Hostname: page-component-5d59c44645-jb2ch Total loading time: 0 Render date: 2024-02-22T01:15:06.325Z Has data issue: false hasContentIssue false


Published online by Cambridge University Press:  29 March 2016


With the recent publication of the new Sappho song by Dirk Obbink, we now have two largely preserved songs by Sappho apparently dealing with prayers for a brother away at sea. Both are examples of a type of song relating not to eros, marriage, or hymns to gods, but to bonds of love and duty between women and their male family; songs whose primary emotional focus is not other women, but male relatives absent from home.

Research Article
Copyright © The Classical Association 2016 

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)



I am very grateful to Professor C. Pelling, Dr A. Kelly, and the anonymous reader for this journal, for their helpful comments and corrections on this article. Any errors remaining are mine. All the translations are mine, except where indicated otherwise.


1 Obbink, D., ‘Two New Fragments of Sappho’, ZPE 189 (2014), 2949Google Scholar. References to fragments of Sappho in this article are to the edition of E.-M. Voigt, Sappho et Alcaeus. Fragmenta (Amsterdam, 1971).

2 See also D. Obbink, ‘Provenance, Authenticity, and Text of the New Sappho Papyri’, paper read at the Society for Classical Studies’ Panel: ‘New Fragments of Sappho’, New Orleans, 9 January 2015, <>, accessed 19 November 2015.

3 Lidov, J. B., ‘Sappho, Herodotus and the Hetaira’, CPh 97 (2002), 203–27Google Scholar; D. Yatromanolakis, Sappho in the Making. The Early Reception (Washington, DC, 2007), ch. 4. See also Biffi, N., ‘Le storie diverse della cortigiana Rhodopis’, GIF 49 (1987), 5160Google Scholar.

4 Obbink (n.1), 29, implies that the new song, naming Charaxus as it does, lends general credence or support to stories about Charaxus. The approach of Ferrari, F., ‘Saffo e suoi fratelli e altri brani del primo libro’, ZPE 192 (2014), 214Google Scholar, without necessarily accepting all the details of the tradition of the Charaxus/Doricha affair, accepts it in general, and aims at aligning the restoration and interpretation of the fragments with it. G. Liberman, ‘Reflections on a New Poem by Sappho Concerning Her Anguish and Her Brothers Charaxos and Larichos’, paper delivered at F.I.E.C. Bordeaux, August 2014, available at <>, accessed 19 November 2015,  accepts the truth of the tradition. The conjecture of West, M.L., ‘Nine Poems of Sappho’, ZPE 191 (2014), 6Google Scholar, of ‘Doricha’ in the fourth stanza of fragment 5 reflects an approach in which surviving fragments are interpreted and restored to align with the stories in the later Sappho tradition.

5 The translations of Herodotus here are those of Yatromanolakis (n. 3), 314–15, with minor alterations.

6 On the translation here, see below, p. 00.

7 See Yatromanolakis (n. 3), 312–16, on the way in which Herodotus develops the theme of the search for fame in particular through erotic means.

8 Posidippus 17 in A. S. F. Gow, and D. L. Page, The Greek Anthology. Hellenistic Epigrams (Cambridge, 1965), quoted by Athenaeus at 13, 596c–d.

9 Thus U. von Wilamowitz-Moellendorff, Sappho and Simonides (Berlin, 1913), 19, sees Posidippus’ irony as deploying a skilful irony ‘directed at discerning readers’. However Lidov (n. 3), 223, sees Posidippus as incapable of such irony, and the Posidippus poem as evidence that Doricha was a figure praised (not attacked) in Sappho. But the point of the Herodotus story is that Sappho is supposed to have disapproved in the strongest terms of her brother's liaison: on this point at least (i.e. Sappho's hostile reaction to the relationship) all the tradition is united. That Posidippus was aware of the Herodotus passage is suggested by his focus on the theme of fame. In any case, it is not likely that Doricha was in fact praised in Sappho: later writers are unlikely to have associated Rhodopis with a figure who was praised.

