979); followed by Jones (1990) 55, who urges the same meaning in a text dated to AD 155. AncientSociety 44, 1-23. doi: 10.2143/AS.44.0.3044797 © 2014 by Ancient Society. All rights reserved. 2 K. RIGSBY The Achaemenid dedication derives from the 420s or the 360s BC (Artaxerxes I or II), copied now in Roman times.3 Robert took the second sentence to be contemporary with the first, continuing it, but the third to be later, perhaps contemporary with the stone itself (II AD). The importance of the inscription was spelled out by Robert with great clarity. His account of the matter was that Βαραδαταῖος is Zeus ‘law-giver’, a statue of Ahura Mazda. The dedication reveals a weakening of the Zoroastrian prohibition of religious imagery. The exclusion from some Anatolian cults represents an effort by a Persian governor to prevent the local Persian elite from going native. This account has been challenged in two important contributions. Gschnitzer argued that Βαραδατεω is not a Roman-era grapheme for Βαραδαταίῳ but a correct Ionic genitive Βαραδατέω, ‘Zeus of Baradates’.4 On the analogy of later instances of such genitives in Anatolian cults, he held that this indicated a private and familial cult of Zeus, founded by someone named Baradates. The implication is that the god was not Ahura Mazda but Zeus, and the statue carries no implications for the history of Zoroastrianism. Briant accepted this interpretation of Baradateo and the irrelevance of Zoroastrianism and offered a wider-ranging study.5 He argued that the second and third sentences, set off by leaves, are separate from the first. What was dedicated in the first sentence was an andrias, thus a statue of a human being, not of Zeus, so there is no religious content at all; the two religious injunctions that follow on the stone, different in genre from the dedication, are later acts, Hellenistic-Roman, and unrelated to the Achaemenid text: rather the stone contains a random compilation of old texts from the temple. The Achaemenid dedication was his focus, and he called for a separate study of the two religious injunctions.6 3 For the fourth-century date: Robert (1975); Chaumont (1990); Debord (1999) 367374; Fried (2004) 129-137. Fifth century: M. Weiskopf in: Hanfmann (1983) 256 n. 10. Briant (1998) leaves the question open. On the anomalous form of the date in line 1 see Thonemann (2009) 392-393. 4 Gschnitzer (1986); this has been widely followed, e.g. by Debord (1999) 367-374; Frei (1996) 24-26, 90-96. 5 Briant (1998); followed e.g. by Dusinberre (2003) 100, 118; but rejected by Debord (1999) 367-374 (who is followed by Fried (2004) 24-26), maintaining that all three sentences are contemporaneous and Achaemenid. 6 Briant (1998) 223 n. 45. A RELIGIOUS ASSOCIATION AT SARDES 3 I argue here that the two injunctions are indeed later than the Achaemenid dedication, of Roman Imperial date, but that the three sentences are in fact related and all concern a statue of Zeus. THE DEDICATION Robert took the reasonable view that the first and second sentences were continuous: ‘he (dedicated)… he orders’. But Briant’s chronological separation of the two is necessary. We have two different genres of text, first the dedication and second the two prohibitions that are framed by good Roman leaves.7 The unexpressed verb of the first sentence would be understood as aorist, as was normal (typically ἀνέθηκεν when expressed); the two laws by contrast use the present tense. Moreover, to separate the two sacred laws, the second and third sentences, from each other creates difficulties. The two are parallel in structure, vocabulary, and intent. The expression ‘these mysteries’ in the third sentence was evidently felt to be a sufficient reference: it can only refer back to the second sentence, without which the third would be unintelligible. The stone as a whole derives from two texts, a dedication of Achaemenid date and a later pair of rulings about the therapeutai. Briant held that an andrias must be of a person,8 so not a statue of Zeus, but a statue dedicated to Zeus. The Greek formulation is familiar: one dedicated (a statue of) a person to a god. But Briant had then to admit that the person honored and portrayed by the statue is not mentioned in the inscription.9 That would be astonishing. To the contrary, we have clear examples of an andrias of a god. This was common enough in Roman times, as Robert noted, in Egypt and 7 Briant (1998) 211, recognized the leaf closing off the third sentence. In an inscription of Imperial date, a leaf divides Greek and Phrygian texts: Drew-Bear e.a. (2008) 109-110. 8 Briant (1998) 215-220. This was normal; in the authors note Cassius Dio’s conscientious ἀνδριάντας ... μᾶλλον δὲ ἀγάλματα of Antinous (29.11.4; cf. IG VII 2712.100102). Debord (1999) 370, doubted the rigor of this vocabulary; but the basis for doubt is worth illustrating. 9 “The absence of the name of the person so represented does not seem to me to constitute a particularly impressive argument: a votive statue can be anonymous” (Briant (1998) 219, giving no example). For a statue like I.Cret. III III 23, Τ. Λάρκιος Κυδικλῆς τὸν θεῖον, we can deduce that uncle and nephew had the same name. 4 K. RIGSBY elsewhere.10 Late texts are perhaps poor support for an instance of Achaemenid date.11 But at all dates one finds dedications of statues to gods in which, as in the Sardes inscription, the andrias is not defined by a genitive. Yet in these instances the statue is of the god, e.g.: Astypalaea ca. 400 BC, Ἀνα[ξί]τιμος Εὐπόλι[ος [τ]ῶι Ἀπόλ[λω]νι δεκάταν ἀνέθηκε τὸν ἀνδριάντα καὶ τὸν οἶκον κατασκευώσας (Peek, I.Dorisch.Ins. no. 89 (AbhLeipzig 62.1 (1969) 43) Cyprus, early IV BC, [Βααλρωμος] ὁ Ἀβδιμίλκων τὸν ἀδργάταν τόδε κατέστασε… τῷ Ἀπόλωνι τῷ Ἀμύκλωι (Masson, ICS 220 = Kition/Bamboula V T69) Cimmerian Bosporus, Hellenistic, Λεύκων Παιρισάδου ἀνέθηκε τὸν ἀνδριάντα Ἀπόλλωνι [Ἰ]ητρῶ[ι ἱ]ερησάμενος and Κάλλων Κάλλωνος [ἱερ]ησάμενος ἀνέθηκεν τὸν [ἀν]δριάντα Ἀπόλλωνι ᾿Ιητρῶι (CIRB 25, 1044) Thera, Hellenistic, two brothers Καρτίνικος Ἄνθης Θεάνορος τὸν ἀνδριάντα Διονύσωι (IG XII.3 419) Egypt, late Hellenistic, Ἐρμοῦθει θεᾷ μεγίστῃ καὶ Ἀνύβι θεῷ μεγάλῳ τὸν ἀνδ[ρι]άντ[α] Πυθιάδ[η]ς in behalf of himself and his wife and children (I.Fayoum III 160) Roman Lycia, three agoranomoi of Bubo, upon leaving office, κατεσκεύασαν τὸν ἀτριάντα… μεγίστῳ θεῷ Ἄρει καθὼς ὑπέσχοντο (F. Schindler, I.Bubon 4, cf. 3 = C. Kokkinia, Boubon (Meletemata 60), Athens 2008, nos. 71, 72). The natural reading of such texts is that the andrias portrayed the god named in the dative, and not an unnamed person, or the human dedicators themselves. The sentence structure is the same when ἄγαλμα is 10 Robert (1975) 313, cited MAMA IV 275A τὸν ἀνδριάντα τοῦ ἀλεξικάκου Ἀπόλλωνος (Dionysopolis, AD 275); see also e.g. I.Hadrianoi 11 τοῦ Διὸς τὸν ἀνδριάντα; Diod. 20.14.6 ἀνδριὰς Κρόνου χαλκοῦς (Tyre). In Egypt: IGR I 1051 τοῦ μεγάλου Σαράπιδος, OGIS 708 τοῦ Πολιέως Σαράπιδος (both Alexandria); IGR I 1136 Σούχου μεγάλου; E. Bernand, I. métriques 106 ἀνδρειὰς μὲν ὅδ’ ἐστι Διός; cf. Amm. Marc. 17.4.21 τῶν θεῶν ἀνδριάντας ἀνέθηκεν. Agalma of a human: e.g. I.Ankara 115.19 (A.D. 150-175). 