Ten
Who Sang Sappho's Songs?
André Lardinois
In recent years the traditional division of Greek lyric into exclusively choral or monodic poets has been called into question. The main subject of inquiry has been the choral poets, like Pindar, Stesichorus, and Alcman. It has been argued that some or even most of their poetry was not performed by choruses, but by the poets themselves or other soloists.[1] In this article I want to focus attention on one of the allegedly monodic poets: Sappho. I will argue that there are among her fragments more chorally performed songs than so far has been acknowledged.
Other scholars already have voiced some uneasiness with the traditional picture of Sappho as a monodist. Hermann Fränkel believed that "among the Lesbians too, then, there were songs fairly close to choral lyric," like
I would like to thank Andrew Ford, Richard Martin, Jan Bremmer, Claude Calame, and Dirk Obbink for their valuable suggestions at different stages of this article, which began as a term paper I wrote for Andrew Ford's 1989 Graduate Seminar on Sappho and Alcaeus at Princeton University. A shorter, oral version was delivered at the 1992 APA conference in New Orleans. Some of the arguments are repeated from my article on Sappho, "Subject and Circumstance," which in many ways complements this article by arguing that Sappho was involved in the setting up of young women's choruses.
Fragments and testimonia of all lyric poets, including Sappho, are cited from Campbell's edition in the Loeb Classical Library, Greek Lyric , unless noted otherwise. Elegists and iambic poets are cited according to West, Iambi et Elegi Graeci .
Sappho fragment 16, "which meditates and argues like choral poetry." More recently, Claude Calame has suggested that Sappho's circle was organized as a young girls' choir which sang or danced to songs composed by Sappho, and Judith Hallett has declared that "many of Sappho's fragments thought to be personal, autobiographical statements might in fact be part of public, if not marriage, hymns sung by other females."[2] Yet, overall the traditional picture has prevailed that Sappho composed songs, essentially about herself (her own emotions), to be performed by herself.
I will first take a look at some of the evidence about Sappho's work. Next I will discuss the applicability to her work of the traditional distinctions between choral and monodic poetry. Special attention will be given to the use of the first-person singular and plural in early Greek poetry, which will also allow us to take a closer look at some of Sappho's fragments (frs. 94 and 96). In a final section I will review the other major fragments and show that they can be interpreted as being performed with the help of choruses.
1. The Ancient Evidence About Sappho's Work
It is commonly acknowledged that at least some of Sappho's poetry was choral. One of Sappho's books, probably the ninth, in the Alexandrian edition of her poems consisted wholly of epithalamia or wedding songs,[3] at least some of which were meant to be performed by age-mates of the bride.[4]
[2] Frankel, Early Greek Poetry 186 nn. 45, 172; Calame, Les chœurs 1:127, 368-69; Hallett, "Sappho and Her Social Context," 141.
[3] A better term is hymenaioi. Epithalamia is the name Hellenistic scholars gave to these poems and is not attested before that period. Originally it referred to songs sung in the evening outside the marriage chamber, but Sappho's songs cover a number of occasions on the wedding day, including the wedding procession and the banquet: see Schadewaldt, Sappho 32-58; Muth, "Hymenaios" 38-40; Page, Sappho and Alcaeus 119-23; Bowra, Greek Lyric Poetry 214-23; and Contiades-Tsitsoni, Hymenaios 68-109. Hymenaios , on the other hand, is an archaic Greek term that was used for all types of song: as Aeschylus suggests, from bridal bath to bridal bed (PV 555-56). See Muth on this passage and the two terms in general, and more recently Contiades-Tsitsoni, esp. 30-32.
[4] Page, Sappho and Alcaeus 112 ff.; Campbell "Sappho" 162. Hymenaioi appear to have been primarily choral songs, performed by a chorus of young men and/or young women (Muth, "Hymenaios" 36) and there are clear indications of choral performance in Sappho's wedding songs, for example the dialogue-form of fr. 114 (Page 119 n. 1, 122,) and the refrain in fr. 111. Still, Homer (Od . 4.17-19) pictures a song performed at a wedding feast that is sung by a soloist and danced to by two tumblers, and Hague has compared Sappho fr. 115 to the eikasia games played at symposia, "in which one person ridiculed another by making a ludicrous comparison" ("Ancient Greek Wedding Songs" 134; my italics). It is possible that particularly during the banquet there were solo performances of songs related to the bride and groom, with or without the accompaniment of a dancing chorus. Fr. 44, if indeed a wedding song, is a possible candidate for such a performance (Contiades-Tsitsoni, Hymenaios 107).
Another type of song that is ascribed to her is religous hymns (test. 21, 47). These need not all have been choral, but some of them appear to have been genuine choral songs, such as fragment 140a, which is composed as a dialogue between a person (or group) impersonating the goddess Aphrodite and a group of young girls.[5]
Page maintained that, apart from these poems, "[t]here is no evidence or indication that any of Sappho's poetry ... was designed for presentation by herself or others (whether individuals or choirs) on a formal or ceremonial occasion, public or private," and that "[t]here is nothing to contradict the natural supposition that, with this one small exception, all or almost all of her poems were recited by herself informally to her companions."[6] One must be wary of relying on "natural suppositions," especially in the case of Sappho, and Page's supposition is actually far from "natural," since generally when scholars find that one or more poems of an archaic Greek poet are choral, they assume that the same holds true for the other poems.[7] Snyder, more carefully, distinguishes between three types of songs: those that are purely public (the wedding songs); those with the conventional form of public poetry (e.g., frs. 1, 2, 16); and those that are purely private (e.g., frs. 31, 94, 96).[8] It is not clear, however, whether she believes that the second group was actually performed in public and/or by others than Sappho herself: "Even though we may not want to go so far as to say that these songs were meant to be performed at some specific occasion, they nevertheless seem in some way connected with familiar rituals of a public character." The question is why we should not go so far as to say that these songs were performed at public occasions, if they indeed follow "the conventional forms of public poetry."[9]
According to the Suds (test. 2), Sappho wrote nine books of "lyric songs" (
) and also "epigrams, elegiacs, iambics, and solo songs (
)." We know that the epigrams were late Hellenistic forgeries,[10] and of her iambics and elegiacs nothing has survived, but the separate mention of solo songs has caused some surprise: "how did these last differ from
[6] Page, Sappho and Alcaeus 119. He has been followed by most interpreters, most recently Parker, "Sappho Schoolmistress" 331.
