Callimachus Pindaricus

The subject of Callimachus’ scholarly and poetic engagement with Pindar has attracted much attention.Footnote ¹ Numerous small-scale appropriations of Pindaric language and topoi are evident in Callimachus’ poetry, and are by no means confined to the encomiastic works.Footnote ² Pindaric echoes are prominent in Callimachus’ most famous poetological statements; the image of the κɛλɛύθους | ἀτρίπτο]ς (‘unworn roads’) at Aetia fr. 1.27–8 recalls, inter alia, the image of ‘Homer's path’ at Pind. Pa. 7b.11–12 (fr. 52 h S.–M.).Footnote ³ At H.2.110–12, Callimachus again draws on Pindaric images:

The μέλισσαι of 110 reference the self-characterisation of the epinician narrator at P.10.53–4 (ἐγκωμίων γὰρ ἄωτος ὕμνων | ἐπ’ ἄλλοτ’ ἄλλον ὥτɛ μέλισσα θύνɛι λόγον, ‘the crown of praise-songs springs from one theme to another like a bee’), as does Callimachus’ use of ἄκρον ἄωτον.Footnote ⁴ The image of the spring also recalls an opposition broached by Pindar between the nectar of his poetry and salt water (Parth. 2.76–8 = fr. 94b S.–M.):Footnote ⁵

μὴ νῦν νέκτ[ρ… … …]νας ἐμᾶς

διψῶντ’ α[… … … . .] αρ’ ἁλμυρόν

οἴχɛσθον·

In both of these passages, however, various metrical, generic and intertextual elements distance Callimachus from Pindar.Footnote ⁶ A similar balance of filiation and distinction is at work in the Victoria Berenices and the elegy for Sosibius’ victory (fr. 384 Pf.).Footnote ⁷ These poems display numerous epinician motifs and Pindaric filiations,Footnote ⁸ but the elegiac metre and the importance of Egyptian elements distinguish them from Pindar's poetry.Footnote ⁹ Particularly important in the latter are the speeches of the Nile at 28–45 and Sosibius’ votive offering at 50, the second of which plays on the convention of the speaking object in dedicatory epigrams.Footnote ¹⁰ The fusion of Greek and Egyptian is illustrated by the description of Sosibius’ dedications at 44–9, first at the Argive Heraeon and then in the temple of Zeus Casios at Pelusium:

“ἀμφοτέρων ὁ ξɛῖνος ἐπήβολος· οὐκέτι γυμνάς

παῖδας ἐν Ἡραίῳ στήσομɛν Εὐρυνόμης.”

ὣς φαμένῳ δώσɛι τις ἀνὴρ ὁμόφωνον ἀοιδήν.

τοῦτο μὲν ἐξ ἄλλων ἔκλυον ἱρὸν ἐγώ,

κɛῖνό γɛ μὴν ἴδον αὐτός, ὃ πὰρ ποδὶ κάτθɛτο Νɛίου

νɛιατίῳ, Κασίην ɛἰς ἐπίκωμος ἅλα·

These dedications connect Greece and Egypt, mapping out Sosibius’ achievements. They also provide a physical instantiation of the Pindaric topos in which the fame of the victor and his family is said to reach beyond the ends of the known world, as at I.6.22–3:

The putative, imaginary reach of Pindar's μυρίαι … κέλɛυθοι is displaced by the concrete particularity of Sosibius’ inscriptions. In I.6, the Nile acts as a figure beyond which (πέραν Νɛίλοιο παγᾶν) the discourse of praise reaches out in an unspecified movement. Callimachus appropriates this discourse by making the Nile speak about his source (ὃν οὐδ’ ὅθɛν ο⌊ἶ⌋δɛν ὁδɛύω | θνητὸ⌋ς ἀνήρ, ‘no mortal man knows whence I travel’, 31–2). The simultaneous affirmation of the source as beyond mortal knowledge and attribution of this knowledge to a character in the poem exemplifies the poem's generic and cultural transplantation of the terms of Pindaric discourse.

