The Onomastic Evidence for the God Hermanubis

The personal name Ἑρμάνουβιϲ and related forms are usually subsumed under the category of "polytheophoric" names characteristic of Roman Egypt. These are names that combine the names of two deities, such as Ἡρακλαπόλλων or Ἑρμαντίνοοϲ. They are attested solely in the Roman period, especially between the second and the fourth centuries, and were predominantly popular among males of the middle and upper classes in certain metropoleis like Hermopolis.[1] The name Ἑρμάνουβιϲ, however, though formally combining two divine names, is somewhat special: unlike other polytheophoric names, it refers to a specific deity that is independently attested in Roman Egypt. In other words, it is the name of the god himself that is a "polytheophoric" formation, and this has some consequences on how the personal names based on this deity are formed. Before discussing these names, it is appropriate first to give a brief summary of our state of knowledge about Hermanubis. After reviewing the name forms based on this deity, the onomastic evidence will be related and compared to his iconographical and documentary attestations.

The divine name Ἑρμάνουβιϲ is known from a handful of epigraphic and literary sources, mostly of the Roman period.[2] Plutarch cites the name as a designation of Anubis in his underworldly aspect (De Is. et Os. 375e), while Porphyry refers to Hermanubis as ϲύνθετοϲ, "composite," and μιξέλλην, "half-Greek" (De imaginibus fr. 8, p. 18.1–2 Bidez). The name has been restored in a second-century BC dedicatory inscription from Delos (ID 2156.2 [Ἑρ]μανού[βιδι νικ]ηφόρωι), which would be its earliest attestation, but otherwise it appears in three inscriptions of the Roman period, two from Egypt and one from Thessalonike (see below). It is clear that the name is a result of the assimilation of the Egyptian god Anubis to the Greek god Hermes, which is well attested in a number of literary, epigraphic, and artistic sources.[3] Although Hermes was traditionally equated with the Egyptian Thoth, his function as psychopompos encouraged his association with Anubis given the latter's comparable funerary role in Egyptian religion as embalmer and guardian of the dead and as leader of the deceased to the tribunal of Osiris. This assimilation resulted in widespread Graeco-Roman representations of the canine-headed Anubis with attributes of the Greek Hermes, such as the distinctive staff known as the kerykeion or winged sandals.[4] As far as I know, none of these representations whose provenance is known was found in Egypt, though this may be due to mere chance.[5] But in Roman Alexandria there emerges a new iconographical type, well represented in coins and sculpture, in which a fully anthropomorphic young god is flanked by a dog and holds the same attributes as the said Anubis, in addition to wearing the calathos headdress.[6] It is this type that art historians have traditionally labeled "Hermanubis." In making this identification, they have been encouraged by Porphyry's statement about the god's "composite" and "half-Hellenic" character. However, it is important to stress that the reference to the fully anthropomorphic type as "Hermanubis" is only a hypothetical convention, and that not a single of the numerous representations of this type bears a legend or inscription identifying it as such.[7] I will return to this question at the end of the paper in light of the onomastic evidence.

There are three forms of the personal name based on Hermanubis in papyri and inscriptions from Egypt, in addition to two "proper" polytheophoric names. (In four instances [1, 5, 17, 21] the ending of the name is abbreviated or not preserved.) The form Ἑρμάνουβιϲ, attested six times (2, 9, 11, 13, 14, 15), simply replicates the god's name without further ado, through the phenomenon of "direct theonymy" (the giving to humans of unaltered divine names) that became widespread in the Roman period.[8] In an odd instance (9) the form Ἑρμανουβιϲ in the nominative is borne by a woman, an Alexandrian nurse (ἡ τροφόϲ).[9] It is highly unusual for female theophoric names to be unmarked for gender in this way.[10] I suspect that because the element -ανουβιϲ in Ἑρμάνουβιϲ was often declined with a dental declension in -ιδ-, it might have been conflated with the similar declension of Greek feminine names in -ίϲ, -ίδοϲ.[11] If this is correct, it may be preferable to read the nurse's name as an oxytone (Ἑρμανουβίϲ). There are two further forms based on the god Hermanubis: Ἑρμανουβᾶϲ (six instances: 3, 4, 7, 8, 12, 20) and Ἑρμανουβίων (four instances: 3, 6, 18, 19). -ᾶϲ and -ίων were, in fact, the commonest Greek suffixes for personal names based on the Egyptian god Ἄνουβιϲ.[12] It is interesting that in a business letter (3) the two forms Ἑρμανουβίων (l.38) and Ἑρμανουβᾶϲ (l.41) are used interchangeably to designate the same individual. In this case, Ἑρμανουβᾶϲ is probably a hypocoristic of Ἑρμανουβίων, since the suffix -ᾶϲ often served this purpose.[13]

