¹ Rudhardt, J., Notions fondamentales de la pensée religieuse et actes constitutifs du culte dans la Grèce classique (Paris, 1992 ²)Google Scholar, 30. Unless otherwise specified, abbreviations are those used in the Guide de l'épigraphiste, 2010, and translations are ours. Other abbreviations used include: Akinci Öztürk and Tanriver (2008) = Öztürk, E. Akinci and Tanriver, C., ‘New katagraphai and dedications from the sanctuary of Apollon Lairbenos’, EA 41 (2008), 91–111Google Scholar; Cabanes and Drini (2007) = Cabanes, P. and Drini, F., Corpus des inscriptions grecques d'Illyrie méridionale et d'Epire, vol. 2: Inscriptions de Buthrotum (Athens, 2007)Google Scholar; Darmezin (1999) = Darmezin, L., Les affranchissements par consécration: en Béotie et dans le monde hellénistique (Nancy, 1999)Google Scholar; Hatzopoulos et al. (2000) = Hatzopoulos, M.B., Petsas, F.M., Gounaropoulou, L. and Paschidis, P., Inscriptions du sanctuaire de la Mère des Dieux Autochtone de Leukopétra (Macédoine) (Athens, 2000)Google Scholar; Ritti et al. (2000) = Ritti, T., Simsek, C. and Yıldız, H., ‘Dediche e καταγραφαί dal santuario frigio di Apollo Lairbenos’, EA 32 (2000), 1–88.Google Scholar

² See the state of art in Papazarkadas, N., Sacred and Public Land in Ancient Athens (Oxford, 2011).CrossRefGoogle Scholar

³ There are, to our knowledge, two general approaches of the phenomenon. Bömer, F., Untersuchungen über die Religion der Sklaven in Griechenland und Rom, vol. 2 (Hamburg, 1960), 149–89Google Scholar is devoted to ἱεροί from all over the Greek world. But his analysis relies on several misleading conceptions, such as an excessive emphasis on the question of ‘what is Greek and what is not’. Debord, P., Aspects sociaux et économiques de la vie religieuse dans l'Anatolie gréco-romaine (Leiden, 1982), 78–83CrossRefGoogle Scholar also provides a comprehensive commentary and bibliography, but several relevant epigraphic pieces of evidence have been published since then. Specific dossiers have also been treated subsequently without being compared to the ‘whole picture’ provided by the different pieces of evidence (see n. 6).

⁴ The notion of ‘social status’ will appear several times in our analysis, to designate the relationship between individuals and the community around them, in terms of freedom and obligations, as it is defined legally or through official texts. For a more developed reflection on ‘social status’ and its connection with the legal sphere, see e.g. Hunter, V. and Edmondson, J., Law and Social Status in Classical Athens (Oxford, 2000)CrossRefGoogle Scholar, especially Chapters 1 and 8.

⁵ It should be stressed from this point that by ‘dedication’ or ‘consecration’, we are not considering curses which imply sanctions on a person. No divine ἄγος is at stake in the texts discussed here, though identical expressions may be used for both kinds of consecration: compare ἀνιεροῦν in Strabo 11.14.16 and in Cnidian curse tablets (IK Knidos 1.147, 148, 149, 151 and 158). Ἀνατίθημι is also attested in these tablets: cf. IK Knidos 1.150.

⁶ On ἱεροί in the mysteries of Andania, see Deshours, N., Les mystères d'Andania: étude d'épigraphie et d'histoire religieuse (Paris, 2006), 77–82CrossRefGoogle Scholar; Pirenne-Delforge, V., ‘Mnasistratos the “Hierophant” at Andania (IG 5.1.1390 and Syll.³ 735)’, in Dijkstra, J., Kroesen, J. and Kuiper, Y. (edd.), Myths, Martyrs, and Modernity: Studies in the History of Religions in Honour of Jan N. Bremmer (Leiden, 2010), 219–35CrossRefGoogle Scholar, at 227–8; Gawlinksi, L., The Sacred Law of Andania: A New Text with Commentary (Boston, 2012), 22–7Google Scholar. In regard to the comparison with Peloponnesian epitaphs mentioning ἱεροί, see Brulé, P. and Piolot, L., ‘Women's way of death: fatal childbirth or hierai? Commemorative stones at Sparta and Plutarch, Lycurgus, 27.3’, in Figueira, T.J. (ed.), Spartan Society (Swansea, 2004), 151–78CrossRefGoogle Scholar. Epitaphs mentioning ἱεροί are not limited to the Peloponnese: see e.g. epitaphs from Samos in IG 12.6.2, 688, 756 and 803. See Themelis, P., ‘Ἀνασκαφὴ Μεσσήνης’, PAAH 156 (2001), 57–96Google Scholar, at 74: an inscription from Messene, dated to the beginning of the third century b.c., shows seven ἱεροί dedicating a bronze statue to Apollo Karneios. On ἱερά in the mysteries from Samothrace, see Karadima-Matsa, C. and Clinton, K., ‘Korrane, a sacred woman in Samothrace’, ZPE 138 (2002), 87–92Google Scholar. See also Rizakis, A.D., Achaïe, vol. 3: Les cités achéennes: épigraphie et histoire (Paris, 2008)Google Scholar, no. 63: an inscription from Achaia shows συνιεροί of a hero dedicating a statue to Thrason son of Xenophon.

⁷ We know of three ἱεροί in Aizanoi, whose precise function is unclear, as discussed in Lehmler, C. and Wörrle, M., ‘Neue Inschriftenfunde aus Aizanoi III. Aizanitica Minora’, Chiron 32 (2002), 571–646Google Scholar, at 576. According to an uncertain conjecture, a ἱερός called Hermas was in charge of δημόσια γράμματα (see MAMA 9 P 28: τοῖς πανηγυριάρχαις καὶ Ἑρμᾷ ἱερῷ τῷ πρὸς δημο[σίοις γράμμασι]). If right, it would suggest that ἱεροί were assistants also in non-religious matters.