11 The Colossus of the Naxians, cited by Robert (1975) 313, has [τ]ο͂ αϝυτο͂ λίθο ἐμὶ ἀνδριὰς καὶ τὸ σφέλας (I.Délos 4; LSAG pl. 55). Briant (1998) 218 n. 27, invokes Hermary (1993) 11-27, as undermining the point; but Hermary, arguing that the extant statue is a late fifth century BC replacement, did not modify the substance or archaic date of the original inscription. This andrias, more than eight meters tall, did not portray a human being. This is the only instance of a god’s andrias cited in LSJ. On the inscription see also Butz (2009). A RELIGIOUS ASSOCIATION AT SARDES 5 used, with no genitive needed to identify the statue as the god’s: e.g. Φιλαινοὺ Θεογενεία ἱερειτεύονσα τὸ ἄγαλμα Ματρὶ Θεοῦν ἀνέθεικε (Phalanna, III BC: ArchEph (1916) 18 no. 271). For the interchangeability of the two terms, a locus classicus is Pindar, ἀνδριαντοποιός… ἐργάζεσθαι ἀγάλματα (Nem. 5.1-2; cf. Plut. Luc. 20.1 ἱεροὺς ἀνδριάντας). Droaphernes dedicated a statue of Zeus. Zeus, however, still presents a problem. Baradateo has not been convincingly explained.12 If with Robert we take the word as a divine epithet, Zeus ‘Law-giver’,13 the word order is abnormal. Robert (1975: 313) gave examples of the uncommon sequence epithet-name; but for that sequence the article is usually required.14 If with Gschnitzer we see ‘Zeus of Baradates’, a family cult with Baradates understood as the man who founded the cult,15 this is a surprisingly early instance of a usage that first appears in inscriptions only in Imperial times.16 Strabo shows it at least by the Augustan period (Μὴν Φαρνάκου in Pontus, 12.3.31),17 and Gschnitzer would trace that cult to Achaemenid times. But for this usage also, the sequence genitive-nominative would be surprising. These difficulties with the text lead me to imagine an extreme solution: that Βαραδατέω is a gloss. These are frequent enough in 12 The Iranian origin of both this name and Barakes has been questioned: on Baradates see Frei (1996) 25, Chaumont (1990) 580-581, and Briant (1998) 214 n. 17. Paz de Hoz (1999) 177, proposed emending to Bagadates; so too Mitchell (2007) 158. Barakes in one of the Avroman parchments: Mayrhofer (1974) 209, who tentatively urged an Iranian root, tentatively followed by Briant (1998) 206 n. 3; earlier, however, Minns (1915) 44, called Barakes “perhaps Semitic”, son of Μαιφόρρης “probably Semitic”, brother of Σωβήνης “perhaps Semitic”. 13 The concept is obvious enough, whether Persian or Greek; at Sinope, Zeus is once hailed as δικαιόσυνος (I.Sinope 120, late Hellenistic), and this epithet was known to the authors (CGF VIII 586). 14 The omission sometimes occurs in magistrate titles (Didyma 252 προστάτης Διδυμέως Ἀπόλλωνος, 281 στεφανηφοροῦντος Διδυμέως Ἀπόλλωνος, I.Sardis 47 ἱερέα μεγίστου Πολιέως Διός; cf. eponymity of the sort ἐπὶ δημιουργοῦ, ἐπὶ κόσμου). An early instance in a dedication is hελλανίοι Διί on a jar rim (IG IV2 1056, early V BC). 15 This explanation of such genitives was first urged by von Gutschmid (1892) 497, and was confirmed by Keil & von Premerstein (1911) 104; explicit in the dedication SEG LVI 1434. Cf. Petzl (1994) 103; Versnel (2011) 64 n. 152. 16 To Gschnitzer’s references add SEG XXX 622 (Thessalonica: Ζεὺς Διόνυσος Γονγύλου); Corsten (2003) 122 n. 46; van Bremen (2010). 17 Cf. the tantalizing reference to Διὶ Φαρναουα in Cappadocia (SEG LII 1464 ter.16, with Debord 2005). 6 K. RIGSBY manuscripts, but rare in inscriptions.18 The similarity of the two names in line 4 is suggestive.19 The stone offers a copy of a text, not the original. And the mason probably was not standing next to the Achaemenid stone as he carved his own, but had a copy of the old text written on papyrus (or whatever medium) — that is, we possess a copy of a copy. Imagine that the Achaemenid stone was hard to read and that on the papyrus copy the patronym was botched (as it is on the extant stone, ΒΑΡΛΚΕΩ); that this line ended with ὕπαρχος; and that a correction had been added in the margin of the papyrus, thus: ΕΤΕΩΝΤΡΙΗΚΟΝΤΑΕΝΝΕΑ ΑΡΤΑΞΕΡΞΕΩΒΑΣΙΛΕΥΟΝΤΟΣ ΤΟΝΑΝΔΡΙΑΝΤΑΔΡΟΑΦΕΡΝΗΣ ΒΑΡΛΚΕΩΛΥΔΙΗΣΥΠΑΡΧΟΣ ΒΑΡΑΔΑΤΕΩ ΔΙΙ The mason misunderstood, and copied the whole fourth line, both the error at the beginning and the correction at the end. The original wording would then have been Δροαφέρνης Βαραδάτεω Λυδίης ὕπαρχος Διί. This would be unexceptional as Greek and as a gesture: a Persian grandee, Droaphernes son of Baradates, dedicates a statue of Zeus. The strictures of Zoroastrianism against statues of gods are then irrelevant (as Gschnitzer and Briant argued). And the dedication offers no evidence for a private or family cult — it is simply a statue of Zeus. But one can hardly press such a hypothetical proposal. It is not evident how ΑΔΑΤ might have come to be miswritten as ΛΚ. And if the father was instead Ba<g>adates, as has been urged, then the putative corrector perpetuated the error Baradates. About the Achaemenid event, then, let us conclude only that what the Persian governor dedicated was a statue of Zeus. This is pertinent to the interpretation of the regulations that follow on the stone. 18 A parallel would be the Hipponium version of the much-copied ‘Orphic’ poem for the dead (photographs at Sacco (2001); for the text, Giannobile/Jordan (2008)): the third line ends ΚΥΠΑΡΙΣΟΣ Σ. The marginal sigma is a correction, intending that the proper ΚΥΠΑΡΙΣΣΟΣ be written; ΚΥΠΑΡΙΣΟΣ Σ was in the immediate archetype, which the copyist reproduced without comprehending the corrector’s intent. In the epigram G. Petzl, I.Smyrna 766.2 (II AD), Μέλητα ποταμόν, Boeckh saw that the extrametrical ποταμόν must be expelled, and Petzl (to whom I owe the reference) saw that it is a gloss. Cf. Herrmann & Malay (2007), no. 85 and p. 115; Petzl (2011) 53. 19 Gschnitzer (1986) 48, however, saw this as evidence that Droaphernes belonged to the family that founded the cult. But Chaumont (1990) 581, wondered whether there is some textual confusion in the two names. A RELIGIOUS ASSOCIATION AT SARDES 7 THE PROHIBITIONS The second and third sentences are of the same genre. But what is their relation to each other, and to the first sentence, the Achaemenid dedication? Robert reckoned that the second sentence derives from the same document as the dedication that precedes it on the stone, hence Achaemenid in date and origin, while the third is later. As Briant saw, the two prohibitions belong together and are post-Achaemenid. But if we give up Robert’s date for the second sentence, we also give up his context and explanation (a governor trying to keep Persians from going native/ Anatolian); and some other explanation is needed. Robert considered the second and third sentences to be a summary rather than the original text.20 Certainly they have been truncated, in that their subjects have been omitted by the copyist; αὐτοῦ in line 8, ‘his therapeutai’, implies that the god had been named earlier in the original text.21 But they can be called summary only in the sense that they are extracts.22 So far as we have them, these can be the exact words of the original. Who gives orders about ritual propriety? A Persian provincial governor, as Robert thought, is conceivable. But the most frequently met authority is a god, through an oracle or a vision, as in the κατ’ ἐπιταγήν / iussu so often seen in dedications.23 The verb προστάσσειν is found of oracular responses; e.g., a poem at Delphi reported - - - Φοῖ]βος . . . προσέταξεν (FD III.1 560). It is used in the healing cults, concrete instructions from Asclepius: προσέταξε I.Cret. I XVII 7, 9, 11; cf. ἐκέλευσεν IG IV.12 126.4; Plut. Mor. 999C. And gods forbid: ἀπαγορεύει ὁ θεός· μὴ [ε]ἰσφέρειν (κτλ.) (Lupu, NGSL 4.7, Athens). But here this seems unlikely: τὸν θεὸν in the predicate of the sentence would seem to exclude the god as the subject of the sentence. The priest in charge of a cult can also give orders: προστάσσοντος τοῦ ἱερέως ἢ τοῦ ἀρχιβάκχου (LSCG 51.66, 20 Robert (1975) 317: “these lines do not translate the authentic document; they summarize its substance”. 