Chapter 5. Tithonus in the “New Sappho” and the Narrated Mythical Exemplum in Archaic Greek Poetry

Lowell Edmunds

carminis fi nem statueram, nunc dubito
—M.L. West [1]

At present, some think that the twelve lines of the “New Sappho” are a complete poem; some think that the poem must have continued, as it appears to do in the Oxford papyrus (P.Oxy. 1787). [2] One way to approach the question of completeness is to focus on the Tithonus exemplum, and to compare it with the typical use of the narrated mythical exemplum in archaic Greek poems. [3] If the “New Sappho” is complete in twelve lines, then it ends with an exemplum. How likely is it that an exemplum will conclude a poem?

The question of such a conclusion has already been addressed apropos of the “New Sappho” by Hans Bernsdorff. [4] He wishes to establish what he calls the “open” conclusion as a stylistic feature of archaic Greek lyric. [5] But, as he has to admit at the outset, in archaic monody there exists only one possible case of a complete poem ending with a mythical narrative. It is Alcaeus fr. 44. [6] Bernsdorff thus has to look elsewhere, and he chooses Pindar and Horace. In Pindar, he finds three “open” endings, two in odes (Olympian 4; Nemean 1) and one in a paean (4). In my opinion, Pindar is the wrong place in which to look. Dionysius of Halicarnassus took Pindar and Sappho to represent opposite kinds of style, and Horace is likely to be reflecting this view in Odes 4.2. [7] Although both Pindar and Sappho are “lyric” poets, they differ in time, place, dialect, meters, and performance venue, thus also, I assume, in the use of the mythical exemplum. In this last respect, some differences are immediately obvious. First, the opening and closing formulas of the Pindaric mythical narrative are strikingly different from those in monody. [8] Second, the myth in Pindar tends to be more allusive and to be complexly related to the historical reality to which it refers. (The mythical exemplum in tragedy, too, tends to stand in a complex relation to the action. [9] ) As for Horace, as one who hoped to play the Lesboum barbiton (Odes 1.1.34), he seems a more likely candidate for comparison with Sappho. It is unclear, however, why his practice in one epode (13) and four odes (1.7; 1.8; 3.11; 3.27) is probative for Aeolic monody. Note that, in each of these poems, Horace concludes not with narrative, as a twelve-line “New Sappho” would conclude, but with quotation of a character in the narrative, whose words directly or indirectly make the point that Horace wishes to make.

My comparanda for the Tithonus exemplum are in archaic poems for solo performance. I will also discuss the narrated mythical exemplum in speeches in Homer, for reasons to be explained later. [10]

First, Sappho and Alcaeus. The former provides at most two exempla. In fr. 16, Helen leaves Menelaus and her daughter and so forth. If J.G. Howie is right, the poem did not end at line 20 but continued to line 33, where there is a coronis. [11] If that was the end of the poem, one can assume that there was a return to the gnome with which the poem begins or to the circumstances which occasioned that gnome. In fr. 17, a prayer or cult hymn [12] to the Hera of Lesbos, [13] Sappho, having evoked Hera’s support of the Atreidae, returns to her petition with νῦν δὲ κ[ἄμοι (11; cf. fr. 1.25 καὶ νῦν). Someone might say that the closing petition is obligatory in a prayer and therefore has no bearing on the question of how the “New Sappho,” which is not a prayer, ended. [14] The formal markers, however, and, one might say, the style of prayer or cult hymn were hardly restricted to ritual occasions but were broadly shared across genres. For example, sixteen of Pindar’s epinicians begin with cult hymns. [15]