⁸ Herrmann, P. and Malay, H., New Documents from Lydia (Vienna, 2007)Google Scholar, no. 54: τῶν ἱερῶν ὅτι ‘μὴ φ[ο]|βοῦ’· ἐπεζήτησεν | ἡ θεὸς καὶ ἦραν τρίφωνα | τὰ τέκνα αὐτῆς | καὶ τὰ ἔγγονα αὐτῶν, | ἔτους σξ᾿, μη(νὸς) Ἀπελ|λαίου βι᾿ (‘from the holy servants (saying): “Have no fear!” The goddess made an inquiry and her [that is an aforementioned woman's] children and their descendants took (the sin) away by means of three voiced animals, the year 260, month of Apellaios, the 12^th’; the editors' slightly revised translation). Yet the presence of ἱεροί is all but certain.

⁹ Herrmann, P. and Varinlioglu, E., ‘Theoi Pereudenoi: eine Gruppe von Weihungen und Sühninschriften aus der Katakekaumene’, EA 3 (1984), 1–18Google Scholar, no. 10: —- ου Μηνὶ Λαβάνᾳ καὶ Μη|νὶ Πετραείτῃ ἐν Περεύ|δῳ Ἀμμία Zηνᾶ Ἀνκυρα|νὴ ὑπὲρ τῆς οἰκίας τῆς | ἠγόρασεν παρὰ Ἀμμίας | Καλλιμάχου ἔδωκα (δηνάρια) οβ᾿ | καθὼς ἐπεσζήτησαν οἱ | θεοί, ἅτινα παρέλαβαν οἱ | εἱεροὶ Ἀπολλώνιος Ἀπολ|λωνίου, Ἀντίοχος Ἀντιόχου, | Γλύκων Ποπλίου (‘On the month of Labanas and of Petraeites, in Pereudos, I, Ammia, daughter of Zenas, from Ankyra, had for the house bought from Ammia, daughter of Kallimakhos, 72 denarii, as the gods required. The sum was given to the ἱεροί Apollonios son of Apollonios, Antiokhos son of Antiokhos, Glykon son of Poplios’). The text is not completely clear as to why the house in question is connected to the gods.

¹⁰ Nevertheless, the priest is appointed to his office through specific processes and an analogy may be proposed between processes through which one became a ἱερός and the τελετή through which the priest had to go.

¹¹ The latter is suggested in a decree from Ephesus: Syll.³ 742.

¹² Darmezin (1999).

¹³ For an overview of modes of manumission, see Zelnick-Abramovitz, R., Not Wholly Free: The Concept of Manumission and the Status of Manumitted Slaves in the Ancient Greek World (Leiden, 2005), 69–99Google Scholar. ‘Sacral manumission’ implies two types of processes: the freedman was either sold or consecrated to a god. In this paper, our attention will bear on the latter. On sacral manumission, see Velissaropoulos-Karakostas, J., Droit grec d'Alexandre à Auguste (323 av. J.-C. – 14 ap. J.-C.): personnes – biens – justice (Athens, 2011), 379–86.Google Scholar

¹⁴ In Boeotia, for instance, deities to whom slaves are dedicated vary from place to place. Any attempt to establish a close relation between the identity of the deity and the process of consecration or the identity of the consecrated persons seems fruitless. Darmezin (1999), 184: ‘Il ne semble donc pas qu'il y ait eu des divinités ‟spécialisées” dans la protection des affranchis.’ Darmezin's suggestion that only girls were consecrated to Artemis Eilithyia, a deity who had a ‘champ d'action spécifiquement féminin’, seems plausible (pp. 184–5).

¹⁵ Darmezin (1999), 180.

¹⁶ On this interchangeability, see Cabanes, P., ‘Epigraphie et affranchis du monde grec: acquis et problèmes’, in Roman, Y. and Le Bohec, Y. (edd.), Epigraphie et histoire: acquis et problèmes. Actes du Congrès des Professeurs d'Histoire Ancienne, Lyon-Chambéry, 21–23 mai 1993 (Lyon, 1998), 53–60Google Scholar, at 59. Accordingly, for the sake of convenience, and since ἱερός is often – though not systematically – used to qualify the status of a new freedman, the term ἱεροί will here be used to designate persons who went through this process in the corpus of Darmezin.

¹⁷ The argument of M. Ricl that, since no word in ἐλευθ- is attested in the Leucopetra dossier (see below), it should be inferred that slaves were donated to a divinity but not freed is unfounded: Ricl, M., ‘Donation of slaves and freeborn children to deities in Roman Macedonia and Phrygia: a reconsideration’, Tyche 16 (2001), 127–60Google Scholar, at 130 and 134–5. Moreover, Zelnick-Abramovitz (n. 13) relied on the fact that freedmen had to render a service to the gods to prove that consecration involved a ‘moral link between slaves and gods; the latter could keep the slaves, give them back to the owners, or make them completely free’.

¹⁸ See e.g. Darmezin (1999), no. 29 = IG 7.3083.

¹⁹ Darmezin (1999), nos. 45, 72, 73.

²⁰ Darmezin (1999), 230. See also Petrakos, V., Ὁ Ὠρωπὸς καὶ τὸ ἱερὸν τοῦ Ἀμφιαράου (Athens, 1968)Google Scholar, 48.

²¹ See the list in Darmezin (1999), 186, and inscriptions nos. 43, 45, 72, 73, 77. No. 79 shows that, if an agreement was needed in such a case, it was to come from members of the family of the ἱεροί, not from a magistrate: παριόν|τος αὐτῇ τῶ ἱαράρχαο Νικίαο Χα|ρώνδαο, συνευδοκίοντος | κὴ τῶ ιουἱῶ αὐτᾶς Εἵρωνος (lines 3–6). See also parallels from the Buthrotum corpus in P. Cabanes, ‘La loi des ateknoi dans les affranchissements d'Epire’, in Liebs, D. and Modrzejewski, J. (edd.), Symposion 1977 (Cologne, 1982), 215–22Google Scholar and Cabanes and Drini (2007), 257–61.