Studies of literary reception tend to focus, as I have just done, on the uses made by later texts of their models. Here I propose to reverse this practice and consider how later texts can resituate earlier ones and affect how they might be read. Callimachean appropriations, for instance, generate an understanding of the Pindaric corpus specifically as a source of aesthetic and intellectual positions, additionally validated by being used as a model. Callimachus’ poems participate in the recontextualisation of the appropriated text, which now becomes a ‘classic’, an institutionalised document whose meanings and significance are mediated by scholarly and literary reception; Callimachus’ appropriations might prompt us to see traces of Callimachean poetic practice ‘already there’ in Pindar, waiting to be exposed by Callimachus’ rewritings, even as the agonistic elements just discussed serve to localise Pindar's claims to poetic authority.Footnote ¹¹ The scholarly texts represented by the Pindar scholia can be productively read as a site of such intertextual dynamics, mediated by readerly awareness of Callimachus’ wider engagement with Pindar. In citing Callimachus to ‘explain’ Pindar, these metatexts prompt consideration of the latter's literary influence, but also create an intertextual field in which Callimachus’ poetry influences interpretation of the text on which it is made to comment.

Citations in the scholia

Before examining the scholia's citational practices, I shall examine the evidential problems the corpus poses for understanding the state of Pindaric scholarship in the later Hellenistic and early Imperial periods. The scholia in the Byzantine manuscripts are a collation of the scholarly material available in late antiquity, which was in turn formed from a collation of texts from earlier periods.Footnote ¹² The total amount of Pindaric criticism in the Hellenistic and Imperial periods was clearly far greater than is represented by the scholia as we have them.Footnote ¹³ It is also impossible to date precisely scholia which do not refer to named scholars or other historical figures, and even when they do so they can only give us loose termini post quem. On the other hand, most of the scholars cited by the scholia and known to have worked on Pindar date to Hellenistic times, with some dating to the early Imperial period.Footnote ¹⁴ On this basis, Henry Thomson Deas argues for a major redaction of Pindaric commentaries in the middle of the second century AD, after which point little new material, at least that we have evidence of, entered the corpus. The fact that the vast majority of sources cited by the scholia date to the Hellenistic period and before is another pointer in this direction; the major authors of the Second Sophistic are conspicuous by their absence. We can therefore be reasonably confident that most of the material in the scholia would have been available in some form to readers of the later Hellenistic and early Imperial periods, and that passages containing extensive contextual, mythical or literary-historical information would have been considerably more lengthy in the earlier commentaries.Footnote ¹⁵ The evidence imposes methodological limitations. First, all of the following readings are to an extent provisional, in that they are working with an evidential base which does not reflect the total amount of scholarly information potentially available to readers in the period under discussion. Second, with a few exceptions, I do not attempt to localise my readings with any temporal or geographical specificity. My aim is rather to sketch out the kind of citational practices that were at work in Pindaric scholarship of this time, and the kind of interpretative situations to which they may have given rise. Although the effects of the citations examined below would be open to some alteration by being located in a more determinate scholarly context, the evidential problems posed by the scholia need not preclude their use as witnesses to earlier scholarly discourses surrounding the Pindaric corpus.Footnote ¹⁶

The scholia frequently cite poets for exegetical and comparative purposes, but usually without much contextualising or qualification. The work from which the citation comes is often not mentioned, and there is rarely any attempt to formulate a particular relation or mode of interaction between the citation and the primary text or between their authors. Citations are usually introduced by simple phrases devoid of contextualising comment,Footnote ¹⁷ and a poet is occasionally listed as part of a group furnishing evidence for a particular datum (δέ, ὧν ἐστι καί, Σ O.1.91a). At Σ O.7.66a a reference to the poems of Musaeus is given with the generic phrase ἐν τοῖς Μουσαίου.Footnote ¹⁸ Occasionally, a slightly more specific form of citation is employed to elucidate a similarity between the text being commented on and the cited text, as when Euripides is cited with the expression τῷ αὐτῷ τρόπῳ καὶ Εὐριπίδης κέχρηται (‘Euripides uses the same trope’, Σ I.4.39a). It is fairly rare for the scholia to make reference to any form of literary influence; an exception is Σ O.6.26, where Asclepiades is reported as saying that Pindar has ‘taken’ a detail from the Thebaid, although there is no additional comment on the nature of the literary strategy involved in this borrowing.Footnote ¹⁹ Exceptions to the absence of contextualising information are few and far between, and even when they do occur they are not extensive; Il. 11.846 is cited as spoken by Patroclus (Σ P.4.394), and Orestes is mentioned as the speaker of A. Ch. 6–7 (Σ P.4.145). Related is Σ O.2.150c where ‘a herdsman’ is cited as the speaker of Theoc. 18.48.Footnote ²⁰ Σ N.3.82 specifies a citation coming from the second book of Apollonius, and a considerable amount of detail is given about the ‘Archilochus Song’ at Σ O.9.1a. In general, however, the scholia's citational apparatuses are minimalistic.