[7] See Davies, "Monody." Like Davies, I do not favor this approach and therefore will refrain from making such an argument.
[8] Snyder, "Public Occasion and Private Passion."
[9] Ibid., 3, 10.
[10] Campbell, Greek Lyric 1:xiii; for their text, 205.
her lyric poetry?" asks Campbell in a footnote to his edition of Sappho's fragments and testimonia.[11] I hope to show that this is more than just a rhetorical question.
As regards the actual performance of Sappho's songs, we have very little information. In later antiquity we hear of performances of her songs both by individuals (a boy; test. 10) and by groups of girls and boys (test. 53), but we do not know how this relates to the original performance context.[12] Some of Sappho's poems seem to have been intended to be recited by herself, like fragment 1, in which she mentions her own name, but such clarity is exceptional: the only other fragments in which Sappho mentions her name are 65, 94, and 133. We cannot be absolutely certain that she sang even these songs herself. Alcman composed several songs (frs. 17, 39, 95b) in which he mentions his own name but which nevertheless may have been performed by a chorus,[13] and both in Pindar's epinikia and later in the parabaseis of Aristophanes (e.g., Nub . 518-62) the chorus or chorus leader can speak in the name of the poet/composer. Sappho further mentions in her poetry that other women sang songs about each other or Aphrodite, and in one case she alludes to a song dance (
) of Atthis.[14] Were these their own compositions or did Sappho compose these songs for them, the same way she composed the wedding songs or the hymns?
The testimonial tradition about Sappho is not uniform either. Horace pictures her as plucking the lyre while singing to herself about her girls (Carm . 2.13.24-25 = test. 18), whereas an anonymous poet in the Anthologia Palatina describes her as leading a dancing chorus of Lesbian women, "her golden lyre in hand" (AP 9.189 = test. 59).[15] Philostratus (Imag . 2.1.1-3), finally, is reminded of Sappho when he sees a picture of a female director (
) leading a band of singing girls (
).[16] We thus have witnesses
[11] Ibid., 7.
[12] Other references to the performance of Sappho's songs in later times are Plut. Quaest. cony . 622c, 711d, but the number of singers is unclear here.
[13] Calame, Alcman 362 f.
[14] Frs. 21, 22, 96.5. In fr. 96.4 the Lydian woman is further said to have compared Atthis to a goddess and it is not unlikely that she did so in a song.
and/or fragments for at least three different types of performances: Sappho sings, with or without her chorus dancing; full choral performances; performances by one of her companions.
Sappho composed songs about young women,[17] and she probably composed her wedding songs for performances by them. The Adonis hymn (fr. 140a), with its reference to 
, may represent another type of song Sappho composed for these girls. Both types of song would have been performed in public.[18] It is generally assumed that Sappho sang the other songs herself in the small circle of her companions,[19] but there is really no evidence for this. No one in antiquity says so, not even Horace, who makes Sappho sing to her own lyre in the underworld.[20] This idea seems to have originated in the French salons of the seventeenth and eighteenth century, whose members believed to have found in Sappho a kindred spirit.[21] In the nineteenth century Sappho's "salon" was interpreted as a school for young girls and, more recently, as a female thiasos , but, as far as we know, there existed no literary "salons," schools for girls, or private thiasoi in archaic Greece.[22]
Modern scholars sometimes make reference to fragment 160 in which the speaker says something like: "I shall now sing these songs beautifully to the delight of my companions" (
).[23] We cannot be sure that this is what Sappho actually said
[18] Page, Sappho and, Alcaeus 119; Campbell, "Sappho" 162.
[19] E.g., Page, Sappho and Alcaeus 119; van Erp Taalman Kip, "Einige interpretatieproblemen" 340; Stigers [Stehle], "Sappho's Private World" 45; Burnett, Three Archaic Poets 209 n. 2.
[20] Horace's picture is reminiscent of Achilles playing all by himself on his lyre in Iliad 9 (on which, see 170 n. 102) rather than of a performance before a group.
[21] Compare Saake, Sapphostudien 15-16; DeJean, Fictions of Sappho 43f., 135-36.
[22] See Lardinois, "Subject and Circumstance" 63-64, 75-79. See Parker, "Sappho Schoolmistress" 339, for some examples of scholars who interpreted Sappho's "circle" as a thiasos . Nineteenth-century scholars: Welcker, Sappho 97; Wilamowitz, "Die griechische Literatur" 26.
[23] E.g., Lanata, "Sul linguaggio amoroso" 66; Winkler, "Double Consciousness" 165; Burnett, Three Archaic Poets 216; Campbell, "Sappho" 162.
(
does not fit the meter) or that Sappho herself is the speaker, but even if this were the case, to whom would she address these words? She does not use a second-person plural (as the speaker does in fr. 141) and therefore may be speaking about her companions in the presence of a larger audience. In the Homeric Hymn to Apollo the speaker similarly asks his dancing chorus, which consists of young girls (
), to remind others how much his singing delighted them (
; 170). If the Anthologia Palatina (test. 59) reflects an authentic tradition and Sappho sometimes performed her songs in public while her chorus danced, fragment 160 may have been part of such a song.[24]
Other possible evidence is fragment 150, in which Sappho calls a house (if 
is the correct supplement for the unmetrical 
) that of "the servants of the Muses" (
).[25] According to Maximus of Tyre, who has preserved the fragment for us, Sappho spoke these words to her daughter, which is probably why most scholars assume that she is speaking here about her own house.[26] Yet, even if this were the case, the fragment does not say that it was in her house, and only in her house, that Sappho and her companions performed their songs. We do not know what she means by the word 
, but we encounter the same term again in a Boeotian inscription where it refers to a theater group.[27] I do not want to deny that Sappho and her companions may have recited songs to each other at her house, but this is by no means evident and, instead, there are good reasons to believe that Sappho composed her songs for public performances.
The closest parallel to Sappho's circle is the groups of Spartan women for whom Alcman composed his songs.[28] These are young girls, at the brink of marriage, who come together to sing in choruses and perform certain rituals. The Spartan evidence strongly suggests that these groups were trained for
[24] For this type of performance in which a soloist sings while the chorus dances, one may compare Demodocus's song about Ares and Aphrodite, which is sung by Demodocus and danced to by a group of young Phaeacians (Od . 8.262-64), the wedding song in Od . 4.17-19, or the execution of the Linos song in Il . 18.569f. For some applications of this type of performances to other archaic Greek poets, see Davies, "Monody" 62-63.
