II. CULT EPITHETS IN MODERN WORK

Divine cult epithets,Footnote ¹⁶ which subdivide and categorize gods according to locality, function, or preferred mode of sacrifice or other type of honour (τιμή), offer a means to the better understanding of ancient Greek polytheism. That is why they have been the object of close study in recent years, culminating in the establishment of a valuable database by a team at Rennes university in France.Footnote ¹⁷ Scholarly efforts have been made to categorize such epithets, and so impose taxonomic order on a huge variety, but this is, as they say, like herding cats. A crucial distinction, already made in antiquity (see §III below) is between poetic epithets and cult titles. This will not concern us much, because so many of the epithets in Lykophron are said by the scholia to be place-specific, and that implies cult. One other main and valuable distinction has been hinted at already: that between epithets which are really, so to speak, the ethnic of the god (thus at 610 ‘Troizenia’ alone designates Aphrodite i.e. she is a ‘citizen’ of the polis of Troizen, see above), and functional or power epithets like Apollo Iatros, the doctor godFootnote ¹⁸ (he is Iatros alone at 1207, 1377).Footnote ¹⁹ But we soon run into trouble at the level of detail; thus although Apollo is the healer god par excellence, Iatros or ‘Doctor’ was also a title of his son Asklepios, unsurprisingly, and also, more surprisingly of Poseidon. Philochoros, quoted by Clement of Alexandria, says Poseidon was Iatros on the Kykladic island of Tenos. We have no idea why.Footnote ²⁰ This raises a general problem, but one specially acute with respect to Lykophron's scholia, namely that on the evidence of the grammarians and scholiasts, functional/descriptive epithets, on the one hand, and local epithets, on the other hand, overlap, sometimes bafflingly. The challenge is to explain the local variation, answers to which might give us a handle on the polis-specific character of Greek religion and its tension with Panhellenic religion.Footnote ²¹ Thus Hesychios says that Artemis is ‘Kaprophagos (‘Boar-eater’) at Samos', and this means that wild boar were abnormally sacrificed to her there, on a principle which Parker has explained with reference to Hera Aigophagos, goat-eating Hera, at Sparta.Footnote ²² Discovery of an inscribed sacred law might confirm these implications.

Let us return to Lykophron's Ἀμφίβαιος, which the scholiast told us was Poseidon among the Kyrenaians'.Footnote ²³ This epithet has been speculatively identified as a synonym for Ἀμφίγειος or Ἀμφίγειος, which is (this is yet more speculation) none other than the familiar Gaieochos, ‘Earth-shaker’ or ‘Earth-holder’.Footnote ²⁴ One might be tempted to dismiss this as a poetic epithet: surely there was no cult to the Earth-shaker any more than to Zeus Cloud-gatherer? But in fact there was cult to the Earth-shaker (I am assuming that is the right translation, rather than e.g. ‘Earth-holder’) at one classical city, namely Sparta, according to Hesychios and Pausanias; that rare thing, a lengthy fifth-century b.c. inscription from Sparta, namely the ‘Damonon inscription’, bears this out by mentioning a festival Gaiaochia in an agonistic context.Footnote ²⁵ (At Athens the cult is not attested before the second century b.c., and in so well-documented a religious centre, this absence in earlier centuries has some weight.Footnote ²⁶ A late sixth-century b.c. potsherd from Mende in Chalkidike bears a heavily restored inscription which might be relevant, but it is not certainly cultic.Footnote ²⁷) It is tempting to connect the Spartan cult with the area's well-known proneness to earthquakes. Sparta was, via Thera, the ‘grandmother’ city of Kyrene. Thus a cult which was appropriate in a colonizing city was inherited by the colony and then the colony's colony, even though Kyrene is not seismically active, and we would expect the Theraians to be more worried about the volcano god Hephaistos than about Poseidon of earthquakes, unless they knew that earthquakes and volcanic eruptions are related (as they are).Footnote ²⁸ So here is an example of a poetic epithet which had real cultic force in one city, and of colonies which adopted cultic name and cult without the specific motive which held for the metropolis.

III. VOCABULARY: EPITHET, EPIKLESIS AND OTHER TERMS

In this paper I shall use ‘epithet’ and ‘epiklesis’ indifferently. Greater precision has been claimed: it is held that ἐπίκλησις indicates actual cult, because of the derivation from ἐπικαλῶ, ‘I invoke’.Footnote ²⁹ This, to be sure, would provide us with a convenient and harmless modern technical term for a cult epithet. But it is not so harmless, if it implies that less loaded or ‘marked’ ancient Greek words for ‘name’, such as ἐπώνυμον or simple ὄνομα, did not sometimes have cultic force. If that is the implication, it is misleading, for two reasons. First, although ἐπίκλησις can certainly mean a cult epithet (for instance, Pausanias distinguishes ἐπίκλήσεις from poetic ‘names’, ὀνόματα),Footnote ³⁰ the word can also, from Homer onward, mean ‘nickname’ or even just ‘name’, with no religious implications.Footnote ³¹ Second, different Greek authors have different preferences or no preference at all: in definitely religious contexts Herodotus once uses ἐπίκλησις and once ἐπώνυμον, and on a third occasion uses both expressions in the course of a single sentence.Footnote ³² The Greek original of English ‘epithet’ is ἐπίθετον. Although LSJ⁹ does not list ‘divine epithet’ as a separate sense of the Greek word, it is used by Hesychios and others in that way.Footnote ³³ And there are many periphrastic expressions such as we have already encountered above (‘god x is honoured at Kyrene’ and the like); despite their looseness and informality, these certainly imply cult.Footnote ³⁴

IV. DIVINE POLYONYMY

Multiple epithets form a subcategory of polyonymy. The Dioskouroi, Kastor and Polydeukes (Latin Castor and Pollux) were polyonymous to an exceptional degree. Other names for them were the Guests (ξένοι, their epiklesis at Sparta, according to the scholiast on Lykophron),Footnote ³⁵ the Tyndaridai, and the Anakes; if we look up to the sky they are the Twins (but there were other Greek candidates for the constellation Gemini, which may anyway represent cultural transference from Babylonian Mash-Mash);Footnote ³⁶ and if we look back to Indo-European prototypes they are the Indo-European Ashvins. This is an aspect of polytheism which we could call poly-polytheism. Multiple asyndetic epithets are occasionally found in inscriptions, with no special emotional charge. In a sacred calendar from fourth-century b.c. Kos, a heifer is to be sacrificed to Hera Argeia Eleia Basileia.Footnote ³⁷ But in literary contexts, accumulation of divine epithets can have special power. This is specially true of prayers. There are multiple epithets – or divine ethnics – in Homeric prayers: ‘King Zeus, Dodonaian, Pelasgian, you who live far off’, Ζεῦ ἄνα, Δωδωναῖε, Πελασγικέ, τηλόθι ναίων (Hom. Il. 16.233), where ‘Pelasgian’ is vague, but not interchangeable with or a mere elaboration of ‘Dodonaian’. That is a prayer by Achilles (‘the most solemn in Homer’, according to Janko),Footnote ³⁸ and it is easy to see why in prayers a supplicant should wish to catch the god in as many manifestations as possible. But the attempt at completeness can bring its own risks: what if you miss one god who turned out to be vital?Footnote ³⁹ A well-known story told by Xenophon about himself in the Anabasis (7.8.4) shows that it had been a mistake, in the opinion of the seer whom he consulted, to have sacrificed to Zeus the King to the neglect of Zeus Meilichios. Burkert remarks that, in Greek prayers, ‘as much as possible, epithets are heaped one upon another …’.Footnote ⁴⁰ Tragedy reflects this type of polyonymy, perhaps taking its cue from epic. Euripides in the Trojan Women (a play well known to Lykophron, and one which features Kassandra prominently) has Zeus invoked by Hekabe as 1288–9 as ‘Son of Kronos, Phrygian lord, ancestor …’, Κρόνιε, πρύτανι Φρύγιε, γενέτα … (and compare Andromache 886, Δωδωναῖα, which may allude to the line of the Iliad quoted above). The same is true of divine invocations in oaths. Kretan inscribed inter-polis oaths of the Hellenistic period are specially rich in this sort of catch-all polyonymy.Footnote ⁴¹

