EÏÏËÓÈο 58.2 (2008) P. COLON. inv. 21351+21376 AND P. OXY. 1787 fr. 1: MUSIC, CULTURAL POLITICS, AND HELLENISTIC ANTHOLOGIES The text* of the new Cologne Sappho papyrus, published by M. Gronewald and R. W. Daniel in 20041 and partly overlapping with P. Oxy. XV 1787 fr. 1 (= Sappho fr. 58 V),2 is significant not only for providing most of the missing text of twelve lines of Sappho fr. 58 V (lines 11-22) but also for our understanding of the transmission of the poems of Sappho in the Hellenistic period. While P. Oxy. 1787 is dated to the second/third century AD,3 P. Colon. 21351+21376 goes back to the early third century BC and con * This article is based on arguments advanced at a lecture I gave at Harvard University in February 2005. It is dedicated to Professor Panagiotis Roilos, who provided incisive suggestions that improved the final text. The following bibliography is cited throughout by author’s name only (and the corresponding date): M. Gronewald - R. W. Daniel, «Ein neuer Sappho-Papyrus», ZPE 147 (2004a) 1-8; M. Gronewald - R. W. Daniel, «Nachtrag zum neuen Sappho-Papyrus», ZPE 149 (2004b) 1-4; M. Gronewald - R. W. Daniel, «Lyrischer Text (Sappho-Papyrus)», ZPE 154 (2005) 7-12; A. S. Hunt, «[P. Oxy.] 1787. Sappho, Book iv», in B. P. Grenfell and A. S. Hunt, The Oxyrhynchus Papyri. Part XV, London 1922, pp. 26-46; E. Lobel and D. Page, Poetarum Lesbiorum Fragmenta, Oxford 1955 (corrected ed. 1963); R. Merkelbach, «Ver- zeichnis von Gedichtanfaengen», ZPE 12 (1973) 86; D. Page, Sappho and Alcaeus: An Intro- duction to the Study of Ancient Lesbian Poetry, Oxford 1955; P. J. Parsons, «[P. Oxy.] 3724. List of Epigrams», in R. A. Coles, H. Maehler, and P. J. Parsons, The Oxyrhynchus Papyri, vol. LIV, London 1987, pp. 65-82; W. Schubart, Papyri Graecae Berolinenses, Bonn 1911; E. G. Turner - P. J. Parsons, Greek Manuscripts of the Ancient World, second ed. revised and enlarged by P. J. Parsons, London 1987; D. Yatromanolakis, «Alexandrian Sappho Revisited», HSCP 99 (1999) 179-195; D. Yatromanolakis, Sappho in the Making: The Early Reception, Cambridge, Mass. 2007; D. Yatromanolakis, Fragments of Sappho: A Commentary, Cambridge, Mass. 2008 (forth- coming). 1. Gronewald and Daniel 2004a and 2004b/2005; P. Colon. inv. 21351+21376 [MP3 1449.01, LDAB 10253]. 2. MP3 1449, LDAB 3899. 3. I agree with Hunt 1922 that the papyrus should probably be dated to the third century AD. Before the publication of P.Colon. 21351+21376, Sappho fr. 58 V was the focus of much discussion (especially lines 25-26): see, among more recent analyses, V. Di Benedetto, «Il tema della vecchiaia e il fr. 58 di Saffo», QUCC n.s. 19 (1985) 145-163; G. Nagy, Pindar’s Homer: The Lyric Possession of an Epic Past, Baltimore 1990, pp. 285-286; J. J. Winkler, «Sappho and the Crack of Dawn (fragment 58 L-P)», Journal of Homosexuality 20 (1990) 227-233; B. Marzullo, «Sapph. fr. 58, 25s. V.», Philologus 138 (1994) 189-193; G. Liberman, «A propos du fragment 58 Lobel-Page, Voigt de Sappho», ZPE 108 (1995) 45-46; T. M. Falkner, The Poetics of Old Age in Greek Epic, Lyric, and Tragedy, Norman/London 1995, pp. 102-107; C. Salemme, «Lo splendore del sole e la bellezza: Saffo 58, 25-26 V.», Orpheus 22 (2001) 185-191. 238 Dimitrios Yatromanolakis stitutes by far the earliest papyrus of Sappho.4 From approximately the same period (third/second century BC)5 we have only an ostrakon that preserves what is now known as Sappho fr. 2 V. Furthermore, another early papyrus with a text of Sappho (fr. 98 V) may come from the first half of the first century BC.6 Five hundred years or so separate P. Oxy. 1787 from P. Colon. 21351+ 21376. The chronologically later papyrus apparently, but not certainly, provides us with the Hellenistic text as constituted in the Alexandrian edition of Sappho. On the contrary, the early third century BC papyrus, with its still epigraphic letter forms (Â, ˙, ı),7 preserves not only what seems to be the end of a composition of Sappho and a twelve-line song by the same poet but also another poem,8 the authorship, date, content, form, and metre of which remain somewhat indeterminate.9 Although the new papyrus was part of the 4. Gronewald and Daniel 2004a compare the hand of the Cologne papyrus (the Sappho fragments) with Turner/Parsons 1987: no. 52, pp. 92-93, P. Hib. I 4, dated to the early third century BC [MP3 1708, LDAB 7002], anonymous fragments of tragedy (TrGF 625). Note that in P. Hib. I 4, the sigma is written in four strokes, but normally in one sequence, while in the fragments of Sappho on the Cologne papyrus the sigma is lunate. 