[26] E.g., Welcker, Sappho 97; Wilamowitz, Sappho und Simonides 73; Kranz, Geschichte der griechischen Literatur 88; Burnett, Three Arctic Poets 211.
[27] IG VII. 2484. See Lanata, "Sul linguaggio amoroso" 67; Galame. Les chaurs 1:367.
[28] Merkelbach, "Sappho und ihr Kreis" 3; Calame, Les chaurs 1:27, 367f.; Lardinois, "Lesbian Sappho" 26-29.
public performances, not for the privacy of the poet's house. This does not necessarily mean that the gifts always had to do both the singing and the dancing. A fresh look at Alcman's poetry might reveal that not all of his "maiden songs" were like fragments 1 and 3, that is, sung by the whole chorus. There is the suggestion of exchanges between the choir and the poet, and of prooemia sung by Alcman himself.[29] It is also possible that such a "monodic"-looking fragment as fragment 59a was actually sung by the poet while his maiden chorus danced.[30] I want to argue for such a variety of performances in the case of Sappho's poetry as well.[31]
Finally, we may question whether any archaic Greek poet, male or female, would have composed poetry for something as intimate as a private group of young, adolescent women. Parallels have been drawn between Sappho's circle and the hetaireia of the Lesbian poet Alcaeus,[32] but there is quite a difference between a gathering of politically active, adult men and a group of young girls. If Sappho's circle had a counterpart in any male organizations, it was in juvenile bands of boy initiates, not in adult clubs of aristocratic warriors. Such groups were, like Alcman's choruses, trained for performances in public.[33]
2. Sappho and the Distrinctions between Choral and Monodic Poetry
Critics of the traditional division between choral poetry and monody have pointed out that it is not very old. Plato (Leg . 764d-e) is the first to mention it, and he speaks about the performances of songs, not about their monodic or choral character.[34] The archaic Greeks themselves do not seem to have been particularly interested in the distinction, for a number of archaic Greek genre names could refer to a poem sung by a soloist or a choral song, such
[29] On exchanges, see Rosenmeyer, "Alcman's Partheneion I" 338, who points to frs. 26, 38, 39, and 40; on prooemia , see Segal, "Alcman" 128.
[30] On the monodic character of Alcman fr. 59a, see Davies, "Alcman fr. 59a P.," review of Calame 387-88, and "Monody" 54-55.
[31] Besides these different kinds of "maiden songs," Alcman, just as Sappho, also composed marriage songs: for the evidence see Contiades-Tsitsoni, Hymenaios 46-67, with Muth's prudent remarks (review of E. Contiades-Tsitsoni 587).
[32] Most, "Greek Lyric Poets" 95-96; Burnett, Three Archaic Poets 209; Gentili, Poetry and Its Public 81.
[33] Bremmer, "Adolescents" 138; Buxton, Imaginary Greece 23-24. It is is worth noting that other female poets in the classical period were credited with having composed songs for young women's choruses as well: Calame, Les chœurs 2:174; Snyder, The Woman and the Lyre 40, 50 (Corinna), 54-55 (Praxilla), 60 (Telesilla).
[34] Davies, "Monody" 57, Lefkowitz, "Who Sang Pindar's Victory Odes?" 191. Cf. Fäber, Die Lyrik 1.16, and Harvey, "Classification" 159 n. 3.
as the skolion ,[35] the epinikion ,[36] and the hymenaios .[37] The differences between choral and monodic poetry that one finds most often cited concern
1. their metrics: the meters of choral songs are said to be more elaborate, the strophic structures longer.
2. their language: monodic poets stay closer to their local dialects, while choral poets make use of a more artificial language, based on Doric and the epic.
3. their contents: choral poets are less intimate and personal than monodic poets.[38]
Note that these differences are all relative: they may be less the result of the number of performers of the song than of the individual poet, the subject of the song, the audience, and so on.
I will now examine how these distinctions relate to Sappho's poetry:
1. There can be no question of any clear, metrical division between Sappho's choral and monodic poetry since we possess wedding songs (frs. 27, 30), as well as supposedly monodic songs (fr. 1), in the same Sapphic stanza.[39]
[35] Harvey, "Classification" 162. Dicaearchus (fr. 88 Wehrli) distinguished three different types of drinking songs: first a song sung by all the guests together, then stanzas sung by each of the guests in turn, and finally songs sung by the experts.
[37] See n. 4 above, p. 151. In general, archaic Greek terms for poetry mark the occasions of the song rather than any formal features: see Calame, "Réflexions" 118; West, Studies 7, 23; Fowler, The Nature of Greek Lyric 90; Gentili, Poetry and Its Public 36.
[38] E.g., Bowra, Greek Lyric Poetry 6; Kirkwood, Early Greek Monody 10-11 (who makes an exception for Sappho: 10); Most, "Greek Lyric Poets" 89; Campbell, "Monody" 161. Segal, "The Nature of Early Choral Poetry" 125, is already more cautious. See for the history of the distinction Davies, "Monody" 58-61.
Sappho also used the dactylic hexameter for wedding songs (frs. 105, 106, 143) and for such a song as fragment 142, believed to be the opening line of one of her "amorous" songs.[40] The idea that choral meters are always complex is based in large part on Alcman's first partheneion, which has a fourteen-line stanza. Yet not all dance songs need have been so intricate as this one, which was clearly composed for a solemn occasion.[41] One should note that Alcman composed three-line stanzas as well.[42]
2. Page in his commentary on Sappho and Alcaeus followed Lobel in his assessment that Sappho wrote in her Lesbian vernacular, "uncontaminated by alien or artificial forms and features," with the exception of some "abnormal" poems.[43] However, this distinction, which has recently been disputed,[44] does not correspond to a division between her choral and supposedly monodic songs. Some of the "abnormal" poems appear to be monodic (notably fr. 44), while some wedding songs are as "uncontaminated" as her supposedly monodic songs (frs. 27, 30). Again, we should be aware of other circumstances that can determine the use of, for example, epic diction. Thus it is to be expected that Sappho is able to use more diction familiar to us from epic in poems composed in the hexameter or in a meter close to it (all "abnormal" poems). They are also more appropriate for a song in which she recounts an epic story, like fragment 44, yet no one would argue on this basis that fragment 44 is a choral song.[45]
[40] Campbell, Greek Lyric 1:157. We, of course, associate the dactylic hexameter mainly with the solo performances of Homer and Hesiod.