For Sappho, then, there are perhaps two examples, in one of which Sappho returns from the myth to the present. In Alcaeus, four relevant examples appear. In the sequence of Voigt numbers, the first is fr. 38A, in which Alcaeus exhorts Melanippus to get drunk with him. He uses Sisyphus as an example of the futility of human attempts to escape death and returns to his exhortation with ἀλλ’ ἄγι (10). Fr. 42 is the curious twelve-line (four Sapphic strophes) synkrisis of Helen and Thetis. It begins ὠς λόγος and it ends by rounding off the synkrisis without an application of the myths. If this poem, with its “open” conclusion, is complete, then it is a parallel to the “New Sappho” understood as a complete poem, which happens also to be twelve lines long. But, from the time of its publication, opinions have differed about the completeness of Alcaeus fr. 42, and one would have to say that it is a parallel not to a twelve-line “New Sappho” but to the problem of a twelve-line “New Sappho.” [16] Even if Alcaeus fr. 42 is complete, it is possible that it could function as a complete poem only within a sympotic context, in which it was part of a chain, attempting to cap the preceding symposiast’s effort. [17] Another fragment of Alcaeus which some have considered complete is 44. [18] The myth here is modeled on the episode in the Iliad in which Achilles asks his mother to intercede with Zeus on his behalf (1.348–427; 493–516). Though sense can be made of the poem only in line 6, and the beginning of the poem is thus uncertain, a coronis at line 8 shows where the poem ended. It ended with the myth. Fr. 44 is thus a secure example of what Bernsdorff calls the “open” conclusion, and the only secure example in Sappho and Alcaeus he can point to. [19] The fourth and last example from Alcaeus is fr. 298 with the myth of Cassandra and Ajax’s impiety, apparently preceded and followed by reference to the impiety of Pittacus, betrayer of his companions. [20] Pittacus is referred to in line 47. For Alcaeus, then, we have one example of a myth concluding a poem (fr. 44), two examples of a return from the myth to the present situation (frs. 38A, 298), and one non liquet (fr. 42).

Elegy contains many mythical exempla but only a few that are narrated and thus relevant to the present question. Mimnermus’ lines on the sun are a description and not a narrative (fr. 12). They are from Nanno and thus presumably had some particular contextual frame. The two fragments on Jason and the Golden Fleece are from a narrative (frr. 11–11A), and it sounds as if the narrative is being used as an argument for something, but context is completely lacking. [21] In the priamel in Theognis 699–718, the foils—heroes and mythical figures—are concise and unnarrated, with one exception, Sisyphus. His story is amplified to the length of five couplets (703–12). [22] The priamel as a whole is framed by a gnome (men consider wealth the greatest aretê, 699–700, 717–718). [23] Another narrated mythical exemplum in Theognis, this time two couplets long, occurs in a pederastic poem (1341–1350). [24] The myth, which is about Ganymede, is part of an argumentum ex Iove. [25] Zeus did it. Why shouldn’t I? Theognis introduces the myth with ἐπεί ποτε καί and ends by addressing someone named Simonides: οὕτω μὴ θαύμαζε. In another poem, Theognis compares his sufferings to those of Odysseus (1123–1128). It is not clear that this poem, which ends with talk of vengeance on the suitors, is complete. [26] I suggest that what we have points to a missing conclusion in which Theognis returned to one of his favorite themes, namely, vengeance on his enemies. [27]

The only example in Theognis—the only example in all of elegy—of an “open” ending is the poem in which Theognis, addressing a boy, compares his pursuit of him to Atalanta’s suitors’ pursuit of her (1283–1294). He introduces the exemplum with ὥς ποτέ φασιν. He concludes: she came to know Aphrodite in spite of her refusal. In a sympotic context, the point of the exemplum is so clear that it is perhaps more effective if left unexpressed: as even Atalanta yielded, so you will yield to me. This poem, too, however, is surrounded by controversy. There are those, like West, who say that it is an amalgam of two poems; there are those who say that it is a single, complete poem. [28]

To conclude on elegy, there is now the remarkable mythical narrative concerning Telephus which constitutes the “new Archilochus” (P.Oxy. 4708). Dirk Obbink in his publication of the fragment suggests that the story may be an exemplum and shows how it could have functioned as such. [29]

The largest and I would say most useful set of narrated mythical exempla is found in the speeches in Homer. But what justifies using these speeches as comparanda for Sappho? There are three reasons. In the first place, mythical exempla have the same introductory elements in these speeches as are found in monody and elegy. Second, the speeches belong to a set of rhetorical genres, and some of these genres, for example, prayer, lament, and supplication, are shared with poetry. [30] Third, as rhetorical utterances, the speeches in Homer are, like poems or songs, performance genres. [31]

Much research has been devoted to narrated mythical exempla in Homer, which are customarily referred to as “paradeigmata.” [32] Scholars have classified them in various ways, discussed their functions, and the distortions or inventions which appear in the myths themselves. My question is a simple, formal one: do speakers return from a narrated myth to the situation which prompted the telling? The answer is yes. M. M. Willcock made the relevant point in an article in 1964: paradeigmatic speeches in Homer have a ring-compositional structure, with the myth in the middle. [33] The same is true of the few paradeigmata in the Homeric Hymns, one of which happens to be about Tithonus. [34] The same is true also of the ainos, of which we have fifteen examples in archaic and classical poetry which survive in their entirety. [35]

