Eckerman ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 2012 Christopher C. Eckerman Was Epinician Poetry Performed at Panhellenic Sanctuaries? Christopher C. Eckerman N 1985 Thomas Gelzer, following Ulrich von Wilamowitz- Moellendorff, Herwig Maehler, and others,1 published a substantial and influential article, which argued that epi- nician poems were first performed either at the sanctuary at which the respective victory was won, immediately after the victory, or at the home polis of the victor, shortly after the victor returned home.2 The great majority of scholars have fol- lowed Gelzer in suggesting two alternate performance contexts for epinician odes. David Campbell, for example, comments on Bacchylides 2, “this brief song seems to have been performed at the Isthmus, the elaborate poem 1 at a later celebration on Ceos.”3 Campbell’s note displays the common assumption that, when a longer and a shorter ode for the same victory are pre- served, then the shorter was composed for on-site performance while the longer was composed for performance in the home of the victor. Similarly Neumann-Hartmann assumes that certain epinician odes were performed on site and devotes a chapter of 1 Cf. H. Maehler, Die Lieder des Bakchylides (Leiden/New York 1982) I.2 10, 30, with further bibliography. 2 T. Gelzer, “Μοῦσα αὐθιγενής: Bemerkungen zu einem Typ Pin- darischer und Bacchylideischer Epinikien,” MusHelv 42 (1985: hereafter ‘Gelzer’) 95–120: “So gibt es … schliesslich zwei Typen von Siegesliedern im engeren Sinn. Davon sind die einen für ein spezifisches Siegesfest ge- dichtet, das zu Ehren des Siegers nach seiner Rückkehr vom Wettkampfort in seiner Heimat gefeiert wird, die andern dagegen zur Darbietung bei einer Siegesfeier am Wettkampfort selber, vor der Heimkehr des Siegers in seine Heimat, bestimmt” (96–97). 3 D. A. Campbell, Greek Lyric IV (Cambridge [Mass.] 1992) 125. I CHRISTOPHER C. ECKERMAN 339 ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 her recent monograph to such odes.4 The assumption of on-site performance, which has become entrenched, deserves a rejoinder in favor of greater skepticism.5 No one has reassessed the data marshaled in detail by Gelzer, however, to query how much evidence the epinician odes really offer for performance at the sanctuaries. Here I reexamine the relevant passages and argue that there is no secure evidence for the performance of any of these odes at Panhellenic sanc- tuaries. My goal is not to depreciate the importance of song culture at victory celebrations on site, since we know that song was an important part of such celebrations. However, we should not assert that some odes were first performed at the Panhellenic sanctuaries and that others were first performed in the home poleis of the victors, since the poems themselves provide no evidence for their performance at the sanctuaries. Consequently, no deductions can be made regarding per- formance location based on the length of any ode: prima facie there is no reason to suggest that shorter odes were regularly performed at the sanctuary and longer odes were regularly per- formed in the home poleis of the victors.6 This paper, then, situates itself among the literature on deixis that Nancy Felson and others inaugurated in relation to lyric poetry.7 By focusing on ad oculos deixis, I hope to make a contribution toward re- covering the specific context in which epinician poetry was first 4 A. Neumann-Hartmann, Epinikien und ihr Aufführungsrahmen (Hildesheim 2009), ch. 3. 5 Although it is rare for scholars to be critical of Gelzer’s arguments, William Race notes, “It is often claimed that shorter epinikia … were im- provisations performed at the site of the victory … but there is no conclusive evidence for such assumptions”: Pindar: Olympian Odes, Pythian Odes (Cam- bridge [Mass.] 1997) 15. 6 Contra e.g. Gelzer 99, “Diese kleinen Formen sind also für diesen Typ üblich.” 7 See The Poetics of Deixis in Alcman, Pindar, and Other Lyric = Arethusa 37 (2004) 253–466. 340 EPINICIAN POETRY AT PANHELLENIC SANCTUARIES ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 performed.8 As both Gelzer and Neumann-Hartmann note, there is no external evidence for on-site performance of epinician poetry, and the evidence accordingly would have to derive from the odes themselves.9 This may, however, not be surprising since we have very little evidence regarding epinician performance in general.10 Working only with Pindar’s and Bacchylides’ extant odes, then, Gelzer suggests that eight of these were composed for on-site performance, and he provides various reasons, based on text-internal references, for his argument.11 As I shall show, however, Gelzer overlooks several passages, within these eight odes, that argue against his thesis, and all the passages that he uses as evidence for on-site performance are problematic. It is best to begin with a reexamination of the eight odes that 8 Felson explains ad oculos deixis by saying, “Ocular deictics in perfor- mance poetry point extra-textually to objects or properties surrounding the discourse and visible in the extra-linguistic context of the utterance”: Are- thusa 37 (2004) 259. 9 Gelzer 97 n.6: “Zur Entscheidung, ob ein Lied am Wettkampfort oder anderswo aufgeführt worden sei, ist man ganz auf die Hinweise im Text selber angewiesen.” Neumann-Hartmann, Epinikien 69: “Um in Erfahrung zu bringen, wo und wann Epinikien am Wettkampfort zur Aufführung gelangten, müssen wir daher von den Siegesliedern ausgehen, die einen Vortrag am Festspielort nahe legen.” 10 For example, it is not clear whether odes were always performed by a soloist or by a chorus, whether odes were performed at the home of the victor or at a public venue, whether some odes were processional, etc. On the paucity of external evidence, see too J. Herington, Poetry into Drama: Early Tragedy and the Greek Poetic Tradition (Berkeley 1985) 27–28. 11 Gelzer 97: “Alle acht im folgenden genannten Lieder enthalten Erwäh- nungen, die als Argument für Aufführung am Wettkampfort interpretiert werden können, nämlich die Nennung von Ort und Gott des Wettkampfs und Umständen der Herkunft der Ehre des Sieges wie Wettkampf, Bekrän- zung und Feier des Siegers und Neuheit seines Glücks, in folgenden Versen: O. 4, 1–5. 