21 Briant (1998) 22 n. 45, wishing to eliminate from the rulings any reference to Zeus, held that αὐτοῦ means ‘here’; that would be otiose, and the other references to this cult group (see below) are explicit, οἱ τοῦ Διὸς θεραπευταί. 22 Frequent enough in inscriptions: from sacred laws, I.Ephesos 10 (II/III AD, κεφάλαιον νόμου πατρίου), two extracts with infinitives depending on some main verb that has been omitted; LSAM 12 (Pergamum: Hellenistic), beginning with δέ (as does IGLSyrie VII 4028D (K. Rigsby, Asylia 218), an extract from a civic decree). 23 The corpus of exvisu inscriptions has been compiled by Renberg (2003). 8 K. RIGSBY Athens); ἐὰν δέ τις ἱέρεια πλείω τῶ[ν γεγρα]μμένων ἐν τῶι νόμωι προστάσσηι (LSCG 107.3, Ios); ὁ ἱερεὺς τοῦ Ἀπόλλωνος τοῦ Ἐριθασέου π[ρ]οαγορεύει καὶ ἀπαγορεύει κτλ. (LSCG 37, Athens).24 Or god and priest together: ἀπαγορεύει δὲ καὶ ἡ θεὸς κ[αὶ ὁ προφήτης] Καλλίστρατος μηθένα ὀ[ρ]γ[εῶνα κτλ. (IG II2 1289.9). So a divine commandment — if not uttered by the god then by his agent — might be the source of the two sentences. But this does not explain the shift from singular verb to plural in the last sentence (this objection would apply also to commands of Droaphernes). Another possibility: in inscriptions as in the Attic orators,25 what most often ‘orders’ (and in the enduring present tense) is the Law, singular or plural. Thus, of sacred matters, ὁ νόμος προστάττει (SEG II 9.5, a thiasos on Salamis); or κελεύει, on oversight of cult (LSCG 32.18) and in a decree of orgeones(LSCG 46.11);26 and in secular rules, ὁ νόμος προστάσσει (Hellmann, Choix d’inscriptions 2.15: the Pergamene astynomoi law). The plural οἱ νόμοι προστάττουσι is common in contexts both religious (SEG XLIV 60.5, a thiasos of Bendis, Athens) and secular (IG II3 359.14; XI.4 1052.9; II2 1008.72 ἀπαγορεύουσιν; V.1 1379.5-6 κελεύοντι; I.Priene 23.6 συντάσσουσιν). In the orators the subject can be omitted: Lys. 1.32, ἀνάγνωθι δέ μοι καὶ τοῦτον τὸν νόμον … ἀκούετε, ὦ ἄνδρες, ὅτι κελεύει. Shifting between singular and plural is found in earnest declarations about the law. So in Isaeus 11: 11-13 ὁ νόμος δηλώσει — τῶν νόμων δεδοκότων, 23-25 ὁ νόμος — τῶν νόμων, 29-30-31 οὐ δίδωσι ὁ νόμος — τῶν νόμων — τοὺς νόμους. Accordingly, at Sardes, ‘the law/laws’ might be the missing subjects of both sentences. Finally, it may be that the neokoroi therapeutai, or a committee of them,27 decided that the law does not allow such participation, and then they signaled this finding to Dorates, whose behavior may have prompted the question. The original document might then have had the structure of 24 See Hitch (2011) 122-124, on the inscribing of priestly exegeses. The orators’ argumentative citation of what the law demands: e.g., Isae. 10.2.13 (κελεύει), 12 (κατὰ τὸν νόμον ὃς οὐκ ἐᾷ); Demosthenes 23.63 (τοῦ νόμου λέγοντος); the new Hypereides (ἀπαγορεύουσιν οἱ νόμοι: Tchernetska (2005) 2); Andoc. 1.110, 115-116. Cf. a reported law of Solon: ἐὰν μὴ ἀπαγορεύσῃ δημόσια γράμματα (Dig. 47.22.1 pr. 1). The expression ‘the law commands’ is frequent in the Ath.Pol., indifferently singular or plural. 26 Cf. NGSL 26.19 (Nacone); IG V.1 5.12, 16 (Sparta); IG XII.7 241.19 (Minoa); I.Lampsakos 9.35, 61. 27 Cf. I.Ephesos 2, οἱ προήγοροι ὑπὲρ τῆς θεοῦ κατε[δι]κάσαντο. 25 A RELIGIOUS ASSOCIATION AT SARDES 9 a statement of their decision about the law followed by a statement of their prohibition in the present case, with the law the subject of the first extant verb and the therapeutai the subject of the second: e.g. (οἱ θεραπευταὶ/ ἔκδικοι ἔκριναν ὅτι ὁ νόμος) προστάσσει… προστάσσουσι δὲ Δορατῃ τῷ νεωκόρῳ… In sum, it is possible that the two prohibitions can have been quoted from a single original document as well as from two of different dates. On this reading, a question has arisen for this association: Can those who tend the statue of Zeus share in certain mysteries? Authority has been consulted, and the answer is: ‘The god/priest/law orders…’ Does this general finding apply to the neokoros Dorates? ‘The laws/the devoteesorder…’ I prefer to think that one text was the source of both sentences: if the question about these mystery cults had already been settled for the cult attendantsand applied to all of them, why was there need to commemorate in stone an individual decision about Dorates? On this theory, the two sentences were copied from a single decree of the cult association, and the excerptor has extracted from it for his purpose the two injunctions that were announced in it — the general legal finding and then its application to the current question. DATE AND CONTEXT The date of these two prohibitions needs to be established independently of the Achaemenid statue.28 Are there external controls on this question? Robert did not fail to see that this cult group has long been known. (a) They honored a prominent citizen of Sardes with a statue (I.Sardis 22): οἱ τοῦ Διὸς θεραπευταὶ τῶν εἰ[ς] τὸ ἄδυτον εἰσπορευομένων καθιερώσαντες ἐστεφάνωσαν Σωκράτην Πολεμαίου Παρδαλᾶν, τὸν πρῶτον τῆς πόλεως, διακείμενον ἐκ προγόνων πρὸς τὸ θεῖον εὐσεβῶς. The devotees of Zeus, of those who enter the inner sanctum, blessed and crowned Socrates Pardalas son of Polemaeus, first man of the city, by family tradition piously disposed toward the divine. 28 The name Dorates is too rare to point to a date. The one other instance in LGPN is on a seal at Cyrene (Δοράτους), probably of the first century AD: Maddoli (1963-64) 129, no. 924. 10 K. RIGSBY As to the date, a man of this name was eponymous priest of Dea Roma (in I.Sardis 91); that usage, introduced after the beginning of Roman rule in 129 BC, was replaced by the stephanephoros around the mid-first century AD.29 We do not expect to find someone called ὁ πρῶτος τῆς πόλεως before the Roman Empire.30 (b) Another Socrates, or the same man,31 paid for a temple and statue of Hera, which his granddaughter repaired at some date after the earthquake of AD 17 (SEG XXVIII 928). (c) The same group, οἱ τοῦ Διὸς μύσται καὶ θεραπευταί, honored someone around AD 100, and (d) an altar apparently of Zeus and mentioning τῶν θεραπευτῶν may be as early as the first century BC.32 Robert took these testimonia to show the long continuity of the group of therapeutai, from Achaemenid times (the Droaphernes dedication) to Roman Imperial. Rather, these other inscriptions can offer an approximate control on the date of our prohibitions — they may be roughly contemporary with these other documents of the therapeutai of Zeus,33 which range from perhaps as early 100 BC (I.Sardis 22) through the first century AD. Thus the documented continuity of this association is not so impressive — perhaps already in the first century BC, and certainly in the early Empire. Our two injunctions about mysteries need be no older than this; they may well be contemporary with each other and with the inscription itself in the second century AD. TENDANCE Ancient polytheism knew groups of subordinates or volunteers who tended sacred things. The terminology and the practice were old: statues and other equipage had always required maintenance. At Sardes 29 Herrmann (1996b) 57-58. E.g., IG IX.1 8, XII.5 292, XII Suppl. 