5. PSI XIII 1300 [MP3 1439, LDAB 3904, and Cribiore 247]. The ostrakon was dated by Medea Norsa to the second century BC, but with Edgar Lobel and Denys Page 1955, I believe it should be assigned to the late third century BC. For the editio princeps and other early studies of the ostrakon, see Page 1955, 35. 6. P.Mil.Vogl. II 40 [MP3 1452, LDAB 3903]. For the dating of this papyrus, see A. F. Moretti, «Revisione di alcuni papiri greci letterari editi tra i P.Mil.Vogl.», APap 7 (1995) 23. For a different dating («third century BC or a little later»), cf. Page 1955, 98 and Lobel and Page 1955, ix (saec. iii/ii a.C.»). I find Moretti’s argument more convincing. 7. Note that the archaic shape of omega (cf. P. Berol. 9875; Schubart 1911, no. 1 [MP3 1537, LDAB 4123]) is not used throughout. Cf. also the form of epsilon, which is occasionally round. In this hand the pi is broad: cf. the shape of the pi in the oldest datable documentary papyrus (c. 331-323 BC), P. Saqqara inv. 1972 GP 3 (Turner/Parsons 1987, no. 79, pp. 136- 139), in which the epsilon is rectangular and the sigma is written in four strokes; see also P. Saqqara inv. 71/2 GP 9 (Turner/Parsons 1987, no. 88, pp. 146-147), dated to the late fourth (or possibly early third) century BC. 8. Gronewald and Daniel 2005. 9. Research on the new papyrus fragments and Sappho fr. 58 V (as well as fr. 59 V) has been intense: V. Di Benedetto, «Osservazioni sul nuovo papiro di Saffo», ZPE 149 (2004) 5-6; W. Luppe, «Überlegungen zur Gedicht-Anordnung im neuen Sappho-Papyrus», ZPE 149 (2004) 7- 9; H. Bernsdorff, «Schwermut des Alters im neuen Kölner Sappho-Papyrus», ZPE 150 (2004) 27- 35; H. Bernsdorff, «Offene Gedichtschlüsse», ZPE 153 (2005) 1-6; M. L. West, «A New Sappho Poem», Times Literary Supplement (24 June 2005a) 7-8; R. Janko, «Sappho Revisited», Times Literary Supplement (23 and 30 December 2005) 19-20; M. Puelma - F. Angiò, «Sappho und Poseidippos: Nachtrag zum Sonnenuhr-Epigramm 52 A.-B. des Mailänder Papyrus», ZPE 152 (2005) 13-15; G. Lapini, «Posidippo, Ep. 51 Austin-Bastianini», Philologus 149 (2005) 233-243; M. L. West, «The New Sappho», ZPE 151 (2005b) 1-9; V. Di Benedetto, «La nuova Saffo e dintorni», ZPE 153 (2005) 7-20; A. Hardie, «Sappho, the Muses, and Life after Death», ZPE 154 (2005) 13-32; C. Geissler, «Der Tithonosmythos bei Sappho und Kallimachos: Zu Sappho fr. 58 V., 11-22 und Kallimachos, Aitia fr. 1 Pf.», Göttinger Forum für Altertumswissenschaft 8 (2005) P. Colon. inv. 21351+21376 and P. Oxy. 1787 fr. 1 239 collection of a private owner, and thus constitutes another case of purchase of ancient scrolls outside Egypt, the texts it provides present numerous intriguing problems and encourage papyrological and historical debate about the way in which texts of Sappho were transmitted in the Hellenistic period. Before the appearance of the new fragments, it had been suggested that in the Hellenistic era and possibly later, her songs were circulated in editions and collections less «standard» than the Alexandrian scholarly edition of the poetry of Sappho.10 The publication of the early Ptolemaic papyrus fragments of songs of Sappho and their juxtaposition with a fragment of a composition that cannot be assigned to Sappho further support this argument. In this article, I shall discuss the new fragments and explore this early Hellenistic collection in the light of cultural practices related to the ancient reception of Sappho and the attribution of compositions to her during the early, highly complex stages of this reception. The first eight lines of the early Ptolemaic papyrus fragments provide what appears to be the end of a composition. I shall call this new piece Sappho fr. 58b. P. Oxy. 1787 fr. 1 preserves nine different lines11 – appar- ently the end of a different composition (which I shall call 58a)12 – before those twelve lines (fr. 58.11-22 V) that partly overlap with P. Colon. 