These forms of the name Ἑρμάνουβιϲ with modified suffixes already separate names based on Hermanubis from other polytheophoric names, which usually simply join two divine names without modifying the original suffix.[14] Indeed, because Hermanubis was conceptualized as a single god, it is not surprising to find personal names such as Ἑρμανοβάμμων (16) and Ἡρακλερμάνουβιϲ (10), in which the name of Hermanubis is joined to that of other gods in the characteristic pattern of polytheophoric names. It is misleading to analyze these names as triple theophoric names, for they imply the juxtaposition of only two gods, Hermanubis and Ammon in one, and Hermanubis and Heracles in the other.[15] In this respect, they are not exceptions to the rule that polytheophoric names combine no more than two deities.

The near majority of the names based on Hermanubis are attested in papyri of the second and third centuries AD. Possibly the earliest example is on an ostracon dated to AD 94 (1), but the name there is abbreviated and Ἑρμανο(υβίου) is Wilcken's resolution.[16] Otherwise, the next attestations are from the second quarter of the second century. There are overall eight instances from the second century, eight from the third, and three from the fourth. One instance is from an inscription dated broadly to the "imperial period." The name disappears from the evidence after the fourth century. What is notable about this chronological distribution is that it corresponds neatly to that of the iconographical representations of the fully anthropomorphic god known conventionally as "Hermanubis." For example, the great number of Alexandrian coins that show a representation of this type on the reverse range from the reign of Trajan (or possibly as early as Domitian) to the Tetrarchic period. Likewise, the statues, statuettes, and carved reliefs of this god have all been assigned to the second or third century AD.[17]

Geographically, names based on Hermanubis are attested in papyri from various areas of Egypt, but it is important to separate, where possible, the provenance of the papyrus from that of the individuals named therein. When this is done, most of the individuals bearing Hermanubis names for whom we can establish an origin turn out to be Alexandrian or Cynopolite (three instances each). This concentration again coincides with the documentary and archaeological evidence for the god Hermanubis. Thus, in the precinct of the Alexandrian Serapeum an inscription of uncertain date was found (SB I 3482) dedicated to "Hermanubis, great god, who listens to prayers and grants favours" (Ἑρμανούβιδι θεῶι μεγάλωι | ἐπηκόωι καὶ εὐχαρίϲτωι).[18] Not far from this dedication a statuette of the anthropomorphic "Hermanubis" was discovered.[19] It is in a suburb of Alexandria, in the sanctuary of Ras el-Soda, that the most famous statue of "Hermanubis" was found (LIMC n. 14). One of the Alexandrian coins representing the god shows him standing inside a temple (LIMC nos. 5a, b), which suggests the existence of such an edifice in the city, probably in the Serapeum precinct given the provenance of the just-mentioned dedication and statuette from there and the frequent iconographical associations between "Hermanubis" and Sarapis in Alexandrian coins.[20] Alexandria, then, clearly emerges as the cult center and probably the birthplace of the fully anthropomorphic Hermanubis, which is consonant with the three Alexandrians bearing his name (5, 6, 9).[21] For the Cynopolite nome, the chief centre of the Egyptian god Anubis, the onomastic evidence is more suggestive than the archaeological and the iconographical. Among the three individuals with a name based on Hermanubis (11, 15, 16), there stands out the father of an underage priest of "Anubis, Leto and associated gods" in the Cynopolite village of Laura (15). Since priesthood was usually hereditary, the father Hermanubis was probably himself a priest of these gods. This onomastic connection of Hermanubis with the Cynopolite nome also has an echo in the iconography: one or two Alexandrian coins personify the Cynopolite nome with a representation of the anthropomorphic "Hermanubis" on their reverse (LIMC nos. 3, 22).