447, I.Iasos 618, I.Stratonikeia 1031, MAMA VI 352, TAM II 146, 838f, 920, 1204 (a woman), TAMV 1835 with Petzl’s comments: none earlier than II AD. Cf. Plut. Mor. 679C, οἵ τε ἄρχοντες… καὶ οἱ πρῶτοι τῆς πόλεως. 31 So Hanfmann & Ramage (1978)178. Robert (1975) 321 n. 50 (doubtfully) and Paz de Hoz (1999) 177, followed the original editors in dating I.Sardis 22 ca. 100 BC, which was based on equating Socrates with a Socrates of that date (quoted below, n. 67). Herr– mann (1996a) 324 n. 29, leaves the question of the date of I.Sardis 22 open. 32 Herrmann (1996a) 321-329. 33 Cf. Herrmann (1996a) 326, 333; Mitchell (2007) 158 (“perhaps between the first century BC and the first century AD”). 30 A RELIGIOUS ASSOCIATION AT SARDES 11 itself, a neokoros of the city’s tutelary goddess Artemis was appointed πρὸς τὴν θεραπείαν καὶ εὐκοσμίαν τῶν κατὰ τὸ ἱερὸν τῆς παρ’ ἡμῖν Ἀρτέμιδος (I.Sardis 4.9, II BC).34 As to statues specifically: in Crete, some persons were thanked for caring for the ancient statues, τὰ ἀρχαῖα [ἀ]γάλματα θαραπεύσαντες (I.Cret. III II 1, late II BC). On Thasos, a retiring neokoros was honored because she had ‘painted and gilded the goddess’, [νεωκο]ρήσασα ἐνέκαυσε [καὶ ἐχρύσ]ωσε τὴν θεόν (I BC).35 At the Ptoion in Boeotia, a benefactor paid for τὴν τῶν ἀγαλμάτων ἐπαγάνωσιν καὶ θεραπείαν (I BC).36 These were onetime voluntary benefactions, duly praised. The therapeutai at Sardes, by contrast, were a continuing and privileged group, permitted to enter the adyton and tend the statue. But whether self-selecting or appointed by the state, they were, like the instances just mentioned, volunteers in their piety. In the Hellenistic period the noun therapeutes is found mostly of devotees of the Egyptian gods.37 In Roman Imperial times it is more widely used, and especially of those who tended divine statues and saw to their clothing, crowning, and cleaning.38 The verb θεραπεύειν is well attested,39 and in Christian polemic θεραπεύειν τὰ εἴδωλα became a commonplace.40 This could become a special privilege: πάντες οἱ περὶ τὸν θεὸν θεραπευταὶ καὶ τάξεις ἔχοντες (Aelius Aristides 48.47). The therapeutai of Zeus of Sardes would fit well with an Imperial date. 34 Cf. IG XII.4 348.47, τῶν οἰκημάτων ἢ τοῦ [τ]εμένευς θεραπειάς (Cos, late IV BC). 35 IG XII Suppl. p. 164, with Pouilloux (1954) 385. IG VII 4149, with Roesch (1982) 226 n. 2. 37 Vidman (1970) 66-72; Robert (1975) 319 n. 41; Pleket (1981) 159-160. 38 So already Foucart (1875) 333, on IG V.2 265, θεραπεύουσα καὶ συνευκ[ο]σμ[οῦ]σα… [πρ]ὸς τὰν τᾶς θεοῦ (Kore) τιμὰν καὶ κ[όσμη]σιν (‘year 85’, first century BC or AD). 39 SEGVIII 467.B.51 (Egypt, III BC), τοῖς τὸ ἱερὸν θεραπεύουσιν; IG II2 1365.26 of Mên (Roman Athens), θεραπευέ[τω] τὸν θεόν; Plut. Mor. 351E of Isis, ἁγνείας τε πάσης καὶ νεωκορίας ἔργον ὁσιώτερον, οὐχ ἥκιστα δὲ τῇ θεῷ ταύτῃ κεχαρισμένον, ἣν σὺ θεραπεύεις. 40 In the Acts of Paul and Thecla, τὰ εἴδωλα ἐθεράπευον (Lipsius, Acta apost. apocr. I 271); in [Justin Martyr] Quaest.resp., θεραπευθέντων τῶν εἰδώλων and τὰ εἴδωλα πανταχοῦ ἐθεράπευον (von Otto, Apol.Christ. V pp. 60, 204); in Barl.&Jos., πλείονος δὲ τιμῆς τοὺς τῶν εἰδώλων ἀξιοῖ θεραπευτάς τε καὶ νεωκόρους (p. 30 Loeb). Cf. Athanasius on pagan νεωκόροι τῶν εἰδώλων (Epist.encyc. 5.7, Hist.Arian. 54.3: Opitz II.1 p. 174, 214). 36 12 K. RIGSBY MYSTERY CULTS AT SARDES Participation in three mystery cults is proscribed for the neokoroi therapeutai. This list favors a date in the Roman Empire rather than earlier. Mysteries of Phrygian Sabazius were known in Athens already in the fourth century BC (Theophr., Char. 27.8). At Sardes, one other reference to the god is extant, a dedication of late Hellenistic date.41 The Attalid monarchs imported mysteries of ‘Zeus the Sabazius’ to Pergamum from Cyzicus towards the mid-second century BC (Royal Corr. 66-67); perhaps they were no earlier in coming to Sardes. But his mysteries were widespread and popular in Asia Minor by Imperial times. Roman Sardes must have long known the mysteries of Sabazius, though this is not otherwise on record. While old among the Greeks, the cult was often regarded as foreign and vulgar. For Aristophanes the rites are alien and irrational, women’s religion (Vesp. 9, Av. 874, Lys. 387-388 τῶν γυναικῶν τρυφή). Theophrastus’ initiate is the ‘Superstitious Man’. Greeks of a certain class evidently regarded the mysteries of Sabazius as both exotic and unmanly. This might help explain the prohibition in the Sardes inscription. But what seems more pertinent is its unparalleled qualification, ‘the mysteries of Sabazius of the censer-bearers’: whatever Sardians thought of traditional Sabazius mysteries, here apparently was a novel and distinct sect. Worship of Phrygian Agdistis is found in Athens already at the beginning of the Hellenistic period.42 But apart from the Sardes inscription, mysteries of Agdistis isto nomine are not attested at any date. Robert thought that these mysteries could be explained to the extent that Agdistis was understood as a name for the Mother of the Gods43 or for Cybele, to whom ecstatic worship was attributed by Greeks from an early date.44 These two goddesses are attested at Sardes as cults of the state, Cybele in the archaic period and Meter in the Hellenistic,45 but under their 41 E. Lane, CorpuscultusIovisSabazii II no. 30 (a text that awaits explication). In Peiraeus, a dedication Ἀνγδίστει καὶ Ἄττιδι: IG II2 4671 = CorpuscultusCybelaeno. 308 with photograph p. 78. 43 Robert (1975) 324. Μήτηρ Θεῶν Ἄνγδιστις: W.M. Ramsay, CitiesandBishoprics I 246, no. 88; E. Haspels, Highlands of Phrygia I 298, no. 8; more examples in Robert (1980) 236-238. 44 Scholars sometimes characterize rituals that included ecstasy, music, and dancing as ‘mysteries’. See however the caution of Burkert (1987) 112. 45 Hdt. 5.102.1, ἱρὸν ἐπιχωρίης θεοῦ Κυβήβης; Gauthier (1989) 47 [SEG XXXIX 1284.5], τοῦ ναοῦ τοῦ ἐν τῶι Μητρώιωι. 