Agocs ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 Article copyright held by the author(s) and made available under the Creative Commons Attribution License CC-BY https://creativecommons.org/licenses/by/4.0/ Message-Stick of the Muses: Lyric Epistolarity and Textuality in Pindar and Bacchylides Peter Agócs T HAS LONG BEEN RECOGNIZED that certain epinicians and skolion-fragments of Pindar and Bacchylides exploit a mise- en-scène that defines the song as something “sent.”1 The epinicians are Pindar Ol. 6, Ol. 7, Pyth. 2, Pyth. 3, Nem. 3, Isth. 2, and Bacchylides 5.2 All seem to date between the early 470s and the late 460s BCE. In these odes, the idea of “sent song” is often expressed by a part of the verb πέµπω: elsewhere, one can infer the “sending” from the context.3 The verb πέµπω is com- mon in other lyric, but this particular usage, which frames the utterance as a message sent from the speaker to a human lau- dandus, is unique to the encomiastic poetry of Pindar and Bac- chylides. Circumscribed by date, context, author, genre, and addressees, the “sending motif” thus constitutes an innovation 1 In Pyth. 2 and Isth. 2 this was noticed by Alexandrian commentators schol. Pyth. 2.6b and schol. Isth. 2.inscr.a, who describe both odes as “sent” (ἀποστολικόν) songs. See D. C. Young, “Pindar Pythians 2 and 3: Inscrip- tional ποτέ and the ‘Poetic Epistle’,” HSCP 87 (1983) 30–48, at 31–32 n.3; A. Tedeschi, “L’invio del carme nella poesia lirica arcaica: Pindaro e Bac- chilide,” StIt 3 (1985) 29–54, at 35 nn.19–20. Beyond the Christian applica- tions of ‘apostolic’, ancient Greek uses of the term are limited to lyric criticism: cf. Phot. Bibl. 322a and Athen. 631C–D. 2 See Tedeschi, StIt 3 (1985) 45–46. The non-epinician fragments (Pind. fr.124ab; Bacchyl. fr.20B–C) will be discussed elsewhere. Other epinicians where pempein occurs (Ol. 7.1–14 and Isth. 5.62–63) are not “sent” odes. 3 E.g. Pyth. 3, Ol. 6, and Isth. 2. I 386 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 in the language of Greek song.4 But the precise nature and significance of this departure have not been fully appreciated. Where the sending motif is noticed at all, it is usually interpreted as a reflection of real-life processes of commissioning, production, and performance, or dismissed as a simple façon de parler, an epistolary fiction of little consequence to our understanding.5 Both these approaches fail to engage with the sending motif’s pragmatic effects. Recently, Spelman, discussing two examples examined below (Pyth. 2 and Nem. 3), has emphasized how the poetic voice’s spatial and temporal distance from the addressee is “typically connected with Pindar qua celebrated professional poet who immortalizes his subject.”6 Apart from associating the sending motif with the idea of song’s permanence, Spelman also notices its recurring opposition, as a trope of absence, to the epinician “arrival motif,”7 the figure that foregrounds the poet’s (or speaker’s) arrival and presence at the scene of the epinician performance, 4 Except for Nem. 3 and Bacchyl. fr.20B, the motif is found in odes dedi- cated to members or affiliates of Sicilian tyrant families. 5 For the evidence see J. Herington, Poetry into Drama. Early Tragedy and the Greek Poetic Tradition (Berkeley 1985) 26 and 189–195, and Tedeschi, StIt 3 (1985) 29–54. Cf. W. Mullen, Choreia. Pindar and Dance (Princeton 1982) 20– 31 (“sending” means the poet was present), Tedeschi 51–54 (he was absent), and M. Spadafora, “Poems Cleaving the Ionian Sea: Modes of Trans- mission of Poetry in the Sicilian odes,” in H. L. Reid et al. (eds.), Pindar in Sicily (Sioux City 2020) 221–240. U. von Wilamowitz-Moellendorff, Pindaros (Berlin 1922), argued that certain epinicians were actually “poetic epistles”; see contra Young, HSCP 87 (1983) 31–33; C. Carey, A Commentary on Five Odes of Pindar (Salem 1981) 23–24; D. A. Schmidt, “The Performance of Bacchyl- ides Ode 5,” CQ 37 (1987) 20–23; K. A. Morgan, Pindar and the Construction of Syracusan Monarchy in the Fifth Century B.C. (Oxford 2015) 268–272; P. Ceccarelli, Ancient Greek Letter Writing: A Cultural History (Oxford 2013) 13–19. On early lyric texts and their circulation see T. A. Hadjimichael, The Emergence of the Lyric Canon (Oxford 2019). 6 H. Spelman, Pindar and the Poetics of Permanence (Oxford 2018) 24. 7 On the “arrival motif” see n.23 below. PETER AGÓCS 387 ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 which he argues is associated with “Pindar’s poetic task as it is related to personal and religious relationships.” This antithesis between arrival and sending, presence and absence, plays an important role in the interpretations presented below, building on and corroborating earlier work on the spatial and temporal pragmatics of the epinician voice. The sending motif emerges as a way of expressing the mediated nature of the poetic utterance.8 But we still need an exploration of the pragmatics and underlying biases of the motif. In this essay, I aim to demonstrate that attempts to explain the epinician sending theme as historically factual, rhetorically redundant, or merely metaphorical miss its significance as a clue to early fifth-century Greek ideas about textuality and the voice. This realisation will then help me to reframe some well-understood aspects and paradoxes of epinician pragmatics in their broader anthro- pological context. Early Greek concepts of textuality differed substantially from our own. They were mainly based on the idea of the text as a form of utterance and voice. I argue that the sending motif should be read with the similarities in mind documented by Day, Steiner, and others between the voice of Pindar’s “ego- centric” epinician poetry and the voice found in the epigraphic corpus of early “speaking object inscriptions” (oggetti parlanti).9 8 I owe much to G. B. D’Alessio’s discussions of Pindaric voice ( “First- person Problems in Pindar,” BICS 39 [1994] 117–139) and time (“Past Future and Present Past: Temporal Deixis in Ancient Greek Lyric,” Arethusa 37 [2004] 267–294). In “The Problem of the Absent I: Lyric Poetry and Deixis in ‘Mediated’ Communication,” AION 42 (2020) 1–30, D’Alessio discusses deictic effects of distance as markers of authorial voice with methods and conclusions similar to mine. 9 The term comes from M. Burzachechi, “Oggetti parlanti nelle epigrafi greche,” Epigraphica 24 (1962) 3–54. See D. Steiner, “Pindar’s ‘oggetti parlanti’,” HSCP 95 (1993) 159–180, and The Tyrant’s Writ. Myths and Images of Writing in Ancient Greece (Princeton 1994) 91–99. On epigraphic poetry and voiced reading in the seventh–fifth centuries see J. Svenbro, Phrasikleia. An 388 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 The sending motif uses similar strategies of spatiotemporal dis- placement to call attention to the paradox of a song that exists simultaneously as a text separable from its author and as per- formed musical utterance. In fact, its main function as a mise- en-scène is to distance the speaker both spatially and temporally from the performance of his own ongoing utterance, which is separated from him and “sent,” even as the performance is projected into the future.10 This self-distancing works differ- ently in each instance, but its effects are always similar. It also enables the epinician speaker both to assume more emphati- cally the voice of an author (as opposed to a performer) and to develop a closer, more intimate style of communication with the recipient of the sent utterance. This style is appropriate to encomiastic poetry, which is all about the poet-patron relation- ship. From an anthropological perspective, the motif is a trope of “entextualization.” It calls attention to the fact and process by which the epinician utterance is entextualized: “rendered … as text, detachable from its local context,” and “made available for repetition or recreation in other contexts.”11 In any given ___ Anthropology of Reading in Ancient Greece (Ithaca 1993), and J. W. Day, “Rituals in Stone: Early Greek Grave Epigrams and Monuments,” JHS 109 (1989) 16–28, “Epigram and Reader: Generic Force as (Re‑)activation of Ritual,” in M. Depew et al. (eds.), Matrices of Genre: Authors, Canons, and Society (Cambridge [Mass.] 2001) 37–57, and Archaic Greek Epigram and Dedication: Representation and Reperformance (Cambridge 2010). 