[41] I would argue the same for Pindar's epinikia and the tragic choruses. Dionysius of Halicarnassus (Comp . 19 = Stesich. test. 28 Campbell) maintained that "the ancient poets, I mean Sappho and Alcaeus, made their stanzas short, so they did not introduce many variations in their few colons, and they used the epode or shorter line sparingly, but Stesichorus, Pindar, and the like made their periods longer and divided them into many meters and colons for the sheer love of variety " (my italics): no mention of monodic versus choral structures.
[42] Fr. 3.3 col. iii Page/Davies and, possibly, ft. 14a: see West, "Greek Poetry" 181. This, incidentally, spoils Davies' neat division between "eastern" and "western" poets ("Monody" 63-64).
[43] Page, Sappho and Alcaeus 327. The so-called abnormal poems are frs. 44, 104a, 105a, 105c, 1066-9, 142, and 143.
[44] Hooker, Language and Text ; Bowie, The Poetic Dialect ; and Nagy, Pindar's Homer 94 n. 60. It appears that Sappho's poetry, just like that of Alcaeus, is a complicated mix of old Aeolic, epic, and her local dialect. The same is essentially true for Alcman's (choral) poetry: Calame, Alcman xxiv-xxxiv.
[45] Rösler, "Ein Gedicht," following Merkelbach ("Sappho und ihr Kreis" 17) and Frankel (Early Greek Poetry 174), has actually argued that fr. 44 was a choral (wedding) song, but on different grounds. For the latest twist in the interpretation of this poem, see Lasserre, Sappho 81-106, and Contiades-Tsitsoni, Hymenaios 102-8, who both argue that it represents a monodic wedding song (cf. n. 4 above, p. 151).
3. This brings us to the contents of Sappho's songs. Many pages have been written about the profoundly personal feelings that Sappho expresses in her lyrics. But can we be sure that these are really her own feelings? Can we be sure that any of the early Greek poems is "personal," for that matter?[46] What is "personality" in such a group-oriented society as archaic Greece? Central to the debate have been poems in which the poet clearly impersonates a character.[47] Some of these we find, interestingly enough, among Sappho's fragments as well.[48] I will not pursue this matter further here. Instead, I will focus on some of the similarities between Alcman's partheneia, Sappho's choral wedding songs, and her so-called love poems.
The discovery of Alcman's partheneia has greatly changed the perception of early Greek choral poetry. Fränkel commented on the first fragment: "the style in the second half of Alcman's maiden song is as simple as that of the monodies of Sappho; in content choral lyric is frequently as personal as monody."[49] Of course, he meant to say that the chorus , not Alcman himself, was as "personal" in this song as Sappho in her poems.[50] But if this is true and the same degree of intimacy can be found in Alcman's choral songs as in Sappho's fragments, we must allow at least for the possibility that Sappho's songs were performed by a chorus of young women, just like Alcman's partheneia and her own wedding songs.
Indeed, the same degree of "intimacy" can not only be detected in Alcman's partheneia but in Sappho's wedding songs as well. In fragment 112 Sappho has a choir of girls sing to the bride: "your form is gracious and your eyes / ... / honey-sweet; love streams over your desire-arousing face."[51] One is hard pressed to find another fragment of Sappho that is so "intimate" as this one. The similarity between Alcman's partheneia, Sappho's own
[46] On this vexed question, see most recently Slings, "The 'I' in Personal Archaic Lyric," and Jarcho, "Das poetische 'Ich.'" For some of the consequences this has for the distinction between choral and monodic poetry, see Russo, "Reading the Greek Lyric Poets" 709-10.
[47] Dover, "The Poetry of Archilochus"; West, Studies 22 f.; Rösler, review of Tsagarakis, and "Personale reale o persona poetica?"; Slings, "The 'I' in Personal Archaic Lyric" 4 f.
[48] Fr. 102 (impersonating a girl speaking to her mother), fr. 137 (dialogue between a man and a woman). For more examples, see Tsagarakis, Self-Expression 77-81.
[49] Fränkel, Early Greek Poetry 170 n. 3.
[50] Compare Bowra, Greek Lyric Poetry 32, on Alcman fr. 3: "We are left with the impression that the whole company is in love with her [Astymeloisa]."
wedding songs, and her fragments about the erotic appeal of young women strongly suggests that the latter could have been performed in public and possibly by others than herself.
If we cannot rely on any formal distinction, how then do we judge which fragments qualify for a choral performance, which for a solo one, and which possibly for a mixed mode? I suggest that we study carefully the situation described in the poems, together with any traces of the addressee and possible identification marks of the speaker. In most cases, however, too little of the poems survives to make even an educated guess as to how they were performed, and we had better accept this conundrum instead of touting these fragments as prime examples of personal lyric.
3. "I" and "We" in Sappho and Other Early Greek Poetry
There is one formal feature that can throw some more light on the possible speaker of Sappho's fragments: the use of the first-person singular or plural. It is often assumed that "I" and "we" are interchangeable in archaic Greek poetry, but the situation is in fact not as simple as that. The latest studies of the Homeric language suggest that single characters normally use a first-person singular in referring to themselves, and that instances in which they use a first-person plural are to be explained as indications that they somehow want to include one or more other persons.[52] This is the case both with the individual heroes and with the poet himself.[53]
The same holds true for the archaic Greek poets. Maarit Kaimio, who examined the use of the first person in tragic choruses, mentions three different ways in which a single poet or performer can revert to a first-person plural: to include the person addressed, to include a third person, or to include a larger group (for example the state or the whole of humanity).[54]
[52] Chantraine, Grammaire homérique 2:33-34; Schwyzer and Debrunner, Griechische Grammatik 243-44, contra Wackernagel, Vorlesungen 1:98 f. Notable exceptions are the possessive pronouns, which I will therefore leave out of consideration. See also Benveniste, "Relationships of Persons in the Verb" 201 f., on the use of the first-person plural in general.
[53] E.g., Il . 2.486 (all mortals), Od . 1.10 (poet and his public); cf. Chantraine, Grammaire homérique 2:34.
