The Iconographic Evidence and Its Background

Sappho’s leap from the Rock of Leucas is arguably the most salient iconic trait in modern visual images of the lyric poetess from the Renaissance onwards.Footnote ¹ On the other hand, only one single, if conspicuously spectacular, representation of the motif is known from antiquity. Its depiction in the apse of the so-called underground ‘basilica’ at Porta Maggiore is the focal point in the iconographic programme of an impressive underground building.Footnote ² More than a century after its discovery in 1917, and almost a century after its first full publication by G. Bendinelli,Footnote ³ the overall interpretation of the hypogeum is still very much an object of debate. It probably was a funerary building and is likely to be dated to the age of Tiberius or Claudius.Footnote ⁴ Its topography, in proximity to the Horti Tauriani and to the large group of burials (more than 700) of the liberti and slaves of the Statilii Tauri, the so-called Monumentum Statiliorum (just within the city walls), founded by either T. Statilius Taurus, consul in 26 BCE and general of Octavian’s land army at Actium, or by a later member of this powerful family, makes its attribution to the Statilii Tauri plausible. Its striking and rich decoration has been read in several ways. Its apparent uniqueness, though, may to a certain extent also derive from the loss of potential elements of comparison. I draw attention here to an intriguingly similar underground funerary chamber adorned with rich stucco decoration that was discovered in the same area of Rome in 1733. This was part of the large burial complex of another prominent senatorial family of the period, that of L. Arruntius L. f., possibly the consul of 22 BCE (and, interestingly, the general of the central division of Octavian’s fleet at Actium and a Quindecemvir sacris faciundis at the time of the ludi saeculares), or of one of his descendants. The building itself and its decoration have vanished long ago. They were destroyed to make way for modern buildings on the Esquiline in the late nineteenth century but they were recorded in the eighteenth century, notably in a series of drawings by Pier Leone Ghezzi between 1733 and 1740 and of etchings by Giovanni Battista Piranesi in the early 1750s.Footnote ⁵ The similarities with the Porta Maggiore hypogeum are not limited to the general lay-out nor to the style of the decoration. It is difficult to make out all the details of the iconography of the panels reproduced by Piranesi, but one of them can be recognised as the abduction of one of the daughters of Leucippus (figure 1).Footnote ⁶ This is the same subject of one of the two main panels in the central nave in the Porta Maggiore hypogeum (figure 2).Footnote ⁷ As noted by Wadsworth in the aftermath of the discovery of the ‘basilica’, ‘if it were possible, it might be most interesting to compare these two monuments especially as they were not far apart geographically’.Footnote ⁸ It is indeed striking to find such similarities in two funerary buildings so chronologically and topographically close, possibly belonging to two families brought to great prominence by members closely involved in Octavian’s/Augustus’ military and ideological enterprise. This should probably make us more cautious in assessing the degree of originality (or even uniqueness) of the so-called ‘basilica’.

Figure 1.

Section of the vault of the Sepulcrum Arruntii, from Piranesi (Reference Piranesi1784) tav. XII.

Figure 2.

Detail of the vault of the central nave of the Porta Maggiore hypogeum (© Soprintendenza Speciale Archeologia Belle Arti e Paesaggio di Roma).

The basilica’s iconographical project must have been based on an eschatological interpretation of a wide range of mythical scenes. Attempts at linking all this to a definite religious/philosophical sect, however, do not seem to me to have been successfully argued and none of the pieces of evidence adduced so far for a more specifically Neo-Pythagorean reading stands up to scrutiny. It would be safer to acknowledge the eclectic nature of the project. Some of its elements can be variously defined as ‘Dionysiac’, ‘Orphic’, or even ‘Pythagorean’, or, more vaguely, as related to mystery cults but it is difficult to identify a distinctively systematic approach behind the whole, and much of it can be paralleled in less elaborate contemporary or later funerary imagery.Footnote ⁹ Even taking into account this eclecticism and the element of comparison provided by the Arruntii building, however, Sappho’s scene, dominating the apse of the Porta Maggiore building, remains, at the present state of knowledge, unique for its iconography.Footnote ¹⁰ The central role of Sappho’s leap, moreover, is unusual also if we look at its relationship to literary sources, and, more specifically, to actual texts attributable to Sappho. In the next section of this paper, it is my intention to explore some neglected pieces of evidence that may help to shed light on its background from this perspective.