10 On spatial effects see N. Felson, “Vicarious Transport: Fictive Deixis in Pindar’s Pythian Four,” HSCP 99 (1999) 1–31, and “Introduction,” Arethusa 37 (2004) 253–266, on “vicarious transport”; and D’Alessio, AION 42 (2020) 1–30. On time, see D’Alessio, Aresthusa 37 (2004) 267–294. Also B. Currie, “Choral Lyric: Pindar and Bacchylides,” in I. de Jong (ed.), Space in Ancient Greek Narrative (Leiden 2012) 285–303; V. Lewis, Myth, Locality, and Identity in Pindar’s Sicilian Odes (New York 2019); L. Kurke and R. Neer, Pindar, Song, and Space: Towards a Lyric Archaeology (Baltimore 2019). 11 For “entextualization” in anthropology see M. Silverstein and G. Urban, Natural Histories of Discourse (Chicago 1996) 21 (from which the quotation comes); K. Barber, The Anthropology of Texts, Persons and Publics: Oral PETER AGÓCS 389 ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 context, entextualization, as a process, may or may not depend on the art of writing,12 but I argue that epinician “sending” engages very strongly with the social fact of literacy. But it never allows the song to be completely subsumed into “litera- ture.” “Sending,” I will argue, is never fully opposed to “perfor- mance.” Instances of the motif are always hedged round with language that alludes to the musical and performative aspects of the song. The sending motif thus captures the paradoxical absent presence of the entextualized Greek poetic voice. I propose to examine the sending motif in the epinician cor- pus, example by example, to demonstrate how it affects each ode’s self-framing through referential gestures that function like the contextualization cues (“here” and “there,” “now” and “then,” “I” and “you”) of ordinary conversation. The readings follow a progression from the simplest examples to the most complex. Section one focuses on two odes for Hieron I (Bacchyl. 5 and Pyth. 3) that exemplify the motif’s basic parameters: how it distorts space and time, distancing the speaker from the performance of his own utterance and projecting it into the song’s own future; how it combines a focus on the song as an object with the language of musical performance and the voice; how it can foreground the recipro- cal relationship between the speaker and his recipient, and how it foregrounds authorship. The next sections, on Pyth. 2 and Nem. 3, focus on temporality and further illustrate the para- doxical tensions created by the sending theme between “here” and “there,” presence and absence, and song-performance and text. Sections four and five focus on Ol. 6 and Isth. 2, drawing out the motif’s relationship with textuality and vocality by examining Pindar’s use of human mediator-figures who behave ___ and Written Culture in Africa and Beyond (Cambridge 2007) 22–23. 12 Barber, Anthropology of Texts, Persons and Publics, discusses a variety of “oral texts” from contemporary Africa, including proverbs and praise poetry. 390 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 like texts. Ol. 6 vividily illustrates the association of writing and memory, and brings out the idea of performance as an interpretation of a scripted song, while Isth. 2 focuses on the role of fixed texts in preserving social memory and family kleos through ritual re-enactment. The paradoxical tensions between presence and absence, “arrival” and “sending,” performance and text, and author and performer reveal themselves to be ele- ments of the image of textuality that the sending motif projects. The concluding section demonstrates how the motif and the tensions it creates fit into the wider early fifth-century discourse of textuality, exploring how these insights contribute to the an- thropology of early Greek literacy. Ultimately the sending motif emerges as a trope of a partial or incomplete textuality de- fined always in relation to the voice. 1. Understanding sending: Bacchylides 5 and Pythian 3 The sending motif’s most basic effect is a certain type of spatio-temporal distortion. Two odes for Hieron of Syracuse, Bacchyl. 5 and Pyth. 3, exemplify this.13 The ring-composition that ties the opening of Bacchyl. 5 (1–16) to its end (191–200) is the simplest example of the sending motif in epinician. Bacchyl. 5 defines itself entirely as sent song (1–16): Εὔµοιρε̣ [Σ]υρακ[οσίω]ν ἱπποδινήτων στρατα[γ]έ, γνώσηι µὲν [ἰ]οστεφάνω̣ν Μοισᾶν γλυκ[ύ]δωρον ἄγαλ̣µ̣α, τῶν γε νῦν αἴ τις ἐπιχθονίων, 5 ὀρθῶς· φρένα δ’ εὐθύδικ[ο]ν ἀτρέµ’ ἀµπαύσας µεριµνᾶν δεῦρ’ ⟨ἄγ’⟩ ἄθρησον νόωι· 13 Pindar and Bacchylides are quoted from the Teubners of Snell and Maehler. Translations were influenced by W. H. Race, Pindar (Cambridge [Mass.] 1997), D. A. Campbell, Greek Lyric IV (Cambridge [Mass.] 1992), and D. L. Cairns, Bacchylides: Five Epinician Odes (Cambridge 2010). Refer- ences to Pindaric scholia are to the Teubner edition of A. B. Drachmann. PETER AGÓCS 391 ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 ἦ σὺν Χαρίτεσσι βαθυζώνοις ὑφάνας ὕµνον ἀπὸ ζαθέας 10 νάσου ξένος ὑµετέραν ἐς κλυτὰν πέµπει πόλιν, χρυσάµπυκος Οὐρανίας κλεινὸς θεράπων· ἐθέλει {δὲ} γᾶρυν ἐκ στηθέ͜ων χέων 15 αἰνεῖν Ἱέρωνα … O lucky general of horse-whirling Syracuse, you will recognize this sweet gift and marvel of the violet-crowned Muses if anyone earthborn [can], fairly for what it is, and, resting your righteously-judging heart without fear from cares, come! turn your mind this way. Truly, your guest-friend, the famous servant of Ourania of the golden headband, having woven a song with the help of the deep-bosomed Graces, sends [it] from the sacred isle [i.e Bacchylides’ home island, Ceos] to your famous city. He wishes to pour a voice out from his breast in praise of Hieron … The song is represented as an “agalma of the Muses” (4), as something woven (9–10), and as an outpouring of the voice.14 It seems to exist in two states at once: as object and as perfor- mance. Agalma evokes dedications and the power of dedicatory ritual to generate charis between giver and recipient, a relation- ship then embedded in the ties of formalized guest-friendship between foreigners (xenia).15 Hieron, the recipient, is invited to abandon his cares of state, turning his mind towards the poet’s utterance and estimating its value.16 These themes are revisited in the closing epode (191–200), where the speaker reiterates his willingness to send a song (195–197): “I am easily persuaded to 14 On this Bereitwilligkeits-Motiv see H. Maehler, Die Lieder des Bakchylides (Leiden 1982) 83, 92. 15 Cf. Day, Archaic Greek Epigram and Dedication 85–129; Maehler, Die Lieder des Bakchylides 87; A. Ford, The Origins of Criticism: Literary Culture and Poetic Theory in Classical Greece (Princeton 2002) 117–118. 16 As Cairns (Bacchylides 218) notes, the closest analogies to this passage in epinician are Pyth. 2.69–71 and Ol. 6.98, both of which I discuss below. 392 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 send (πέµπειν) Hieron a fame-bringing tongue (εὐκλέα … γλῶσσαν).” If the proem glosses the sending of song as an outpouring of breath and voice, the laudator here declares his willingness to send his “friend” an object, his “tongue.”17 Nothing better expresses the in-betweenness of a performance- text than this mailing of a tongue. The ode’s use of spatio- temporal deixis establishes a more authorial tone and a focus on the song as an agalma, as currency in an exchange of charites between poet and laudandus. A similar spatio-temporal dynamic is apparent in Pyth. 3, also for Hieron, which is undated and lacks any occasional context apart from a brief mention (73–74) of Pythian victories won in 482 and 478 BCE.18 This explains how it found a place among the Pythians, but does not explain what the ode actually was in its first historical context. It frames itself as consolation rather than praise. Hieron is apparently dying from a chronic illness.19 Pindar tries to comfort him with a series of thoughts structured around antinomies: near and far, possible and impossible, death and immortality, human ambition and self-knowledge.20 His powerlessness to help Hieron is established when he wishes Chiron were alive to heal him (1–3):21 an opening wish that, while conventional and human, is also hybristic in its neglect of 17 The metaphor is also found in Pindar (Ol. 9.41–42, Nem. 4.85–88). 18 Like Pyth. 2, Pyth. 3 cannot be dated any more securely than the decade from the early 470s to 467 when Hieron died. A date perhaps around 470 seems likeliest: see schol. Pyth. 3.tit. ii, 61; inscr.a; B. Gentili, Pindaro. Le Pitiche (Milan 1995) xli–xlii; Young, HSCP 87 (1983) 30–48; Tedeschi, StIt 3 (1985) 47; see also D’Alessio, AION 42 (2020) 18–20. 