INTRODUCTION
Archaic Greek epos is full of references to different places. Achilles the Phthian, Odysseus the Ithacan, Helen the Spartan … nearly every character receives an origin in a particular locality, and the genre abounds in catalogues of different places and peoples, from the Iliad’s Catalogue of Ships to the circuitous journeys of the Homeric Hymn to Apollo. An influential theory of Nagy holds that these details derive from local traditions around the Greek world: Homeric epos ‘synthesizes the diverse local traditions of each major city-state into a unified Panhellenic model that suits most city-states but corresponds exactly to none’.Footnote 1 Nagy attributes this synthesis to a phase of increased intercommunication between Greek states in the eighth century, adducing the rise of the Olympic games, the Delphic oracle and the alphabet. According to this theory, the Iliad and Odyssey poets drew on a number of local traditions in various polities, eliding local details for a wider appeal: city-specific hero cults are suppressed; foundation legends for cities in Asia Minor involving Achilles are omitted; the practice of giving cuts of meat to the gods is ignored, because the choice of cut varied locally.Footnote 2 Following Griffin, the other poems of the Epic Cycle are considered ‘radically different’ from the Iliad and Odyssey, lacking ‘a truly Panhellenic status’.Footnote 3 This approach encourages us to posit local traditions behind particular stories and details in epos—by which the poems themselves may then be interpreted.
Nagy’s theory has seen widespread—though not uncritical—acceptance; many have enthusiastically posited original local traditions behind epos.Footnote 4 But it raises certain issues. Some are matters of detail. The Hymn to Apollo may not be a fusion of Delian and Pythian sections.Footnote 5 Griffin’s view that the lost Cyclic poems are utterly unlike the Iliad and Odyssey is controversial,Footnote 6 and surviving evidence shows a similar focus on the ‘Panhellenic’ Trojan expedition, rather than epichoric stories. Panhellenic ideology is less prominent before the Persian Wars; given the Greek dialects were mutually intelligible, Panhellenic identity may always have existed as a potential,Footnote 7 but the Olympic Games and Delphic oracle are unlikely to have had more than regional significance in the eighth century.Footnote 8
Other problems are more general. A poem can be local (or Panhellenic) in at least three ways: content, origin and intended audience. The content might refer to a certain locality or a ‘Panhellenic’ combination of different places, or to no particular place. It might originate in one polity, where a poet lives and composes; or in many, if the poet moves around. It might be performed for a purely local audience, or a wider one (as at a regional festival). Certainly, a tradition developing in a particular place may well be particularly concerned with local characters,Footnote 9 and a poet performing for a local audience might exploit local material; but nothing prevents a tradition forming in, say, Argos about events in Thebes. For any particularly local-seeming feature—say, the role of Anchises’ descendants in the Homeric Hymn to Aphrodite—one may hypothesize a local origin in the Troad, or performance to local aristocrats who claimed descent from Anchises.Footnote 10 But it would be naive to assume that every element of local content derives straightforwardly from local origin and audience.
Secondly, should we really envisage a number of more-or-less isolated local hexameter traditions in different polities? The surviving poems appear exclusively Panhellenic (featuring characters from around the Greek world);Footnote 11 evidence for comparable epichoric hexameter is absent.Footnote 12 One might argue that truly local hexameter lacked popularity, vanishing without trace; but several early elegiac poems related epichoric history or foundation narratives,Footnote 13 without initially suffering unpopularity. Furthermore, local(ly originating) hexameter traditions imply that bards stayed in particular polities, passing down techniques and stories to successors there; but bards depicted in epos, and the (admittedly not highly reliable) biographical traditions for archaic poets, suggest considerable mobility.Footnote 14 In the Odyssey, although Demodocus and Phemius stay in particular locations (Phaeacia, Ithaca), at 17.380–7 bards are one of the craftsmen typically invited abroad. The persona of the Works and Days feigns ignorance of sailing, but travelled to Chalcis to perform at Amphidamas’ funeral; his father moved to Ascra from Aeolian Cyme. The narrator of the Hymn to Apollo (174–5) lives in Chios, performs in Delos, and travels widely. We find no notion of local poets passing down traditions in a particular polity. The existence of the oral tradition implies continuous transmission, but not necessarily in one particular place.
Thirdly, on Nagy’s model, the Mycenaean collapse created isolated units in which local traditions formed, before increasing connectivity in the eighth century brought them back together.Footnote 15 More recent archaeological evidence challenges this idea of a Greek ‘dark age’;Footnote 16 although the crisis at the end of the Late Bronze Age disrupted connections within the Greek world, the different areas were not so isolated.Footnote 17 Pottery distributions demonstrate continued exchange in the Early Iron Age within the Greek world and beyond.Footnote 18 Trade may not precisely reflect population movement, but does suggest that any local traditions would have been in contact with each other.