5–12; O. 11, 7f. 12–14; P. 6, 3–9. 16–19; P. 7, 4. 10–12; N. 2, 1–5. 14f. 24f.; Ep. 2, 3–7.11–14; Ep. 4, 1–6. 14–16; Ep. 6, 1–3. 6. 14–16.” Cf. M. Hose, “Bemerkungen zum 4. Epinikion des Bakchylides,” in A. Bagordo and B. Zimmermann (eds.), Bakchylides: 100 Jahre nach seiner Wiederentdeckung (Munich 2000) 161–168, at 162. CHRISTOPHER C. ECKERMAN 341 ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 are frequently used in support of the case for on-site perfor- mance. Like Bacchylides 1, Bacchylides 2 was composed for Argeius of Ceos, and it contains two passages deemed highly supportive of the on-site hypothesis.12 The first relates to the lacuna at the very beginning of the ode, while the second re- volves around the phrase Μοῦσ’ αὐθιγενής in line 11. The brief poem may be cited in full.13 ἄ[. . . . .] σεµνοδότειρα Φήµα ἐς Κ[έον ἱ]εράν, χαριτώ- νυµ[ον] φέρουσ’ ἀγγελίαν, ὅτι µ[ά]χας θρασύχειρ<ος> Ἀρ- 5 γεῖο[ς ἄ]ρατο νίκαν, καλῶν δ’ ἀνέµνασεν ὅσ’ ἐν κλε[εν]νῷ αὐχένι Ἰσθµοῦ ζαθέαν λιπόντες Εὐξαντίδα νᾶ- σον ἐπεδείξαµεν ἑβδοµή- 10 κοντα [σὺ]ν στεφάνοισιν. καλεῖ δὲ Μοῦσ’ αὐθιγενὴς γλυκεῖαν αὐλῶν καναχάν, γεραίρουσ’ ἐπινικίοις Πανθείδα φίλον υἱόν. Report, giver of majesty, [sped/speed (?)] to holy Ceos, carrying the message of gracious name, that Argeius won the victory in the bold-handed fight and reminded us of all the fine achieve- ments we had displayed at the famous neck of the Isthmus when we left the sacred island of Euxantius and won seventy garlands; and the authigenes Muse summons the sweet skirl of the pipes as she honors with victory-songs the dear son of Pantheides. An interpretation of a performance context for this ode may be 12 See e.g. Maehler, Die Lieder des Bakchylides I.2 10: “Das kurze Lied B. 2, das sicher für die Aufführung am Isthmos gleich nach der Siegerehrung be- stimmt war…, zeigt einen einfachen, dreiteiligen Aufbau.” 13 For text of Bacchylides, I follow Campbell’s Loeb edition (Greek Lyric IV: 1992) and base my translations closely on those of Campbell. For text of Pindar, I follow Race’s Loeb edition (1997) and base my translations closely on his. 342 EPINICIAN POETRY AT PANHELLENIC SANCTUARIES ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 affected by how we supplement the lacuna at the beginning of the poem. Many scholars, including Schroeder, Taccone, Wila- mowitz, Steinkopf, and Maehler, fill the lacuna with an im- perative, ἄ[ϊξον], and interpret the opening to be a command, during performance at the Isthmus, to Φήµα, to speed from the Isthmus to Ceos to spread news of the victory.14 This is the interpretation of Gelzer too, and when the passage is taken thus, it may provide evidence that this ode was not first per- formed at Ceos; scholars then deduce that the ode must have first been performed at the Isthmus, the site of victory.15 I suggest an alternative interpretation that, on the one hand, retains the imperative ἄϊξον but, on the other, positions the first performance not at the Isthmus but at Ceos. Assuming for the moment that Bacchylides did use the imperative in the lacuna, ἄϊξον, ὦ σεµνοδότειρα Φήµα, ἐς Κέον ἱεράν (“speed, Report that gives majesty, to holy Ceos”), nonetheless it is not clear whence the performers’ voices arise. Is the chorus at the Isthmus or at Ceos? It may be plausible, perhaps even prefer- able, to assume that this is an invocation and that a chorus on Ceos, not at the Isthmus, calls on Φήµα to come to Ceos with news of the victory. Given Pindaric and Bacchylidean com- paranda, it becomes more likely that the phrase “speed to Ceos” is an invocation, rather than an exhortation to flight. There are many such invocations where a divinity is called upon to be present at an epinician celebration, but there are no passages where the chorus urges a divinity to leave its side. As Maehler notes, “Mehrere Siegeslieder beginnen mit einem Anruf an Göttinnen wie Phema (B. 10), Nika (B. 11), Hora (Pind.N.8), auch an die Ἁµέρα, die sehr wahrscheinlich in B. 7.2 zu ergänzen ist.”16 Although Bacch. 10 also begins with an invocation of Φήµα, there is no reason to suggest that that ode 14 Cf. Maehler, Die Lieder des Bakchylides I.2 28, and R. Jebb, Bacchylides. The Poems and Fragments (Cambridge 1905) 251. 15 Cf. T. Hubbard, “The Dissemination of Epinician Lyric,” in C. J. Mackie (ed.), Oral Performance and its Context (Leiden 2004) 71–93, at 77. 16 Die Lieder des Bakchylides I.2 28. CHRISTOPHER C. ECKERMAN 343 ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 was performed on site. The invocation of Φήµα in Bacch. 2, then, may not derive from a need for Φήµα to leave the Panhellenic sanctuary and report news of the victory at Ceos; rather, the invocation of Φήµα may participate in a generic encomiastic topos: Panhellenic victories should be widely ‘reported’ and celebrated through epinician poetry, such as the currently performed ode.17 The invocation hypothesis explains how the chorus can use the imperative while the first performance occurs on Ceos, not at the Isthmus.18 Accordingly, if the chorus in Bacch. 2 beckons Φήµα to hurry to Ceos, it would be a powerful dramatic con- ceit that enlivens performance at Ceos, with Φήµα bringing news to the home polis of the victor from the site of victory. And through Bacchylides’ ode, Φήµα does arrive and bring news of the victory, for, as Hubbard notes in relation to Bacch. 10, Bacchylides’ poem is an agent of Φήµα.19 Alternatively, the lacuna may be supplemented with an aorist indicative. Gelzer does not notify his readers that the indicative is an important possibility to consider, although many Hel- lenists, including Levi, Blass, Festa, and Jebb, preferred this over the imperative.20 If the lacuna did contain the aorist in- dicative, Φήµα has already sped from the Isthmus to Ceos, where Bacchylides’ ode is being performed, and this would mirror reality on Ceos, since the people at the celebration know they are there to celebrate Argeius’ victory. The opening of Bacch. 