When yellow leaves, or none, or few do hang
Upon those boughs which shake against the cold,
Bare ruined choirs, where late the sweet birds sang.
Shakespeare’s image of the absence of song neatly sums up the familiar notion that the essential misery of old age is its distance from the worlds of youthful gaiety. Over twenty years ago, Vincenzo Di Benedetto argued for the literal presence of this motif in Sappho. He posited a pointed contrast between youth and old age in fr. 58V and fr. 21V (both known only from Oxyrhynchus papyri), and inferred in the mention of the Muses’ gifts in fr. 58.11 the presence of the theme which denied song to the aged poet, although he cited only one poem, by Horace—Odes 3.15, to an aging meretrix—as a parallel for the actual mention of song in this context (the denial of the pleasures of love, of course, is easily paralleled).  Di Benedetto reaffirmed his reading shortly after the publication of the Cologne papyrus (P.Köln inv. 21351+21356) in which 58.11 is the first of the twelve overlapping lines of the two papyri we call the New Poem. For lines 1 and 2 of the New Poem, where the original editors had suggested supplements that made Sappho represent herself as a musical performer, he proposed instead imperative verbs (1. γεραίρετε…2. χορεύσατε) attributing all the musical activity to the paides (Di Benedetto 2004:5–6; Gronewald and Daniel 2004a:7). Di Benedetto’s understanding became the basis for M. L. West’s commentary on the New Poem:
And West’s supplements for the opening lines modified only the details of Di Benedetto’s in order to underscore the opposition they both inferred.
ὔμμες πεδὰ Μοίσαν ἰ]οκ[ό]λπων κάλα δῶρα, παῖδες,
σπουδάσδετε καὶ τὰ]ν̣ φιλάοιδον λιγύραν χελύνναν·
(Cf. Gronewald and Daniel: 1. φέρω…2. λάβοισα). This has now become the common version of the New Poem, taken as the starting point for further articles in ZPE and disseminated in a variety of translations.
I recall this (recent) history to emphasize that the interpretation depends entirely on the supplements. With the second person verbs, Sappho begins by distancing herself from the activity of her companions, and then resigns herself to the human condition (Di Benedetto), and finds consolation for it (West), by reflecting on the fate of Tithonus as a mythological paradigm for the necessity of aging. But the Cologne papyrus does not actually offer any evidence that confirms this interpretation, because, despite all the gains, nothing in the material outside the brackets gives any indication of the verb structure in the first two lines. Certainly, we learn in the poem that the speaker has aged and does not dance (and Alcman 26 PMG, the oft-cited parallel, also says no more than that); but it remains possible that the speaker is emphasizing not the distinction between herself and her addressees, but between singing (which she does) and dancing (which she cannot do).  I regard this, in fact, as more probable.
The unusual circumstance of the discovery of the same text in a second papyrus offers the opportunity for a new appraisal of what we should look for when we begin the combined tasks of interpreting and supplementing. For the Cologne papyrus offers not only more words, but a new and different context. Papyrus books themselves constitute ancient contexts. Ancient sources attribute different meters to different books of Sappho, and some papyri that apparently derive from continuous texts present groups of fragments that can conform to a single meter or type. So it is reasonably certain that there was a metrical basis for the arrangement of poems into books in a standard ancient edition.  In both sources, the New Poem is preceded by lines in the same meter, but they are not the same lines, so the two papyri are not fragments of copies reproducing the same edition. In this case, therefore, we have the opportunity to take into account the characters of two different contexts. Taken together, they will show that it is more likely that the aging speaker is not expressing resignation or searching for consolation in the common fate of mankind, but rather confidently asserting that her continued devotion to singing assures her of a kind of immortality.
I will begin with the context in the new find. The status of the three poems on the Cologne papyrus has been thoroughly discussed by the original editors and others, especially Alex Hardie (2005) and John Lundon (2007a) (although their discussions proceed from the assumptions of the Di Benedetto / West reading of the New Poem). Despite much that must remain in doubt (for example, whether the poems are complete, or the temporal relation of this papyrus to the Alexandrian edition), for my purposes some points can be reasonably assumed. If the Cologne papyrus represents an alternate metrical edition—even though we have no other reason to think that there was more than one—it would be the end of its book, since the scribe of the New Poem left the remainder of its column blank; it is odd, therefore, although not decisive, that there is no subscription. But it is more likely that this papyrus represents some kind of anthology or partial compilation; the poems of Sappho were popular and would have circulated outside whatever complete edition was available when it was written. What is crucial for establishing the context is the third poem, an anonymous lyric written in the space after the New Poem, in a second, nearly contemporaneous hand. It is not Lesbian; its affinities, which are fully explored by Dee Clayman in this volume, seem to be Hellenistic. However, the association of the three poems does not appear to be unmotivated (which would be possible, as Lundon notes, if it were, say, a symposiast’s handbook); for the third poem not only shows imitations of Sappho, it shares certain subject matter with the two preceding it. At the least, all three contain references to music and hints of a concern with the underworld or some kind of life beyond the normal limit of death:
P.Köln 1–8: .νέρθε δὲ γᾶς … ἔχοισαγ γέρας
… ὠς νῦν ἐπὶ γᾶς …— λιγύραν, —…ἔλοισα πᾶκτιν…ἀείδω;
New Poem: [Μοίσαν]…δῶρα, —… φιλάοιδον λιγύραν χελύνναν …
—myth of Tithonus ἔχοντ’ ἀθανάταν ἄκοιτιν
anon. lyric: Hermes? — …ἄπνους ?… —Orpheus…— [εὔ]
φθογγος λύραν 
What the third poem provides, therefore, is a reading of the poems that precede it. Whatever the purpose or origin of the collection (and even if it was part of an edition), to the person who added the third poem, the essential point about old age in the New Poem would have been its connection with the theme of death. This too, of course, is a familiar complaint about old age; indeed, it is easier to find parallels to the connection of age and death than to the separation of age and music.  Mimnermus (2.57W) pairs death and old age as the burden of the “black fates” that stand beside us:
… Κῆρες δὲ παρεστήκασι μέλαιναι,
ἡ μὲν ἔχουσα τέλος γήραος ἀργαλέου,
ἡ δ’ ἑτέρη θανάτοιο· …
and Anacreon (50/395.1–2, 7–8 PMG) describes his aging self in terms quite similar to Sappho, and bewails his condition as a sign of impending death:
πολιοὶ μὲν ἡμὶν ἤδη
κρόταφοι κάρη τε λευκόν …
διὰ ταῦτ’ ἀνασταλύζω
θαμὰ Τάρταρον δεδοικώς·
More important for us, however, is that in the first poem (with its mention of being under the earth, having honor, and singing) and in the third (with its mention of Orpheus), music is not only something not lost in old age but even represents a claim on a kind of immortality.  So even if the anthologist is quoting only so much as fits his purposes, it seems unlikely that the initial lines of the New Poem represented the poet’s resignation from the musical world, in contrast to the paides. It is the benefits of music in the face of mortality that seems to have piqued the original compiler’s interest.