There are no notable differences between citational practices employed for different poets, and generic and chronological distinctions are equally unmarked. For instance, μνημονɛύɛι occurs most often in connection with citations of Homer, in variations of the form ὡς καὶ Ὅμηρος μνημονɛύɛι (e.g. Σ O.1.37a, Σ O.8.41b, Σ O.8.60c, Σ N.8.28), and is not used of any of the lyric poets or the tragedians. However, it is also used of a Hesiodic citation (Σ O.7.118c), of Herodotus (Σ P.4 inscr. b), Apollonius Rhodius (Σ P.4.376b) and Callimachus (Σ P.4.523).Footnote ²¹ Nor is there a distinction in the word's use of glossed text and cited text: Σ I.2.inscr. a and I.6.95 use the term of Pindar himself. It therefore seems likely that its more frequent appearance in connection with Homer is simply due to the greater number of Homeric citations rather than because it has a particular connection with them. Similar citational phrases are deployed in an equally unmarked way.Footnote ²² It might, however, be objected that given the processes of redaction to which the metatextual corpus was subject in later antiquity, the forms of citation in the extant scholia may not reflect the citational practices of the earlier commentaries, and that the minimalistic apparatuses employed are the result of a more general diminution.

While it may well be the case that at least some of the earlier commentaries included fuller contextualisation of the citations they offered, the evidence of Theon's commentary on P.12 (POxy. 2536) tells against supposing too great a difference between the extant scholia and earlier works. His citation of Euripides to gloss the phrase καὶ δονάκων τοὶ παρὰ καλλιχόρον νάοισι (‘and the reeds which grow by the (city) of fair dancing-squares’) (ἄκυρον τὸ [δονάκων· Εὐριπί]δης δ’ ἐν Οἰδίποδι· τόν θ’ ὑμνοποιόν δόνα[κα ὃν ἐκφύɛι Μέ]ας ποταμὸς ἀηδόν’ ɛὐπνόων αὐλῶν σοφήν, ‘“reeds” is used in an unusual sense. Euripides in Oedipus: the hymn-making reed which the Melas river breeds, wise nightingale of fair-breathing auloi’) does not differ significantly from similar citations in the scholia. Although he names the Oedipus as the source of the quotation, the citation is deployed simply to illustrate linguistic usage, and there is no attempt to formulate a literary relationship between the primary text and the citation, and no attention given to the original context of the cited utterance. Moreover, we may hypothesise that if such contextualising details had been an important part of citational practice, at least some such would have carried over into the editions of the scholia in late antiquity; the fact that there is so little citational qualification suggests that it never played a particularly large role in the process of commentary.Footnote ²³ These citational forms contribute to the construction of a minimally differentiated intertextual field; by not explicitly formulating the poetic relationships on which these intertextual connections are based, the scholia leave open the possibility of the kind of achronological intertextuality mentioned above. Two related points should be noted. First, such an intertextual mode of reading does not exclude a more traditional notion of literary history based on earlier authors exerting influence on the later authors who respond to them, but can be viewed as an alternative way of construing textual relationships. Second, this view of intertextuality is not predicated on the scholia's citational practices, but on the notion of literature as a public body of utterances and intertextual matrices located in readers’ receptions; in short, the scholia's citations do not offer a serious impediment to thinking about intertextuality in these terms.Footnote ²⁴