2, then, need not provide any evi- dence in favor of on-site performance whether the imperative or the indicative is chosen as a supplement. Maehler provides 17 This problematizes Gelzer’s thesis that the ‘Botschaft’ and ‘Zeugnis’ motifs are peculiar to the ‘on site’ category of poems. See Gelzer 99–101, and Hose, in Bakchylides 161–162. 18 Contra Jebb, Bacchylides 251, who claims, “If ἄιξον, ὦ were read, it would imply that the poet himself was at the Isthmus.” 19 Hubbard, in Oral Performance 73. 20 Gelzer 99; for discussion see Maehler, Die Lieder des Bakchylides I.2 28. 344 EPINICIAN POETRY AT PANHELLENIC SANCTUARIES ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 numerous comparanda that lead him to favor the imperative,21 and, given those comparanda, I find it most likely that Bac- chylides did use the imperative rather than the indicative in the lacuna. Furthermore, given the epinician comparanda in favor of invocation rather than exhortation to flight, I find it most plausible that a chorus on Ceos invokes Φήµα to be present at the victory celebration; nonetheless, the aorist indicative should not be ruled out. Μοῦσ’ αὐθιγενής in line 11 is also a topic of concern and has been taken to provide substantial evidence for the on-site hypothesis. αὐθιγενής, according to LSJ, means ‘born on the spot’, ‘born in the country’, or ‘native’. As a compound ad- jective formed from the deictic adverb αὖθι and the verb γίγνοµαι, the semantics of the adjective are unproblematic, meaning ‘born right here’. Thus Μοῦσ’ αὐθιγενής means ‘Muse that is born right here’. For the original audience, which knows whether the performance is occurring at the Isthmus or at Ceos, the adjective is unproblematic. The important point to stress, however, is that ‘born right here’ is not a declaration of performance at the Isthmus any more than it is a declaration of performance at Ceos. For some scholars, including Wilamo- witz, Maehler, and Gelzer, however, the phrase refers to a Muse that is ‘born on the spot [of victory]’ and accordingly the adjective refers to on-site performance.22 However, ‘of victory’ or the like is a peculiar and unwarranted addition to the semantics of αὐθιγενής. Critics who do not favor on-site performance have assumed that the adjective means ‘born right here’, i.e. ‘kindred’, ‘native’—thus Jurenka, Festa, Taccone, and Jebb. With Μοῦσ’ αὐθιγενής, Bacchylides would refer, then, either to himself metaphorically as a Muse native to Ceos ( ‘born right here’) or 21 Die Lieder des Bakchylides I.2 28, 205–206. 22 Campbell, Greek Lyric IV 125, follows Gelzer but also considers the other possibility: the Muse is authigenes “since B. is composing at the site of the games; but perhaps ‘the compatriot Muse’ with reference to the Cean origin of B. and the victor.” CHRISTOPHER C. ECKERMAN 345 ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 to a Muse that Bacchylides has brought to life ‘on the spot’ during performance at Ceos. Bacchylides is fond of program- matic statements linking himself metaphorically with a Muse,23 and he may see himself here as the mouthpiece of the Muse and hence the Muse is ‘kindred’, reflecting that Cean Bac- chylides has close ties to the Cean victor specifically and Ceos generally. Bacchylides stresses his Cean identity in Bacch. 3 where he refers to himself as a “Cean nightingale” (Κηίας ἀηδόνος, 98) and in Bacch. 10 where he refers to himself as a “clear-voiced island bee” (να̣σιῶτιν … λιγύφθογγον µέλισσαν, 10). In epinician poetry, moreover, the end of a poem is frequently a place for programmatic statements where the poet asserts his preeminence,24 and Bacchylides’ stressing his close bond to the Muse would generically be more expected than a statement that asserts that a Muse was brought to life at the site of victory, which is otherwise unparalleled.25 Whether we in- terpret Μοῦσ’ αὐθιγενής to mean ‘Muse born on the spot’ or ‘native/kindred Muse’, the phrase provides no evidence for the on-site hypothesis. Like the first stanza of Bacch. 2, that of Bacch. 4 is often taken to offer strong evidence for on-site performance, spe- cifically in lines 4 and 5:26 ἔτι Συρακοσίαν φιλεῖ πόλιν ὁ χρυσοκόµας Ἀπόλλων, ἀστύθεµίν θ’ Ἱέ[ρω]να γεραίρει· τρίτον γὰρ π[αρ’ ὀµφα]λὸν ὑψιδείρου χθονὸς 5 Πυ[θ]ιόνικος ἀ[είδε]ται ὠ[κυ]πόδων ἀρ[ετᾷ] σὺν ἵππων. ἔ[λακε δ’] ἁδυεπὴς ἀ[να]- [ξιφόρ]µιγγος Οὐρ[αν]ίας ἀλέκτωρ 23 Cf. 3.98 (nightingale), 4.8 (cock), 5.10 (servant of Urania), 10.10 (bee). 24 E.g. Bacch. 3.96–98. 25 On programmatic statements in epinican poetry see I. Pfeijffer, “The Image of the Eagle in Pindar and Bacchylides,” CP 89 (1994) 305–317. 26 Cf. Maehler, Die Lieder des Bakchylides I.2 64, and Bacchylides: A Selection (Cambridge 2004) 100–101. 346 EPINICIAN POETRY AT PANHELLENIC SANCTUARIES ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 [ποτὲ µ]έν· ἀλλ’ ἑκ[όν]τι νόῳ 10 [νῦν νέο]υς ἐπέσεισ⌊εν⌋ ὕµνους. Golden-haired Apollo still loves the Syracusan city and honors its righteous ruler …[see below]… and the sweet-voiced cock of lyre-ruling Urania (cried out once before?) but (now) with willing mind he has showered on him (new) songs of praise. This stanza has been heavily reconstructed, and presumably we should not invest heavily in any dogmatic position that may de- rive from its interpretation. Regardless, Campbell’s translation may serve as a standard way of interpreting lines 4 and 5: Since for the third time he is hymned by the navel of the high- ridged land as a Pythian victor, thanks to the excellence of his swift-footed horses. The evidence for on-site performance has been written into the text with the provision of ἀ[είδε]ται as a supplement in line 5. ἀ[είδε]ται is not unquestionable as a supplement because multiple letters need to be provided and because it creates a performance context for epinician poetry that is not elsewhere clearly recognized in the epinician corpora. As Neumann- Hartmann too notes, the supplement is not a given.27 In short, it may be that ἀ[είδε]ται is not the correct supplement, and if it is not, then there is no evidence for on-site performance in this ode. Regardless, for the sake of scholarly dialogue, I shall take it as an attractive supplement and analyze the passage with it. Scholars generally take π[αρ’ ὀµφα]λὸν ὑψιδείρου χθονὸς with ἀ[είδε]ται, following a traditional syntactic rendering of the passage, but as S. R. Slings has shown, “Bacchylides’ style is still an oral style,” and a pragmatic approach to under- standing his language is often more helpful than a traditional grammatical approach.28 Bacchylides begins line 4 by seeming to assert that something has happened for a third time (the 27 Epinikien und ihr Aufführungsrahmen 38. 