[54] Kaimio, The Chorus 30, lists five exceptions, all of which in my opinion can be explained: Solon 7.6 Diehl (= 19.6 West) and Xenophanes 2.12 Diehl (= 2.12 West) are instances of possessive pronouns (see n. 52 above); Anacreon frs. 357 and 395 are indeed troublesome cases, but "us" in fr. 357.6 can be explained as encompassing the poet and Kleoboulos (or even the whole of humanity), while in fr. 395-1 Anacreon may be suggesting that not only his own hair is turning gray but that of his audience as well. The final exception Kaimio mentions is Sappho fr. 121, but this may actually be a choral song. Stobaeus, who cites the fragment, says that it refers to the relative ages of marriage partners and it is possible that the fragment is derived from a wedding song. In any case, it is highly unlikely that Sappho spoke this fragment "in her own voice."
Choral poets, on the other hand, use the first-person plural as well as the first-person singular to refer to the group as a whole. A quick glance at the evidence shows that they actually use the first-person singular more often than the first-person plural: Alcman's partheneion fragment 3 has only first-person singulars (nine in total), in fragment 1 in the majority of cases. This is also true for the remains of Pindar's partheneia, paeans, and dithyrambs (Paean 6.128 is an exception). In tragedy, according to Kaimio, there is also a preponderance of the use of the first-person singular in self-references of the chorus. This use of the first-person singular by a chorus can be explained in several ways: the chorus is perceived as one body, or each of its members is believed to be speaking for him- or herself, or the first-person singular represents the experiences of another person (e.g., Sappho) with whom the chorus identifies itself.[55]
In other words, where the number of speakers is concerned, the first-person plural is marked and the first-person singular unmarked in archaic Greek poetry.[56] A first-person singular can refer to a soloist or a chorus in virtually all circumstances, but a first-person plural only to a chorus or a soloist who wants to include others. It is therefore possibly revealing to study the use of the first-person plural in Sappho's poems. In fragments 27 and 30 (two wedding songs) and fragment 140a.1 (the hymn for Adonis) the speaker refers to itself with a first-person plural and is therefore, most likely, a chorus. By analogy, fragments 6, 19, and 121 are probably spoken by a chorus as well. Fragments 5, 21, 24a, 38, 147, and 150 are either spoken by a chorus or by a soloist (not necessarily Sappho) who wants to include one or more other persons.
Among the major fragments in Sappho's corpus there are two that make extensive use of first-person plurals: 94 and 96. In fragment 96 (a song for Atthis about a woman in Lydia), the study of the first-person speaker can be combined with an examination of the situation described in the poem. Before taking a closer look at this poem we must determine, however, where exactly it ends. Some scholars have suggested that the poem ends at line 20,[57] but the echo of lines 4-5 in line 21 makes it quite clear that the poem continues.[58] Besides, the strophes 24-26 and 27-29 (and perhaps 21-23, if
[55] Kaimio, The Chorus 251; cf. Calame, Les chœurs 1:436-39.
[56] For the terms "marked" and "unmarked," see Crystal, Dictionary of Linguistics 211-12; cf. Nagy, Pindar's Homer 5-8, and Martin, Language of Heroes 29-30.
[57] Theander, "Studia Sapphica II" 67, followed by Schadewaldt, Sappho 120, and Kirk-wood, Early Greek Monody 118. Page, Sappho and Alcaeus 95 n. 2, is skeptical of any break.
[58] Saake, Zur Kunst Sapphos 174; Burnett, Three Archaic Poets 311; and Gentili, Poetry and Its Public 83.
The first thing to be noticed is the persistent use of the first-person plural by the speaker:. . 
(3), 
. [. .] (18), 
(21), and presumably 
in line 27. Of the verb ending in line 3 we cannot say very much, except that the "we" contained in it contrasts itself with the "you" in line 4, who is probably Atthis.[60]
in line 18, probably the subject of the infinitive 
, could be an inclusive "we" (speaker or speakers + Atthis), as 
in line 27 seems to be ("and for us ... she [Aphrodite] poured nectar"). But this can hardly be the case with 
in line 21. Again there is a contrast between the speaker ("we") on the one hand and Atthis on the other: "it is not easy for us to rival goddesses in loveliness of figure, but you have...."[61] Burnett comments about these lines that "the singer praises Atthis with the voice of a group" and "[t]he plurality is undoubted, and more important, the playful self-denigration—so like that of the girls of Alcman's Partheneion (or Theocritus's Helen )—is a sign that the group here hails Atthis as its leader,"[62] but she does not draw the obvious conclusion that the speaker is therefore most likely a group.
These words are in many ways reminiscent of Alcman's first partheneion, in which the chorus compares its leader, Hagesichora, to goddesses (though falling short of an equation; 96 f.) and her companion, Agido, to the Sun (41), while at the same time playing down their own beauty (64 ff.) and singing talents (85-87; cf. 100-101). To Hailer goes the credit of first having noticed the agonistic quality of Sappho fragment 96 and its resemblance to Alcman's partheneia.[63] Not only is the plural speaker of the poem and the way it contrasts itself to Atthis and the Lydian woman suggestive of a choral performance, but also the actions described in the poem. The woman overseas is thought of as dancing in Lydia right
[60] Page, Sappho and Alcaeus 92; Campbell, Greek Lyric 1:123 n. 1; Burnett, Three Archaic Poets 302-3; Hague, "Sappho's Consolation for Atthis" 29.
[62] Burnett, Three Archaic Poets 312.
[63] Hallett, "Sappho and Her Social Context" 140.
now,[64] while in the past she enjoyed the singing and dancing (
; 1.5) of Atthis. It would be very effective to think of the speakers of this poem as performing a song dance at the same time as the woman in Lydia and like Atthis in the past.[65]
In fragment 94 the use of "we" is more complex. Sappho is probably singing the song herself since her name is mentioned in line 5, although we cannot exclude the possibility of another soloist or a chorus impersonating her (see above). In the first line, Sappho or the girl who left her speaks in the first-person singular (
, 1; 
, 2).[66] In lines 4-5 the girl speaks (again) and uses a first-person plural (
; 1. 4). At first it might seem that this first person refers just to herself ("o, how we suffer"), but the echo of these words in line 11 (
) makes it clear that she is probably speaking both for Sappho and for herself.[67] The first-person plural in line 8 (
) is exclusive and probably refers to Sappho and her companions;[68] "we" in line 26 (
) refers again to Sappho and the girl, or to Sappho, the girl, and her companions. These companions, together with the girl and Sappho, may have formed the chorus that is mentioned at the end of line 27.[69] Their inclusion in line 8 strongly suggests that they were present at the performance of this song too, either in the audience (as commonly envisioned) or as a chorus supporting Sappho while she was singing.