The Oxyrhynchus papyrus tells a different kind of story. This copy does appear to come from a standard edition of Sappho organized by meter; fragment 58 is one of over 25 items in Voigt’s edition (frr. 58–87) that are apparently from the same roll and can be interpreted as in the same meter. But because some partial correlation of meter and content can be observed throughout the corpus of Sappho’s poetry, the metrical arrangement may also offer some contextual guidance.
I should stress that I am not arguing that specific meters have specific values or that there are necessary restrictions on what can be said in any meter. But, despite how little actually survives, I do think we can observe some clusters that indicate that for certain kinds of songs or certain types of occasion or performance situations, some meters were typical or at least preferred to others. Conversely, there may have been themes or occasions for which some meters were inappropriate. There are, for example, numerous prayers for divine presence or assistance in the sapphic stanza; of the 19 fragments by Sappho in Voigt that are substantial enough to comment on, seven are prayers and three more may be prayers or contain reports of prayers. Among the 135 items in other meters, only one looks similar (fr. 86). Alcaeus’s sapphics have the same tendency: 5 out of 8 are prayers. On the other hand, there appear to be no overt references to the Graces or Muses in the sapphic stanza, no references to the afterlife, and apparently no poems which are self-reflectively concerned with the art of singing.  The remnants of the second book were in glyconics with double dactylic expansions (the meter that used to be called aeolic dactyls). Two major examples survive, the description of Andromache’s bridal procession and a Hymn to Artemis that is possibly by Sappho; both of these are characterized by extended narrative and a style and content we connect with the Homeric tradition. Looking ahead, in Book 5, which seemed to exhibit more metrical variety, we see that the stanzas in some poems are similar to each other in that they incorporate thrice-repeated glyconics. In the two partially readable fragments of poems of this type, 94 and 96, a woman is being reminded of the joy that remembered pleasures can bring; what remains of a third, fr. 95, contains a motif that also occurs in fr. 94 (but nowhere else), someone who wishes she were dead, and that poem also seems to feature a narrative of the past.
The New Poem, as well as the preceding poem in the Cologne papyrus, would have come from Book 4, which, as best as can be deduced, contained poems written in distichs and perhaps tristichs (see Voigt ad fr. 88). It appears that the individual items were short; two consecutive fragments (numbers 62 and 63) have the left margin preserved, and three instances of a coronis mark them off as 12 and 10 lines long, respectively; at 12 lines the New Poem would fit this pattern. The basic meter of both Books 3 and 4 was an aeolic meter with choriambic expansion, rather than the dactylic expansion of Book 2. The meter of the poems in Book 4 was a headless hipponactean with a double-expansion:
^hipp2c = x l k k l l k k l l k k l k l l
gly2c = l x l k k l l k k l l k k l k l
Throughout these fragments we find references to music and to the afterlife, and indications of direct address. We can also pick up an assertive or argumentative tone: people are compared, and there are expressions of strong disapproval. The typical features of these fragments are most fully represented in a poem from Book 3 which we know from several quotations, fr. 55 :
κατθάνοισα δὲ κείσῃ οὐδέ ποτα μναμοσύνα σέθεν
ἔσσετ’ οὐδὲ †ποκ’†ὔστερον· οὐ γὰρ πεδέχῃς βρόδων
τὼν ἐκ Πιερίας· ἀλλ’ ἀφάνης κἀν Ἀίδα δόμῳ
φοιτάσῃς πεδ’ ἀμαύρων νεκύων ἐκπεποταμένα.
But when you die you will lie there and no memory of you
will linger in later time, for you have no share in the roses
that come from Pieria. Unnoticed in Hades’ house as well,
you will range among the shadowy dead, flown from our midst.”