28 S. R. Slings, “Information Unit and Metrical Unit,” in I. L. Pfeijffer and S. R. Slings (eds.), One Hundred Years of Bacchylides (Amsterdam 1999) 61– 75. CHRISTOPHER C. ECKERMAN 347 ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 temporal denotation ‘third’ will only become apparent as the sentence unfolds). As the line continues, we learn that some- thing seems to have happened at the omphalos of the high- ridged land, namely Delphi. At the end of line 4 there is period end, and, pragmatically, the audience will construe meaning for this line before the next period begins. Given epinician norms, the audience will interpret this line, as an information unit, to mean that Hieron has now won his third victory at the Pythian games, since they know that the currently performed hymn celebrates his victory (470 B.C.) in the chariot-race at Delphi and that he won twice previously at Delphi in the horse race (482 and 478). In addition, para + dative or accusative is a formulaic way to name the place of victory in an epinician ode;29 word order and epinician norms, then, encourage the audience to link τρίτον with π[αρ’ ὀµφα]λὸν ὑψιδείρου χθονὸς as an information unit and to take π[αρ’ ὀµφα]λὸν ὑψιδείρου χθονὸς as a reference to the place of victory. Accordingly, the audience may supply a verb such as ‘he won’ at the end of line 4 to make sense of the preceding material in the period. Though it would not be unparalleled, it presumably would come as something of a surprise if in line 5, with a new metrical period, the audience has to realign their understanding of the previous information unit. A traditional syntactic rendering of the line, however, requires such a reinterpretation of the infor- mation unit in line 4. As line 5 develops, the victor is referred to as a Pythian victor and we learn that he is being sung. In fact, the seemingly enjambed Πυ[θ]ιόνικος, with its emphasis on winning, eases the grammatical difficulty of the preceding line where the audience members were, perhaps, encouraged to supply ‘won’ for themselves; at this point they may no longer need to supply a verb ‘won’ since that verbal idea is explicitly supplied in the verbal noun Πυ[θ]ιόνικος. However, as a new metrical unit and information unit begin, Πυ[θ]ιόνικος will also 29 Cf. W. J. Slater, Lexicon to Pindar (Berlin 1969) 411–412, and D. Gerber, Lexicon in Bacchylidem (Hildesheim 1984) 187–188. 348 EPINICIAN POETRY AT PANHELLENIC SANCTUARIES ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 be taken with ἀ[είδε]ται: Hieron not only won at Pytho (line 4), but he is also being sung as a Pythian victor (line 5). Ac- cordingly, we might modify Campbell’s traditional syntactic translation quoted above, and pragmatically translate as: For, for the third time, he is being sung [at e.g. Syracuse], for being a victor at Pytho, at the omphalos of the high-ridged land. More can be said in favor of rethinking the traditional inter- pretation of the syntax and semantics of lines 4 and 5. If we take the lines as postulated here, then the passage means that Hieron has won three times at Delphi and is now being celebrated in song. Alternatively, if we take the prepositional phrase π[αρ’ ὀµφα]λὸν ὑψιδείρου χθονὸς with ἀ[είδε]ται, the sentence asserts that Hieron is being sung for the third time at Pytho. Accordingly, Bacchylides would not be stressing the ac- complishments of Hieron, but rather would be commenting on the fact that Hieron had been sung at Delphi twice previously in addition to this third performance at Delphi. However, it would be outside epinician norms for Bacchylides to stress Hieron’s previous celebrations at Delphi rather than to stress his accomplishments at Delphi. Accordingly, based on metrical structure, pragmatic linguistics, and epinician topoi (i.e., the panegyrist’s job is to celebrate the victor, not the fact that the victor has been celebrated three times at Delphi in song), it may be best to take the prepositional phrase as separate from ἀ[είδε]ται; there would be, then, no evidence in this poem for on-site performance. If, however, we choose to retain the supplement ἀ[είδε]ται and do not choose to follow Slings and his interpretation of Bacchylides in terms of metrical units and sense units (feeling that this interpretation contains special pleading), but prefer a traditional interpretation of the grammar, following, for example, the translation of Campbell, ἀ[είδε]ται still need not refer to Bacchylides’ immediate epinician ode. Bacchylides may be saying that Hieron has in the past been sung at Delphi and continues to be sung at Delphi, without explicitly articu- lating what he has in mind. He may refer here to extem- poraneous singing, such as the so-called Archilochus song that CHRISTOPHER C. ECKERMAN 349 ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 was performed on site (cf. Ol. 9.1). Alternatively, the passage may refer to a reperformance scenario of epinician poetry at Delphi. That is to say, Hieron may be sung at Delphi with epinician odes that had been composed previously. There are, then, multiple ways to interpret ἀ[είδε]ται, and lines 4 and 5 need not refer to the immediate performance of Bacch. 