In Love with Phaon

In a recent contribution, I have suggested that an eschatological interpretation of Sappho’s leap might have been known already in late fifth-century-BCE Athens. In a fragment of a play of the comedian Pherecrates, The Miners, a woman comes back from the Underworld and describes the blessed life of its residents. An interlocutor interrupts her lengthy description, noting that she might cut it short and plunge (κολυμβάω) back again into Tartarus (fr. 113.20–21 K-A). I have argued that the woman in this episode is likely to have been either Sappho, coming back from the realm of the dead, or a character inspired by Sappho.Footnote ¹¹ This might have been based on representations of the Underworld and metaphorical images present in lyric poems of Sappho herself. A fragment of Anacreon (376 PMG = 94 Gentili, 23 Leo) shows that the motif of the leap from the Rock of Leucas (here, too, the verb κολυμβάω is used) was connected to love and suffering already by the late sixth or early fifth century. Since the motif was later linked to Sappho, and Anacreon conspicuously showcased his poetic debt to Sappho, it is not unlikely that the leap metaphor might have already been used by Sappho herself.Footnote ¹²

But was there any connection between this and her unrequited love for Phaon in Sappho’s own songs? Or, indeed, was there any mention of Phaon at all? The evidence is apparently conflicting. There is only one source apparently implying that Phaon was the subject of songs by Sappho. In all other texts that mention her love for Phaon the story is told as part of her biography, but without any reference to her poems. In some of them, indeed, the Sappho in love with the young boy is even presented as a different character from the famous poetess herself.

The only source that would seem to mention explicitly lyric songs composed by Sappho for Phaon is a short story inserted in the collection of Unbelievable Tales (περὶ ἀπίϲτων) that goes under the name of Palaephatus. Palaephatus was a Peripatetic ‘historian’, a direct pupil of Aristotle, active in the fourth century BCE, author of various mythographic and paradoxographic works characterised by their rationalising approach, mentioned and quoted by several ancient sources. Whether the collection of tales handed down to us under this name is actually the work of this same fourth-century author or it is a later work nobilitated by being tagged to a famous author has been the subject of a long debate. Nicola Festa, editor of the work in Teubner’s Mythographi Graeci, was of the opinion that this was a later work that may have included original materials.Footnote ¹³ More recently, Greta Hawes and other scholars, including Renee Nünlist, the editor of Palaephatus’ fragments and testimonies in Brill’s New Jacoby, are ready to accept that the collection, too, goes back substantially to the real Palaephatus.Footnote ¹⁴ At any rate. it is frequently assumed that our witness about the presence of Phaon in Sappho’s songs may go back to the fourth century BCE.Footnote ¹⁵

Here is the whole short story, with a critical apparatus limited to its last sentence.

7 ἐφ’ ὧι τὸν codd: ἐφ’ ὧι <διὰ> τὸν Tollius Reference Tollius1649 || 8 ἆιϲμα Tollius Reference Tollius1649: αἷμα codd. [Palaephati] et [Eudocia]; ἐμελοποίηϲεν Festa Reference Festa1902; ἅλμα W. Hoerschelmann Reference Hoerschelmann1891

There are two points that need addressing in this passage. The first is a relatively minor textual issue. The transmitted text of the final sentence, with its crucial reference to Sappho’s songs,Footnote ¹⁶ is obviously corrupted. All manuscripts read αἷμα ἐποίηϲε(ν). The meaning of this would be ‘she made blood’, which makes no sense in the context. Most editors until Festa printed ἆιϲμα ἐποίηϲε, which Festa defines as the vulgate reading and attributes to his siglum ‘E’, i.e. what he calls ‘personata E(udocia)’, an author using the pseudonym of Eudocia. The reference is to the so-called Eudocia’s Violarium, a large compilation of stories and biographies attributed to the Byzantine Empress Eudocia Macrembolitissa (eleventh century) but in fact a forgery assembled by the Greek scribe and erudite Constantine Palaeocappa in the sixteenth century and based, among other sources, also on Palaephatus. However, Palaeocappa’s own autograph, preserved in Par. gr. 3057, has the same reading as all manuscripts of Palaephatus. The vulgate reading does appear (with the wrong accent, ᾄσμα) in the eighteenth-century first edition of this work by de Villoison.Footnote ¹⁷ The correction, though, goes back already to the 1649 edition of Palaephatus by Tollius,Footnote ¹⁸ who also proposed to read <διὰ> τὸν ἔρωτα (as correctly reported by Festa). This is a very easy correction, even if it entails a certain linguistic clumsiness, since ἆιϲμα should be understood not as the direct object of the verb but as a predicate to τὸν ἔρωτα αὑτῆϲ (literally ‘on whose account she often made her love into a song’). I cannot find precise parallels for this construction, but it would be difficult to argue that it is not possible.Footnote ¹⁹ I suppose that this difficulty must have led Festa to his conjecture ἐμελοποίηϲεν, accepted, to my knowledge, by all subsequent editors. This is linguistically much easier, and its origin can be explained if we assume that the corruption to αἷμα ἐποίηϲε originated not as a palaeographical error but as the result of a phonetic mistake with emelopiisen corrupted into ema epiisen (though Festa did not state this). It seems difficult, at any rate, to envisage alternative textual solutions not entailing a reference to Sappho’s song.Footnote ²⁰ The fact remains that this unique piece of information rests upon a somewhat uncertain conjecture.Footnote ²¹