10 Cf. Ferrari (n. 4), 9. On the chronology, see Lidov (n. 3), 218–19 (‘strains the chronology, although not to the breaking point’) and the studies cited there. As the anonymous reader has pointed out to me, the chronology of Sappho is not sufficiently certain to rule out that she wrote the poem at a later date: even if Sappho was born c.620–600, she could still have written a poem about Charaxus by 570, when she was thirty or older.

11 Cf. Lidov (n. 3), 216–17.

12 See POxy. 1800 fr.1 (Sappho T 252), from a biographical writing about Sappho: ‘Sappho had three brothers: Eriygius, Larichus, and Charaxus, the eldest, who sailed to Egypt and associated with one Doricha, and spent very large sums on her’; Suda entry on Aesop and Rhodopis (= T 254) claiming that Charaxus married Rhodopis and had children by her; Ov. Her. 15.63–70, 117–20. See also D.S. 1.64.14 recording a pyramid which some say is the tomb of Rhodopis, built by her lovers at their common expense, but without mention of Sappho or Charaxus. Lidov (n. 3), 221, has a full account of the testimonia.

13 Yatromanolakis (n. 3), 322–37.

14 Kerkylas from the Suda entry on Sappho. The name is almost certainly a joke name from one of the comedies about Sappho: see H. N. Parker, ‘Sappho Schoolmistress’, in E. Greene (ed.), Re-reading Sappho. Reception and Transmission (Berkeley, CA, 1996), 147. On Phaon and Leukas, see G. Nagy, ‘Phaethon, Sappho's Phaon, and the White Rock of Leukas: “Reading” the Symbols of Greek Lyric’, in E. Greene (ed.) Reading Sappho. Contemporary Approaches (Berkeley, CA, 1996), 35–57.

15 On the use of the phrase ἐν μέλει, Yatromanolakis (n. 3), 328, compares other uses of the phrase in Herodotus, e.g. Hdt. 5.95.2: in his narrative of the struggle of the Athenians and Mytileneans for Sigeium, Herodotus notes that Alcaeus describes ἐν μέλει how he was forced to flee and abandon his arms to the Athenians – a song independently attested by Strabo (Alcaeus frag. 401B V). This suggests that Herodotus would not have used the phrase in 2.135 unless he believed that there really was some such song by Sappho.

16 Lidov (n. 3), 203 and 227–30. Arguing against Attic comedy as Herodotus’ source: Yatromanolakis (n. 3), 324–5, and see also Ferrari (n. 4), and Dale, A., ‘Sapphica’, HSPh 106 (2011), 67Google Scholar.

17 Lidov (n. 3), 203, concludes that there is no Sappho song underlying the tradition at all. In contrast, Yatromanolakis (n. 3), 328 ff., argues that there was a Sappho song about the relationship between Charaxus and Doricha, but that this has become assimilated with, or overlaid by, an oral tradition about Rhodopis.

18 As Chris Pelling points out to me, the aorist κατεκερτόμησε seems to indicate a single song with much derision, rather than many songs with derision. The word, especially with its ‘kata–’ prefix, is unusually strong, and is used on only one other occasion by Herodotus (1.129.2), for Harpagus’ taunting of Astyages, having avenged himself on Astyages for having made him eat his own son.

19 The start of the sentence (‘As for Charaxus’) shows that Herodotus has switched here to talking about Charaxus, and Rhodopis is not mentioned at all in the sentence, so it would be odd if she is meant. Ferrari (n. 4), 12, suggests that it would be better to understand here that Rhodopis is meant, on the grounds that the songs we have do not, in fact, attack Charaxus. But not even Doricha (if she is the addressee in frag. 15 [see below]) seems to be addressed in surviving songs in a way which could be described as ‘κατεκερτόμησε’, and of course there is the problem that Sappho did not attack Rhodopis at all, but, if anyone, Doricha. Lidov (n. 3), 221, suggests that later writers interpreted ‘μιν’ as referring to Rhodopis, leading to a search for a female figure in Sappho's songs derided by the poet.