19 On Hieron’s illness see Pyth. 1.50–52, with schol. Pyth. 1.89ab and schol. Pyth. 3.117. 20 See D. C. Young, Three Odes of Pindar. A Literary study of Pythian 11, Pythian 3, Olympian 3 (Leiden 1968) 27–68, 116–120; Gentili, Pitiche xli–xlii, 81–82. 21 Young, Three Odes 29–31; Gentili, Pitiche 82 n.1; Morgan, Pindar 273– 274, 298–299. PETER AGÓCS 393 ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 death’s universality. The mythic exempla of Coronis and Asclepius (5–60) that follow argue that death must be accepted. In the ode’s second half (80–115) Pindar asserts that song has the power to grant a weak immortality to certain fortunate individuals. Just at the join between negative mythic exempla and positive moralizing (61–76) he returns to his opening theme of incapacity: “Do not, my soul, strive after immortal life!” (61). What follows is the only place in the text where epinician motifs accumulate,22 as Pindar turns the familiar arrival motif on its head.23 Were Chiron alive, he would have persuaded him to send Hieron a healer like Asclepius. Had the singer himself crossed the sea to Syracuse bearing “golden health and a komos (victory revel) to add radiance to the crowns from the Pythian games,” he would have looked to his sick friend like a light of salvation and glory more radiant than the sun. But this is make-believe. Instead, Pindar offers prayer to Cybele, the goddess whose rites take place in front of his porch in Thebes (77–79). From this point, he asserts that the only cure for dying is immortality in song. Only here does he address Hieron. Until now, no addressee was indicated except the speaker’s own soul. Now Hieron is drawn into a Hesiodic wisdom-sermon. Pindar’s absence is thus one of the ode’s great themes. He is in Thebes and Hieron in Syracuse, and he can give only song and consolation; but his closing argument speaks with the directness of a letter. Although Pyth. 3 is never formally “sent,” clearly it, and not the poet, will travel to Syracuse.24 As both these examples suggest, the sending motif develops a broader tendency in epinician to frame the lyric utterance as a 22 Young, Three Odes 44–50; Morgan, Pindar 270–280, 282–286; Gentili, Pitiche xli. 23 On the “arrival motif ” see E. L. Bundy, Studia Pindarica I–II (Berkeley 1986) 27–28; E. Cingano, “Commento,” in Gentili, Pitiche 366. 24 Cf. Spelman, Pindar 171–173. 394 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 kind of composition-in-performance, in a style that has been described as “fictive spontaneity” or even “oral subterfuge.”25 As D’Alessio has shown, this is not always accompanied (especially in Pindar) by a focus on what we might call “deictic simultaneity,” where the time of the utterance (which he calls “coding time”/CT) coincides, as in ordinary speech, with the time of its reception (“receiving time”/RT).26 Sometimes, the ode projects itself as a performance into the future, disrupting the natural mimetic quality of a frame based on deictic simul- taneity. This is what the sending motif does in both Bacchyl. 5 and Pyth. 3, where the epinician speaker’s utterance is both spatially distanced and somehow temporally prior to its own future performance. Elsewhere, the sending motif intensifies these spatio-temporal distortions, as the next section will demonstrate. 2. Pythian 2: “arrival” and “sending” Bacchyl. 5 and Pyth. 3 are at least consistent: sending is the only enunciative context established for the utterance. They radicalize an existing tendency of Greek praise-poetry to distort what one might call the normal, naturalistic system of deictic cues in lyric, where “I” addresses “you” in a speech-context defined by deictic simultaneity (CT = RT). But in other odes the sending theme creates extraordinary deictic inconsistencies. Pyth. 2 (1–8) is perhaps the most obvious example: 25 See e.g. Carey, Commentary 5; A. Miller, “Pindaric Mimesis: The Asso- ciative Mode,” CJ 89 (1993) 21–59, at 21–23 nn.1–3; R. Scodel, “Self- correction, Spontaneity, and Orality in Archaic Poetry,” in I. Worthington (ed.), Voice into Text. Orality and Literacy in Ancient Greece (Leiden 1996) 59–79; I. L. Pfeijffer, Three Aeginetan Odes of Pindar (Leiden 1999) 34–37; A. Bonifazi, Mescolare un cratere di canti: pragmatica della poesia epinicia in Pindaro (Alessandria 2001) 60–73; A. D. Morrison, The Narrator in Archaic Greek and Hellenistic Poetry (Cambridge 2007) 67–73. 26 D’Alessio, Arethusa 37 (2004) 269–270. PETER AGÓCS 395 ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 Μεγαλοπόλιες ὦ Συράκοσαι, βαθυπολέµου τέµενος Ἄρεος, ἀνδρῶν ἵππων τε σιδαροχαρ- µᾶν δαιµόνιαι τροφοί, ὔµµιν τόδε τᾶν λιπαρᾶν ἀπὸ Θηβᾶν φέρων µέλος ἔρχοµαι ἀγγελίαν τετραορίας ἐλελίχθονος, εὐάρµατος Ἱέρων ἐν ᾇ κρατέων 5 τηλαυγέσιν ἀνέδησεν Ὀρτυγίαν στεφάνοις, ποταµίας ἕδος Ἀρτέµιδος, ἇς οὐκ ἄτερ κείνας ἀγαναῖσιν ἐν χερσὶ ποικιλα- νίους ἐδάµασσε πώλους. Syracuse, great city, sanctuary of Ares deep-founded in war, divine nurse to men and horses who delight in steel! To you I come from wealthy Thebes, bringing this song, news of the earth-shaking four-horse chariot in which Hieron, man of fine chariots, crowned with far-shining garlands of victory the brow of Ortygia, the seat of Artemis of the river, not without whose help he broke in gentle hands those fillies of his with em- broidered reins. The speaker announces his arrival in Syracuse, bringing from Thebes a song (τόδε µέλος, 3–4) that is also fresh news (ἀγγελία, 4) of Hieron’s chariot-victory.27 Compare this open- ing, with its emphatic use of the arrival motif, to the close of the poem’s third triad where, as the ode enters its final movement, the speaker breaks off his praise (56–67), bidding Hieron fare- well and sending him the song (67–71): … χαῖ- ρε· τόδε µέν κατά Φοίνισσαν ἐµπολάν µέλος ὑπέρ πολιᾶς ἁλός πέµπεται· τὸ Καστόρειον δ’ ἐν Αἰολίδεσσι χορδαῖς θέλων ἄθρησον χάριν ἑπτακτύπου 70 φόρµιγγος ἀντόµενος. Farewell. This song is being sent like Phoenician cargo over the hoary salt sea: but as for the Kastoreion in Aeolian strings, look 27 On the angelia see Carey, Commentary 21; L. Nash, The Aggelia in Pindar (New York 1990); Cingano in Gentili, Pitiche 339–340; and Day, Archaic Greek Epigram and Dedication 201–228. 396 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 you gladly upon it, the gift of the seven-thundering lyre, as you go down to meet it. In both places τόδε µέλος (3–4, 67–68) must describe the cur- rent song.28 In 3–4 that song arrives in Syracuse; at 67–71 the speaker could be anywhere but Syracuse. “Sending” (CT) logically precedes “arrival” (RT). The proem is thus temporally inconsistent with 67–71.29 Scholarly discussions (if they notice this problem at all) are divided between those that think the proem’s arrival motif is a fiction, while the sending confirms the poet’s historical absence, and those who think that the proem attests Pindar’s presence in Syracuse while the sending is fictional.30 There is no empirical basis for either judgement. Both “arrival” and “sending” take place in the vocalized present of the utterance, and each exists in its own natural spatial parameters (“Syracuse” vs. “Thebes”). Hellenistic scholars certainly noticed the inconsistency and ac- cording to the scholia tried a different solution, resolving the problem with a λύσις ἐκ τοῦ προσώπου (“solution from the person/character speaking”: the critical practice, familiar from the Homeric scholia, of changing the ascription of a given pas- sage to resolve some perceived infelicity).31 They comment that the speaker at Pyth. 2.1–8 cannot logically be the poet, who was 28 Cf. Nem. 4.44–45, Isth. 3/4.37–41, Bacchyl. 13.221–225, and Nem. 3.76 (discussed below); also Pyth. 10.65 and Pyth. 12.3–5. 29 Tedeschi, StIt 3 (1985) 30–35; Cingano in Gentili, Pitiche 366. 30 Cingano in Gentili, Pitiche 390–391; Carey, Commentary 25; Tedeschi, StIt 3 (1985) 32–35, esp. nn.10–11. 