The second point is of much greater consequence. All editors of Sappho and all scholars who have dealt with the implications of this source (with the single exception of Wilamowitz, impressively ignored in subsequent scholarship on this issue)Footnote ²² have either not noticed or failed to draw attention to the fact that, even if we accept the hypothesis that our collection of Unbelievable Tales actually goes back to the fourth century BCE, Phaon’s story belongs to a group of seven tales appended to the rest of the collection only in a late branch of the manuscript tradition that all editors consider to be an interpolation which has nothing to do with the rest of the collection (they also lack the rationalising approach characteristic of Palaephatus and of the rest of the work).Footnote ²³

Summing up, our only source mentioning Phaon as the object of Sappho’s songs is a late, possibly even Byzantine, interpolation in a mythographic compendium, whose uncertain phrasing rests upon a conjecture. Other sources, starting with the passage from Menander’s Leucadia (fr. 1 Austin, Blanchard, 258 K.–T.: our earliest evidence of the circulation of the love story with Phaon), involve Sappho as a character in the story, not as its source, and make no reference at all to poems composed about Phaon among the works of the poetess. A significant group of witnesses, indeed, attribute the unhappy love story to a different character, a namesake of the famous Sappho.

Another Sappho?

What might be the earliest source mentioning another Sappho is a textually problematic passage of Book 13 of Athenaeus’ Deipnosophistae, where one Sappho from Eresos features within a survey of famous hetairai (‘courtesans’) (596e):

τῆϲ ἑταίραϲ A: τῆϲ ἑτέραϲ (Musurus) 〈ὁμώνυμοϲ〉 Casaubon; τῆϲ 〈ποιητρίαϲ ὁμώνυμοϲ〉 ἑταίρα Kaibel; ἡ ἐξ Ἐρέϲου δέ τιϲ ἑταίρα Σαπφώ Meursius; fort. τιϲ ἑτέρα <ὁμώνυμοϲ ἑταίρα> Σαπφώ vel etiam τιϲ ἑτέρα Σαπφώ (D’Alessio) || Νύμφιϲ A: Νυμφόδωροϲ Wilamowitz, cf. 6.265c, 7.321f, 8.331e

The identification of the source from which Athenaeus derives this piece of information could in theory help to provide a chronological context. The manuscript has the name Nymphis, a third-century-BCE author whom Athenaeus quotes elsewhere for a work on his hometown, Heraclea on the Black Sea, a topic that does not seem relevant here. On the other hand, Athenaeus quotes several times a Nymphodorus of Syracuse, twice attributing to him a Voyage around Asia: in the first passage (Book 7, 321f–322a, 572 F 5 BNJ ²) regarding a poetess from Lesbos, and in the second one (Book 13, 609e, 572 F 7a BNJ ²) when dealing with the beauty of the women of Tenedos (an Aeolian island not far away from Lesbos). Wilamowitz hence plausibly conjectured that Nymphis in our passage is a corruption for Nymphodorus. Nymphodorus’ date is uncertain: he must be earlier than Pliny the Elder, who quotes his geographical works, and is usually thought to belong to the early Hellenistic period.Footnote ²⁴ One of the main issues of this text is whether Nymphodorus (or Nymphis) stated that the poetess Sappho was a prostitute and that she was in love with Phaon,Footnote ²⁵ or whether he thought that there were two different Sapphos, the poetess and the prostitute. The text is undoubtedly defective and some sort of correction is necessary: the presence of the meaningless genitive τῆς ἑταίρας is more easily accountable if there was a mention of two different homonymous individuals (as in the text as reconstructed by Casaubon and Kaibel). A solution such as that of Meursius (ἡ ἐξ Ἐρέϲου δέ τιϲ ἑταίρα), though, could be taken as attributing to Nymphodorus the mention of just one Sappho, ‘a certain hetaira’. Even so, however, it would be very difficult to suppose that Nymphodorus/Nymphis (and Athenaeus) would have referred so vaguely to the poetess as τιϲ ἑταίρα. The indefinite pronominal adjective (‘a certain hetaira’) implies that we are dealing with a potentially unfamiliar character, and the lack of any mention of the fact that this Sappho was also a poetess makes it very unlikely that the reference was to the same individual as the famous lyric author. Whether or not the issue of homonymy was present in the text, it seems reasonable to assume that Nymphodorus/Nymphis mentioned this character as a different Sappho, as it more explicitly happens within the parallel passage of Ael. VH 12.19 πυνθάνομαι δὲ ὅτι καὶ ἑτέρα ἐν τῇ Λέϲβῳ ἐγένετο Σαπφώ, ἑταίρα, οὐ ποιήτρια: ‘I know that there was also another Sappho in Lesbos, an hetaira, not a poetess.’