Lykophron extends this technique to descriptive narrative, and characteristically takes it to extremes. But for an example of multiplied divine epithets in Lykophron which do form part of a genuine narrated prayer, see 359–60, where Kassandra invokes Athena as ‘helper’, ἀρωγός, to save her from her intending ravisher Ajax. We shall return to this important passage. See also 536–7 where Zeus, too, is an ἀρωγός with many epithets.

Hymns regularly celebrate a god as ‘having many names’, πολυώνυμος. Among the Orphic hymns, that to Hekate is specially dense in epithets, and there is much Hellenistic evidence, both literary and inscribed. The literary model for all such hymns must be the shorter Homeric hymns, for instance no. 1 to Hermes; but all this evidence (prayers as well as hymns) must reflect real-life religious behaviour.Footnote ⁴² Some evidence suggests awareness of the puzzles generated by plural divine naming. Prometheus says his mother is Gaia and Themis, ‘a single form with many names’, πολλῶν ὀνομάτων, μορφὴ μία.Footnote ⁴³ An unusual metrical inscription from Lykian Oinoanda, dated to the second century a.d., is an oracular response concerning the nature of god.Footnote ⁴⁴ It speaks of the god as ‘admitting no name, having many names’, οὔνομα μὴ χωρῶν, πολυώνυμος.

V. LYKOPHRON'S CULT EPITHETS: FIRST MENTIONS AND LITERARY FUNCTION

I now turn to Lykophron in particular. The Alexandra is a 1474-line iambic poem ascribed to a known tragic poet Lykophron of Euboian Chalkis, a member of the Pleiad, who lived in the early third century b.c., in the first part of the reign of Ptolemy II Philadelphos, who was sole ruler from 282–246 b.c. But – and this has been perceived as a problem since antiquity – the Alexandra, the main part of which is one long prediction after the event, foretells in a famous passage Roman ‘sceptre and monarchy over land and sea’ (1229). This, unless with some modern scholars we dilute it to the status of vague compliment,Footnote ⁴⁵ is impossible at the date at which Lykophron of Chalkis was active, because Rome then had no overseas province. So we must posit either an early second-century b.c. deutero-Lykophron (so that the Alexandra becomes pseudonymous or wrongly attributed); or else large-scale interpolation. I prefer the first of these two solutions, a second-century poem. The date problem can, however, be largely ignored here, because even if we think in terms of interpolation, the candidate sections (towards the end of the poem) contain little or no material relevant to our topic. But in one way the date might be relevant: if we regard cult epithets as proliferating in the course of the Hellenistic age, the likelier a later date for the poem becomes. Not many cult epithets in Lykophron are inconceivable before 200 b.c. But we shall consider one possible example later (below, p. 113 for Komyros).

Now for the poem itself. It culminates in an extended prophecy of east-west conflict, put into the mouth of Priam's most beautiful daughter,Footnote ⁴⁶ Kassandra, who is called in the poem by her Spartan epiklesis Alexandra; hence the poem's title. The poem begins with a short narrative introduction by the guard who has been set by Priam to watch over Kassandra in her stone cell. He describes how she emerged and prophesied. First, she predicts the fall of Troy to the Greeks, culminating in Lokrian Ajax's attempted rape of herself, Kassandra. The second, very long, section tells of the unhappy homecomings, nostoi, of the Greeks; it also recounts their pan-Mediterranean wanderings and founding of new cities when they cannot get home. This is pan-Mediterranean, because some heroes go to Asia Minor or Cyprus. There is a strong western (Italian and Sicilian) slant to the longest of the nostoi: Odysseus' adventures are retold at length from this viewpoint. Another sub-narrative locates Diomedes, Odysseus' accomplice in the Palladion theft, in SE Italy and the Adriatic. The woes of all returning or non-returning Greeks are presented as collective punishment for the crime of Kassandra's assault by one man, Ajax (see most explicitly 365, ‘in requital for the sin of one man’, ἑνὸς δὲ λώβης ἀντί …). Her speech closes with a Herodotean reprise of the entire east-west conflict in myth and history, including prehistoric Greek colonization of Asia Minor.

The piled-up, asyndetic, multiple cult epithets, which are such a characteristic feature of the poem (above, p. 91), have been little studied in recent years.Footnote ⁴⁷ The latest study, while full of subtle literary insights, does not concern itself with the evidence of epigraphy (and thus with the cultic realities); and it restricts itself almost entirely to Apollo, Athena and Aphrodite.Footnote ⁴⁸ Many epithets in the poem are not specially euphonious,Footnote ⁴⁹ so that a merely decorative explanation seems implausible. I deal first with a reductivist dismissal of my problem. It might run: ‘Lykophron is full of riddles, especially riddles involving names, and this is just another such riddle, and not worth stopping over’. I do not dispute the riddles, which are indeed everywhere (thus the name of the priapic god Orthanes at 538 refers to Priam's son Paris, because of his sexual excesses); but my subgroup is a specially formed set of name clusters, and it is the contention of this article that some at least of the information they conceal is of religious interest, even importance, and can be checked against documentary sources.