Earlier, I mentioned introductory elements as a shared feature of exempla in monody and epic speeches. The same is true of the elements by which the speaker rounds off the exemplum and returns to the present. Here are a few instances: In Iliad 20, Aeneas concludes his genealogical discourse to Achilles: ἀλλ’ ἄγε μηκέτι ταῦτα … (244). In Iliad 24, Achilles concludes the Niobe exemplum: ἀλλ’ ἄγε δὴ καὶ νῶϊ … (618). Compare the way Alcaeus turns from the Sisphyus exemplum to his exhortation to Melanippus: ἀλλ’ ἄγι … (fr. 38A.10). Achilles’ concluding formula in the speech to Priam is really a combination of two formulae which can occur separately. With καὶ νῶϊ one can compare νῦν δὲ κ[ἄμοι in Sappho fr. 17.11 (cf. fr. 1.25 καὶ νῦν). A favorite way to round off a paradeigma in Homer is with a phrase introduced by ὥς, e.g ὣς καὶ ἔγω, and one can compare in Theognis the phrase οὕτω μὴ θαύμαζε which I have already cited. [36]

To move toward a conclusion, a survey of the evidence for the narrated mythical exemplum in archaic monody and elegy shows only a single indubitable case of an “open” conclusion, against a few cases of an A-B-A structure, where A is the present situation and B is the myth. Alcaeus fr. 42 has the “open” conclusion, which may be owing to its place in a sympotic chain. As for the Theognidea, lines 1287–1294 (on Atalanta) would also be an example of an “open” ending if it were certain that these lines constitute a single poem. Another exemplum in Theognis is followed by an explicit return to the situation which prompted the myth. In Homer, the speaker who uses a paradeigma will regularly apply it to the present situation. In Alfred Heubeck’s formulation, the argument of the paradeigma in Homer is either “Thus someone acted then, thus you should act now” or “Thus it happened once before, take warning.” [37]

The preponderance of the evidence, such as it is, especially if one admits speeches in Homer, as I think that one can, leads one to expect a return from the myth to the present, an A-B-A structure. One would expect Sappho, then, to return from Tithonus to her present situation, and one notes in lines 23–24 of the Oxyrhynchus papyrus present tenses which would indicate a return to the present time of the enunciation. Further, one notes καί μοι, which resembles a rounding-off formula. I cited an example of this formula from Sappho fr. 17.11. Penelope, too, uses it at Odyssey 19.524 to round off the story of Pandareus’ daughter, the nightingale, and return to her own situation.

Penelope’s myth illustrates something else. As I said in the abstract of the presentation from which this paper has emerged, “The narrator finds a particular point of contact between the myth and the situation to which he or she applies it.” The tertium comparationis is never in doubt or left as allusion or riddle. [38] In the “New Sappho,” Tithonus clearly illustrates the gnome that mortals inevitably grow old. If this exemplum has some further, different meaning, this idea will be found not in the myth itself but in Sappho’s perspective on the myth. The peculiar imperfect ἔφαντο suggests that Sappho’s perspective has indeed changed, presumably because of self-understanding that post-dates the perspective of the subjects of ἔφαντο. She knows something that they didn’t know.

What ought to be in concluding lines is some kind of self-consolation, as the first editors of the “New Sappho” thought. [39] Alex Hardie has put together, from other fragments of Sappho (including, obviously, the one in the Cologne papyrus) and from a vast array of related material, the elements that one would expect to find in the conclusion, even though Hardie appears to believe, in spite of himself, that the twelve lines are a complete poem. These elements are devotion to the Muses in the form of song and dance, with related eschatological hopes, all as part of the choric education with which the poem begins. [40]

My reading is a historicizing one. It aims at establishing the outline of the poem that Sappho’s audience might have expected. There is another kind of reading, equally important, which I would call a possible reading. For example, when Richard Janko in an article in the Times Literary Supplement explains the Tithonus exemplum as in itself consolatory he gives a reading which might have become possible at some point in antiquity and is possible for us. [41] Likewise, when Bernsdorff cites Horace for the practice of Sappho, I think that he is really talking about a way of reading Sappho which he learned from Horatian innovation in the use of the mythical exemplum. The later poet enables new ways of reading the earlier poet. [42]

This distinction I have made between the way or ways we can read the “New Sappho” and the way the twelve lines would have appeared to Sappho’s audience needs further discussion, as an anonymous reader’s reaction to this paper shows. He or she states: “We moderns have no trouble making sense of the twelve-line poem and its final mythical exemplar. Sappho’s audience would have been more equipped to do so as well.” In this way, he or she collapses the distinction between reception contemporary with Sappho and later reception such as Janko’s. The deductive form of his or her argument, of the modus ponens type, is not the usual one in discussion amongst classicists, but it happens to be convenient for illustrating what I consider a mistaken point of view. It can be restated thus:

  • If we moderns can make sense of the twelve-line “New Sappho,” then Sappho’s audience could make sense of it, too.
  • We can make sense of it.
  • Therefore Sappho’s audience could make sense of it.

