Chapter III - Poetic Memory

Summary

This chapter considers how the language of memory and knowledge indexes tradition. In Homeric epic, characters’ memories coincide with the audience’s recollection of intertextual and intratextual episodes (e.g. Aeneas’s flight from Achilles, Heracles’ labours, Diomedes’ wounding of Ares) and sometimes mark selective retellings of tradition (e.g. Agamemnon on recruiting Odysseus). On occasion, characters’ knowledge even extends proleptically to the future (e.g. Hector on Achilles’ death). Few comparable cases of characters’ mythical recall are visible elsewhere in archaic epic or lyric poetry because of our fragmentary evidence and differences in narratological presentation. But lyric poets also index tradition through the memories of their narrators, evoking both other myths (e.g. Theognis on Odysseus) and their own wider cycles of song (e.g. Sappho). They also appeal directly to the audience’s knowledge (e.g. Pindar on Ajax, Bacchylides on Thebes). From Homer onwards, memory and knowledge proved recurring but varied indices of allusion.

III.1 Introduction

In this chapter, our focus is the indexical potential of memory, occasions where the reminiscences of narrators or characters in the fictional world coincide with those of a poet’s audience. In later Hellenistic and Roman poetry, this frequently involves an alignment between characters’ autobiographical memories and the external audience’s knowledge of the literary tradition, as when Ovid’s Ariadne recalls her Catullan past (Fast. 3.471–6) or Mars quotes his Ennian self (Met. 14.812–15).¹ But it can also extend beyond the purely autobiographical to embrace the recollection of more distant literary passages beyond an individual character’s fictional life: in Apollonius’ Argonautica, Medea’s recollection of the pleasures of life simultaneously recalls Asclepiades’ epigrammatic description of them (μνήσατο μὲν τερπνῶν ὅσ΄ ἐνὶ ζωοῖσι, Ap. Rhod. 3.813 ~ ἐν ζωοῖσι τὰ τερπνά, 2.3 HE = AP 5.85.3). Her memory does not index an earlier literary treatment of her own life, but rather an unrelated text on a similar theme.²

To this indexical potential of memory, we can also add another sphere of personal cognition: knowledge. Just as characters recall events from the literary tradition, so too do they often ‘know’ or ‘recognise’ things that would strike an audience as familiar from the literary past. In Lucan’s De Bello Civili, a frenzied matron prophetically ‘recognises’ the disfigured trunk of Pompey at the very same time that an audience recognises the echo of Priam’s own Pompey-like ‘nameless corpse’ from the Aeneid (agnosco, 1.685–6 ~ Aen. 2.557–8).³ And in Ovid’s Metamorphoses, Ulysses’ complaint that Ajax does not ‘know’ the relief work of Achilles’ shield (neque … novit, Met. 13.291) immediately precedes his near-quotation of the Homeric shield ecphrasis – there is no doubt where Ulysses and Ovid ‘know’ these details from (Met. 13.292–4 ~ Il. 18.483–9).⁴

In comparison to the indexical hearsay of the last chapter, these allusive gestures are dependent not on the external and circulating news of others, but rather on the first-hand, embedded experiences of literary characters. Yet they function in a similar manner, prompting an audience to recall and recognise their own ‘memories’ of the literary tradition. In the sections that follow, I will explore how these allusive tropes are already manifest in our earliest Greek poetry.

Before turning to archaic poetry, however, it is worth acknowledging that later Greek writers often employ the language of memory and knowledge when quoting other works, a practice which demonstrates their strong indexical potential, at least by the classical period. In a fragment of Philippides, a poet of New Comedy, a quotation of Euripides’ Stheneboea is preceded by the instruction to ‘remember Euripides’ (Εὐριπίδου μνήσθητι, fr. 18.2 K–A), while in Aristophanes’ Frogs, Dionysus explicitly claims that he is ‘recollecting’ an iambic verse of Hipponax (ἴαμβον Ἱππώνακτος ἀνεμιμνῃσκόμην, Ran. 661).⁵ The same phenomenon is also visible in prose works: in Plato’s Meno, Socrates precedes his quotation of Theognis by asking his interlocutor whether he ‘knows’ what the poet says (οἶσθ’, Meno 95c9–e2), while elsewhere memories are invoked at points of intratextual back reference, inviting audiences to recall earlier material from the same text (ἀναμνήσθητι, ἀναμνήσω, Symp. 201a2–3 ~ 197b3–9; ἀναμιμνῃσκόμενος, Lysias 1.17 ~ 1.14).⁶ Most significant of all, however, is the famous fragment of Antiphanes’ Poiesis (fr. 189 K–A), which thematises the activation of memory and knowledge in a literary context. The speaker claims that the ‘stories’ of tragedy are so ‘familiar’ to the audience (οἱ λόγοι | ὑπὸ τῶν θεατῶν εἰσιν ἐγνωρισμένοι, 2–3), that a poet need only ‘remind’ them of each tale (ὥσθ’ ὑπομνῆσαι μόνον | δεῖ τὸν ποιητήν, 4–5), and that as soon as someone says ‘Oedipus’, ‘they know all the rest’ (τὰ δ’ ἄλλα πάντ’ ἴσασιν, 6). By the classical period, the discourse of recollection and knowledge was intimately integrated into the practice of literary citation and referencing. In the sections that follow, I will argue that we can trace this discourse even further back in time to the poetry of the archaic period.

III.2 Epic Recall

Memory is central to early Greek poetics, both as a prerequisite for its production and as a primary function of its performance.⁷ Oral poets’ ability to recall, embellish and creatively retell their inherited tradition is heavily reliant on their own powers of memory,⁸ while a key goal of the epic genre itself is to preserve the memory of the heroic exploits of a bygone era, acting as a community’s storehouse for past deeds which articulate shared values and ethics.⁹ In a primarily oral society, where such a past could not easily be recorded, preserved and consulted through writing, epic song was a major vehicle for the transmission of a society’s (ever-changing) heritage, values and identity: a vehicle for the transmission and preservation of cultural memory.

The centrality of memory to early Greek epic is readily apparent from our extant texts, especially in the prominent position they attribute to the Muses as inspirers of epic song, the daughters of ‘Memory’ (Mnemosyne) herself. In the famous invocation at the start of the Catalogue of Ships in Iliad 2, the narrator admits that he could not name all those who came to Troy unless the Muses were to ‘recall’ them for him (μνησαίαθ’, Il. 2.492), while Hesiod’s Theogony begins with a miniature Hymn to the Muses which includes a prominent description of their birth from Mnemosyne (Theog. 53–62), as well as an emphasis on their powers of knowledge (ἴδμεν … ἴδμεν, Theog. 27–8).¹⁰ Crucially, the Muses are a distinctive feature of Greek poetry, with no parallel in Near Eastern traditions, where literary creation and preservation were instead associated with writing.¹¹ Their prominence from Homer onwards highlights the core and unique role of memory in early Greek poetics.

This emphasis on recollection is further reflected in epic’s concern to preserve κλέα ἀνδρῶν, as well as epic characters’ own interest in their future renown and immortality (§ii.2). Heroes aspire to be remembered for all posterity, especially by means of a prominent tomb¹² or by the report of others (Od. 8.241–5). And even poets themselves wish to be ‘remembered’, like the narrator of the Homeric Hymn to Apollo (μνήσασθ’, HhAp. 167). Yet it is especially in the wider corpus of the Homeric Hymns that memory’s close connection with song emerges.¹³ At the start and end of many Hymns, the narrator foregrounds his powers of recollection (μνήσομαι),¹⁴ while elsewhere in the Hymn to Apollo, the Delian maidens are said to ‘recall’ the men and women of old by singing (μνησάμεναι ἀνδρῶν τε παλαιῶν ἠδὲ γυναικῶν, HhAp. 160; cf. μνησάμενοι, 150); song is figured as an act of both recollection and commemoration. In these and many other cases throughout archaic epic, ‘remembering’ comes to stand as a near-synonym for ‘singing’ itself.¹⁵

Set against this emphasis on memory, early Greek poetry also displays a reciprocal concern and almost perverse fascination with its opposite: forgetfulness. Material sites of memory repeatedly fail to preserve an individual’s kleos for long,¹⁶ and Homer’s heroes constantly fight against the overbearing threat of oblivion. In the Iliad, Achilles has a famous choice between an anonymous long life and the renown of a heroic, premature death (Il. 9.410–16), while in the Odyssey, Odysseus’ fame is reliant on his safe nostos (‘homecoming’), which is repeatedly threatened during his adventures. He is repeatedly ‘recalled’ by other characters, almost in an attempt to keep him and his story ‘alive’,¹⁷ but numerous obstacles raise the threat of forgetfulness, including the Lotus-Eaters (Od. 9.94–7), Circe (Od. 10.235–6) and especially the Sirens, whose ability to enchant passers-by mirrors the power of song (Od. 12.39–46).¹⁸ In the Greek world, moreover, Helen’s Egyptian drugs in Sparta threaten obscurity, bringing a ‘forgetfulness of every ill’ (κακῶν ἐπίληθον ἁπάντων, Od. 4.219–30),¹⁹ while even the Muses are agents of oblivion as much as recall. In the Theogony, Mnemosyne is said to have given birth to them specifically as ‘forgetfulness of ills and relief from cares’ (λησμοσύνην τε κακῶν ἄμπαυμά τε μερμηράων, Theog. 55), while a poet who sings ‘quickly forgets his anxieties and does not remember his sorrows at all’ (αἶψ’ ὅ γε δυσφροσυνέων ἐπιλήθεται οὐδέ τι κηδέων | μέμνηται, Theog. 102–3).²⁰ This reflects a key ambivalence surrounding ancient perceptions of the power of song: it could commemorate and memorialise some deeds but also omit others, consigning them to oblivion.

Memory and its opposite, therefore, were of central importance for early Greek poetry. Modern scholars, too, have been no less interested in exploring the power and significance of memory’s various facets in these poems, bolstered by the recent explosion of interest in memory studies in the humanities more generally. Especially productive has been the application of concepts from cognitive psychology to both Homeric epics,²¹ alongside the fruitful examination of the social and cultural features of remembrance.²² Yet more can still be said on the self-reflexive and indexical character of memory in early Greek epic. Already in these texts, as in later Graeco-Roman literature, memory and knowledge play an important indexical role, a means of both gesturing to and incorporating other traditions.²³

In the sections that follow, we shall begin once more with the Iliad and Odyssey, exploring how the language of memory, forgetting and knowledge serves to signpost both inter- and intratextual references within each poem (§iii.2.1–2). After establishing the general contours of this pattern, we shall turn to cases in which characters’ reminiscences appear to involve tendentious and partial misrememberings of tradition (§iii.2.3), as well as those in which characters exhibit an uncanny and proleptic knowledge of future events (§iii.2.4). We shall close by exploring some larger questions, as well as the evidence for indexical memory elsewhere in archaic Greek epic (§iii.2.5).

III.2.1 Intertextual Memories

In both Homeric poems, characters repeatedly recall events from their own past which were also familiar from the larger mythical tradition. Whenever a character remembers or reminds another of an earlier experience, the audience are similarly invited to recall their own knowledge of this episode. Such cases of indexical memory are generally less agonistic than the appeals to hearsay we encountered in the previous chapter, but they nevertheless serve an encyclopaedic and incorporative function: through characters’ reminiscences, the poet gestures to the broader web of tradition within which he situates his own work.

Mortal Memories

On the human plane, such cases of indexical memory point to recent episodes of the Trojan war expedition or the heroes’ own lives. On some occasions, such memories are reported indirectly by the narrator. When Peisistratus reminisces about his brother Antilochus in Odyssey 4, for example, his speech is indexed not only by an appeal to anonymous hearsay (φασί, Od. 4.201: §ii.2.3), but also by the narrator’s introductory emphasis on his act of memory (Od. 4.187–9):

μνήσατο γὰρ κατὰ θυμὸν ἀμύμονος Ἀντιλόχοιο,

τόν ῥ’ Ἠοῦς ἔκτεινε φαεινῆς ἀγλαὸς υἱός.

τοῦ ὅ γ’ ἐπιμνησθεὶς ἔπεα πτερόεντ’ ἀγόρευεν·

He recalled to his mind excellent Antilochus, whom the splendid son of bright Dawn had killed. Remembering him, he spoke winged words.

Peisistratus’ recollection of past events within the fictional world of the narrative precipitates and coincides with the audience’s own recall of a familiar episode from the Trojan war tradition. As we have noted before (§ii.2.3), Antilochus’ death was narrated in the Aethiopis of the Epic Cycle (Aeth. arg. 2c GEF). But the tradition evidently pre-dated it: Memnon’s periphrastic introduction here by the matronymic ‘son of Dawn’ (Ἠοῦς … υἱός) suggests that he was a familiar figure of myth,²⁴ while the traditionality of the whole fabula is also presupposed by Iliadic allusions to it.²⁵ Indeed, when Peisistratus goes on to note that Menelaus surely knew Antilochus (ἴδμεναι, Od. 4.200), the overall message is reinforced: Antilochus was a familiar and memorable figure of myth.²⁶

More often in Homer, however, such instances of indexical memory occur in character speech, especially in two-person dialogues where one individual challenges another’s memory of the past. When Achilles encounters Aeneas in Iliad 20, for example, he asks his adversary whether he remembers the previous time (ἤδη … καὶ ἄλλοτε) he was routed from the foothills of Mount Ida (Il. 20.187–96):

ἤδη μὲν σέ γέ φημι καὶ ἄλλοτε δουρὶ φοβῆσαι.

ἦ οὐ μέμνῃ ὅτε πέρ σε βοῶν ἄπο μοῦνον ἐόντα

σεῦα κατ’ Ἰδαίων ὀρέων ταχέεσσι πόδεσσι

καρπαλίμως; τότε δ’ οὔ τι μετατροπαλίζεο φεύγων.

ἔνθεν δ’ ἐς Λυρνησσὸν ὑπέκφυγες· αὐτὰρ ἐγὼ τὴν

πέρσα μεθορμηθεὶς σὺν Ἀθήνῃ καὶ Διὶ πατρί,

ληϊάδας δὲ γυναῖκας ἐλεύθερον ἦμαρ ἀπούρας

ἦγον· ἀτὰρ σὲ Ζεὺς ἐρρύσατο καὶ θεοὶ ἄλλοι.

ἀλλ’ οὐ νῦν ἐρύεσθαι ὀΐομαι, ὡς ἐνὶ θυμῷ

βάλλεαι·

I claim that I put you to flight with my spear at another time before now too. Don’t you remember when I drove you away from your cattle when you were all alone, sending you hurrying down the hills of Ida with your swift feet? You did not look back at all then as you fled. From there you escaped to Lyrnessus, but I sacked it, rushing in pursuit with the help of Athena and father Zeus, and I led the women away as captives, robbing their day of freedom from them – though Zeus and the other gods saved you. But I do not think they will save you now, as you imagine in your heart.