We first need to ask if the gods are the only entities so treated in the poem. Hellenistic poets were fond of making lists, a practice known as pinakography, and specially associated with Kallimachos. Lykophron arguably uses the device in one other way, namely for geographical specification. Four neighbouring regions, Euboia, Boiotia, Thessaly and Lokroi, are alluded to by named enumeration of cities, rivers, mountains and other local landmarks (373–5, 644–7, 900–7 and 1146–9). Euboia and Lokroi are specially important in the poem, because returning Greeks were shipwrecked off Euboia, and because the tribute of the Lokrian maidens is presented as punishment for Kassandra's assault by Lokrian Ajax: all the Lokrian towns will mourn their daughters. But this is not the same as strings of divine epithets, because it is standard rhetorical amplification, parts serving for a whole. Nor are the places listed asyndetically, but are linked with normal τε καί expressions, in the manner of Hesiod's list of Okeanids in the Theogony;Footnote ⁵⁰ and that is also true of the various named sibyls and sirens with which the guard compares Kassandra at the poem's end (1463–6). So from now on I concentrate on multiple cult epithets. The numerical prize goes to Athena: there are six at 355–9, an emotional passage describing Kassandra's assault by Ajax, when she appealed to Athena's statue the Palladion, which turned away in horror without helping. The second prize goes to Demeter, who – as we have seen – has five epithets at 152–3; but this is equalled by Athena – again – at 519–20.

A system of sorts is discernible in Lykophron's use of cult epithets. If this is a game, it is a game with rules. Only occasionally does the cluster include the most usual divine name, and the approximate system seems to be as follows. At the first occurrence of a god or goddess, the first name in the first-occurring list is often (though by no means always) closest to being a standard title, and is the least enigmatic. The limiting case of this is where the first appearance of a god in the poem uses the standard name with no obfuscation at all; this happens with Zeus, Hades and Ares. About half of the canonical Olympian twelve, and one or two others, can be accounted for in this way. Let us now consider Lykophron's treatment of the Olympians and some others (not all: for reasons of space, not every minor deity can be treated); we may begin with the three who are given their regular names.

Zeus gets special treatment (naturally): he is father of Herakles at 42, his first but indirect appearance, then ‘Zeus’ without mystification at 80, where he is bringer of torrential rain, cf. the full Rainy Zeus, Ζεὺς Ὄμβριος, at 160 (for this title see below §VI), and 622 where he is again associated with irrigation (ἀρδηθμός) and referred to straightforwardly: Διός. He is apostrophized by Kassandra as ‘Saviour Zeus’, Ζεῦ σῶτερ, at 512. (But note that ‘Zeus’ is sometimes Agamemnon and conversely; see below §VIII.)

Hades is mentioned early in the poem as a god (fought by Herakles as in the Iliad) not as a place, and with his regular name Ἅιδης, 51. But Hades is thereafter used for the place: 197, 404, etc. He is designated by the epithet Plouton near the poem's end (1420).

Ares is just that, Ἄρης, at 249–50, where he is personified (‘and now Ares, the dancer, sets fire to the land, leading the song with a bloody tune on his trumpet shell’, though it might be objected that those lines are just a fancy way of saying ‘war breaks out’); and also at 518. But he will receive three out-of-the way epithets at 937–8; for one of them (Μάμερτος) see n. 116. The specially blunt treatment of Ares in the earlier passages may be due to the violence which marks so much of the poem; or it may have something to do with the rarity of epikleseis for Ares generally.Footnote ⁵¹

Now we may consider gods who are not actually named on their first occurrence, but are nevertheless referred to transparently. First, Dionysos. The familiar ‘Bacchus’ at 206 (compare already ‘Bacchic mouth’, βακχεῖον στόμα, at 28, and ξεινοβάκχη, ‘madly in love with a stranger’, at 175) ushers in four much more obscure epithets at 207–12, including ‘He who trips’, Σφάλτης (207), a reference to the myth of Telephos (see below, §VI). The other three are ἐνόρχης Φιγαλεὺς Φαυστήριος (212), ‘Potent one [lit. ‘uncastrated’],Footnote ⁵² Phigaleian, Torch-god'.

Athena. The equally familiar ‘Pallas’ at 355 (the goddess' first appearance) precedes no fewer than five much harder epithets for Athena at 357 and 359 (Βουδείαν Αἴθυιαν Κόρην, ‘the Ox-binder, the Seagull goddess, the Maiden’), including one – Κόρη– which normally denotes another goddess, Persephone. This is six, the highest score (above). Lists later than the first make no concessions. At 519, ‘Thrice-born goddess’ (Τριγέννητος θεά) is the first in a list, but this time the title is by no means perspicuous.Footnote ⁵³

Apollo. The very familiar Δελφίνιος, associated particularly with Miletos where there was an important Apolline sanctuary, the Delphinion, is Apollo's epithet at 208. That is his first mention in the poem – unless we count the two implied references at 6, δαφνηφάγων φοίβαζεν, ‘laurel-eating, she Phoibized’ (i.e. prophesied), in which case Phoibos will be the first epithet, and is also very familiar; indeed, like Bacchos/Dionysos, it is almost another name for Apollo rather than an epithet. Lykophron never calls him by his two best-known epithets Πύθιος or Δήλιος, his ‘ethnics’ at Delphi and Delos; but for Boiotian Πτῷος see 265.Footnote ⁵⁴

Hekate is obviously the ‘dog-slaying goddess’ at 77, her first mention. Perhaps because she is not an Olympian in any sense (chthonian, and not one of the big twelve) and thus out of the mainstream already, her disguises are never hard to penetrate. She will be Brimo and ‘she of the three shapes’ (Τρίμορφος) at 1176, and ‘Pheraian’ at 1180. These are all familiar designations:Footnote ⁵⁵ under the name Enodia, Hekate was the ‘patron divinity’ of Thessalian Pherai.

Aphrodite is a halfway case. Arguably, her first mention is as the Homerically familiar Κύπρις, 112, but there it means ‘marriage’, and so approximates to a category which will be considered below: non-riddling non-periphrastic divine names for natural entities. But it does not quite fit that category because it is strictly an ethnic not a name, and so mildly periphrastic. If Kypris be disallowed, her first mention is as ‘the Kastnian’, 403, which is not obscure by Lykophron's standards (we saw at n. 13 that it also makes appearances in Kallimachos and a post-classical Pamphylian inscription) but is not common currency either, as are – for example – ‘Kythereian’ or ‘Kyprian’.

Other gods do not fit the suggested scheme nearly so well. Artemis does not feature explicitly until very late in the poem, though she hovers in the background of the Iphigeneia narrative which begins at 187. At 1331, she is unmistakably present in the description of the Amazon Antiope as ‘she who subdues with the bow, Orthosia’ (a hint at her patron Artemis under the epithet Orthia, see below), even before the next line, where Antiope's fellow Amazons are ‘sisters, virgins of Nepounis’ (= Artemis of Etruscan Nepe).

Poseidon is introduced as Proteus' father at 125 (cf. Zeus as father of Herakles, a much better-known paternal relationship). Then he is ‘Aigaion’ at 135 and ‘Naumedon’ at 157.

Hera is ‘second mother’, i.e. stepmother to Herakles, at her first mention: 39.