²² Darmezin (1999), 196.

²³ Cf. n. 96 below.

²⁴ Darmezin (1999), 222.

²⁵ This was already suggested in Motte, A., L'expression du sacré dans la religion grecque (Louvain-la-Neuve, 1996)Google Scholar, 128, with regard to the general meaning of the adjective ἱερός: ‘Il signifie souvent la simple appartenance ou la réservation d'une chose à tel dieu, ou encore la simple relation à un sanctuaire’ (our emphasis).

²⁶ For such sanctions, see Darmezin (1999), nos. 126 and 131. Quotation from p. 224.

²⁷ Darmezin (1999), no. 123. Κουρία is therefore not used in the same meaning as κυριεύειν in the inscription from Pergamum, see n. 96.

²⁸ Darmezin (1999), no. 133, lines 22–5.

²⁹ Cabanes and Drini (2007), no. 26.

³⁰ Darmezin (1999), no. 138.

³¹ For this koinon's institutions, see Cabanes and Drini (2007), 242–8. The inscriptions were found in three locations: two in the city theatre and one in a Roman tower, whose blocks belonged to previous temples of Asclepius and Zeus. See the introductory remarks in Cabanes and Drini (2007), 63–5.

³² The order can change. In addition, ἐλεύθερον, ἱερόν, ἀνέφαπτον, or even the whole sentence ἀφίημι ἐλεύθερον, can be omitted. These interchangeable variants seem to have no consequence for the meaning and must be consequently read as abbreviations, as proved by some awkward combinations such as no. 123, ἀφίητι ἱερὸν | καὶ ἀνατίθητι Φιλόστρ|ατος Στίλπωνα παρὰ | Δία Σωτῆρα, ἀνέφαπτον (cf. nos. 124, 125, 143); cf. Cabanes and Drini (2007), 273.

³³ For our purpose, it would be tempting to address the question of the Roman manumissio sacrorum causa, which is mentioned by the grammarian Festus (s.vv. Puri, probi, profani, sui auri: dicitur in manumissione sacrorum causa). Buckland, W.W., The Roman Law of Slavery: The Condition of the Slave in Private Law from Augustus to Justinian (Cambridge, 1970), 447–8Google Scholar makes a brief comparison between this type of Roman manumission and Greek practices, but does not assimilate them. The role of manumissio sacrorum causa within the Roman system itself is unclear; for instance, the connection with the manumissio vindicta is problematic: see Fabre, G., Libertus: recherches sur les rapports patron-affranchi à la fin de la république romaine (Rome, 1981)Google Scholar, 19 n. 147.

³⁴ The main publications of these texts are Ricl, M., ‘Les ΚΑΤΑΓΡΑΦΑΙ du sanctuaire d'Apollon Lairbenos’, Arkeoloji Dergisi 3 (1995), 167–95Google Scholar and Ritti et al. (2000), who present both edited and unedited texts, and Akinci Öztürk and Tanriver (2008), with only hitherto unedited texts.

³⁵ When manumission is mentioned in a καταγραφή, it seems to be a separate process. See Akinci Öztürk and Tanriver (2008), no. 13: Τίτος Φλάβις Ἀχιλλεὺς καταγράφω τὸν ἐμαυτοῦ δοῦλον ὀνόματι Ἐπίκτητον Ἡλίῳ Λαρμηνῷ ὃν κὲ ἐπύησα ἐλεύθερον διὰ τῶν ἐν Μοτέλλοις ἀρχείων (‘I, Titus Flavi(u)s Achilleus, assign to Helios Larmenos my slave named Epiktetos whom I also made free through the archives in Motella’). We would also propose ‘through the agency of the magistrates’, but the meaning is not completely clear. No. 14 clearly states that a certain Zosimos will be ἱερός and free: the quality of ἱερός can be obtained from the καταγραφή while freedom comes from a procedure in Motella.

³⁶ For other meanings of the term, see LSJ s.v.: ‘engrave’, ‘describe’, ‘enroll’, ‘register’, and so on.

³⁷ Akinci Öztürk and Tanriver (2008), no. 2: καταγράφω δὲ τῷ Zήνωνι ἐργαστήριον κὲ τὸ δίστεγον κὲ ἄρμεν[α- ]α σὺν εἰσόδοις κὲ ἐξόδοις; Akinci Oztürk and Tanriver's translation, slightly modified (‘Ι also assign to Zenon a workshop and a two-storied house and tools [for …] together with (their) incomes and outgoings (?) …’ ). See also no. 7.

³⁸ Cases of καταγράφω ἱερόν: Ritti et al. (2000), nos. 8, 11 and 49; Akinci Öztürk and Tanriver (2008), nos. 1, 5, 7, 8, possibly 9, 11, 15.

³⁹ SEG 42.1185, line 4.

⁴⁰ See Ricl, M., ‘Society and economy of rural sanctuaries in Roman Lydia and Phrygia’, EA 35 (2003), 77–101Google Scholar, at 91 n. 91.

⁴¹ See e.g. I.Tralles 6, lines 9–12: παλλακεύσα|σα κατὰ χρη|σμὸν | Διΐ. This has nothing to do with sacred prostitution, about which see Debord (n. 3), 78. According to Budin, S., The Myth of Sacred Prostitution in Antiquity (Cambridge, 2008), 193–6Google Scholar, the παλλακή in this inscription is a ‘cult functionary’. The status of a παλλακή is not unproblematic. It may indicate a specific status that someone grants to herself in her own carrying out of a cult: being a παλλακή would therefore fall in the category of personal devotion rather than in official cultic organization. On the other hand, the hypothesis that it designates a specific sacerdotal function cannot be dismissed, as the term and its use seem more specific than ὑποτακτικός.

⁴² The dynamic of familial piety is patent in Akinci Öztürk and Tanriver (2008), no. 14: a θρεπτός goes through a καταγραφή on the altar of Artemis, which had been dedicated by the father of the author of the καταγραφή. For other examples of familial commemoration, see the entry of Akinci Öztürk and Tanriver's article in EBGR 2008, no. 1.