The whole poem, or at least the preserved part (Sappho's speech to the woman who leaves her), is, I would suggest, concerned with choral performances. Most of the "pleasant things" of which Sappho reminds her, the stringing of flower wreaths (12 f.), putting on garlands (15 f.), wearing
[66] Gomme ("Interpretations" 255-56), Burnett ("Desire and Memory" and Three Archaic Poets 292), Snyder (The Woman and the Lyre 26), and Greene ("Apostrophe and Women's Erotics" 239-40) assume that the girl speaks the first line, contra Wilamowitz (Sappho und Simonides 50), Page (Sappho and Alcaeus 82), Saake (Zur Kunst Sapphos 189), and Robbins ("Every Time I Look at You"), who opt for Sappho.
[68] Page, Sappho and Alcaeus 78.
perfumes (18 f.), and going to holy places (25, 27), where there is a "chorus" (? 
; 27) and "sound" (
; 28), agree with the activities of a chorus; and one can even read a linear progression into them, starting with the preparations and leading up to musical performances at temples and other places.[70] In that case Sappho would be reminding a girl of previous performances perhaps at the very moment that she and her choir, of which the girl no longer was part, were performing again a song dance, just as in fragment 96.[71]
4. Some Major Fragments: Fragments 1, 2, 5, 16, 17, 31, 58, and 95
So far I have provided positive arguments why certain fragments of Sappho probably were composed for choral presentations. In the following paragraphs, dealing with some of the other major fragments, I will allow myself more latitude. I will reverse Page's "natural supposition" and consider if there is any evidence or indication that these songs may have been performed with the help of choruses.
Fragment 1 was most probably sung by Sappho herself or by someone impersonating her: her name is mentioned in line 20. It is possible, however, that she was accompanied by a group of dancers, just as in fragment 94. West has argued that Sappho deliberately left the name of her beloved unmentioned so the song could be performed on different occasions.[72] This certainly would depersonalize the song. It would also lend special significance to the idea of the repetition of her love feelings in the poem (with every new performance there is the pretense of a new love).[73]
[71] Fr. 94 has been identified as a "farewell song," which involves memories of previously shared experiences: see most recently Rauk, "Erinna's Distaff ."
[72] West, "Burning Sappho" 310.
Fragment 2 is an obvious candidate for a choral performance either by the chorus itself or by Sappho and her chorus. If 
(hither ... to this temple) is the correct reading in line 1, and there seems to be no better alternative,[74] we are probably present at a real shrine, however dreamlike this shrine is subsequently represented.[75] Athenaeus quotes the final lines of our fragment: "Come, Cypris, pouring gracefully into golden cups nectar that is mingled with our festivities,"[76] and adds what appears to be an adaptation of Sappho's subsequent line: "for these my companions and yours" (
). If this was still part of Sappho's poem, the hetairai associated with the speaker were probably present at the scene as well.[77] We might add that lines 13f., about Aphrodite pouring nectar for the participants in the festivities, is reminiscent of fragment 96.26f., where the speaker, whom I identified as a chorus, remembers how Aphrodite poured nectar for them and for Atthis.[78] On the basis of some broad similarities I would argue for a similar interpretation of fragment 17 (the so-called "Hymn to Hera"). Here the singers may be mentioned in line 14 (
).[79] These two fragments together with fragment 140a (the Adonis hymn) suggest that at least some of Sappho's (choral) poetry was composed for ritual occasions, not unlike Alcman's partheneia.
In fragment 5, a poem about her brother Charaxus, Sappho uses, after an initial 
(? l. 1), the first-person plural (
; 7), probably to include
[74] See Page, Sappho and Alcaeus 36; see ad loc. Voigt, Sappho et Alcaeus , and Campbell, Creek Lyric vol. 1.
[75] Page, Sappho and Alcaeus 42, and Merkelbach, "Sappho und ihr Kreis" 28: contra West, "Burning Sappho" 317; McEvilley, "Sappho Fragment Two"; and Burnett, Three Archaic Poets 261 f.
the other members of her family and/or her friends.[80] If the poem is a propemptikon or "send-off" poem, as several scholars have suggested,[81] it is almost certainly performed in public.[82] I believe that this song was sung by Sappho (or someone impersonating her) in public, while her chorus danced. The philoi included in 
in line 7 may refer to these dancers or to members of Sappho's family in the audience.[83]
Fränkel already identified fragment 16 as possibly a choral song.[84] Its opening priamel, followed by a mythical example and praise of the "laudanda," resembles the structure of Pindar's epinikia .[85] Hallett added that the isolation of a few distinctive features (Anactoria's step and face in ll. 17, 18) resembles the individual compliments paid to the chorus members in Alcman's first partheneion.[86] Segal, finally, observed that the desire of the
[81] Merkelbach, "Sappho und ihr Kreis" 24 n. 1; Bowra, Greek Lyric Poetry 210; and Snyder, The Woman and the Lyre 17-18. Governi, "Su alcuni elementi propemptici," has adduced parallels from greetings and farewell scenes both in Homer (Od . 6.180, 184-85; 8.461; 15-111-12, 128) and in other Greek poetry (Theog. 691-92) for several lines in the poem (3-4, 6-7).
[83] Ft. 2o appears to derive from a similar song. In fr. 15, Sappho strikes a more critical note about her brother. If this is meant as a satirical poem, as I assume, its delivery is again best pictured in public where it would have effect. The same holds true for those poems in which she vilifies her rivals or girls who went to them: frs. 57, 68a, 71, 131, 133a, 144, 155, 178 (?), 213. I do not exclude the possibility that some of the figures mentioned in this poetry are poetic personae, similar to the stock characters presented in Archilochus's iambics (on which see West, Studies 25-28, and Nagy, Pindar's Homer 430-31.
[84] Fränkel, Early Greek Poetry 172.