Achilles invites Aeneas to recall their previous encounter as a parallel for the present, establishing expectations about the outcome of this latest meeting. Besides its paradigmatic force, Achilles’ recollection also invites Homer’s audience to recall their own memory of this episode from the larger epic tradition.²⁷ According to Proclus’ summary, this encounter featured in the Cypria, alongside Achilles’ sacking of Lyrnessus, Pedasus and other surrounding settlements (Cypr. arg. 11c–d GEF).²⁸ And here too, there are good grounds for supposing that this encounter, like much else in the Cypria, pre-existed the Iliad. Achilles’ raids appear early in art²⁹ and were a crucial element in the larger fabula of the Trojan war, as the occasion for Achilles’ acquisition of Briseis as his war prize. Within the Iliad, too, they are a recurring point of reference. The narrator mentions how Achilles had previously captured two sons of Priam, Isus and Antiphus, while they were out herding their sheep (Il. 11.104–12: note ποτ’, 104; πάρος, 111), a prior history which prompted Agamemnon to recognise them (γιγνώσκων, 111). Within Iliad 20 itself, moreover, Aeneas has already offered his own summary of the episode (Il. 20.89–96):

οὐ μὲν γὰρ νῦν πρῶτα ποδώκεος ἄντ’ Ἀχιλῆος

στήσομαι, ἀλλ’ ἤδη με καὶ ἄλλοτε δουρὶ φόβησεν

ἐξ Ἴδης, ὅτε βουσὶν ἐπήλυθεν ἡμετέρῃσι,

πέρσε δὲ Λυρνησσὸν καὶ Πήδασον· αὐτὰρ ἐμὲ Ζεὺς

εἰρύσαθ’, ὅς μοι ἐπῶρσε μένος λαιψηρά τε γοῦνα.

ἦ κε δάμην ὑπὸ χερσὶν Ἀχιλλῆος καὶ Ἀθήνης,

ἥ οἱ πρόσθεν ἰοῦσα τίθει φάος ἠδ’ ἐκέλευεν

ἔγχεϊ χαλκείῳ Λέλεγας καὶ Τρῶας ἐναίρειν.

Not now for the first time shall I stand against swift-footed Achilles, but at another time before now too he put me to flight with his spear from Ida, when he came after our cattle and sacked Lyrnessus and Pedasus. But Zeus saved me, rousing my spirit and swift knees. Otherwise I would have been slain at the hands of Achilles and Athena, who went ahead to protect him and urged him to kill the Leleges and Trojans with his bronze spear.

Despite Achilles’ polemical suggestion that Aeneas may have forgotten the event, the Trojan is all too mindful of it. Indeed, his account overlaps with that of Achilles in many details (underlined above), even down to his speedy flight (λαιψηρά τε γοῦνα, 20.93 ~ ταχέεσσι πόδεσσι | καρπαλίμως, 20.189–90), and it too is indexed in temporal terms (οὐ … νῦν πρῶτα …, ἀλλ’ ἤδη … καὶ ἄλλοτε, 20.89–90). Given the ‘cursory manner’ of Aeneas’ account, Anderson has argued that ‘the Iliadic allusions derive from an earlier tradition which was ultimately codified in the Kypria’.³⁰ He takes this argument no further, but additional support for his case can be found in the verbal echoes between Aeneas’ and Achilles’ narratives, which suggest a consistent and uniform fabula underlying both passages. The Trojan prince is driven to Mount Lyrnessus (Λυρνησσόν, 20.92 ~ 20.191, same sedes), which Achilles sacks (| πέρσε, 20.92 ~ | πέρσα, 20.192), and he is saved only by Zeus (Ζεὺς | εἰρύσαθ’, 20.92–3 ~ Ζεὺς ἐρρύσατο, 20.194). Especially significant, however, is the repeated emphasis on Achilles’ routing of Aeneas with his spear (δουρὶ φόβησεν, 20.90 ~ δουρὶ φοβῆσαι, 20.187, same sedes). These are the only two appearances of this phrase in extant Greek literature before the Imperial period (Quint. Smyrn. 8.151), a fact which suggests that the formula could have been specifically associated with the fabula of this episode. By redeploying the phrase twice here, Homer alludes to an established tradition surrounding the early years of the Trojan war and marks the parallel between the two heroes’ present (νῦν, 195) and previous (τότε, 190) encounters. Indeed, this current confrontation proves to be a close replay – or ‘doublet’ – of the earlier meeting.³¹ Although Achilles hopes that the gods will not save Aeneas this time (195–6), Poseidon ultimately intervenes to ensure that the Trojan hero escapes alive once more (Il. 20.288–339, cf. 20.194).³²

When Achilles asks Aeneas whether he can remember this event, therefore, Homer’s audience are invited to draw on their own knowledge of the larger Trojan war tradition. By having the heroes recall their earlier encounter, Homer effectively cites his model for the present scene: Aeneas and Achilles meet again, as they previously had on Mount Ida. Through the language of memory, Homer gestures to his encyclopaedic control of the whole tradition, replaying an earlier episode with a self-conscious sense of déjà vu.

Such recall of past events can also be activated through the language of knowledge and forgetting. In the Odyssey, Penelope asks Antinous whether he is unfamiliar with a past occasion when Odysseus saved his father Eupeithes, after he had joined Taphian pirates (ἦ οὐκ οἶσθ’ ὅτε, Od. 16.424). Scholars suspect that this episode may have been invented for its immediate context,³³ but even if that were true, it builds on the traditional associations of the Taphians as pirates and Odysseus’ allies, details with which not only Antinous but also Homer’s audience would have been familiar.³⁴ Similarly, in the Odyssean Underworld, Odysseus realises that Ajax has not forgotten the anger he felt because of his defeat in the contest for Achilles’ arms (οὐδὲ θανὼν λήσεσθαι ἐμοὶ χόλου εἵνεκα τευχέων | οὐλομένων, Od. 11.554–5), an event that was a central part of his mythical fabula (§ii.3.1), and again familiar to us from the cyclic tradition (Aeth. arg. 4d; Il. Parv. arg. 1a, fr. 2 GEF).³⁵ In the Iliad, meanwhile, Glaucus opens his account of Bellerophon’s exploits by claiming that ‘many men know his ancestry’ (πολλοὶ δέ μιν ἄνδρες ἴσασιν, 6.151), marking the familiarity of the tale that follows,³⁶ and both Achilles and Patroclus are criticised for forgetting the advice they received from their fathers before departing to Troy (σὺ δὲ λήθεαι, Il. 9.259 = 11.790), nodding to the traditions of pre-war recruitment as attested in the Cypria (arg. 5 GEF) and elsewhere.³⁷ Through the language of forgetting, memory and knowledge, the Homeric poet indexes a range of episodes from the wider Trojan war cycle, marking his control of his mythical repertoire.

Divine Memories

It is especially on the divine plane that we encounter such cases of indexical memory. Gods, too, can recall recent mythical events, as when Zeus opens the Odyssey by remembering the revenge of ‘far-famed’ Orestes (τηλεκλυτός), introducing an analogy that will underlie the whole poem (μνήσατο, ἐπιμνησθείς, Od. 1.29–31).³⁸ But more regularly, the gods look back to a more distant age, reflecting their more enduring powers of memory.³⁹ Such access to the distant past renders them apt models for the poet who similarly recalls remote myths and legends through the patronage of the Muses.

A favourite subject of divine recollection is the Greek hero Heracles, whose exploits are a recurring presence in Homer, Hesiod and archaic Greek poetry more generally.⁴⁰ Indeed, the frequency and consistency of his appearances, alongside the developed formulaic system attached to his name, suggest a well-established tradition surrounding the hero,⁴¹ much of which likely went on to shape or influence the later Heracles epics that we know of.⁴² The gods’ frequent recollections of this former age set the current events at Troy in a broader mythological perspective.

In Iliad 8, Athena is frustrated by Zeus’s support of the Trojans and complains that he no longer remembers her previous support of his son Heracles (Il. 8.362–9):

οὐδέ τι τῶν μέμνηται, ὅ οἱ μάλα πολλάκις υἱὸν

τειρόμενον σώεσκον ὑπ’ Εὐρυσθῆος ἀέθλων.

ἤτοι ὃ μὲν κλαίεσκε πρὸς οὐρανόν, αὐτὰρ ἐμὲ Ζεὺς

τῷ ἐπαλεξήσουσαν ἀπ’ οὐρανόθεν προΐαλλεν.

εἰ γὰρ ἐγὼ τάδε ᾔδε’ ἐνὶ φρεσὶ πευκαλίμῃσιν,

εὖτέ μιν εἰς Ἀΐδαο πυλάρταο προὔπεμψεν

ἐξ Ἐρέβευς ἄξοντα κύνα στυγεροῦ Ἀΐδαο,

οὐκ ἂν ὑπεξέφυγε Στυγὸς ὕδατος αἰπὰ ῥέεθρα.

He has no memory at all of the fact that I very often saved his son when he was worn down beneath Eurystheus’ tasks. Time and again, he would cry aloud to heaven, and Zeus sent me from heaven to help him. If only I’d known all this in my wise mind when Eurystheus sent him to the house of Hades the Gatekeeper to bring the hound of hateful Hades up from Erebus, then he would not have escaped the steep streams of the Stygian water.

Athena recalls how frequently she stood by Heracles’ side: the emphatic adverb combination (μάλα πολλάκις, 362) and the pair of iterative verbs (σώεσκον, 363; κλαίεσκε, 364) combine to render Zeus’s ingratitude all the more alarming. But the emphasis on frequency also highlights how traditional an element this is of Heracles’ fabula. Athena’s patronage of the hero and his labours are attested throughout archaic Greek epic,⁴³ while the specific exploit she recalls here, the theft of the dog Cerberus from the Underworld, was also traditional at an early date (Od. 11.623–6, Hes. Theog. 310–12).⁴⁴ When Athena recalls this episode, she refers to an incident that not only Zeus should remember, but also Homer’s external audience, from frequent (πολλάκις) tellings.

Recollection of the wider contours of this myth, however, complicates the immediate context of Athena’s speech. The goddess does not utter her complaint directly to Zeus, but rather to his wife Hera. An audience spurred to ‘recall’ the Heracles tradition would be all too aware that Hera was Heracles’ persistent enemy throughout his life, a consistent feature of the mythic fabula.⁴⁵ As the exegetical scholia note, Hera kept opposing Athena’s attempts to save him (ἠναντιοῦτο αὐτῇ σῳζούσῃ τὸν Ἡρακλέα, Σ bT Il. 8.362 ex.), which makes Athena’s recollection particularly ill judged (εὐπρεπὴς … οὐκέτι). In the wider context of Book 8, as Hera attempts to recruit Athena in a plot to thwart Zeus’s control of the war, this reminder is – as Kelly remarks – ‘neither tactful nor predictive of a successful alliance’.⁴⁶ It foreshadows the pair’s ensuing failure: as soon as they set out towards Troy, Zeus spots them, sends Iris to intercept their chariot and threatens dire punishment (8.397–483), sticking to the threat that he made at the start of the book (8.1–40). Athena’s recollection thus not only acknowledges the familiarity of the Heracles myth but also invites an audience to supplement their broader knowledge of it, adding a further resonance to the goddesses’ scheming.

Such indexical gestures to established tradition are also apparent in other divine recollections of events from this earlier generation. In the Iliadic theomachy, Poseidon complains that Apollo no longer remembers the woes that the pair endured in their year-long service to Laomedon (οὐδέ νυ τῶν περ | μέμνηαι, Il. 21.441–2), referring to the story of Laomedon’s deceit, which precipitated Heracles’ campaign against Troy.⁴⁷ Earlier in the poem, meanwhile, Zeus awakes after the Δίος Ἀπάτη and accuses Hera of failing to remember their past conflict: he hung her up in the air by her feet and bound her wrists with an unbreakable gold band in punishment for her treatment of Heracles (Il. 15.18–33). Hera’s enmity against the hero is a well-established feature of myth, as we have already noted, but Zeus’s passing reference to Heracles’ visit to Cos (Κόωνδ’ εὖ ναιομένην, 15.28) evokes a whole further episode of that hero’s adventures, in which he almost lost his life against the Meropes, the local inhabitants of the island.⁴⁸ Hera’s hanging, meanwhile, fits into a larger tradition of the succession myth and potential threats to Zeus’s rule, a major narrative thread that underlies the Iliad.⁴⁹ In the first book of the poem, we have already heard of Hephaestus’ punishment for attempting to help his mother in the past (Il. 1.590–4: note ἤδη … καὶ ἄλλοτ’, 590), as well as Achilles’ instruction to his mother Thetis to remind Zeus (μνήσασα, Il. 1.407) of the time when she freed him from the bonds devised by the other Olympians (Il. 1.396–406), a story that he has ‘often heard’ her tell before (πολλάκι … ἄκουσα, 396).⁵⁰ When Zeus frames his criticism of Hera with references to memory (ἦ οὐ μέμνῃ ὅτε, 15.18; τῶν σ’ αὖτις μνήσω, 15.31), Homer thus indexes tradition once more: not only the fabula of Heracles, but also the wider myth of divine discord and past threats to Zeus’s dominion.

There is more at stake in this final reminiscence, however. As with the encounter of Aeneas and Achilles in Iliad 20, we know that Zeus’s addressee Hera is all too mindful of these past events. Hypnos had already reminded her of the Heracles story in the previous book, citing Zeus’s extreme anger on that occasion as a reason to avoid lulling him to sleep again in the present (Il. 14.249–62).⁵¹ Hera’s response is telling: she asks if Hypnos really thinks that Zeus would help the Trojans just as he grew angry for his son (ἦ φῂς ὣς Τρώεσσιν ἀρηξέμεν εὐρύοπα Ζῆν | ὡς Ἡρακλῆος περιχώσατο παῖδος ἑοῖο; 14.265–6). Her rhetorical question implies the answer ‘no’, but as the audience listen on, they are invited to note the similarities between Hera’s present and past tussles with Zeus. By the time we reach Zeus’s recollection in Book 15, we have witnessed a close replay of the Heraclean episode, as Hera tricks Zeus again and he now responds with similar rage.⁵² The Δίος Ἀπάτη is thus framed by two separate accounts of the Heracles narrative which together represent her present deception of Zeus as a replay of her earlier resistance over Heracles. Zeus’s repeated language of memory not only signals an allusion to the Heracles fabula but also introduces it as a narrative doublet for the immediate action.

Characters’ recollections of their past, therefore, coincide with and precipitate the audience’s own recall of the same episodes from the larger mythical and literary tradition. Through such acts of recall, the poet maps out the larger contours of myth against which he situates his poem. In particular, he frequently gestures to earlier moments that act as models or doublets for the present myth, including Achilles and Aeneas’ previous encounter, Orestes’ revenge and Heracles’ sack of Troy. Through such an encyclopaedic vista, these recollections emphasise the interconnected strands of myth.

III.2.2 Intratextual Memories

In all of the foregoing cases, we have been dealing with an inevitable degree of speculation, tracing the contours of potential pre-Homeric traditions from internal and post-Homeric evidence. Many of our examples seem very plausible, but given the state of our evidence, absolute certainty is impossible. Nevertheless, these cases of intertextual ‘poetic memory’ in Homer are supported by instances where memory and knowledge function similarly to index intratextual connections within each poem.

Remembering Diomedes

Most striking of all is another divine recollection, in this case from the Iliadic theomachy. Ares asks Athena whether she remembers the time when she supported Diomedes as he fought against the war-god (Il. 21.394–9):

τίπτ’ αὖτ’, ὦ κυνάμυια, θεοὺς ἔριδι ξυνελαύνεις

θάρσος ἄητον ἔχουσα, μέγας δέ σε θυμὸς ἀνῆκεν;

ἦ οὐ μέμνῃ ὅτε Τυδεΐδην Διομήδε’ ἀνῆκας

οὐτάμεναι, αὐτὴ δὲ πανόψιον ἔγχος ἑλοῦσα

ἰθὺς ἐμεῦ ὦσας, διὰ δὲ χρόα καλὸν ἔδαψας;

τῶ σ’ αὖ νῦν ὀΐω ἀποτεισέμεν ὅσσα ἔοργας.

Why are you driving the gods together again in strife, you dogfly, with your fierce daring, and why has your great heart sent you forth this time? Don’t you remember when you sent Diomedes the son of Tydeus to wound me, and openly grasped his spear yourself, driving it straight at me and tearing my beautiful skin? So now I think you will pay for all you did then.