In other sources, too, the Sappho who falls in love with Phaon is an hetaira or, at any rate, a different individual from the poetess. A second text, though evidently later, has great relevance:

schol. Liban. Ep. 257 = Paus. Att. φ 4 Erbse (I) ≅ Phot. φ 100 Th. (II) ≅ Et. Gen. AB s.v. Φάων (gl. 164 Calame) = Et. Sym. FCVE s.v. Φάων (III) ≅ Suda φ 89 A. (IV). Cf. Apostol. 17,80 = Arsen. 52,16.

Phaon: (a noun used) for handsome and arrogant people: for many say that Sappho, not the poetess but <…> from Lesbos, fell in love with Phaon, and having been unsuccessful threw herself down from the Rock of Leucas.

Hartmut Erbse traced back this passage to the work of Pausanias, a lexicographer active at the time of Hadrian, and the author of an Ἀττικῶν ὀνομάτων Συναγωγή (a Collection of Attic Words), which was widely used by later sources.Footnote ²⁷ Pausanias was a highly learned scholar, who made abundant references to many early authors, based on his obvious first-hand familiarity with their current editions, at a time when many of these editions (including that of Sappho) were available in the libraries of most important and even of minor provincial urban centres. It is very unlikely that an erudite source of this calibre would flatly deny that Sappho ‘the poetess’ was in love with Phaon if there had been any actual poems on the subject in the standard edition of Sappho’s poems current at his time.Footnote ²⁸ An alternative Sappho is linked to Phaon also in the double biography (entries Σ 107 and 108) provided in the Byzantine encyclopaedic lexicon Suda, where the entry of the famous poetess is followed by the one of another Sappho:

In Suda the famous poetess (entry Σ 107) has nothing to do with Phaon, while a different Sappho, a harp player (a category of female musicians often linked with the role of entertaining at symposia, and in several ways contiguous with hetairai), is said to have been in love with Phaon and to have thrown herself into the sea from the Rock of Leucas. Very interestingly, she too is said to have composed lyric poetry, suggesting that there must have been poems attributed ‘by some’ to this other Sappho.

The story of Sappho and Phaon belongs to the construction of Sappho’s persona and seems to have become an important trait of her reception from an early date. In the Hellenistic period, when a definite textual body of poems attributed to Sappho started to be available within authoritative editions/collections, the lack of correspondence between the anecdotal tradition and the content of the poems themselves became a problem and must have played a role in the development of the double-Sappho theory.Footnote ²⁹ It is certainly no coincidence that most of the sources on Sappho the hetaira/harp player, when they do provide additional details (Nymphis/Nymphodorus, Pausanias, Suda), turn out to be all centred around the Phaon story. This very strongly suggests that this story might have been entirely absent from the consolidated corpus of Sapphic poetry, and this must have been one of the reasons (if not the reason) behind the development of the double-Sappho theory.Footnote ³⁰

The Third Cologne Poem: a Missing Link?

There were other avenues to keep the textually sanctioned Sappho and the protagonist of the Phaon story together. The Epistle of Sappho to Phaon is a remarkable example of a possible alternative. The dialectic tension between lyric and elegiac tradition in the Epistle has recently provided key insights for its understanding.Footnote ³¹ Here, too, the importance of the (arguable) lack of lyric poems for Phaon in Sappho’s standard edition might have played a more important role than hitherto surmised. In the Epistle, Sappho’s unhappy love story with Phaon in fact coincides with the end of her career as a lyric poet, and this may explain why we have no lyric poems of Sappho for Phaon in her standard editions.

In the next section of this paper I am going to argue that there were other, earlier attempts at dealing with the issue and that we may indeed have a neglected piece of evidence that can help us to figure out how the story of a pseudo- or deutera-Sappho took shape, bringing to light what might be one of the missing links with the apse of Porta Maggiore’s hypogeum.

The clue is provided by what is to date the most ancient Sappho papyrus. A first portion was published more than twenty years ago, in 2004, from a cartonnage of unknown provenance acquired by the University of Cologne (P. Köln 429). It is dated on palaeographical grounds to the first half of the third century BCE and preserves in two columns two poems in the same metre and dialect, the second of which (58c Neri = 58.10–21 Voigt) was already known also from a papyrus roll with Book 4 of the later ‘standard’ edition of Sappho (POxy. 1787). The two poems share several thematic and linguistic features. The first, very fragmentary poem (58b Neri) seems to focus on the destiny of the poetess in the Underworld. The second, much better preserved, and already partially known as Sapphic, is the so-called ‘Tithonus’ poem (58c), featuring the poetess’s considerations on old age. The language, metre, content, and style of the first poem are perfectly compatible with an attribution to Sappho, too.