Hermes' first mention is as Kadmilos (162, cf. 219, Kadmos), a Samothracian epithet, well enough attested for Hermes, but hardly transparent (see further below, §VI). Thereafter he has several epithets, all out-of-the-way. See below, §VI for Boiotian ‘White Hermes’, at 680, an epithet which is combined with ‘Nonakrian, Three-headed’. The first of these is ArkadianFootnote ⁵⁶ and no doubt conjures up Styx and Hermes' chthonian role; the other is said by Philochoros to relate to Hermes' role as showing the way, i.e he stood at a road junction.Footnote ⁵⁷ The line in question immediately and appropriately precedes Lykophron's narrative of Odysseus' visit to the Underworld.

Hephaistos features only once in the poem, and that doubtfully, as Kandaon, 328. This is certainly obscure, so much so that the name may in fact refer to Ares, or even to Orion. The name Hephaistos occurs en clair at 1158, but as a synonym for fire; see below for this category.

Hestia nowhere features in the poem. Like Ares, she has few epikleseis anywhere in the Greek world (she shares Boulaia and its equivalents with one or two other gods and goddesses, notably Zeus, see §VI).

Herakles is the ‘lion begotten [by Zeus from Alkmene] in three nights’, τριεσπέρου λέοντος, at 33, Kassandra's very first sentence. At 801, ‘Herakles’ en clair is not the god/hero at all, but Herakles son of Alexander the Great and Barsine, a young man who was treacherously killed by Polyperchon of Tymphaia in 309 b.c. (Diod. Sic. 20.28.2–3). This, incidentally, is Lykophron's only mention by name of a historical individual.

In a separate category are divine names such as Ἠώς for Dawn (who at 16–19 is personified as ‘leaving Tithonos’, so that this means ‘at daybreak’, compare above on Ares), cf. the similar connotation of Τιτώ at 941; Thetis for the sea (22); and Hephaistos for fire (1158). These names are given in non-periphrastic form, but are already periphrastic in a different way. (See above for Κύπρις.)

We should next ask whether these accumulated epithets perform a particular literary function in the poem; in particular, whether we can detect what Homeric scholars call significant ‘denomination.’Footnote ⁵⁸ One section in particular stands out, the highly charged auto-narrative of Kassandra's sexual assualt by Ajax (348–72), and it has been convincingly argued that the gods are here designated in deliberately pointed ways.Footnote ⁵⁹ The first of Athena's epithets is neutral and indicative, as we saw: it is Pallas, 355. The next line (356) raises the emotional temperature: μισονύμφου Λαφρίας Πυλάτιδος. Of these the first is not quite a divine epiklesis, but it almost functions as one (cf. Κόρην at 359): ‘marriage-hater’, μισόνυμφος, at 356 refers both to Athena's own origin from the head of Zeus rather than by normal female birth,Footnote ⁶⁰ and to the role she played in the trial scene of Aeschylus' Eumenides. There she upheld the rights of the male against the female (‘I praise the male in all respects’, τὸ δ᾽ ἄρσεν᾽ αἰνῶ πάντα, Eum. 737, where however she adds ‘except for joining in marriage’, πλὴν γάμου τυχεῖν, perhaps a reference to her narrow avoidance of rape by Hephaistos). Λαφρία, goddess of spoils/booty, indicates Athena's role as war goddess (also used of her at 985 and 1416). λαφύρα is one of several words for booty, spoils taken from the living, as opposed to σκῦλα, taken from the dead. (Compare Artemis and Apollo Laphria/ios at Kalydon in Aitolia, a notorious piratical centre.Footnote ⁶¹) Athena was Πυλᾶτις as guardian of city gates. So the three epithets of 356 are not chosen arbitrarily; they all ‘call attention to crucial points of the scene’:Footnote ⁶² Kassandra herself as marriage-avoiding virgin and as living booty; Athena as protectress. Much further on in the poem (1207), Apollo's medical epithets Ἰατρός and Τερμινθεύς (below, 111 and n. 93) are very apt in their context – the Apolline oracle which ordered the fetching of Hektor's bones to Thebes for the avoidance of plague (1205, and for the bones and Theban cult of Hektor see nn. 49 and 129).

In the same section, Apollo's three epithets at 352 have also, but less compellingly, been seen as marked: the god's sexual appetite is held to be alluded to by Θοραῖος, on the hypothesis that that obscure epithet is derived from θορός, ‘semen’; and Ὠρίτης, ‘god of the seasons’, is held to hint paradoxically at the unseasonable nature of his attempt to have sex with Kassandra.Footnote ⁶³ For the third epithet, Πτῷος, see above, n. 49.

Again, consider the sequence Βοαρμία Λογγᾶτις Ὁμολωὶς Βία, at 520 (Athena again); the context is the fight between the Spartan Dioskouroi and the Messenian Apharetidai. The cult epithets have been held to be specially appropriate here because of the last in particular, Βία, which combines Athena with Ares and Enyo (named at 518–19) so as to indicate the violence which characterizes the episode. Boarmia is said by the scholiast to be Boiotian, and may be something to do with the yoking of oxen.Footnote ⁶⁴ But perhaps it should rather – compare above for the Ajax section – be taken in a passive and apotropaic sense: the goddess who does not wish to be yoked in marriage.Footnote ⁶⁵ This would be thematically appropriate, given that the fight between the two pairs of brothers broke out (on one version of the myth) over the nubile Leukkipides. ‘Longatis’ is, however, not easily amenable to any such interpretation. The scholiast again connects it with Boiotia, but the name may rather derive from a poorly attested Sicilian city called something like Longane.Footnote ⁶⁶ And generally, we must acknowledge that Lykophron's divine epithets are not always so well fitted to their literary contexts as are those discussed above. Kassandra's agony at 348–72 seems to have called for specially resonant treatment of Athena and Apollo. Other individual occurences of cult epithets can be explained contextually, to be sure;Footnote ⁶⁷ but there were more than three Olympians, and many puzzles remain.

The conclusion of this section is that Lykophron can often be shown to have devoted care and thought to the manner in which the cult epithets are introduced (easy first-time mentions); and that sometimes – but only sometimes – the poet chooses and positions cult epithets with sensitive regard to their narrative context. They are, then, not all randomly enigmatic. This conclusion encourages a further stage of inquiry: into their historical authenticity.

VI. THE EVIDENCE OF EPIGRAPHYFootnote ⁶⁸

Historians should be curious to know whether all, or some, or any, of these Lykophronic cult epithets are rooted in reality. The main control is, as always with ancient Greek religion, epigraphy.Footnote ⁶⁹ But inscriptions have in general (not just in the matter of cult epithets) been virtually ignored by all modern commentaries on Lykophron. To list omissions every time would be tedious polemic; it may be assumed that none of the epigraphic material cited below, or anywhere in the present article, is to be found in the existing commentaries unless I say so explicitly. I now look at some relevant examples.