42 A RELIGIOUS ASSOCIATION AT SARDES 13 proper names, and there is no evidence of mysteries. Agdistis is on record in Lydia otherwise only once, in a private cult of Zeus at Philadelphia (TAMV 1539.51 = LSAM 20, I BC), as guardian of the group’s regulations (like Nemesis?); no mysteries are involved. The uniqueness of the mysteries of Agdistis in the Sardes inscription suggests again a novelty cult. Ma was the great goddess of Comana in Cappadocia and of Comana in Pontus.46 Robert thought the Sardes inscription to be the only evidence for her mysteries at any date. While this is literally correct, it may also be incomplete. According to Origen (C.Celsum 6.22), Celsus discussed various mysteries but ignored those of the barbarians: those of the Egyptians, in which many are initiated, or those of the Cappadocians for Artemis of Comana. The locale Comana47 and the fact that Ma and Artemis were frequently equated by Imperial writers48 show that Celsus was referring to Ma.49 Comana’s goddess of war may well be on record as early as the Bronze Age. But she seems to be unmentioned by any Greek source before the first century BC.50 The only other testimony to her in Lydia is a tiny private altar found near Hyrcanis.51 So a preImperial date for Ma at Sardes seems unlikely, and mysteries of Ma at any date are almost unheard of. Thus all three of these mysteria are either unparalleled or almost so. Their common denominator is recherché novelty. And it is conspicuous that the therapeutai are not forbidden to participate in other mysteries 46 See now Debord (2005). Cf. τῇ ἐν Κομάνοις θεῷ, SEG LII 1464 ter.10. 48 From Strab. 12.2.3 to Procop. Wars 1.17.13. 49 This is recognized by Hartmann (1928) 89. 50 Early testimonia: a dedication at Susa SEG VII 10 (II/I BC or later, to judge from the script); Ivantchik (2004) 1-14, argues that the Olbian dedication CIRB 74 is as old as 100 BC. The claim that Sulla introduced the cult of Ma from Cappadocia to Rome in 88 BC (e.g. Lex.topog.urb.Rom. I (1993) 193; rightly absent from Richardson (1982) 57-58) is based on a misunderstanding of Plut. Sull. 9.7 (Sulla’s dream of a war goddess, θεὸν ἣν τιμῶσι Ῥωμαῖοι παρὰ Καππαδόκων μαθόντες, εἴτε δὴ Σελήνην οὖσαν εἴτε Ἀθηνᾶν εἴτε Ἐνυώ, “a goddess whom the Romans, having learned of her from the Cappadocians, honor, whether as Luna or Minerva or Bellona”). Perhaps Sulla in his memoirs (cf. Plut. Sull. 6.6) reported his dream of an unnamed Cappadocian goddess; in any case the passage reflects the application to this story of a scholarly syncretism of war goddesses and an idle deduction about origins (Roman cults of all Plutarch’s candidates predated 88 BC). It happens that no temple of Bellona in Rome is said to belong to Ma (see the sources quoted in Hartmann (1928) 80-81), and though Sulla was in Cappadocia in 92, no source says that he imported the cult or knew the name of Ma. 51 TAM V 1305 (17.5 cm. wide), late Hellenistic or Imperial. 47 14 K. RIGSBY which must have long been available in Sardes, for example those of Dionysus or Isis. The prohibition against sharing these mysteries reflects a background and an anxiety. The background, the religious behavior implied by the regulations, is the collecting of multiple initiations to mysteries and the pursuit of the exotic. The quest for foreign mysteries was in full spate in the Roman Empire. By contrast, the evidence for mysteries attached to eastern cults before Imperial times is vanishingly small.52 Thus an Imperial date for the mysteries named in the Sardes prohibitions is more likely. FORBIDDEN CULT PARTICIPATION If the quest for novel mysteries is characteristic of the Roman Empire, the group’s response is timeless: suspicion and anxiety in the face of the new and alien in religious practice. It is, to be exact, reactionary, and reflects the suspicion which a respectable and established group felt towards the unfamiliar and unproven, and towards the competition which it threatened. We see this attitude, with its implied aspersions on class and gender, throughout antiquity, from Demosthenes’ attack on the rituals of Aeschines’ mother to the Romans’ sporadic restrictions on foreign cults. If there was more of this conservative reaction in the Empire, it is because there was more to react against. And yet, forbidding participation in cult is virtually unheard of.53 Select persons could of course be exclude from access to sacred space,54 52 See Dunand (1975) 12-15; Burkert (1987), esp. 40-41. The aretalogies call Isis the creator of initiations (μυήσεις) in general, just as she was also of sacred precincts (τεμένη), but they do not credit her with mysteries of her own; cf. Grandjean (1975) 103-104. Burkert calls Tibullus 1.3.23-26 “the oldest literary witness for mysteries of Isis” (40). But the passage concerns the annual ten-day period of purification for worshippers; purification is not initiation, and it featured in all ancient rites, only more elaborately for the Egyptian gods. Perhaps as early as I BC: I.Prusa ad Olympum 1028, ἄρρητα βεβήλοις of Isis (on it see Burkert 26); SEG XI 974.18-26, mysteries of the Syrian Goddess at Thuria. 53 Burkert (1987) 4: “in the pre-Christian epoch the various forms of worship, including new and foreign gods in general and the institution of mysteries in particular, are never exclusive”; 48-51 on multiple initiations and priesthoods. Long ago Roussel (1916) 253-255, 267, observed that the several associations that honored the Egyptian gods on Delos were not exclusive; cf. Cumont (1949) 407. 54 See the examples in E. Lupu, NGSL 14-21, 72-73. A RELIGIOUS ASSOCIATION AT SARDES 15 temporarily or permanently, for reasons of pollution, criminality,55 nationality, gender, etc., or even for impurity of heart, or for atheism.56 Cults shared by the members of a group could exclude non-members, e.g. a cult maintained by a family or a tribe.57 Any group could define who was allowed in and who was not.58 Cities, being self-governing, might do likewise, denying a temple to a foreign god, just as they could deny land-ownership to a foreign person. To take up residence in Attica, Astarte needed permission (IG II3 337). The Roman state could require state certification in order to form a Bacchic group, or ban eastern gods from within the pomerium. Those were bans on admission to one’s own rites and places. A civic parallel to the Sardes prohibitions, by contrast, would be a law that citizens, when abroad, could not worship certain gods. That is unthinkable. The rule in Greek religion was: When in doubt, salute. The Romans’ reported treatment of Druids is more akin to the Sardes prohibitions: Augustus forbade Roman citizens from participating in Druid cult, Claudius banned it outright (Suet., Claud. 25). Whatever the law that Rome applied to Christians stated, it did not legislate belief or ritual (the first would require mind-reading; the second would require a typology of ritual acts, those which were legal and those which were not). What the law addressed was membership and participation (especially financial) in a group: that was testable in court. These rare prohibitions by governments concerned matters of public order or the protection of borders or of membership; they derive from the interest of the state in monitoring the behavior and ensuring the safety of its citizens and subjects. Where, then, do we find one cult forbidding participation in another cult? (1) Nock was able to cite only one example, a fifth-century AD story that a Mithraic pater (πατὴρ ὢν τῆς Μιθριακῆς τελετῆς) became an Eleusinian hierophant, even though he was devoted to other gods 55 Thus banning from all the city’s public places, including temples. The latter is aggressively expressed in LSAM 16 (Gambreium, III BC): those convicted of impiety are not to sacrifice to any god for ten years (26, θύειν μηθενὶ θεῶν ἐπὶ δέκα ἔτη). 