21351+21376. I shall call these twelve lines fr. 58c:13 105-114; L. Bettarini, «Note linguistiche alla nuova Saffo», ZPE 154 (2005) 33-39; M. Magna- ni, «Note alla nuova Saffo», Eikasmos 16 (2005) 41-49; A. Nicolosi, «Recuperi di lirica greca arcaica da papiri», A & R 50 (2005) 80-94 [87-94]; J. Danielewicz, «On lines 8 and 18 of the New Sappho (P. Köln 21376)», Eos 92 (2005) 133-136; V. Di Benedetto, «Il tetrastico di Saffo e tre postille», ZPE 155 (2006) 5-18; J. Danielewicz, «Bacchylides fr. 20A, 12 S.-M. and Sappho, P. Köln fr. I-II, 12», ZPE 155 (2006) 19-21; L. Edmunds, «The New Sappho: öÊ·ÓÙÔ (9)», ZPE 156 (2006) 23-26; R. Rawles, «Notes on the Interpretation of the “New Sappho”», ZPE 157 (2006·) 1-7; R. Rawles, «Musical Notes on the New Anonymous Lyric Poem from Köln», ZPE 157 (2006b) 8-13. The present article was completed in December 2006.G. Lieberg, «Bemer- kungen zum neuen Sappho-Papyrus und zu Sappho, fr. 44a, 11-12 V.», Hermes 134 (2006) 237- 238. 10. Yatromanolakis 1999. 11. These nine lines are printed in Edgar Lobel’s 1925 edition as ten lines (™·ÊÔܘ ÌÂÏáÓ ‰ fr. 1.1-10 Lobel = fr. 58.1-10 V). I have examined the original papyrus and confirmed that Hunt 1922, pp. 28-29, was right in printing nine lines, while Lobel in his ™·ÊÔܘ ̤ÏË (Oxford 1925, p. 26) was not correct in holding that there are traces of an additional line at the beginning of P. Oxy. 1787 fr. 1. 12. I call it 58a, since this papyrus was published earlier than the Cologne papyrus. 13. For the beginning of line 7 Gronewald and Daniel 2004a propose the emendation <Ù·Ü>Ù· ÛÙÂÓ·¯›Û‰ˆ (or <kÓ ‰b> ÛÙÂÓ·¯›Û‰ˆ); cf. West’s Ùa <ÌbÓ> ÛÙÂÓ·¯›Û‰ˆ. I would suggest the emendations <q> Ùa or Ù¿<‰Â> (cf. Sappho fr. 150.2 V) or <˙a> Ùa (paleographically plau- sible); even Ùa <ÓÜÓ> ÛÙÂÓ·¯›Û‰ˆ. <Ù·Ü>Ù· ÛÙÂÓ·¯›Û‰ˆ is attractive. For line 10 Gronewald and Daniel 2004b provide the restoration öÚˆÈ ‰¤·˜ ÂåÛ·Ó‚¿ÌÂÓ’ (ÂåÛÔÓ‚¿ÌÂÓ’ would be ex- pected), accepted by Di Benedetto 2005 (above, n. 9). On the papyrus fragments, after öÚˆÈ, it is 240 Dimitrios Yatromanolakis å]ÔÎ[fi]ÏˆÓ οϷ ‰áÚ· ·Ö‰Â˜ ]. ÊÈÏ¿ÔȉÔÓ ÏÈÁ‡Ú·Ó ¯ÂχÓÓ·Ó. 2 ] ÔÙ’ [ö]ÔÓÙ· ¯Úfi· ÁÉÚ·˜ õ‰Ë âÁ]¤ÓÔÓÙÔ ÙÚ›¯Â˜ âÎ ÌÂÏ·›Ó·Ó, 4 ‚¿Ú˘˜ ‰¤ Ì’ ç [ı]ÜÌÔ˜ ÂfiËÙ·È, ÁfiÓ· ‰’ Ôé ʤÚÔÈÛÈ, Ùa ‰‹ ÔÙ· Ï·›„ËÚ’ öÔÓ ùÚ¯ËÛı’ úÛ· Ó‚ڛÔÈÛÈÓ. 6 †Ù·† ÛÙÂÓ·¯›Û‰ˆ ı·Ì¤ˆ˜. àÏÏa Ù› ÎÂÓ Ô›ËÓ; àÁ‹Ú·ÔÓ ôÓıÚˆÔÓ öÔÓÙ’ Ôé ‰‡Ó·ÙÔÓ Á¤ÓÂÛı·È. 8 ηd Á¿Ú [Ô]Ù· T›ıˆÓÔÓ öÊ·ÓÙÔ ‚ÚÔ‰fi·¯˘Ó AûˆÓ öÚˆÈ ..[.]·.ÂÈÛ·Ó‚¿ÌÂÓ’ Âå˜ öÛ¯·Ù· ÁĘ ʤÚÔÈÛ·[Ó 10 öÔÓÙ· [Î]¿ÏÔÓ Î·d Ó¤ÔÓ, àÏÏ’ ·sÙÔÓ û̈˜ öÌ·Ú„Â ¯ÚfiÓˆÈ fiÏÈÔÓ ÁÉÚ·˜, ö¯[Ô]ÓÙ’ àı·Ó¿Ù·Ó ôÎÔÈÙÈÓ. 12 After these lines, the Cologne papyrus preserves a metrically different poem, the dialectal and metrical features of which do not seem to suggest that it is by Sappho (or Alkaios). The terminus ante quem for this composition is the early third century BC. For the sake of clarity, I shall call this poem an «ano- nymous composition.» However, P. Oxy. 1787 fr. 1+2 transmits a different text after fr. 58c. I shall call this text fr. 58d (= Sappho fr. 58.23-26, followed by fr. 59 V). In fr. 59 V, Voigt, adopting an idea proposed by Lobel in his ™·ÊÔܘ ̤ÏË (Oxford 1925) as well as in his and Page’s 1955 critical edition, prints the editorial symbol of initium carminis at the beginning of line 1 because Lobel believed that there was a coronis and a paragraphos between fr. 58 and fr. 59 V.14 For all that, there is no coronis between these lines in the Oxyrhynchus papyrus.15 The paragraphos after the last line of fr. 58 V may well not represent the cross-bar of a coronis. Given the appearance of other paragraphoi without coronis in the same papyrus, there is no compelling reason not to take this horizontal line as a paragraphos that marks off a strophic couplet. From fragment 58a (= 58.1-10 V) very little is preserved. As a conse- possible to read the letter ‰ (Ê seems to me impossible to read: see West 2005b Ê..·ıÂÈÛ·Ó ‚¿- ÌÂÓ’); after ‰ I would read Â. Magnani 2005 (above, n. 9) proposes öÚˆÈ ‰›[ÊÚ]ÔÓ ÂåÛ·Ó‚¿ÌÂÓ’, but Ó at the end of ‰›[ÊÚ]ÔÓ is unlikely: before ÂÈÛ·Ó one can read the upper left part of a probably round letter (ı, Ô, or lunate Û) and before that round letter it is possible to read part of an ·. If this is correct, Magnani’s restoration and Janko’s öÚˆÈ Ï·[Ï]¿ÁÂÈÛ·Ó (above, n. 9) seem improbable. 