In this kind of argument, as everyone knows, everything depends on the truth of the premises. The first premise here is, I submit, untenable. It assumes that the mind will always, across millennia, respond to the same poetic work in the same fundamental way. A counter-argument from common sense suffices. The workaday classicist constantly discovers not the fundamental similarity of the ancient to the modern mind but its alterity. One could also point to the current vogue of reception studies, which is showing how the meaning of ancient texts has differed in different times and places. A stronger counter-argument against the idealist, transhistorical pretension of the syllogism I have analyzed can be found in Hans-Georg Gadamer’s Truth and Method. On his account, the understanding of an ancient text comes about not in spite of but because of and within our historical distance from that text.


Several speakers besides me at the two Sappho panels held in San Diego in 2007 addressed the question of the completeness of the “New Sappho.” Dee Clayman described the Cologne papyrus as the remains of a Hellenistic florilegium on old age, mortality and song, comparing the roughly contemporary “New Posidippus” (P.Mil.Vogl. VIII 309) with its thematic arrangements under nine headings. The “New Sappho” might be a truncated edition of the poem, she said, enough for the compiler’s purposes. André Lardinois, too, compared the “New Posidippus,” and said that, in the unstable conditions of archaic poetry, two different versions might have been transmitted. Both he and Clayman were thinking materialistically, so to speak, i.e. of transmission in writing. Gregory Nagy, thinking in terms of performance, spoke of the ending of the archaic poem as always already fungible. Deborah Boedeker, too, argued for the possibility of different versions in performance. The position I have taken in this paper can be reconciled with the positions of Clayman and Lardinois, I believe. As for Nagy’s and Boedeker’s view, one would have to imagine, for the twelve-line “New Sappho,” a performance context in which the audience, if I am right about the precise point of contact between performer and his or her situation, is content with Tithonus as an exemplum illustrating the old saw about the inevitability of aging. The only way around this conclusion, or to counter this conclusion, is to argue that the Tithonus myth, as in the twelve-line “New Sappho,” contains a latent, positive, consolatory idea. So Richard Janko argued in his TLS article, [43] and so, but finding a different idea, did Joel Lidov, though Lidov stopped short of saying that the “New Sappho” is a complete poem, preferring “complete poetic statement.” To repeat, I myself do not expect a mythical exemplum in such a context to be a riddle or to be allusive. (The myth in Pindar is another story.)


Andersen, Ø. 1975. Paradeigmata: Beiträge zum Verständnis der Ilias. Diss., University of Oslo.

———. 1978. Die Diomedesgetalt in der Ilias. Symbolae Osloensis Supplement 25. Oslo.

Austin, N. 1966. “The Function of the Digressions in the Iliad.” Greek, Roman, and Byzantine Studies 7:295–312. Reprinted 1978: Essays on the Iliad: Selected Modern Criticism (ed. J. Wright) 70–84. Bloomington, IN.

Bastianini, G., and A. Casanova, eds. 2007. Atti del convegno internazionale di studi “I papiri di Saffo e di Alceo”: Firenze 8–9 giugno 2006. Studi e Testi di Papirologia n.s. 9. Florence.

Bernsdorff, H. 2005. “Offene Gedichtschlüsse.” Zeitschrift für Papyrologie und Epigraphik 153:1–6.

Bierl, A. 2003. “‘Ich aber (sage), das Schönste ist, was einer liebt!’ Eine pragmatische Deutung von Sappho Fr. 16 LP/V.” Quaderni Urbinati di Cultura Classica 74:91–124.

Bonifazi, A. 2004. “Relative Pronouns and Memory: Pindar Beyond Syntax.” Harvard Studies in Classical Philology 102:41–68.

Borges, J. L. 1964. “Kafka and his Precursor.” Other Inquisitions, 1937–1952. Trans. R. L. C. Simms. 106–108. Austin, TX. Originally published 1952: Otras inquisiciones, 1937–1952. Buenos Aires.