56 So at least Lucian, Alex. 38: the new cult excludes Christians and Epicureans. 57 E.g. Athens, IG II2 1214.16, rites reserved for members of the deme; Cos, IG XII.4 103.3 ἔδο[ξ]ε ταῖς φυλαῖς αἷς μέτεστι τῶν ἱερῶν Ἀπόλλωνος καὶ Ἡρακλεῦς ἐν Ἁλασάρναι; cf. 348.7 τοὶ τῶν ἱερῶν κοινωνεῦντες, 52 οἷ[ς] μέτεστι τῶν ἱερῶν. Compare Herodotus on the Caunians, τούτοισι μὲν δὴ μέτεστι, ὅσοι δὲ ἐόντες ἄλλου ἔθνεος ὁμόγλωσσοι τοῖσι Καρσὶ ἐγένοντο, τούτοισι δὲ οὐ μέτα (1.171.6). 58 So the Lindians, μ]ὴ μετέχωντι τῶν ἐν Λίνδωι ἱερῶν οἳ μὴ καὶ πρότερον μετεῖχον (Syll.3 340, with Bresson (1988) 145-154). 16 K. RIGSBY (than the Eleusinians) and had taken an oath not to preside over other rites (θεοῖς ἑτέροις καθιέρωται καὶ ὀμώμοκεν ἀρρήτους ὅρκους ἑτέρων ἱερῶν μὴ προστήσεσθαι).59 This prohibition, as Nock noted, need not have been a general feature of the Mithraic initiates, or older than its alleged time, when pagans were appropriating Christian usages. And (in the telling) it went unenforced. (2) The pose of henotheism that was commonly struck in prayers and hymns could go to extremes. ‘Pray to Isis; do not pray to a[nother god]’, says a Demotic hymn in a second-century AD papyrus.60 But this, like all such brave talk about a one and only god in polytheism, is the momentary rhetoric of praise, not theology and not a rule for actual conduct in the world.61 A modern analogy is the rhetoric of advertising: ‘the one and only cleanser for every need’ (εἷς θεός), ‘accept no substitute’. Such exclamations should not be taken literally or as law. (3) In the first century BC an Egyptian club that honored Zeus stipulated (along with more general requirements for obedience) that no member was ‘to make factions or to leave the brotherhood of the president for another’.62 This is an exclusivity, but again different from that at Sardes in not being selective: resignation in order to join any other group is banned. The Roman Empire witnessed the rivalry of initiatory cults, even of the ‘same’ god. An acclamation of the second century AD: ‘now we are the foremost of all the Baccheioi’ (νῦν πάντων πρῶτοι τῶν Βακχείων, IG II2 1368.26). But rivalry alone cannot explain singling out three particular cults as forbidden. (4) Judaism and Christianity offer no strict similarity.63 The Sardes association did not say, Thou shalt have no other gods. It specifies 59 Eunap. VS 7.3.4 (475); Nock (1933) 292-293; cf. Burkert (1987) 50-51. Kockelmann (2008) 33. 61 See Versnel (2011) 280-307. 62 P.Lond. VII 2193.13, μηιδὲ σχί<σ>ματα συνίστασ[θαι] μηιδ̣’ ἀπ[ο]χ̣ωρήισε[ιν ἐκ] τ̣ῆς τοῦ ἡγ[ου]μ̣ένου φράτρας εἰς ἑτέραν φράτραν, with Nock e.a. (1936) 39-88. The Order of the Star founded by King Jean of France in 1352 stipulated that no member could be a member of another order: de Laurière (1729) 465, “Et se il y a aucuns qui avant ceste compaignie ayent empris aucun Ordre, il la devront lessier, se il pevent bonnement”. The new order failed miserably, meeting only once. This was in the context of the vast proliferation of knightly orders: “There was not a prince or great noble who did not desire to have his own order” (Huizinga (1954) 87). Pierre d’Ailly (ca. 1400) complained of the proliferation of religious orders that “this leads to a diversity of usages, to exclusiveness and rivalry, to pride and vanity” (Huizinga 153). 63 Horsley, New Documents, p. 23 on the Sardes inscription, invokes the exclusivity of Judaism and Christianity and concludes that it was “not confined to these groups”; 60 A RELIGIOUS ASSOCIATION AT SARDES 17 three mystery cults amid what would have been a sea of mystery cults. It is this selectivity that makes the gesture look less like Judeo-Christian exclusivity and more like the Roman state’s prohibition of joining Christianity. The rulings applying to the neokoroitherapeutai at Sardes do not seem to fit well with any of these exclusivities. What was their motive? One could imagine various objections on their part: to keep members from ritual contamination (which seems unlikely); from ‘double-dipping’, an excessive pursuit of pious initiations; irresponsible flightiness; insufficient concentration on one’s duties to the cult group. But such motives would apply to joining any other cult, not just the three that are named. Other mysteries will have been available in Sardes64 as in any other Roman city; far more visible and popular mystery cults — certainly we can expect Isis or Dionysus in Roman Sardes — were not forbidden by these rulings. What the three forbidden cults share is obscurity in extant testimonia: that is sufficient evidence of their novelty in Sardes. To the therapeutai, I suggest, mysteries of Agdistis and of Ma and of a factional sect of Sabazius-worshippers were unknown and intrusive: other mysteries will have been seen as established and time-honored, and therefore legitimate. Such conservatism65 marked the Imperial age, and perhaps especially Roman Sardes — the autochthonous Sardians, as they liked to call themselves.66 followed by Ascough (2006) 174-175. So already Robert (1975) 326, “such exclusivity is in force only among the Jews and Christians”. 64 The record however is slim: ‘mysteries’ of Hermes and Heracles in the gymnasium (I.Sardis 21, with Herrmann’s caution, (1996a)340 n. 75: Hellenistic), and in Imperial times a mysterionof Attis (I.Sardis 17.6), mystaiof Apollo (1996a, 318-321), and some unidentified mysteries (1996a, 317-318 and I. Sardis 62 with 1996a, 341). In the third century AD a cult group at Sardes, unfortunately unidentified, received the provincial governor’s permission to proceed with their traditional ‘mysteries and sacrifices and libations’ (the surviving text does not reveal what had been the impediment): Petzl (2009) 377-386. 65 Well put by Kraabel (1992) 254: “conservatism, reinforcing the piety of the past”; but accepting Robert’s Achaemenid date he concluded “it is likely that this kind of exclusiveness is nothing new”. 66 I.Sardis 13 and 63-66, with Herrmann (1993) 238-243; πρωτόχθων in IGUrbRom I 85, SEG XXXVI 1095, 1096; Σάρδιες ἀρχαῖαι in Anth.Gr. 7.709.1. In AD 26 (Tac. Ann. 4.55) the Sardians informed the Senate that they had colonized Italy (the Etruscans) and the Peloponnesus (Pelops): the ancient capital of the Lydians was older than Greece and Rome. 18 K. RIGSBY LEGITIMATION This conservative attitude can be the link between the Achaemenid dedication and the later cult regulations. What do the statue-dedication and the prohibitions have in common, that caused them to be extracted and reinscribed on one stone? The stone is a building block, not a statue base. Briant67 held that this was an accidental collocation, and that only the base of the statue, not the statue itself, had survived from Persian times. But the lack of a referent would render the second sentence unintelligible: whose statue did the therapeutai crown? The unnamed god of the regulations is the Zeus that Droaphernes dedicated; I would conclude that one authority was responsible for inscribing this stone, the devotees of Zeus. They were devoted to the care of his statue; and they had a text that showed its antiquity, a dated dedication