20 Lidov (n. 3), 220.

21 Yatromanolakis (n. 3), 328. He also notes (333–5) that a brother of Sappho who went to Naucratis is absolutely plausible, given what we know of the involvement of Mytilene abroad at this period, and in Naucratis in particular.

22 For example Hermesianax (writing in the early third century bc and a contemporary of Posidippus) depicts not only Alcaeus expressing in his songs his desire for Sappho but also Anacreon (fl. c.450 bc) competing with Alcaeus and with Lesbian women for the attentions of Sappho: Hermeseniax frag. 7, 47–56. Posidippus may well have engaged in this sort of romantic re-imagining of the lives of Greek authors in his Aesopia – Ath. 13.596c (on which see above). Yatromanolakis is well aware of this, of course, and indeed explores this kind of Hellenistic invention in his book (n. 3), 348 ff.

23 Lidov (n. 3), 217–18: ‘[Hellenistic scholars] had the biographical information from Herodotus; the puzzle was to find a name to pin it to.’

24 See Lidov (n. 3), 222.

25 Yatromanolakis (n. 3), 330–2, believes on the basis of his autopsy of the fragment that Doricha is at least a possible restoration in Sappho frag. 15. In contrast, Lidov (n. 3), 218, accepts that the name Doricha appeared in Sappho's songs, but denies (224–5) that the name Doricha can be restored in Sappho frag. 15.

26 Lidov (n. 3), 225, notes that Sappho's brother Larichus has a name with the same rare ‘-rich-’ ending, suggesting that Doricha too may be a Mytilenean name, rather than that of a Naucratean courtesan.

27 Sappho's songs included strong invective against women: see frag. 55 (apparently to a rival in poetry who Sappho predicts will die unremembered). On Sappho's supposed rivals, see T 219 (Gorgo and Andromeda), and frag. 130 (Andromeda).

28 Thus Lidov (n. 3), 225, doubting whether Doricha is even addressed in a hostile way here. Reliable restorations are only possible for one stanza of this fragmentary song. A proposed restoration in line 5, to produce the almost identical text to line 5 of frag. 5 (‘and [may he] atone for all his previous mistakes’), relies on a hypothetical juxtaposition of a separate papyrus fragment with the text of the main fragment, which is itself driven by a desire to identify frag. 15 as a ‘brothers’ song, because of the mention of the name Doricha: see Yatromanolakis (n. 3), 330. But see, however, Ferrari (n. 4), 10–12, arguing that the extra fragment is correctly joined to lines 5–7.

29 Charaxus and Larichus probably belonged to the Mytilenean nobility: an Erygius, son of Larichus, is mentioned as a Mytilenean general of Alexander the Great: Arr. 3.11.10; D.S. 17.57.3.

30 Sappho T 203: ‘The lovely Sappho in many passages (πολλαχοῦ) praises her brother Larichus because he poured the wine for the Mytileneans in the prytaneion’ (Ath. 10.425a). See also schol. on Iliad 20.234: ‘for it was the custom, according to Sappho, for handsome young men of noble birth to pour the wine’. Depending on how much weight one puts on πολλαχοῦ, this may suggest a number of poems in which Larichus was mentioned. This depiction of a younger family member, whom the singer looks to in order to bring honour to the family, is fully in keeping with the Larichus in the new song.

31 As well as the passages cited above, see Ov. Her. 16.67, which has Sappho say of Charaxus: ‘he hates me because I gave him much good advice out of loyalty’.

32 See frag. 213A, a fragmentary lyric commentary mentioning Charaxus several times.

33 See Yatromanolakis (n. 3), 333–4. Lidov (n. 3), 232, suggests that a Sappho song composed for a wedding and containing ritual abuse of the groom could be the origin of a story that Charaxus was attacked in Sappho.