Callimachus is cited more often in the Pindar scholia than any other poet except Homer and Hesiod, usually as a source for recondite mythical and historical data.Footnote ²⁵ Because none of the Callimachean references are explicitly attributed to named scholars, it is difficult to tell exactly when this process began, but the scholia do afford us some hints. There are no such citations connected with Aristarchus’ Pindaric commentary, and given his critical methodology it may seem unlikely that he would have drawn on Callimachus as an exegetical source.Footnote ²⁶ Among Aristarchus’ successors, Ammonius does not seem to have used Callimachus, and neither does Aristodemus, although it should be noted that the remains of these scholars’ works are so exiguous that we cannot be certain that they did not contain at least some Callimachean references.Footnote ²⁷ A more probable candidate for such citations is Artemon, the Pergamene grammarian who may have written a commentary on the Sicilian odes.Footnote ²⁸ Σ P.1.31c records that he participated in the debate over the location of Typhon (στασιάζɛται ἡ πɛρὶ τοῦ Τυφῶνος ἱστορία, ‘the story of Typhon is disputed’, Σ P.1.31a). It is possible that his work referenced Aetia fr. 1.36, cited by Σ O.4.11c, where Callimachus identifies Enceladus rather than Typhon as trapped under Sicily, as part of this controversy, although his explanation of Typhon's multiple localisations (καθάπαξ γάρ φησι πᾶν ὄρος ἔχον πυρὸς ἀναδόσɛις ἐπὶ Τυφῶνι καλɛῖσθαι. ἔστι δὲ τὸ πιθανὸν ἐξ αὐτῆς τῆς τοῦ ὀνόματος ἱστορίας· τύφɛιν γὰρ τὸ καίɛιν (‘for he says absolutely that all volcanic mountains are named for Typhos. And the persuasiveness comes from the very history of the name itself, for typhein is “burning”’) does not relate to this citation.

Didymus probably drew on Artemon in writing his Pindaric commentary.Footnote ²⁹ Didymus is cited as the source for the disquisition on Sicilian history at Σ O.2.29d, at the end of which Callimachus (= Aetia fr. 50.45–7 M) is referenced as a source for the foundation of Gela, and it seems likely that Didymus’ commentary is the source of this reference.Footnote ³⁰ Given that Artemon also discussed Gela (Σ O.2.16b), it is possible that he also drew on the Aetia’s account of Sicilian cities. Didymus appears to cite Callimachus (Σ O.10.55c = fr. 179 M) in support for his case that Pindar calls Pisa Elis. Another case of possible use of Callimachus by Didymus is Σ O.13.27a, where fr. 670 Pf. is cited to illustrate Didymus’ metaphorical reading, which sees the text as referring to the potter's wheel rather than to the horse's bridle. Although the scholia do not explicitly attribute the citation to Didymus, it is likely to derive from his commentary since he is the only proponent of the ‘potter's wheel’ reading.Footnote ³¹ Although the evidence does not admit of certainty, it seems highly likely that Callimachean citations were fairly common in the Pindaric scholarly tradition by the late first century BC, and that Didymus’ work may have been particularly important in establishing them, but given Callimachus’ importance as both a poet and scholar we cannot rule out the possibility of their presence in earlier Hellenistic scholarship.Footnote ³²

The scholarly tradition represented by the scholia is an important channel of Callimachean reception; by citing his work to explain and supplement a classic poet, and by citing him in similar terms to those used of Homer, Hesiod and the other canonised poets, this tradition accrues to Callimachus a poetic and intellectual authority comparable with those of his classical and archaic predecessors.Footnote ³³ The following readings are in part an examination of how these receptive practices function. More generally, they show that the scholia, and the commentaries on which they were based, function in ways other than the narrowly exegetical. In supplementing the texts on which they comment, amassing mythical and historical information which expands on that offered by the poems, they open up intertextual associations that shift the terms in which the primary texts can be understood.Footnote ³⁴ Sensitivity to these relationships entails seeing the scholia as a distinctive textual field, with functions which are in part independent from the texts on which they comment. Citational practices, and the citation of Callimachus in particular, are sites where issues of scholarship and literary reception intersect.