With a formula that we have repeatedly encountered as an index of intertextual connections beyond both Homeric poems (ἦ οὐ μέμνῃ ὅτε), Ares invites Athena (and the audience) to ‘recall’ an episode from earlier within the very same poem: Diomedes’ aristeia in Iliad 5.⁵³ In that episode, Diomedes had been advised by Athena only to fight Aphrodite among the immortals (Il. 5.124–32), an injunction which he claimed he was still mindful of when later reproached by the same goddess (μέμνημαι, Il. 5.818). Despite his recollection of these instructions, however, both he and Athena soon disregarded them as Diomedes went on to attack Ares, the god of war himself, and wounded him with Athena’s help (Il. 5.855–9):

δεύτερος αὖθ’ ὡρμᾶτο βοὴν ἀγαθὸς Διομήδης

ἔγχεϊ χαλκείῳ· ἐπέρεισε δὲ Παλλὰς Ἀθήνη

νείατον ἐς κενεῶνα, ὅθι ζωννύσκετο μίτρῃ·

τῇ ῥά μιν οὖτα τυχών, διὰ δὲ χρόα καλὸν ἔδαψεν,

ἐκ δὲ δόρυ σπάσεν αὖτις·

Then Diomedes, good at the war-cry, charged at Ares with his bronze spear; and Pallas Athena leaned on the spear, driving it into the bottom of Ares’ belly, where the skirt-piece was belted. There he struck and wounded him, tearing his beautiful skin, and he drew the spear out again.

In Iliad 21, Ares explicitly invites Athena to recall this episode. The recollection is reinforced verbally by the repetition of οὖτα (5.858) in οὐτάμεναι (21.397) as well as the more pointed repetition of the whole phrase διὰ δὲ χρόα καλὸν ἔδαψεν/ἔδαψας (5.858 ~ 21.398), an expression which is found nowhere else in extant Greek literature.⁵⁴ The uniqueness of the phrase suggests that we could even treat it as a direct quotation of the earlier scene, or at least a quotation from a specific and recognisable fabula of Diomedes’ theomachic hybris. After all, the frequency with which later writers refer to the ‘Aristeia of Diomedes’ as an independent and recognisable part of the epic suggests that it would have been a self-standing and familiar episode of tradition.⁵⁵ But in any case, there is a particular irony in Ares’ reminiscence here, which unwittingly foreshadows the outcome of this present clash: as before, Ares will be defeated by Athena’s intervention (21.403–14).⁵⁶

This intratextual example, in which we can actively point to the incident recalled, lends strength to other cases noted above where we no longer have an early epic treatment of the episode in question. Events both beyond and within the poem are ‘recalled’ in the same manner, suggesting the continuum of larger mythological traditions. Specifically ‘Iliadic’ events are treated no differently than those belonging to other parts of the Trojan war tradition. All episodes are conceived as different paths, οἶμαι, within the broader network of song.⁵⁷

Remember, Remember …

This conclusion can be strengthened by numerous other intratextual back references which are similarly flagged through the language of memory and knowledge, tying the threads of the narrative together. In the Iliad, Diomedes’ charioteer Sthenelus does not forget the instructions he had received from Diomedes a short while earlier to steal Aeneas’ horses (οὐδ’ υἱὸς Καπανῆος ἐλήθετο συνθεσιάων, Il. 5.319 ~ Il. 5.259–73), while in the Odyssey, Odysseus does forget Circe’s advice that Scylla cannot be fought or defended against and vainly arms against her (λανθανόμην, Od. 12.226–7 ~ Od. 12.119–20). In Iliad 9, meanwhile, Diomedes remarks that ‘the young and old of the Argives know all this’, that he was earlier rebuked by Agamemnon (ταῦτα δὲ πάντα | ἴσασ’ Ἀργείων ἠμὲν νέοι ἠδὲ γέροντες, Il. 9.35–6), a back reference to ‘Agamemnon’s ill-judged censure’ of Tydeus’ son in Iliad 4 (Il. 4.370–400).⁵⁸ Diomedes marks this intratextual knowledge as familiar to the whole community through the totalising polar expression ‘young and old’ (ἠμὲν νέοι ἠδὲ γέροντες).⁵⁹ It is knowledge shared by everyone, not only Diomedes’ internal audience, but also Homer’s external one. Such intratextual links connect small chains of narrative together, inviting audiences to recall recent episodes in the plot and more clearly follow their development.⁶⁰

At times, this intratextual function of memory even appears to draw self-conscious attention to the structuring of the narrative itself. At the start of Odyssey 5, for example, Athena ‘remembers’ the many woes of Odysseus (μνησαμένη, Od. 5.6) and bemoans how ‘nobody remembers’ him any longer (οὔ τις μέμνηται, Od. 5.11–12), repeating the earlier words of Mentor at Od. 2.233–4. Such a repeated emphasis on the failure to remember Odysseus in the poem’s opening books may self-reflexively draw attention to the narrative delay of the ‘Telemachy’ which dominates Odyssey 1–4, with its unexpected focus on Ithaca and Telemachus, rather than Odysseus.⁶¹ After these opening four books, it is indeed as if the poet and audience have themselves ‘forgotten’ the poem’s alleged protagonist.⁶²

More significantly, however, this indexical function of memory also conveys the sense that events within each Homeric poem are already becoming traditional, joined to the larger map of myth. Just as we saw the language of hearsay attached to the events of the Trojan war (§ii.2.3), so too is the language of recollection. In Iliad 1, Achilles begins his summary of events to his mother by remarking, ‘You know. Why should I tell the tale to you who know all the details?’ (οἶσθα. τίη τοι ταῦτα ἰδυίῃ πάντ’ ἀγορεύω; Il. 1.365) – a question that not only marks Thetis’ privileged divine knowledge, but also self-consciously acknowledges the audience’s familiarity with his coming words; they have already heard the story that he is about to repeat (Il. 1.6–349).⁶³ At points, characters even consider the future recall of their contemporary events, looking ahead to the reception of Homeric song. When Agamemnon tells Achilles that ‘long will the Achaeans, I think, remember the strife between me and you’ (Ἀχαιοὺς | δηρὸν ἐμῆς καὶ σῆς ἔριδος μνήσεσθαι ὀΐω, Il. 19.63–4), he lays implicit claim to the preservation of the Iliad itself, with its opening topic of the quarrel between the two warriors (ἐρίσαντε, Il. 1.6; ἔριδι, 1.8).⁶⁴ Similarly, Hector claims that there should be a ‘memory’ of consuming fire around the Greeks’ ships (μνημοσύνη, Il. 8.181), implicitly pointing to the immortalising power of Homer’s own words to preserve and commemorate this significant turning point in the narrative.⁶⁵ In the Odyssey, meanwhile, Odysseus suspects that ‘these dangers, too, I think, we shall someday remember’ (μνήσεσθαι, Od. 12.212) – a claim that hints at the future poetic fame of his Apologoi, just as the Phaeacians’ repeated requests for Odysseus to ‘remember’ them point to their future preservation in song (Alcinous: μεμνημένος, Od. 8.244; Arete: μεμνημένος, Od. 8.431; Nausicaa: μνήσῃ, Od. 8.462). In sum, poetic memory is not only about nodding to other traditions and poems which the poet subsumes within his work, but also a means for Homer to mark out his own place in this tradition – to ensure that future generations too will remember the events that he narrates, just as he and his characters remember other episodes of the mythical past.

III.2.3 Selective Recall

In many of the cases that we have explored above (§iii.2.1–2), indexical memory gestures to a wider canon of myth, incorporating broader traditions and details that reflect on the present poem. These signposts often introduce allusions that seem less agonistic than many of the instances of indexical hearsay that we have explored before. But indexical memory is not entirely free from combative posturing. We have already noted the competitive aspect in characters’ challenges to their addressees’ memories, revealing an anxiety surrounding the fallibility of individuals’ powers of recall. But beyond this, there are also cases of indexical recall which introduce a selective and partial reshaping of tradition. We have already noted Athena’s omission of Hera’s enmity, but here we shall explore two more complicated cases, one from the Odyssey and one from the Iliad. Appeals to memory authorise departure from tradition, while also inviting audience members to supply what is left untold from their own knowledge.

Recruiting Odysseus

The first passage comes from Odyssey 24, when Agamemnon’s shade addresses the newly deceased Amphimedon’s ghost. After recognising the suitor and inquiring how he died, Agamemnon appeals to their former xenia and asks whether he remembers the time when the Atreidae came to Ithaca to recruit Odysseus for the expedition against Troy, employing the same introductory phrase that we have seen repeatedly before (Od. 24.115–19):

ἦ οὐ μέμνῃ ὅτε κεῖσε κατήλυθον ὑμέτερον δῶ,

ὀτρυνέων Ὀδυσῆα σὺν ἀντιθέῳ Μενελάῳ

Ἴλιον εἰς ἅμ’ ἕπεσθαι ἐϋσσέλμων ἐπὶ νηῶν;

μηνὶ δ’ ἐν οὔλῳ πάντα περήσαμεν εὐρέα πόντον,

σπουδῇ παρπεπιθόντες Ὀδυσσῆα πτολίπορθον.

Don’t you remember when I came there to your house with godlike Menelaus to urge Odysseus to accompany us to Ilium in his well-benched ships? It took us a whole month to complete our journey over the wide sea, since it was only with great difficulty that we won over Odysseus the sacker of cities.

Agamemnon’s question evokes the traditions surrounding the mustering of Greek troops for the Trojan expedition, an episode which Amphimedon does indeed remember (μέμνημαι τάδε πάντα, Od. 24.122).⁶⁶ Like Achilles’ raids in the Troad, these events were also treated in the Cypria (Cypr. arg. 5 GEF) and alluded to in the Iliad, where Achilles’ recruitment by Nestor and Odysseus is twice mentioned (Il. 9.252–9, 11.765–90). Agamemnon’s question here, however, emphasises the specific difficulties involved in recruiting Odysseus, who seems to have shown some reluctance: the whole expedition to win him over took a whole month (μηνὶ δ’ ἐν οὔλῳ, 24.118); Odysseus was only persuaded with difficulty (σπουδῇ, 24.119) and deceit (παρπεπιθόντες, 24.119);⁶⁷ and the Atreidae had to stay at Amphimedon’s house, rather than at Odysseus’ own, suggesting some friction in their relationship (24.115).⁶⁸ This emphasis on Odysseus’ reluctance seems to hint at a specific tradition of his unwillingness to join the Trojan expedition, an episode also familiar to us from the Cypria.⁶⁹ In that poem, according to Proclus’ summary, Odysseus refused to join the mission and even feigned madness to avoid it, only to be tricked by Palamedes into revealing his sanity when the life of his son Telemachus was threatened (Cypr. arg. 5b GEF). The reason for his reluctance was apparently a prophecy by the seer Halitherses, indicating that Odysseus would not return from Troy until the twentieth year (cf. Od. 2.170–6).⁷⁰

The figure of Palamedes is, of course, notably absent from the Odyssey, which could suggest that this tale is simply a post-Homeric invention, and perhaps even an embellished extrapolation from this very passage.⁷¹ However, aspects of Palamedes’ character suggest a figure of considerable antiquity,⁷² and one can easily understand why Homer would have muted his presence in the poem: as another figure of cunning and guile who had outwitted even Odysseus, he would be a rival claimant to the title of πολύμητις ἀνήρ. In addition, any mention of Odysseus’ vengeful and deceitful murder of Palamedes (Cypr. arg. 12b, fr. 27 GEF) would considerably impair our estimation of the poem’s protagonist. Palamedes’ absence is thus, in all likelihood, a pointed case of Homeric exclusion.⁷³ Agamemnon’s memory of the incident, like Homer’s, is selective.

Regardless of Palamedes’ involvement, however, the traditionality of Odysseus’ feigned madness is reinforced by the fact that it reflects a facet of Odysseus’ character that is already well established in Homer: his devotion to his family.⁷⁴ On several occasions in the Iliad, Odysseus describes himself as the ‘father of Telemachus’ (Τηλεμάχοιο πατήρ, Il. 2.260; Τηλεμάχοιο φίλον πατέρα, 4.354), uniquely defining himself in terms of his son, rather than the usual heroic practice of one’s father.⁷⁵ This same concern with family is at the heart of the recruitment episode, in which Odysseus not only tries to stay at home but also abandons his ruse to save his son. Both these Iliadic scenes, moreover, can be seen to evoke the context of Odysseus’ maddened ploughing: in Book 2, Odysseus goes on to claim that any man becomes impatient who is parted from his wife even for a single month (ἕνα μῆνα, Il. 2.292–3), a sentiment which parallels his initial reluctance to go to the war, especially if ‘one month’ was the traditional duration of his delay (μηνὶ … οὔλῳ, Od. 24.118). In Book 4, meanwhile, he has just been rebuked by Agamemnon for not entering the battle more quickly (Il. 4.336–48), just as he shirked from battle on Ithaca. As Scodel remarks, by mentioning his son in this context, the poet again ‘links Telemachus with a question of whether Odysseus is eager to fight’.⁷⁶ Although, as ever, certainty is impossible given our limited evidence, it is likely that the tradition of Odysseus’ reluctance and Palamedes’ resolution of the impasse pre-dated the Odyssey. After all, we know from the Hesiodic Catalogue that Odysseus was not bound to participate in the Trojan war by the oath of Tyndareus, unlike Helen’s former suitors (Hes. fr. 198.2–8, 204.78–84); he thus had more reason to avoid participation than most.

By alluding to the episode through the language of memory, therefore, Agamemnon once more indexes the recollection of another episode from the larger Trojan war tradition. In this case, however, we may also have a case of partial misremembering, and not just because of Palamedes’ omission. As we have seen above (§iii.2), the Homeric epics tread a fine line between the opposite poles of memory and oblivion, and any act of memory is always liable to be partial, gradually eroded by the passage of time. In the case of this episode, it is worth noting that, outside the Odyssean Underworld (here and Od. 11.447–8), Agamemnon is not known to have featured in other early versions of the embassy to Odysseus. According to Proclus (Cypr. arg. 4–5 GEF), the embassy in the Cypria comprised Menelaus, Nestor and Palamedes,⁷⁷ while in Apollodorus’ Epitome, Agamemnon is said to have sent a herald to each king, avoiding the dirty work of negotiation himself (Epit. 3.6). Judging by other Iliadic scenes, such delegation was his usual modus operandi: he sent the heralds Talthybius and Eurybates to take Briseis from Achilles (Il. 1.318–48), dispatched Phoenix, Ajax and Odysseus to speak on his behalf in the embassy to Achilles (Il. 9) and delegated the initial pre-war recruitment of Achilles to Nestor and Odysseus (Il. 9.252–9, 11.765–90).⁷⁸ Later in Apollodorus’ Epitome, meanwhile, it is Menelaus, Odysseus and Talthybius who go to Cyprus to recruit Cinyras, the local king who offers a gift of breastplates to the pointedly ‘absent’ Agamemnon (Ἀγαμέμνονι … οὐ παρόντι, Epit. 3.9; for this gift, cf. Il. 11.20–8: §iv.2.1). In the case of Odysseus’ recruitment too, it is thus likely that Agamemnon did not traditionally play a direct role.⁷⁹ Agamemnon’s ‘recollection’ here appears to rewrite tradition, effacing any memory of Palamedes and substituting Agamemnon in his place.⁸⁰

For an audience versed in tradition, Agamemnon’s indexical appeal to memory would encourage recollection of this suppressed detail. Just as Agenor’s indexical φασί effaces the tradition of Achilles’ immortality in the Iliad (Il. 21.569: §ii.2.4), so too does Agamemnon’s reminiscence conceal Palamedes’ role in a cloud of forgetfulness, subtly acknowledging the Odyssey’s partisan presentation of events. The appeal to memory invites audiences to recall this omitted detail and acknowledge Homer’s more positive presentation of Odysseus as the sole πολύμητις ἀνήρ. Memory, just like hearsay, not only marks allusive references but also signposts particularly contentious points of tradition, inviting audiences to recall other competing versions.