34 On Larichus, see n. 30 above. Milne, H. J. M., ‘A Prayer for Charaxus’, Aegyptus 13 (1933), 171–8Google Scholar, argues that the fragmentary song frag. 20 is also a brothers’ song, noting parallels of wording with frag. 5. With the publication of the new song, we can also see wording overlaps with frag. 20: e.g. μεγάλαι ἀήται in frag. 20.4, cf. line 11 of the new song (see Obbink [n. 1], 35, n. 7). See also Sappho frag. 213 – a fragmentary commentary which refers to a further Sappho song mentioning Larichus and Erygius. It is worth noting that Alcaeus, too, had a ‘brothers’ song (frag. 350), addressed to the brother Antimenidas, on the occasion of his return after serving as a mercenary abroad. As with Sappho's ‘brothers’ songs, a key theme is the effect of the brother's deeds on the family honour.

35 Though note Alcaeus frag. 117b, 26–7, a fragmentary papyrus song ascribed to Alcaeus (but which might be by Sappho) dealing with ships and a voyage, which contains the lines ‘whatever one gives to a prostitute (πóρνα) is the same as throwing it into the waves of the grey sea. Whoever consorts with prostitutes, this is what happens to him. [Afterwards] he suffers disgrace and much wretched misery’ (26–31).

36 See in general Parker (n. 14) on the inventions of the later tradition, and the risks to our reading of Sappho.

37 Obbink (n. 2), 4–5. On the basis of the existence of these lines, Obbink has now renumbered the lines, so that the previous line 1 is now line 4. In this article, the original line numbers are followed.

38 Obbink (n. 1), 42, suggests the infinitive ἔλθην represents an original subjunctive or optative, just as ἐξίκεσθαι does in line 7. To my mind, the parallel with line 7 makes this the better interpretation, though Obbink notes that ἐξίκεσθαι follows a verb of prayer, and the construction may be more difficult after θρύλησθα.

39 West (n. 4), 9 (supported by Ferrari [n. 4], 2–3), amends to ἐπ᾽ἄρηον in line 14, and translates δαίμων as ‘fortune’, so: ‘Those whose fortune the ruler of Olympus chooses to turn around from hardship for the better, they come out blessed and prosperous.’ But Obbink (n. 2), 6, defends ἐπάρωγον on the basis of close parallels in Theocritus 17. Further support for this vocabulary in the context of a prayer for a missing man is in Soph. El. 254 (Electra/Chrysothemis praying for the assistance of dead Agamemnon as ἀρωγός [‘helper’]). For δαίμων as ‘divine helper’, see below on Aesch. Cho. 115–21.

40 It is possible that the missing opening contained a framing narration – ‘she said to me’, ‘I replied’, as in Sappho frag. 94.

41 See Nuenlist, R., ‘Das Schiff soll unversehrt sein, nicht voll! Zu Sapphos neuem Lied über die Brüder’, ZPE 191 (2014), 13Google Scholar.

42 Hom. Od. 16. 183–7 (Telemachus takes Odysseus, restored to him by Athena, as himself a god), 16. 475–87 (suitors speculate whether Odysseus disguised as a beggar is a god; in fact, Odysseus himself will take the role of avenging deity), and 23.63 (after Odysseus has killed the suitors, Penelope, not believing in his return, ascribes it to ‘one of the immortal gods’).

43 See the texts cited in the first section above.

44 Dale (n. 16), 67–71, while accepting (against Lidov) that the later tradition about Charaxus and Larichus stems from actual content in Sappho's songs, notes certain similarities with Hesiodic paraenetic poetry, and suggests that Charaxus and Larichus might be ‘typical’ or ‘generic’ addressees, like Hesiod's brother Perses – i.e. whose key function (whether or not they really existed) is to act as recipient of generic paraenetic sentiments. While there may well be strong generic elements in the address to the brothers in frags. 5 and 9a (see further below on how these songs conform to a pattern of prayers by women for male relatives), the specificity of detail suggests that the addressees will have been identifiable by the original audience as real, known figures, and it was partly this ability to identify the figures as relatives of the singer which gave the songs (especially the new song) their particular emotional resonance. The fact that an Erygius, son of Larichus, is mentioned as a Mytilenean general of Alexander the Great (see n. 29) makes it very likely that the individuals identified as the brothers of Sappho were real Mytilenean figures.