4, even if we choose to accept the supplement. Finally, and most importantly, the Bacchylides papyrus itself preserves textual evidence for the performance of this ode in the home polis of the victor, Syracuse, and not on site. At line 14 the papyrus clearly reads παρεστίαν, which Maehler emends to πάρεστιν δ’ ἐν for metrical and syntactical reasons.30 Editors have objected to παρεστίαν because of a presumed problem in responsion between 4 τρίτον γὰρ παρ’ (⏑ ‒ ‒ ⏑) and 14 παρ’ ἑστίαν (⏑ ‒ ⏑ ‒), although, as Maehler points out, Bacchylides has the same anaclastic freedom of responsion in his aeolic base at 19.15. It has also been argued that the infini- tives in lines 16 (ἐρέπτειν) and 18 (ἀείδειν) must be dependent on a finite verb. For two reasons, then, Maehler prints πάρεστιν δ’ ἐν. As Carmine Catenacci and Marialuigia Di Marzio have argued, however, there is no need to emend the text for either the metrical or the syntactical reason: “In realtà nessuno dei due argomenti è decisivo.”31 I should stress that they did not preserve παρ’ ἑστίαν because they believe that the first performance of this ode occurred at Syracuse. Rather, they follow the opinio communis and assume that the ode was first performed on site because they do not problematize the supplement ἀ[είδε]ται in line 5. Their maintaining the reading preserved in the papyrus is independent of my interpretation here and is based on their own independent syntactical and metrical arguments. Their acceptance of the opinio communis for 30 See Maehler, Die Lieder des Bakchylides I.2 75, and C. Catenacci and M. Di Marzio, “Il gallo de Urania (Bacchilide, Epinicio 4),” QUCC 76 (2004) 71– 89, at 79, both with discussion of previous conjectures. 31 QUCC 76 (2004) 79. 350 EPINICIAN POETRY AT PANHELLENIC SANCTUARIES ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 the performance context of this ode places them in a difficult position because they have to assume that the phrase παρ’ ἑστίαν must refer to a victory celebration that is to come in the future at Syracuse rather than to an immediate performance at Syracuse. There is, however, no grammatical signal in the passage that suggests that παρ’ ἑστίαν should refer to a future performance rather than to the immediate performance. They are willing to accept this problem, however, because they are confident that the papyrus’ reading should be preserved. The interpretation put forth here, namely that παρ’ ἑστίαν refers to immediate performance at Syracuse, removes this problem from their independent arguments made for respecting the transmitted text and thereby strengthens their position.32 In short, the evidence provided by Bacch. 4 for a first per- formance context on site must be taken with greater skepticism than it has been. That evidence is provided via a questionable supplement that sits within a broader section of tattered papyrus. Moreover, via pragmatics, there is an alternative in- terpretation for the semantics of the phrase, and, furthermore, even if ἀ[είδε]ται is accepted and the pragmatic interpretation is disregarded, there is still multivalence, since it is not clear that Bacchylides refers to his own ode when he says that Hieron is sung at Delphi. Finally, a controversial emendation of a clear passage of Greek must be made to expunge evidence of performance at the home of the victor in line 14. I think that the supplement ἀ[είδε]ται is incorrect and that it has caused substantial confusion, leading scholars to favor the on-site hypothesis. Following Catenacci and Di Marzio, we should accept παρ’ ἑστίαν in our editions of Bacchylides and reckon that line 14 provides explicit testimony for Bacch. 4 being per- formed at Syracuse. In addition to Bacch. 2 and 4, Bacch. 6 is regularly con- 32 We need not take παρ’ ἑστίαν literally and assume that the perfor- mance occurs beside Hieron’s hearth, though a literal interpretation should not be ruled out. παρ’ ἑστίαν may mean no more than that the ode is per- formed in Syracuse, Hieron’s metaphorical hearth. CHRISTOPHER C. ECKERMAN 351 ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 sidered to provide strong evidence for the on-site hypothesis. Two passages (lines 6–8 and 14–15) are relevant to the perfor- mance context of this ode: Λάχων Δ∆ιὸς µεγίστου λάχε φέρτατον πόδεσσι κῦδος ἐπ’ Ἀλφεοῦ προχοαῖς [◡– –] δι’ ὅσσα πάροιθεν 5 ἀµπελοτρόφον Κέον ἄεισάν ποτ’ Ὀλυµπίᾳ πύξ τε καὶ στάδιον κρατεῦ[σαν] στεφάνοις ἐθείρας νεανίαι βρύοντες. 10 σὲ δὲ νῦν ἀναξιµόλπου Οὐρανίας ὕµνος ἕκατι Νίκ[ας,] Ἀριστοµένειον ὦ ποδάνεµον τέκος, γεραίρει προδόµοις ἀοι- 15 δαῖς ὅτι στάδιον κρατήσας Κέον εὐκλέϊξας. Lachon by the speed of his feet latched on to the highest glory from great Zeus at the mouth of Alpheus, (adding to the fine achievements?) for which in earlier days young men, their hair luxuriant with garlands, sang of vine-nurturing Ceos as the winner in sprint and boxing at Olympia; and now to you, wind- footed son of Aristomenes, thanks to Victory the hymn of song- ruling Urania gives praise in an ode sung before your house, since by winning the sprint you brought fame to Ceos. When Bacchylides mentions young men singing here, he ex- plicitly refers to former singing (6 ποτ’), and the passage, accordingly, makes no connection to his currently performed ode. We must decide, nonetheless, how to take the locative Ὀλυµπίᾳ in 6. Should it be connected with the verb ‘sung’ (ἄεισάν) or with the following participial phrase (πύξ τε καὶ στάδιον κρατεῦ[σαν])? The metrical unit and information unit do not end until κρατεῦ[σαν], and, accordingly, line 6 goes together as a unit, and Ὀλυµπίᾳ may be construed either with the verb or with the participle. The former interpretation sug- gests that young men sang at Olympia in the past, while the 352 EPINICIAN POETRY AT PANHELLENIC SANCTUARIES ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 latter does not. The grammatical ambiguity of the phrase is ex- hibited by Jebb’s prescription: “Join Ὀλυµπίᾳ with ἄεισαν, not with κρατεῦσαν. These tributes of song were paid by young men of Ceos at Olympia; the occasion would be a festal proces- sion, escorting the Cean victor to the temple of the Olympian Zeus, where he would give thanks; or it might be a banquet.”33 Whenever editors emphatically prescribe one interpretation over another, as Jebb does here, it tells us that the interpre- tation of the syntax is in doubt, or there would be no need for such a prescriptive gloss.34 Contra Jebb, Bacchylides may have intended his audience to link Ὀλυµπίᾳ with κρατεῦ[σαν], in which case the passage provides no evidence for young men singing at Olympia. If one does choose to link the locative with the main verb, however, it is important to note that the passage only asserts that Ceans have sung