Moreover, all epos—even poems purportedly from a non-Ionic-speaking area, like the Hesiodic poems—largely display Ionic dialect, with Aeolic elements (sometimes attributed to an Aeolic phase of the tradition, or the replacement of archaisms with dialect forms).Footnote 19 Can we really postulate a line of, say, Knossian poets immersed in Cretan Doric, but passing on an Ionic Kunstsprache? Archaic poetry is never in pure spoken local dialect; literary Lesbian, Ionic and Doric always includes archaisms, borrowings and complications.Footnote 20 On one explanation, the distinctive formal genres (epic, elegiac, choral lyric, etc.)—and literary dialects—developed in separate regions, then spread whilst maintaining dialect associations.Footnote 21 Alternatively, each area originally used its own dialect for all genres, literary dialects becoming fixed through particular models.Footnote 22 The issue is further complicated by the uncertain prehistory of the dialects and their relationship with Mycenaean,Footnote 23 making it a difficult starting point to evaluate the status of local traditions.Footnote 24
Wholesale re-examination of these controversies requires book-length treatment.Footnote 25 The more modest aim here is to examine three significant epic characters—Nestor, Iris and Sarpedon—and the evidence for their prehistory, drawing on evidence from etymology, formulaic diction, narrative and non-epic depictions. Did they develop in local traditions, before being absorbed into epos? In each case, the situation is more complex than anticipated.
I. NESTOR
Nestor was popular in early epos, appearing extensively in the Iliad and hosting Telemachus in Pylos in the Odyssey. He is also mentioned in the Catalogue of Women and the Epic Cycle. Although numerous studies explore his links with local traditions, many are extremely speculative; the issues therefore deserve reconsideration.Footnote 26
1. Pylian tradition?
Nestor is a Pylian, so we might suppose that he originates in a Pylian local tradition,Footnote 27 encouraged by the Iliad’s narratives of his youth: 7.132–57 (his fight with Ereuthalion at Pheia), 11.670–761 (battle with the Eleians) and 23.629–45 (Amarynceus’ funeral games). It might seem easy to suppose that such exploits derive from local Pylian stories. Some, however, judge them ad hoc creations, fitting Nestor’s rhetorical needs:Footnote 28 in Book 7, encouraging the Greeks to volunteer against Hector by recounting his own duel; in 11, urging Patroclus to similar exploits; in 23, displaying his past athleticism. But rhetorical function does not prove untraditionality. In the case of the Ereuthalion narrative, a mention at 4.219 anticipates the story at 7.136–56, suggesting that it is not an innovation.Footnote 29 Moreover, the narratives mention many specific place names, unlikely to be the poet’s inventions.
The specificity of the Elean battle is itself problematic, however. The Pylians go from Pylos to the city of Thryoessa in half a day (11.711, 723–6); next morning, the Epeians are routed as far as Buprasium (754–8). But Messenian Pylos is more than seventy miles from Thryoessa;Footnote 30 Buprasium is similarly unrealistically far north from the Alpheios.Footnote 31 Some scholars delete certain of the offending details,Footnote 32 or accept Strabo’s argument (8.3.26–9) that the Homeric Pylos was in Triphylia.Footnote 33 But outside Strabo (and Didymus, perhaps Strabo’s sourceFootnote 34 ) this Triphylian Pylos is scarcely if at all mentioned,Footnote 35 unlike neighbouring Lepreum; Strabo himself says little about it apart from arguing that it is Homeric Pylos, and saying it was synoecized into Lepreum some time after the Second Messenian War (8.3.30). Even if we were to accept the Triphylian Pylos, there remains the implausible distance to Buprasium. It seems preferable to accept that the narrative is unrealistic. Perhaps this emphasizes the superhuman valour of the Pylians and Eleans (though nothing else in the story particularly supports this); or the original details might have become confused; or the original composer might have employed local place-names, but not a genuinely local story.
A further factor bearing on the existence of local Pylian tradition is the region’s history. Around 1200 b.c. the Mycenaean palace of Pylos at Ano Englianos was burnt, one of a series of destructions comprising the ‘Late Bronze Age Collapse’. It was not rebuilt; subsequent habitation at the site was scanty at best.Footnote 36 However, a reduced population apparently continued to live in the region, building a small tholos tomb near Korifasi in the Protogeometric period and a settlement and archaic temple near Romanos, closer to the coast.Footnote 37 That the name Pylos survives itself demonstrates some continuity, in Thucydides (4.3.2) referring to the coastal promontory near the former palace; by then the area was sparsely populated, presumably due to the Spartan occupation. In the second century a.d., Pausanias records Pylian local traditions (4.36), including a ‘house of Nestor’, tomb and cave where Nestor kept his cattle. But given the long Spartan rule, these may be innovations created after Messenian independence.Footnote 38 In short, the archaeology is at least compatible with stories about Nestor being passed down in the region until the Spartan conquest.