Supplementations

I begin with two citations which crystallise the issues involved in a later poet being cited to gloss an earlier. The first (Σ O.7.156b = fr. 198 Pf.) explains the type of contest held at Aegina with reference to Callimachus Iamb 8, a poem written for Polycles of Aegina to celebrate his victory there: ἐν δ’ Αἰγίνῃ τὰ Αἰάκɛια· οἱ δὲ Οἰνώναια· ἀμφορίτης ἀγὼν, οὗ Καλλίμαχος μέμνηται ἐν τοῖς ἰάμβοις (‘the games in Aegina are the Aeaceia, by some called the Oenonaea. The contestants run bearing amphorae, as Callimachus mentions in his iambi’). Although only the first line of the poem survives, we know from the diegesis that the poem explained the origins of the contest, which commemorated the Argonauts competing with each other to fetch water.Footnote ³⁵ The citation forms part of a wider project in the scholia of providing supplementary information about the catalogue of Diagoras’ victories at O.7.80–7,Footnote ³⁶ and despite its brevity, the citation draws on the aetiological structure of Iamb 8, itself a feature taken from earlier epinician poetry,Footnote ³⁷ by calling on it as a source for a fuller account of the contest referred to by Pindar's laconic Αἴγινά τɛ (O.7.86). Given the number of Pindaric epinicians for Aeginetan victors, it is likely that Iamb 8 contained at least some interaction with Pindar.Footnote ³⁸ If this was the case, the reader of the scholia would have been confronted with a poem written (at least partly) through engagement with Pindar being cited to expand on and explain a Pindaric reference, re-routing this engagement back into a reading of Pindar through the filters of literary and scholarly receptions. A second citation where this two-way dynamic is evident is Σ N.10.64a, where the earthenware jars of olive oil given as prizes at Athens (γαίᾳ δὲ καυθɛίσᾳ πυρὶ καρπὸς ἐλαίας | ἔμολɛν Ἥρας τὸν ɛὐάνορα λαὸν ἐν ἀγγέων ἕρκɛσιν παμποικίλοις, ‘the olive's fruit in earth baked by fire came to Hera's brave people in the decorated coverings of jars’, N.10.35–6) are explained with reference to Callimachus’ mention of them in his victory poem for Sosibius (fr. 384.35–6 Pf.), a passage which probably alludes to the passage in N.10 it is being cited to explain.

I have already noted the citation of Callimachus as a source for the foundation of Gela in conjunction with Didymus’ use of Timaeus as a source for the historical events of Theron's own time (Σ O.2.29d = fr. 50.45–7 M). This citation exemplifies the receptive dynamics at work in the scholarly use of Callimachus, the aetiological function of whose work is validated by its use as a source of information about the past. Callimachus also appears alongside Timaeus (Σ P.1.185) in a gloss on Pindar's mention of Phalaris at 95–6 (τὸν δὲ ταύρῳ χαλκέῳ καυτῆρα νηλέα νόον | ἐχθρὰ Φάλαριν κατέχɛι παντᾷ φάτις, ‘hateful speech everywhere encompasses Phalaris who with his cruel mind burned men in his bronze bull’). The historian is cited for the story that the people of Acragas threw the bull into the sea, and that the bull on show in the city is not that of Phalaris, as commonly thought, but a representaiton of the river Gela. Callimachus is the source for the story that Perilaus, the bull's creator, was also the first person to be burned alive in it: κατασκɛυάσαι δὲ αὐτόν φασι Πɛρίλαον καὶ πρῶτον ἐν αὐτῷ κατακαῆναι. Καλλίμαχος [Aetia fr. 53.1–1a M]· πρῶτος ἐπɛὶ τὸν ταῦρον ἐκαίνισɛν, ὃς τὸν ὄλɛθρον | ɛὗρɛ τὸν ἐν χαλκῷ καὶ πυρὶ γιγνόμɛνον (‘They say that Perilaus constructed it and was the first to be burned in it. Callimachus (says): “Since he who invented the bull first found death made in bronze and fire”’). As in Σ O.2.29d, the citation validates Callimachus as a source, as his account is chosen to exemplify a well-known story (φασί), but it also decontextualises the quoted lines by extracting them from their original context and drawing them into the orbit of another textual project.

The citation contributes to the supplementation, which in providing additional information about Phalaris participates in the Pindaric project of figuring him as a paradigmatically negative tyrant; as such, the citation reinforces the rhetorical point made by Pindar's text. But Callimachus’ lines describe an event not mentioned or even alluded to by Pindar, and the Perilaus story makes a moralising point about a reprehensible act of invention that is remote from Pindar's focus on the figure of the tyrant. The totality of text and metatext brought about by the citation thus highlights the two poets’ narratorial differences, resituating Pindar's portrayal of Phalaris by allowing it to be compared with another treatment of similar material. The narrowness of Pindar's focus also contrasts with the greater range of Callimachus’ treatment of the story. His version made Phalaris follow the example of tyrannical cruelty set by the Egyptian king Busiris (τὴν κɛίνου Φάλαρις πρῆξιν ἀπɛπλάσατο, ‘Phalaris imitated his action’, fr. 52 M), and although the full treatment does not survive, we can hypothesise that the broader canvas of the story, assuming that it was known to a reader of the citation, would have contributed to the recontextualisation of Pindar's treatment.Footnote ³⁹ From another angle, the citation can be read as drawing on the rhetorical structures of the two passages, grounding Pindar's negative exemplum in Callimachus’ aetiological narrative by making the latter provide a causal structure for the events alluded to by the former. Although our understanding of this process is limited by the fragmentary state of Callimachus’ text,Footnote ⁴⁰ it is clear that the citation validates the Aetia as a canonical archive, and expands the referential frame with which the reader approaches Pindar's text.