31 L. Prauscello, Singing Alexandria. Music between Practice and Textual Trans- mission (Leiden 2006) 40–51; on the λύσις ἐκ τοῦ προσώπου see D’Alessio, BICS 39 (1994) 117–118 n.3; B. Currie, “The Pindaric First Person in Flux,” ClAnt 32 (2013) 243–282, at 256–257; F. Schironi, “The Speaking Persona: Ancient Commentators on Choral Performance,” in M. Foster et al. (eds.), Genre in Archaic and Classical Greek Poetry. Theories and Models (Leiden 2019) 109–132. PETER AGÓCS 397 ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 not present at the performance (οὐκ αὐτὸς ὁ Πίνδαρος ἥκων εἰς τὰς Συρακούσας ταῦτά φησιν, οὐ γὰρ ἀφῖκται πρὸς τὸν Ἱέρωνα, ἀλλ’ ἡ ᾠδὴ ἀποστολικὴ, schol. Pyth. 2.6b) but must be identified either with “the person who carried the poem to the tyrant” (ἐκ τοῦ διακοµίσαντος … ἐστιν ὁ λόγος), or with the chorus: “for he [Pin- dar] used to send odes by means of a chorus” (ἢ ἐκ τοῦ χοροῦ· διὰ γὰρ χοροῦ ἔπεµπε τοὺς ὕµνους). The verb πέµπεται (68) could indeed describe an act of “bringing” (cf. φέρων, 3) or “escort- ing” a µέλος, and the present tenses of ἔρχοµαι and πέµπεται could thus simultaneously be true.32 For Pyth. 2 this is a neat solution;33 but, as the next section will show, it cannot explain what happens in Nem. 3. Nor can it answer the question why it is that the journey in Pyth. 2 begins in the epinician moment (RT) and works its way back to a point closer to (if not entirely identical with) the time of composition (CT). The ode’s emplotment is bizarre. It develops in reverse from performance (the default instance in lyric) to composition and sending.34 And it is complicated by the speaker’s sudden declaration at 62–63 that he is prepared to “embark on a garlanded ship to praise your (Hieron’s) excellence” (εὐανθέα δ’ ἀναβάσοµαι στόλον ἀµφ ̓ ἀρετᾷ / κελαδέων). The “fortunate voyage,” which may mean nothing more than that the speaker is devoted to his task of praise, is projected into the future. In the opening, τόδε µέλος describes a present act of speech, which first is pushed into its own future, before suddenly emerging at 67–71 as an “epistle,” just before the speaker makes his paradoxical injunction to the victor: “look and see!” (ἄθρησον). “Seeing” involves the laudandus’ physical presence at, and indeed his response to, the performance, even as voyaging or 32 On πέµπω see LfrgE s.v. I.2–3, with G. Fatouros, Index verborum zur früh- griechischen Lyrik (Heidelberg 1996) s.v.; Tedeschi, StIt 3 (1985) 32–35; Mullen, Choreia 30–31. 33 Accepted by Tedeschi, StIt 3 (1985) 34–35. 34 Prauscello, Singing Alexandria 42–43; cf. Tedeschi, StIt 3 (1985) 32. 398 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 sending over the sea recalls Nem. 5.2–5, where the speaker calls upon his “sweet song” to “go forth from Aegina on every ship.” There the song-medium is the messenger who diffuses the victory-news far and wide. At 67–71 the ode is sent not from the victor’s city but to it, aimed at the addressee. It is also “Phoenician merchandise.” The scholia interpret this as a reference to Pyth. 2 itself, the epinician commissioned by Hieron in exchange for money; this, in their view, is opposed to the Kastoreion, a song sent gratis. This view is still popular today.35 But there are other more natural explanations for the phrase. It points to the distance the ode must travel,36 and its rich exotic nature. The skills of Phoenician craftsmen were famous from Homer onwards.37 And writing itself was a Phoenician invention, a product of Pindar’s own Kadmeian Thebes.38 This leaves the problem of the Kastoreion. While a dominant school of thought proposes to read it in reference to a separate hyporchema for Hieron, I follow those who see a reference to Pyth. 2 itself, in particular to its melody, which presumably modelled itself on a traditional “Kastor-tune” in an Aeolic musical or metrical frame. On this view, Pindar at Pyth. 2.67–71 again invokes the reciprocal exchange of gifts and honors (charis) between poet and patron, employing antithetical 35 Schol. Pyth. 2.125a–b; cf. B. Gentili, “Pindarica III,” QUCC N.S. 40 (1992) 49–55, at 52–53, and Pitiche xlviii n.2; cf. Cingano in Pitiche 391; J. Péron, Les images maritimes de Pindare (Paris 1974) 149–153; R. Campagner, “Reciprocità economica in Pindaro,” QUCC N.S. 29 (1988) 77–93, at 87; Prauscello, Singing Alexandria 42 n.128. 36 Cingano in Gentili, Pitiche 391, notes that πολιᾶς ἁλός evokes Homeric sea-journeys. 37 Hom. Il. 6.290, 23.743, Od. 4.615–619, 15.118, 15.425. See L. Kurke, The Traffic in Praise. Pindar and the Poetics of Social Economy2 (Berkeley 2013) 51– 52; and P. Wilson, “Dancing for Free: Pindar’s Kastor Song for Hieron,” ClAnt 38 (2019) 298–362, at 310–311. 38 On the passage as an allusion to writing see Spadafora, in Pindar in Sicily 236–238. PETER AGÓCS 399 ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 diction typical of rhapsodic endings. The µέν … δέ phrase expresses reciprocal division of roles and different aspects of the work (viewed once as text/object and once as song/process), rather than contrasting two songs or parts of the same one.39 To put it another way, Pyth. 2.67–71 reverts from “text” to “performance.” This reversion to performance-language in Pyth. 2.69–71 limits the song’s autonomy as text. It is not only an object to be sent or traded, or indeed a mere “message” (ἀγγελία, 4), but also a traditional Kastoreion. Hieron is to meet the ode and “gaze” upon it (θέλων ἄθρησον … ἀντόµενος, 69–71). The idea of vision could apply equally to performance or reading, but Hieron will experience, if not the speaker’s presence, at least that of the song. Who will perform it is, however, as unclear as who will bear it to its destination. Pyth. 2 thus moves from presence (arrival), to absence (text), and finally to an imagined future performance of a scripted song, all the while maitaining the default orality (imaginary composition-in-performance) typical of lyric. Whether and how this reflects historical cir- cumstances of commissioning, composition, performance, and transmission is anyone’s guess. But the historical reality of sent song is less important for our understanding of the poets’ 39 For the hyporcheme interpretation see n.35 above; also Wilson, ClAnt 38 (2019) 317–322. For the other view see H. Lloyd-Jones, “Modern Inter- pretation of Pindar: The Second Pythian and Seventh Nemean Odes,” JHS 93 (1973) 109–137, at 123; Carey, Commentary 48–49; G. W. Most, The Measures of Praise (Göttingen 1985) 96–101; Morgan, Pindar 193–194; T. Phillips, “Epinician Variations: Music and Text in Pindar, Pythians 2 and 12,” CQ 63 (2013) 37–56, at 45–52. On the musical aspects: H. Fränkel, Early Greek Poetry and Philosophy (Oxford 1975) 435 n.18, and “Schrullen in den Scholien zu Pindars Nemeen 7 und Olympien 3,” Hermes 89 (1961) 385–397, at 394– 395; Cingano in Gentili, Pitiche 391–392; Herington, Poetry into Drama 28–29, 182; L. Prauscello, “Epinician Sounds: Pindar and Musical Innovation,” in P. Agócs et al. (eds.), Reading the Victory Ode (Cambridge 2012) 59–82, at 79– 80; Phillips 37–56. 400 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 decision to employ the sending motif in a few special contexts than the idea of textuality itself. The ode’s paradoxes— particularly its tensions of “here” and “there,” presence and absence, performance and sent text—enact the song-text’s complex relation to the voice it encodes. 3. Nemean 3: temporalities of performance and text The clash between arrival and sending is enacted in Nem. 3 in a way that complicates the ode’s temporality still further.40 In this epinician for Aristokleidas, an Aeginetan pancratiast, the sending motif occurs late, in 71 ff. The ode itself begins with an elaborate hymnic proem (1–13) in which the speaker, the Muse (his addressee), and a group of men described as “youths, builders of sweet-singing komoi ” (µελιγαρύων τέκτονες κώµων νεανίαι, 4–5) perform an odd variation on the arrival motif. The speaker invokes the Muse to come to Aegina: a komos is waiting by the “water of Asopos” for “voice” (ὄπα, 5). The ancient scholia show uncertainty about this Asopos. Aegina is riverless, but two Asopos rivers (one near Nemea and the other in Boeotia) were identified as possible settings. The speaker, however, tells his Muse to come to Aegina because the youths are waiting “by the Asopian water,” which points to a scene on the island.41 Victory, he says, thirsts for nothing more than song, of which he prays the Muse will “grant an abundance from my own mind” (τᾶς ἀφθονίαν ὄπαζε µήτιος ἁµᾶς ἄπο, 9). He then commands her to “begin … a worthy praise-song” to Zeus that he, the composer, intends to “share out between their 40 Tedeschi, StIt 3 (1985) 35–39. 