Braswell, B. K. 1971. “Mythological Innovation in the Iliad.” Classical Quarterly n.s. 21:16–26.

Campbell, D. A., ed. 1982. Greek Lyric I. Cambridge MA. Corrected second edition 1990.

Canter, H. V. 1933. “The Mythological Paradigm in Greek and Latin Poetry.” American Journal of Philology 54:201–224.

de Jong, I. J. F. 1997. “ΓΑΡ Introducing Embedded Narratives.” New Approaches to Greek Particles (ed. A. Rijksbaron) 175–185. Amsterdam Studies in Classical Philology 7. Amsterdam.

Demoen, K. 1997. “A Paradigm for the Analysis of Paradigms: The Rhetorical Exemplum in Ancient Imperial Greek Theory.” Rhetorica 15.2:125–158.

Edmunds, L. 1985. “The Genre of Theognidean Poetry.” In Figueira and Nagy 1985:96–111.

———. 2001. Intertextuality and the Reading of Roman Poetry. Baltimore, MD.

———. 2006. “The New Sappho: ἔφαντο (9).” Zeitschrift für Papyrologie und Epigraphik 156:23–26.

Faraone, C. A. 2005a. “Exhortation and Meditation: Alternating Stanzas as a Structural Device in Early Greek Elegy.” Classical Philology 100:317–336.

———. 2005b. “Catalogues, Priamels, and Stanzaic Structure in Early Greek Elegy.” Transactions of the American Philological Association 135:249–265.

Ferrari, F. 1989. Teognide: Elegie. Milan.

Figueira, T., and G. Nagy, eds. 1985. Theognis of Megara: Poetry and the Polis. Baltimore, MD.

Fraenkel, H. 1927. Review of R. Oehler, Mythologische Exempla in der älteren griechischen Dichtung (Aarau 1925). Gnomon 3:569–576.

Fraenkel, H. 1962. Early Greek Poetry and Philosophy. Oxford.

Gerber, D. E. 1991. “Early Greek Elegy and Iambus 1921–1989.” Lustrum 33:7–225.

———, ed. 1999. Greek Elegiac Poetry from the Seventh to the Fifth Centuries BC. Loeb Classical Library. Cambridge, MA.

Gonis, N., et al., eds. 2005. The Oxyrhynchus Papyri 69. London.

Gotteland, S. 2001. Mythe et rhétorique: les exemples mythiques dans le discours politique de l’Athènes classique. Paris.

Gronewald, M., and Daniel, R. W. 2004a. “Ein neuer Sappho-Papyrus.” Zeitschrift für Papyrologie und Epigraphik 147:1–8.

———, 2004b. “Nachtrag zum neuen Sappho-Papyrus.” Zeitschrift für Papyrologie und Epigraphik 149:1–4

Hardie, A. 2005. “Sappho, the Muses and Life after Death.” Zeitschrift für Papyrologie und Epigraphik 154:13–32.

Henderson, W. H. 1983. “Theognis 702–712: The Sisyphus-exemplum.” Quaderni Urbinati di Cultura Classica n.s. 15:83–90.

Heubeck, A. 1954. Der Odyssee-Dichter und die Ilias. Erlangen.

Holzberg, N. 2002. The Ancient Fable: An Introduction. Trans. C. Jackson-Holzberg. Bloomington, IN.

Howie, J. G. 1977. “Sappho Fr. 16 (LP): Self-Consolation and Encomium.” Papers of the Liverpool Latin Seminar 1 (ed. F. Cairns) 207–235. Liverpool.

Hunter, R. 2007. “Sappho and Latin Poetry.” In Bastianini and Casanova 2007:213–225.

Jacoby, F. 1957. Die Fragmente der griechischen Historiker ed. 2. vol. 1a. Leiden.

Jaeger, W. 1945. Paideia 2nd edition. vol. 1. Trans. G. Highet. New York.

Janko, R. 2005. “Sappho Revisited.” Times Literary Supplement December 23.

Jurenka, H. 1914. “Neue Lieder der Sappho und des Alkaios.” Wiener Studien 36.201–243.

Maronitis, D. N. 2004. Homeric Megathemes: War—Homilia—Homecoming. Trans. D. Connolly. Lanham, MD.

Martin, R. 1989. The Language of Heroes: Speech and Performance in the “Iliad.” Ithaca.

Meyerhoff, D. 1984. Traditioneller Stoff und individuelle Gestaltung: Untersuchungen zu Alkaios und Sappho. Hildesheim.