14. A coronis along with a paragraphos occurs at fr. 3.2-3, 14-15, 24 and fr. 35 col. ii. 3-4 of the same papyrus (P. Oxy. 1787). 15. Luppe 2004; cf. West 2005b (above, n. 9), 4, who believes that the three very frag- mentary lines of fr. 59 V constitute part of a poem that starts in line 23 of fr. 58 V (West 2005b, 4, 7-9). Di Benedetto 2004 and 2006 (above, n. 9) holds that lines 23-26 of fr. 58 V represent a complete poem by Sappho. P. Colon. inv. 21351+21376 and P. Oxy. 1787 fr. 1 241 quence, it is almost impossible to understand the plot of the song, unless we attempt to supplement it heavily. A reference to «success» (ÚfiÎÔ„ÈÓ) or a closely related notion16 at the very end of the fragment does not help suf- ficiently.17 Before the publication of the Cologne papyrus, the lines of fr. 58a were deemed part of fr. 58c,18 but most scholars currently believe that the first line of fr. 58c constitutes the beginning of a song, the twelve lines of which are transmitted in both the Oxyrhynchus papyrus and in the new Cologne papyrus.19 The metre of fr. 58a seems to be the same as that of 58b (preserved only in the Cologne papyrus) and 58c/d. It is an acephalous hipponactean inter- nally expanded by two choriambs (^hipp2c) or a hagesichorean with cho- riambic expansion (hag2c):20 √ – ´ ´ – – ´ ´ – – ´ ´ – ´ – –. That is the principal metre that may possibly be assigned to the fourth book of the Alexandrian edition of Sappho. However, as Hunt incisively observes, He- phaistion (p. 36 Consbruch) does not attribute this Aeolic metre (which, as Hephaistion states, was frequently used by Sappho) to a certain book of the Alexandrian edition of her poetry.21 Because the first three books of that edition were characterized by metrical homogeneity (the first book consisted of poems in Sapphic stanzas, while the second and third books were com- posed of poems in two-line stanzas and in the so-called ™·ÊÈÎeÓ ÙÂÛÛ·- 16. See ÚÔÎÔ‹ in LSJ9. 17. In the previous line, Ò]Ó˘ÌÔÓ, if correctly supplemented, might perhaps be connected with the reference to ÛÙ[‡]Ì·[ÙÈ] (or ÛÙ[‡]Ì·[ÛÈ] according to Hunt 1922, p. 42) in the last line of the fragment. 18. See F. Preisshofen, Untersuchungen zur Darstellung des Greisenalters in der frühgriechi- schen Dichtung, Wiesbaden 1977, pp. 57-64. 19. C. Gallavotti (Saffo e Alceo: Testimonianze e frammenti, vol. I, third revised ed. Naples 1962 [second ed. 1957, first ed. 1947]) had conjectured that line 11 of fr. 58 V is the beginning of a song; cf. Page 1955, p. 129, who quotes and discusses fr. 58 starting with line 11. These twelve lines are considered a complete song by numerous scholars: see, among others, Di Benedetto 2004, 6; Luppe 2004, 8; West 2005a and 2005b, 3-4; A. Carson, «The Beat Goes On», New York Review of Books (20 October 2005) 47; Janko 2005, and Bernsdorff 2005 (above, n. 9); Puelma and Angiò 2005 (above, n. 9) trace a possible allusion in Posidippus 52.5-6 A-B to Sappho fr. 58c and d, viewed as a continuous poem (= fr. 58.11-26 V). Posidippus 52 A- B (aptly discussed by L. Battezzato, «Song, Performance, and Text in the New Posidippus», ZPE 145 (2003) 33-42) is preceded by an epigram (51 A-B) with a reference to «Sapphic songs» (™·[ÊáÈ’ ôÈÛÌ]·Ù·; cf. ™·[ÊÔܘ ôÈÛÌ]·Ù·, a supplement proposed by Lapini 2005 (above, n. 9), 239), but such possible Hellenistic poetic allusions to archaic Greek compositions can be easily traced (and have all too often been traced in the past by scholars) but are difficult to ascertain or substantiate (Di Benedetto 2006 (above, n. 9), 10 refutes Puelma and Angiò’s hypothesis). 20. The term hagesichorean was coined by M. L. West (Greek Metre, Oxford 1982, p. 30) after Alkman fr. 1.57 PMG ^AÁËÛȯfiÚ· ÌbÓ ·≈Ù·. 