[85] For a detailed comparison, see Howie, "Sappho Fr. 16," esp. 209-14. The sirnilarity was already noted by Fränkel, "Eine Stileigenheit" 90 f., and Early Greek Poetry 186; and Bundy, Studia Pindarica 5-6. Stern's objection ("Sappho Fr. 16" 349), that the priamel is voiced too personally for choral poetry, is answered by Bundy (6 n. 19), if Pindars epinikia are choral (see n. 1 above).
speaker in this fragment to make her observations "known to every one" (
Fragment 31 can go either way. The poem certainly contains a great number of first-person singular statements, but these could refer to a chorus as well as a soloist. The emotions described can be summarized by what Alcman's chorus says about its chorus leader: 
(she wears me out; fr. 1.77).[88]
Just as in this partheneion or in Sappho fragment 96, a triangle is set up between the speaker, the girl she is in love with, and a third person with whom the girl is involved (in this case a man). Note, for example, the structural opposition between that man, who "appears to be the equal of the gods" (
1-2a), and the speaker, who in lines 15-16 "appears to be little short of dying" (
). This echo, already noted by Wilamowitz, contradicts Winkler's assertion that the man is "not an actor in the imagined scene."[89] Better Snyder: "[the man is] a foil for the exposition of the speaker's feelings; he is calmly 'godlike' in response to the woman's sweet talk and charming laugh, whereas the speaker, in the same situation, is instantly struck dumb."[90] Both in Alcman fragment 1 and in Sappho fragment 31 the rivals for the affection of the beloved are compared to gods (Alcman fr. 1.41, Sappho fr. 31.1) and they are together with the beloved (Alcman fr. 1.78-79, Sappho fr. 31.3-4), while the speakers are unable to be in her presence. In both poems the speakers are also resigned to this fact. (Sappho fr. 31 continues in l. 17 with the words "but all can be endured," 
.)
As for the occasion on which this song was performed, I would not want to exclude the possibility that it was sung at a wedding, as Wilamowitz
[87] Segal, "Eros and Incantation" 64. Burnett, Three Archaic Poets 285 n. 19, adduces two parallels for the expression, one from a Pindaric hyporchema (fr. 105a.1 Maehler) and one from an epinikion of Bacchylides (ft. 3.85).
[89] Winkler, "Double Consciousness" 179. See Wilamowitz, Sappho grid Simonides 57 (cf. Robbins, "Every Time I Look At You" 259).
declared.[91] The opening line is certainly reminiscent of the traditional makarismos of the groom.[92] Most modern interpreters, starting with Page, have discredited this view. Page's main objection is that it would be inappropriate for Sappho (or, presumably, any other speaker) to speak about the intensity of her passions for a bride on her wedding day, but this could be our modern sensitivity.[93] In fragment 112 of Sappho a chorus describes a bride in very glowing terms, and when it says that "eros streams over her desirable face" (
) it is by no means clear that this is supposed to have an effect on her husband only.[94] Snyder objects that "a wedding song must have chiefly to do with the bride and the groom, not with the speaker's passion for one of them," but as Most remarks: "It is in fact the beauty of the unnamed girl that is the burden of the poem and the justification for its composition and performance: every detail Sappho provides is designed to testify, not to the poet's susceptibility, but to the girl's seductiveness."[95] In
[91] Wilamowitz, Sappho und Simonides 58. He was followed by Snell, "Sapphos Gedicht" 82; Schadewaldt, Sappho 98; Merkelbach, "Sappho und ihr Kreis" 6; Fränkel Early Greek Poetry 176; and Lasserre, "Ornements érotiques" 22. Welcker, "Oden" 89-90, already suggested that this poem "veranlasst ist durch die Heirath einer geliebten Schülerin," and "es mag auch eine Huldigung, Preis der Schönheit in dem hohen Ausdruck dieses Entzückens versteckt sehn" ("[this poem] is occasioned by the marriage of a favorite pupil," and "there may also be a celebration, praise of beauty hidden behind the intense expression of her enchantment"). Winkler, "Double Consciousness" 178-80, compares Sappho fr. 31 to Od . 6.158-61 (Odysseus's praise of Nausicaa), which in turn has been compared to a wedding song (Hague, "Ancient Greek Wedding Songs" 136-38).
[93] Page, Sappho and Alcatus 30ff. This was also noted by Merkelbach, "Sappho und ihr Kreis" 9, and Segal "Eros and Incantation" 69 n. 19: "we know too little of what conventions on archaic Lesbos would or would not have permitted." There is some evidence to suggest that a bride might be expected to be the object of widespread erotic admiration at ancient Greek weddings: see Seaford, Reciprocity and Ritual 36 n. 25.
[94] Segal, "Eros and Incantation" 69, already drew the parallel between fr. 112 and fr. 31.
Alcman's partheneia and in Sappho fragment 112 the women are similarly praised through a declaration of the effect their beauty has on the speaker. Fragment 31, whether performed at a wedding or not, is really an enkomion .[96]
Fragment 58 is generally not considered one of the major fragments, but it is significant because, like fragments 21 and 22, it is suggestive of exchanges between Sappho (or another soloist) and the chorus. Line 11 mentions paides with beautiful gifts, either of the deep- or violet-bosomed Muses.[97] The speaker (a woman) says that she is overcome by old age and no longer able to do like the young fawns (probably to dance).[98] A similar-looking poem is preserved among Alcman's fragments. Here the speaker (Alcman himself, according to Antigonus, who preserved the fragment) addresses a group of "honey-tongued, holy-voiced girls," telling them that "his limbs no longer can carry" him.[99] I believe that Sappho in this fragment conjures up the same image and that the paides of line 11 make up the chorus that is dancing while she (or another performer) is singing.[100]
[96] Lasserre, "Ornements érotiques" 23, who argues that frs. 47 and 130 were part of similar enkomia , and further notes that the "I" person in these poems has more in common with the speaker in choral than in monodic poetry (7 n. 6, 21; cf. Nagy, Pindar's Homer 371). Burnett, Three Archaic Poets 235, and Race, " 'That Man' in Sappho fr. 31" 98-101, compare fr. 31 to Pindar's encomium for Theoxenus (fr. 123 Maehler), where there is also a third person who "meets the liquid glance that gleams from Theoxenus's eye and fails to swell with passion," whereas the speaker, I, "like wax in sun's high heat melt" (Burnett's translation). Compare also Odysseus's words to Nausicaa in Od . 6.160-61, which, just as Pindar's, are not intended as a declaration of love but as praise.
[97] See Di Benedetto, "Il tema della vecchiaia" 147-48, and Voigt, Sappho et Alcaeus ad loc. It is not unlikely that this line constitutes the actual beginning of the poem: Di Benedetto 147; Gallavotti, Saffo e Alceo 1:113. Page, Sappho and Alcaeus 129, also starts the poem on this line.