41 See Tedeschi, StIt 3 (1985) 38–39; Pfeijffer, Three Aeginetan Odes 247– 248; G. A. Privitera, “Pindaro, Nem. iii, 1–5, e l’acqua di Egina,” QUCC N.S. 29 (1988) 63–70; D. Fearn, “Aeginetan Epinician Culture: Naming, Ritual, and Politics,” in D. Fearn (ed.), Aegina: Contexts for Choral Lyric Poetry (Oxford 2011) 175–226, at 188–189; M. Cannatà Fera, Pindaro. Le nemee (Milan 2020) 53–55, 310, 337–338. Most modern interpretations assume a fountain in Aegina town. PETER AGÓCS 401 ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 (i.e. the komos’) voices and the lyre” (ἐγὼ δε κείνων τέ νιν ὀάροις λύρᾳ τε κοινάσοµαι, 10–12). The komasts thus wait to perform a song the Muse has not yet begun, and which the speaker has yet to compose.42 Nemean 3 thus defers its own performance.43 The next roughly sixty lines first praise Aegina and Aristokleidas (14–21) and then narrate two myths, the first of Heracles (21–28) and then of the Aiakidai (29–63).44 At line 64 the speaker breaks off: “It is from here that the far-shining light of the Aiakidai is fixed: Zeus, yours is the blood, yours the contest, that this hymn has struck as it loudly proclaims a joy for the land in the voices of young men” (τὸν ὕµνος ἔβαλεν ὀπὶ νέων ἐπιχώριον χάρµα κελαδέων, 64–66).45 Returning from myth to occasion, the ode is suddenly already complete: the deferred performance of the komos has ended before it even began. A little later, after a brief second praise of Aristokleidas, the sending theme intrudes (Nem. 3.76–84): τῶν οὐκ ἄπεσσι· χαῖρε φίλος· ἐγὼ τόδε τοι πέµπω µεµιγµένον µέλι λευκῷ σὺν γάλακτι, κιρναµένα δ’ ἔερσ’ ἀµφέπει, πόµ’ ἀοίδιµον Αἰολίσσιν ἐν πνοαῖσιν αὐλῶν, ὀψέ περ. ἔστι δ’ ἀιετὸς ὠκὺς ἐν ποτανοῖς, 80 ὃς ἔλαβεν αἶψα, τηλόθε µεταµαιόµενος, δαφοινὸν ἄγραν ποσίν· κραγέται δὲ κολοιοὶ ταπεινὰ νέµονται. 42 Pfeijffer, Three Aeginetan Odes 264. 43 D’Alessio, Arethusa 37 (2004) 26; A. Burnett, Pindar’s Songs for Young Athletes from Aegina (Oxford 2005) 140. 44 Burnett, Pindar’s Songs 13–44; I. Polinskaya, A Local History of Greek Polytheism: Gods, People, and the Land of Aegina, 800–400 BCE (Leiden 2013) 126–163. 45 ἔβαλεν, a perfective aorist (Pfeijffer, Three Aeginetan Odes 373–374), marks completion. Cf. C. Carey, “The Victory Ode in Performance: The Case for the Chorus,” CP 86 (1991) 192–200; and D’Alessio, Arethusa 37 (2004) 26. 402 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 τίν γε µὲν, εὐθρόνου Κλεοῦς ἐθελοί- σας, ἀεθλοφόρου λήµατος ἕνεκεν Νεµέας Ἐπιδαυρόθεν τ’ ἄπο καὶ Μεγάρων δέδορκεν φάος. From which [sc. virtues] you are not far. Farewell, my friend. I send you this [thing] mixed from honey with white milk, the dew mingling and busy about it: a drink of song accompanied by the Aeolian breaths of auloi, late though it be. Among flying things the eagle is swift, who snatches suddenly, searching from afar, the bloodied prey with his feet. Cawing jackdaws range in the lower sky. Upon you, thanks to your victorious resolution, the light of victory has shone from Nemea and Epidauros and Megara through the favor of Kleo of the fair throne. The utterance which began with a komos waiting on Aegina, and which has only just completed itself, ends with no performers in sight, the poem late, and the speaker absent. As at Pyth. 2.67–71, the idea of sending is introduced by χαῖρε, “farewell,” and the similarities between the two passages are compelling. But in Nem. 3 the sending comes as the culmination of a larger plot or quasi-narrative46 built around the speaker’s lateness. The idea of lateness, foreshadowed in the proem’s waiting komos and enacted in the ending’s sending-claim and the eagle simile that follows, culminates in δέδορκεν … φάος (84), which specifies what poet and victor have achieved together.47 As it ends, the song again completes itself as sent utterance, recontextualizing terms and themes established in the proem.48 The performance is still indefinitely deferred. 46 Cf. M. Lowrie, Horace’s Narrative Odes (Oxford 1997) 28–30 n.26. 47 The perfect works similarly to the aorist ἔβαλεν at Nem. 3.65 (n.45 above). 48 ὀψέ περ (80), τηλόθε (81), and µεταµαιόµενος (81) = µαιόµενοι (5); move- ment to Aegina (of Muse = of song); emphasis on the speaker’s position and plans; the healing power of victory (16–18) = refreshing nature of the “singing drink”; finally, closing victory-catalogue = victory-announcement (15–18). PETER AGÓCS 403 ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 In both Pyth. 2 and Nem. 3 Pindar’s use of χαῖρε (Pyth. 2.67– 68, Nem. 3.76) reinforces this gesture of closural summation while drawing attention to the charis-relationship between poet and laudandus.49 In both songs we see rich, paradoxical in- stantiations of the lyric sphragis in diction borrowed from the hymn.50 Paradoxical, because the lyric speaker is most present to his addressee when his absence is most apparent. The reciprocity conveyed through this trope of mediated communi- cation51 can also betray nuances of social tone. Pyth. 2.67–71 emphasizes distance and respect: the value of the poet’s “gift” depends on Hieron.52 Nem. 3.76–84 is sympotic, and the speak- er the laudandus’ philos: χαῖρε is both a farewell and a toast.53 It fits into the ring-composition patterns that bind ending to proem, since the “drink” will satisfy victory’s “thirst” for song.54 In Nem. 3, then, we are closer to the normal epinician pattern where victory summons song to memorialize itself.55 Like Pyth. 2.67–71, Nem. 3.76–79 is shot through with musical diction. Of the ingredients of the magic drink, honey and dew are associated with inspiration and immortality, while dew suggests male fecundity.56 Epic speech “flows like honey.” In Pindar, song gushing like honey or dew is a metaphor for 49 Most, Measures of Praise 96–101; cf. R. Wachter, “Griechisch χαῖρε: Vorgeschichte eines Grusswortes,” MusHelv 55 (1998) 65–75; C. Calame, Masks of Authority: Fiction and Pragmatics in Ancient Greek Poetics (Ithaca 2005) 26–30. 50 Carey, Commentary 48. 51 See D’Alessio, AION 42 (2020) 1–30. 52 Cf. Morgan, Pindar 207. This is close to Bacchyl. 5. 53 S. Instone, “Problems in Pindar’s Third Nemean,” Eranos 91 (1993) 13– 31, at 27, and schol. Nem. 3.132a. 54 D. Boedeker, Descent from Heaven: Images of Dew in Greek Poetry and Religion (Chico 1984) 90–91. 55 Kurke, The Traffic in Praise 74–139. 56 Boedeker, Descent from Heaven 10–51, 54–60. 404 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 vocality.57 Although “mixing” may remind one, as it did the scholiast (schol. Nem. 3.132a), of the sympotic κρατήρ, Pindar also associates it with song’s aesthetic, thematic and structural properties, and with his own compositional decisions.58 The “drink” is thus definitely song, surrounded by its accom- paniment, as the dew surrounds and works upon the other ingredients. The “Aeolian breaths of auloi” must, like the Aeolian strings of Pyth. 2, point to music. The metaphor’s real target is, however, clarified by the description of the drink as a πόµ’ ἀοίδιµoν (79). The Homeric epithet aoidimos may describe “a drink composed of song,” or one “worthy of/full of com- memoration” or “famed in song.” The memory-drink develops the symbolism of Pindar’s cocktail-recipe, for honey and milk make a µελίκρητον: a libation for deceased humans or chthonic powers.59 It may be that the “drink of sung fame” alludes (however strangely in an ode for a living victor) to funeral commemoration even as it places song in the sphere of kleos. At the end, both Nem. 3 and Aristokleidas’ victories are a light shone by Kleo (83–84). The ode attains realization only in performance; that performance being deferred, the singing drink becomes a sent script for future recitals. It is an im- 57 H. Usener, “Milch und Honig,” RhM 57 (1902) 177–195, at 179 n.10; C. Bonner, “Dionysiac Magic and the Greek Land of Cockaigne,” TAPA 41 (1910) 175–185, at 180–185; J. H. Waszink, Biene und Honig als Symbol des Dichters und der Dichtung in der griechisch-römischen Antike (Opladen 1974); Boedeker, Descent from Heaven 80–99; D. Steiner, The Crown of Song: Metaphor in Pindar (Oxford 1986) 45–46 (cf. Pfeijffer, Three Aeginetan Odes 401–402); Cannatà Fera, Nemee 340. 58 See Pfeijffer, Three Aeginetan Odes 398–399; Carey, Commentary 171–172; Most, Measures of Praise 196–199; Boedeker, Descent from Heaven 93–94. 