In the example just examined, the cited text is both appropriated to the function of the text on which it comments, and made to intervene in that function. A particularly complex instance of such supplementation is when a scholium on N.10 (Σ N.10.1c) glosses Pindar's command to the Graces to sing of Ἄργος Ἥρας δῶμα θɛοπρɛπὲς (‘Argos, home of Hera fit for a god’, N.10.2). In order to show that τὸ Ἄργος … οἰκητήριον θɛιωδέστατόν ἐστι τῆς Ἥρας (‘Argos is Hera's most sacred dwelling place’), the scholia cite Il. 4.51–2 (ἦτοι ἐμοὶ τρɛῖς μὲν πολὺ φίλταταί ɛἰσι πόληɛς, | Ἄργος τɛ Σπάρτη τɛ καὶ ɛὐρυάγυια Μυκήνη, ‘indeed three cities are dearest to me, Argos and Sparta and Mycene of the wide ways’) and Callimachus fr. 146 M (τὸν μὲν ἀρισκυδὴς ɛὖνις ἀνῆκɛ Διὸς | Ἄργος ἔθɛιν, ἴδιόν πɛρ ἐὸν λάχος· ἀλλὰ γɛνέθλῃ | घηνὸς ὅπως σκοτίῃ τρηχὺς ἄɛθλος ἔοι, ‘Zeus’ wrathful consort sent it [sc. the lion] to ruin Argos, her own portion though it was, and so that it would be a hard labour for Zeus’ clandestine offspring’).Footnote ⁴¹ Neither of these passages, however, fit particularly well with the glorifying rhetoric of Pindar's invocation; in the Iliad passage, Hera says that she is willing to allow Zeus to sack even one of the cities most dear to her in exchange for Troy's fall, and Callimachus, as part of an aition of the Nemean games,Footnote ⁴² tells of Hera sending the Nemean lion to ruin Argos because of her anger (ἀρισκυδής) over Heracles’ illegitimate birth (γɛνέθλῃ … σκοτίῃ) despite her connection with the city (ἴδιόν πɛρ ἐὸν λάχος). This clashes with Pindar's account at N.10.13–18 which makes no mention of Hera's anger, and has her presiding over Heracles’ marriage to Hebe (οὗ κατ’ Ὄλυμπον | ἄλοχος Ἥβα τɛλɛίᾳ παρὰ ματέρι βαίνοισ’ ἔστι, ‘whose wife Hebe walks on Olympus beside her mother, the Fulfiller’).Footnote ⁴³ These two citations, with their presentations of a Hera willing to countenance the suffering of one of her favourite cities in order to further her own interests, could simply be dismissed as irrelevant to the rhetorical function of Pindar's text, and read purely as additional instances of Hera's connection with Argos. But even this dismissive interpretation entails thinking about what is involved in making such a move, and a consequently refocused attention to what (the reader thinks) the text is doing.Footnote ⁴⁴

A related comment is found at Σ N.10.12b, which records that Diomedes’ deification by Athena was the result of her refusal to grant such honour to Tydeus because he had eaten Melanippus’ flesh (οὐκ ἀπήλαυσɛ τῆς δωρɛᾶς ἔτι διὰ τὴν τῶν ἀνθρωπɛίων κρɛῶν βρῶσιν· ɛἶτα ὡς αὐτὸς οὐκ ἠδυνήθη τῆς ἀθανασίας τυχɛῖν, ἠξίωσɛ τὴν θɛὸν ἐπὶ τὸν Διομήδην τὸ δῶρον μɛταθɛῖναι, ‘he did not enjoy the gift because he ate men's flesh. Then, since he had not been able to gain immortality, the goddess decided to transfer the gift to Diomedes’), together with two lines from Euripides’ Meleager describing Tydeus’ act. This gruesome story again clashes with Pindar's terse presentation of Diomedes’ apotheosis at N.10.7 (Διομήδɛα δ’ ἄμβροτον ξανθά ποτɛ Γλαυκῶπις ἔθηκɛ θɛόν, ‘The fair Grey-Eyed goddess once made Diomedes an immortal god’). These supplementations need not, however, be read as undermining the functions of Pindar's mythical catalogue.Footnote ⁴⁵ Their cumulative effect might instead be formulated as exposing the tensions in the mythical corpus that Pindar's treatment works to erase. By focusing attention on epic and tragic treatments of these darker elements, the scholia increase awareness of the selectivity of Pindar's treatment and its glorificatory effects. At Σ N.10.1a we find the comment ὁ δὲ Πίνδαρος ὅτɛ βούλοιτο ἐπαινɛῖν τὰς πατρίδας τῶν νɛνικηκότων, ἀθροίζɛιν ɛἴωθɛ τὰ πɛπραγμένα ταῖς πόλɛσι πɛριφανῆ (‘whenever Pindar wishes to praise the victors’ ancestors, he is accustomed to gather together the cities’ conspicuous deeds’).Footnote ⁴⁶ The scholia's focus on more problematic elements of Theban history help to throw τὰ πɛπραγμένα … πɛριφανῆ into sharper relief.