Let us start with Linear B, and with Δρύμας, an epiklesis of Apollo at 522 (Apollo and Poseidon built the walls of the first Troy) and said by Tzetzes to be Milesian, Ἀπόλλων, παρὰ Μιλησίοις. A gold vase from Pylos (Tn 316) has the name Drimios, di-ri-mi-jo, who is son of Zeus. This is thought by Mycenaean specialists to be Apollo (who may also have been designated in Linear B by some form of the name Paion). Rougemont, in a very full recent discussion of cult epithets in Linear B, does not seek to explain or derive it.Footnote ⁷⁰ The vowels υ and ι are not identical, but Linear B expertsFootnote ⁷¹ nevertheless offer etymologies for di-ri-mi-jo from either δριμύς, ‘sharp’, ‘piercing’; or else δρυμός, ‘copse’, ‘thicket’ (compare below for Hylates).Footnote ⁷² Lykophron has, however, played virtually no part in the argument hitherto, though see n. 74. (In regard to spelling, we should recall – see p. 94 above on Kastnia – that Lykophron sometimes gives epithets in slightly eccentric forms.) There is a slight link between Miletos and Messenian Pylos, because Neileus son of Kodros, a younger kinsman of Pylian Nestor, founded Miletos, according to a myth treated in detail by Lykophron (1379; see also Hdt. 9.97). So Lykophron, when narrating an episode of vast antiquity (a episode from long before the main Trojan War), seems to have chosen a venerable epithet indeed. In the same line, however (522), Poseidon is ‘Prophantos, lord of Kromna’; these designations are not noticeably ancient.Footnote ⁷³

A different (but even more speculative) approach to the epithet Δρύμας might start from the Phokian place name Drymos. An inscribed Hellenistic agreement about financial matters, between Drymos and the Oitaian federation, is best interpreted as having an amphiktionic aspect. This would bring us to Apollo by another route, because one of the creditors was his sanctuary at Delphi. There is no reason why this approach should exclude the other, Mycenaean, explanation: in Lykophron, things are often neither settled nor stable. Indeed, the Phokian city has featured in explanations of the Mycenaean word di-ri-mi-jo.Footnote ⁷⁴ Demeter was ‘Erinys’ in Lykophron, as we saw in §I (152, cf. 1040, where she is ‘Telphousia’, a reference to the cult of Demeter Erinys at Arkadian Telphousa, and 1225, where ‘Onkaian pit’ is a further reference to this Demeter, named from another Arkadian city, Onkai or Onkeion). It is thought that a Mycenean goddess e-ri-nu, named on Linear B tablets from Knossos on Krete, preceded Demeter, and was eventually assimilated to her.Footnote ⁷⁵ This example, however, does not – unlike Drimios/Drymas – provide new and unexpected evidence for a cult title in Lykophron which was otherwise puzzling.

I now turn to Greek epigraphy as conventionally understood, that is, inscriptions in the alphabetical Greek script. (For some syllabic Cypriot texts, see n. 92.) In the most rewarding cases, inscriptions not only confirm the historical existence of a cult epithet, but show that it was at home in the geographical area to which Lykophron's scholiast assigns it; in other cases they confirm existence only. A good number of the better-known cult epithets used by Lykophron, but as always unaccompanied by the actual name of the god, are naturally epigraphically attested, some of them many times over: [Apollo] Delphinios (208) at Miletos and its numerous colonies,Footnote ⁷⁶ [Apollo] Zoster or Zosterios (1278) in Attica and elsewhere,Footnote ⁷⁷ [Apollo] Ptoios (265, 352) at the Ptoion in Boiotia,Footnote ⁷⁸ [Dionysos] Bacchos (206, 273) at Knidos in Karia,Footnote ⁷⁹ [Poseidon] Erechtheus at Athens,Footnote ⁸⁰ [Zeus] Ombrios, the ‘Rainy one’ (160) at Athens,Footnote ⁸¹ [Zeus] Boulaios, ‘of the Council’ (435) at many places,Footnote ⁸² Zeus Phyxios, ‘of fugitives or exiles’ (288) at Thessaly and elsewhere (above, n. 12). I shall not linger over these, because they help to solve no puzzles. My concern is with the light thrown by inscriptions on the more obscure epithets in the poem, of which there are many.

More relevant for this purpose are those inscriptions which attest a cult title which Lykophron attributes to a different god from the inscriptions; indeed, the surprising Zeus Erechtheus may fall into this category, see n. 80. For example (and this whole §VI is intended to be illustrative not complete), Homolois (520) is Athena ‘among the Thebans’ (Tzetzes; Σ incorrectly says ‘Athenians’), but that epithet was usually applied, mutatis mutandis, to Athena's father Zeus.Footnote ⁸³ At 1331, the name Orthosia (attested epigraphically as an epithet of Artemis at several places, such as Athens and Rhodes) is transferred to her follower, the Amazon Antiope.Footnote ⁸⁴ Hoplosmios (‘armed’) is attested epigraphically as an epithet of Zeus. But Hoplosmia is clearly Zeus' consort Hera at 858 and less clearly at 614. (We are told by the scholiast on 614 that this was a cult title of Athena at Elis, but this may be an error for Hera.Footnote ⁸⁵) Hera Hoplosmia features in the modern epigraphic literature in connection with a short and partly unintelligible Archaic Greek inscription from Paestum (Posidonia) in S. Italy. But the epithet, as opposed to the name Hera, is not present, and the connection seems to be a mere modern guess. So I do not claim it as corroboration of Lykophron.Footnote ⁸⁶

Again, Lykophron does not always confer divine epithets in a straightforward way. At 856 the ‘recesses of Lakinia’ means the grove at Kroton in S. Italy which Thetis gave as a present to Hera, whose well-established epithet Lakinia occurs epigraphically – but also in other types of evidence. The rare word σμίνθοι (‘mice’ or ‘rats’) at Lycoph. 1306, the first foundation of Troy, inevitably evokes the aetiology of Apollo Smintheus (cf. Hom. Il. 1.39), epigraphically attested at the excavated temple site near Alexandreia Troas.Footnote ⁸⁷

In the poem, we sometimes meet unrecognizably different forms of familiar, or at least attested, cult epithets. This may be because they were genuine local variants. Poseidon Amphibaios of Kyrene illustrates this, if he is really a doubly disguised version of Gaieochos, in which case cultic evidence, corroborating the literary sources, is provided by an inscription from Sparta, the grandmother city of Kyrene (see above n. 25). But sometimes the reason for selecting the different form is opaque to us. Thus Tzetzes tells us that Keramyntes (663) is Herakles Alexikakos, ‘driver-away of harm’,Footnote ⁸⁸ and that he was an apotropaic Herakles who drove away the Keres (the epithet Alexikakos was more usually applied to Apollo or even Zeus). Mythographers knew of a cult of Herakles Alexikakos, and there is epigraphic evidence for it from Athens, Chaironeia in Boiotia, and probably also from Epidauros.Footnote ⁸⁹ But it can hardly be claimed that these inscriptions shed much direct light on the epithet Keromyntes, and it is to examples of such direct corroboration that I now turn. The gods are treated in (English) alphabetical order.