21. Hunt 1922, p. 26. Cf. Page 1955, p. 114, who is cautious in describing the possible metre of the fourth book. 242 Dimitrios Yatromanolakis ÚÂÛηȉÂηۇÏÏ·‚ÔÓ – or Aeolic dactylic pentameter – and ™·ÊÈÎeÓ ëÎ- ηȉÂηۇÏÏ·‚ÔÓ, respectively),22 it seems plausible that the fourth book was also (wholly?) homogeneous in metre, in contrast to the fifth book, which was probably composed of metrically non-homogeneous poems was probably composed of metrically non-homogeneous poems in three-line stanzas (with glyconics predominating).23 The fourth book may have con- sisted of poems written in distichs (and perhaps tristichs).24 Certainly the isometric lines preserved in the Cologne papyrus are divided into couplets by paragraphoi.25 In fr. 58b, a female figure, being now on earth (è˜ ÓÜÓ âd ÁĘ öÔÈÛ·Ó), takes a paktis or a lyre and sings,26 while there are also references to the Á¤- Ú·˜ that possibly this woman will have in the afterlife («below the earth»).27 But it should be stressed that much remains uncertain with regard to the plot of the song. Music, song, immortality, and old age are the themes connecting fr. 58b with fr. 58c in the Cologne collection of poems. Fr. 58c consists of twelve headless hipponacteans with choriambic expansion and mainly exploits the theme of old age, the immortality of a goddess and her constantly aging lover, and possibly song and music-making in the face of mortality. The first two lines remain fragmentary, even after the publication of the Cologne pa- pyrus. Although supplementing an archaic fragmentary text implies a desire to «rewrite» the plot of a song or to reenact the archaic process of composing a poem and imitate such a complex and influential poet as Sappho, I would venture to suggest, with all related caveats, the following restoration: 22. Hephaistion p. 23 Consbruch (™·ÊÈÎeÓ ÙÂÛÛ·ÚÂÛηȉÂηۇÏÏ·‚ÔÓ, ˇz Ùe ‰Â‡ÙÂÚÔÓ ¬ÏÔÓ ™·ÊÔܘ Á¤ÁÚ·Ù·È; cf. pp. 130 and 379 Consbruch) and p. 34 Consbruch (™·ÊÈÎeÓ ëÎηȉÂηۇÏÏ·‚ÔÓ, ˇz Ùe ÙÚ›ÙÔÓ ¬ÏÔÓ ™·ÊÔܘ Á¤ÁÚ·Ù·È). 23. The evidence about the fifth book is scant. On the metres of the books of the Alexandrian edition of Sappho, see Hunt 1922, p. 26; Lobel’s 1925 edition (above, n. 11, pp. xiii-xvi), Page 1955, pp. 114-116, and Voigt’s testimonia 226-243. For the so-called book of Epithalamia and related papyrological evidence, see, further, Yatromanolakis 1999, 187-194. 24. On tristichs, see Voigt’s apparatus in Sappho fr. 88 V. 25. The same is true for P. Oxy. 1787, in cases where the left-hand side of the papyrus is preserved. 26. On multiforms of this theme, see Yatromanolakis 2007, pp. 259-272 (with n. 406 on p. 260). 27. For the overall theme, see Sappho frs. 65, 147, 32, and 55 V (with discussion in D. Yatromanolakis, «A Lyric Epos», Hellenika 56 (2006) 381-388; cf. West 2005b, (above, n. 9), 2- 3, with ample supplements for the new fr. 58b and for fr. 65 V, and Di Benedetto 2005, (above, n. 9), 7ff.). Hardie 2005 provides a detailed, but often farfetched, analysis of what he calls a «cult of the Muses» in Sappho, a cult that promised special posthumous benefits for the member of the so-called (reconstructed) «Sapphic circle». P. Colon. inv. 21351+21376 and P. Oxy. 1787 fr. 1 243 ÓÜÓ ô‰Â· MÔ›Û·Ó å]ÔÎ[fi]ÏˆÓ οϷ ‰áÚ·, ·Ö‰Â˜, Ê›ÏËÌÌ› Ù ÊÒÓ·]Ó ÊÈÏ¿ÔȉÔÓ ÏÈÁ‡Ú·Ó ¯ÂχÓÓ·Ó.28 Gronewald and Daniel proposed ʤڈ Ù¿‰Â MÔ›Û·Ó å]ÔÎ[fi]ÏˆÓ οϷ ‰áÚ·, ·Ö‰Â˜, | Ï¿‚ÔÈÛ· ¿ÏÈÓ Ùa]Ó ÊÈÏ¿ÔȉÔÓ ÏÈÁ‡Ú·Ó ¯ÂχÓÓ·Ó,29 which is interesting, since it creates a contrast between the young age of the children and the ÁÉÚ·˜ of the singing voice, which, although no more young and no more with nimble knees, resorts to music-making and singing – an idea that is echoed in the myth of Tithonos. The Homeric Hymn to Aphrodite 218-238, possibly the oldest of the long Homeric hymns30 and almost contemporary with Sappho fr. 58 (P. Oxy. 1787 fr. 1 + P. Colon. inv. 21351+21376), provides an intriguing narrative of the myth of the goddess Eos who seized âÈ›ÎÂÏÔÓ àı·Ó¿ÙÔÈÛÈÓ Tithonos31 and asked Zeus that the mortal man become immortal. Zeus fulfilled her wish, but Eos had forgotten to request eternal youth for him and ultimate avoidance of old age. So long as Tithonos was youthful, he enjoyed the love of Eos and lived by the waters of Ocean at the ends of the earth (âd ›ڷÛÈ Á·›Ë˜).32 But when the first grey hairs appeared on his head and chin – a motif in later pederastic poetry – Eos did not wish anymore to lie in his bed, but took excellent care of him in her ̤Á·Ú· by providing him with food, ambrosia, and beautiful