2. Ionian tradition?
Nestor’s name has the root *nes- (cf. νέομαι, ‘return’), paralleled in Mycenaean ne-e-ra-wo /Ne(h)elāwos/ (PY Fn 79.5).Footnote 39 His father Neleus/Νηλεύς may show the same root—perhaps even an -ευς suffix remodelling of /Ne(h)elāwos/.Footnote 40 This cannot prove that Nestor was a real king at Pylos; we have little comparative evidence for Mycenaean onomastics elsewhere, and the Mycenaean ne-e-ra-wo does not seem to have been of high status. But Mühlestein used it as evidence, not just for Nestor’s Pylian origin, but also to validate the historical tradition by which Nestor’s descendants, particularly Neileus, founded Miletus and the Ionian cities of Asia Minor.Footnote 41
The reality of EIA population movements is too controversial to be evaluated here,Footnote 42 but there is good evidence for a tradition, by the Archaic period, that Pylians founded certain Ionian cities: for Colophon, Mimnermus fr. 9–10 West, and for Miletus, an inscribed bronze bucket from the first half of the sixth century, dedicated in the Samian Heraion by a priest of Ne(i)leus;Footnote 43 the obvious reason for the cult would be Neileus’ role as founder. Hence one may argue that the real focus for local traditions about Nestor was in Ionian cities where he was considered the ancestor of the city founder and of certain aristocratic families—especially the Neleids in Miletus.Footnote 44
Given the Ionian epic language, an Ionian origin for Nestor and his stories may seem appealing. It might seem strange that Ionian traditions would preserve obscure place-names and details of battles from Bronze Age Pylos, but this fits the idea that these have been corrupted in transmission. However, the Milesian founder was called Νείλεως, not the Homeric form Νηλεύς.Footnote 45 The suffix -εύς is explicable as a shorter form of the name,Footnote 46 but the difference between -ει- and -η- (contracted -εε-) shows that Νηλεύς derives from a dialect in which long ē from contraction merges with inherited long ē, unlike Attic-Ionic. This suggests strongly that Νηλεύς—and Nestor—entered epos from a non-Ionic source.Footnote 47
3. Wider connections
Nestor’s connections go beyond Pylos and Ionian Asia Minor. Nestor’s father Neleus is Pelias’ twin, born at Iolkos to Poseidon and Tyro (Od. 11.235–53, [Hes.] frr. 30, 30a M–W). Neleus’ daughter Pero features in the story of Melampus (Od. 11.281–97, 15.225–42); he became ruler of Argos (Od. 15.240, [Hes.] fr. 131 M–W), possibly dying at Corinth.Footnote 48 If we keep positing local origin on the basis of associations of content (as with Pylos) or appropriate audience (as with Miletus/Ionia), we must now add Thessalian, Argive and perhaps Corinthian traditions.
One might explain this multiplicity by diachronic layering. Thus West supposes that Nestor’s traditions originated in Pylos, spread to Thessaly in the twelfth-eleventh centuries, then passed on to Aeolic Asia Minor, eventually to Ionia. Mühlestein prefers a migration Thessaly–Pylos–Ionia.Footnote 49 The contradiction reflects different starting points—for West, the supposed Achaean, Aeolic and Ionic phases of epic language; for Mühlestein, a euhemeristic approach to Neleus’ exile to Messenia and Ionian migration legends. Besides their incompatibility, both neglect some details (such as Pero, and Heracles’ sack of Pylos).
Moreover, heroic exploits themselves are rarely confined to a particular location. The very ‘Pylian’ narratives we started with—the Eleian war and the funeral games at Buprasium—involve Eleian heroes like the Molione. Are they then influenced by Eleian local tradition?Footnote 50 If we suppose that these characters originated in distinct local traditions—with an all-Pylian or all-Eleian cast—then no trace of these original, ‘pure’ local stories actually survives. This is not to completely deny any connection between content and origin; but a situation in which isolated, distinct local strands have been superficially combined seems improbable.
4. Nestor’s traditional language
If Nestor’s youthful Peloponnesian exploits represent his original, local tradition, his elderly appearance at Troy is a secondary, ‘Panhellenic’ development. Nestor’s traditional diction, however, shows him as an older, adviser figure.Footnote 51 Νηληϊάδαο γέροντος,Footnote 52 and frequently γέρων, γεραιός, fits the ‘old man’ at Troy, not the young Pylian. The formula Νέστωρ –⏑⏑–⏑ Γερήνιος οὖρος Ἀχαιῶν,Footnote 53 though likely referring to a Messenian town,Footnote 54 combines this with his status within the larger body of Achaeans; similarly the formula ὦ Νέστορ Νηληϊάδη μέγα κῦδος Ἀχαιῶν.Footnote 55 Even λιγὺς Πυλίων ἀγορητής,Footnote 56 indisputably referencing his Pylian origin, fits Nestor’s Iliadic portrayal as wise adviser better than his warlike Pylian reminiscences. The description of Nestor’s great age, shared between the Iliad and the Odyssey, is thus probably traditional.Footnote 57 Furthermore, each of his paranarratives begins εἴθ’ ὣς ἡβώοιμι βίη τέ μοι ἔμπεδος εἴη | ὡς ὁπότε …,Footnote 58 presumably a traditional formula. Now, formulae may be transferred between characters and reworked over time; but such rich formulaic diction for Nestor the aged adviser is good evidence that Nestor has long held this role; he was not recently borrowed as a young warrior from Pylian local tradition.Footnote 59
5. Iconography and epigraphy
A possible eighth-century depiction of Nestor is identified through his Elean antagonists, the Molione (who, being conjoined twins, are easier to distinguish in the abstract Geometric style). An LG oinochoe produced and deposited at Athens may depict their flight from Nestor (Il. 11.750–2),Footnote 60 perhaps indicating the importance of stories about Nestor in Athens; Coldstream suggested a connection with Athenian aristocrats who claimed Neleid ancestry.Footnote 61 On another interpretation, however, the oinochoe might reflect an artistic stylization, not conjoined twins.Footnote 62 Unambiguous depictions of Nestor appear in the early sixth century,Footnote 63 with wide geographical range, focussing on his involvement at Troy.