Miller, A. M. 1986. From Delos to Delphi: A Literary Study of the Homeric Hymn to Apollo. Leiden.

Nagy, G. 1973. “Phaethon, Sappho’s Phaon, and the White Rock of Leukas.” Harvard Studies in Classical Philology 77:162–177.

Oehler, R. 1925. Mythologische Exempla in der älteren griechischen Dichtung. Aarau.

Page, D. L. 1955. Sappho and Alcaeus: An Introduction to the Study of Ancient Lesbian Poetry. Oxford. Corrected reprint 1959.

Pedrick, V. 1983. “The Paradeigmatic Nature of Nestor’s Speech in Iliad 11.” Transactions of the American Philological Association 113:55–68.

Race, W. H. 1982. “Aspects of Rhetoric and Form in Greek Hymns.” Greek, Roman, and Byzantine Studies 23:5–14.

———. 1990. Style and Rhetoric in Pindar’s Odes. American Classical Studies 24. Atlanta, GA.

———. 1992. “How Greek Poems Begin.” Yale Classical Studies 29:13–38.

Rawles, R. 2006. “Notes on the Interpretation of the ‘New Sappho’.” Zeitschrift für Papyrologie und Epigraphik 157:1–7.

Rösler, W. 1980. Dichter und Gruppe: Eine Untersuchung zu den Bedingungen und zur historischen Funktion früher griechischer Lyrik am Beispiel Alkaios. Munich.

Slater, W. J. 1996. Review of M. M. Willcock, Pindar Victory Odes: Olympians 2, 7 and 11, Nemean 4, Isthmians 3, 4 and 7 (Cambridge 1995). Bryn Mawr Classical Review 96.02.03.

Taalman Kip, A. M. van Erp. 1987. “Alcaeus: ‘Aias and Kassandra’.” Some Recently Found Greek Poems (eds. J. M. Bremer, A. M. van Erp Taalman Kip, and S. R. Slings) 95–127. Mnemosyne Supplement 99. Leiden.

Tarditi, G. 1969. “L’asebeia di Aiace e quella di Pittaco.” Quaderni Urbinati di Cultura Classica 8:86–89.

Thomas, R. 2000. Herodotus in Context: Ethnography, Science, and the Art of Persuasion. Cambridge.

Vetta, M. 1975. “Forma e immagini del teognideo (Theogn. 1283–94).” Prometheus 1:209–224.

———, ed. 1980. Theognis: Elegiarum liber secundus. Rome.

———. 1981. “Poesia e simposio (A proposito di un libro recente sui carmi di Alceo [= Rösler 1980]).” Rivista di Filologia e di Istruzione Classica 109:483–495.

Weil, H. 1862. “Über Spuren strophischer Composition bei der griechischen Elegikern.” Rheinisches Museum 17:1–13.

West, M. L. 1974. Studies in Greek Elegy and Iambus. Berlin.

Willcock, M. M. 1964. “Mythological Paradeigma in the Iliad.” Classical Quarterly n.s. 14:141–154.

Zagagi, N. 1980. Tradition and Originality in Plautus: Studies of the Amatory Motifs in Plautine Comedy. Hypomnemata 62. Göttingen.


[ back ] 1. IEG2, on Simonides fr. 20.12.

[ back ] 2. I am grateful to Michele Caprioli for corresponding with me about this paper.

[ back ] 3. Narrated” refers to the telling of a story or some part of a story. My category of exemplum thus excludes references to a god or hero which are limited to a name and epithet and/or relative clause. The word “mythical” refers to traditional stories about gods and heroes.

[ back ] 4. Bernsdorff 2005. Stehle, this volume, n3, takes the twelve lines of the new Sappho as a complete poem. (Though Stehle’s article is about time, she does not discuss the odd imperfect in line 9, on which see Edmunds 2006.)

[ back ] 5. Bernsdorff 2005:2n5 on the term “open”: it is from Walter Wili.

[ back ] 6. All references to Sappho and Alcaeus are to Voigt’s edition.

[ back ] 7. As Hunter 2007:217–219 argues.

[ back ] 8. Edmunds 2006:n2: The typical opening in Pindar is relative pronoun + aorist + ποτε + aorist participle (or a subset of these). See Bonifazi 2004, who points out that “γάρ is the particle that introduces mythical sections without a relative pronoun” (47), citing two places in Pindar and referring to de Jong 1997. The typical closing is the notorious break-off, with return to the first-person: Race 1990:41–57.