clothing. However, old age was lurking for the immortal Tithonos, and when he eventually could not even lift any of his limbs, she locked him away in a ı¿Ï·ÌÔ˜ with shining doors. The narrative ends with a reference to Titho- nos: ÙÔÜ ‰’ q ÙÔÈ ÊˆÓc ®ÂÖ ôÛÂÙÔ˜ – like the ÊÒÓ· of the loud-voiced ¯¤- 28. Cf., e.g., Alkaios fr. 70.10-11 and 129.11-12 V. Another possibility would be ÓÜÓ Ù’ ô‰Â· MÔ›Û·Ó å]ÔÎ[fi]ÏˆÓ οϷ ‰áÚ·, ·Ö‰Â˜, | Ê›ÏËÌÌ› Ù ÊÒÓ·]Ó ÊÈÏ¿ÔȉÔÓ ÏÈÁ‡Ú·Ó ¯ÂχÓÓ·Ó; for Ù ... Ù in the Lesbian poets, cf. Sappho fr. 44.15 (with Voigt’s apparatus criticus) and Alkaios fr. 140.10-11 V. On ÊÒÓ·]Ó ... ¯ÂχÓÓ·Ó, see Sappho fr. 118 V, with discussion in Yatromanolakis 2007, pp. 7-10; Hunt 1922, p. 42, incisively observes that «¯ÂχÓÓ@Ó is gen. plur.; cf. [P. Oxy.] 1231.14.8, n.». If Sappho fr. 58d (= fr. 58.23-26 V) is part of fr. 58c, Ê›ÏËÌÌÈ in my restoration would respond to the singing voice’s love of delicacy in line 25 of fr. 58 V (on this line, see Yatromanolakis 2007, p. 128 n. 264); cf. οϷ and οÏÔÓ in these self- referential poetic statements. 29. Gronewald and Daniel 2004a, 7, with the alternative supplements ö¯ˆ for the beginning of line 1 and öÏÔÈÛ· for the beginning of line 2. Other restorations are possible: see, for example, West 2005a and 2005b, 5 and the similar supplements in Di Benedetto 2004, 5 (both provide imperative verbs for the first and/or the second line). 30. Cf. M. L. West, Homeric Hymns, Homeric Apocrypha, Lives of Homer, Cambridge, Mass. 2003, p. 16, who accepts that «[w]e shall not go far wrong if we place him (i.e., the poet of the hymn) in the last third of the seventh century». 31. The theme of being equal to gods and heroes or of conversing/being associated with gods is part of the overall poetics of Sappho: see, e.g., frs. 31, 96, 23, 1, 134, and test. 218 V. 32. Cf. the formulaic phrase Âå˜ öÛ¯·Ù· ÁĘ in Sappho fr. 58c. 10. 244 Dimitrios Yatromanolakis Ï˘ÓÓ·È that love to sing the ‰áÚ· of the violet-bosomed Muses,33 while for Tithonos Ô鉤 ÙÈ ÎÖ΢˜ | öÛı’ Ô¥Ë ¿ÚÔ˜ öÛÎÂÓ âÓd ÁÓ·ÌÙÔÖÛÈ Ì¤ÏÂÛÛÈÓ. Similarly, through the exploitation of versions of the myth of Tithonos, the singing voice in fr. 58c may allude to the possibility of being granted such an eternal poetic voice by the Muses – an idea reflected in other fragments of Sappho.34 What is certain is that the song focuses on old age pressing ir- reversibly upon the singing voice, who reflects on this aspect of her life: «but what might I do?» (†Ù·† ÛÙÂÓ·¯›Û‰ˆ ı·Ì¤ˆ˜. àÏÏa Ù› ÎÂÓ Ô›ËÓ; 58c.7).35 Already before the time of the children of fr. 58c, people used to say or believe (öÊ·ÓÙÔ)36 that Tithonos, the husband of an immortal wife, would escape death: he did so, but did not escape old age. From fr. 58d only two lines are preserved in relatively good condition (= fr. 58.25-26 V). Although there are problems with the possible semantic nuances of especially the second of these lines (58.26 V), it is clear that the singing voice places some emphasis on the following point in the song: öÁˆ ‰b Ê›ÏËÌÌ’ à‚ÚÔÛ‡Ó·Ó. Was this part of fr. 58c? 33. Note that Hunt 1922, pp. 28-29, articulated line 12 of fr. 58 V differently, on the basis of the mark of elision in the Oxyrhynchus papyrus: ]ÊÈÏ’¿ÔȉÔÓ (Hunt 1922, p. 28) and t] ʛϒ, ôÔȉÔÓ ÏÈÁ‡Ú·Ó ¯ÂχÓÓ·Ó (Hunt 1922, p. 29). Note further that Sappho fr. 121 V includes a reference to Ê›ÏÔ˜, and the female speaking voice addresses a younger male figure and advises him to take the bed of a youthful woman (àÏÏ’ öˆÓ Ê›ÏÔ˜ ôÌÌÈÓ Ï¤¯Ô˜ ôÚÓ˘ÛÔ ÓÂÒÙÂÚÔÓØ | Ôé ÁaÚ ÙÏ¿ÛÔÌ’ öÁˆ Û‡Ó <Ù’> ÔúÎËÓ öÛÛ· ÁÂÚ·ÈÙ¤Ú·; cf. Yatromanolakis 2007, p. 352). It is certainly possible that the plot of the first two lines of fr. 58c may be different from that envisaged in current restorations. On ·Ö‰Â˜ in line 1 of fr. 58c and different reconstructions of its original context, see D. Yatromanolakis, «Alcaeus and Sappho», in F. Budelmann (ed.), The Cambridge Companion to Greek Lyric, Cambridge 2008 (forthcoming). It is also worth posing the question whether fr. 58b was (or perhaps was transmitted as) part of fr. 58c (οϷ, MÔÖÛ’ can be read as ı·Ï¿ÌÔÈÛ’: the trace in the Cologne papyrus is very ambiguous; cf. West 2005b, 3, who remarks that «[e]lsewhere in Sappho (six times) the Muses are always plural» and, more subjectively, that «this last-minute apostrophe of the Muse does not look natural»). ı·Ï¿ÌÔÈÛ’ or ı·Ï¿ÌÔȘ would impart entirely different contextual nuances to the song. 34. See Yatromanolakis 2006 (above, n. 27). On Tithonos turning into a cicada in later versions of the myth (Hellanikos fr. 140 Fowler) and Sappho fr. 58c, see, from a different perspective, Janko 2005 and cf. Rawles 2006a (above, n. 9), 1-7. 