The inscribed ‘Nestor’s Cup’ from Pithecusae has a claim to ownership by ‘Nestor’ and two hexameters threatening anyone who drinks from it with being seized by Aphrodite.Footnote 64 This may refer to Nestor’s Iliadic cup (11.632–7), closely parallel to Mycenaean items;Footnote 65 but it is hard to imagine two vessels as different as the Pithecusan skyphos and the Mycenaean/Iliadic cup. The scholarship divides over whether the inscription constitutes a genuine curse or a joke,Footnote 66 and on the relationship with the Homeric scene. More convincingly, Nestor’s cup was a traditional motif referred to by both the Iliad and the inscription,Footnote 67 given the Mycenaean parallel, the skyphos’ date (c. 725, rather early for a fixed written version of the Iliad at Pithecusae) and the anomaly of the Iliad passage itself—only Nestor, comparatively old and weak, can lift it.Footnote 68
6. Conclusions
The hypothesis that Nestor’s character originated in an isolated local tradition—whether Pylian or Ionian—in the 400 years after the fall of the Pylian palace, being then integrated into a new, ‘Panhellenic’ Greek epos in the eighth century, is untenable. Though it is certainly conceivable that stories about a local king, surviving in the Pylos area despite the Early Iron Age population decrease, entered epos before the Spartan conquest, Nestor’s rich traditional diction in epos indicates a long development in the role of a wise, aged advisor, not the young, bold warrior of the Iliad’s Pylian exploits. The idea that Nestor developed in an Ionian tradition, bound up with an ancestral connection to founders of Ionian cities, faces linguistic objections over Neleus’ name (and doubts regarding Pylos-(Athens)-Ionia migration).Footnote 69
In any case, theories about strictly local development neglect the fundamental diversity of areas with which Nestor (and his family) is connected; the ‘Pylian’ stories themselves narrate wars and competitions with other Peloponnesian regions. The principle that local content implies locally-developing tradition would result in a vast number of different hypothetical local traditions about Nestor; though it would be possible to explain this by means of multi-stage migration, this approach appears rather arbitrary. Given the introductory comments about Early-Iron-Age interconnectedness, we might instead propose that traditions about Nestor did not develop in an isolated region, but in an environment of wider communication and connection. That is, the epic tradition in which Nestor’s character developed, even if performed mainly around some part of Ionia, Aeolia or elsewhere, was constantly in contact with other regions of the Greek world, absorbing characters and local details, and known to audiences there.
II. IRIS
My second case-study is Iris, the divine messenger of the Iliad. Whilst not (explicitly) mentioned in the Odyssey, she appears occasionally in the Homeric Hymns and Theogony. While the issue with assigning Nestor to a particular local tradition turned out to be the complexity of many different local connections, Iris shows precisely the opposite: a lack of any apparent local associations in the Archaic period. Erbse considered Iris an invention of the Iliad poet; but more recent work on her etymology and formulaic diction demonstrates her diachronic depth in the tradition.Footnote 70
1. Absence from cult
According to Nagy, ‘the Homeric concept of the Olympian gods … incorporates, yet goes beyond, the localised religious traditions of each city-state’.Footnote 71 We might therefore expect ‘Panhellenic’ epos to emphasize common ground between local cults, focussing on deities that are widely worshipped and excluding those who exist only in particular localities. With Iris, however, far from being worshipped everywhere, one is hard-pressed to find traces of her in any local cult whatsoever.
Iris is difficult to identify, sharing as she does most of her attributes with Nike.Footnote 72 Her earliest certain depictions are an early sixth-century altar from Amyclae, a limestone metope from Paestum and a seventh/sixth-century terracotta metope from Thermos, Aetolia.Footnote 73 Early sixth-century Attic vases show Iris leading a divine procession at the wedding of Peleus and Thetis;Footnote 74 towards the end of the century a wider variety of themes are attested: Iris running, at the judgement of Paris, bringing Antilochus to Troy, stopping a chariot, and with Heracles and Geryoneus.Footnote 75 Iris is usually represented in situations that relate to mythological narrative, suggesting that she was primarily associated with poetry. The only evidence for cult is a fragment of the third-century antiquarian Semos of Delos, recording the sacrifice of a certain kind of cake to Iris on the island of Hecate.Footnote 76 It seems unlikely that this tradition—one of many such Delian customs—is a survival of some substantial archaic cult, from which Iris might have been adopted into the epic pantheon.