that proved an origin older than the coming of the Greeks to Lydia. By the second century AD, the statue set up in Persian times was quite possibly the oldest dated monument in Sardes; and in ancient religion, age brought respect. This, I suggest, is what caused them to cite the dedicatory text on the occasion of a dispute about proper conduct: that text was their proof of the antiquity and therefore the legitimacy and superiority of their tendance of the ancient statue, by contrast with new fads for strange mysteries. Sardes was besieged by Antiochus III and extensively destroyed in 215/4, and substantial rehabilitation and rebuilding were begun in 213.68 Late in the third century BC the great temple of Artemis, the city’s chief god, was reconfigured, divided into two halves, a space for Artemis to the west and one for Zeus to the east.69 After the earthquake of AD 17, significant rebuilding again had to be done to the temple. In one of these two repairs, a wall was built across that prevented communication of the two halves. It may be that the statue dedicated by Droaphernes, whatever had been its original site, was moved to this space belonging to Zeus on one of those occasions,70 to be inside the adyton as the second 67 Briant (1998) 223-224. Gauthier (1989). 69 As Zeus Polieus: οἱ ἐν τῷ ἱερῷ τοῦ τε Πολιέως Διὸς καὶ τῆς Ἀρτέμιδος οἰκοῦντες (I.Sardis 8.133, Augustan: the temple residents are allowed to erect a statue of a benefactor). His priest is secondary eponym ca. 100 BC: ἐπὶ ἱερέω[ς] τῆς μὲν Ῥώμης Σωκράτου, τοῦ δὲ Διὸς τοῦ Πολιέως Ἀλκαίου (OGIS 437, now Laffi (2010)). 70 So Robert (1975) 321 suggested. A fragment of a large statue of Zeus seems to derive from this area; but the matter is complex, see Hanfmann (1983) 93, 119–121, 132, 68 A RELIGIOUS ASSOCIATION AT SARDES 19 sentence says. That location, in the city’s chief temple, would help explain the self-importance of the statue’s therapeutai. We can suspect that the therapeutai believed that their group and their privileged duties dated from the creation of the statue in Achaemenid times. What was the truth? Voluntary associations proliferated in the Hellenistic period. Commonly they were non-citizens who banded together under the banner of a favored god in order to fulfill the various goals which for citizens were fulfilled by the subdivisions of the polis; we know them best on Delos and Rhodes.71 The Roman Empire seems to witness more associations in which piety was the chief motive; and it witnessed also the archaizing cultivation of inherited or invented traditions. These seem to be the characteristics of the therapeutai of Zeus at Sardes. The caretakers of his statue look less like the social and commercial club of Poseidoniasts on Hellenistic Delos, and rather more like the altar guild of a modern Christian church, self-consciously and ostentatiously pious and old-fashioned. At Roman Olympia a group of persons who claimed to be descendants of Phidias were uniquely privileged to clean the famous statue of Zeus.72 That ancestry was probably fictitious; in any case, they were volunteers. I suggest that service groups of this sort are more likely to be found in Imperial times than before. CONCLUSION On this interpretation, the group that toward Roman times formed itself to tend an ancient statue of Zeus would likely be among the most conservative people in the city, ostentatious in their services to an old cult statue, a privilege which gave them a sense of superiority — self-dramatizing atavists in an already conservative city. In the second century AD, the therapeutai reacted to the possible intrusion of the new and exotic among their own number: one member, Dorates, has in the manner of the age gotten himself initiated into some exotic new mysteries, or talked of doing so; the group has reacted with alarm and sought a judgment. fig. 176. It would follow that our inscribed block came originally from the east end of the temple — which perhaps is testable. F.K. Yegül is preparing a full account of the building. 71 See especially Fraser (1977) 46-70. 72 Paus. 5.14.5; one member is known, a Roman citizen: Τίτον Φλάουιον Ἡράκλειτον, τὸν ἀπὸ Φειδίου, φαιδυντὴν τοῦ Διὸς τοῦ Ὀλυμπίου (I.Olympia 466). 20 K. RIGSBY The three sentences extracted for display on this stone can be described as authority, policy, and execution: 1) The group was in charge of the cult statue in the temple, and on its base was the ancient dedicatory inscription. They saw this text as their founding document, and believed that their privileged service was as old as the statue itself: the inscription justified their existence and their authority. So in the course of the proceeding about Dorates, this ancient text was quoted. 2) The god or his priest or the law now made a general finding about these novel mysteries: the god’s attendants are not to participate. 3) The laws or the therapeutai apply the policy to the case at hand: they order Dorates to abstain. Hostility to the new and strange in religion is documented from at least the fifth century BC onwards. Such innovations were seen by many as charlatanism, not respectable, female, lower class. Legitimate religion meant, first, old religion. When the Ptolemies wanted to guarantee the authenticity of private cults of Dionysus, a consumer-protection law, they framed a simple standard: the cult must be three generations old.73 But that posture in the face of the new and dubious was an attitude of the political community, the polis, the traditional locus of one’s identity: the citizens had every right to guard their boundaries, to decide who of men and gods was allowed to settle with them. At Sardes we see something different, the pious devotees of a statue viewing another cult with alarm and forbidding themselves to join it. The conservative reaction against the new in religion that we know so well in the ancient world, a part of one’s identity as a respectable citizen, is here taken up and made a part of another sort of identity, the therapeutai of Zeus, a chosen identity that rested on service rather than on location or birth. This is what is novel and striking in this inscription. These arguments about the inscription and its chronology are admittedly atmospheric — that the prohibitions seem to breathe the air of the Roman Empire rather than earlier,74 and that the collocation of these texts is not fortuitous but expresses the concerns of a single group for their proper tendance of a statue. If that is correct, this episode at Sardes 73 Lenger, C.Ord.Ptol. 27; cf. LSAM 73.7, third generation citizenship required for eligibility for a priesthood (Theangela, Hellenistic). As the Christians would say, πᾶν τὸ ἀρχαιότητι διαφέρον, αἰδέσιμον (Basil, PG 31.165; quoted at Council VII: Mansi XIII 252). 74 Cf. Athanassiadi (2010) 40: “dans le vocabulaire hellénistique, on manque absolument de termes pour désigner la notion d’intolérance religieuse”. A RELIGIOUS ASSOCIATION AT SARDES 21 anticipates, in its small way, the identity politics based on religious adhesion which through the third and fourth centuries will grow to become such a decisive feature of public and private life. Kent RIGSBY DukeUniversity [email protected] BIBLIOGRAPHY Ascough (2006): Richard S. ASCOUGH, ‘Voluntary Associations and the Formation of Pauline Communities’, in: A. Gutsfeld & D.