The Greeks at Aulis

A similarly selective treatment of the mythical past is visible in the Iliad. As the Greek army start disbanding in response to Agamemnon’s ‘testing’ speech in Iliad 2, Odysseus rallies them by recalling an event from before the start of the war (Il. 2.299–304):

τλῆτε, φίλοι, καὶ μείνατ’ ἐπὶ χρόνον, ὄφρα δαῶμεν

ἢ ἐτεὸν Κάλχας μαντεύεται, ἦε καὶ οὐκί.

εὖ γὰρ δὴ τόδε ἴδμεν ἐνὶ φρεσίν, ἐστὲ δὲ πάντες

μάρτυροι, οὓς μὴ κῆρες ἔβαν θανάτοιο φέρουσαι·

χθιζά τε καὶ πρωΐζ’, ὅτ’ ἐς Αὐλίδα νῆες Ἀχαιῶν

ἠγερέθοντο κακὰ Πριάμῳ καὶ Τρωσὶ φέρουσαι·

Endure, my friends, and wait a little longer, until we can learn whether Calchas’ prophecy is true or not. We know this well in our minds – you were all witnesses to it, those whom the fates of death have not since carried off. It seems just like yesterday or the day before when the ships of the Achaeans were gathering at Aulis, bringing trouble for Priam and the Trojans.

He goes on to recall an omen that they witnessed while sacrificing to the gods at Aulis: a terrible blood-red-backed snake appeared near the altar and devoured eight sparrow chicks alongside their mother, before disappearing or being turned to stone (Il. 2.305–20).⁸¹ Calchas immediately interpreted this omen to mean that the Greeks would sack Troy in the tenth year of the war, a prophecy that Odysseus recalls now to stop the Achaeans disbanding the war effort on the cusp of victory (Il. 2.321–32). This event appears to have been a well-established feature of the pre-war tradition.⁸² Like many of the episodes we have discussed above (§iii.2.1/3), it was treated in the Cypria (Cypr. arg. 6 GEF). And already in Iliad 1, the importance and traditionality of Calchas’ pre-war prophecies have been suggested by his introduction as the man who guided the Greek ships to Troy with his art of prophecy (Il. 1.71–2) and by Agamemnon’s scathing criticism of his ever-unfavourable prophecies (Il. 1.106–8).⁸³ By introducing his account of Aulis as something which he and his audience have witnessed (μάρτυροι, 2.302) and know (ἴδμεν, 2.301), Odysseus reinforces the sense that this is indeed a familiar and traditional episode,⁸⁴ an impression strengthened by a string of further indexical markers: the temporal phrase χθιζά τε καὶ πρωΐζ’ ὅτε (2.303) marks the event as fresh in the Greeks’ memories (‘it seems just like yesterday or the day before’),⁸⁵ while Calchas goes on to predict that the fame of this omen will never die (κλέος οὔ ποτ’ ὀλεῖται, Il. 2.325), a phrase which not only self-referentially marks the Iliad’s role in preserving that κλέος,⁸⁶ but also the fame and reputation that the tale has already acquired in tradition. Indeed, by recalling events in Aulis, the poet paves the way for the subsequent Catalogue of Ships (Il. 2.494–779), a passage which evokes the initial mustering of the Greek contingent at Aulis.⁸⁷

There is one detail, however, that complicates the simplicity of Odysseus’ appeal to knowledge. According to Proclus’ summary of the Cypria, this snake and sparrow portent took place many years before the Greeks even arrived at Troy, during the army’s first gathering at Aulis. Rather than immediately reaching Troy after this mustering, they mistakenly landed in Mysia, attacked Telephus and his men and returned home after being scattered by a storm (Cypr. arg. 7 GEF).⁸⁸ Proclus does not specify the time frame of this first abortive ‘Teuthranian’ expedition, but according to Apollodorus, it added an extra ten years to the whole expedition: the Greeks set out to Mysia in the second year after Helen’s rape and only gathered again in Aulis eight years later, where they were helped by Telephus’ local knowledge to reach Troy (Apollod. Epit. 3.18–20). Scholars have long debated whether these events are presupposed by the Iliad.⁸⁹ Their details seem to have been well established at an early date: Telephus is mentioned in passing as the father of Eurypylus in the Odyssey (Od. 11.519–20), his birth and flight from the Greeks are narrated in the Hesiodic Catalogue (Hes. fr. 165) and his encounter with Achilles appears to have been treated in the Little Iliad (Il. Parv. fr. 4 GEF). In recent years, moreover, Archilochus’ Telephus elegy (fr. 17a Swift) has added further evidence that the myth was a familiar part of the epic tradition by at least the seventh century.⁹⁰ Yet the Iliad is conspicuously silent on this episode: it makes no direct mention of Telephus, and it is Calchas – not Telephus – who is said to have guided the ships to Troy through his own art of divination (Il. 1.71–2).⁹¹ In the Odyssey, moreover, these extra ten years are incompatible with the poem’s internal time frame, in which Odysseus returns to Ithaca after twenty years, ten spent wandering and ten at Troy. Even so, however, there is one detail in the Iliad that seems to presuppose the Teuthranian expedition: Helen’s complaint that she has now been in Troy for twenty years (ἐεικοστὸν ἔτος, Il. 24.765–6), a total that is difficult to explain without presupposing the additional ten-year delay in Mysia (λέγεται τὸν πόλεμον εἰκοσαετῆ γενέσθαι, Epit. 3.18).⁹² It is only a small hint, but it is enough to suggest that the audience of the Iliad could have been aware of the Teuthranian campaign.⁹³

In that case, we should ask how this larger tradition affects our interpretation of Odysseus’ recall of Calchas’ prophecy in Iliad 2. With the knowledge of hindsight, it seems that Calchas’ calculation only determined how long the Greeks would spend in Troy once they had actually arrived there, but this was not the only possible way of interpreting his words. Like many ancient oracles and prophecies, Calchas’ speech is polysemous. The crucial word is the adverb αὖθι (ὣς ἡμεῖς τοσσαῦτ’ ἔτεα πτολεμίξομεν αὖθι, Il. 2.328). Taken with its locative meaning (‘there’), it indicates that the Greeks will fight for ten years in Troy, and so it is fully compatible with the Teuthranian expedition before that time. However, if we foreground its temporal meaning (‘forthwith/immediately’), the prophecy tells that the Greeks will fight for ten years from the moment of the portent, a time frame that leaves no space for the Teuthranian campaign. Within the immediate context of Iliad 2, however, evoking a prophecy that preceded a failed and lengthy expedition is not especially auspicious. Indeed, as Hunter has remarked, if ‘the audience of the Iliad were aware that this portent was elsewhere connected with an abortive first Trojan expedition, then this can only have increased a sense that Odysseus was manipulating “the facts” for rhetorical effect’.⁹⁴ Odysseus’ evocation of knowledge, like Agamemnon’s of memory, is pliable and selective. He avoids explicit mention of the many years of hardship endured even before they reached Troy, but in evoking the communal knowledge of his Greek audience, he invites Homer’s external audience to recall this other episode, with all its additional baggage.

III.2.4 Proleptic Knowledge

Internal characters’ repeated references to memory, knowledge and forgetting thus had a strong indexical potential in both Homeric poems, triggering an audience’s recall of other episodes from the larger tradition, even those that had been suppressed or pointedly reshaped. Before turning to the phenomenon in the wider epic tradition, however, it is worth dwelling on a distinctive aspect of indexical knowledge: the tendency for Homer’s characters to exhibit knowledge which transcends the expected limits of their immediate circumstances, displaying an uncanny familiarity with events of the mythical future.

A simple example of this phenomenon occurs in Iliad 10. After Diomedes has chosen Odysseus to accompany him on his night mission, the Ithacan hero insists that Diomedes should not say too much about him, since ‘you are saying these words among Argives who know’ (εἰδόσι γάρ τοι ταῦτα μετ’ Ἀργείοις ἀγορεύεις, Il. 10.250). As Maureen Alden has noted, ‘What Odysseus thinks the Argives know on this occasion is that joint action by himself and Diomede[s] is a common theme in the tradition, and that he also has a number of solo night missions to his credit.’⁹⁵ Indeed, shortly before this, Diomedes has asked how he could possibly ‘forget’ Odysseus as his ideal partner (πῶς ἂν … λαθοίμην, Il. 10.243), making the very same point. What Alden does not acknowledge, however, is the fact that most of these collaborations and nocturnal missions are events that take place after the action of the Iliad. Their joint theft of the Palladium (Il. Parv. arg. 4e GEF) and wounding of Polyxena (Cypr. fr. 34 PEG),⁹⁶ as well as Odysseus’ capture of Helenus (Il. Parv. arg. 2a GEF) and disguised expedition in Troy (Il. Parv. arg. 4b–d GEF; Od. 4.240–58) all take place after the death and burial of Hector; only their joint slaying of Palamedes occurs earlier than the events of the Iliad (Paus. 10.31.2 = Cypr. fr. 27 GEF).⁹⁷ Odysseus thus presents the Greeks as having an anachronistic knowledge of his expertise and companionship with Diomedes from previous tellings of the myth. Their knowledge becomes aligned with that of Homer’s audience.

Such proleptic knowledge is a recurring element of both Homeric poems. In the Odyssean Nekyia, the newly deceased Elpenor already ‘knows’ (οἶδα) what lies in store for Odysseus after his Underworld trip – that he will make a return visit to Circe on Aeaea before continuing his homeward voyage (Od. 11.69–70). Such knowledge is strictly anachronistic – and the first that Homer’s audience has heard of this plot detail.⁹⁸ As with the Iliadic Argives’ larger knowledge of Odysseus and Diomedes’ teamwork, so too here, Elpenor’s knowledge derives from the larger tradition, or at least from an atemporal familiarity with the whole of the poem that is still in progress. In the Iliad, meanwhile, both Hector and Agamemnon claim with unerring accuracy that they know full well that Troy will fall, an event that lies not only in their future, but even beyond the scope of their current poem (Il. 4.163–5 = 6.447–9):

εὖ γὰρ ἐγὼ τόδε οἶδα κατὰ φρένα καὶ κατὰ θυμόν·

ἔσσεται ἦμαρ ὅτ’ ἄν ποτ’ ὀλώλῃ Ἴλιος ἱρὴ

καὶ Πρίαμος καὶ λαὸς ἐϋμμελίω Πριάμοιο.

I know this well in my mind and heart: the day will come when sacred Ilios will be destroyed, along with Priam and the people of Priam of the good ash spear.

These repeated verses provide a complementary and contrasting insight into what Troy’s fate means to both the Greeks and the Trojans, Agamemnon’s assertive declaration serving as a foil for Hector’s later pathetic acknowledgement.⁹⁹ But the knowledge they express here again transcends their usual mortal limits. Agamemnon could be referring back to the Aulis prophecy which Odysseus recalled several books earlier, but Hector, as far as we are aware, has not been privy to any such divine message. Moreover, it is striking that after these verses both speakers utter alternative visions of the future which contradict this confessed ‘knowledge’: Agamemnon goes on to fear that Menelaus will die and the expedition be abandoned in ignominy (Il. 4.169–82), while Hector changes his tune to pray that his son Astyanax may rule mightily over Troy and be deemed superior to his father, a source of continuing joy for his mother – an image incompatible with his previous vision of Troy’s ruin (Il. 6.476–81).¹⁰⁰ Their prophetic knowledge almost seems to be a quotation of the mythical tradition, of which they themselves in character remain uncertain.¹⁰¹

Even more striking than this prophecy of Troy’s general doom, however, is Hector’s dying prediction of Achilles’ future death (Il. 22.356–60):

ἦ σ’ εὖ γιγνώσκων προτιόσσομαι, οὐδ’ ἄρ’ ἔμελλον

πείσειν· ἦ γὰρ σοί γε σιδήρεος ἐν φρεσὶ θυμός.

φράζεο νῦν, μή τοί τι θεῶν μήνιμα γένωμαι

ἤματι τῷ ὅτε κέν σε Πάρις καὶ Φοῖβος Ἀπόλλων

ἐσθλὸν ἐόντ’ ὀλέσωσιν ἐνὶ Σκαιῇσι πύλῃσιν.

Yes, I see what will be – I know you well; I wasn’t going to persuade you, since your heart is truly like iron in your breast. But take care now, in case I become a cause of divine wrath against you on the day when Paris and Phoebus Apollo destroy you at the Scaean gates, despite your bravery.

Hector shows an intimate awareness of the details of Achilles’ death, the clearest in the whole poem. Throughout the Iliad, we have received increasingly precise premonitions of Achilles’ fate, especially from his own horse Xanthus (Il. 19.416–17), and Achilles himself has admitted that he ‘knows full well’ that he will die thanks to the insight of his divine mother Thetis (εὖ νυ τὸ οἶδα καὶ αὐτὸς ὅ μοι μόρος ἐνθάδ’ ὀλέσθαι, Il. 19.421).¹⁰² But Hector’s remarks here transcend such a general awareness to specify the precise details of Achilles’ fate: he will die at the hands of Paris and Apollo at the Scaean gates. Scholars often note that the dying were thought capable of supernaturally prophetic speech in antiquity, the same kind of precognition also displayed by Patroclus when he predicts Hector’s impending demise at Achilles’ hand (Il. 16.852–4).¹⁰³ But as Currie has argued, it is striking that Hector here – despite his misreading of the future at other times in the poem – matches the record of traditional mythology precisely.¹⁰⁴ In both the Aethiopis and later artistic depictions, it is both Paris and Apollo who are responsible for the hero’s death (Aeth. arg. 3a GEF), while the Scaean gates are mentioned in the context of Peleus, Thetis and Achilles in a highly fragmentary papyrus ascribed to the Hesiodic Catalogue (Σκαιῇσι πύλῃσι, Hes. fr. 212b.5).¹⁰⁵ Crucially, Hector prefaces this prediction by emphasising his own knowledge (γιγνώσκων), marking his privileged understanding of Achilles’ whole fabula. His knowledge transcends what a character should logically know within the plot.

Besides evoking episodes of the mythical past, therefore, characters’ declarations of knowledge can also have a proleptic edge, looking forward to future events that reach beyond the strict confines of narrative logic. Such indexing of tradition is even more self-conscious than retrospective nods elsewhere, since it involves characters’ familiarity with events of which they should strictly have no awareness. Of course, in a world that believes in prophecy, these moments could perhaps be taken as naturalistic descriptions of plausible human behaviour within the story world. But there are a number of factors that differentiate these episodes from the usual mantic mode elsewhere in Homer. First, they are not spoken by seers or prophets: except for those at death’s door, these words are spoken by ordinary mortals from whom we would not expect such spontaneous, intuitive divination.¹⁰⁶ Second, their predictions are direct, unmediated by a dream or the interpretation of omens.¹⁰⁷ And third, their prolepses are both specific and precise, lacking the normal ambiguity of literary prophecies.¹⁰⁸ Even in antiquity, their proleptic knowledge of future events would have been arresting and unusual. Through his characters’ words, the prophet-like poet looks ahead to future mythical events.¹⁰⁹ Characters’ knowledge can look both forwards and backwards to incorporate the whole story of the Trojan war.