2. Etymology and diction
Iris’ antiquity is clear from her name. Synchronically functioning as the Greek word for ‘rainbow’,Footnote 77 the initial *w- is revealed by Homeric hiatus, and digamma in early inscriptions.Footnote 78 Traces of *w- in Homeric formulae demonstrate that Iris is no creation of the Iliad poet. *wīr-is is an i-stem derivative of a -ró- noun (like ἄκρις ‘peak’ from ἄκροςFootnote 79 ), the root being either *u̯ieh 1 - ‘wrap, envelope, turn’ or *u̯eih 1 - ‘be eager, strive, have fresh energy’ (both form -ró- derivatives).Footnote 80 At first sight, *u̯ieh 1 - seems a better match, given semantic parallels like Latin arcus, Proto-Germanic *regnabugô deriving the name of the rainbow from its curved shape.Footnote 81 But in Greek we otherwise only find *u̯eih 1 - (e.g. ἵεμαι, ἵρηξ, ἴς, Ἰάνασσα, Ἰάνειρα), which therefore seems more likely.Footnote 82 Moreover, as Massetti has shown, Iris has a parallel figure in Vedic mythology, Venás, whose name may be a -nó- derivative of *u̯eih 1 -, and who shares a number of characteristic features: a messenger of the gods, who flies and is assimilated to a bird of prey, with golden wings.Footnote 83 Given these parallels, these figures are surely cognate, from a common prototype in Proto-Indo-European mythology.
Iris’ antiquity is confirmed by her role and diction in epos. Her appearances in the Iliad often take the form of messenger/divine visit scenes,Footnote 84 with considerable repeated diction. Some is not specific to Iris, but part of broader formulaic languageFootnote 85 —although this still shows her integration into the traditional storytelling of the Iliad. Specific, however, is the repeated phrase by which Zeus orders her to bring a message, Ἶριν δ᾽ ὄτρυνε χρυσόπτερον ἀγγελέουσαν· | “βάσκ᾽ ἴθι Ἶρι ταχεῖα …”Footnote 86 ‘He urged on Iris, the golden-winged, to bear a message: “Swift Iris, go forth …”’. Here χρυσόπτερος parallels Venás’ pakṣé hiraṇyáye ‘golden wing’,Footnote 87 alongside the imperative βάσκε, itself only found in the combination βάσκ᾽ ἴθι, in these four instances and at Il. 2.8 (Zeus sends a dream to Agamemnon) and 24.336 (to Hermes, not long after being spoken to Iris). As Hainsworth suggests, βάσκε is an archaic formation (< *g w n̥-skō, cognate with Sanskrit gácchati), surviving in the fossilized context of Iris’ traditional language.Footnote 88 Iris also has unique formulae: the epithet ποδήνεμος ὠκέα Ἶρις ‘wind-footed swift Iris’ (9x Il., Hom. Hymn Ap. 107)Footnote 89 ; ἀελλόπος ἀγγελέουσα (see n. 85), ἀελλόπος ‘storm-footed’ occurring only in this formula in the Iliad;Footnote 90 and πόδας ὠκέα ‘swift-footed’ (9x Il., Th. 780), although the masculine is shared with Achilles.Footnote 91
3. Beyond the Iliad
Other than the potential indirect reference of Od. 18.6–7, Iris is not used as the gods’ messenger in the Odyssey. This is not typical of early epos in general, which is more consistent with the Iliad. Iris appears at Hom. Hymn Dem. 314–24, Hom. Hymn Ap. 102–4 (bringing Ilithyia to Delos) and in the Cypria, informing Menelaus of Helen’s departure.Footnote 92 She is described in the Theogony as the daughter of Thaumas and Electra (both linked to the sea), the sister of the Harpies Aello and Ocypete; Theogony 268–9, describing their speed on a par with winds and birds, may apply to Iris as much as to her sisters. In Theogony 780–9, Iris is the messenger from Zeus to Styx (ἀγγελίην πωλεῖται, Theog. 781), fetching a jugful of Stygian water for oaths.