59 E.g. Hom. Od. 10.518–520 = 11.26–29. See Pfeijffer, Three Aeginetan Odes 220–221, 403–404. For a dew-libation to the dead see Pyth. 5.96 with C. Dougherty, The Poetics of Colonization (Oxford 1993) 112; Boedeker, Descent from Heaven 96; C. Segal, “Messages to the Underworld: An Aspect of Poetic Immortalization in Pindar,” AJP 106 (1985) 199–212. PETER AGÓCS 405 ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 mortality-serum that promises an afterlife in memory. Like Pyth. 2, Nem. 3 defines the circumstances of its own reception by enacting them in a way that distorts the ode’s spatio-temporal frame, hinting, through the weird, inauspicious device of a funeral libation offered to a living man, at the prospect of unending memory in the voice of a reperformable text. The emphasis on temporality and deferral created by the sending motif in Nem. 3 thus intersects with the broader social purpose of the epinician ode: to keep the laudandus’ memory and that of his family and community alive through time. 4. Olympian 6: the poetic message-stick The text as medium of communication between poet and patron is only implied in Pyth. 2 and Nem. 3. In contrast, Olym- pian 6, an exuberant epinician celebrating the mule-cart victory of Hagesias (in 472 or 468),60 is much more explicit about it, using a man called Aineas as a proxy for the poet’s script. A member of Hieron’s circle61 who also belonged to the Iamid γένος (a line of seers who oversaw a famous divination-cult at Olympia), Hagesias apparently enjoyed citizens’ rights in Syracuse and Arcadian Stymphalos. The song is distinguished in Pindar’s epinician corpus by the intensity of its framing and by its use of the device of imaginary travel to bind a number of disparate themes and demands together. These include the settings of Hagesias’ life (Syracuse and Stymphalos), the family’s past and present in Greece and Sicily, and the con- flicting demands of aristocratic self-assertion and loyalty to an 60 On dating: schol. Ol. 7.inscr.a–b; see Zs. Adorjáni, Pindars sechste olym- pische Siegesode (Leiden 2014) 52–56; P. Giannini, “Olimpica VI,” in B. Gentili, (ed.), Pindaro. Le Olimpiche (Milan 2013) 142. 61 Schol. Ol. 6.30c infers that Hagesias performed the seer’s duties for Hieron’s armies, while schol. Ol. 6.165 alleges he was slain “after Hieron lost power” (he died in office). 406 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 authoritarian monarchy.62 The ode’s setting jumps from Syracuse to Pitane in Laconia and onward to Stymphalos, even as it moves from a komos celebrated by the victor and his fellow- citizens back to the birth of Iamos, before returning to embrace present-day preparations for Hagesias’ return to Syracuse.63 Half in Stymphalos (present) and half in Syracuse (future per- formance), Ol. 6, as Pindar says of Hagesias himself, has “two anchors” (100–101). The most significant passage in terms of the sending motif is the speaker’s instructions to Aineas in the final triad (87–105): ὄτρυνον νῦν ἑταίρους, Αἰνέα, πρῶτον µὲν Ἥραν Παρθενίαν κελαδῆσαι, γνῶναί τ ̓ ἔπειτ ̔, ἀρχαῖον ὄνειδος ἀλαθέσιν λόγοις εἰ φεύγοµεν, Βοιωτίαν ὗν. 90 ἐσσὶ γὰρ ἄγγελος ὀρθός ἠϋκόµων σκυτάλα Μοι- σᾶν, γλυκὺς κρατὴρ ἀγαφθέγκτων ἀοιδᾶν· εἶπον δὲ µεµνᾶσθαι Συρα- κοσσᾶν … Spur on, Aineas, your comrades, first to praise Hera Parthenia in song, and then to find out whether we have escaped the old insult “Boeotian pig” with a true account. For you are a true 62 Recent discussions: C. Calame, “Metaphorical Travel and Ritual Performance in Epinician Poetry,” in Reading the Victory Ode 303–320; L. Athanassaki, “Recreating the Emotional Experience of Contest and Victory Celebration: Spectators and Celebrants in Pindar’s Epinicians,” in X. Riu et al. (eds.), Approaches to Archaic Greek Poetry (Messina 2012) 173–219, at 204– 210; Adorjáni, Pindars sechste olympische Siegesode; M. Foster, “Hagesias as synoikister: Seercraft and Colonial Ideology in Pindar’s Sixth Olympian Ode,” ClAnt 32 (2013) 283–321, and The Seer and the City: Religion, Politics, and Colonial Ideology in Ancient Greece (Berkeley 2017); Spadafora, in Pindar in Sicily 225–230. 63 Cf. Adorjáni, Pindars sechste olympische Siegesode 56–66; Bonifazi, Mescolare un cratere 143–149; D’Alessio, Arethusa 37 (2004) 271; Calame, in Reading the Victory Ode 303–320. PETER AGÓCS 407 ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 messenger, a message-stick of the Muses with their gorgeous tresses, a sweet mixing-bowl of loud-voiced songs. Tell them to remember Syracuse … One peculiarity of Ol. 6 is its use of helpers to assist the real and metaphorical movements of song and singer. Aineas (un- mentioned before) apparently leads the choral group (komos, cf. 98) that will carry the song from Stymphalos to Syracuse. Perhaps he is an Arcadian relative of Hagesias.64 The scholiast identifies him as a χοροδιδάσκαλος charged with interpreting Pindar’s intentions to the komos.65 “Pindar” instructs Aineas about the themes on which the komos, as they compose their song in performance, must touch, cycling as in a recusatio through a catalogue of topics enacted as soon as they are mentioned. The mention of the well-attested Stymphalian Hera-cult sets the scene there, while the latter comment seems to prefigure future reception in Syracuse.66 The rest of the komos’ epinician program is laid out subsequently. First, however, we have Aineas: a “true messenger,” a “mes- sage-stick of the Muses with their gorgeous tresses,” and a “sweet krater of loud-voiced songs.” The messenger implies 64 The name is attested both for Stymphalos (Xen. Hell. 7.3.1, Anab. 4.7.13) and within the lineage of the Iamidai (Paus. 6.2.4, 8.10.5): see S. Hornblower, Thucydides and Pindar: Historical Narrative and the World of Epinikian Poetry (Oxford 2004) 183–184. Cf. Adorjáni, Pindars sechste olympische Siegesode 49–50 nn.68–69; G. O. Hutchinson, Greek Lyric Poetry (Oxford 2001) 413– 415. 65 Schol. Ol. 6.148a, cf. 149a: the scholiast adds that Pindar employed Aineas “because he himself had a weak voice.” On Aineas see Adorjáni, Pindars sechste olympische Siegesode 48–51; Bonifazi, Mescolare un cratere 133–143; A. Uhlig, Script and Song in Pindar and Aeschylus (diss. Princeton 2011) 78–80. 66 On Hera Parthenia (a debated topic: is Pindar referring here to one song or two?) see Giannini in Gentili, Olimpiche 468; Adorjáni, Pindars sechste olympische Siegesode 278–279. Following Y. L. Too, “Hera Parthenia and Poetic Self-reference in Pindar, Olympian 6.87–90,” Hermes 119 (1991) 257– 264, I assume that Ol. 6 includes the “Hymn to Hera” by mentioning it. 408 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 oral transmission; σκυτάλα Μοισᾶν then suddenly shifts every- thing to writing. The scholia see reference to the σκυτάλη Λακωνική (“message-baton”).67 Skytale and “honest messenger” both allude to an authoritative medium and dismiss concerns about the impact of corruption, interpolation, or theft on the transmitted text. As the message-sender relies on the mes- senger, and the messenger on his voice, so the entextualized voice relies for its articulation on a human proxy. The Muses and the bellowing bowl express this oral element. Aineas’ description as a “krater of loud-voiced songs” contains a promise of musical intoxication and reestablishes our komastic and sympotic focus, while the idea of “mixing” attests a mastery of song-themes, forms, and myths. The figures of poet and inter- mediary momentarily merge, just as the voices of poet, Aineas, and komos overlap.68 Inscribed in Aineas’ memory, the poet’s text remains, figuratively at least, an oral object; while Aineas’ name, inscribed into the victory ode, records its “successful transmission.”69 But it is not Aineas who performs the ode on arrival. Rather, he assumes certain functions of a text, and perhaps even im- personates one. The evidence for fifth-century choral training suggests that it involved dictation (ὑπολέγειν) and repetition, and had much in common with contemporary music peda- gogy, as the trainer’s recital of text and melody fixed itself in the singers’ minds.70 This ambiguity (text plus recitation; oral 67 For exhaustive bibliography on the skytale-question: Adorjáni, Pindars sechste olympische Siegesode 285–288. See S. R. West, “Archilochus’ Message- stick,” CQ 38 (1988) 42–48; T. Kelly, “The Myth of the Skytale,” Cryptologia 22 (1998) 244–260. My view follows D. Steiner, The Tyrant’s Writ: Myths and Images of Writing in Ancient Greece (Princeton 1994) 114–115. 68 Calame, in Reading the Victory Ode 309–311. 69 M. J. Schmid, “Skytála Moisân: Song and Writing in Pindar,” Minerva 12 (1998) 57–81, at 78. 