III.2.5 Mapping Epic Memory

As the foregoing examples have demonstrated, poetic memory was already a well-established feature of Homeric poetry. Characters’ recollections and knowledge of other episodes in their fictional world repeatedly map onto the recall of both earlier and later episodes from the epic tradition. Of course, not every mention of ‘memory’ will necessarily have such indexical potential. When characters ‘recall’ general nouns, such as ‘battle’, ‘valour’ and ‘food’, we would be hard pressed to interpret these indexically.¹¹⁰ But in every instance where Homeric characters recall events (of the past or future), often alongside a temporal ὅτε (or ὡς/ὅσα), they appear to index a familiar episode from the larger cycle of epic myth.

So far, we have focused almost entirely on the Iliad and Odyssey as rich sources for such cases of indexical memory. When we turn to the broader corpus of early Greek epic, by contrast, it is striking how few parallels we can find. In Hesiod’s Works and Days, we only encounter repeated exhortations to Perses to ‘remember’ the instructions he has received (μεμνημένος, Op. 298, 422, 616, 623, 641, 711, 728; cf. HhAphr. 283: Aphrodite to Anchises), while in the Theogony, characters only remember a handful of events from within the narrative: the Cyclopes recall Zeus’s favour (ἀπεμνήσαντο, Theog. 503); Zeus recollects Prometheus’ deception (μεμνημένος, Theog. 562); and the Hundred-Handers remember their friendship with Zeus (μνησάμενοι, Theog. 651). The best non-Homeric example occurs at the start of the seventh Homeric Hymn (to Dionysus), which does not launch into its narrative with the usual hymnic relative clause, but rather with an act of memory (Hh. 7.1–4):

ἀμφὶ Διώνυσον Σεμέλης ἐρικυδέος υἱόν

μνήσομαι, ὡς ἐφάνη παρὰ θῖν’ ἁλὸς ἀτρυγέτοιο

ἀκτῇ ἐπὶ προβλῆτι, νεηνίῃ ἀνδρὶ ἐοικώς

πρωθήβῃ·

I shall recall how Dionysus, the son of glorious Semele, appeared on a protruding headland by the shore of the barren sea, looking like a young man in the prime of youth.

As we have already observed, the Homeric Hymns frequently foreground their narrator’s engagement with memory, especially at their opening and close (§iii.2), and scholars frequently suggest that this specific Hymn’s phrasing is a simple variant for the imperatival ἔννεπε/ἔσπετε found in other hymnic introductions.¹¹¹ But it is noteworthy that this foregrounding of memory precipitates an immediate dive into the Hymn’s narrative: such framing suggests that the subsequent story of Dionysus’ capture and revenge was a familiar story. After all, the delocalised and distilled nature of the hymn’s narrative certainly seems to presuppose a fuller pre-existing tradition of Dionysian epiphany and retribution.¹¹²

Of course, the precise nuance of μνήσομαι here could be debated: the verb hovers ambiguously between the poles of ‘making mention of’ something (i.e. introducing it into – and inscribing it within – collective memory) and actively ‘recalling’ it (i.e. drawing it from pre-existing memories). In this case, however, the latter interpretation is encouraged by a close Iliadic parallel which combines the same verb (μνήσομαι) and same conjunction (ὡς) in a context which clearly refers to the recollection of prior information (Il. 9.646–8):¹¹³

ἀλλά μοι οἰδάνεται κραδίη χόλῳ, ὁππότε κείνων

μνήσομαι, ὥς μ’ ἀσύφηλον ἐν Ἀργείοισιν ἔρεξεν

Ἀτρεΐδης, ὡς εἴ τιν’ ἀτίμητον μετανάστην.

But my heart swells with anger whenever I remember this, how the son of Atreus insulted me among the Argives, as if I were some worthless migrant.

In his final dismissive response to the embassy, Achilles cites Agamemnon’s original misstep as the cause of his continuing rage. His recollection intratextually looks back to the first book of the poem. It is thus plausible that the hymnic poet’s own recollection, phrased in similar language, looks out intertextually to prior Dionysiac traditions in a similar fashion, presenting the ensuing account as established and authoritative.

Even if we count this example, however, our extant remains of early Greek epic offer slim pickings when it comes to indexical memory beyond the Iliad and Odyssey. As far as our evidence goes, it seems to be an almost exclusively Homeric phenomenon, far more so than in the case of indexical hearsay. This is a significant finding, and one that could lend support to those scholars who picture Homeric epic as uniquely ‘meta-Cyclic’ or ‘meta-epic’, positioning itself against larger traditions in an extremely self-conscious manner.¹¹⁴ However, it is likely that this apparent Homeric monopoly on indexical memory is largely a result of the narrative form of the Iliad and Odyssey, rather than any unique self-reflexivity. This allusive mode relies above all on the presence of character speech in extended mythical narratives, precisely what we find repeatedly in Homer, but rarely in the rest of extant archaic Greek epic. If other early Greek epics survived in their entirety, our picture would likely be very different. It is well known that Proclus’ summaries of the Epic Cycle downplay the significance of character speech, making it very difficult to see what role it played there.¹¹⁵ Rather than claiming ‘poetic memory’ as something originally or distinctively ‘Homeric’, then, it is better to see it as particularly tied to Homer’s blend of the mimetic and diegetic modes.

Before concluding, however, we should address two potential objections to many of these cases of indexical memory in Homer and archaic epic. The first is the potentially formulaic nature of the language in which they are expressed. Many of the above examples have been introduced by a single recurring phrase, ἦ οὐ μέμνῃ/οἶσθ’ ὅτε (Il. 15.18, 20.188, 21.396; Od. 16.424, 24.115), while the close structural parallels between Od. 1.29–31 and Od. 4.187–9 might similarly suggest formulaic scaffolding.¹¹⁶ If so, the allusive readings above may put too much weight on what were simply traditional modes for introducing narratives and character speech. However, I do not think this is the case. For a start, we should not overplay the formulaic nature of all these examples: with Od. 1.29–31 and Od. 4.187–9, for example, we must stress that these are the only two places in Homer which combine the verbal forms μνήσατο and ἐπιμνησθείς; a fact which should make us hesitate before classing them as formulaic.¹¹⁷ But in any case, this ‘formulaic’ objection relies on an outdated view of Homeric formulae as empty place-fillers, convenient building blocks devoid of meaning. Recent scholarship, by contrast, has highlighted the connotative meaning embedded in recurring formulae based on their repeated appearances (cf. §i.2). In this case, we could plausibly argue that the traditional resonance of ἦ οὐ μέμνῃ ὅτε lies precisely in its evocation of other stories and characters: traditionally, this phrase functions as a longhand ‘cf.’. In short, the formulaic quality of this language bolsters, rather than inhibits, its indexical value.

The second possible objection to this analysis is the fact that most of these Homeric instances of indexical memory are extremely naturalistic. Characters within this fictional world remember events or people from their own past, which is a completely natural human process. In comparison to the Ovidian Ariadne’s overtly metaleptic memories, these Homeric examples are generally less marked, less in-your-face.¹¹⁸ Given this greater subtlety, it might seem an overinterpretation to read these Homeric memories indexically. However, once again, I would challenge this reasoning. First, the instances of proleptic knowledge above (§iii.2.4) involve the citation of mythical traditions which do extend beyond the natural limits of a character’s available knowledge, not all of which can simply be explained away by ancient views on the supernatural insight of the dying. Moreover, it is striking that in every instance where characters recollect other events, these events belong either to earlier moments of the same text or to other traditional moments of the epic tradition. It would be overly sceptical to dismiss this mapping as a complete coincidence and deny its allusive significance.

Indeed, rather than seeing this difference between Homer and Ovid as a reason to dismiss our Homeric interpretations, it would be better to see it as an indicator of this index’s diachronic development. While indexical memory functions smoothly and seamlessly in Homer, it gradually becomes more overt and artificial over time. The embeddedness of the Homeric process certainly fits with an overarching incorporative aesthetic: the Homeric text subsumes all past traditions within itself, a practice which again fits with Scodel’s concept of Homer’s ‘rhetoric of traditionality’ (cf. §ii.2.3).¹¹⁹ By casting the literary and mythical past as ‘memories’ of his characters, the poet maintains the pretence that everything is familiar and traditional – even, as we have seen, when those memories prove pointedly selective.

III.3 Lyric Recall

Archaic lyric poets were no less concerned with memory and the immortalising aspects of poetry than their epic counterparts. They too fostered a close relationship with the Muses: they boast of being the Muses’ attendants, messengers, interpreters, helpers and even sons,¹²⁰ and they often talk of their poetry or their poetic ability as a ‘gift of the Muses’.¹²¹ Like epic poets, they invoke the Muses as a source of inspiration, to grace them with their presence, to begin a new song, to give lovely charm to their poetry and – in epic style – to sing on a certain subject or answer a specific question.¹²² But it is, above all, because of the Muses’ ability to know everything (ἴσθ’ … πάντα, Pind. Pae. 6.54–5) and to bestow metaphorical immortality in song that they are frequently summoned, thanks to their close association with memory.¹²³ More generally, lyric poets are also deeply invested in preserving the memory of whatever they narrate, including places (Ol. 6.92, Pyth. 9.88), gods (Nem. 7.80), laudandi (Nem. 7.14–16; Isth. 8.62) and themselves (Thgn. 100 = 1164d). Just like epic poets, they are embedded in an elaborate system of literary commemoration and preservation.¹²⁴

The anxiety of forgetfulness also underlies much lyric poetry, where song again proves the antidote to eternal oblivion.¹²⁵ In Pindar’s epinicia, oblivion is aligned with silence, darkness and obscurity as a foil to the commemorative ‘light’ of song. In Nemean 7, the poet acknowledges that ‘great feats of strength live in deep darkness if they lack hymns’ (ταὶ μεγάλαι γὰρ ἀλκαί | σκότον πολὺν ὕμνων ἔχοντι δεόμεναι, Nem. 7.12–13), and they are only preserved ‘if, by the grace of shining-crowned Mnemosyne, recompense for labours is found in the famous songs of poetry’ (εἰ Μναμοσύνας ἕκατι λιπαράμπυκος | εὕρηται ἄποινα μόχθων κλυταῖς ἐπέων ἀοιδαῖς, Nem. 7.15–16). It is the shining light of poetic Memory that ensures one’s legacy in the face of gloomy forgetfulness.¹²⁶ Similarly, Sappho remarks that an unknown addressee will lack any remembrance (μναμοσύνα, fr. 55.1) after dying and will wander ‘unseen’ (ἀφανής, fr. 55.3) in the house of Hades because she has ‘no share in the roses of Pieria’, the birthplace of the Muses (οὐ γὰρ πεδέχῃς βρόδων | τὼν ἐκ Πιερίας, fr. 55.2–3); by apparently failing to mention the addressee’s name, Sappho ensures her Muse-less and forgotten fate.¹²⁷ By contrast, the poet is confident that she and another addressee will still be remembered in the future (μνάσασθαί τινα φα<ῖ>μ’ ἔτι κἄτερον ἀμμέων, fr. 147) and that she will not be forgotten even after death (οὐδ’ ἀποθανούσης ἔσται λήθη, Aristid. Or. 28.51 = fr. 193).¹²⁸ Lyric poets were concerned to preserve both their subject matter and their own name from the threat of eternal oblivion.

Despite this prominent concern with memory, however, extant lyric poetry offers few direct parallels for the kinds of indexical memory that we have identified in Homer. As in the larger corpus of archaic Greek epic, there are very few cases in which characters’ reminiscences overlap with the audience’s recall of the literary and mythical past. Some of the examples that follow will thus be more speculative and open to debate than those we have encountered before. But given the well-established presence of indexical memory in Homeric epic, it is worth considering the various ways in which the device may also function in our lyric corpus – an exploration which opens up a number of interesting possibilities. In particular, the different narratological frame of lyric seems to invite other kinds of indexical memory to operate. The stronger presence of the first-person narratorial voice means that the narrator’s own memories, rather than those of internal characters, can serve as an allusive trigger. In addition, lyric poets’ more explicit acknowledgement of their audiences means that they could also appeal directly to their auditors’ knowledge of the literary past – a stark contrast to the indirectness of Homer.

In the following sections, we will consider these various aspects of indexical memory in lyric, beginning with our limited evidence for the overlap of characters’ and audiences’ memories (§iii.3.1). We will then consider the recollections of lyric narrators, which can both blur with events of the mythical past (§iii.3.2) and evoke episodes of a contemporary poetic present (§iii.3.3). And we will close by exploring lyric poets’ explicit evocation of their audiences’ memories and knowledge (§iii.3.4).

III.3.1 Mythical Recall

When we turn to extant lyric poetry’s treatment of myth, we find nothing precisely comparable to the Homeric cases of indexical recollection that we have explored above. There is no clear case of a character recalling an event from the mythical or poetic past. But there are a few glimpses of possible examples in several fragments and testimonia, which are worth exploring for what they tell us about lyric’s engagement with indexical memory.

At times, the lyric narrator notes that characters did or did not remember an aspect of the mythical past, a comment that invites an audience to recall their own knowledge of the myth in question. In Isthmian 8, for example, Pindar claims that the gods ‘remembered’ the pre-eminence of the Aeacids (ἐμέμναντ’) at the time when (ὅτ’) Zeus and Poseidon quarrelled over marrying Thetis (Isth. 8.24–31). This opening emphasis on memory not only acknowledges the traditional excellence of the Aeacid line (including Achilles, Ajax and Peleus),¹²⁹ but also serves as a springboard into the narrator’s own recollection of the marriage of Thetis and Peleus. The combination of a verb of memory with the temporal conjunction ὅτε is almost a Pindaric rebranding of the common epic formula ἦ οὐ μέμνῃ ὅτε; the phrase has been redistributed but retains its allusive function.

Scholars have long debated whether Pindar is here following a familiar tradition or innovating, especially on the points of Zeus and Poseidon’s quarrel over Thetis and Themis’ subsequent prophecy revealing the danger of Thetis’ offspring for its father.¹³⁰ If these were Pindaric inventions, the poet’s appeal to divine ‘memory’ may partly authorise this departure (especially when compounded by the concluding φαντί, Pyth. 8.46a). There is some evidence, however, that these elements are not complete fabrications. As Anne Pippin Burnett notes, Themis’ role might already be suggested by an early sixth-century dinos of Sophilos (LIMC s.v. ‘Peleus’ 211), on which Themis follows immediately after Cheiron in the wedding procession.¹³¹ Regardless of the degree of Pindaric innovation, however, it is significant that this myth is introduced as an act of character memory, precipitating the audience’s own similar recall of the mythic past.¹³²

In other lyric instances, a character reports their own memories or challenges that of another. An extremely scrappy fragment of Stesichorus’ Geryoneis offers a particularly tantalising case (fr. 18):

— ⏕ — ⏕ μ]ι̣μνε παραὶ Δία

       παμ[βασιλῆα ⏑ —.

⏑ ⏑ — γλαυκ]ῶ̣πις Ἀθάνα

⏕ — ⏕ —]ς ποτὶ ὃν κρατερό-

       φρονα πάτρω’ ἱ]πποκέλευθον·

“⏕ — ⏕ —]ς μεμναμένος α[

       ⏑ ⏑ — —]

⏕ — ⏕ Γαρυ]όναν θ[αν]άτου

… remained beside Zeus, king of all; … grey-eyed Athena … to her strong-minded, horse-driving uncle: ‘ … remembering … Geryon … (from/of) death …

From what survives, we seem to have the start of an exchange between Athena and her uncle Poseidon – two gods who were no strangers to indexical memory in the Iliad (§iii.2.1). Athena begins her speech by appealing to Poseidon’s memory (μεμναμένος). The surrounding context is frustratingly lacunose, but the most plausible supplements offer an intriguing situation (fr. 18.6–9, suppl. Page and Barrett):¹³³

ἄγ’ ὑποσχέσιο]ς μεμναμένος ἅ[ν-

      περ ὑπέστας]

μὴ βούλεο Γαρυ]όναν θ[αν]άτου

       [ῥῦσθαι στυγεροῦ].