35. The formation of the verb ÛÙÂÓ·¯›Û‰ˆ evidently places special emphasis on the frequent groaning of the singing voice in the face of old age. Questions constitute a constant feature of laments: see M. Alexiou, The Ritual Lament in Greek Tradition, second revised ed. D. Yatroma- nolakis and P. Roilos, Lanham and Oxford 2002, pp. 161-165. Instead, here the question takes the form of self-consolation and subtly alludes to the tradition of threnodic questions. On Sappho and laments, see Battezzato (above, n. 19), 38-39 (with references to earlier studies), D. Yatroma- nolakis, «Ritual Poetics in Archaic Lesbos: Contextualizing Genre in Sappho», in D. Yatromano- lakis and P. Roilos, Towards a Ritual Poetics, Athens 2003, pp. 43-59, and 2008 (forthcoming). 36. On the imperfect öÊ·ÓÙÔ here and â¢ıfiÌÂı· in the episode of Phoenix – who narrates the story of Meleager for the young Achilles – in Iliad 9.524, see Edmunds 2006 (above, n. 9). On the meaning of öÊ·ÓÙÔ, see Gronewald and Daniel 2004a, 8 («Die Form öÊ·ÓÙÔ ist im Epos bezeugt und hat hier die Bedeutung “glauben”, vgl. LSJ s.v. ÊËÌ› II b»). P. Colon. inv. 21351+21376 and P. Oxy. 1787 fr. 1 245 Gronewald and Daniel argued that the last line of fr. 58c provides an ending that is «sehr abrupt»37 and that fr. 58d was possibly part of 58c: according to these two scholars, the Cologne papyrus may provide only an excerpt from a longer song, since the papyrus constitutes an anthology of different compositions.38 If their suggestion is in the right direction, the singing voice returns to her own perspective (frs. 58c + 58d) after the nar- ration of the myth of Tithonos: öÁˆ ‰b Ê›ÏËÌÌ’ à‚ÚÔÛ‡Ó·Ó, ... ] ÙÔÜÙÔ Î·› ÌÔÈ | Ùe Ï¿ÌÚÔÓ öÚˆ˜ àÂÏ›ˆ ηd Ùe οÏÔÓ Ï¤ÏÔÁ¯Â.39 Although fr. 58c can certainly stand by itself and represent a more or less complete poem,40 a return to the poet’s perspective after a reference to a myth would be easily paralleled in the work of archaic poets. Can fr. 58c, as preserved in the Cologne papyrus, and fr. 58c/d of the Oxyrhynchus papyrus represent two performative versions of the same song, one of them shorter and more reflective on the (potentially also sympotic) theme of the inescapability of old age? 41 Can fr. 58c reflect a stage in the transmission of Sappho’s poetry during which Sappho’s songs were performed as skolia in symposiastic contexts? Can the complex politics of the reception of the songs and music- making of Sappho in the classical period be somewhat related to the performative transmission of certain of her compositions?42 As I have argued elsewhere,43 our earliest source concerning performance of Sappho’s poetry at Attic symposia or related venues is visual (an Attic red- figure vase of the early fifth century BC), and not, as has most often been 37. Gronewald and Daniel 2004a, 2; cf. Magnani 2005 (above, n. 9), 43-44. 38. For other early examples of such anthologies, see, e.g., P. Berol. 9772, dated to the second century BC (excerpts from comedy and tragedy; Schubart 1902, no. 6c and discussion in p. x; Schubart also cites P. Hibeh I 7, a third century BC anthology; cf. P. Hibeh II 224), P. Strasbourg W.G. 306-307, dated to c. third/second century BC (P. Strasbourg W.G. 307 in Turner and Parsons 1987, no. 30 and discussion in p. 60; they cite P. Hamburg 118 and 119, as analyzed by H. Lloyd-Jones in an addendum in G. W. Bond, Euripides: Hypsipyle, Oxford 1963, pp. 157-160). Cf. also P. Ross. Georg. 1.9 and P. Schubart 28 (P. Berol. 13680r), both dated to the second century BC. For later cases like P. Tebt. I 1 and 2, dated to the second/first century BC, see F. Pordomingo, «Les Anthologies de P.Tebt. I 1 et 2», in Atti del XXII Congresso Internazionale di Papirologia, Firenze, 23-29 agosto 1998, vol. 2, Florence 2001, pp. 1077-1093. 