Fragment 95 portrays a conversation Sappho (or another woman) had with a woman in the past (probably Gongyla, whose name is mentioned in l. 4). This situation is reminiscent of fragment 94 and it may have been performed under similar circumstances.[101]
Most of these reconstructions are only suggestions. Ultimately it is impossible to prove that a particular song was sung by a chorus or by Sappho herself, with or without the help of choral dancers, but I hope to have shown that a choral performance of these songs is at least a serious possibility.
5. Sappho's Public Poetry
I have argued that three modes of performances can be detected in Sappho's poems, all public:
She sang while a chorus of young women danced (e.g., frs. 1, 5, 94, 95, 160?). The young women did both the singing and the dancing (most epithalamia, frs. 2, 16, 17, 31?, 96).
Exchanges between Sappho or another soloist and the group (frs. 21, 22, 58, 140a).
It is possible that, besides these more or less choral songs, there were genuine monodic songs, performed only by Sappho herself: after all, the Suda speaks of Sappho's monodies as well as lyric songs (above). If so, we must ask ourselves where these monodic songs were performed. We really know of only one occasion where more or less monodic poetry was performed in the archaic period: symposia.[102] It could be that Sappho composed some songs for symposia, as did Praxilla (see below), but I can find no trace of them in the remaining fragments, with the exception perhaps of some poems that were composed for the wedding banquet.[103]
[101] Cf. Page, Sappho and Alcaeus 85.
We may conclude that a monodic performance, by which I mean a single performer accompanying herself on the lyre and singing to a group, is not a likely option for Sappho's poetry. I have argued that Sappho's poems about young women do not fit the monodic mold any more than her wedding songs do or Alcman's partheneia. I have further argued that there are traces of a plural voice in these poems and many parallels with choral poetry (in particular Alcman). The testimonia and Sappho's own poems speak about a variety of performances but not about monodic performances (with the exception of the Suda , which mentions monodic songs in addition to nine books of other lyric songs). Finally, I disputed the idea that any archaic Greek poet, male or female, would have composed poetry for delivery to a group of young girls in the privacy of her own home, and I suggested that this view of Sappho, which is commonly accepted, first originated in the French salons of the seventeenth and eighteenth century.
By arguing for a more public delivery of Sappho's poems, I do not want to deny the important differences in tone and subject matter between Sappho and most male poets,[104] but instead of seeing this as a difference between a public (male) and a private (female) world (as do Stigers [Stehle], Winkler, and Snyder), I would like to suggest that this reflects a difference between two distinct public voices.[105] Only in this way can we make sense of the many similarities between Alcman's partheneia and Sappho's poetry. I believe that
[104] Here recent scholarship that has adopted a feminist perspective has done much to further our understanding of Sappho's poetry: duBois, "Sappho and Helen," and Sowing the Body 26-27 (with an important caveat on p. 29); Stigers [Stehle], "Sappho's Private World"; Stehle, "Sappho's Gaze"; Winkler, "Double Consciousness"; Snyder, "Public Occasion and Private Passion"; and Greene, "Apostrophe and Women's Eratics." See also Lasserre, "Ornements érotiques" 30; Hallett, "Sappho and Her Social Context" 138; Burnett, Three Archaic Poets , esp. 225-26, 288; Svenbro, "La stratégie de l'amour"; and Race, "Sappho, Fr . 16." Skinner, "Woman and Language," in my opinion, goes too far in insisting on the independence of Sappho's poetry.
[105] These public voices appear to be gender-specific as far as contents and performance are concerned, but they are not necessarily composed by only men or only women. Alcman composed "maiden songs," as did Pindar (frs. 94a-104d Maehler), Simonides, and Bacchylides (Calame, Les chœurs 2:167-74), while Praxilla was credited with writing skolia (Snyder, The Woman and the Lyre 55-56), a genre that was closely associated with the male domain of the symposium.
in the future it will be fruitful to compare Sappho's poetry more closely with Alcman's partheneia,[106] the poetry of other female poets,[107] and the public voices of Greek women in general. Anthropological studies of women's public or poetic voices in other cultures may be illuminating as well.[108] One of the public speech genres associated with women both in archaic and in rural Greece today was the lament.[109] There are echoes of this speech genre in Sappho's hymns (fr. 140a), in her wedding songs (fr. 114), and in a series of songs that are preserved among her "other" poetry.[110] No matter how one reads Sappho's songs, it is important to realize that most of them probably were intended to be performed in public with the help of choruses.
[106] For a beginning, see Calame, Les chœurs 1:361 ff., 2:94-97; Hallett, "Sappho and Her Social Context" 139-42; and Cavallini, Presenza di Saffo 17-20. The differences between Alcman's partheneia and Sappho's poetry to which Stehle ("Romantic Sensuality Poetic Sense" 147-49) and Skinner ("Woman and Language" 186-87) have pointed are noteworthy but do not measure up against the many similarities. See Lardinois, "Subject and Circumstance" 73 n. 59.
[107] Here Rauk's article, "Erinna's Distaff , "deserves special mention. He sees in Sappho fr. 94 and Erinna's Distaff a generic type of farewell addresses of women. We may also point out that Praxilla (fr. 747), like Sappho (fr. 140a), composed a hymn for Adonis, a typical women's cult (Burkert, Greek Relyion 177; Winkler, Constraints of Desire 188 f.), and that Telesilla (fr. 717), like Sappho (test. 21, 41; ft. 44a), composed a hymn to Artemis, a goddess presiding over various facets of women's lives including the initiation of young girls (Burkerr 151 and Calame, Les chœurs 1:174f.)
[108] For example, on Crete women can compete with men in witty, poetic responses (mandinadhes ), mostly of a sexual nature (Herzfeld, The Poetics of Manhood 142-46), not unlike the exchange found in Sappho fr. 137. On Madagascar, women are associated with direct, open expressions of anger (Keenan, "Norm-Makers, Norm-Breakers" 137-39); cf. Homer Il . 20.252-54 and, perhaps, some of Sappho's satirical poetry. Nagy, Poetry as Performance chap. 4, compares Sappho ft. 1 to female initiation songs of the Navajo and Apache.
[109] On archaic Greece, see Alexiou, The Ritual Lament , esp. 4-14, and Martin, Language of Heroes 87; on rural Greece today, see Alexiou, esp. 36-5[1] , and Caraveli, "The Bitter Wounding:"