Come now, remember the promise that you made, and don’t desire to rescue Geryon from hateful death.

These supplements offer a plausible reconstruction, with good epic and lyric parallels,¹³⁴ and they also fit the larger context well: it is natural that Poseidon would be inclined to save his grandson Geryon, just as he saves his sons the Moliones from Nestor in the Iliad (Il. 11.750–2).¹³⁵ Moreover, Poseidon and Athena are traditional rivals, not only in their competition for Athens, but also in their opposing treatments of Odysseus in the Odyssey; this is the latest in a series of ongoing confrontations. In any case, if this reconstruction is along the right lines, Athena here invites Poseidon to recall a promise that he had once made not to interfere in Geryon’s death, a promise that may well have featured earlier in the narrative.¹³⁶ Athena’s appeal to Poseidon’s memory would then be akin to the intratextual recall of the Iliadic theomachy, where Ares challenges Athena’s own memory of Diomedes’ aristeia (Il. 21.394–9: §iii.2.2). This, of course, can remain no more than a tentative possibility, not only because of our dependence on supplements, but also because there is no guarantee that this ‘promise’ would have appeared earlier in the poem. In comparable cases from epic and drama, such ‘reminiscences’ sometimes invoke a promise or oath that has not in fact been mentioned previously.¹³⁷ The ‘memory’ here would then establish a fact as newly traditional, inscribing it into tradition, rather than looking back to anything pre-existing. On current evidence, we cannot determine the truth, but this example at least shows the potential for indexed memories in narrative lyric.

The other most tantalising example of mythical recall in lyric also concerns the Heracles tradition: Archilochus’ treatment of Deianeira’s rape by Nessus and the centaur’s subsequent death at Heracles’ hands. For this poem, we have no direct text at all, only a number of indirect references (frr. 286–9). Among these is the critique of Dio Chrysostom, who claims that some objected to the manner in which Archilochus portrayed the episode (Archil. fr. 286 = Dio Chrys. 60.1):

ἔχεις μοι λῦσαι ταύτην τὴν ἀπορίαν, πότερον δικαίως ἐγκαλοῦσιν οἱ μὲν τῷ Ἀρχιλόχῳ, οἱ δὲ τῷ Σοφοκλεῖ, περὶ τῶν κατὰ τὸν Νέσσον καὶ τὴν Δηιάνειραν, ἢ οὔ; φασὶ γὰρ οἱ μὲν τὸν Ἀρχίλοχον ληρεῖν ποιοῦντα τὴν Δηιάνειραν ἐν τῷ βιάζεσθαι ὑπὸ τοῦ Κενταύρου πρὸς τὸν Ἡρακλέα ῥαψῳδοῦσαν, ἀναμιμνήσκουσαν τῆς τοῦ Ἀχελῴου μνηστείας καὶ τῶν τότε γενομένων, ὥστε πολλὴν σχολὴν εἶναι τῷ Νέσσῳ ὅτι ἐβούλετο πρᾶξαι· οἱ δὲ τὸν Σοφοκλέα πρὸ τοῦ καιροῦ πεποιηκέναι τὴν τοξείαν, διαβαινόντων αὐτῶν ἔτι τὸν ποταμόν.

Can you solve this crux for me, whether or not some are right to criticise Archilochus, and others Sophocles, for their treatment of Nessus and Deianeira? Some say that Archilochus is talking nonsense when he makes Deianeira sing a long speech to Heracles while she is being assaulted by the centaur, reminding him of the wooing of Achelous and the events that happened then – with the result that Nessus had plenty of time to do what he wanted. Others say that Sophocles introduced the shooting of the arrow before the right moment, when they were still crossing the river.

Dio’s anonymous critics considered the length of Deianeira’s appeal to Heracles inappropriate in context, giving her assailant all too much time to have his wicked way with her.¹³⁸ To prompt such critical censure, Deianeira’s speech must have been an account of some length. The participle ῥαψῳδοῦσαν certainly suggests as much, figuring Deianeira as an epic rhapsode, stringing out an extensive recitation.¹³⁹

Most significant for us here, however, is the content of her speech. According to Dio, she ‘reminded’ Heracles of her earlier wooing by Achelous and the events that took place at that time (ἀναμιμνήσκουσαν), recalling a previous occasion on which Heracles had faced another bestial foe to secure Deianeira’s hand in marriage.¹⁴⁰ We know little more about the speech than what Dio gives us, but a Homeric scholion provides the further detail that Archilochus depicted Achelous in a taurine form (Archil. fr. 287). In that case, we might suspect that Deianeira’s report included a key detail known from many later accounts, a detail which is first explicitly attested in Pindar: that Heracles tore off one of Achelous’ horns in the skirmish.¹⁴¹

We do not have any original verses from this poem, and we do not even have direct evidence of its genre or metre. But Ewen Bowie has plausibly argued that Archilochus’ poem was a narrative elegy,¹⁴² and it is tempting to suppose that Dio’s summary reflects, at least in part, the basic language and structure of Archilochus’ original. In that case, given Dio’s emphasis on Deianeira’s ‘reminding’ of Heracles, the captured maiden could have explicitly prompted Heracles to recall the former occasion of the conflict with Achelous, perhaps even introducing it with the formula ἦ οὐ μέμνῃ ὅτε, a phrase that we have seen repeatedly in Homer. The Achelous episode would have provided a natural model for Heracles’ current situation, marking Nessus as a doublet of the river, another rival for Deianeira’s affections.¹⁴³ In addition, it would presage his coming defeat: like Achelous, Nessus is no match for Heracles. Just as the Iliadic Achilles recalled his former encounter with Aeneas, so too here would Deianeira recall a former tussle for her love as a paradigm for the present.

Admittedly, this is a speculative case, dependent on the language of Dio’s summary, but it is the closest we come to the Homeric usage of this device in lyric. It may be significant, then, that this potential instance occurs in a strongly epicising context. As Bowie notes, ‘On the scant evidence we have, this is a poem in which elegy handled material usually treated in hexameter epic, and did so in the same way as hexameter epic.’¹⁴⁴ Indexical memory within character speech may have thus carried a distinctively epic resonance, a resonance which would have been all the stronger if Stesichorus’ Geryoneis did indeed contain a comparable example. Scholars have long remarked on the epicising nature of Stesichorean lyric, and an epicising narrative is more likely to exhibit extended character speech, the prime host for indexical recall.¹⁴⁵ Indeed, if we had a complete text of other Stesichorean poems, such as the Oresteia, Nostoi or Games for Pelias, we might well expect to find further cases.

There are thus only very limited hints of indexed character memories in lyric, whether in the voice of the narrator or characters. As with the larger corpus of archaic Greek epic, the major reason for this must be our general dearth of extended mythical narrative in lyric – a dearth which results in part from the fragmentary state of our evidence, forcing us to rely on scrappy fragments and second-hand testimonia.¹⁴⁶ But it also reflects a larger compositional strategy of the genre: Greek lyric poets frequently introduce myths in passing and in summary form, as paradigms for their present. In so doing, they rarely give direct voice to the characters of the mythical past – and even when they do, rarely at any length.¹⁴⁷ With this dominant approach to myth, it is unsurprising to find fewer cases of indexical memory in extant lyric. After all, every possible case we have explored concerns a fuller mythical narrative, each of which has a distinctively epic hue. We have already noted the epicising nature of Stesichorus’ lyric and of Archilochus’ Heracles narrative, while Pindar’s account of the gods and Thetis concerns the origins of the whole Trojan war, a key epic theme. Indexical memory may thus have had not only a close association with narrative, but also more specifically with epic traditions.

III.3.2 Personal and Mythical Memory

Besides these cases of internal characters’ memories within poems, we can also identify instances where a lyric narrator foregrounds their own memory at a moment of allusion to mythical precedents and prototypes. Here too, the myths recalled have a distinctively epic tinge. In Nemean 9, Pindar uses an act of recall (μνασθείς, v. 10) as a springboard into his account of Adrastus, Amphiaraus and the expedition against Thebes, evoking Theban myth (Nem. 9.10–27). In another unplaced fragment, he bids Apollo ‘remember’ that Heracles set up an altar to him and Zeus on Paros (μνάσθηθ’ ὅτι, fr. 140a.62–8), recalling the hero’s sojourn on the island during his quest for Hippolyte’s belt before his initial expedition to Troy, a tale that likely dates at least to the time of Archilochus.¹⁴⁸ In a poem of the Theognidea, meanwhile, the speaker’s personal memory precipitates a summary account of archetypally epic adventures (Thgn. 1123–8):

μή με κακῶν μίμνησκε· πέπονθά τοι οἷά τ’ Ὀδυσσεύς,

    ὅς τ’ Ἀΐδεω μέγα δῶμ’ ἤλυθεν ἐξαναδύς.

ὃς δὴ καὶ μνηστῆρας ἀνείλετο νηλέι θυμῷ

    Πηνελόπης εὔφρων κουριδίης ἀλόχου,

ἥ μιν δήθ’ ὑπέμεινε φίλῳ παρὰ παιδὶ μένουσα,

   ὄφρα τε γῆς ἐπέβη †δειμαλέους γε μυχούς†

Don’t remind me of my misfortunes: I have suffered the kinds of things that Odysseus did, he who returned after coming up from the mighty house of Hades. With a pitiless spirit, he gladly slaughtered the suitors of Penelope, his wedded wife, who waited for him for a long time, staying by the side of her dear son, until he set foot on his land …¹⁴⁹

The speaker’s wish not to be ‘reminded’ of his ills segues into the recall of a mythical figure who has endured such suffering: the epic Odysseus, an archetypal endurer (πέπονθα, 1123 ~ πάθεν, Od. 1.4). After launching into the exemplum with an act of recall, Theognis focuses on two major episodes of the hero’s fabula: the descent to the Underworld and his slaughter of the suitors.¹⁵⁰ Such an Odyssean analogy fits into the larger narratorial posturing of the Theognidea,¹⁵¹ or – if Bowie is right to ascribe the poem to Archilochus – the iambicist’s similar Odyssean persona.¹⁵² Most crucial for my current purpose, however, is how this mythical exemplum is once again introduced with the language of memory: the speaker’s recall of his own ills prompts the recollection of an epic exemplar of such suffering. Personal memory transitions to mythical memory.

A similar blurring of personal and mythical recall occurs in Tyrtaeus’ elegy on ἀρετή, in which the poet begins with a catalogue of mythological exempla introduced by another verb of memory (fr. 12.1–9):¹⁵³

οὔτ’ ἂν μνησαίμην οὔτ’ ἐν λόγῳ ἄνδρα τιθείην

    οὔτε ποδῶν ἀρετῆς οὔτε παλαιμοσύνης,

οὐδ’ εἰ Κυκλώπων μὲν ἔχοι μέγεθός τε βίην τε,

    νικῴη δὲ θέων Θρηΐκιον Βορέην,

οὐδ’ εἰ Τιθωνοῖο φυὴν χαριέστερος εἴη,

    πλουτοίη δὲ Μίδεω καὶ Κινύρεω μάλιον,

οὐδ’ εἰ Τανταλίδεω Πέλοπος βασιλεύτερος εἴη,

    γλῶσσαν δ’ Ἀδρήστου μειλιχόγηρυν ἔχοι,

οὐδ’ εἰ πᾶσαν ἔχοι δόξαν πλὴν θούριδος ἀλκῆς·

I would not recall a man nor include him in my poetry¹⁵⁴ for his prowess in running or wrestling, not even if he had the size and strength of the Cyclopes or could outrun Thracian Boreas, nor if he were more handsome than Tithonus in form or richer than Midas and Cinyras, nor if he were more royal than Pelops, son of Tantalus, or had Adrastus’ smooth persuasive tongue, nor if he had a reputation for everything except for furious valour.

In this opening priamel, the poet exalts ‘furious valour’ (θοῦρις ἀλκή) as the pinnacle of excellence (ἀρετή), dismissing other candidates for the title (extraordinary strength, speed, beauty, wealth, royalty and eloquence), which are each represented by a famous mythical hero. Tyrtaeus insists that he would ‘not recall nor include in my poetry’ a man who even outstripped these mythical forebears, preferring instead the man who is good in battle, an opposition reinforced by verbal repetition (ἄνδρα, fr. 12.1 ~ ἀνὴρ ἀγαθὸς … ἐν πολέμῳ, fr. 12.10 = fr. 12.20, ἀνὴρ … πολέμου, fr. 12.43–4). As H. James Shey has highlighted, however, these exempla are carefully selected to emphasise the dangers of the other traits: ‘In every instance, the mythological characters of the priamel possess aretaí which cause harm to themselves or others, or which are unable to save them from harm, unhappiness, or bad reputation’: Polyphemus’ brawn was outwitted by Odysseus’ brains, Tithonus’ beauty eventually withered into an extreme old age and so on.¹⁵⁵ The larger tradition lying behind each name implicitly indicates why Tyrtaeus most highly values θοῦρις ἀλκή.

To build on Shey’s point, we could note how the opening emphasis on memory and λόγος encourages an audience to recall these wider traditions and to fill in the rest of each story. As a common tactic of praeteritio, the speaker invites his audience to recall what he claims he will leave unspoken. Of course, if we wanted to, we could easily find a mythical candidate who equally exemplifies the dangers of θοῦρις ἀλκή: Telamonian Ajax exhibits this very trait in Homer (e.g. Il. 11.566), and as we have already seen, he too comes to an ignominious end (§ii.3.1). Tyrtaeus, however, avoids pointing us in that direction and rather encourages us to recall the fabulae surrounding the characters he does name. In this poem, memory has shifted from a character’s embedded speech to the narrator’s own voice. His power of memory controls which myths are recalled or not.

As Ernst-Richard Schwinge has emphasised, however, this priamel is not purely ethical, for it also has a larger poetic and generic significance. Tyrtaeus is not just dismissing specific myths associated with other potential ἀρεταί, but also a collection of myths with a distinctively epic timbre.¹⁵⁶ Most of the heroes he cites are familiar from the epic tradition; the values they represent are valorised in epic;¹⁵⁷ and the very language in which they are expressed also draws heavily on the epic tradition: verse-ends, in particular, exhibit a whole host of familiar epic idioms,¹⁵⁸ while the phrase Ἀδρήστου μειλιχόγηρυν (fr. 12.8) may also draw on an epic formula associated with lost Theban epic.¹⁵⁹ The poet’s opening appeal to his own memory thus triggers the recall of a host of epic traditions as a foil for his elegiac poem, with its new attitude towards ἀρετή. In what follows, he articulates an alternative poetics distanced from epic. By beginning with his own act of memory (μνησαίμην), he even usurps the traditional role of the epic Muses (μνησαίαθ’, Il. 2.492). The poet’s memory evokes and appropriates the epic tradition – in many ways, a foreshadowing of the Roman recusatio.

Indexical memory in lyric, therefore, was not restricted to internal characters’ recollections of their fictional autobiographies, but also extended to the memory of poetic narrators, especially in elegy. In both of the foregoing cases, however, it is worth emphasising again that the myths recalled have a distinctively epic resonance: Theognis’ Odysseus and Tyrtaeus’ catalogue of epic figures. Once more, indexical memory seems particularly associated with the epic past; it is as if lyric poets were specifically acknowledging the epic heritage of this indexical device.

III.3.3 Memories of the Moment

Another major distinction between lyric and epic poetry also helps to explain lyric poets’ apparently limited use of indexical memory, namely their far greater concern for their immediate present. Lyric poets frequently picture themselves and their audiences in their own contemporary world, focusing more on personalised reminiscences of specific occasions from the recent – not mythical – past. When discussing such contemporary affairs, the language of memory will inevitably have a different valence to that found in continuous mythical narratives.