Three conclusions flow from this onomastic survey. First, names based on Hermanubis stand somewhat apart from other polytheophoric names common in the same period in Roman Egypt because, unlike the latter, they refer to a specific and independently attested god: a hybrid-named Hellenized version of Anubis. This is visible in the "onomastic behaviour" of Hermanubis names, such as their receptivity to various suffixes and their composition with names of other gods in "proper" polytheophoric formations. Second, the precise chronological and geographical overlap between the onomastic evidence and the representations of the fully anthropomorphic Alexandrian god known conventionally as "Hermanubis" supports art historians' designation of the latter as "Hermanubis" (for which there is otherwise no positive support; see above). Finally, the connection of Hermanubis with Cynopolis evidenced by some of the personal names, and especially the connection with a priest of Anubis, suggests that Hermanubis continued to be closely associated with the traditional Anubis. It may not be warranted, therefore, to separate too forcefully, in the manner of Jean-Claude Grenier (1990: 265, 268), the more traditional funerary Anubis of the Isiac cult from a more "universalizing" Hermanubis.[22]

I conclude by noting that a name based on Hermanubis is attested once outside Egypt in an inscription from Dion in Macedonia dated to the imperial period (22). The form of the name there is Ἑρμα-νούβιοϲ, which is not known from Egypt. The name is probably to be related to the cult of Hermanubis attested by a third-century dedication in Thessalonike (across the Thermaic golf from Dion), which was set up by a cultic association of Hermanubis on behalf of certain individuals. This document and the personal name, along with a statue of the god from the Sarapeum of Carthage (LIMC n. 15), suggest that the cult of Hermanubis enjoyed some popularity outside its original home in the Roman period.[23]

Appendix 1: List of names based on Hermanubis

Appendix 2: Two literary instances of the name Hermanubis

As a complement to the onomastic survey above, two literary attestations of the name Hermanubis are worth signaling. The first is a passage from Greogry of Nazianzus' autobiography in iambic verse known as De vita sua.[24] The context is Maximus the Cynic's failed attempt to supplant him as bishop of Constantinople in 380. Before the fruitless coup, according to Gregory's account, Peter, bishop of Alexandria and a supporter of Maximus, instructed some Egyptian sailors in Constantinople to spy on Gregory and aid Maximus. Gregory names and lambastes them as follows (De vita sua 834–843, in Patrologia Graeca XXXVII, col. 1087):

Paul Gallay suggests that the names in 838–839 are fictional and part of the satire: an opportunity to associate the Egyptians, and implicitly the Egyptian church, with their pagan past and its odious animal worship.[25] At any rate, most of the names, Ϲτιππᾶϲ and Ῥόδων excepted, are characteristically Egyptian.[26] The occurrence of Ἄνουβιϲ near Ἑρμάνουβιϲ suggests the near equivalence of the names. The adjective κυνώδειϲ in 840, however, need not imply that Hermanubis refers to the canine-headed god, since it could have been triggered only by the name Ἄνουβιϲ (πιθηκόμορφοι, at least, does not correlate with any of the theophoric names cited).

The other literary instance similarly exploits the name Hermanubis for satiric effect. This is a curious, though poorly executed, anonymous poem in the Greek Anthology (AP XI 360), a diatribe against some strategos:

The form Ἑρμανούβηϲ is not attested in documents. There is likewise no instance of Ἀνούβηϲ (but cf. the Copticized Πανούβηϲ in P.Erl. I 108.ii.50 [Theb.; IV]). It is attractive, therefore, to emend Ἑρμανούβηϲ to Ἑρμάνουβιϲ (the final syllable of the name, being in anceps position, could be either short or long). The implied context of the poem is obscure. There is some hesitation among translators whether Ἑρμανούβηϲ is the subject ("the general H. turned into a dog" [Budé]) or predicate ("the general [unnamed] turned into the dog H." [Loeb]) of ἐγένετο. But surely the satire would be more biting if the general was actually named Hermanoubes/Hermanubis: the poet then would be playing on the associations of his name with Anubis the dog on the one hand (κύων) and Hermes the thief on the other (Ἑρμᾶϲ ... ἀϲημοκλέπταϲ). Whether κύων is a general insult or connotes the protective duties of the strategos is unclear, and so is the nature of the implied relationship between the strategos and the two thievish Hermeses.