[ back ] 9. For example, a character may contest the relevance of a myth to him- or herself. Zagagi 1980:32–46 studies places in tragedy in which a character whom the chorus has attempted to comfort with a mythical exemplum contests the appropriateness of the exemplum by insisting that “his own situation has surpassed … the unhappy events of the myth in question” (33).

[ back ] 10. For an inventory of narrated mythical exempla, I started with Oehler 1925. I am not, however, the new Oehler whom we need. Canter 1933 might be useful for purposes other than mine. He gives an inventory of exempla arranged by topic (206–219), with a few comments, e.g. on comparative frequency in Greek and Latin, on their usage.

[ back ] 11. Howie 1977. (Please note the conditional form of this proposition.) See Bierl 2003:121n112 for a list of publications since Howie’s which take the view that fr. 16 continued after the sixth strophe.

[ back ] 12. See Race 1992:28 and n49 on the term “kletic hymn,” which could also be used of Sappho fr. 17. He does not distinguish between kletic hymn and prayer. On the distinction between cult hymn and rhapsodic hymn, see Miller 1986:1–5; Race 1992:28–31.

[ back ] 13. Hera of Lesbos: cf. Alcaeus fr. 129.6; Sappho test. 59 (from the Palatine Anthology 9.189).

[ back ] 14. Race 1982:10–14 on the request at the end of the hymn, at which point “the hymnist tries to establish the closest connection between him and the god.” Race 1992:19n17 for bibliography on hymnal openings. Race 1992:28–29 (diagram) shows clearly that the poet or singer returns at the end, with a salutation or a request, to the divinity named or addressed at the outset.

[ back ] 15. Race 1992:30.

[ back ] 16. Complete: The majority opinion. See Rösler 1980:223 and n269 for references. Maronitis 2004:79–80 gives a formalistic argument for unity. Campbell 1982:259: “16 is certainly the last line.” Incomplete: see references in Rösler 1980:224n270. Rösler’s own arguments for incompleteness: 224–227; 233–235 (would have been ten strophes long).

[ back ] 17. For the sympotic destination of fr. 42: Jurenka 1914:229; Rösler 1980:221 (fr. indicates the “non-political conversational material of the hetaireia”); cf. Page 1955:280n1; Vetta 1981:486–487.

[ back ] 18. Page 1955:282: “ends abruptly in the eighth line, leaving the sequel to the memory of the audience.”

[ back ] 19. Bernsdorff 2005:2 cites Meyerhoff 1984:52–53.

[ back ] 20. Thus already Tarditi 1969:86–96. Taalman Kip 1987:125: “ … [W]e cannot even guess how Alcaeus made the switch from myth to reality at the end of the fragment. We can only say something about the way the myth is introduced. If we assume (as I think we must) that this happens at v. 4, we may ascertain that the conclusion of the story comes first, serving as a link between reality and the story proper.”

[ back ] 21. The context in Strabo 1.2.40 does not help.

[ back ] 22. Faraone 2005a and 2005b has recently written about the unit of five couplets as a characteristic building-block of elegy, which he calls a “stanza.” In both articles, he refers to Weil 1862, the first to perceive these units. Weil called them “strophes” and also spoke of responsion. He devoted less than a page (8) to Theognis and referred only to lines 1135–1150, without noticing the “choral” introductions to the mythical exempla in lines 703–712 and 1123–1128.

[ back ] 23. See Henderson 1983 on the framing of the exemplum.

[ back ] 24. On the attribution of these lines to Euenos of Paros, either the contemporary of Socrates or an older one, which was proposed by four different scholars between 1907 and 1934, see Vetta 1980:121–123.

[ back ] 25. Cf. Theognis 1345–1350 (discussed in this paper); Aristophanes Clouds 1079–1082 (Weaker recommends argumentum ex Iove if Pheidippides is accused of adultery); Euripides Trojan Women 948–950 (Helen uses this argument); Plato Euthyphro 5e–6a (Euthyphro uses it apropos of prosecuting his father); etc.

[ back ] 26. Campbell 1982:339n1 sums up the position of the doubters: “Many assume that the poem is incomplete.”

[ back ] 27. For this theme: Edmunds 1985:103.

[ back ] 28. Amalgam: West 1974:165–167; IEG2 app. crit. on 1288–1294; Vetta 1975; Vetta 1980:81–82. Not an amalgam but self-consistent: Ferrari 1989:316–320; Gerber 1991:213–214; Gerber 1999:371n4. With clairvoyance denied West and Vetta (and earlier scholars cited by Vetta), an anonymous reader tells me: “The Atalanta exemplar in Theognis 1287–1294 is perfectly comprehensible as it is.” He or she provides no explanation.