45 Obbink (n. 1), 42. Talk of a Zeus-sent daimon who makes a person wealthy would fit a trade voyage.

46 Ferrari (n. 4), 4.

47 Obbink (n. 1), 41, suggests a series of theoretical possibilities for addressees, including Rhodopis/Doricha, and the singer's circle addressed collectively in the singular, or one of them. My proposal below is that a sister, a member of Sappho's circle, is a possible fit for the addressee of this song.

48 Obbink (n. 2), 7–8, agrees with West (n. 4), 9. Ferrari (n. 4), 4, and also Liberman (n. 4), 4.

49 West (n. 4), 7. The few surviving letters of the line preceding line 1 of the new song might be compatible with the restoration of ‘mother' (West [n. 4], 9; Obbink [n. 1], n. 29). Whether or not this is the case, the addressee could have been named in another line. A mother is also addressed in frag. 102, though the speaker here is probably not the Sappho-persona.

50 See Bettenworth, A., ‘Sapphos Amme: ein Beitrag zum neuen Sapphofragment (Brothers Poem)’, ZPE 191 (2014), 1519Google Scholar, citing as one example Penelope's companion Eurycleia. But the nurse figure is perhaps more likely to give wise advice than to receive it (especially from a young woman). Thus Eurycleia (Hom. Od. 4. 750–7) is the one who comforts Penelope and bids her pray to the gods for Telemachus’ return.

51 Nuenlist (n. 41) emphasizes this distinction: ‘Das Schiff soll unversehrt sein, nicht voll’ (‘May the ship be intact, not full’).

52 Thus Ferrari (n. 4), 3.

53 Other examples: frag. 98 (whose subject is, in the first instance, hair adornment, naming Cleis), frag. 22 (bidding [Abanthis] to sing of Gogyyla), and frags. 71 and 130 (chiding named women [Mika, Atthis] for seeking friendship elsewhere). Such dialogues in Sappho may have helped inspire the dialogues of women in Herodas and Theocritus.

54 E. Stehle, Performance and Gender in Ancient Greece. Nondramatic Poetry in Its Setting (Princeton, NJ, 1997), 282: a circle of friends, suggesting a parallel institution to Alcaeus’ hetaireia. But Stehle sees the erotic poems, including frag. 96, as not performed before the circle.

55 What Stehle (n. 54), 302, calls ‘a play of intersubjectivity’.

56 Obbink (n. 1), 43, cites Hom. Il. 3.308 (‘Zeus maybe knows [whether his son will be victorious in a duel], and the other immortal gods’), and Hom. Od. 15.323.

57 Though we may be meant to sense the clichés of traditional moralizing here (see C. B. R. Pelling, ‘Sappho’, in C.B.R. Pelling and M. Wyke, Twelve Voices from Greece and Rome. Ancient Ideas for Modern Times [Oxford, 2014], 43), the singer may be doing her best to put a brave face on things to console her interlocutor.

58 See J. Davidson, The Greeks and Greek Love (London, 2007), 402–3, on complexity of perspective in Sappho: ‘richocheting, the poem shifts quickly from one point of view to another’.

59 See Pelling (n. 57), 32–4.

60 See Stehle (n. 54), 287.

61 See ibid., 302–3 (on frag. 5), suggesting that the Sapphic circle might be partly bound by family and/or political connections. I suggest the new song provides further support for this idea.

62 Pelling (n. 77), 33, on frag. 94: ‘Sappho was being, or at least projecting herself as being, so very mature, as she reminded the girl of those beautiful times’.