(tr. A. Miller 1996:58)
The poem is strongly opinionated, in fact censorious, uses direct address and implies a dialogue context; it refers to poetry and the Muses, and these are involved in the judgment it passes; and it refers to life after death. Many of the other fragments are too scrappy to say much about, but I indicate some of the significant words in the accompanying chart (which excludes fr. 58) We find an invocation of the Graces in fr. 52, and the correct way to approach them is the subject of fr. 81; in fr. 56 a parthenos is praised for her exceptional sophia, which I would take to mean musical skill. In the half-dozen readable words of fr. 70 we find mention of a chorus, and perhaps harmony. For the dialogic or disputatious element, we see that fr. 71 addresses Mica and condemns her for choosing the friendship of the house of Penthilus, then says something about singing. Music is not always involved. Someone is censured for rustic dress in fr. 57. In fr. 60 we find something about “fighting with me” and “you know well.” In fr. 65, Sappho is in the vocative, we see “you,” and fame and the underworld are mentioned. Fr. 62 presents itself as a response to other speakers. Fr. 68 contains a number of moral judgments: we find “sinful” “does not restrain koros”, someone is described as no longer without guile; “sinful” in fr. 69 and “shameful” or “ugly” in fr. 64 are among their very few readable words.
Fragments of Sappho, Books 3 and 4 (not including fr. 58)
|Frag.||length||censure / comparison||direct address/
|music or afterlife|
|55||4 lines||οὐδέ μναμοσύνα
|2nd pers.||βρόδων τῶν ἐκ
κἀν Ἀίδα δόμῳ
|56||2 lines||οὐδ’ … ἔσσεσθαι …
|57||=2 lines||ἀγρωΐωτις …. οὐκ
|60||c. 18 wds||ἔμοι μάχεσθαι, σὺ…|
|62||c. 20 wds||μύγις…εἰσάιον,
|63||c. 30 wds||(prayer?)||?μὴ πεδεχην…μηδὲν
|65||15 wds||Ψάπφοι σε ?||πἀντᾳ κλέος,
σ’ ενν Ἀχερ[οντ…
|67||c. 12 wds||2nd pers.|
|68||c. 20 wds||αλίτρα,
κόρον οὐ κατισχε.[,
|69||c. 2 wds||ἀλίτρα|
|70||c. 8 wds||]αρμονίας (aut n. pr.)
χόρον, … λίγηα.[
|71||c. 15 wds||κα[κό]τροπ’ … μέλ[ος] τι γλύκερον||σ’ ἔγωὐκ ἐάσω||μέλος …
|72-80||c. 20 wds total|
|81||4 lines||ἀστεφανώτοισι δ’
|2nd pers. imper.||Χαριτες|
|82||1 line||εὐμορφοτέρα Μ. τὰς Γ.|
|86||c. 12 wds|
(The word count is of words which can be read at all, including conjunctions and particles. Frr. 61, 66, 83–85 are too scrappy even to list. Frr. 58–86 are from P.Oxy. 1787; in some cases this is the only clue to their meter. Voigt notes that 64, 65, 73, and 86 may possibly be from tristichs in which the third verse was shorter than the first two.)
The first poem in the Cologne papyrus, with its apparent claim for a singer’s fame below the earth, would fit here as well. The unreadable lines in fr. 58 before the New Poem seem to end with “success to/by the mouth,” which could also conform. The only fragment that certainly seems to strike a tone different from what I have mentioned is fr. 63, in which someone appeals to Dream, as if to a god, and seems to seek relief from anxiety; but the poem also mentions a concern with “sharing” and “the blessed,” presumably a reference to the afterlife. 
These scattered examples may not seem like much, but it is important to note that there is in fact very little to read (the second column indicates how much can be read of each). Even if some of the examples appear dubious, it is striking that these key features turn up so readily in so little material. Consider again, that in the Sapphic stanzas, of which there are much more copious remains, discussions of singing as an activity and mentions of the Muses or Graces are rare or missing.  Dialogue appears throughout the corpus, but the opinionated and assertive tone is especially peculiar to these fragments, especially in combination with musical activity.  The Sappho that I infer in these fragments is the Sappho described by Aelius Aristides (Orat. 28, 51 = Sa. 193 L-P, 55 V Test.):
οἶμαι δέ σε καὶ Σαπφοῦς ἀκηκοέναι πρός τινας τῶν εὐδαιμόνων δοκουσῶν εἶναι γυναικῶν μεγαλαυχουμένης καὶ λεγούσης ὡς αὐτὴν αἱ Μοῦσαι τῷ ὄντι ὀλβίαν τε καὶ ζηλωτὴν ἐποίησαν καὶ ὡς οὐδ’ ἀποθανούσης ἔσται λήθη.
I think you must have heard Sappho too boasting to some of those women reputed to be fortunate and saying that the Muses had made her truly blessed and enviable, and that she would not be forgotten even when she was dead.
There are, of course, also references to Eros and Aphrodite among these fragments, in common with all the other books. What we do not find are any hints of a need for resignation in the face of mortality or any expectation that the end of life must necessarily be a dead-end.
Taken together, then, the contexts of the two papyri suggest that we should be looking for a forthright assertion by the speaker—forthright enough to counter any accusation to the contrary—that she enjoys the virtues of a musical life, in particular the expectation of a reward in the afterlife. We should not expect any concession that a person who practices the arts of the Muses will be seriously inferior or lacking in comparison to any one else—whatever the appearances. The speaker of the new poem will be altogether unlike the unfortunately doomed unmusical addressee of fr. 55. Does the text support such a reading?