[ back ] 29. In Gonis and Obbink 2005:20–21.

[ back ] 30. Martin 1989:44: “the major rhetorical genres available for the heroic performers are prayer, lament, supplication, commanding, insulting, and narrating from memory.” Further, “these discourse types constitute poetic ‘genres’ outside epic” (94). Martin shows the lyric basis of Hector’s speech at Iliad 7.235–241. He cites (n81) Nagy’s Pindar’s Homer (at that time forthcoming) in order to distinguish his own point (“a relation to lyric”) from Nagy’s chronology of hexameter, which begins in shorter meters of the kind which we call “lyric.”

[ back ] 31. Martin 1989:37 (muthoi are public; performance before an audience); 160 (“performance” a more inclusive term for “speeches”); 225 (historical basis of the speeches: “performance of self”). The use of mythical exempla in Athenian oratory has been seen as continuous with the practice of the Homeric heroes: Gotteland 2001:11. Cf. Thomas 2000:257 on Antiphon DK 87 B54 in particular and sophists’ use of myth. I have not seen Demoen 1997.

[ back ] 32. Andersen 1975 and 1978; Austin 1966:300–304 (discusses the mythical exemplum in relation to a larger “paradigmatic logic” informing historical digressions in the Iliad); Braswell 1971; Heubeck 1954; Pedrick 1983. For older discussion, see Heubeck 1954:23n33. Pietatis causa, I cite also Fraenkel 1927 and Jaeger 1945:40–43 (see also Index s.v. “Example”).

[ back ] 33. Already Heubeck 1954:25: “normhafte Dreistufenaufbau.”

[ back ] 34. Homeric Hymn to Apollo 300–375 (Typhoeus) is a rare example of a narrated mythical excursus (not an exemplum) in the poet’s own voice. Note transition formulas at 307 and 375. Cf. West on Hesiod Theogony 22. In the Homeric Hymn to Hermes 414–462 the poet introduces into his narrative a cosmogonic song sung by Hermes which succeeds in softening the anger of Apollo at the theft of his cattle. The examples most relevant to the question of the mythical exemplum in the new Sappho are Aphrodite’s precedents for her union with Anchises: Tros (Homeric Hymn to Aphrodite 200–217) and Tithonus (Homeric Hymn to Aphrodite 218–238). The latter story also explains why the mortal Anchises cannot become her husband. Note transition formulas at 202 and 218.

[ back ] 35. Holzberg 2002:20. In nine of the fifteen, the speaker draws the moral from the tale.

[ back ] 36. Theognis 1349 οὕτω μὴ θαύμαζε. Cf. ὥς at Iliad 11.762; 14.328; 18.120; 19.134; 23.643. For the imperative cf. Iliad 1.274; 15.31.

[ back ] 37. Heubeck 1954:23–24. (I have somewhat simplified the second kind of argument.)

[ back ] 38. Pace Rawles 2006:3.

[ back ] 39. Gronewald and Daniel 2004b:3-4.

[ back ] 40. Hardie 2005. Cf. Nagy 1973:177: “As a coda to this poem, the last two verses amount to a personal and artistic manifesto.”

[ back ] 41. Janko 2005. (Cf. Hardie 2005:28n100.) Rawles 2006:7 assumes the concept of what I am calling a possible reading: “A reading of the poem which absorbed the element of the story in which Tithonus is metamorphosed into a cicada is necessarily conjectural, since it cannot be certainly demonstrated that this element was available to Sappho and her audience. It can at best be shown that it is compatible with what we do know and with how the poem may in any case be read.”

[ back ] 42. As Borges said, “ … [E]ach writer creates his precursors”: Borges 1964[1952]. For discussion of this idea see Edmunds 2001:159–163.

[ back ]43. It is not my purpose in this paper to argue against such interpretations. It seems necessary, however, to point out that reliance on Hellanicus fr. 140 Fowler = FGrH 4 F 140 as an early authority for metamorphosis of Tithonus into a cicada is risky. This fragment, from a scholiast on Iliad 3.168, runs: “Day (῾Ημέρα) fell in love with Tithonus, son of Laomedon, brother of Priam, by whom she had a son Memnon, and, after she had regaled him with a long life, she changed him into a cicada.” One has to agree with Jacoby 1957:466: “mehr als die genealogie wird H nicht gehörchen.” The scholiast offers not an alternate version of the myth but nonsense.