63 See M. Williamson, Sappho's Immortal Daughters (Cambridge, MA, 1995), 139–40.

64 Obbink (n. 1), 35: ‘the power of hymnic song, framed in the poem, to help secure Charaxos’ safety’.

65 See Sappho T 252 and 253.

66 See, in particular, frags. 6, 208, 249.

67 Sappho frag. 20 (a very fragmentary song) is also a ‘storm’ song: Milne (n. 34) suggested that it is also a brothers poem. There is a parallel of vocabulary with the new song, with μεγἀλαι ἀήται appearing in both.

68 Cf. Alcaeus frag. 38, discussed below.

69 Obbink (n. 1), 34.

70 See Obbink (n. 2), 8: ‘otherwise, he comes out of the blue at the end’. Few openings of Sappho songs are preserved. Typical openings include (i) an address to a god/prayer (but this is ruled out for the new song, in which the prayer comes later, ‘embedded’ in the dialogue); (ii) an address to a personal addressee (one of the group); (iii) a common-place/topos. Here some version of (ii) is called for.

71 Frag. 22 (Abanthis); frag. 49 (Atthis); frag. 71 (Mica); the addressee in frag. 30 is apparently not named.

72 Obbink (n. 1), 41.

73 Frag. 94 may have begun in this way, if ‘honestly, I wish I were dead’ can be taken as the words of the singer addressed to the audience, outside the reported dialogue.

74 See D. Obbink, ‘New Poems by Sappho’, TLS, 5 February 2014, where Gregory Hutchinson is credited with pointing out the similarity.

75 Cf. Alcaeus frag. 6, which also begins with a ‘situational’ description and, like the new Sappho song, is also a ‘storm’ poem: ‘This wave in turn, like the previous one rolls in: it will give us much toil to bale’. The rest of the song calls on the audience to adopt an attitude of robust resistance to the metaphorical storm of factional strife which threatens to engulf them.

76 Translation from J. Michie, The Odes of Horace (Harmondsworth, 1967).

77 Sappho could be seen as achieving the change of register in the second and third stanzas of the new song by drawing on topoi of storms and resignation in the face of the gods more typical of the poetry of the male symposium.

78 Obbink (n. 1), 40, notes that POxy. 2289, frag. 5.1–2, probably contains letters from the first stanza of the new song, including in the fourth line, perhaps σέ (‘you’).

79 Obbink (n. 2), 9, proposes a hypothetical exempli gratia restoration of the first stanza drawing on Sappho T 203, and, less reliably, Ov. Her. 15 and Anacreon. The restoration attempts to pack a variety of information on Sappho brothers from other sources into the missing first stanza. However, it is plausible (given the introduction of Larichus in the final stanza) that the criticism of the addressee's attitude to Charaxus in the first surviving stanza could have been contrasted with an attitude that the singer suggests she should adopt towards Larichus in the previous (lost) stanza.

80 On the relationship between the papyri published by Burris et al. and that published by Obbink, see Burris, S., Fish, J., and Obbink, D., ‘New Fragments of Book 1 of Sappho’, ZPE 189 (2014), 128Google Scholar, 1; Obbink (n. 1), 33; and now Obbink (n. 2), 1–4. They are from the same manuscript of Book 1 of Sappho.