⁴³ An interesting case of someone qualified as ἱερός in an honorific decree unfortunately had to be dismissed. First published in Robert, L., Hellenica VI: Inscriptions de Lydie (Limoges, 1948), 49–50Google Scholar, the inscription seemed to concern a certain Apellas the second Loukios, a ἱερός who had been δεκάπρωτος, στρατηγός and γραμματεύς. Robert subsequently assumed that the lacunary text was to be interpreted not as ἱερόν but as ἱερόνομον: see BE 1973, 413; L. and Robert, J., La Carie: histoire et géographie historique avec le recueil des inscriptions antiques, vol. 2: Le plateau de Tabai et ses environs (Paris, 1954)Google Scholar, 295 n. 1, and TAM 5.2.266. Other inscriptions, however, show more clearly some ἱεροί: see Robert (1948), 49–50.

⁴⁴ The recipients of these fines vary from one inscription to the other. Akinci Öztürk and Tanriver (2008), no. 3 mention these three recipients together.

⁴⁵ Ricl (n. 34), no. 30 = Ritti et al. (2000), no. 29, lines 8–10: μηδινὸς ἔχοντος ἀν[θρ|ώ]που ἐξουσίαν κατὰ τοῦ Ἀ[πο|λ]λωνίου; Ricl (n. 34), no. 32 = Ritti et al. (2000), no. 31, line 9: ἐφάψασθαι ὡς δούλης. Akinci Öztürk and Tanriver (2008), no. 14, lines 7–9: εἴ | τις δὲ ἐπενκαλέσει τοῦ Zωσί|μου ὡς εἰς δουλίαν ἀνθρώπου.

⁴⁶ For cases of Hellenistic sacred manumissions from Macedonia, cf. SEG 43.388 from Edessa: Εὐρυνόα Ἀρισ|τοκλείδου ἀ|νατίθησιν τὴ|ν αὑτῆς παιδ|ίσκην Εὐτυχί|δα, τὸ γένος Σύ|ραν, Παρθένωι̣ (first half of the second century b.c.); EAM 115 from Kelle (Eordaia): ἔτους γ̣ʹ [---] | στρατηγοῦντ̣[ος Λ(ευκίου) Καλ]|πορνίου Πείσω[νος] | Εὔδικος Ταυρίωνο̣[ς] | Βρυναῖος ἀφῆκεν ἐλε|υθέραν Μέλισαν εὐ|χὴν Ἡρακλῇ Κυνα|γίδᾳ (57–55 b.c.). Cf. Hatzopoulos et al. (2000), 36, with reference at n. 6; Youni, M., ‘Maîtres et esclaves en Macédoine hellénistique et romaine’, in Anastasiadis, V.I. and Doukellis, P.N. (edd.), Esclavage antique et discriminations socio-culturelles (Bern, 2005), 183–95.Google Scholar

⁴⁷ For the topography of the area, see Hatzopoulos, M.B., ‘Herodotos (8.137–8), the manumissions from Leukopetra, and the topography of the middle Haliakmon valley’, in Derow, P. and Parker, R. (edd.), Herodotus and His World: Essays from a Conference in Memory of George Forrest (Oxford, 2003), 203–18.CrossRefGoogle Scholar

⁴⁸ Conversely, ἱερόδουλοι are present: cf. Section 3.

⁴⁹ See Ricl (n. 17) and (n. 40). For previous literature, see the references collected in Hatzopoulos et al. (2000), 33–5. The editors interpreted the inscriptions as a consistent corpus of slave manumissions, a view which has not met with general acceptance: see Chaniotis, A., ‘From woman to woman: female voices and emotions in dedications to goddesses’, in Prêtre, C. (ed.), Le donateur, l'offrande et la déesse: systèmes votifs dans les sanctuaires de déesses du monde grec (Liège, 2009), 51–68CrossRefGoogle Scholar, at 55.

⁵⁰ Cf. I.Leucopetra 134 (uncertain date).

⁵¹ I.Leucopetra 69, lines 3–12: σώματα ἃ ἠγόρασα παρὰ Αὐ|ρηλίου Φορτουνάτου καὶ Κλαυδίου | Σωτῆρος, ὧν καὶ τὰς ὠ|νὰς παρέσχον σοι πολλάκις ἐ|πὶ εὐχαριστηρίωις οἷς παρέσχου | τῷ ἀνδρί μου Κλ(αυδίῳ) Ἀγάθωνι, ἃς καὶ | ἔδωκά σοι, χαρίζομαί σοι διὰ | ταύτης μου τῆς ἐπιστολῆς, ὧν | σωμάτων καὶ τὰ ὀνόματα ὑ|πέγρα̣ψα κτλ.

⁵² A particularly expressive case in this sense is provided by I.Leucopetra 12, lines 13–17: προσμενοῦσι | δέ μοι τὸν ζῶ χρόνον ὑπηρετοῦντα τῇ θεῷ | τὰς ἐθίμους ἡμέρας, μετὰ δὲ τὴν ἐμὴν τε|λευτὴν μηδένα εἶνε κύριον ἢ τὴν θεὸν μόνη|ν. Hatzopoulos et al. (2000), 90 state that ‘apparemment Phlauios Eutrapélos avait emprunté 1.000 deniers pour acheter Phélix. Ne pouvant rembourser cette dette, il passa un accord avec le sanctuaire de la Mère des Dieux, selon lequel ce dernier assumait l'obligation du remboursement, recevant en contrepartie la totalité des biens de Phlauios Eutrapélos.’ However, Ricl (n. 17), 147 proposes that the letters ΙCA at line 11 ‘should be read as ἴσα and understood as referring to χειρόγραφα τ[ὰ] | ὑπάρχοντα immediately preceding it. The donor is simply stating that he is depositing with the Goddess the number of documents equivalent to the number of the donated slaves … I would then put a full stop or a semi-colon after ἴσα and treat the following phrase as a separate clause dealing with the donor's debt of 1,000 denarii and its repayment by the Goddess.’