70 Plut. De audiendo 46B; Herington, Poetry into Drama 24, 30; Prauscello, Singing Alexandria 44–48; E. Hall, The Theatrical Cast of Athens. Interactions be- PETER AGÓCS 409 ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 and written mnemosyne working together) is reflected in Aineas’ poetic portrait and the descriptions applied to the ode itself. The word komos can mean a song in performance, an occa- sion, and a type of performing group.71 The following verses (92–100) hint at the journey that Aineas, the komos, and the ode must undergo. Pindar tells Aineas to tell the komasts to “re- member” (that is: “mention”) Syracuse and Ortygia (92). Men- tion of Hieron follows (93–97): “May future time not crash his [Hieron’s] happiness; and may he welcome with lovely acts of friendship the komos of Hagesias, as it returns from home to home journeying on from Stymphalian walls, leaving behind the mother of Arcadia rich in flocks” (97–100). Hieron’s intru- sion makes us ask who the real recipient of this ode is. Aineas thus mediates not only between text and komos, but between Hagesias and Hieron. It is Hieron who will receive and sanc- tion the komos brought to him by the victor after a journey from Stymphalos to Syracuse that retraces in reverse the ode’s initial fantastic voyage from Syracuse to Laconia (22–28). Ol. 6 too ends before it begins. If it is unlikely that, as the scholiast to Pyth. 2 suggests, Pindar sent his odes by means of choruses, one can see how this passage could have supported that view. The final triad of Ol. 6 is virtually our only internal evidence for the commissioning process of epinician lyric from composition through chorus training to performance. Though riddled with fictional and metaphorical elements, it reflects the tensions be- tween sending and performance seen in all “sent” songs. 5. Isthmian 2: reading and reenactment These tensions between sending and performance are even more prominent in the closing lines of Isth. 2 (43–48), an ode addressed to Thrasyboulos, nephew of the Emmenid tyrant Theron of Akragas, where the speaker calls on Nikasippos, his ___ tween Ancient Greek Drama and Society (Oxford 2006) 39–48. 71 Agócs, in Reading the Victory Ode 198–201, with bibliography. 410 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 proxy, to “read out” his message to the patron. The scholiast summarizes the problem: “This epinician is for Xenocrates of Acragas and is sent to Thrasyboulos via a certain Nikasippos; hence it is an ἀποστολικός.”72 It commemorates Xenocrates— Thrasyboulos’ father, Theron’s brother, and Hieron’s brother- in-law, now deceased—by recalling old victories won by a charioteer named Nikomachos. The Emmenids’ houses, Pindar says, are “not unfamiliar with lovely komoi, nor with songs that boast like honey” (30–32). Until the final epode (43–48), how- ever, the ode is addressed to Thrasyboulos. Then, in a sudden seemingly unmotivated apostrophe, Pindar addresses Nikasip- pos. Like Aineas in Ol. 6, Nikasippos has been described as a chorodidaskalos or as the man who performed Pindar’s ode solo in Akragas.73 But internal evidence tells us nothing about Isth. 2’s first performance or Nikasippos himself. His role is as vague as Aineas’ in Ol. 6. Both are best conceived as textual figures for the ode’s mediated nature and the poet’s absent presence. As the sudden appearance of Aineas in Ol. 6 is surprising, so in Isth. 2 is Nikasippos’ epiphany (43–48): µή νυν, ὅτι φθονεραὶ θνατῶν φρένας ἀµφικρέµανται ἐλπίδες, µήτ’ ἀρετάν ποτε σιγάτω πατρῴαν, µηδὲ τούσδ’ ὕµνους· ἐπεί τοι 45 οὐκ ἐλινύσοντας αὐτοὺς ἐργασάµαν. ταῦτα, Νικάσιππ’, ἀπόνειµον, ὅταν ξεῖνον ἐµὸν ἠθαῖον ἐλθῃς. 72 Schol. Isth. 2.inscr.a. On this ode see Tedeschi, StIt 3 (1985) 46–47; Spadafora, in Pindar in Sicily 224–225; D’Alessio, AION 42 (2020) 16–18. 73 G. A. Privitera, Pindaro: Le Istmiche (Milan 1982) 165–166. For the “solo” performance hypothesis, see C. Catenacci, “Ἀπονέµειν/‘leggere’: Pind. Isthm. 2, 47; Soph. fr. 144 Radt; Aristoph. Av. 1289,” QUCC N.S. 62 (1999) 49–61; F. Ferrari, “Intorno all’ombelico del mondo: le prospettive del rito nelle Pitiche di Pindaro,” SemRom 2 (2000) 217–242, at 232 n.60. E. M. Stehle, “The Construction of Authority in Pindar’s Isthmian 2 in Perfor- mance,” in E. J. Bakker (ed.), Authorship and Greek Song: Authority, Authenticity, and Performance (Leiden 2017) 8–33, argues for a choral voice. PETER AGÓCS 411 ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 May he never, then, because mortal minds are overshadowed by envious expectation, silence either his father’s glory nor these very songs, for I did not craft them to stand uselessly. These words, Nikasippos, read out when you to my honoured friend arrive. Pindar, till now performing his friendship with Thrasyboulos in a lyric mode of directness and familiarity, shifts abruptly to the third person (σιγάτω, 44), withdrawing both speaker and ut- terance from a frame in which, until now, they have carefully anchored themselves.74 Nothing explains it until the address to Nikasippos at the end (47–48). Much depends on the meaning of his command to Nikasip- pos: ταῦτα ἀπόνειµον (47). The preceding verses (43–46) are “user instructions” directed through him to a suddenly-distant Thrasyboulos. But is it just these he must deliver, or the ode’s entire text? The second interpretation gains plausibility from the self-referential habits of the closural sphragis and the likely meaning of ἀπόνειµον, whose basic sense (cf. LSJ s.v.) is fair ap- portionment: it might thus mean “impart these words whenever you arrive in Akragas.” Taken iteratively, it could imply the song’s diffusion of the patron’s kleos. But a fragment of Sopho- kles scholia (fr.144 Radt) suggests that ἀπονέµειν might mean “read” in the sense of voiced reading before a group.75 In Isth. 2, voiced (rather than silent) reading thus arguably overwrites performance. Song, reconfiguring itself as recitation, begins to behave like a true poetic epistle enacted by reading out. In fact, 74 Privitera, Istmiche 31. 75 Schol. Isth. 2.68. Ferrari, SemRom 2 (2000) 232; Catenacci, QUCC N.S. 62 (1999) 57–59; J. Svenbro, “La lecture à haute voix. Le témoignage des verbes grecs signifiant ‘lire’,” in C. Baurain et al. (eds.), Phoinikeia Grammata. Lire et écrire en Méditerranée (Namur 1991) 539–548, at 540–545, and Phrasikleia 35 n.47, 109–116; Steiner, The Tyrant’s Writ 123–124; P. Chaintraine, “Les verbes grecs signifiant ‘lire’,” Παγκάρπεια. Mélanges Henri Grégoire II (Brussels 1950) 115–126. 412 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 of course, it is open both to voiced reading and to solo or choral performance. Pindar’s instructions for use addressed to Thrasyboulos (43– 46) align Isth. 2 with the traditional values of kleos-song. “There- fore, because envious hopes hang around the minds of mortals, let him (sc. Thrasyboulos) not keep his paternal arete in silence, nor these songs (τούσδ’ ὕµνους, 45), for I did not make them to stand around idly.” “These songs” can refer only to Isth. 2—in which case the performance of the ode is again projected into the future.76 But one wonders if they do not also include all the songs composed over about two decades by Pindar for the Emmenidai.77 The tense of ἐργασάµαν would then attest the completion not of one ode, but an entire oeuvre. In that case, CT and RT are again contrasted within the utterance, with the latter projected outside the frame, this time in reference to a corpus of praise poetry intended for systematic re-use.78 Apart from a sense of personal involvement conveyed by the first person and τοι, the lines evoke the familiar comparison of poems to art-objects, especially statues, found not least in the Nem. 5 proem. There, the speaker rejects the sculptor’s art on the grounds that statues “standing idly on their single bases” cannot spread word of the victor’s kleos. In Isth. 2, he wears this artisan’s badge proudly. His ὕµνοι have the durability of finely- crafted heirlooms, and can project the oikos’ fame.79 They 76 D’Alessio, Arethusa 37 (2004) 278. 77 L. Athanassaki, “Performance and Re-performance: The Siphnian Treasury Evoked (Pindar’s Pythian 6, Olympian 2 and Isthmian 2),” in Reading the Victory Ode 134–157, at 154–157. Cf. Athanassaki, in Approaches to Archaic Greek Poetry 197–202; also D’Alessio, Arethusa 37 (2004) 286, and Spelman, Poetics of Permanence 65–66. 