Notes

1. See Benaissa 2009.

2. Although it was not common in traditional Greek religion to combine the names of two gods in this manner, the double determination of Hermanubis has some formal parallels in the earlier period. The most obvious is the god Ἑρμαφρόδιτοϲ (see Ajootian 1990), attested from the fourth century BC onwards, but his name implies the paradoxical union of two different gods (Hermes and Aphrodite) rather than an assimilation in the manner of Hermanubis. A more apt parallel is provided by Ζηνοποϲειδῶν, an epiclesis of the Carian god Zeus Osogoa, who received the additional determination "Poseidon" as a result of his representation with features such as the trident (see Blümel 1987, 128). In Egyptian religion, the tendency to unite the names of two gods was long established and is well known (cf. Leclant 1975). A comparable form to Hermanubis from the Roman period is Ἡλιοϲέραπιϲ, which is not attested as a personal name (pace Hopfner 1946, 45). Schwartz 1947 also postulates a syncretic funerary god Ἑρμηρακλῆϲ, but this is based purely on the iconographical evidence of some seals, without any support from textual sources (the personal name Ἑρμηρακλῆϲ, two instances of which have emerged since Schwartz's thesis, could be simply an example of a polytheophoric name and need not imply the recognition of an actual god of that name).

3. The literary and epigraphic sources associating Anubis with Hermes are collected by Grenier 1977, 53–59, 95–98. Bernand, Inscriptions métriques n. 73 (Abydos; I–II) = GVI 1090, should also be added, as the Κυλλήνιοϲ Ἑρμῆϲ (9) leading the deceased to serve Osiris of Abydos is implicitly conflated with Anubis (see Dunand 1975, 158). Kákosy 1990, 145 points to some pre-Ptolemaic Egyptian statuettes, in which the feet or teeth of Thoth are modeled as jackal heads, as evidence that the later rapprochement of Anubis and Hermes was based on the earlier assimilation of Thoth to Anubis; but although Thoth and Anubis were closely associated in funerary contexts, I am not persuaded that these statuettes "wollten sicher den synkretischen Gott Thoth-Anubis darstellen" or that they are relevant to the question of the origin of the Graeco-Roman Hermanubis.

4. See Leclant 1981, 871, and the references in Grenier 1977, 137–138. In the amulet published in Bonner 1950, n. 36 (Plate II) = LIMC I s.v. Anubis n. 19, Anubis is described as holding in his right hand "a tall scepter resembling a Roman military standard, with palm leaves at top, and two garlands, the lower with ties" (Bonner 1950, 259). But this looks to me like a version of the kerykeion-cum-palm known from a number of coins of the fully anthropomorphic Hermanubis; see Grenier 1990, 265, and cf. the coins under 8(a). Whether the appellation "Hermanubis" was given to the canine-headed Anubis with Hermetic features is uncertain; Apuleius, Met. XI 11, refers to the god attolens canis cernices arduas and laena caduceum gerens simply as "Anubis."

5. LIMC I s.v. Anubis n. 13, found in Carthage or Cherchel, is the only representation for which an Alexandrian origin has been postulated, but on no firm basis (see Picard 1956, 176–179, who assigns to it an Alexandrian provenance and a date in the third century BC on the basis of the floral decoration).

6. See Grenier 1990.

7. Cf. Leclant 1981, 873.

8. On "direct theonymy" see Parker 2000, 57–8, with further bibliography. For the accentuation Ἑρμάνουβιϲ rather than Ἑρμανοῦβιϲ (Preisigke and some editors), see Clarysse 1997, 180.

9. While the letter was found in the Fayum, the sender is clearly writing from Alexandria; cf. Bagnall and Cribiore 2006, 389–390.

10. For some observations on the marking of gender in Greek personal names, see Davies 2000, 20–21.

11. Cf. Gignac, Grammar, II 55–57; Mayser, Grammatik, I.2 22–23. For masculine names in -ιϲ, -ιδοϲ/-ιοϲ and feminine names in -ίϲ, -ίδοϲ, see Pape-Benseler, Wörterbuch der griechischen Eigennamen3, xviii-xix.

12. See e.g. Preisigke, Namenbuch, 33–34; Foraboschi, Onomasticon, 34–35.

13. See e.g. the "index des suffixes" in Masson 1990, II 632–633 s.v. -ᾶϲ ("suffixe d'hypocoristique et de sobriquet masculin").