⁵³ On the other hand, while χαρίζομαι hints at an action pleasing the deity, its appearance in texts of the Leucopetra corpus with economic relevance suggests that one should not interpret this verb as necessarily indicating a gift to the goddess.

⁵⁴ I.Leucopetra 53, lines 2–6: ἐχαρισόμην κοράσιον ὀνό|ματι Συνφέρουσαν Μητρὶ Θε|ῶν Αὐτόχθονι τὸ κὲ ἀπούλ < ωλ > ον | τὸ αὐτὴ ἀτῇ ἀναζητή|σ̣εις.

⁵⁵ Chaniotis (n. 49), 57.

⁵⁶ Cf. the formula κατ' ἐπιταγήν (I.Leucopetra 9?, 34, 101, 151, 154, 164); κατὰ | κέλευσιν τῆς θεοῦ (I.Leucopetra 131; cf. IG 10.2.2, line 34 = A.N. Oikonomides, Sylloge inscriptionum Graecarum et Latinarum Macedoniae, 2 vols. (Chicago, 1980²) = SEG 49.838; from Lynkestis, Vašarejca); see also I.Leucopetra 78, where the dedication is the consequence of the goddess' order (lines 4–6: καθὼ[ς] ἐκέλ̣ευ|σας ἀγοράσε με σω[μ]άτι|α) accepted by the donor (lines 10–11: ὁ|μολογ[ῶ).

⁵⁷ See Chaniotis, A., ‘Ritual performances of divine justice: the epigraphy of confession, astonishment, and exaltation in Roman Asia Minor’, in Cotton, H.Μ., Hoyland, R.G. and Price, J.J. (edd.), From Hellenism to Islam: Cultural and Linguistic Change in the Roman Near East (Cambridge, 2009), 115–53CrossRefGoogle Scholar; cf. also Chaniotis (n. 49), 57–8, and SEG 50.597, p. 188.

⁵⁸ I.Leucopetra 47: ἔτους · ελσ Σεβαστοῦ | τοῦ καὶ αντ, Λαδόμα | Ἀμύντου ἐχαρίσατο | τῇ θεῷ ὑὸν ἴδιον ὀνόμα|τι Π<α>ράμονον, ὃν ὑπέσ|χετο ὄντα ἐν νόσῳ, ὑ|π<α>ιρετοῦντα μηδενὶ ἑτέρῳ | ἢ μόνῃ τῇ θεῷ. ὁ προγε|γραμμένος Παράμονος | παρῆν καὶ συνεπέδωκεν αὑτόν· | ἱερωμένης Αἰλίας |Αὐρηλιανῆς, ἐπι|μελουμένης Αὐρηλίας | Σαπφοῦς. Beside the editors' commentary, see for this text Tataki, A.B., Ancient Beroea: Prosopography and Society (Athens, 1988), 487–8Google Scholar; Ricl, M., ‘Legal and social status of threptoi and related categories in narrative and documentary sources’, in Cotton, H. M., Hoyland, R. G., Price, J. J. and Wasserstein, D. J. (edd.), From Hellenism to Islam: Cultural and Linguistic Change in the Roman Near East (Cambridge, 2009), 93–114CrossRefGoogle Scholar, at 109; Chaniotis (n. 49), 60.

⁵⁹ According to the editors, ‘Paramonos demeure, bien entendu, libre’: Hatzopoulos et al. (2000), 117. One must recall, however, that according to them, all the consecrations are manumissions.

⁶⁰ Hatzopoulos et al. (2000), 37–8.

⁶¹ On Hellenistic ἱεροδουλία, see Otto, W., Beiträge zur Hierodulie im hellenistischen Ägypten (Munich, 1949)Google Scholar; Delekat, L., Katoche, Hierodulie und Adoptionsfreilassung (Munich, 1964)Google Scholar; Debord, P., ‘L'esclavage sacré: état de la question’, in Actes du colloque 1971 sur l'esclavage (Paris, 1972), 135–50Google Scholar; Scholl, R., ‘Ἱερόδουλος im griechisch-römischen Ägypten’, Historia 34 (1985), 466–92Google Scholar; Legras, B., Les reclus grecs du Sarapieion de Memphis: une enquête sur l'hellénisme égyptien (Leuven, 2011), 13–21, 162–5.Google Scholar

⁶² See Budin (n. 41), Chapter 7.

⁶³ Strabo 8.6.20: ‘The temple of Aphrodite was so rich that it owned more than a thousand temple-slaves (ἱεροδούλους) serving as courtesans (ἑταίρας), whom both men and women dedicated (ἀνετίθεσαν) to the goddess.’

⁶⁴ On the cult of this god, which was typical of Southern Phrygia and central Pisidia, cf. Labarre, G., ‘Les origines et la diffusion du culte de Men’, in Bru, H., Kirbihler, F. and Lebreton, S. (edd.), L'Asie mineure dans l'Antiquité: échanges, populations et territoires (Rennes, 2009), 389–414.Google Scholar

⁶⁵ IGLS 1.1 (= OGIS 383; Nemrud Dağ); IGLS 1.47 (with corrigenda in IGLS 3, p. 681; Arsameia); SEG 12.554 (revised text of IGLS 1.51; Selik, near Samosata); SEG 53.1763, ANd (Ancoz); SEG 53.1770–1 (Zeugma); SEG 53.1776 (revised text of SEG 26.1623; Sofraz Köy). See discussion in Waldmann, H., Die kommagenische Kultreformen unter König Mithridates I: Kallinikos und sein Sohne Antiochos I (Leiden 1973)CrossRefGoogle Scholar; Crowther, C. and Facella, M., ‘New evidence for the ruler cult of Antiochus of Commagene from Zeugma’, in Heedeman, G. and Winter, E. (edd.), Neue Forschungen zur Religionsgeschichte Kleinasiens (Bonn, 2003), 41–80Google Scholar; Wagner, J. and Petzl, G., ‘Relief- und Inschriftfragmente des kommagenischen Herreschkultes aus Ancoz’, in Heedeman, G. and Winter, E. (edd.), Neue Forschungen zur Religionsgeschichte Kleinasiens (Bonn, 2003), 85–96.Google Scholar

⁶⁶ Cf. Scholl (n. 61), 468.