As I remarked above, the crucial moment for interpretation is the first couplet, because it is there that the poem answers the question of to whom the gifts of the Muses belong. Since παῖδες can be vocative without being the subject, we can look for a supplement in which the gifts of the Muses are what the speaker lays claim to, not what she regrets. That is possible because we can separate singing and lyre-playing, to which age is no barrier, from dancing, to which it is. We could return to the supplements suggested by the original editors (2004a:7), which have been largely ignored in the subsequent discussions:
φέρω τάδε Μοίσαν ἰ]οκ[ό]λπων κάλα δῶρα, παῖδες,
λάβοισα πάλιν τὰ]ν̣ Φιλάοιδον λιγύραν χελύνναν·
This at least makes Sappho the lyre-holder, but “bringing” the gifts establishes no line of thought and opens an unanswered question about the context —bring where? They call these lines a prelude, preempting the search for a connection (they assume that the full poem ended with the lines about the love of the sun and constituted a “personal declaration” on the part of the poet). As an example of a more purposeful opening I suggest:
νῦν δὴ μ’ ἔτι Μοίσαν ἰ]οκ[ό]λπων κάλα δῶρα, παῖδες, (or νῦν μ’ ἤδεα Μοίσαν…)
φίλημμι δὲ φώνα]ν̣ φιλάοιδον λιγύραν χελύνναν·
For the first line: “now I have still the Muses’ lovely gifts”  or “now the lovely gifts of the Muses are sweet to me”, and for the second, “and I love the song-loving voice of the resounding lyres.” This sets up the emphasis on the Singer’s present condition that is the basis for the contrast between singing and dancing. The opening couplet exemplifies a structural feature that I find characteristic of the first six lines: the first line of each couplet is general—here the Muses’ gifts—and the second specifies what is immediately relevant—here the lyre and song.
Two other problems are solved in this example as well. The noun φώναν removes the hard-to-explain definite article of other supplements; and by taking λιγύραν χελύνναν as a genitive plural, it avoids the problem of a doubled adjective for one noun.  For the semi-personification of the lyre, compare fr. 118, χέλυ…φωνάεσσα. The plural suggests that the lyres belong to the Muses, singing in harmony; that is, that the singer identifies her song with theirs.
Line 3 begins the explanation for the emphasis on singing in the first couplet. She loves the lyre and song, for she has gotten old. This description fills two couplets:
νῦν γὰρ μ’ ἄπαλον πρὶν] π̣οτ’ [ἔ]ο̣ντα χρόα γῆρας ἤδη 3
κατέσκεθε, λεῦκαι δ’ ἐγ]ένοντο τρίχες ἐκ μελαίναν·
βάρυς δέ μ’ ὀ [θ]ῦμος πεπόηται, γόνα δ’ [ο]ὐ φέροισι, 5
τὰ δή ποτα λαίψηρ’ ἔον ὄρχησθ’ ἴσα νεβρίοισι.
γὰρ Di Benedetto 1985 ἄπαλον e.p. πρὶν Di Benedetto 2004 κατέσκ. West 2005 λεῦκαι e.p. P.Oxy. 1787
“For now age has taken hold of my once lovely flesh” (cf. fr. 2 for the repeated initial adverb).  The introduction of old age is directed to the goal of explaining something. I stress this point because commentary so far has rushed to describe this as a paratactic symptom-list, like the list of afflictions in the phainetai moi fragment, or even in Anacreon’s little poem, where his white hair is complemented by his aging teeth. But these lines have more internal structure than those. The first of these two couplets describes external appearances. Sappho begins, again, with a general statement, that her exterior has been ruined by age. Gronewald and Daniel’s supplement ἄπαλον is apt, but not for the reason they give. They cite ἁπαλὸν χρόα from Archilochus 118.1W, where they assume a contrast with Snell’s ὄγμοις as a metaphor for “wrinkles” (which they also supply in the next line here).  Such a use of the word would make the line specific, smooth skin set against wrinkled, parallel to the contrast of black hair and white in the next line, and so would conform to a reading of the lines as a unstructured list. But ἄπαλος is one of Sappho’s own general purpose words for youthful attractiveness, and she uses it as an absolute quality; she applies it not only to hands weaving wreaths (fr. 81.5; cf. Alcaeus 45.5 and inc. auct. 16.2 V= Sappho 93 L-P), and a neck decorated with flowers (fr. 94.16), but also to whole persons: ἀπάλας Γυρίννως (fr. 82a), a child collecting flowers (fr. 122), and sleeping companions (fr. 126).  So I take this line to be the general introduction to the youthful appearance that she has lost. In the next line, she is more specific: she has that most notorious visual mark of age, white hair.