In 2005 the same editors published a further portion of the second column contributing a third lyric poem (PKöln. 430). This was added by a possibly slightly later hand immediately after Sappho’s ‘Tithonus’ poem and is linked through imagery and keywords to the two preceding poems. Most scholars have realised that language, style, and metre of the third poem are incompatible with an attribution to archaic Lesbian poetry.Footnote ³² It is a rare example of lyric poetry produced between the fourth and the third century BCE. Here is the text as reconstructed by Gronewald and Daniel (Reference Gronewald and Daniel2005):

Weaver of whispers, deceiver, fabricator of fables (…), treacherous boy (…), my companion, I go away (…) without breathing (I abandon) (…) the light and of the stars (…) the fire-blazing (… of the) sun (…) I hear (…) the son of Oeagrus, Orpheus (…) all animals (…) the lovely (…) [good]-sounding lyre (…) holding (female subject) as a helper (…)

Without providing here a complete status quaestionis, I will focus straight away on what I see as the most obvious reading of this very fragmentary text, after dealing with only a few preliminary issues. One of the first questions to address is that of the function of the dicolon, the two superposed dots that appear at least at ll. 3 and 8 and possibly at ll. 4 and 5. This is attested in ancient papyri to signal changes of speaker, but also as a punctuation mark and/or to indicate internal articulations of metrical sequences. This is the function it has, for example, in an early-second-century-BCE papyrus preserving a text that has several points of contact with our poem, the so-called Fragmentum Grenfellianum, known also as the Maiden’s lament (PGrenf. 1), a Hellenistic lyric poem giving voice to the monologue of a woman abandoned by her lover.Footnote ³³ The alternative explanation, that of its use as an indicator of change of speaker, was formulated as a possibility already by the first editors and endorsed by Rawles (Reference Rawles2006). The overall interpretation of this poem as a dialogue, though, with (in Rawles’ reconstruction) a hymnic opening, followed by an exchange with an address by a speaker to a fellow performer, seems contrived. This is true also of the articulation envisaged by Gangutia (Reference Gangutia Elícegui2010) 17, who posits a change of speaker between ἑταῖρε and ἀφέρπω, i.e. immediately before the dicolon (and so effectively ignores the actual evidence on which this interpretation should have been based), and by Martínez Nieto (Reference Martínez Nieto2011), who argues for a female speaker declaring her intention to abandon her lover and uttering a magical charm that would have been followed by the words of her lover (who, in the meantime, would have changed his mind). The remains of the whole text in my opinion can be interpreted much better as a coherent speech by a single speaker, as envisaged by most other interpreters: the poem opens with an address to an individual characterised in negative terms (ll. 1–2) and continues with a statement of the speaker’s intention to take leave from his/her partner (l. 3 ἑταῖρε ἀφέρπω). The last preserved section contains a section on Orpheus and his lyre, an instrument possibly held by a female speaker.

Who is the individual addressed in these lines, and who is the person addressing him? Some of the adjectives of the first two lines can all very well be applied to the gods of love (cf., for example, Aphrodite δολοπλόκοϲ (‘weaver of deceits’) in Sapph. 1, and Eros μυθοπλόκοϲ (‘weaver of stories’) in Sapph. 188) as well as to a lover. If they were addressed to the speaker’s lover, the Fragmentum Grenfellianum would provide excellent parallels for δόλιε and ἐπίβουλε, matched by ἐπιβούλωϲ (‘treacherously’) at v. 5, and φρεναπάτηϲ (‘mind-deceiver’) at v. 18.Footnote ³⁴ In the Cologne papyrus the use of ἀφέρπω, though, settles the question. The basic meaning of the verb is ‘to go away’. It may be understood also as a euphemism for ‘dying’, a meaning for which we have only one example in a pseudo-Pythagorean Doric fragment transmitted by Stobaeus and attributed to a pseudepigraphic work of the obscure and certainly fictional Pempelos.Footnote ³⁵ In our text, however, after the series of deprecatory address terms, and after the vocative ἑταῖρε, it would be very difficult not to understand it in its primary sense, ‘I go away’ (from you). Death may be the final goal of the speaker’s movement (and, as I am going to argue, it actually is, in a way), but the most immediate implication at this stage is that the speaker is ‘taking leave’ from a harmful addressee. And it is from a lover not from a divinity that the speaker can take leave.Footnote ³⁶