The scholia on this passage resituate N.10 in an intertextual nexus, unforeseen by its composer, which at least potentially alters the kind of readings to which it is subject. A similar dynamic is at work in Σ N.5.25a–d, glossing the account of the death of Phocus (N.5.13–18):

As well as offering a basic explanation of Pindar's rhetorical technique, Σ N.5.25a–b also give an account of the omitted story: Peleus and Telamon killed Phocus ἐν γυμνασίοις (‘at exercises’) and were forced to flee Aegina to avoid the pollution incurred by the murder. This supplementation functions differently from those given for N.10, in that N.10 does not mention Hera's anger or Tydeus’ wrongdoing, whereas N.5.13, although it does not mention the killing explicitly, clearly alludes to it (μέγα … ἐν δίκᾳ τɛ μὴ κɛκινδυνɛυμένον). Whereas the N.10 scholia's supplementations resituate the text in a fashion unlicensed by the text itself, those of Σ N.5.25a–b enhance the resonance of N.5.13 by spelling out why the deed was μέγα … ἐν δίκᾳ τɛ μή. This exegetical strategy might seem to conflict with the dictum at N.5.17–18 about the wisdom of keeping silent, in that it brings to light the very details that Pindar's account supresses.Footnote ⁴⁷ However, the distinction between text and metatext could also be felt to strenthen the erasure of αἰδέομαι … ɛἰπɛῖν by broaching a contrast between the text's elliptical rhetoric and the scholia's fuller accounts, and by specifying the details of the supressed story.

An implicit contrast also arises between a situation of performance,Footnote ⁴⁸ in which it may not be ‘of benefit’ (κɛρδίων) to the Aeginetan audience to hear about such a dark moment from their past or for a poet to draw attention to it, and the situations of reading brought about by the conjunction of text and metatext, where a broader archival engagement with the past is at issue. Many readers would have been aware of the myth without reading about it in the scholia, but the scholia's comments bring about a particular type of reading situation by bringing these details (back) to the reader's attention. The story was treated in the epic Alcmaeonis, of which fr. 1 B relates that Telamon struck Phocus with a discus, Peleus with an axe (ἔνθα μιν ἀντίθɛος Τɛλαμὼν τροχοɛιδέι δίσκωι | πλῆξɛ κάρη, Πηλɛὺς δὲ θοῶς ἐνὶ χɛιρὶ τινάξας | ἀξίνην ἐύχαλκον ἐπɛπλήγɛι μέσα νῶτα, ‘then godlike Telamon struck him in the back with a curved discus, and Peleus swiftly brandishing a well-made bronze axe in his hand struck him in the middle of the back’). Σ N.25a attributes the weapons differently (φασὶ γὰρ Πηλέα καὶ Τɛλαμῶνα ἐν γυμνασίοις ἀνɛλόντας Φῶκον, τὸν μὲν δίσκῳ τύψαντα, τὸν Πηλέα, τὸν δὲ Τɛλαμῶνα σιδήρῳ τὰ μɛτάφρɛνα, ‘they say that Peleus and Telamon killed Phocus while at exercise, Peleus striking him with a discus, and Telamon striking him in the back with an iron [sword/axe]’), which may be a consequence of an imperfect recall of the Alcmaeonis passage, or indicate use of a different source. In either case, the account of the details of the story are likely to prompt recall of or recourse to the Alcmaeonis, a text which was likely not as well known in the Hellenistic period as the Iliad and the Odyssey. As in the case of N.10, the scholia's citations allow for an increased awareness of Pindar's rhetorical strategies in comparison with other texts.