⁶⁷ See on this purpose the invitations to contextualizing ἱεροδουλία within specific space and time contexts by Scholl (n. 61), 466, 468, 487.

⁶⁸ On whether ἱερόδουλοι in Egyptian sanctuaries were charged with proper cult offices or simple administration (which seems more plausible), see the discussion by Scholl (n. 61) including previous bibliography.

⁶⁹ Scholl (n. 61), 488–92; Depaw, M., A Companion to Demotic Studies (Brussels, 1997), 136–7Google Scholar; Legras (n. 61), esp. 13–21, 162–5.

⁷⁰ On κατοχή in Ptolemaic Egypt, see Legras (n. 61). On the administrative charges of the κάτοχος Ptolemy of Glaucias in the Memphite Serapeum, in the mid second century b.c., see Legras (n. 61), esp. 180–2.

⁷¹ I.Leucopetra 39 mentions the dedication (line 4: ἀνατίθημι τῇ θεῷ) of a three-year-old child by the woman who bought her at her birth and brought her up, probably for the purpose of the consecration, as the name Theodote may suggest (cf. Chaniotis (n. 49), 59). No. 109 has a man offer a two-year-old baby he has brought up (lines 5–6: δωροῦμαι τῇ δεσ̣ποίνῃ | μου θρεπτόν μου). In no. 112 a woman offers a child (lines 1–2: δωροῦμε παῖδα τῇ δεσποίνῃ μου | [Παράμ]ονον, <ὃν> ἀνέθη̣[κα). The female donor of no. 113 offers a girl she has brought up so that she shall stay at the sanctuary on the customary festival days (lines l2–16: δωροῦμε | τῇ δεσποίνῃ μου θρεπτήν | μου Ἀλεξάνδραν, ἐτῶν κεʹ,| προσμενούσης τῇ θεῷ τὰς ἐ| < θί > μους ἑορτάς·). No. 117 has a lady, Theodote (chronology makes an identification with the homonymous consecrated baby of no. 39 impossible), offer a servant and her son (line 3: χαρίζομε), adding the παραμονή clause valid till her death (line 5: ἐφ' ᾧ προσμίνωσίν μοι παρὰ τὸν τῆς ζοῆς χρόνον); no. 151 is of different nature, since here the offering of a child to the goddess is represented as a restitution ordered by Zeus Hypsistos (perhaps in a dream?), (lines 2–6: κατ' ἐπιτα|γὴν Θεοῦ Ὑ[ψί]στου, | μετὰ υἱοῦ Παραμό|νου τὴν ἐπιτ[αγ]ὴν | ἀπέδωκεν τῷ θεῷ).

⁷² On θρεπτοί, see especially, for Phrygia, Ricl, M. ‘Legal and social status of threptoi and related categories in the Greek world: the case of Phrygia in the Roman period’, in Νεολληνική κληρονομία στους Σέρβους 1 (Belgrade, 2005), 145–66Google Scholar; and, for Lydia, M. Ricl, ‘Legal and social status of threptoi and related categories in the Greek world: the case of Lydia in the Roman period’, in Sobria ebrietas: mélanges offerts à Miron Flašar. Recueil de travaux de la Faculté de philosophie, série A: Les sciences historiques 20 (Belgrade, 2006), 293–321; Ricl (n. 58) for a comprehensive survey.

⁷³ Hatzopoulos et al. (2000), 60 maintain that being a ἱερόδουλος ‘équivalait presque à un affranchissement’. As seen above, this statement can be accepted in many cases but there are acts in Leucopetra where a passage of the slave's property from a private owner to the sanctuary is a more suitable interpretation than manumission.

⁷⁴ Two texts contribute to the understanding of the tasks entrusted upon these low-rank cult personnel: the ἱερόδουλος donor of I.Leucopetra 39 is a λυχνάπτρια; I.Leucopetra 131 informs us that the offered person will serve as αὐλητής. On high and low-rank personnel of the cult in Macedonian, Phrygian and Lydian sanctuaries of the Imperial period, see Ricl (n. 40).

⁷⁵ Dedications of human beings to local sanctuaries in Imperial Macedonia are attested from a few scattered locations beside the two major corpora, from Leucopetra and from the city of Beroea (I.Beroea 48–56: recipient gods are Artemis Agrotera, Artemis Eileithyia, Syria Parthenos, Dionysos Agrios (?) Erikryptos/Kryptos Pseudanor).

⁷⁶ IG 10.2.2.18c: ἔτους δ̣λυʹ μηνὸς Δίου· ἐ|γὼ Διονυος [sic] ἱερόδουλ(ος) | θεᾶς Πασικράτας ἀνέ|θηκα υἱόν μου κὲ δοῦ|λον ὀνόματι Φίλητον, | [ὃ]ν̣ ἠγόρασα μετὰ κὲ τῆς <θεᾶς> | εἰς τὴν χρῆσιν αὐτοῦ | ὡς τὸν ζωῆς χρόνον. See commentary in Ricl, ŽAnt 32 (1982), 165–70, for the interpretation of υἱόν μου κὲ δοῦ|λον being one person (cf. SEG 32.636), rather than an unnamed consecrated son accompanied by a manumitted slave.

⁷⁷ For cases where ἱερόδουλος appears in accusative, thus expressing the status acquired by the person through consecration, see the texts mentioned above from Commagene and Pisidia, in Section 3.1. In the inscription from Episkopi, however, the use of the verb ἀφίημι, which is usually related to manumissions, makes the interpretation of the formula much more complicated.