But who is the speaker in our poem? And who is the perfidious partner? According to most scholars, based on the feminine participle at the nominative, ἔχοιϲα at l. 13, the speaker in our text is a woman. This cannot be considered as completely certain, but, within the context, it seems indeed probable. At l. 5 ἄπνουϲ (nū, rather than lambda, is the most convincing reading, as the first stroke is practically vertical) is a two-ending adjective and can be feminine as well as masculine. If we are dealing with a female speaker, the use of ἑταῖροϲ for a heterosexual partner can be paralleled in Ar. Eccl. 911–13,

This is particularly relevant for our poem as the Aristophanes passage is rightly considered an early occurrence of the lyric motif of the lament of the abandoned woman, modelled upon the famous ‘quasi-Sapphic’ ἐγὼ δὲ μόνα καθεύδω (Sapph.? 168B), paralleled also in Fragmentum Grenfellianum 35.Footnote ³⁸

The adjective ἄπνουϲ (‘breathless’) has been understood already by the first editors as referring to ‘death’ or ‘apparent death’. The term may echo Sappho’s attention to the technical description of medical symptoms and can be paralleled in the Hippocratic corpus, On Diseases 2.20, where the symptoms of an ἄπνοοϲ individual include ψῦχοϲ, ‘cold’, and ἱδρώϲ, ‘sweat’, as in Sapph. 31 (describing a near-death experience, cf. v. 15) and lead to death within three days.Footnote ³⁹ Following this, προ̣[λείπω or προ̣[λίποιμι would be a very attractive supplement expressing the speaker’s intention to abandon the world of the living (ed. pr., ‘I abandon (…) the light and of the stars (…) the fire-blazing (of the) sun (…)’; προ̣[λιποῖϲα would be another exempli gratia alternative). As noted already in the editio princeps, this would find a close parallel in the words of Praxilla’s Adonis in the Underworld (where the verb used is the simple λείπω):

‘Sillier than Praxilla’s Adonis’: used of stupid people. Praxilla of Sicyon was a lyric poetess, as Polemon says. In her hymns (or ‘songs’, a variant in the manuscript tradition) this Praxilla represents Adonis as being questioned by those in the Underworld about the most beautiful thing he left behind when he came, and giving as his answer:

We find an even closer turn of phrase, with the same verb supplemented here, προλείπω, in the words uttered by Aphrodite when describing her descent into the Underworld in search of Adonis in the new Orphic fragments of the Sinai Palimpsest (fr. B, vv. 11–13):Footnote ⁴⁰

The verbs λείπω and προλείπω are not infrequent in funerary contexts with the Sun as their object,Footnote ⁴¹ προλείπω is attested in the ‘Orphic’ golden tablet from Thurii (OF 487.1 B.: ἀλλ’ ὁπόταν ψυχὴ προλίπηι φάοϲ ἀελίοιο, ‘when a soul abandons the light of the sun’), and in OF 339.6 B. (quoted by Proclus, probably from the Rhapsodies): ὁππότε δ’ ἄνθρωποϲ προλίπηι φάοϲ ἠελίοιο,/ψυχὰϲ ἀθανάταϲ κατάγει Κυλλήνιοϲ Ἑρμῆϲ/γαίηϲ ἐϲ κευθμῶνα πελώριον (‘when a man abandons the light of the sun/Hermes Cyllenius leads the immortal souls/in a giant cave of the earth’). The presence of the Moon as the object is limited to Praxilla and the Sinai Palimpsest, but the Moon is likely to have been in a gap in the Cologne papyrus, too, and the stars occur only in Praxilla and the Cologne papyrus (in the Sinai Palimpsest they are represented by the οὐράνιοϲ πόλοϲ). This suggests a conceptual link between the three passages, which becomes even more evident if we consider that all three of them involve (or may involve, in the case of the Cologne papyrus) catabatic experiences of characters related to each other (the pairs Aphrodite/Sappho and Adonis/Phaon).

Against this background, therefore, the fact that the very rare adjective πυριφεγγήϲ (‘fire-blazing’), found at l. 7 in the papyrus, is attested almost only in Orphic contexts acquires additional interest.Footnote ⁴² It occurs in the Orphic Hymns (52.9), in the Orphic Lithica (173), in the Orphic Argonautica (214), and in the famous Hymn to Zeus included in the Rhapsodies, 243.22 B in a variant text quoted by Proclus, who also uses the adjective in one of his Hymns (2.6).Footnote ⁴³ If we consider that this section immediately precedes the mention of Orpheus himself, the idea that the speaker is here linking her/his taking leave from a lover with language connoted as Orphic and catabatic is more than simply likely.Footnote ⁴⁴