The third couplet proceeds from the outward appearance others can see to the physical consequences felt by the singer herself. In its first lines she says, in general terms, that her thumos is heavy. This is a physical incapacity, not a psychological affect. Thumos in the Lesbians refers to the organ which generates desires that can be acted on and gives impulse to motion. Di Benedetto pointed this out in his notes on the Cologne papyrus,  but since it has become common to translate βάρυς… [θ]ῦμος as “heavy-hearted” and give it the value of a state of mind,  it will be useful to review the evidence. The thumos explains action,
Alcaeus 308b.1–2 V χαῖρε Κυλλάνας ὀ μέδεις, σὲ γάρ μοι
or is connected to the expectation of action; it is found on either side of prayer:
Sappho 5.3–4 V κὤσσαϊϝ]ο̣ι̣ θύμω<ι> κε θέληι γένεσθαι
πάντα τε]λέσθην, (cf. Sa 1.26–7; 60.5–6)
Alcaeus 129.9–12 V …. ἄγ̣[ι]τ̣’ εὔνοον
θῦμον σκέθοντες ἀμμετέρα[ς] ἄρας
ἀκούσατ’, ἐκ δὲ τῶν[δ]ε̣ μ̣ό̣χ̣θ̣ων
ἀργαλέας τε φύγας ῤ[ύεσθε·
When the thumos is repressed, a person is unable to act. So the birds afflicted with a cold thumos drop their wings:
Sappho 42 V ταῖσι < l > ψῦχρος μὲν ἔγεντο θῦμος
πὰρ δ’ ἴεισι τὰ πτέρα
It is relief from such impairment that Sappho seeks in Poem 1:
Sappho 1.3–4 V μή μ’ ἄσαισι μηδ’ ὀνίαισι δάμνα,
By the end of the poem she is ready to act. Alcaeus says that something so stirred Helen’s thumos that she went into motion and followed Paris to Troy:
Alcaeus 283.3–6 V κἈλένας ἐν στήθ[ε]σιν [ἐ]πτ[όαισε
θῦμον Ἀργείας· Τροΐ̣ω<ι> δ’ [ἐ]π̣’ ἄν[δρι
ἐκμάνεισα ξ̣[ε.]ναπάτα<ι> ‘πι π[όντον
But in contrast, when the same verb is used to describe the disturbance that renders Sappho incapable of action, it does not act on her thumos:
Sappho 31.6–7 V ……. τό μ’ ἦ μὰν
καρδίαν ἐν στήθεσιν ἐπτόαισεν,
The general sensation that βάρυς…[θ]ῦμος describes in the New Poem is what we call not a “heavy heart” but a “slowing of the blood” (or a low thyroid-count), and Sappho explicates it immediately as a problem of motion: she cannot move her limbs. That is the general difficulty. In the next line she gets to the point to which the whole description has been building: specifically, she cannot dance. And that is why she emphasizes lyre-playing and singing.
But, as elsewhere in Sappho, the logical development of the argument has become the vehicle for a shifting tenor.  The explicit statement of the first three couplets, that the Singer maintains her attachment to song, even as age has deprived her of dance, has turned our attention to the deficits of old age. She makes this topic explicit at the start of the next couplet:
τὰ <μὲν> στεναχίσδω θαμέως· ἀλλὰ τί κεν ποείην;
ἀγήραον ἄνθρωπον ἔοντ’ ού δύνατον γένεσθαι.
“I often bewail this, but what might I do?” στεναχίζω is a striking word choice here. Greek gives us στένω, the frequentative στενάζω, its poetic lengthened form στενάχω, and this further lengthened form στεναχίζω. The word is frequent in Homer, where it is clearly very emphatic, and occurs in no other archaic or classical author. What has moved Sappho, a writer who rarely uses a long word where a short one will do, and who avoids otiose syllables, to adopt this blatant epicism? To use such a word here is to call attention to the manner of speech itself. In addition, the whole phrase has something formulaic about it. Compare the lament for Adonis, fr. 141:
κατθνάσκει, Κυθέρη’, ἄβρος Ἄδονις· τί κεν θεῖμεν;
ἀγήραον ἄνθρωπον ἔοντ’ ού δύνατον γένεσθαι.
“Adonis is dying, what might we do?” and the cry of the Furies when they think that Athena’s judgment has utterly vanquished them: στενάζω· τί ῥέξω; (Aeschylus Eumenides 788). It is the cry not of someone confronting gray hair and arthritis, but confronting death or ruin. On the one hand the language is disproportionate to what has been described, perhaps humorously exaggerated; on the other, it brings us to the true subject of the poem.  The gnome on the inevitability of old age, which completes the couplet, is a reminder not only of the necessity of losing our youth but of the inevitability of death. It is in this context that the role of Dawn must be appreciated. For from this invocation of mortality at its center, the poem now works back to its beginning. Tithonus too did not move his limbs, but he was carried to the ends of the earth by a god’s favor (with a repeated use of φέρω):
καὶ γάρ π̣[ο]τ̣α̣ Τίθωνον ἔφαντο βροδόπαχυν Αὔων
ἔρῳ [….]εἰσανβάμεν’ ἐς ἔσχατα γᾶς φέροισα[ν,
And when gray old age took from Tithonus his youthful looks, he too still had an immortal companion:
ἔοντα̣ [κ]ά̣λ̣ο̣ν καὶ νέον, άλλ’ αὖτον ὔμως ἔμαρψε
χρόνωι π̣ό̣λ̣ι̣ο̣ν̣ γῆρας, ἔχ[ο]ντ’ ἀθανάταν ἄκοιτιν.
The final position establishes Dawn as a parallel to the Muses in the first line. The participle ἔχοντ’ should not be read concessively (“even though he had…”; ὔμως contrasts with the previous verbs), as would be required by the logic of the consolatory interpretation; that degree of emphasis on a concessive meaning would be difficult without a particle—καί or περ—and lacks a parallel in Sappho.  Rather, the emphasis shifts from the loss to what Tithonus has and will always have. In other versions of the story, in particular the one related by Aphrodite in the Hymn to Aphrodite, Tithonus is marked by the continuing sound of his voice (237) in the context of his weakness and his rejection from Aphrodite’s bed. Although the hymn is the nearest parallel to Sappho temporally and marks Tithonus as in between the human and divine, we need not assume that Aphrodite’s speech there (explaining to her lover, somewhat illogically, why she will not marry him) determines a negative image of Tithonus for the whole early tradition; the formulaic references in Homer (Iliad 11.1–2=Odyssey 5.1–2) describe him as Dawn’s continuing bed mate. Nor should we assume a necessary opposition between elderly voices and the capacity for song. The value of the voice of the aged was also an available tradition. The old men of Iliad 3 are likened to cicadas (to which Tithonus is also likened in the fifth century), which become emblems of the Muses in the opening of the Phaedrus.  The Old Men of the Agamemnon make clear that singing is what is left when strength and mobility are gone (72–75, 104–109):
ἡμεῖς δ’ ἀτίται σαρκὶ παλαιᾷ
τῆς τότ’ ἀρωγῆς ὑπολειφθέντες
ἰσόπαιδα νέμοντες ἐπὶ σκήπτροις.