Comparison with the winds matches Iris’ traditional epithets (ποδήνεμος, ἀελλόπος) and the simile at Iliad 15.170–2, where Iris is compared to snow or hail driven by the North Wind, and 23.198–212, where Iris visits the winds so that they will kindle Patroclus’ funeral pyre. Perhaps these connections reflect Iris’ nature as personified weather phenomenon. Moreover, Alcaeus fr. 327 V. identifies Eros as the son of Iris and Zephyrus.Footnote 93 Though this may allude to Iris’ visit to the winds in Iliad 23, it is also possible that it reflects a traditional association.Footnote 94
If Iris is well-integrated in the tradition, why is she absent from the Odyssey? On the one hand, Iris is not the exclusive messenger of the gods in the Iliad. Other gods are sent when appropriate: Thetis to Achilles (24.120–42), because she is his mother; Hermes to Priam (24.333–469, 679–94), who can put the guards to sleep at 445. The disguised Hermes’ extensive conversation with Priam resembles Athene visiting Telemachus as Mentes in Odyssey Book 1. Revealing also is the Homeric Hymn to Demeter, where Iris is first sent to Demeter to convince her to return; when she fails, Zeus sends the other gods in turn (325–30). Zeus then sends Hermes to Hades (334–57), presumably because of his role as psychopomp. One could make the case that Hermes is sent to Calypso in Odyssey Book 5 rather than Iris because Ogygia is so remote that it requires him rather than Iris;Footnote 95 but Iris is happily sent to Styx (Theog. 780–9), though far away and in the underworld (775); and Odyssey 5.29 (Ἑρμεία, σὺ γὰρ αὖτε τά τ ̓ ἄλλα περ ἄγγελός ἐσσι) is incongruous if Iris is the default messenger.Footnote 96 Perhaps we have to do more with an idiosyncratic difference between different poems. If so, this is a good example of a difference within the poetic tradition for which there is no evidence that it represents a difference between local traditions.
4. Conclusions
The striking feature about Nestor was his connection to a variety of different locations and regions, challenging the idea of an isolated Pylian local tradition. Iris poses the opposite problem: despite being an ancient, established part of the poetic tradition with Indo-European roots, she has no connections with local cult. Naturally, it is hard to disprove the possibility that she was transmitted in a local tradition; but if so, this has left no detectable trace, neither on her character in epos nor on local cult. Moreover, her earliest certain depictions in art occur with a relatively wide distribution, apparently deriving from poetry, not cult.
So Iris seems, in her characteristics, to have been ‘Panhellenic’—at least, not associated with a particular local context—from the outset. If her non-local quality is an ancient inheritance, is it not plausible to suppose that other ‘Panhellenic’ or non-localized aspects of the epic gods (and the epic world in general) might also be ancient, not eighth-century innovations resulting from merging features of local traditions? At least in this respect, the non-localized world of epos is apparently no recent development.
III. SARPEDON
In the Iliad, Sarpedon leads the Lycians, one of the most prominent Trojan allies. In the Hesiodic Catalogue of Women (frr. 140–1 M–W), he is Minos’ brother, who becomes king of Lycia and fights at Troy;Footnote 97 otherwise he is absent from epos. But does he come from a local tradition—Lycian, or Greek? And if so, what evidence is there for the date of his entrance into epos?
1. Etymology
Sarpedon’s name is linked with relative certainty to south-west Anatolia. In addition to the form /zrppedun…/ found on the Xanthos Inscribed Pillar (TL 44, see below), probably his name in Lycian B, we find a corresponding pattern of Lycian A names in /hrppid-/, with a wider pattern of correspondence between Lycian A /hr-/, Lycian B /zr-/, Carian /š(a)r-/, likely from */sr-/.Footnote 98 A very close parallel is Carian Pēdōn. Since neither Lycian A /hr-/ (corresponding to Αρ-/Ερ-) nor B /zr-/ (corresponding to Σερ-) are attested in Greek transcriptions as Σαρ-, whereas we have examples of Carian /šar-/ transcribed Σαρ- (Σαρυσσωλλος, Σαρυασσις),Footnote 99 it seems likely either that the Greek form is based on an older pre-form with different phonetic realization,Footnote 100 or (more plausibly) directly on Carian.Footnote 101 Either way, the etymology supports an origin in south-west Anatolia.Footnote 102
Admittedly, Sarpedon’s name is not unique, being shared with the Gorgons’ island in the Cypria,Footnote 103 an Atlantic island in Stesichorus,Footnote 104 a Cilician promontory,Footnote 105 a Thracian capeFootnote 106 and a Thracian hero killed by Heracles.Footnote 107 The Thracian Sarpedon might be a later creation or variant based on the Iliadic Sarpedon. The place-names are less clear; they do not obviously come from Caria or Lycia;Footnote 108 perhaps similar-sounding names were assimilated to the more familiar character’s name. But they are relatively marginal, and an unlikely source for Sarpedon’s name. The question then arises: what evidence supports a Lycian (or Carian) tradition about Sarpedon, from which he might have been absorbed into Greek epos?
2. Local Lycian tradition?
It has been proposed that Sarpedon originated in local ‘Lycian’ tradition,Footnote 109 and that the return of his body to Lycia (Iliad 16.666–83) reflects local cult.Footnote 110 Cult is later attested at the ‘Sarpedoneion’, a building in Xanthos described by Appian.Footnote 111 But it is not clear that this existed early. Appian’s Sarpedoneion, if it is building G on the acropolis, dates from 460 without evidence for an earlier shrine.Footnote 112 Sarpedon’s name is found in a long inscription in Lycian B (alongside Greek and Lycian A), the late fifth-/early fourth-centuryFootnote 113 Xanthos Inscribed Pillar (TL 44), which is metricalFootnote 114 and might refer to the Trojan War.Footnote 115 But given our limited understanding of Lycian B, it is unclear whether this reflects an ancient epichoric poetic tradition about Sarpedon, or Greek cultural influence; the accompanying Greek inscription draws on fifth-century Athenian ideology.Footnote 116 A reference to the Trojan War would imply at least partial influence from the Trojan saga, rather than an original, separate tradition about Sarpedon from which he might hypothetically have been transferred to Troy in the eighth century.