Line 8 has always been a stumbling block for the reconstruction of our text. The παϲ sequence is preceded by a trace compatible with a sigma or a nu, and it is difficult, if not impossible to find a word ending in -ϲπαϲ suitable to this context. Μost scholars have thus opted for understanding this as a form of the adjective πᾶϲ. If, based on the single instance of l. 3 (ἑταῖρε ἀφέρπω),Footnote ⁴⁵ we suppose that the scribe made consistent use of scriptio plena, without marking elisions, παϲ should be interpreted as a masculine nominative form, and, since it immediately precedes the first-person verbal form ἀκούω, it would seem natural to infer that the speaker of the whole poem (or of this sequence if we suppose that the text was a dialogue) was a male character. There are, however, several problems with this hypothesis. The first is that at l. 13 we find the Aeolic form (and a frequent trait also of the non-strictly Aeolic lyric diction) ἔχοιϲα. If the scribe and the composer were consistent, we would expect to find παῖϲ here. The composer of this poem and/or its scribe, though, were not consistent in adopting purely Aeolic forms for other lexemes and morphological features. The second problem is that the expression πᾶϲ ἀκούω is very difficult to understand in this context. Most translators understand it along similar lines: ‘ich bin ganz Ohr (?)’ (Gronewald and Daniel, with a significant question mark), ‘io sono tutt’orecchie’ (Lundon), ‘soy todo oídos’ (Gangutia), ‘completely I hear’ (Bierl). When joined to a verb such as ἀκούω, the adjective (or the pronoun) πᾶϲ would naturally be understood as a collective, but this would, of course, be incompatible with the first-person singular form of the verb. The only parallel quoted by the first editors (and by subsequent scholars) is Xenophanes fr. 21 B 24 D. K. οὖλοϲ γὰρ ὁρᾷ, οὖλοϲ δὲ νοεῖ, οὖλοϲ δέ τ’ ἀκούει, ‘he sees in his entirety, in his entirety he perceives, and in his entirety he hears’. This refers to the conception of a non-anthropomorphic god, who, having no separate organs of perception, experiences perception in his entirety. Puglia proposes an alternative interpretation, ‘del tutto [innocente] sento [i tuoi biasimi]’ (‘entirely [innocent] I hear [your accusations]’), supplementing e.g. the adjective ἀναίτιοϲ before πᾶϲ, and implying an adverbial use of the adjective πᾶϲ, modifying the previous, now lost adjective.Footnote ⁴⁶ Similar difficulties apply to the possibility that what we have here is the elided πᾶϲ(α). The most likely solution, thus, is to assume that this was an elided verbal form.Footnote ⁴⁷ The ending of the infinitive active of the sigmatic aorist is elided only in the Lesbian poets (cf. Sapph. 31.7–8 as reconstructed by all modern editors, and Alcaeus. 349a2, unanimously accepted reconstruction of a text based on indirect tradition) and (fairly frequently) in comedy, precisely the two genres that seem to have more clearly influenced the lyric lament of the abandoned maiden. A very suitable supplement in our context would be δι|α]ϲ̣πᾶϲαι, with a verb, διαϲπάω (‘to tear apart’), particularly appropriate to describe the sad destiny of Orpheus, the mythical poet evoked in these lines: ‘I hear that [the women of Thrace] tore apart the son of Oeagrus’.Footnote ⁴⁸ The verb is applied to Orpheus’ death in various sources (cf. Isoc. Bus. 39, [Apollod.] Bibl. 1.15) and more particularly in connection with the women of Thrace in Lucian Adv. Indoct. 11.3. This would fit very well also the traces preserved at the end of the line (θρη [or θρα[), inviting a supplement such as δι|α]ϲ̣πᾶϲ’(αι) ἀκούω Θρα̣[ικίαϲ γ]υ[ν]α̣[ῖκαϲ … Οἰά|γ]ρου κόρον Ὀρφέα.Footnote ⁴⁹ The ‘Lesbian myth’ reported by Lucian Adv. Indoct. is interestingly relevant to our context in that it mentions the death of Orpheus torn apart by the Thracian women (with διαϲπάω, as in my reconstruction here), the journey of Orpheus’ head and of his mourning lyre towards Lesbos, the burial of his head in a sanctuary of Dionysus and the dedication of the lyre to Apollo,Footnote ⁵⁰ and the magical power of the lyre to enchant wild beasts.