κύριός εἰμι θροεῖν ὅδιον κράτος αἴσιον ἀνδρῶν
ἐντελέων· ἔτι γὰρ θεόθεν καταπνεύει
πειθώ, μολπᾶν ἀλκάν, σύμφυτος αἰών·
In content, the 12 lines of the New Poem assert positively what fr. 55 asserted negatively: having a share in the world of music—the roses of Pieria, the song even if not the dancing—assures one of continuing companionship, even in the afterlife. The singer makes an assertion, and the tone is confident.
Is the New Poem a complete poem? It has a beginning, middle and end, and the general structure of ring composition. But there are also readings that incorporate the lines that follow in the Oxyrhynchus copy (as Boedeker and Lardinois illustrate in this volume)—two fragmentary lines that allow for a variety of connections with the New Poem, and two completed by a quotation in Athenaeus; objections have also been raised to making the end of the poem coincide with the end of a paradigmatic myth. I think that the root of the problem, in this case, is the notion of a poem that stands by itself. I would prefer to say that these 12 lines are a complete statement. The second person address and the context of Book 4 suggest that there is a dialogic situation; the poet affirms what another might deny or devalue.  In fact, the lines from Athenaeus that are part of what follows in the Oxyrhynchus papyrus assert the value of the here and now, and could well be a contrastive reply to these lines. But it could also be the statement of the person who is censured in fr. 55, just as the New Poem could be a response to the kind of anxiety expressed by the speaker of fr. 63, who appears to fear not having a share in the blessed. So I propose that what we are looking at are songs for performances that involved some kind of give and take, whether between the chorus and the chorodidaskalos negotiating their roles or between choral members (or leaders) enacting a rivalry, in which relatively short compositions were exchanged in a kind of lyric conversation.  As relatively independent units, they may have been combined in any number of ways. Whatever their original arrangement, they could have been rearranged for re-performance. If they were composed for the same type of occasion—in other words, belong to a single genre—the similarities among the songs might have suggested different ways of grouping them to subsequent performers, readers, or editors. Certainly, the Alexandrian standard edition of Sappho may have collected such items without regard for their sequence in any actual performance. If the poems were composed for an occasion in which the participants used both the meters of Books 3 and 4, the system of collection, based on meter, would have separated them. (Sappho fr. 137 V, the exchange between Sappho and Alcaeus quoted by Aristotle, while probably apocryphal, suggests the possibility that Athenian readers could have found dialogue using related meters.)
Such exchanges may have a direct descendant in the bucolic amoebaea of Theocritus, but there is now another possible confirmation of the existence of dialogic exchanges. In a poem mourning the death of the parthenos Nikomache, disappointing her many suitors, Posidippus tells how Fate took, among her other delights, Nikomache’s continuing exchange of Sapphic song or conversation at the loom.
πάντα τὰ Νικομάχης καὶ ἀθύρματα καὶ πρὸς ἑῷαν
κερκίδα Σαπφῴους ἐξ ὀάρων ὀάρους
ᾤχετο Μοῖρα φέρουσα προώρια· …
(Posidippus 55.1–3 A-B)
All Nicomache’s delights, all her conversations à la Sappho,
one after the other, at the sound of the morning shuttle,
Fate has taken away prematurely.
The editors refer the words generally to feminine talk about love and other subjects.  But Posidippus may be suggesting a more specifically literary activity, if my assumption about the contents of Books 3 and 4 is correct, for there could be no better description of such a collection then to call it Σαπφῴους ἐξ ὀάρων ὀάρους, conversations in song by Sappho about the rewards of a life in which the Muses and Graces and what they concern themselves with play a major part, composed for performance by or with young women. 
Acosta-Hughes, B., and S. A. Stephens, 2002. “Rereading Callimachus’ Aetia Fragment 1.” Classical Philology 97:238–255.
Bastianini, G., and C. Gallazzi, eds. 2001. Epigrammi: P.Mil.Vogl. 8. 309. With C. Austin. Papiri dell’ Università degli studi di Milano 8. Milan.
Bergk, T., ed. 1878–82. Poetae Lyrici Graeci. Leipzig.
Bernsdorff, H. 2004. “Schwermut des Alters im neuen Köln Sappho-Papyrus.” Zeitschrift für Papyrologie und Epigraphik 150:27–35.
Brandt, H. 2003. Wird auch silbern mein Haar: eine Geschichte des Alters in der Antike. Beck’s archaologische Bibliothek. Munich.
Broger, A. 1996. Das Epitheton bei Sappho und Alkaios: eine sprachwissenschaftliche Untersuchung. Innsbrucher Beiträge zur Sprachwissenschaft. Innsbruck.