⁷⁸ SEG 2.396; Roman period: [Αὐρ]ηλία Φιλίππα [ἡ] | [π]ρὶν Εὐροδίκης ἀ|φίημι παιδίσκην ὀ|νόματι Ἀριάγνην | θεᾷ Ἀρτέμιδι Γαζω|ρίᾳ ἱερόδουλον το|[--]Η̣Μ ΩΥΡΟ[--]. Cf. Scholl (n. 61), 468.

⁷⁹ Delacoulonche, A., ‘Mémoire sur le berceau de la puissance macédonienne des bords de l'Haliacmon et ceux de l'Axius’, Archives des missions scientifiques et littéraires 8 (1859), 67–288Google Scholar, no. 29; cf. Hatzopoulos (1987), 410–11: ἔτους CΓΞC [γξσʹ(?)] Σεβαστοῦ, | μηνὸς Ὑπερβερεταίου λʹ | Οὐλπία Εὐπορία καὶ Αὐ|ρήλιος Διονύσιος ἠξί|ωσαν οἱ θρέψαντες κα|λῶς δουλευθέντες ὑπὸ | θρεπταρίου εἰδίου ὀνόμα|τι Ὀνησίμαν περὶ ἔτη ιηʹ ἀνατιθή|μειν θεᾷ Ἀρτέμιδι Γαζωρίᾳ ταύτη[ν] | εἶν<αι> δούλην τῆς θεᾶς πρὸς τ[ὰ] | ἐκτὸς ἐλευθέραν, μέχρ[ι –].

⁸⁰ A good practical synthesis of the status of ἱερόδουλοι and δοῦλοι of a god is given by Ricl (n. 40), 90: ‘they legally became slaves of divinities protected by their divine patrons; yet with respect to the public authorities and private individuals they were considered personally free. They had property and personal rights, as well as legal capacity, but not complete freedom of movement or freedom to change their status.’ A different interpretation of being δοῦλος of a god is proposed by Scholl (n. 61), 487, who thinks that the formula could simply point at the piety and devotion of a person towards a god. This reading, however, seems to fit better with cases where a person refers to himself as a servant of the god rather than with those where this definition is the effect of an act of consecration.

⁸¹ EAM 59b, lines 3–7: προσμ|[ε]νεῖ τῷ ἥρῳ | καὶ εἶναι ἐ|λ̣ευθέρα|ν ναοῦ etc. Cf. EAM 59a, line 2.

⁸² I.Leucopetra 43: ἔτους ζκσʹ Σεβ<α>στοῦ τοῦ | καὶ γμτʹ, εἱερωμένης | Αἰλίας Μητρῶς κὲ | ἐπιμελουμένου Αὐ|ρηλίου Ἀσσκληπιάδου· | Κρισπίνα Μητρὸς | Θεῶν ἀπελευθέρα | ἐχαρισόμην Μη|τρὶ Θεῶν Αὐτόχθο|νι δούλην ὀνό|ματι Ἐλπιδίαν π<ρ>ο|<σ>μένουσαν τὰς ἐθί|μους ἡμέρας, τὸν | δὲ κατάλοιπον χρό|νον προ<σμ>ενῖ ἐμοὶ | καὶ Διονυσίῳ τὸν | ζώομεν χρόνον, | μετὰ δὲ τὴν ἡμε|τέραν τελευτὴν | μηδένα εἶνε | κυριώτερον | ἢ τὴν θεόν.

⁸³ Ricl (n. 40), 90–1 n. 88.

⁸⁴ Hatzopoulos et al. (2000), 111.

⁸⁵ SEG 50, p. 189.

⁸⁶ Ricl (n. 17), 142.

⁸⁷ Ricl (n. 17), 143 with n. 61 tentatively pointed to a second-century b.c. inscription from Lycian Oenoanda (SEG 27.932). In addition to coming from a very different context, however, the text is highly mutilated and the relevant passage is entirely in lacuna; for the uncertainty of the restoration, see L. Robert, Bull. Épigr. 1978, 462.

⁸⁸ Ritti et al. (2000), 51.

⁸⁹ Horsley, G.H.R., The Greek and Latin Inscriptions in the Burdur Archaeological Museum (London, 2007)Google Scholar, nos. 20 and 50.

⁹⁰ TAM 5.1.483a, lines 15–17.

⁹¹ Herrmann and Malay (n. 8), no. 84. The ἀφιδρύματα τῶν θεῶν and θεοί are, in our opinion, images of the gods and not persons playing the role of gods. It could be that ἱερόδουλοι actually were in charge of carrying the images of the gods. On ἀφιδρύματα, see Pirenne-Delforge, V., ‘Des marmites pour un méchant petit hermès! ou comment consacrer une statue’, in Estienne, S., Jaillard, D., Lubtchansky, N. and Pouzadoux, Cl. (edd.), Image et religion dans l'antiquité gréco-romaine: actes du colloque de Rome, 11–13 décembre 2003 (Naples, 2008), 109–10.Google Scholar

⁹² Ricl (n. 40), 90. Ricl is more careful in other passages. A similar confusion is evident in Bömer (n. 3), 151–2.

⁹³ Strabo (11.14.16) considers dedicating slaves to a god to be οὐ θαυμαστόν.

⁹⁴ Darmezin (1999), no. 198: Κλοινιζόας (…) Ἑρμαίου Ὀνοβ[ά]|ρου Μνανδρασέως ἀπέλυ|σεν τῆι Μητρὶ Ὀρείαι ἱεροδού|λας Ἀκιεροῦν καὶ Ἀπιονιθεῖν | τὰς ἑαυτοῦ παιδίσκας … (‘Kloinizoas […] freed to Mater Oreia, as sacred slaves, his own servants Akieros and Apionitheis’ ). We disagree with Darmezin's translation of ἀπέλυσεν, as we see no reason to translate it by ‘he abandoned’ rather than ‘he freed’.

⁹⁵ We will not discuss the ἱερὸς παῖς mentioned in the inscription of the Iobacchoi (IG 2².1368), which is a specific case.