In the papyrus poem, too, it would seem that the story of Orpheus’ tragic destiny led, eventually, to the mention of the magic power of his lyre. And this is likely to be precisely the same lyre that at line 13 a female individual holds as her ‘helper’ (ϲυ]νεργόν). Most scholars, since the first edition, have identified this female individual with the speaker of our text. Even so, with the exception of Clayman (and, of course, of Tsantsanoglou, who thinks that this is actually a poem by Sappho), they have been reluctant to take into consideration the hypothesis that the poem might have been in the voice of Sappho herself, allowing only, in one case, for the possibility that some readers might have understood it in this way.Footnote ⁵¹ But, as Clayman notes: ‘in the context of the first two poems, however, what other female would be holding a lyre?’Footnote ⁵² By the third century BCE the tradition that Orpheus’ lyre and his head, after it had been torn asunder by the women of Thrace, had been brought by sea to Lesbos was already well established, as an aition for the island’s musical and poetic excellence.Footnote ⁵³ In our text, the woman who holds the lyre as her helper inherits this tradition, and which other female individual could have been constructed as the heiress to this tradition more appropriately than ‘Sappho’?

If the poem is in the voice of Sappho, who might then have been her ἑταῖροϲ? Scholars, even when they have envisaged the possibility that the individual addressed was not a deity but a lover, have been remarkably consistent in considering his identification with Phaon as on the whole unlikely.Footnote ⁵⁴ Even Clayman, who correctly, in my opinion, sees the poem as composed ‘in the voice of Sappho’, rejects the identification on the ground that Sappho’s descent to Hades would have evoked Orpheus’ own descent, which she maintains must have been linked to his homosexuality, thus ruling out a representation of a heterosexual ‘Sappho’.Footnote ⁵⁵ Clayman accepts, therefore, that the first lines are addressed not to ‘Sappho”s lover but to a deity.

And yet, who else might have been addressed by a ‘Sappho’ taking her leave from him and ready to abandon the world of the living? The epithets of ll. 1–3 square very well with Phaon’s traditional representation. The poem’s addressee is a devious seducer,Footnote ⁵⁶ prone to deceit, and so young as to be called a παῖϲ. All these features will be very familiar to the reader of Sappho’s Epistle to Phaon, where the abandoned Sappho warns Sicilian women about her former lover’s deceitful lies: nec uos decipiant blandae mendacia linguae (‘let the lies of his charming tongue not deceive you’, v. 55) and dwells on his very young age (quid mirum si me primae lanuginis aetas/abstulit, atque anni quos uir amare potest? ‘how strange is it if the age of a young man’s first beard carried me away, and years that could attract male lovers’, vv. 85–6).Footnote ⁵⁷

It is thus possible to offer a coherent reading of the remains of this poem as a speech in the voice of Sappho, who declares her intention to take leave from her partner, a deceitful youth (Phaon himself) and, eventually, to abandon the world of the living, bringing with her a lyre connected to (and possibly inherited from) Orpheus. It is a poem developing the figure and the story of Sappho beyond the corpus of the Sapphic poems transmitted from the Archaic period in an exemplary case of creative reception. The Cologne anthology shows how this was subtly intertwined with the transmission process of the ‘genuine’ poems, a process through which the whole tradition, Archaic, Classical, and Hellenistic, was subject to a complex and varied reinterpretation.Footnote ⁵⁸ The Cologne poems acquire new potential meanings through their arrangement within a book roll, whereby the ‘new’ (pseudo-)Sappho re-orients the reading of the ‘old’ one(s), and, at the same time, gains ‘authority’ not only through its imitative style, but also thanks to its circulation within a written sequence. How this complex transmission interacted with the formation of the standard Alexandrian edition of Sappho’s ‘genuine’ works is a fascinating subject, which is largely covered by darkness.

The conspicuous echoes of Orphic poetry show that in the Cologne papyrus this story was charged with eschatological implications, which, if my former argument about Pherecrates’ Miners is correct, may go back already to the late fifth century. The Cologne poem can be seen as providing a missing link to the majestic iconography of the hypogeum. It would be Orpheus’ lyre that Sappho is holding when facing her farewell to the world. And Orpheus’ lyre was the subject of astrological/eschatological speculations, alluding to the celestial ascension of the soul, in an Orphic poem that went by the title Lyra. As we know from an enigmatic passage in a commentary on Verg. Aen. 6.119 in the manuscript Par. lat. 7930 (OF 417 Bernabé), Varro autem dicit librum Orfei de vocanda anima liram nominari, et negantur animae sine cithara posse ascendere, ‘Varro says that a book of Orpheus on the evocation of the soul was called Lyre, and it is said that it is not possible for the souls to ascend without a lyre’.Footnote ⁵⁹ In our case, with (possibly) Orpheus’ lyre in her hand, Sappho is equipped for her fatal step, a step, though, that would have brought her to a blessed life. This Hellenistic poem thus provides the key, or a key, to understanding the magnificent Sappho of the apse of the Porta Maggiore ‘basilica’ as part of a process of creative representations of Sappho going back at least to the third century BCE, and probably already to the fifth.