in the past at Olympia, and that that has nothing to do with the immediate performance context of this ode. Moreover, it is not clear that the young men are singing an epinician ode that has been composed for immediate performance on site. They may be singing an im- promptu song or they may be re-performing an epinician ode that had been composed on an earlier occasion. If we follow Bacchylides literally, they are not even singing praise for a specific victor but rather are praising Ceos itself. With this passage, then, Bacchylides brings into his ode a depiction of generic song culture at Olympia. The depiction of celebration and song at the Panhellenic festivals is, unsurprisingly, an epinician topos (cf. Ol. 5.4–7, 9.1–4, 10.73–77; Nem. 6.37–38), but it provides no evidence for the on-site performance of any of the individual odes. While the first stanza, then, provides no evidence for the performance of this song on site, the second stanza suggests that this ode may have been performed at Ceos. For example, McDevitt, without hesitation, avers on προδόµοις in line 15, 33 Jebb, Bacchylides 295. 34 See too Hubbard, in Oral Performance 78, on this passage. CHRISTOPHER C. ECKERMAN 353 ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 “This indicates that Ode 6 was sung in Ceos,”35 and Jebb states “This short ode was sung before the house of Aristomenes, Lachon’s father, in Ceos (v. 14).”36 Their assertions derive from the fact that the chorus seems to claim to be performing the ode before the house of the victor in the penultimate line, with the phrase προδόµοις ἀοιδαῖς. In his scholarly commentary, Maehler, following Kenyon, argues that the ode would have been performed on Ceos before the house of the victor’s father.37 In his later, student commentary, however, he changes his mind and leans toward interpreting προδόµοις in reference to a performance before the temple of Zeus at Olympia, though he cites no comparanda for this interpretation.38 There is, however, a Pindaric parallel that suggests that προδόµοις ἀοιδαῖς should refer to performance in the home polis of the victor and not to performance at the sanctuary. At Nem. 1.19– 20 Pindar says that his song is sung at his patron’s “courtyard gates” (ἔσταν δ’ ἐπ’ αὐλείαις θύραις / ἀνδρὸς φιλοξείνου κα- λὰ µελπόµενος, “I have taken my stand at the courtyard gates of a generous host as I sing of noble deeds”). As Christopher Carey notes, “the absence of mention of civic space in most victory odes strongly suggests that state involvement was inter- mittent at most and that most celebrations took place at a private house.”39 Both Bacch. 6 and Nem.1, then, seem to have been performed at the homes of the victors, and Bacch. 6 provides no evidence for the on-site hypothesis in either of the passages discussed above. To return to line 6, there may be one good reason for con- struing Ὀλυµπίᾳ with κρατεῦ[σαν] rather than with ἄεισάν 35 Bacchylides 143. 36 Bacchylides 203. 37 Die Lieder des Bakchylides I.2 127, 131–132. 38 Bacchylides 129–130; so too Gelzer 99, “mit Gesang vor dem Tempel.” 39 C. Carey, “Pindar, Place, and Performance,” in S. Hornblower and C. Morgan (eds.), Pindar’s Poetry, Patrons, and Festivals: From Archaic Greece to the Roman Empire (Oxford 2007) 199–210, at 203. 354 EPINICIAN POETRY AT PANHELLENIC SANCTUARIES ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 ποτ’. Bacchylides would be stressing the continuity of Cean athletic victories at Olympia and their celebration in song in their home polis, continuity which has been played out most recently with Lachon’s victory at Olympia celebrated in this song, which, on the evidence of lines 14–15, seems to have been performed at the home of the victor. Since Bacchylides is developing parallelisms between former Cean athletic victors and his current patron, the reference to this poem currently being performed on Ceos before the home of the victor, προδόµοις ἀοιδαῖς, may encourage the audience to envision an analogous performance context for the celebrations of the previous Cean victors. Regardless of Bacchylides’ intentions, I presume that some members of the audience would have construed the locative Ὀλυµπίᾳ with the main verb and imagined a reference to generic song culture at Olympia, while others would have construed it with the participle and imagined the accomplishments of Ceos in past competitions at Olympia. In addition to Bacchylides 2, 4, and 6,40 Gelzer suggests that five poems of Pindar should be classified as performed at the site of the victory. The Pindaric passages have generally been treated as marginal cases, and the majority of the argument in favor of on-site performance has revolved around the three passages of Bacchylides discussed above. Gelzer does not dis- cuss in detail the Pindaric passages and how they may be used as evidence in support of his thesis, but footnotes a list of pas- sages that he takes as evidence in support of his argument.41 It is worthwhile, then, to examine these passages in some detail. 40 While positing that Bacch. 6 was composed for performance at Ceos, A. McDevitt, Bacchylides: The Victory Poems (London 2009) 145, suggests that Bacchylides composed Bacch. 7, Bacch. 6’s “companion poem,” for perfor- mance at Ceos because of the proem: “personified ‘Day’ is … identified with the particular day of the prize-giving at the end of the festival; this sug- gests very strongly that this ode was intended to be sung in the celebrations at Olympia itself.” As McDevitt notes, this is speculative. 