I begin with Aphrodite. We have already (above p. 94 and n. 13) discussed her epithet Kastnia or Kastnietis, attested by Lykophron (403, 1234), Kallimachos and, spectacularly, by a Pamphylian inscription in Greek, of Roman date. Frustratingly, there is no mention of Pamphylia or any other region in the scholiastic tradition here, although Zeus's epithet ‘Drymnian’ at 536 is there said to be Pamphylian (so that that part of S. Asia Minor was within the scholiast's range of knowledge); this last epithet has so far resisted explanation.Footnote ⁹⁰

Apollo is referred to as Κερδῷος at 208. This cult of Apollo is epigraphically very well attested in Thessaly – and nowhere else. But no region is mentioned in connection with it, by either scholiast or Tzetzes. It is now thought that the meaning of the epiklesis is not ‘god of gain’ as has usually been assumed (beginning with Tzetzes), but ‘le Rusé’, from an adjective for cunning, applied to foxes.Footnote ⁹¹

Apollo is very appropriately called Ὑλάτης (from a place called Ὕλη near Cypriot Kourion according to the scholiast, but perhaps really ‘God of the Grove’, Silvanus) at 448, in the course of the opening sentence of the long Cypriot section of the poem (447–591). The important sanctuary of Apollo Hylates at Kourion on the south coast of the island has yielded many Greek dedicatory inscriptions dating from the middle of the second century b.c. to the Roman Imperial period.Footnote ⁹² There were three other sanctuaries to Apollo Hylates on Cyprus (at Paphos, at Dhrymou near Paphos and at Chytroi near modern Nikosia), and relevant Cypriot syllabic dedications from these places date from as early as the fourth century b.c. They have the epithet in the form u-la-ta.

Apollo is Λέψιος at 1207 and Λεψιεύς at 1454; these are ethnics of the Aegean island of Lepsia, W. of Karia. An inscription found on the island, and supplemented with virtual certainty, enables the identification of the sanctuary of Apollo Lepsieus. Also at 1207, Τερμινθεύς (again Apollo) represents ‘terebinth Apollo’ (named after a healing drug, cf. ‘Doctor’, Ἰατρός, at the beginning of the same line, 1207) whose temple features in an important Milesian inscription: τοῦ Ἀπόλλωνος τοῦ Τερβινθέως.Footnote ⁹³

Athena is Βούδεια at 359, as we have seen. The mention of a month Βουδιών in an inscribed sale list from fourth-century b.c. Kyzikos in Asia Minor has led to a conjecture that behind this lurks the Boudeia of Lykophron, and of Stephanus of Byzantion, who says that Athena Boudeia was a Thessalian cult. At 1261, Παλληνίς is the Athena worshipped in the Attic deme Pallene; she is well attested in fifth-century b.c. Attic epigraphy. The League of Athena Pallenis is known to us, not from a surviving inscription on stone, but from one quoted by Athenaeus, probably drawing on the Hellenistic epigrapher Polemon of Ilion.Footnote ⁹⁴ The cult was of lasting importance.

Dionysos is hidden behind the epithet Σφάλτης at 207. This refers to the myth of Telephos of Mysia (NW Asia Minor), who was tripped up (from σφάλλω) by Dionysos on the occasion of the first and misdirected Greek expedition against Troy. Lykophron says that this tripping up happened not because of divine anger against Telephos – the usual story – but because Agamemnon had prayed to Zeus, who therefore helped the Greeks. A mid second-century b.c. inscription from Delphi, a dedication of a stoa, purports to give the hexameter text of an oracle of Apollo (‘Loxias’) telling Agamemnon to sacrifice to Σφαλεώτας the ἄναξ, ‘king’, where the specification of sacrifice means that Σφαλεώτας looks like a cultic title. (The common Apolline epithet Loxias is suggested at the start and close of the poem, at 14 and 1467, but the correspondence hardly amounts to significant epigraphic corroboration.) This remarkable inscription shows that the mythological tradition, as recorded at length in the scholia and by Tzetzes, was correct to associate the epithet with Telephos of Mysia (‘correct’ in the sense that this association was taken for granted in the Hellenistic period, i.e. Lykophron's own time). Dionysos Ταῦρος, ‘the Bull’, is similarly attested both in Lykophron (209) and in a verse inscription from Delphi: the cult paian of Philodamos (340 b.c.). Θέοινος (Dionysos as Wine-god, see Lycoph. 1247) is not yet attested epigraphically, but the related Attic genos of the Theoinidai is so attested, in a late Hellenistic inscription.Footnote ⁹⁵

Herakles is called ‘Palaimon’ at 663. The epithet is epigraphically attested for Herakles at Boiotian Koroneia.Footnote ⁹⁶ The scholia and Tzetzes do not mention Boiotia, but merely explain it by reference to the root for wrestling, παλ-.

Next in the alphabet comes Hermes. The ‘shining god’, Φαιδρὸς θεός, at 680 is said by the scholiast to be ‘honoured as White Hermes among the Boiotians’, παρὰ Βοιωτοῖς λευκὸς Ἑρμῆς τιμᾶται. One attractive possibility is that this is an ephebic cult at Tanagra, in which case Tanagraian ephebic inscriptions, influenced by Athenian institutions, might become relevant.Footnote ⁹⁷ But there is no hard epigraphic evidence for such a cult, and we can only hope that some will appear. We have seen in §V that Hermes is called Kadmilos by Lykophron (162). The restoration Καδμ]ίλου at SEG 32.1077, Velia in Italy, date unclear, is very optimistic.

I will come later (§VIII) to the important epigraphic evidence for Kassandra's Spartan cult under the epiklesis Alexandra, and to the related cult of [Zeus] Agamemnon there. Lykophron mentions the second specifically, and was (it will be argued) well aware of the first.

Poseidon is teasingly alluded to in 767 by the name Ἱππηγέτης, and though Poseidon Hippios is familiar, and he is known by other ἱππ- titles, this is an otherwise unattested cult epithet. It is said by the scholiast and Tzetzes to denote Poseidon ‘among the Delians’. There is very slight epigraphic support from Delos for this Poseidon the ‘Horse-driver’ (or ‘Horse-begetter’, perhaps referring to his mating with Demeter and fathering of Arion): a very fragmentary inscription, perhaps naming Poseidon, has been found on a mosaic near what may (or may not) have been the site of the hippodrome.Footnote ⁹⁸