Calame, C. 2001. Choruses of Young Women in Ancient Greece: Their Morphology, Religious Role, and Social Function ed. 2. Trans. D. Collins and J. Orion. Lanham, MD.
Di Benedetto, V. 1985. “Il tema della vecchiaia e il fr. 58 di Saffo.” Quaderni Urbinati di Cultura Classica n.s. 19:145–163.
———. 2004. “Osservazioni sul nuovo papiro di Saffo.” Zeitschrift für Papyrologie und Epigraphik 149:5–6.
———. 2006. “Il tetrastico di Saffo e tre postille.” Zeitschrift für Papyrologie und Epigraphik 155:5–18.
Edmunds, L. 2006. “The New Sappho: ἔφαντο (9).” Zeitschrift für Papyrologie und Epigraphik 156:23–26.
Ferrari, F. 2007. Una mitra per Kleis: Saffo e il suo pubblico. Biblioteca di Materiali e discussioni per l’analisi dei testi classici 19. Pisa.
Geissler, C. 2005. “Der Tithonosmythos bei Sappho und Kallimachos: Zu Sappho fr. 58 V, 11–22 und Kallimachos, Aitia fr. 1 Pf.” Göttinger Forum für Altertumswissenschaft 8:105–114.
Gronewald, M., and R. W. Daniel, 2004. “Ein neuer Sappho-Papyrus.” Zeitschrift für Papyrologie und Epigraphik 147:1–8.
Hardie, A. 2005. “Sappho, the Muses and Life after Death.” Zeitschrift für Papyrologie und Epigraphik 154:13–32.
King, H. 1989. “Tithonos and the tettix.” Old Age in Greek and Latin Literature (ed. T. M. Falkner and J. de Luce) 68–89. SUNY Series in Classical Studies. Albany.
Lidov, J. B. 2002. “Sappho, Herodotus, and the Hetaira.” Classical Philology 97:203–237.
Lobel, E., ed. 1925. ΣΑΠΦΟΥΣ ΜΕΛΗ: The Fragments of the Lyrical Poems of Sappho. Oxford.
Lundon, J. 2007. “Il nuovo testo lirico nel nuovo papiro di Saffo.” I papiri di Saffo e di Alceo (ed. G. Bastianini and A. Casanova) 149–166. Studi e Testi di Papirologia, N.S. 9. Florence.
Miller, A. M., trans. 1996. Greek Lyric: An Anthology in Translation. With introduction and notes. Indianapolis.
Page, D. L. 1955. Sappho and Alcaeus: An Introduction to the Study of Ancient Lesbian Poetry. Oxford. Corrected reprint 1959.
Pfeijffer, I. L. 2000. “Shifting Helen: An Interpretation of Sappho, Fragment 16 (Voigt).” Classical Quarterly 50:1–6.
Pretagostini, R. 2003. “L’epigramma per Nicomache (Posidippo, P.Mil.Vogl. VIII 309, IX 1–6).” Il papiro di Posidippo un anno dopo. Atti del convegno internazionale di studi: Firenze 13–14 giugno 2002 (ed. G. Bastianini and A. Casanova) 121–128. Florence.
Preisshofen, F. 1977. Untersuchungen zur Darstellung des Greisenalters in der frühgriechischen Dichtung. Hermes Einzelschriften 34. Wiesbaden.
Rawles, R. 2006a. “Musical Notes on the New Anonymous Lyric Poem from Köln.” Zeitschrift für Papyrologie und Epigraphik 157:8–13.
———. 2006b. “Notes on the Interpretation of the ‘New Sappho’.” Zeitschrift für Papyrologie und Epigraphik 157:1–7.
West, M. L. 2005. “The New Sappho.” Zeitschrift für Papyrologie und Epigraphik 151:1–9.
[ back ] 2. Rawles 2006b and Geissler 2005 have explored the possibility of continued singing, but within the limits of West’s supplements and the emphasis he introduces on the contrast between Sappho and the paides; they depend on finding an allusive relationship to the Hymn to Aphrodite to argue that Sappho’s continued ability to sing at least mitigates her sense of loss as she compares her diminished abilities to theirs (Geissler stresses the similarity of the paides and the Muses). These articles were not available to me when I first developed the thesis of this paper for an oral presentation, but their agreement on the prominence of singing reinforces my conviction that the question of the supplements needs to be reopened. Another recent commentator, Ferrari 2007:179–185 also recognizes that the Singer is still singing and, in contrast to the paides, no longer dancing. He retains the focus on the misery of old age, considering the poem a “pathography” (like fr. 31V), in which she accepts its necessity and finds consolation in the implied performance situation: “dalla gioia di poter ancora guidare con la musica e il canto un cora di ragazze adolescenti… (2007:185). [ back ]
[ back ] 3. There is a full discussion in Page  1959:112–116; the most important testimonia are 1, 2, 28–32 Campbell (1982 ). See also Lobel 1925:xii–xv.
[ back ] 4. Rawles has suggested (2006a) that the initial line refers to Hermes (rather than Eros, as Clayman and some others take it) but argues for the Hermes of the Homeric Hymn, inventor of the lyre (“breathless” in contrast to the flute), not Hermes psychopompos, the escort of the “breathless” to the underworld, as seems to me more likely. Although the evidence allows only speculation on who is meant here, the mention of Orpheus in a Hellenistic poem would imply by itself all the motifs involved. [ back ]
[ back ] 5. The evaluation of old age in the poets is studied in detail by Preisshoffen 1977 (see also Brandt 2003:29–38). Both the contrast with youth and the connection with death are typical, and they are not mutually exclusive, although only Mimnermus finds old age worse than death. Old age is commonly associated with the fact of mortality, as one of the markers of the boundary between men and gods.