The Lesbian poet Sappho, for example, shows an emphatic concern with the memories of (what she depicts as) her personal, lived experience.¹⁶⁰ In one fragment, the narrator addresses a departing woman and bids her ‘remember me, for you know how we looked after you’ (μέμναισ’, οἶσθα γὰρ ὤς <σ>ε πεδήπομεν, fr. 94.8) and goes on to ‘remind’ her (ὄμναισαι, fr. 94.10) of all that they experienced with a catalogue of past loves and festivities (fr. 94.10–29).¹⁶¹ In another, a woman who has departed to Lydia ‘remembers gentle Atthis with longing’ (ἀγάνας ἐπι-|μνάσθεισ’ Ἄτθιδος ἰμέρῳ, fr. 96.15–16), while in the famous priamel on τὸ κάλλιστον, the speaker’s description of Helen ‘reminds’ her of another absent female friend, Anactoria (]με νῦν Ἀνακτορί[ας ὀ]νέμναι-|[σ’ οὐ] παρεοίσας, fr. 16.15–16). These and other fragments evoke a network of fond female farewells, in which memory played a key role in preserving social bonds, apparently a far cry from the functioning of poetic memory in the heroic world of archaic epic.

The same social and contextual aspects of memory are also active in the work of many other lyric poets. In a ‘ship-of-state’ poem by Sappho’s Lesbian contemporary Alcaeus, the poet encourages his addressees to ‘remember’ some previous object or event (μνάσθητ̣ε τὼν πάροιθε μ̣[, fr. 6.11), recalling the turbulent and stasiotic life of his hetaireia on Lesbos.¹⁶² Archilochus bids his friend Glaucus to ‘remember’ the land of Thasos (γῆς ἐπιμνήσαιο τ[ῆσδε], fr. 96.3); Alcman hopes to preserve the ‘memory of those present’ (ἔστι παρέντων μνᾶστιν † ἐπιθέσθαι †, fr. 118); and Pindar claims that Hieron will be ‘reminded’ of the battles in which he previously stood steadfast (ἀμνάσειεν, Pyth. 1.47). Lyric poets’ frequent focus on the present and recent past differs strikingly from epic poetry’s immersion in the distant world of myth.

Even here, however, it is possible that these emphatic appeals to memory may have often served an indexical role, recalling recent poetry and songs on contemporary events. Sappho, in particular, is a likely candidate for such poetic self-reference. As we have already seen, she is insistent elsewhere that she and her group will be remembered in the future, unlike the anonymous addressee of fr. 55, a claim which asserts the commemorative power of her poetry (§iii.3). And indeed, the women ‘recalled’ in frr. 16 and 96 do seem to have been regular fixtures in her larger poetic corpus. Atthis features repeatedly in other extant fragments (frr. 8, 49, 90(10^A).15, 130.3–4) and appears alongside Anactoria in later lists of Sappho’s companions (test. 219, 253, 263 = Ov. Her. 15.17–19).¹⁶³ Whatever precise relationship Sappho had with these women, they were evidently a recurring feature of her poetry.¹⁶⁴ As Sappho recalls these absent friends and her former experiences with them, we may thus be invited to recall their presence in her other songs.

The language of these poems certainly encourages us to pursue such cross references. In fr. 94, for example, the events which Sappho recalls resonate richly against her wider extant corpus, with numerous echoes of language and theme (fr. 94.7–29):

χαίροισ’ ἔρχεο κἄμεθεν

⸏μέμναισ’, οἶσθα γὰρ ὤς <σ>ε πεδήπομεν·

—

αἰ δὲ μή, ἀλλά σ’ ἔγω θέλω

ὄμναισαι [… (.)] . [. . (.)] . ε̣αι

10⸏ὀ̣σ̣[… … … .] καὶ κάλ’ ἐπάσχομεν.

—

πό̣[λλοις γὰρ στεφάν]οις ἴων

καὶ βρ[όδων …]κ̣ίων τ’ ὔμοι

⸏κα . . [… … .] πὰρ ἔμοι περεθήκαο

—

15καὶ πό̣⸤λλαις ὐπα⸥θύμιδας

πλέκ⸤ταις ἀμφ’ ἀ⸥πάλαι δέραι

⸏ἀνθέων ἐ̣[… …] πεποημμέναις

<—>

καὶ π … . . [       ]. μύρωι

βρενθείωι . [            ]ρ̣υ[. .]ν

20⸏ἐξαλ<ε>ίψαο κα̣[ὶ ⸤βασ⸥]ι̣ληίωι

<—>καὶ στρώμν[αν ἐ]πὶ μολθάκαν

ἀπάλαν παρ̣[    ]ο̣ν̣ων

⸏ἐξίης πόθο̣[ν      ] . νίδων,

<—>

κωὔτε τις[     οὔ]τ̣ε̣ τι

25ἶρον οὐδ’ ὐ[      ]

⸏ἔπλετ’ ὄππ̣[οθεν ἄμ]μες ἀπέσκομεν,

<—>

οὐκ ἄλσος . [           ] . ρος

                  ]ψοφος

⸏                  ] … οιδιαι

Go, farewell, and remember me, for you know how we looked after you.¹⁶⁵ But if you don’t, I want to remind you … and the good times we enjoyed. For you put on many wreaths of violets and roses and … together by my side; and … many plaited garlands made from flowers around your tender neck; and … with floral, regal perfume you anointed yourself … and on soft beds … tender … you satisfied your longing … There was no … nor shrine from which we were absent, no grove … sound …

The reminiscence moves in vivid snapshots, progressing at first through a scene of increasing intimacy (from head, to neck, to body, to bed), before expanding back out to more communal spaces and activities.¹⁶⁶ At one level, this sequence of memories provides a ‘naturalistic’ recollection of a past experience, but for all these different details we can identify a range of parallel moments elsewhere in Sappho’s poetry. Garlands are a recurring feature of many fragments, as are flowers – and especially roses.¹⁶⁷ Particularly suggestive are the connections with fr. 81, in which Sappho tells Dica to put lovely flowered garlands (στεφάνοις … ἐράτοις, 81.4) around her locks with her ‘tender’ hands (ἀπάλαισι χέρσιν, 81.5), paralleling the ‘tender’ garlanded neck in fr. 94 ([ἀ]πάλαι δέραι, v. 16). More generally, the adjective ἀπαλός recurs often in Sappho’s poetry, especially of her companions. Besides its appearances in fr. 94 (vv. 16, 22) and fr. 81, it describes Gyrinno (fr. 82a), a tender girl picking flowers (fr. 122), and a tender companion on whose bosom someone might sleep (fr. 126).¹⁶⁸ Indeed, if the πόθος which is satisfied in fr. 94.23 refers to a ‘longing’ for sleep, as some scholars have suggested,¹⁶⁹ fr. 126 would provide a particularly close parallel for Sappho’s recollection here. But even if fr. 94 conceals a reference to erotic πόθος (as is more likely),¹⁷⁰ this too finds numerous parallels elsewhere in Sappho’s corpus (frr. 22.11, 36, 48.2, 102.2). Finally, the transition to the shrine (ἶρον, v. 25) and grove (ἄλσος, v. 27) also maps onto other aspects of Sappho’s poetry, especially fr. 2’s ecphrasis of the ‘holy temple’ (ναῦον | ἄγνον, 2.1–2) and ‘grove’ (ἄλσος, 2.2).¹⁷¹ The language and details of Sappho’s reminiscences reverberate repeatedly against her wider poetic corpus, evoking other past songs and performances.

Sappho’s poetry in general is well known for its repetitive nature; throughout her corpus, she repeatedly returns to the same images, vocabulary and motifs.¹⁷² One effect of such repetitions is the creation of a consistent speaking voice, conjuring a sense of ‘Sappho’ as a distinctive and recognisable personality.¹⁷³ But the degree of repetitions and mappings that we have traced in fr. 94 do more than simply establish such an authorial persona: they also seem to offer a summary and distillation of many of the key themes and motifs of her poetry. Of course, our perception could be skewed by the vagaries of transmission. This poem is itself rather fragmentary, and very little now remains from the nine books of Sappho’s poetry that once comprised her Alexandrian edition.¹⁷⁴ But at least as far as we can judge from what survives of this poem and her other extant fragments, this recollection provides almost a ‘table of contents’ for many of Sappho’s wider literary concerns. We should not suppose that this recollection looks back to another specific Sapphic poem which treated the same occasion(s); indeed, such stale repetition would be unparalleled within her larger corpus. Rather, these memories reflect a composite of experiences from other poems, evoking a familiar but disjointed Sapphic world.

Such a strategy of self-citation would fit within a wider phenomenon of Sappho’s poetry which has attracted recent critical interest: her deployment of ‘song cycles’, sequences of interrelated but discontinuous poems on the same topic.¹⁷⁵ With the publication of Sappho’s (unprovenanced) Brothers Poem, scholars have focused particularly on a family cycle centred on the actions of her brother Charaxus, where – as Anastasia-Erasmia Peponi states – individual Sapphic poems serve as ‘snapshots’ or ‘vignettes’, ‘distinct and self-standing facets of a narrative that was never explicitly organized as such’.¹⁷⁶ But we can equally identify traces of other cycles in Sappho’s corpus, including one which appears to have charted various stages of her relationship with Atthis, from loving intimacy (fr. 96, cf. fr. 49a) to bitter separation (fr. 130.3–4).¹⁷⁷ Within such sequences of songs, Sappho’s recollections and self-citations would gain added point, highlighting the larger connections between her poems. Indeed, positioning fr. 94 as a whole against her wider corpus adds even further to our interpretation of it. Sappho’s female interlocutor stresses that she is now leaving ‘against her will’ (Ψάπφ’, ἦ μάν σ’ ἀέκοισ’ ἀπυλιμπάνω, 94.5) – a claim which reverses Aphrodite’s promise in fr. 1 that her similarly anonymous beloved would soon love her ‘even against her will’ (κωὐκ ἐθέλοισα, 1.24). This verbal parallel reinforces a key theme of fr. 1, the cyclicality of desire (cf. fr. 1.21–4): grudging union gives way to grudging departure. As a whole, therefore, fr. 94 invites us to trace links with Sappho’s broader poetic corpus. The emphatically repeated appeal to memory (μέμναισ’, οἶσθα, v. 8; ὄμναισαι, v. 10) reinforces this invitation, spurring Sappho’s audiences to recall her cycles of other related poems.

Such self-citation is also likely in Alcaeus’ appeal to memory. His extant poems foreground their future reception less insistently than Sappho’s,¹⁷⁸ but at various points he acknowledges their enduring appeal, as when he claims that the weapons which he has just described ‘cannot be forgotten’ (τῶν οὐκ ἔστι λάθεσθ’, fr. 140.14) – a remark that ‘figures the poetic memorability of his own description’.¹⁷⁹ It is thus very possible that his recollection in fr. 6 similarly indexes his wider poetic corpus, although the precise reference in this case is obscured by the papyrus’ fragmentary state, leaving the crucial object of memory concealed: μνάσθητ̣ε τὼν πάροιθε μ̣[ (‘remember the previous …’, fr. 6.11). Yet even so, all plausible supplements carry a possible indexical force. Hunt’s μ̣[υθῶν] (‘previous words’) would be particularly self-referential, gesturing to Alcaeus’ previous poetic speech, while the alternative μ̣[όχθων] (‘previous toils’) would evoke the hardships which Alcaeus complains of and prays to escape elsewhere (cf. ⸤μ̣ό̣⸥χ̣θ̣ων, fr. 129.11).¹⁸⁰ In any case, a connection with previous events – and their poetic articulation – is reinforced by the poet’s opening remark that a wave comes upon the ship ‘again’ in the manner of a ‘previous one’ (⸤τόδ’ αὖ⸥τε κῦμα τὼ π⸤ρ⸥οτέρ̣⸤ω † νέμω † | στείχει⸥, fr. 6.1–2). The poet explicitly draws a connection with previous suffering at sea, perhaps evoking a larger cycle of ship-of-state songs, akin to Sappho’s Charaxus or Atthis cycles. By explicitly recalling past events later in the fragment, Alcaeus invites his audience to look back to other poems of his corpus.

Similar arguments can be advanced for the other examples above. Hieron’s martial achievements, for example, could have been celebrated elsewhere in song, especially given the frequency with which the tyrant appears to have patronised literary commemorations of his accomplishments. By ‘reminding’ Hieron of his past military success, Pindar could simultaneously recall earlier poetic celebrations of it (ἀμνάσειεν, Pyth. 1.47). As for Archilochus fr. 96, the addressee of Archilochus’ injunction to remember – Glaucus – reappears elsewhere in his poetry much as Atthis does in Sappho’s (frr. 15, 48.7, 105.1, 117, 131.1); the poet’s invitation to recall Thasos may well look back to Glaucus’ relationship with the island in other poems.¹⁸¹ Of course, these final suggestions can be no more than tempting conjectures on current evidence, but from the work of Sappho and Alcaeus we can conclude that the indexical memory of archaic lyric was not restricted to the realm of myth. Poets’ ‘contemporary’ memories looked not only to their immediate social contexts (real or imagined), but also to the wider construction and articulation of their worlds in song.

III.3.4 Audience Knowledge

As we have seen above (§iii.3.1–3), lyric poets rarely invoke audience’s memories of events. When they do, it is more often through the narrator’s own recollection, rather than those of its internal characters. When we turn to cases of indexical knowledge, however, it appears that lyric’s capacity for more direct engagement between narrator and audience revitalised this allusive mode. Lyric poets occasionally assert their own knowledge of the poetic past, as when Alcaeus claims that he ‘knows for certain’ that one should not move gravel (οἶδ’ ἦ μάν, fr. 344.1), advice that he may have drawn directly from a poem by his Lesbian contemporary Sappho (μὴ κίνη χέραδος, fr. 145 ~ χέραδος μὴ … | κίνεις, Alc. fr. 344.1–2).¹⁸² More frequently, however, lyric poets appeal directly to their audience’s knowledge of the literary and mythical past. In these cases, we can trace the significance of the allusions more clearly than with lyric poets’ indexical memory. Poets appeal to their audience’s familiarity with tradition, explicitly evoking what ‘you all know’.

One such appeal to audience knowledge is Pindar’s evocation of Ajax’s suicide in Isthmian 4 (Isth. 4.35–6b):

                                    ἴστε μάν

35bΑἴαντος ἀλκὰν φοίνιον, τὰν ὀψίᾳ

ἐν νυκτὶ ταμὼν περὶ ᾧ φασγάνῳ μομφὰν ἔχει

36bπαίδεσσιν Ἑλλάνων ὅσοι Τροίανδ’ ἔβαν.

Surely you know of Ajax’s bloodied valour, which he cut through late in the night with his own sword, bringing reproach on all the sons of the Greeks who went to Troy.