-A. Koch (eds.), Vereine, Synagogen und Gemeinden im kaiserzeitlichen Kleinasien (Studien und Texte zu Antike und Christentum, 25), Tübingen 2006. Athanassiadi (2010): Polymnia ATHANASSIADI, Vers la pensée unique, Paris 2010. Bresson (1988): Alain BRESSON, ‘Richesse et pouvoir à Lindos’, in: S. Dietz & I. Papachristodoulou (eds.), Archaeology in the Dodecanese, Copenhagen 1988, p. 145-154. Briant (1998): Pierre BRIANT, ‘Droaphernès et la statue de Sardes’, in: M. Brosius & A. Kuhrt (eds.), Studies in Persian History (Achaemenid History, 11), Leiden 1998, p. 205-226. Burkert (1987): Walter BURKERT, Ancient Mystery Cults, Cambridge (MA) 1987. Butz (2009): Patricia A. BUTZ, ‘The Naxian Colossus at Delos: “Same Stone”’, in: AmbrosiaVII (BCH, Suppl. 51, 2009), p. 77-87. Chaumont (1990): Marie-Louise CHAUMONT, ‘Un nouveau gouverneur de Sardes’, Syria67 (1990), p. 579-608. Corsten (2003): Thomas CORSTEN, ‘Prosopographische und onomastische Notizen’, EpigAnat 35 (2003), p. 113-122. Cumont (1949): Franz CUMONT, LuxPerpetua, Paris 1949. Debord (1999): Pierre DEBORD, L’AsieMineureauIVesiècle, Bordeaux 1999. Debord (2005): Pierre DEBORD, ‘La déesse Ma et les hirondelles blanches’, REG 118 (2005), p. 15-30. de Laurière (1729): Eusèbe Jacob DE LAURIÈRE, Ordonnances des rois de Francedelatroisièmerace II, Paris 1729. Drew-Bear e.a. (2008): Thomas DREW-BEAR, Alexander LUBOTSKY & Mevlüt ÜYÜMEZ, ‘Three New Phrygian Inscriptions’, Kadmos 47 (2008), p. 109-116. Dunand (1975): Christiane DUNAND, ‘Les mystères égyptiens’, in: Mystèreset syncrétismes (Etudes d’histoire des religions, 2), Paris 1975. Dusinberre (2003): Elspeth DUSINBERRE, Aspects of Empire in Achaemenid Sardis, Cambridge 2003. Foucart (1875): Paul FOUCART, ‘Inscription inédite de Mantinée’, Annu.Assoc. ét.gr. 9 (1875), p. 327-341. Fraser (1977): Peter M. FRASER, RhodianFuneraryMonuments, Oxford 1977. 22 K. RIGSBY Frei (1996): Peter FREI, Reichsidee und Reichsorganisation im Perserreich2, Freiburg 1996. Fried (2004): Lisbeth S. FRIED, ThePriestandtheGreatKing, Winnona Lake 2004. Gauthier (1989): Philippe GAUTHIER, Nouvelles inscriptions de Sardes II, Geneva 1989. Giannobile & Jordan (2008): Sergio GIANNOBILE & David R. JORDAN, ‘On the Text of the Hipponium Tablet’, GRBS 48 (2008), p. 287-294. Grandjean (1975): Yves GRANDJEAN, Unenouvellearétalogied’IsisàMaronée, Leiden 1975. Gschnitzer (1986): Franz GSCHNITZER, ‘Eine persische Kultstiftung in Sardeis’, in: W. Meid & H. Trenkwalder (eds.), Im Bannkreis des Alten Orients (InnsbBeitrKult, 24), Innsbruck 1986, p. 45-54. Hanfmann (1983): George M.A. HANFMANN e.a., Sardis from Prehistoric to RomanTimes, Cambridge (MA) 1983. Hanfmann & Ramage (1978): George M.A. HANFMANN & Nancy H. RAMAGE, SculpturefromSardis, Cambridge (MA) 1978. Hartmann (1928): A. HARTMANN, ‘Ma’, RE XIV (1928), col. 77-91. Hermary (1993): Antoine HERMARY, ‘Le colosse des Naxiens à Délos’, REA 95 (1993), p. 11-27. Herrmann (1993): Peter HERRMANN, ‘Inschriften von Sardeis’, Chiron 23 (1993), p. 233-266. Herrmann (1996a): Peter HERRMANN, ‘Mystenvereine in Sardeis’, Chiron 26 (1996), p. 315-348. Herrmann (1996b): Peter HERRMANN, ‘Epigraphische Notizen’, EpigrAnat 27 (1996), p. 55-62. Herrmann & Malay (2007): Peter HERRMANN & Hasan MALAY, NewDocuments fromLydia (DenkWien, 340), Vienna 2007. Hitch (2011): Sarah HITCH, ‘Embedded Speech in the Attic LegesSacrae’, in: A.P.M.H. Lardinois (ed.), Sacred Words: Orality, Literacy, and Religion, Leiden 2011, p. 113-141. Huizinga (1954): Johan HUIZINGA, TheWaningoftheMiddleAges, New York 1954. Ivantchik (2004): Askold S. IVANTCHIK, ‘Dedication to the Goddess Ma from Olbia’, AncientCivilizationsfromScythiatoSiberia 10 (2004), p. 1-14. Jones (1990): Christopher P. JONES, ‘Lucian and the Bacchants of Pontus’, EMC N.S. 9 (1990), p. 53-63. Keil & von Premerstein (1911): Josef KEIL & Anton VON PREMERSTEIN, Bericht übereineZweiteReiseinLydien (DenkWien, 54.2), Vienna 1911. Kockelmann (2008): Holger KOCKELMANN, PraisingtheGoddess, Berlin 2008. Kraabel (1992): Alf T. KRAABEL, ‘Paganism and Judaism: The Sardis Evidence’, in: J.A. Overman & R.S. MacLennan (eds.), Diaspora Jews and Judaism, Atlanta 1992, p. 21-34. Laffi (2010): Umberto LAFFI, Il trattato fra Sardi ed Efeso (Studi ellenistici, 22), Pisa 2010. Maddoli (1963-64): Gianfranco MADDOLI, ‘Le cretule del Nomophylakion di Cirene’, AASA 41-42 (1963-64), p. 39-145. A RELIGIOUS ASSOCIATION AT SARDES 23 Mayrhofer (1974): H. MAYRHOFER, ‘Zu den Parthernamen der griechischen Awroman-Dokumente’, in: Ph. Gignoux & A. Tafazzoli (eds.), Mémorial JeandeMenasce, Louvain 1974, p. 205-213. Minns (1915): Ellis H. MINNS, ‘Parchments of the Parthian Period from Avroman’,JHS 35 (1915), p. 22-65. Mitchell (2007): Stephen MITCHELL, ‘Iranian Names and the Presence of Persians in the Religious Sanctuaries of Asia Minor’,PBA 148 (2007), p. 151-171. Nock (1933): Arthur D. NOCK, Conversion, Oxford 1933. Nock e.a. (1936): Arthur D. NOCK, R. COLLINS, & Theodore C. SKEAT, ‘The Gild of Zeus Hypsistos’, HThR 29 (1936), p. 39-88. Paz de Hoz (1999): Maria PAZ DE HOZ, DielydischenKulte, Bonn 1999. Petzl (1994): Georg PETZL, ‘Die Beichtinschriften Westkleinasiens’,EpigrAnat 22 (1994), p. 1-174. Petzl (2009): Georg PETZL, ‘Bedrohter Kultvollzug: Hilfe von höherer Stelle’, in: A. Martínez Fernández (ed.), EstudiosdeEpigrafíaGriega, La Laguna 2009, p. 377-386. Petzl (2011): Georg PETZL, review of Sencer ŞAHIN & Mustafa ADAK, StadiasmusPatarensis, Gnomon 83 (2011), p. 50-55. Pleket (1981): Harry PLEKET, ‘Religious History as the History of Mentality’, in: H.S. Versnel (ed.), Faith, Hope and Worship (Studies in Greek and Roman Religion, 2), Leiden 1981, p. 152-192. Pouilloux (1954): Jean POUILLOUX, Etudesthasiennes III, Paris 1954. Renberg (2003): Gil RENBERG, ‘Commanded by the Gods’: an Epigraphical Study of Dreams and Visions in Greek and Roman Religious Life (Ph.D. diss., Duke University, 2003). Richardson (1982): Lawrence RICHARDSON, ANewTopographicalDictionaryof AncientRome, Baltimore 1982. Robert (1975): Louis ROBERT, ‘Une nouvelle inscription grecque de Sardes’, CRAI (1975), p. 306-330 (reprinted in Operaminoraselecta V, p. 485-509). Robert (1980): Louis ROBERT, Atraversl’AsieMineure, Athens 1980. Roesch (1982): Paul ROESCH, Etudesbéotiennes, Paris 1982. Roussel (1916): Pierre ROUSSEL, LesculteségyptiensàDélos, Paris 1916. Sacco (2001): Giulia SACCO, ‘Γῆς παῖς εἰμι’, ZPE 137 (2001), p. 27-33. Sokolowski (1979): Franciszek SOKOLOWSKI, ‘Τὰ ἔνπυρα’, ZPE 34 (1979), p. 65-69. Tchernetska (2005): Natalie TCHERNETSKA, ‘New Fragments of Hyperides’, ZPE 154 (2005), p. 1-6. Thonemann (2009): Peter THONEMANN, “Estates and the Land in Early Hellenistic Asia Minor,” Chiron 39 (2009), p. 363-393. van Bremen (2010): Riet VAN BREMEN, ‘Adrastos at Aphrodisias’, in: R.W.V. Catling & F. Marchand (eds.), Onomatologos, Oxford 2010, p. 440-455. Versnel (2011): Hendrik VERSNEL, CopingwiththeGods, Leiden 2011. Vidman (1970): Ladislav VIDMAN, Isis und Sarapis bei den Griechen und Römern, Berlin 1970. von Gutschmid (1892): Alfred VON GUTSCHMID, Kleine Schriften III, Leipzig 1892.