Komyros, i.e. Zeus (459), is one of the most epigraphically rewarding of the cult epithets in Lykophron. Its elucidation will take us to inland Karia in Asia Minor. In a section about Telamonian Ajax, Herakles (‘the lion’) sacrifices and prays successfully to his father Zeus, … καταίθων θύσθλα Κωμύρῳ λέων | σφῷ πατρὶ λάσκε τὰς ἐπηκόους λιτάς (459–60). The scholiast on 459 says that Zeus was honoured as Komyros at Halikarnassos, θύσθλα Κωμύρῳ· τῷ Διί· Κώμυρος γὰρ ὁ Ζεὺς ἐν Ἁλικαρνασσῷ τιμᾶται, and all modern commentators rightly repeat this, in modern language. But it is not enough. We have here a satisfying case where epigraphy confirms the information in the scholiast. ‘Halikarnassos’ stands for ‘Karia’ as part for whole, because a festival of the Komyrion or Komyria (Κομύριον, Κομύρια, sanctuary τὸ Κομύριον) is well attested epigraphically in inland Karia, at Panamara,Footnote ⁹⁹ a sanctuary site. The evidence is a series of inscriptions published in the late nineteenth and early twentieth centuries by G. Deschamps and G. Cousin in the Bulletin de Correspondance Hellénique. The first editors of the inscriptions immediately (1888) made the connection with Lykophron, and called Panamara a place of pilgrimage. Holzinger (1895) and Ciaceri (1902), in their commentaries on the poem, were therefore in a position to know this evidence, but did not mention it, and their successors have followed suit.Footnote ¹⁰⁰ Note the spelling Κυμώριος at Bargylia, N. of Halikarnassos. (This inscription may indicate that the Panamaran cult was indeed Karia-wide, and did indeed attract ‘pilgrims’ from the Halikarnassos region.) It is thought that Komyros was originally a Karian god who was gradually assimilated to Zeus. The Komyria was/were a long-lived mystery (initiatory) festival.Footnote ¹⁰¹ The female festival was the Heraia, the male the Komyria, but the priest of Zeus Panamaros officiated at both. The inscriptions are of Imperial Roman date, but Lykophron entitles us to push the cult back into the Hellenistic period.

The local Karian epigraphic spelling of this non-Greek name is usually with Κομ-, not Κωμ-, but this seems to have fluctuated (see above for Κυμ- at Bargylia). In any case, the lengthening of omicron to omega may be due to Lykophron, who needed a long syllable at that point in the line. Compare the lengthened iota of Σικανός (1029, contrast Callim. Hymn 3.57. A partial model for Lykophron was Hom. Od. 24.307, Σικανίης); or the fluctuation within the poem itself between Σιθῶνος and Σιθόνων (583, 1357, cf. LSJ).Footnote ¹⁰² In all these examples, choice was dictated by metre.

The Karian epiklesis might be used as an argument (like the poem's lengthy attention to Cyprus) in favour of an Alexandrian milieu for the poet, because of extensive Ptolemaic Egyptian links with Karia, and actual Ptolemaic control of Stratonikeia itself in the early third century.Footnote ¹⁰³ There are two other cult epithets in the poem which can loosely be called Karian: [Athena] Myndia at 950 and 1261, the ethnic of Myndos near Halikarnassos (IACP no. 914), and [Apollo] Λέψιος or Λεψιεύς at 1207 and 1454, from the island of Lepsia to the west of Karia (above, with n. 93).

I will stay with Komyros a moment longer in order to return to the poem's date (cf. above, p. 101). It can perhaps be argued that Greek knowledge of Komyros, an obscure inland Karian god fused with Zeus over a period of time, is more likely to feature in a Greek poem at the end of the third century b.c. than the beginning, given that Stratonikeia was not founded until the 270s. Hellenism came to inland Karia in a big way under Rhodian influence in the mid third. But as we saw, there was a Ptolemaic presence in the Stratonikeia region in the early third, so the argument cannot be pressed.

Zeus in the form Γογγυλάτης (435) is a particularly difficult epiklesis. The epithet is said (LSJ, following the scholia and Tzetzes) to mean ‘hurling balls of fire’, from γόγγυλος, ‘round’. That is, perhaps, the god of the thunderbolt.Footnote ¹⁰⁴ A first-century a.d. inscription from Thessaloniki attests a possibly Thracian god ‘Zeus Dionysos Gongylos’.Footnote ¹⁰⁵ This does not solve the problem of the epithet's meaning or etymology, nor is the context of 435 obviously Thracian, but it might indicate that Lykophron did not pluck the epithet out of the air.

A sceptic might say, ‘we grant you Komyros and those others, but how many difficult cult epithets in Lykophron can be so paralleled?’ Obviously, I have chosen some of the more satisfying cases. Many, it must be conceded, remain impenetrable. Tzetzes says that Aithiops and Gyrapsios (536–7) are epithets of Zeus at Chios. GrafFootnote ¹⁰⁶ draws a blank, merely citing Lykophron and A.B. Cook for Gyrapsios. Zeus's epithet Promantheus at 537 is supposed to be related to a Sanskrit word for fire.Footnote ¹⁰⁷ For these and others there is no epigraphic corroboration. But we can hope and expect that the discovery of new inscriptions, and further work on existing ones, will allow more progress.

Another challenge is to find pattern to or significance in the geographical locations of the epithets, assuming that our information is reliable. The Zeus chain at 536–7 runs ‘Drymnios Promantheus Aithiops Gyrapsios’, and Tzetzes says that these are Pamphylian, Thourian, Chian and Chian. It is not obvious what links these places; or why Apollo is Derainos (n. 15), i.e Abderite, in the Mopsos episode, which is located in Kilikia. But we should not despair. We have seen that Hylates is Apollo, and the epithet is used about Cyprus where there are (n. 92) several sanctuaries with dedications to Apollo Hylates. At 589, ‘Golgo's Queen’ is Aphrodite. Golgian, from another Cypriot city, is an accepted epithet for her, even in Latin poetry (Theoc. Id. 15.100; Catull. 36.14; 64.96). More subtle is the designation of Odysseus as ‘thief of the Phoenician goddess’, that is, of the Palladion (658). The scholiast and Tzetzes say this was an epithet of Athena at Korinth (a city famously open to eastern cultural penetration through its harbour at Kenchreai). Perhaps this is an oblique reminder of the role played by an invented Phoenician in Odysseus' lying story, told soon after his arrival in disguise at Ithaka (Od. 14.287–91). The poet's own acquaintance with Alexandria may be relevant to some epithets. At 449 and 958 Zerynthia is Aphrodite on Samothrace (cf. 1178, where Zerynthia denotes Hekate), which was a Ptolemaic overseas possession – like Cyprus, and indeed parts of Karia.

VII. THE SOURCES OF THE CULTIC INFORMATION IN POEM AND SCHOLIA

I hope to have shown that at least some of the cultic information in the scholia is impressively corroborated by epigraphy, and we must now ask where this information actually came from. Wentzel's thesis was that the paroemiographer Diogenianus, Athenaeus, Clement of Alexandria, the scholia to Lykophron and Pausanias all drew on a single lost source, a Hellenistic collection or sylloge of cult epithets. Methodologically, this resembles Fraser's ingenious theory that the various lists of Alexander the Great's city foundations all go back to a single lost treatise of Ptolemaic Alexandrian origin, an anti-Seleukid liber de urbibus Alexandri magni. Footnote ¹⁰⁸ But Wentzel was thinking, not of political propaganda, but of scholarly short cuts. Pseudo-learned and (no doubt) genuinely learned folk made frequent use of such mythological handbooks, but sought to conceal this shameful dependence.Footnote ¹⁰⁹ Wentzel thought that the hypothetical sylloge dated to the first century b.c. Theon of AlexandriaFootnote ¹¹⁰ was perhaps the earliest to use it, also in the first century b.c.; he wrote commentaries on several Hellenistic poets, including Lykophron. Some of Wentzel's parallels between his extant ancient authorities are too well known to be probative; some do not quite work. An example of the latter: Enorches at 212 (above, n. 52) is said by the scholiast on Lykophron to be an epithet of Dionysos on Lesbos, but Hesychios says it was Samos. Nevertheless, Wentzel's main idea is attractive and was closely and well argued.