41 See n.11 above. CHRISTOPHER C. ECKERMAN 355 ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 Gelzer argues that Ol. 4 was performed on site because Pindar says he was a witness of the games at Olympia (line 3 ὑψηλοτάτων µάρτυρ’ ἀέθλων), but Pindar’s claim to have been present at Psaumis’ victory at Olympia provides no evidence that the ode was performed at Olympia. Pindar could have seen Psaumis win at Olympia and then composed his ode to perform at a later time in a different venue. For example, Pindar in Ol. 10 also makes reference to seeing his patron Hagesidamus win at Olympia, but nothing in that ode suggests that it was performed on site.42 Gelzer suggests too that Ol. 11 was performed on site, citing lines 7 and 12–14 as evidence, and Willcock asserts that the brief Ol. 11 is “the most secure example” of odes performed on site.43 At line 7, however, Pindar only generally mentions Olympic victors (making no claim to be at Olympia), while at 12–14 he only refers to the Olympic olive crown (again, making no claim to be at Olympia). As Neumann-Hartmann and Hubbard note, it has also been suggested that Ol. 11 was performed on site on the basis of lines 16–19 (ἔνθα συγκω- µάξατ’· ἐγγυάσοµαι / µή µιν, ὦ Μοῖσαι, φυγόξεινον στρατὸν / µήδ’ ἀπείρατον καλῶν / ἀκρόσοφόν τε καὶ αἰχµατὰν ἀφίξε- σθαι), in which Pindar tells the Muses to go to Western Locri, the home polis of the victor.44 Must Pindar, however, apostro- phize the Muses from the site of victory when he urges them to go to Locri? He elsewhere makes reference to ‘sending’ his odes to the home of the victor,45 and, as Andrew Miller has shown, he regularly merges composition time and performance time to create greater vividness.46 Pindar may, then, be composing 42 παῖδ’ ἐρατὸν ⟨δ’⟩ Ἀρχεστράτου / αἴνησα, τὸν εἶδον κρατέοντα χερὸς ἀλκᾷ / βωµὸν παρ’ Ὀλύµπιον/ κεῖνον κατὰ χρόνον (99–102). 43 M. M. Willcock, Pindar, Victory Odes: Olympians, 2, 7, and 11, Nemean 4, Isthmians 3, 4, and 7 (Cambridge 1995) 55. 44 For further discussion and bibliography see Neumann-Hartmann, Epi- nikien und ihr Aufführungsrahmen 32–33; Hubbard, in Oral Performance 77. 45 Ol. 7.8, 9.25, Pyth. 2.68, Nem. 3.77. 46 A. Miller, “Pindaric Mimesis: The Associative Mode,” CJ 89 (1993) 356 EPINICIAN POETRY AT PANHELLENIC SANCTUARIES ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 these lines anywhere, e.g. Thebes, while exhorting the Muses to go to Locri.47 Nothing, then, in the ode provides secure evi- dence for the on-site hypothesis. Gelzer also cites Pyth. 6, which has regularly been used as evi- dence for on-site performance, but Andrew Morrison, in brief, and I, in detail, have already addressed the problems of that assumption. Accordingly, I take it as a given that there is no reason to think that Pyth. 6 was performed at Delphi.48 As Mor- rison and I note, the evocation of Delphic space does not necessitate that the ode was performed at Delphi, and, follow- ing Nancy Felson, I argue that Pindar “vicariously transports” his audience to Delphi during performance at the victor’s home polis, Acragas, through his vivid narration of Delphic space. I shall add here that Pyth. 6 is a longer ode than the other ones that Gelzer classifies as composed for performance on site and, accordingly, would typologically be an outlier for the on-site hypothesis. Gelzer suggests that Pyth. 7, for Megacles of Athens, would ___ 21–53. 47 In the companion poem to this ode, Ol. 10, Pindar refers to a much- discussed “interest” (9 τόκος) that he needs to pay that may derive from his delinquency in fulfilling his encomiastic obligation for his patron. Since the time of the scholiasts, it has been claimed that Ol. 11 is the “interest” used to pay Pindar’s debt, and many editors (e.g. Dissen, Gildersleeve, Fennell, Far- nell) have followed the scholia in assuming that Ol. 11 would have come no sooner and presumably even later than Ol. 10 (i.e. it could not have been first performed on site). If Pindar does not perform Ol. 11 on site, but sends the poem to Locri from Thebes or elsewhere, then there is no clear tem- poral relationship between Ol. 10 and 11. In fact, the “interest” in Ol. 10 presumably refers to the “excellence of the ode itself,” as W. S. Barrett notes, “Pindar’s Odes for Hagesidamos of Western Lokroi: Olympians 10 and 11”, in M. L. West (ed.), Greek Lyric, Tragedy, and Textual Criticism (Oxford 2007) 55; for further discussion of Ol. 10 and the interest hypothesis, see most recently (with further bibliography) C. Eckerman, “Pindar’s κοινὸς λόγος and Panhellenism in Olympian 10,” RhM 151 (2008) 37–48. 48 A. D. Morrison, Performances and Audiences in Pindar's Sicilian Victory Odes (London 2007) 43; C. Eckerman, “Pindar’s Pythian 6: On the Place of Per- formance and an Interpretive Crux,” RhM 154 (2011) 1–8. CHRISTOPHER C. ECKERMAN 357 ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 have been performed on site at Delphi because the temple of Apollo is evoked in the ode. Pindar’s evocation of the temple, however, does not necessitate placing the performance at Del- phi. Pindar seems to evoke Apollo’s temple to instill symbolic capital in his patron, Megacles, because Megacles’ family, the Alcmaeonidae, funded its repair. There has been substantial debate regarding the ode’s performance context in relation to Megacles’ ostracism from Athens, but the ode and the histori- cal context offer no clear answer.49 Neither Pyth. 7 nor what we know of Megacles’ life provides evidence for the ode being per- formed on site. The last poem that Gelzer claims for on-site performance is Nem. 2, citing lines 1–5, 14 ff., and 24 ff. Lines 1–5 and 14 ff., however, provide no evidence in support, and lines 24 ff. provide evidence in opposition. At 24 Pindar apostrophizes the Athenian victor’s fellow citizens (ὦ πολῖται). The apostrophe would make little sense were this ode performed on site since it would not be performed among the fellow citizens of the victor (nor is the sanctuary at Nemea a polis). In addition to the apostrophe, Pindar exhorts the citizens to celebrate the victor’s homecoming, nostos, and, as Leslie Kurke has shown, nostos imagery positions the victor in his home community after his victory.50 The imagery of nostos and the apostrophe to the victor’s fellow citizens, then, situate the performance of this ode in the victor’s home polis. In addition, like Bacch. 6, which is a short poem and refers to performance in the polis of the victor, this ode too shows that Pindar composed short epinician poems for performance in the victor’s polis. Again, one should hesitate to claim that short poems can be linked with on-site perfor- mance. 49 For discussion see L. Athanassaki, “Song, Politics, and Cultural Memory: Pindar’s Pythian 7 and the Alcmaeonid Temple of Apollo,” in L. Athanassaki and E. Bowie (eds.), Archaic and Classical Choral Song (Berlin 2011) 235–268, at 238–239. 