78 Cf. Spelman, Poetics of Permanence 275–276. 79 Kurke, The Traffic in Praise. On the role of monodic reperformance in Isth. 2 see also T. Phillips, “Pindar’s Voices: Music, Ethics and Reperfor- mance,” JHS 137 (2017) 142–162. PETER AGÓCS 413 ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 cannot, however, be allowed to exist only as fixed ἀγάλµατα gathering dust in a sanctuary or a library. Pindar places Thrasyboulos, the guardian of the Emmenid song-hoard, on a level with the men who brought those songs into being with victories and patronage. He is responsible for keeping their memory alive. This ode, the last of Pindar’s Emmenid epinicians, thus memorializes the poet’s friendship with his patrons. 6. Conclusion: paradoxes of the written voice The sending motif ’s characteristic features—those constant distortions of spatial and temporal deixis, the paradoxical emplotment of song’s relationship to performance, the use of mediator-figures who behave like texts, and the use of song- texts as sources of social capital and memory—are all connected to the encomiastic function of the poetry and to its claims of social power. The theme is associated with proems and endings, privileged sites of self-referential commentary in choral song. It conveys the ode’s mediated nature and calls attention to authorship.80 Choral songs from Alcman onwards tend to describe their authors in the third person, often heightening this indirectness with kennings like Bacchylides’ “famous servant of Ourania” (5.13–14), or references to places (e.g. Thebes or Keos) associated with the historical poet. The sending motif reverses this familiar device, displacing the deictic origo of the utterance from what might be described as “chorus-time” to “poet-time,” while generating spatial distance as well.81 This allows the poet’s voice to separate itself at least partly from that of the performer(s). “Sending” thus can lend the encomiastic voice a more authorial tone. As something that can be separated from its creator and “sent,” it is also liable to 80 Cf. Schmid, Minerva 12 (1998) 77–78; D’Alessio, AION 42 (2020) 1–30. 81 Cf. C. Calame, The Craft of Poetic Speech in Ancient Greece (Ithaca 1995) 21–25, 49–51, 54; D’Alessio, AION 42 (2020) 1–30. 414 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 persist unchanged as a “script” underpinning future perfor- mances.82 As our examples all demonstrated, reading, private recitation, and solo or choral (re)performance are for Pindar potentially equivalent things: parallel ways of (re)using text that are able to restore it to its natural motion, vitality, and voice. As a framing device, lyric epistolarity combines the beautiful artefact with the agility of sung fame, complementing praise- song’s repertory of imagery derived from various objects—from Lydian headbands to golden cups, victor-statues, and inscribed stelai to the portico of a temple—that expresses its exalted view of its own materiality, craftsmanship, permanence, and value.83 But unlike these archetypes of artifice and craft, the sending theme evokes the dual nature of entextualized song. “Sent,” the song is an object; “delivered,” it becomes a performance. The framing effects of lyric epistolarity are not unlike those frequently created in the epigraphic corpus of early speaking- object inscriptions (oggetti parlanti). As Svenbro and Day have illustrated, early (7th–5th-cent.) dedicatory and funerary epi- grams often position themselves in relation to their readers by referring to the person behind the inscription in the third person, even as the object interpellates the reader in the first: “Mantiklos dedicated me.”84 Sometimes this speaking “I” addresses us without reference to an author, ordering the passerby to stop and look, or to pause, remember, and weep.85 82 On the notions of “script” and “scriptory poetics” see Uhlig, Script and Song. 83 Cf. e.g. Nem. 7.77–79 and 8.14–15. Ford, Origins of Criticism 93–130. Cf. Steiner, HSCP 95 (1993) 159–161, and The Tyrant’s Writ 91–99. 84 CEG 326, quoted here, on a statuette of Apollo in the MFA in Boston (inv. 03.997), is a particularly rich early (late 7th-cent.) example of the type: see Day, Archaic Greek Epigram and Dedication 33–48. 85 Steiner, The Tyrant’s Writ 91–99; Svenbro, Phrasikleia 29–43 (noting that after 550 the “egocentric” element is diminished but the temporal displace- ment unchanged). Cf. Young, HSCP 87 (1983) 35–40, and Uhlig, Script 16. PETER AGÓCS 415 ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 In their framing and grammar, these epigrams are thus similar to the sending motif in epinician: they distance their deictic center/origo from the moment of their own conception and inscription (CT), in order to accommodate themselves to the time of their future readers (RT).86 Indeed, the poets were quite familiar with epigraphic language.87 In both the inscribed text and the choral performance, the absent composer (who in the case of epigram is not to be identified with the inscriber) depends for his presence and vocality on his living, breathing proxy: the reader or performer. Similar protocols of distancing and fram- ing are apparent in sixth- and fifth-century letters, and in the proemic syntax of early prose treatises down to the time of Herodotus and Thucydides.88 Early Greek writing tends naturally to inhabit the future reader’s time-frame rather than the author’s, and early texts frame themselves as written ut- terance reanimated by a living voice. This sense of vocality remained basic to Greek thinking about literacy. Ultimately, writing remained a technology of vocal reproduction. In short, the sending motif is a trope of partial textuality that extends and intensifies the normal pragmatic relations implied in any text composed for choral performance by a professional poet, and reflects the ode’s existence as a material, en- textualized object within the self-imposed limitations of a poetic style rooted in the song culture. In the end, it tells us little about 86 See n.26 above. 87 Cf. Nem. 8.46–48, Pyth. 6.1–17, Isth. 8.61–65, with Day, JHS (1989) 22– 24. On Pindar’s engagement with epigraphic culture: Day 16–28, and Archaic Greek Epigram and Dedication 62–63; Steiner, HSCP 95 (1993) 167–178; R. Thomas, “Fame, Memorial, and Choral Poetry: The Origins of Epi- nikian Poetry – An Historical Study,” in S. Hornblower et al. (eds.), Pindar’s Poetry, Patrons, and Festivals (Oxford 2007) 141–166, at 151–152. 88 See Svenbro, Phrasikleia 149–150; D’Alessio, Arethusa 37 (2004) 290 n.83; Ceccarelli, Ancient Greek Letter Writing 9–10; Day Archaic Greek Epigram and Dedication 112–120. 416 MESSAGE-STICK OF THE MUSES ————— Greek, Roman, and Byzantine Studies 62 (2022) 385–416 the process of creation, composition, sending, and rehearsal through which any given ode passed, about how and when writing entered that process, or whether Pindar sent his patrons a chorus, a courier, a wax tablet, or a book.89 Instead, this motif calls attention to the verbal object behind the vocalised utterance, and to the composing voice whose utterance it con- tains and enacts. Writing, for Bacchylides and Pindar, was a way of stabilizing the voice—the mimetic presence—of a song. Readers, like Nikasippos, or performers like the men of Aristo- kleidas’ or Aineas’ komoi, lent their voices to the text. When enacted, the performance-text became a second-order mimesis: it imposed one voice on another, forcing the singer or reader to assume a role. The entextualized speakers inhabit these odes even as the performers (and eventual readers) inhabit the entextualized roles they “script.”90 This amounts to a hitherto unrecognized poetics of the entextualized song: as such, the sending motif should take its place among contemporary discussions of “materiality” and “affect” which have helped refocus critical interest on the physical properties and poetics of text.91 November, 2022 Dept. of Greek and Latin University College London p.agocs@ucl.ac.uk 89 Schmid, Minerva 12 (1998) 70. 90 Calame, Craft of Poetic Speech 20–24, 48–52; Steiner, HSCP 95 (1993) 179. 91 This paper was written as part of the Hungarian National Research Development and Innovation Office “FK_18” research grant no. 128492. I thank Jess Lightfoot as well as Lucia Athanassaki, Árpád Miklós Nagy, Ábel Tamás, Kata Pártay, Judit Beszkid, my GRBS editors and anonymous readers, and numerous others who read drafts.