Let us, then, provisionally accept the hypothesis of a sylloge. Might the poet Lykophron, rather than the poet's scholiasts, have used some such compilation? Wentzel does not consider this radical possibility. If this were so – and it is not obvious how one might go about proving it – that would reduce the poem's originality, but not the value of its epikletic material, or the interest of the way in which the poet exploited, deployed and arranged it. But we do not have to imagine the poet taking a mindlessly short cut, nor (at the other extreme) possessing exhaustive first-hand knowledge of dozens of cult places. The poet could have drawn on a range of scholarly material; and it was not necessary to travel to every polis in the Greek world to discover what cult epithet was used where. We know that, for example, the chronographer Apollodoros (second century b.c.), studied divine epithets.Footnote ¹¹¹ And many of the local historians of the Hellenistic period were keen to advertise the achievements of their local gods, for instance by collecting lists of epiphanies; today, we know about these historians above all from inscriptions praising them.Footnote ¹¹² But vast amounts of such (now mostly lost) erudition were once available in full and in detail on papyrus rolls, whether in great libraries like those at Alexandria, Antioch or Pergamon, or on Hiero II's giant book-barge at Syracuse,Footnote ¹¹³ or in the smaller libraries at Rhodes or Sicilian Tauromenion,Footnote ¹¹⁴ or even perhaps in the temple archives at Italian Lokroi. A final thought: Lobel once suggested that Euphorion wrote a commentary on one of his own poems,Footnote ¹¹⁵ rather like T.S. Eliot supplying notes on The Waste Land. Lykophron might have done the same.

VIII. LOCAL AND PANHELLENIC RELIGION

Finally, I turn to the relation between local and Panhellenic religion, and ask how far Lykophron's epikleseis help with this topic. By Panhellenic I mean the Mediterranean and Black Sea regions. As we saw above (p. 95), the epikleseis do not reflect the newer eastern Hellenistic colonial world. To be sure, Lykophron uses epithets said by the scholia to be at home in the east: Phoenicia, Cyprus, Pamphylia, Cilicia; add Egypt: Osiris is the ‘Memphite leader’ at 1294. But these regions were known to Herodotus and indeed Homer. Nevertheless I hope to have shown that, although this poem was composed in a particular area (S. Italy, I think),Footnote ¹¹⁶ some of the information tying Lykophron's names to various and far-flung cult places is high-grade religious evidence. In Lykophron the gods are Panhellenic, the epikleseis often local, even where they do not actually take the form of ethnics.

A classic study by Sourvinou-Inwood explored the tension between Panhellenic and local religion in S. Italian Lokroi.Footnote ¹¹⁷ She showed that in Lokroi, Persephone's rape by Hades caused her to exercise a marital function normally exercised by Aphrodite. How do epikleseis in Lykophron throw light on or exemplify local versus Panhellenic religion? I end with Kassandra herself and her spouse, as she calls him, Agamemnon; in what follows I am indebted to an important PhD dissertation by Giulia Biffis.Footnote ¹¹⁸ It seems that what Sourvinou-Inwood argued for Persephone was also true of the cult paid, on the evidence of 1126–40, to Kassandra herself in S. Italy, in Daunia (modern N. Puglia); this episode is connected thematically to the Lokrian Maidens section which immediately follows it. Lykophron never says explicitly that Kassandra received cult in Sparta/Amyklai, but is surely well aware of the fact. The connection between Daunia and Sparta runs via Taras/Tarentum, Sparta's most famous historical colony. Daunia is not far north of Taras, and contacts between Daunia and Taras were particularly close;Footnote ¹¹⁹ in Lykophron, Kassandra's Daunian cult is described straight after the prophecy that Agamemnon ‘will be called Zeus by crafty Spartans’.Footnote ¹²⁰

Alexandra is the only name the poet ever gives to Kassandra, 30 (the last line of the guard's preamble). Alex-andra, ‘she who wards off men’, is originally a military name, the female equivalent of Alexandros, her brother Paris.Footnote ¹²¹ But it also evokes Kassandra's sexual rejection of Apollo, and Tzetzes explains it in terms of her virginity: παρὰ τὸ ἀλέξειν τοὺς ἄνδρας· παρθένος γὰρ ἦν. Pausanias and Hesychios tell us that Alexandra is an epiklesis of Kassandra at Sparta, in fact at nearby Amyklai. It is rightly entered as such in Wentzel's Appendix of epikleseis. Since Wentzel wrote, inscribed sherds and plaques, found in excavations from the 1960s at Amyklai, confirm this: ΑΛΕΞ- and ΑΓ-. There was, in fact, joint cult of Kassandra and Agamemnon.Footnote ¹²² The cult was possibly chthonian.Footnote ¹²³ Now, as we just saw, Kassandra's spouse (ἀκοίτης, 1123), Agamemnon, was called Zeus at Sparta. He was perhaps divinized there only as a function or adjunct of Kassandra, who seems interestingly to be the senior partner in the joint cult. Kassandra was, in fact, heroized.Footnote ¹²⁴ So the problem of Zeus Agamemnon should not, as is usual, be treated separately from the heroine status of Kassandra/Alexandra. Agamemnon is like Enyo, Eilythuia, Komyros or Derainos: a minor local figure merged with a major one, in this case Zeus.Footnote ¹²⁵ We have, then, an adaptation of Panhellenic deities to local cultic purposes, compare Sourvinou-Inwood on Lokrian Persephone; and the adaptation is expressed in epikleseis for which Lykophron and scholia are the chief, but not the only, literary sources (n. 120 for Staphylos). For the Amyklaian location, Lykophron seems to have followed Pindar, Pythian 11. In Pindar, who was himself perhaps following Stesichoros, the killing of Agamemnon and Kassandra by Klytemnestra happened at Amyklai, not at Mykenai as in Homer, or Argos as in Aeschylus' Agamemnon.

But there is a further and Italian aspect to the cult of Kassandra. It is in this Italian, but (via Taras) Spartan-influenced, aspect that the marriage rejector Alexandra comes into her own: Lykophron makes Kassandra predict that she herself will be given pre-marital cult by girls in Daunia who wish to avoid marriage because of some blemish in their suitors (1126–41, esp. 1131–4). But their repugnance is surely temporary and simulated. In other words, Kassandra is initiatory in both Sparta – where she ‘wards off men’ – and in Italy: by a familiar sort of reversal, the mythical female who rejects marriage becomes its patron.Footnote ¹²⁶