14. The only exceptions are two instances ending in -ηράκλειοϲ (Ἑρμηράκλειοϲ [SB VI 9396.3], Ϲαραποηράκλειοϲ [SB XXII 15632.1]) and one ending in -απολλώνιοϲ (Ἡρακλαπολλώνιοϲ [P.Leit. 11.1; see BL VIII, 168–169]).

15. The ending -άμμων was attached to the names of most deities that appear in theophoric names (see Dunand 1963), so that its union with Ἑρμάνουβιϲ is unlikely to have a specific significance. One may adduce some numismatic representations in which "Hermanubis" is shown with a radiating crown and the horn of Ammon; see LIMC V s.v. Hermanubis nos. 10a, b. The name Ἡρακλερμάνουβιϲ is more intriguing and difficult to account for. If the papyrus is from the Heracleopolite nome as the ed. princ. suggests with a question mark (no internal evidence confirms this), the name may simply juxtapose the chief god of the nome (Heracles = Egyptian Harsenuphis) with Hermanubis, without necessarily implying an affinity between the two. For what it is worth, a coin from the reign of Antoninus Pius (Milne, Catalogue of Alexandrian Coins 1678 [Plate II]) shows jugate busts of Heracles and "Hermanubis." For a representation of the dog-headed Anubis with a club, see LIMC I s.v. Anubis n. 35.

16. The form Ἑρμανούβιοϲ does not actually occur in Egyptian documents, though Ἀνούβιοϲ is occasionally attested. The Macedonian inscription discussed below (22) is the only instance of the form Ἑρμανούβιοϲ.

17. See Grenier 1990.

18. On this inscription see Fraser 1972, II 413–414 n. 576. For another inscription from Egypt mentioning Hermanubis, see SB I 238.

19. Fraser 1972, I 269. It is not clear to me to which of the statuettes in LIMC this corresponds (n. 21?).

20. On the associations of "Hermanubis" and Sarapis in iconography, see Grenier 1990, 268.

21. The acting epistrategus Hermanubis attested in 13 (and possibly 14) may also originate from Alexandria.

22. Cf. Plut., De Is. et Os. 375e: Ἄνουβιϲ ... ὅτε καὶ Ἑρμάνουβιϲ ὀνομάζεται, τὸ μὲν ὡϲ τοῖϲ ἄνω τὸ δ᾿ ὡϲ τοῖϲ κάτω προϲήκων.

23. Addendum, January 2010: The female name Ἑρμανουβίαινα (referring to an ἀϲτή, i.e. a citizen of Alexandria or one of the three Greek cities of Egypt) is now attested in P.Mich. inv. 1960.3 (17 March 291), published by L.H. Blumell, ZPE 165 (2008) 186–190. I have not been able to consult D. Stefanovic, "The Iconography of Hermanubis," in H. Györy (ed.), Aegyptus et Pannonia III. Acta Symposii anno 2004 (Budapest 2006) 271–276.

24. For a recent text and translation of this work, see Lukinovich and Martingay 1997. Ch. Jungck's edition with commentary (Diss. Basel 1974) was not available to me.

25. Gallay 1943, 166 n. 1 ("A notre avis, ces noms sont fictifs et leur accumulation est un trait satirique, d'autant qu'ils sont rapprochés à des dieux égyptiens à forme de singe ou de chien"). In Orat. 4, which incidentally is more favourable to Egypt and Peter, Gregory emphatically contrasts Egypt's "past" animal worship with its conversion to Christianity; see Smelik and Hemelrijk 1984, 1987.

26. For the otherwise unattested Ϲτιππᾶϲ, "tow-man," cf. the form Ϲιππᾶϲ attested several times in the fourth-century Douch ostraca; see O.Douch I 37.1 n. for references, to which add O.Douch III 228.3. Gregory may also be playfully alluding to Aristophanes' nickname for the politician Eucrates: καὶ ϲὺ κυρηβιοπῶλα Εὔκρατεϲ Ϲτύππαξ (fr. 716 K.-A.; the name is attested in a fifth-century BC inscription from Cyprus, cf. LGPN I s.n. and Bechtel 1898, 72, and as a slave name in P.Lille I 27.9 [III BC]). The reading of Μάξιμοϲ ⟨ϲ⟩τιπ̣π̣α̣ϲ̣, "Maximus the tow-worker(?)," in P.Oxy. XXXIV 2716.4 (302/3) is highly uncertain.

Works Cited