39. The two lines are transmitted by Athenaios 15.687b (= Klearkhos fr. 41 Wehrli). For these lines as part of the text of P. Oxy. 1787, fr. 1 + 2.1, see Hunt 1922, p. 42. 40. See, especially, the cogent analysis in Bernsdorff 2005 (above, n. 9). 41. Yatromanolakis 2007, p. 360 n. 341 (with reference to a 2005 lecture) and p. 344 n. 259. It should be stressed that the absence of gendered words in the poem would further facilitate its performance at symposia and related male gatherings. The theme of old age was exploited by Alkaios (fr. 50 V) and Anakreon (fr. 395 PMG), poets well-known for their sympotic songs. Cf. Mimnermos frs. 1, 2, 4, and 5 W2. 42. The ancient reception of Sappho is investigated in Yatromanolakis 2007. 43. Yatromanolakis 2007, pp. 51-164, 213-216; cf. D. Yatromanolakis, «Visualizing Poetry», CP 96 (2001) 159-168. 246 Dimitrios Yatromanolakis assumed, textual narratives in Aulus Gellius or Aelian.44 Furthermore, Al- kaios’ songs were performed in Attic symposia, as a number of sources of the classical period suggests.45 The Theognidea and the songs of Praxilla also seem to have been performed, improvised, and reperformed on related occasions.46 A well-known example from the Theognidea may suffice to sup- port this idea:47 lines 949-950 – transmitted as part of a poem starting in line 949 and ending in line 954 – are repeated in a different context as lines 1278ab. The transmission of the poems of Solon has also been viewed as possibly reflecting different performative versions.48 I believe that, as far as the evidence permits us to see, the question of whether fr. 58c was a shortened performative version of a longer song remains open; it is none- theless, I argue, worth pursuing it somewhat further. If Sappho fr. 58c was indeed a shortened version that was sung in sym- posia as a kind of skolion, we might want to adduce some supportive sources for such a process of transmission. The best example comes from the songs of another, almost coeval Lesbian poet. An Attic four-line skolion, originally extracted from a more extensive song of Alkaios (fr. 249 V), was being transmitted as an independent composition in (probably) classical Athenian symposia.49 It was eventually included in the collection of Attic skolia that 44. On very late textual sources and their rhetoric and multiforms, see Yatromanolakis 2007, pp. 81-88 and 86 n. 113 (cf. p. 63 n. 29). Among more recent citations of only some of these late textual sources as purported evidence for the performance of Sappho in symposia of the archaic period, see G. Nagy, Poetry as Performance: Homer and Beyond, Cambridge 1996, p. 219 (with reference to early drafts of my D. Phil. thesis); H. P. Foley, «“The Mother of the Argument”: Eros and the Body in Sappho and Plato’s Phaedrus», in M. Wyke (ed.), Parchments of Gender, Oxford 1998, p. 40; and J. Svenbro, Phrasikleia, Ithaca, NY 1993, p. 146. Scholars have confined themselves to such citations of very late sources, without attempting to contextualize them. 45. The sources are examined in Yatromanolakis 2007, pp. 210, 213-216, 341-342. 46. Cf. Yatromanolakis 2007, pp. 215, 343-344 and 343 n. 256 (with references to earlier studies of the Theognidea). 47. On different versions of poems included in the Theognidea, see M. Vetta, Theognis. Elegiarum liber secundus, Rome 1980. 48. Cf. A. Lardinois, «Have we Solon’s verses?», in J. H. Blok and A. Lardinois (eds.), Solon of Athens: New Historical and Philological Approaches, Leiden 2006, pp. 15-35; see also C. Fa- raone, «Stanzaic Structure and Responsion in the Elegiac Poetry of Tyrtaeus», Mnemosyne 59 (2006) 48-50. Traces of formulaic composition can be found in archaic melic and elegiac com- positions: cf. M. L. Lord’s Addendum to chapter 2 of A. B. Lord’s posthumous volume, The Singer Resumes the Tale, M. L. Lord (ed.), Ithaca, NY 2005, pp. 62-68 and Yatromanolakis 2007, pp. 202-205. 49. Cf. W. Rösler, «Die Alkaios-Überlieferung im 6. und 5. Jahrhundert», in Proceedings of the VIIth Congress of the International Federation of the Societies of Classical Studies I, Budapest 1984, pp. 187-190 and Yatromanolakis 2007, pp. 210, 214, 341-342.