If the form of the name is Carian, not Lycian, why is Sarpedon a Lycian in the Iliad? The poet refers specifically to the Lycian river XanthosFootnote 117 (2.877, etc.), clearly distinguishing the Carians (2.867–75). That said, the people the Greeks called Λύκιοι really called themselves the Trm̃mili. The Greek name derives from the Late Bronze Age Lukka peoples, a ‘conglomerate of independent communities’ in western Anatolia with a fractious relationship with the Hittites; it is usually assumed that the Lycians are culturally their descendants.Footnote 118 The Lukka lands, however, may be a wider region, including part of (later) Caria.Footnote 119 So Lukka/Λύκιοι seemingly became restricted to a narrower region over time (cf. the term Aḫḫiyawa/Ἀχαιοί). Perhaps, then, Sarpedon became known to the Greeks from a Carian source before the term Λύκιοι became narrowed to the Trm̃mili.
3. Intermediary Milesian/Rhodian traditions?
According to Strabo and Apollodorus, Sarpedon founded Miletus.Footnote 120 Some scholars therefore posit a local foundation narrative there featuring Sarpedon,Footnote 121 perhaps related to Herodotus’ statement (1.147) that certain Ionian cities adopted Glaucus’ descendants as kings. Further, the killing of Amisodarus’ sons Atymnius and Maris (apparently genuinely Anatolian names, and listed among Sarpedon’s hetairoi) by Antilochus and Thrasymedes, Nestor’s sons (Il. 16.317–29, 419), has been taken to allude to a conflict between Sarpedonic and Neleid phases of Miletus.Footnote 122 Local tradition held that the male Cretan/Carian population had been massacred when Ionians arrived.Footnote 123 On this theory, the Iliad poet borrowed some characters from a local foundation narrative in which Neleids fought Carians over Miletus.Footnote 124
A parallel argument is that Sarpedon’s fight with Tlepolemus in Iliad 5 reflects Rhodian epic tradition. A duel originally in Lycia and narrated by Rhodian poets would be translated to Troy by the Iliad poet.Footnote 125 The evidence for archaic-period conflict between Rhodes and Lycia comes from a vague statement, φασὶ δὲ ἀεὶ Λυκίους Ῥοδίοις ἐχθρεύειν;Footnote 126 the testimony that Cleobulus (sixth-century ruler of Lindos) led a military expedition against Lycia;Footnote 127 and the foundation of Phaselis.Footnote 128 There is some plausibility to the idea that the poet associated Tlepolemus and Sarpedon on the basis of the geographical proximity of Rhodes and Lycia, or contemporary conflict; but hypothesizing a poem about an equally hypothetical war over Phaselis featuring Sarpedon seems overly speculative.Footnote 129
4. Cretan Sarpedon?
In various traditions from [Hes.] fr. 141 M–W onwards,Footnote 130 Sarpedon is Minos’ brother, son of Zeus and Europa, and leads migration from Crete to Lycia. In Strabo and Apollodorus, these events are also linked to Miletus’ foundation. There is a contradiction in parentage with Iliad 6.196–9 (Sarpedon the son of Zeus and Bellerophon’s daughter Laodameia). The Catalogue version is problematic for Sarpedon’s involvement at Troy: he would be a contemporary of Minos, Idomeneus’ grandfather (Il. 13.450–3). Later sources (Diod. Sic. 5.79.3) made the Trojan Sarpedon the grandson of the Cretan Sarpedon, but the solution in Apollodorus (Bibl. 3.1.2)—that Sarpedon lived for three generations—probably goes back to the Catalogue: fr. 141.20 M–W, ]ν̣ μερόπων ἀνθρώπων, resembles the description of Nestor’s three-generation life at Il. 1.250–1, τῷ δ’ ἤδη δύο μὲν γενεαὶ μερόπων ἀνθρώπων | ἐφθίαθ’, and Teiresias at Melampodia fr. 276 M–W, ἑπτά τ’ ἐπὶ ζώειν γενεὰς μερόπων ἀνθρώπων.Footnote 131 Sarpedon’s long life is generally seen as an innovation, reconciling conflicting versions in the tradition,Footnote 132 although the repeated phraseology (and thus the reattribution of characters to different eras) is likely traditional. But this does not tell us whether the Iliad poet knew the Cretan association of Sarpedon.Footnote 133 Aceti argues that the Iliad poet is innovating with Sarpedon’s lineage, since the majority of sources contradict this,Footnote 134 but the other sources are all later, some very considerably.