50 The Traffic in Praise: Pindar and the Poetics of Social Economy (Ithaca 1991) 15–61. 358 EPINICIAN POETRY AT PANHELLENIC SANCTUARIES ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 This paper has readdressed the evidence mustered in favor of the on-site performance hypothesis of Pindar’s and Bacchyli- des’ extant epinician odes, but it has not addressed the logisti- cal problems associated with assuming first performance at the sanctuaries. Given that all the Panhellenic festivals were rela- tively short, it is hard to imagine that Pindar or Bacchylides would have had time to compose, choreograph, prepare a chorus, and perform an ode before a festival was over and the guests of the festival left. As Gelzer notes in relation to the Olympic festival (109), Dort dauerte es im 5. Jahrhundert fünf Tage. Die Wettkämpfe begannen am zweiten und endeten am fünften, an dem auch die Kranzverleihung und die Siegesfeiern stattfanden. Für das Dichten und Einstudieren der Lieder standen also höchstens drei, im knappsten Fall weniger als ein Tag zur Verfügung … In Olympia fanden am letzten Tag die Läufe und die athletischen Agone statt. Wenn die Festordnung in Nemea dieselbe war, so wäre N. 2 für den Pankratiasten Timodemos am selben Tag aufgeführt worden, an dem er seinen Sieg errang. The constraints of the Nemean festival very well could have meant that Nem. 2 would have to be won and celebrated in song on the same day if the on-site performance hypothesis is correct. As I noted above, there is good textual evidence (ὦ πολῖται) that Nem. 2 never was first performed at Nemea; re- gardless, for the proposed on-site category as a whole, it is difficult to imagine how Pindar’s and Bacchylides’ odes could have been performed under these conditions.51 As the odes themselves attest (cf. Ol. 5.4–7, 9.1–4, 10.73–77, Nem. 6.37–38), song played a prominent role in the celebration of athletic competitions at the Panhellenic sanctuaries, and I find Hubbard’s thesis of the reperformance of epinician poems at Panhellenic sanctuaries provocative.52 Reperformed odes 51 Cf. Eckerman, RhM 154 (2011) 7. 52 Hubbard, in Oral Performance 71–93; on reperformance of epinician see too B. Currie, “Reperformance Scenarios for Pindar’s Odes,” in Oral Perfor- CHRISTOPHER C. ECKERMAN 359 ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 could have been performed on site without the logistical diffi- culties necessitated by the thesis of on-site first performance. And we must keep in mind that victories would have been cele- brated on site with other forms of singing in addition to formal victory odes (cf Ol. 9.1 ff.). Finally, it would be an impoverish- ing claim to assert that epinician odes were never first per- formed at Panhellenic sanctuaries. It may be that Pindar and Bacchylides did compose odes for victors that would be per- formed on site immediately after the victory. We might even conjecture that they could have composed odes in advance so that they would have pieces ready for performance on site if called upon, but there is no evidence in favor of this hypothesis. Two scholiasts (on Nem. 3.1c and 6.64d) affirm the perfor- mance of singing in honor of victories at the sanctuaries, though their evidence may derive only from the epinician texts themselves. Since the victors held extravagant celebrations, which included sacrifices, banquets, and song,53 it would be strange, a priori, to exclude epinician performances from these celebrations. In terms of deixis, the extant epinician odes position their first performance at the home of the victor, when they provide information regarding first performance, and, at the same time, they provide internal cues that suggest reperformance at mul- tiple venues.54 The work of Hubbard, Currie, Felson, and others has productively shifted our inquiry to questions that consider not only where the odes debuted but also where they were reperformed. As Felson notes,55 if ocular deictics once anchored an ode by pointing to the site and occasion of its first performance, such deictics—together, in the case of victory odes, with the naming of the victor, his ___ mance 49–69; Carey, in Pindar’s Poetry, Patrons, and Festivals 199–210; Morri- son, Performances and Audiences 11–19. 53 E.g. Athen. 3D–E; Plut. Them. 5.4; Pind. Ol. 5.4–7. 54 See Currie, in Oral Performance 49–69. 55 Arethusa 37 (2004) 264–265. 360 EPINICIAN POETRY AT PANHELLENIC SANCTUARIES ————— Greek, Roman, and Byzantine Studies 52 (2012) 338–360 family, and his homeland—would be expected to “rigidly desig- nate” those lost features in all possible subsequent receptions of the ode. Thus the ode would be intelligible to new, non-local audiences who could re-imagine a lost historical performance context and even imaginatively occupy the homeland space. They, as well as later auditors/readers, would re-experience, vicariously, what it was like to be among that original audience, which included fellow citizens of the victor. The findings put forth here suggest that both Pindar and Bac- chylides forefront performance in the home polis of the victor for a specific purpose: to stress the importance of the victor’s connections to his polis, and this connection will be activated both when the odes debut and when they are reperformed.56 The purpose of this paper, then, has not been to degrade the importance of song culture at Panhellenic festivals but rather to inject greater skepticism in relation to a position that has ossified and become, I believe, inappropriately entrenched in scholarly discourse. A reexamination of Pindar’s and Bacchyli- des’ epinician odes has found no evidence for the common assertion that many of these poems were first performed at the Panhellenic sanctuaries, and three of Gelzer’s eight odes have been shown to provide strong textual evidence that they were performed rather in the polis of the victor: Bacch. 4 (παρ’ ἑστίαν), 6 (προδόµοις ἀοιδαῖς), and Pind. Nem. 2 (ὦ πολῖται). March, 2012 Department of Classics University of Oregon Eugene, OR 97403 eckerman@uoregon.edu 56 I thank the anonymous referee for bringing this point to my attention.