⁹⁶ IvP 2.251, lines 24–6: ‘The priest shall take care of the good order in the sanctuary as seems good and right to him, having authority over the hieroi paides (κυριεύοντα τῶν ἱερῶν παίδων).’

⁹⁷ IG 12.6.169, line 38: μὴ ἐξουσία δὲ ἔστω τῶν ἱερῶν παίδων καπηλεύειν̣.

⁹⁸ Rehm, A., Didyma, vol. 2: Die Inschriften (Berlin, 1958)Google Scholar, no. 40.

⁹⁹ ID 1409, line 107.

¹⁰⁰ See also ID 372: μισθωτοῖς τοῖς θάψασι τὸν ἱερὸν παῖδα Χρήσιμον.

¹⁰¹ Hatzopoulos et al. (2000), 30. The existence of such initiation rites in archaic Thessaly and Macedonia remains, however, far from being proved. The applicability of the formal pattern proposed by Van Gennep, A., Les rites de passage (Paris, 1909)Google Scholar, to the study of initiation rites in ancient Greek religion has been recently questioned in the collective volume Dodd, D. and Faraone, C.A., Initiation in Ancient Greek Rituals and Narratives (London, 2003)Google Scholar: see in particular F. Graf, ‘Initiation: a concept with a troubled history’, 3–24; D. Dodd, ‘Adolescent initiation in myth and tragedy: rethinking the Black Hunter’, 71–84; and B. Lincoln, ‘The initiatory paradigm in anthropology, folklore and history of religions’, 241–54. A review of the debate and a more favourable position towards the utility of A. Van Gennep's paradigm are available in Dowden, K., ‘Van Gennep et l'initiation dans la mythologie grecque: mort prématurée d'un paradigme?’, Gaia 14 (2011), 171–9.Google Scholar

¹⁰² For a brief reflection on rites of passage in connection with Roman manumissio, see Fabre (n. 33), 20–1.

¹⁰³ I.Beroea 1 = SEG 27.261 = NGSL 14; c. 180 b.c. Cf. the commentary in Gauthier, Ph. and Hatzopoulos, M.B., La loi gymnasiarchique de Béroia (Athens, 1993), 65–8Google Scholar; Lupu, E., Greek Sacred Law: A Collection of New Documents (NGSL) (Leiden, 2005), 260–8.Google Scholar

¹⁰⁴ For the social classes excluded from the gymnasium, see Tataki (n. 58), 424–7; Gauthier and Hatzopoulos (n. 103), 78–87, esp. 79–81 for slaves and freedmen.

¹⁰⁵ Commenting on the gymnasiarchic law of Beroea, Tataki (n. 58), 425–6 (with references) suggested that ‘the regulation probably indicates the existence in Macedonia of a social rank with restricted political rights, similar to that attested in neighbouring Thessaly, at Sparta and at Gortyn’. More cases of this in-between condition, which one could not reduce to either freedom or slavery, are discussed by Lotze, D., Metaxu Eleutheron kai Doulon: Studien zur Rechtsstellung unfreien Landbevölkerung in Griechenland bis zum 4. Jahrhundert v. Chr. (Berlin, 1959)Google Scholar and Bürger und Unfreie im vorhellenistischen Griechenland: Ausgewählte Aufsätze (Stuttgart, 2000)Google Scholar; Ducat, J., Les Pénestes de Thessalie (Paris, 1994)Google Scholar; Luraghi, N. and Alcock, S.E., edd., Helots and their Masters in Laconia and Messenia: Histories, Ideologies, and Structures (Washington DC, 2003)Google Scholar; and Cartledge, P., ‘The Helots: a contemporary review’, in Bradley, K. and Cartledge, P. (edd.), The Cambridge World History of Slavery, vol. 1: The Ancient Mediterranean World (Cambridge, 2009), 74–90.Google Scholar

¹⁰⁶ Cf. above, n. 71.

¹⁰⁷ As regards denominations in relation to gender and age, it appears from the dossier that παιδίον refers to slaves consecrated in their childhood, regardless of their gender; παιδάριον is used for male slaves, while κοράσιον and παιδίσκη are attested for female slaves respectively up to twenty years and from eighteen years on. The overlap between eighteen and twenty years is to be interpreted in relation to motherhood: cf. Hatzopoulos et al. (2000), 42–3. This is proven by the fact that, with the exception of the generic terms σώματα, σωμάτια, δούλοι, inscriptions from Leucopetra only mention παιδίσκαι consecrated together with their offspring. Other evidence from contemporary Macedonia confirms this assumption. A twenty-two-year old κοράσιον is consecrated to Dionysus Pseudanor in Beroea together with her two younger brothers, aged twelve and sixteen, which suggests that she was not yet associated with a man (I.Beroea 55). Conversely, the only eighteen-year-old παιδίσκη from Leucopetra (no. 84; a.d. 234) may well be a married woman: cf. no. 92 (a.d. 239), where the consecrated 25-year-old παιδίσκη is mother of two children of ten and eight years; or no. 95 (a.d. 241), where another παιδίσκη aged 23 is mother of six and two-year-old daughters.

¹⁰⁸ The 50-year-old slave of no. 93 is an aberrant case unless we think of a mistake.

¹⁰⁹ Inscriptions from Leucopetra suggest that women could have their first child when they were between fifteen (or perhaps earlier) and twenty years old. A slightly larger generation gap is suggested by no. 69 (a.d. 219), attesting the consecration of a whole family: the grandmother Neike (60 years), the mother Alexandra (40 years) and the children Paranomos (twenty), Helene (eighteen) and Alexandra (twelve).

¹¹⁰ Parker, R., Miasma: Pollution and Purification in Early Greek Religion (Oxford, 1983), 48–53Google Scholar; see p. 49 for cases whereby pregnant and breastfeeding women are temporarily excluded from a specific cult. One cannot rule out the possibility that such an interdiction also existed in relation to the cult of the Mother of the Gods in Leucopetra.