Stories of Cretans settling in south-west Anatolia are sometimes attributed to Bronze Age links between Minoan Crete and Miletus;Footnote 135 this would mean the survival of a Middle Bronze Age tradition, and Sarpedon’s links with Minos and Rhadamanthys could be the result of MBA borrowing from Anatolian into Minoan poetry.Footnote 136 However, little evidence suggests great antiquity for Sarpedon in Greek tradition. Two Corinthian vases from the early mid sixth century do have ΣΑΡΠΑΔΟΝ, apparently /ā/ for Ionic η.Footnote 137 Lycian /e/ is sometimes transcribed as Greek /a/,Footnote 138 but at face value the Corinthian form (vs. epic η) implies that the Ionic form has been through the *ā > ē sound change. However, if Sarpedon’s name originally had ā, it would be difficult to explain the correspondence with Carian Pēdōn (see above); and Pindar has the ē-vocalism (Pyth. 3.112).
Further, Sarpedon’s traditional diction lacks distinctive formulaic epithets or fossilized linguistic forms.Footnote 139 One proposed exception is ταρχύω ‘bury’, used of his burial at Il. 16.456 (= 674), elsewhere once of a grave marker (7.85). This has been linked with Hittite tarḫ-mi ‘conquer’, ‘be superior’, Luwian Tarhund-, Lycian Trqqñt (both names of gods), on the hypothesis that it referred to an Anatolian ritual granting the deceased ‘heroic or divine’ status, coming into epos with Sarpedon from Lycian tradition.Footnote 140 Without attestations in Anatolian in connection with funerary rites, however, the semantic gap seems too wide; even if accepted, it does not necessarily imply that Sarpedon entered the tradition early. Moreover, the onomastic evidence is all from the first millennium.Footnote 141 As such, Sarpedon is probably not a Bronze Age survival in epos, nor as well integrated into epos as, say, Nestor.
5. Conclusions
From etymological evidence, Sarpedon genuinely originates in south-west Anatolia, although pin-pointing specific details proves challenging. The form of the name suggests a Carian source, but there was certainly a local cult to Sarpedon at Xanthos in Lycia from the fifth century, with a distinctive Lycian form of the name; presumably the figure had cross-regional importance. What little we know about local Xanthian beliefs might suggest that Sarpedon was associated with Troy there as well, perhaps due to later interaction with Greek epos.
On internal evidence, Sarpedon seems a relatively recent entrant to epos and the Trojan War. It has been proposed that this happened via a Milesian or Rhodian tradition. I have argued against a Rhodian Sarpedon, although an awareness of the geographical proximity of Rhodes and Lycia might have led the poet to match Sarpedon and Tlepolemus. The evidence that Sarpedon was regarded as pre-Ionian founder of Miletus is stronger, and might explain the Carian form of the name. But the Sarpedon of the Iliad is clearly Lycian, not Milesian; even if the character has originally entered the mythological tradition from Ionia, we must assume direct Lycian influence at some point.
Of the three characters, Sarpedon is the best candidate for a model whereby ‘Panhellenic’ epos subsumes epichoric traditions around the eighth century. Whether from Lycian, Milesian, or other Ionian or Carian tradition, he seems to have entered epos from south-west Anatolia shortly before the composition of the surviving poems—plausibly, the eighth/ninth century. That said, Sarpedon is somewhat problematic for the term ‘Panhellenism’, because (at least in the Iliad) he is not exactly ‘Hellenic’. The fact that the poets were equally happy to adopt a character who hardly fits the Persian-War-era ideology of Panhellenism suggests that they were more interested in a supra-local, broad cast of characters than articulating a particular identity. Moreover, Sarpedon cannot be pinned down to a single ‘major city-state’. Lycia was probably not a political unity before 430,Footnote 142 nor equivalent to a single Greek city state; and Sarpedon seemingly had regional—not purely local—importance.
IV. CONCLUSION
These three case studies challenge the idea that archaic hexameter draws on local poetic traditions of different city states, subsuming them into a Panhellenic whole. In the case of Nestor, postulating an origin in an isolated local tradition—whether Pylian or Ionian—is problematic: a single character may have a wide range of connections to different parts of the Greek world, and it is methodologically dubious to try to isolate individual local strands and assign them to various different local origins. Iris has a long history in the epic tradition, but no local associations: non-local elements in epos are not all recent developments. Sarpedon provides the strongest case for a relatively recent entry into epos, but appears to have had a regional, not just local, importance, and controverts the ‘Hellenic’ element of ‘Panhellenic’. Plausibly, the epic tradition involved ‘Panhellenic’ characters with a wide range of local links even before the eighth century, alongside completely non-local characters like Iris. The formation of the epic tradition was evidently longer and more complicated than the theory of a predominently eighth-century Panhellenicizing synthesis suggests.
A more comprehensive survey might identify characters which better fit a model of isolated local traditions. But this evidence suffices to encourage a degree of caution, should anyone be tempted to posit the existence of a local poetic tradition from local-seeming content. This is certainly not to deny that different traditions, poetic and non-poetic, did exist in the various different settlements of the Aegean, and some may well have been strongly bound to a particular locale; but it is important to emphasize that these were not isolated entities, but interacting in complex ways. Such complexity fits well, both with the interconnectivity revealed by past decades of archaeological research and the mobility of epic poets implied in the poems themselves.
Open access