[ back ] 6. Hardie 2005 locates in the poems of the Cologne papyrus and elsewhere in Sappho (especially in 55, to be discussed below, and 147) evidence for the possibility of an actual cult of the Muses, which promised special treatment after death. His treatment of the New Poem is based on West’s, but even so, he stresses that its emphasis is on mortality. [ back ]
[ back ] 8. Ferrarri 2007:188 reports a forthcoming study by E. Puglia which introduces the theme of youth too to this fragment by joining fr. 87 (13) LP (not in V) to it, despite a gap of 7 syllables.
[ back ] 9. The imperative of “sing” occurs in fr. 21 and has been supplied in fr. 22; singing as accompaniment of a wedding procession appears in fr. 27 and 30.
[ back ] 10. Reproach and baseness certainly seem to be subjects in fr. 3, but the speaker seems to be expressing regret more than any harshness of her own. See Lidov 2002 who argues against the imputation of any reproach to Sappho’s brother or to “Doricha” in any of the poems, particularly fr. 3 and 15. [ back ]
[ back ] 11. For the syntax of an opening with a dative of possession and no copula, cf. Sappho 110 V.
[ back ] 12. Both difficulties were noted ad loc. in the e.p. The genitive was assumed by Diehl and Campbell in their editions of Sappho. The alpha of χελύνναν has a macron in P.Oxy. 1787, but such macrons appear irregularly and are not useful guides; there is a macron over the alpha of an accusative singular in 58.22, and of a genitive plural in 71.3. [ back ]
[ back ] 13. The supplement for the start of line 3 printed in the text in the Introduction, ἔμοι δ’, supposes a change in topic, from the paides to the singer; Di Benedetto’s earlier suggestion of the causal particle, not mentioned in the app. crit., seems to me clearer, although an adversative does not alter the logic.
[ back ] 14. The text is cited by Hephaestion, who gives the nom. sing. ὄγμος, in which case the phrase could be a rude specification of the loss of youthful bloom; see the note of Campbell (1982 ):ad loc.
[ back ] 15. It is not clear whether ἀπάλαν in 94.22V modifies something specific in the following gap or generally characterizes νε]α̣νίδων in the next line (Theander; v. app. crit). Alcaeus 39a.3V does have γῆρας two lines before ἀπάλων. For further discussion of the word, see also Broger 1996.91–92. [ back ]
[ back ] 17. The psychological interpretation is defended at length by Bernsdorff 2004; he takes it to be an innovation of Sappho’s. His argument relies on the emotional uses of forms of βαρὺ στενάχειν in Homer, and on the conception in other poets that anxieties are typical of old age. He does not examine Lesbian uses of θυμός. Hellenistic usage differs. [ back ]
[ back ] 18. I have in mind particularly the way the example of Helen in fr. 16 serves logically to illustrate the gnome and applies to Sappho’s desire for Anactoria, but rhetorically it also serves to make Helen a paradigm for the beauty of Anactoria, described in the final surviving stanza; see Pfeijffer 2000.
[ back ] 19. Ferrari 2007:181 points out the correspondence of στεναχίσζω and Anacreon’s ἀνασταλύζω (395.7, quoted above); note that in Anacreon also the verb introduces the theme of death as the problem of old age.
[ back ] 20. The only concessive participle I have found in Sappho is 1.24, κωὐκ ἐθελοίσα—which not only has the particle, but is also not needed for the basic logic of the conditional sentences to which it is attached; it abbreviates volens nolens as a supplementary emphasis. The necessity of the concessive meaning is stressed by West (2005b:6) to avoid banality under his interpretation. As Rawles (2006b:2) points out, in West’s reading it is the similarity of the “feelings and emotions” supposed for Tithonus and Sappho that are the grounds for the paradigm.
[ back ] 21. Preisshoffen 1977:13–19 provides a nuanced analysis of the myth of Tithonos in the Hymn, where he is situated between the good fortune of immortalized Ganymede and the long-lived but mortal dryads. Preisshoffen emphasizes the myth’s importance for the theme of the necessity of mortality. King 1989 offers a structuralist study of the myth, in which all versions come into play simultaneously. She observes that the Hymn’s specification of his voice (237 φονὴ ῥεῖ ἄσπετος) may reflect what is in other versions a contrast between old age and the continuing song of the cicada. The Hellenistic use of the myth, in which the transformation to a cicada figures prominently, is explored by Acosta-Hughes and Stephens 2002:251–253 and Geissler 2005:11–12.
[ back ] 22. Di Benedetto (2006:11) argues that the four lines following the New Poem in the Oxyrhynchus papyrus (=58.23–26), which contain the quotation from Athenaeus, could be an independent composition and suggests that it may be part of a dialogue. But Lowell Edmunds (2006), in addition to observing the absence of a return from the myth to the occasion, has pointed out that there is no parallel for the use of the imperfect ἔφαντο to introduce an exemplum. A dialogue might provide a specific reference for the past tense, but in the absence of either a generic model or a certain local explanation, the problem of the poem’s occasion or genre can only remain open. [ back ]
[ back ] 24. Bastianini and Gallazzi 2001 ad loc. That view is defended by Pretagostini 2003. But given the combination of death and Sapphic singing, it is tempting to see an even more specific allusion in πρὸς ἑῷαν.