Pindar directly invokes his audience’s acquaintance with Ajax’s ἀλκή, another element familiar from the epic tradition (cf. §iii.3.2 above on Tyrtaeus’ θοῦρις ἀλκή). Here, however, the poet does not just evoke Ajax’s character in general, but rather a specific episode of his fabula: his ignominious suicide after losing to Odysseus in the contest for Achilles’ arms, familiar from cyclic epic and archaic art onwards.¹⁸³

It is unclear, however, how stable the details of Ajax’s death were in the early tradition. As Spelman has highlighted, at least in later tradition, the timing of his suicide differs from that in Pindar’s account.¹⁸⁴ In Sophocles’ Ajax, the eponymous hero kills himself during the daytime, a version of events that seems to be found in other later treatments of the myth.¹⁸⁵ It is thus possible that this appeal to the audience’s knowledge may look to more precise precedent than the epic tradition in general, invoking a specific version in which Ajax killed himself at night. If so, the scholion to this passage names a plausible candidate: in discussing the polyvalence of the phrase ὀψίᾳ ἐν νυκτί (‘late in the night’, 35b–6), it notes that ‘the details of the story also agree with those who take the expression as denoting the pre-dawn hours; for the author of the Aethiopis says that Ajax took his own life towards dawn’ (τοῖς δὲ τὸν ὄρθρον ἀκούουσι καὶ τὰ ἀπὸ τῆς ἱστορίας συνᾴδει· ὁ γὰρ τὴν Αἰθιοπίδα γράφων περὶ τὸν ὄρθρον φησὶ τὸν Αἴαντα ἑαυτὸν ἀνελεῖν, Σ Isth. 4.58b = Aeth. fr. 6 GEF).¹⁸⁶ From this scholiastic citation, scholars have argued that Pindar is making a direct reference to the Aethiopis, marking it as familiar to his audience.¹⁸⁷

However, significant caution is necessary here. First, we should note that Ajax’s suicide also featured in the Little Iliad (Il. Parv. arg. 1b GEF). Proclus’ summary of that epic does not specify its precise timing, but a nighttime setting is again most plausible: the suicide immediately followed Ajax’s maddened attack on the Achaeans’ livestock, an event that always takes place at night elsewhere.¹⁸⁸ In his recent case for a specifically Aethiopic reference in Isthmian 4, Spelman dismisses this possibility, considering it ‘significant’ that the scholia only invoke the Aethiopis as Pindaric precedent, with no mention of the Little Iliad.¹⁸⁹ But such an argument from silence is of limited value, especially when discussing ancient habits of scholarly citation, which – just as today – were never exhaustive. Moreover, there are good grounds for supposing that Sophocles’ daytime suicide was a specific innovation of the tragic stage, dependent on the restrictions of tragic staging and the common dramatic motif of a ‘single day’ of action.¹⁹⁰ I thus consider it likely that the traditional epic version of the myth included a nighttime hunt and nocturnal suicide, and that it was only the lasting influence of Sophocles’ play that overrode this tradition. After all, even in the Sophoclean drama, Ajax’s failed attempt to take revenge on the Greek commanders (which likely derives from the Little Iliad) took place during the night (νυκτὸς … τῆσδε, Aj. 21).¹⁹¹ It is only the suicide that is delayed into the next day, to allow a protracted exploration of its consequences.¹⁹² The Aethiopis’ late-night suicide may well not be as distinctive as scholars assume.

On this occasion, therefore, I do not think our evidence is sufficient to argue for an intertextual link with a specific text. A precise epic may be intended, but on current evidence, it would be overly rash to argue for a direct link with the Aethiopis over the Little Iliad.¹⁹³ The most we can plausibly say is that Pindar is evoking his audience’s knowledge of the epic tradition, whether or not he has a specific text in mind. Yet in any case, the subsequent verses’ celebration of Ajax’s enduring fame through Homeric verse (Isth. 4.37–42) seems to attach these epic traditions to the Homeric canon.¹⁹⁴ By stressing Ajax’s honour ‘among mankind’ (τετίμακεν δι’ ἀνθρώπων, 37), Pindar emphasises the hero’s reception among a range of poetic audiences.¹⁹⁵ The universalising and communal aspect of this noun looks back to Pindar’s opening appeal to his audience’s knowledge (ἴστε). Their familiarity with Ajax’s fate proves the success of the epic tradition in preserving his name and reputation, a model for Pindar’s own immortalisation of Melissus’ achievements (43–5).¹⁹⁶

Such an appeal to group knowledge may well build on Homeric poetry: there too, speakers frequently address the knowledge of the whole community (e.g. Il. 2.301, 9.35–36, 10.250, 20.203–4), a knowledge which – as we have seen – often extends to that of Homer’s own audiences (§iii.2). In Pindar’s lyric, however, this appeal to his audience’s collective knowledge has become more pointed: through the second-person plural verb, he addresses them directly.

Pindar’s only other use of the expression ἴστε μάν occurs in the closely related Isthmian 3 and bears a similar indexical force (Isth. 3.13–16):

                        ἀνδρῶν δ’ ἀρετάν

σύμφυτον οὐ κατελέγχει.

ἴστε μὰν Κλεωνύμου

δόξαν παλαιὰν ἅρμασιν·

He does not disgrace the innate excellence of his kinsmen. Surely you know of Cleonymus’ ancient reputation for charioteering.

The relation between this poem and Isthmian 4 has been long debated. Uniquely in Pindar’s corpus, these two poems address the same victor in the same metre, which has prompted some scholars to join them.¹⁹⁷ However, most scholars now accept their independence on a variety of metrical and structural grounds: Isthmian 3 was composed for a chariot victory at Nemea, shortly after Melissus’ earlier success in the Isthmian Games, celebrated in Isthmian 4.¹⁹⁸ When Pindar mentions ‘twin prizes’ in Isth. 3.9 (διδύμων ἀέθλων), he thus refers to the two ‘crowns’ that Melissus has won (cf. στεφάνους, 3.11), as well as the pair of poems which celebrate these achievements (cf. ὑμνῆσαι, 3.7; ἀγαναῖς χαρίτεσσιν, 3.8). In the verses quoted above, however, Pindar looks beyond these two victories to the larger reputation of Melissus’ ancestors for chariot victories. This is again marked as something with which Pindar’s audience should already be familiar (ἴστε μάν). And here too, it seems that they would have been: Isthmian 4 had already recalled the ‘ancient fame’ of his clan, the Cleonymidae (φάμαν παλαιάν, Isth. 4.22), a fame which Pindar there specified as deriving from earlier chariot victories (Isth. 4.25–7):

ἅ τε κἀν γουνοῖς Ἀθανᾶν ἅρμα καρύξαισα νικᾶν

ἔν τ’ Ἀδραστείοις ἀέθλοις Σικυῶνος ὤπασεν

τοιάδε τῶν τότ’ ἐόντων φύλλ’ ἀοιδᾶν.

That fame heralded their chariot’s victory both on the hills of Athens and in Adrastus’ games at Sicyon, and granted them leaves of song such as these from poets of that time.

It is likely that the ‘leaves of song’ (φύλλ’ ἀοιδᾶν, Isth. 4.27) mentioned here are the source of the knowledge that Pindar invokes in Isthmian 3, especially given the verbal echoes between these passages (δόξαν παλαιὰν ἅρμασιν, Isth. 3.16 ~ φάμαν παλαιάν, Isth. 4.22, ἅρμα Isth. 4.25). Pindar expects his audience to be familiar with this family’s reputation from its earlier poetic celebrations, whether composed by Pindar himself or another epinician poet.¹⁹⁹ In Isthmian 3, the emphatic ἴστε μάν gestures indexically to this poetic precedent, reinforced by the adjective παλαιάν, which further emphasises the antiquity of this fame – it is an established feature of the epinician canon. By indexing this precedent through the plural ἴστε, Pindar again evokes his audience’s communal, shared knowledge of past song. He sets his own poetry within a broader epinician tradition, just as he situates Melissus’ victory within a larger framework of familial success.

This emphasis on an audience’s collective knowledge of tradition is a recurring feature of Pindar’s poetics. In fr. 188, the poet claims that ‘you recognise the well-known utterance of Polymnestus, the man from Colophon’ (φθέγμα μὲν πάγκοινον ἔγνωκας Πολυμνάστου Κολοφωνίου ἀνδρός), referring to a poetic predecessor of the seventh century.²⁰⁰ Unlike our previous examples, he employs a singular verb (ἔγνωκας, perhaps directed to a specific addressee), but the communality of this knowledge is still conveyed by the adjective πάγκοινον: the poet’s song is ‘common to all’. In Pythian 3, meanwhile, the poet recognises that both he and his audience ‘know’ of Nestor and Sarpedon ‘from resounding verses’ (ἐξ ἐπέων κελαδεννῶν … γινώσκομεν, Pyth. 3.112–14), explicitly acknowledging their shared epic heritage.²⁰¹ And in Isthmian 2, he claims that Thrasybulus’ family is ‘not unfamiliar’ with epinician poetry (οὐκ ἀγνῶτες, Isth. 2.30–2), a litotic expression which underscores how frequently the Emmenidae were recipients of poetic praise (cf. εὐδόξων … ἀνδρῶν, Isth. 2.34).²⁰² Indeed, this claim concludes a list of Xenocrates’ earlier victories which had begun with a similar reference to a ‘not unknown’ Isthmian chariot victory (οὐκ ἄγνωτ’ … Ἰσθμίαν ἵπποισι νίκαν, Isth. 2.12–13), a phrase which may again look to earlier literary celebrations of former achievements.²⁰³ The emphatically repeated litotes reinforces the sense that Pindar’s audience, too, should be familiar with these events from earlier song.

Such appeals to the audience’s knowledge thus gestured to traditions of myth and poetry with which they would be familiar, against which the poet could situate his own work.²⁰⁴ But as with indexical hearsay in lyric, this appeal to audience knowledge could also invite audiences to supplement a myth with their broader familiarity of tradition. In Bacchylides’ ninth epinician, for example, a poem composed for the Phliasian athlete Automedes, the poet opens an allusive catalogue of Asopus’ daughters by appealing to his audience’s knowledge (Bacchyl. 9.47–56):

στείχει δι’ εὐρείας κελε[ύ]θου

μυρία πάντᾳ φάτις

σᾶς γενεᾶς λιπαρο-

    ζώνων θυγατρῶν, ἃς θε̣[ο]ί

σὺν τύχαις ᾤκισσαν ἀρχα-

    γοὺς ἀπορθήτων ἀγυιᾶν.

τίς γὰρ οὐκ οἶδεν κυανοπλοκάμου

   Θήβας ἐΰδμα[τον πόλι]ν,

ἢ τὰν μεγαλώνυ]μον Αἴγιναν, μεγ[ίστ]ου

   Ζην]ὸ̣ς̣ [ἃ πλαθεῖσα λ]έ̣χει τέκεν ἥρω

Countless reports travel along a broad path in every direction about your family, your bright-girdled daughters, whom gods settled with good fortunes as the founders of unsacked streets. For who does not know of the well-built city of dark-tressed Thebes, or of great-named Aegina, she who approached the bed of most mighty Zeus and bore the hero … ?

After commencing here with Thebes and Aegina, the subsequent fragmentary lines appear to mention Aegina’s son Aeacus (father of Peleus and Telamon) and continue with a list of other Asopids, before ending in a closural ring composition (9.64–5). The opening emphasis on the family’s fame and renown (μυρία … φάτις, v. 48; [μεγαλώνυ]μον, v. 55) emphasises the traditionality of the catalogue that follows, a familiarity that is reinforced by Bacchylides’ appeal to the audience’s knowledge. The rhetorical question (‘For who does not know … ?’, v. 53) implies that everyone is expected to be familiar with this myth.²⁰⁵ And indeed, the list of Asopus’ daughters, all of whom had been wooed by gods and had become the eponyms of cities, appears to have been an established legend. A fragment of Corinna offers a similar list of nine Asopids, containing much overlap with Bacchylides (fr. 654 col. ii–iv).²⁰⁶ And as Douglas Cairns has argued, both Corinna and Bacchylides seem to be following an earlier version of the myth, perhaps that by the Corinthian Eumelus or the Hesiodic Catalogue of Women.²⁰⁷ We know, moreover, that an Asopid ancestry was an important feature of the Phliasians’ local mythology and part of a larger debate as to whether the ancestor of these illustrious city-nymphs was the Asopus river in Boeotia or its namesake in Phliasian territory.²⁰⁸ As part of their claim, the Phliasians dedicated both a statue group of Zeus and Aegina in Delphi (Paus. 10.13.6) and a group of Zeus, Asopus and five Asopids (including Thebes) at Olympia (Paus. 5.22.6). In asking who is not familiar with these famous cities and their Asopid ancestry, Bacchylides indexes his engagement with a familiar and politically charged local myth.

Besides evoking a well-known myth, this invitation for an audience to recall their knowledge of the Asopids also invites them to supplement the bare details that Bacchylides offers, especially in relation to the first name he mentions: Thebes. The city is described here as ‘well-built’ (ἐΰδμα[τον], 54) and introduced as an example of ‘unsacked streets’ (ἀπορθήτων ἀγυιᾶν, 52). The earlier part of Bacchylides’ poem had already recalled the failed expedition of the Seven against Thebes (9.10–20), an event which on the face of it reinforces this assessment: the city stood strong and repulsed its assailants.²⁰⁹ But any mention of the expedition of the Seven cannot fail to evoke thoughts of its sequel, the successful sacking of the city by the Epigonoi (§iv.2.3). In appealing to his audience’s knowledge of the myth, Bacchylides’ silence on this point resonates all too loudly. ‘Yes’, we are invited to reply, ‘we do know what happened to Thebes’. Scholars have seen a political purpose underlying this suppression of the Epigonoi myth, a way to downplay and negate Argive achievement while simultaneously ‘super-imposing a skewed pro-Phliasian genealogy’ onto Thebes.²¹⁰ But despite the explicit silence, the poet’s appeal to his audience’s knowledge gives space for the lingering shadow of tradition to rear its head, undermining any simple patriotism.

Indexical appeals to audience knowledge, just like indexical hearsay, therefore, can invite audiences to fill in the gaps of a story with their knowledge of tradition, complicating a simple treatment of myth by evoking elements that remain untold.

III.4 Conclusions

Memory and knowledge both functioned as significant indices of allusion in archaic Greek poetry. In many ways, this indexical device is a foil and complement to indexical hearsay. Whereas the latter evokes external traditions that are circulating on the airwaves of fama, ready to be picked up by observant listeners, indexical memory involves a more internal and personal act of preserving, retaining and transmitting knowledge. But as with hearsay, these metaphors of allusion are an apt model for the nature and process of poetic composition and performance.

In comparison to indexical hearsay, we have encountered more variation and divergence in the use of this indexical mode across time and genres. It is most prevalent in Homeric epic, where characters repeatedly urge their interlocutors to recall earlier events of tradition, simultaneously inviting Homer’s audiences to recall their own knowledge of episodes both within and beyond his poetry. Yet there are relatively few precise parallels for this in the rest of early Greek epic and archaic lyric. The extremely fragmentary state of much of our evidence must play some role in this absence, but it is striking that even in Pindar’s completely extant epinicians and Hesiod’s extant didactic works, poetic memory is not as productive. The principal reason for this seems to be the way in which these poems treat myth, and their relative dearth of character speech. Lyric poets in particular rarely tell a mythical narrative in its own right but rather introduce one as an exemplum or point of comparison for events of the real world. When we return to the mimetic world of tragedy, it is perhaps no surprise that cases of indexical memory appear to flourish once more in a mythical context.²¹¹

Despite its limited presence in lyric character speech, however, lyric poets adapted this indexical mode into various new forms, taking advantage of the greater prominence of the lyric narratorial voice. We have seen instances where the narrator’s personal memories overlap and blur into recollections of myth and poetry, as well as cases where poets recall their own past poems, highlighting links across their cycles of song. In addition, lyric poets directly appeal to their audience’s knowledge of the poetic tradition, a more overt and direct signposting of tradition.

This allusive index was thus already deeply engrained in Greek poetry from the very beginning. It was primarily used to gesture to and incorporate other mythical narratives, marking the poet’s mastery of tradition. But we have also noted cases of misremembering, where a character’s memory is pointedly selective, inviting audiences to fill in the gaps. In both cases, the device evokes wider traditions within which each poet situates himself and his work.