81 Though χάραν in line 6 might be a word-play reference to Charaxus: see Burris et al. (n. 80), 24.

82 West (n. 4), 6.

83 For attempts to restore the last stanzas, see West (n. 4), 5–7; Ferrari (n. 4), 4–9. Crucial to the restoration of the fourth stanza is the interpretation of τοκεγχρω in line 13. Burris et al. (n. 80), 25, understand κέγχρω as a genitive of κέγχρον (‘millet’) with the idea that this could be some kind of millet-powered percussion instrument heard by the brother in a dance, in the desired future state. This seems far-fetched, and West (n. 4), 6, and Ferrari (n. 4), 7–8, prefer the ingenious suggestion of Blass: τό κ’ἐν χρῶ<ι> – ‘which would, under the skin’. But this creates an even bigger challenge than the mysterious millet, since reading κε supposes a missing conditional clause (‘he would have…but instead he…’). But the structure of the song is two framing stanzas, each with an address to gods, with intervening stanzas, each of which sets out a prayer contrasting the evils of the past with a wish/prayer for the future either in infinitive or optative (‘may he atone for his wrongs’, ‘may he pay his sister greater honour’). A conditional clause in the fourth stanza would break this pattern and create an odd excursus, disrupting the flow and structure of the song. Perhaps, then, ‘millet’ is right but, rather than a joyous sound from the future state wished for in the prayer, ‘the sound of millet’ is something the brother hears in exile – perhaps the sound of grain being threshed in an agricultural exile. There is a potential close parallel with Alcaeus frag. 130B: Alcaeus, in rural exile because of ‘mutually destructive citizens’ (ἀλλαλοκάκων πολίταν), longs to hear the sound of the assembly being summoned, but instead only hears the ritual songs of the women at the festival.

84 Williamson (n. 63), 138–9.

85 Stehle (n. 54), 283 (and see also Williamson [n. 63], 139), sees this in frag. 5. She suggests, plausibly in my view, that ‘symposium-like’ circles of women such as that of Sappho might have been bound at least partly by their position in relation to Mytilenean faction, determined by their family and marriage relationships. Stehle also notes (285) that the ‘political’ element in Sappho is typically muted, focusing on its intersection with areas of women's interest (as in frag. 98).

86 Ibid., 296: Sappho's erotic songs are ‘always mediated by distance’.

87 Songs about absent family members feature in both Sappho and Alcaeus. See Yatromanolakis (n. 3), 335–6, on Mytilenean involvement in trade and founding cities. They may also have been on campaign with the city (the war with Athens over Sigeum), fighting as mercenaries (like Alcaeus’ brother – Alcaeus frag. 350), in exile (including Sappho herself, according to the Parian marble [Sappho T 251]), or, as women, marrying into Lydian families (the plausible scenario of frag. 96).

88 See also Sophocles’ Antigone, whose declaration (Soph. Ant. 105) that she would rather lose a husband or a child than a brother illustrates in extreme form the close ties of loyalty and dependency which could develop between an unmarried woman and her brother.

89 See e.g. Soph. El. 164–7, 187–8.

90 Cf. new Sappho song, line 14.

91 Perhaps adding a phrase meaning ‘with good fortune’. See Obbink (n. 1), 42, comparing ‘τύχηι σὺν ἔσλαι’ in Sappho frag. 20.3 (which he suggests, following Milne [n. 34], is a brothers poem), and Aesch. Cho. 138–9 (σὐν τύχηι τινί).

92 Though the discussion in the Electra plays must be due at least partly to the special difficulties of a prayer by Electra to her father, murdered by her mother.

93 Traditional prayer forms often use injunction to prayer in the place of prayer: e.g. Psalm 95.1–2: ‘O come, let us sing unto the Lord: let us make a joyful noise to the rock of our salvation. Let us come before his presence with thanksgiving, and make a joyful noise unto him with psalms.’

94 Stehle (n. 54), 288 ff. suggests that certain of Sappho's erotic poems were created as written texts presented as gifts to female recipients. See also Pelling (n. 57), 43: ‘I dare say that it is a poem to be sent to Charaxus, or at least one that he is expected to see and hear.’

95 Another example close to the scenario in the new Sappho is Eminem's song ‘Mockingbird’, a monologue by the singer addressed to his daughter Hailey but published to a wide audience. In it the singer uses the description of a family scene to explain his shame at not being a proper father to Hailey, in a similar way to the use of the family dialogue in the new Sappho to show the reliance and hopes placed by the family in the brothers. The freedom and inventiveness in the use of dramatic description and shifts of addressee and voice, which we perceive in the fragments of Sappho, are also characteristic of modern popular music.

96 As suggested by Chris Pelling to me.