Σ N.5.25b cites Callimachus Aetia fr. 26.20–1 M (ἔκλυɛ <—>, τῶν μηδὲν ἐμοὺς δι’ ὀδόντας ὀλίσθοι, | Πηλɛύς, ‘Peleus heard … of which may nothing slip through my teeth’) as a comparandum for Pindar's elliptical reference, but they note that ‘Pindar knows about Phocus’ death, but shies away from mentioning it’ (οἶδɛν οὖν ὁ Πίνδαρος τὸν Φώκου θάνατον, ἀλλ’ ἐκτρέπɛται ɛἰπɛῖν), and that the structural parallel is inexact: μήποτɛ δὲ καὶ τὸ παρὰ Καλλιμάχῳ [fr. 26.20–1 M] οὕτως ἀποδοτέον, ὅτι αἱ γυναῖκɛς ὠνɛίδιζον αὐτῷ τὸν Φώκου θάνατον (‘Callimachus’ version should not be understood in these terms, because [there] the women berated him about Phocus’ death’). Although the lines referred to by this last phrase do not survive, it is clear that Callimachus’ account involved the women of Aegina lambasting (ὠνɛίδιζον) Peleus for his deed, and was consequently less elliptical than Pindar's treatment; whereas Pindar omits a narrative of the killing altogether, Callimachus’ text includes the women's reactions to its aftermath, whether in direct or indirect speech, to which the narrator's silence (μηδὲν ἐμοὺς δι’ ὀδόντας ὀλίσθοι) is contrasted.Footnote ⁴⁹ The fact that the scholium registers the difference between the two authors’ use of the figure encourages the reader to consider the difference more closely. Whereas in Pindar's text the ellipse is enunciated by the primary narrator, Callimachus’ treatment plays on the difference between the primary narrator, who refuses to mention the abuse directed at Peleus, and the secondary narrator, the Aeginetan women responsible for it. This distinction is obviously a construct of Callimachus’ text, and the primary narrator is being disingenuous, because he, as the figure responsible for the narrative, has mentioned the women's abuse. The comparison enabled by the citation highlights the greater complexity of Callimachus’ passage, which appropriates the moralising tone of the Pindaric narrator but also subtly undermines it by integrating it into a multi-narrator/focaliser structure. As well as bringing about an intertextual supplementation of the N.5.13–18, the scholia on the passage also bring the receptive dynamics of the Callimachus passage to readers’ attention.

Interpreting receptions

So far I have been examining citations that can be broadly understood as supplementary, their primary role being to provide information additional to that of the text in order to aid the reader's understanding. I now want to consider some cases of citations aimed at establishing a more general conceptual connection between the main text and the cited text. The first is Σ N.4.10a, which glosses a reflection on poetry's immortalising power being dependent on the Graces (ῥῆμα δ’ ἐργμάτων χρονιώτɛρον βιοτɛύɛι, | ὅ τι κɛ σὺν Χαρίτων τύχᾳ | γλῶσσα φρɛνὸς ἐξέλοι βαθɛίας, ‘a discourse lives for a longer time than deeds, which a tongue seizes from the mind's depths with the Graces’ favour’, N.4.6–8):

The citation is drawn from an account of the Graces now fragmentary, our main source for which is the set of scholia on the passage (Scholia Florentina ad Call. fr. 5–9¹⁸ M), which inform us that after giving an aetion for the Parian sacrifices to the Graces and explaining why these were conducted without music (frr. 5–7 M with scholia) Callimachus presented various accounts of the Graces’ birth; according to some, they were the daughters of Zeus and Hera, or Zeus and Eurynome, or Zeus and Euanthe.Footnote ⁵⁰ After the primary narrator has catalogued these accounts, the scholia tell us Clio informed the narrator that the Graces were actually the children of Dionysus and Coronis, a Naxian nymph. The background to the citation of fr. 9.13–14 M is therefore complex. By offering different accounts of their birth, Callimachus’ foregrounds the Graces both as divinities and as constructs of various types of poetic and scholarly discourse.Footnote ⁵¹ Moreover, there is a contrast between the invocation of N.4.6–8, which leaves the Graces uncontextualised, and Callimachus’ account of the various stories of their birth.Footnote ⁵² In the light of the differences between Callimachus’ treatment of the Graces and their invocation in N.4, the scholium alerts the reader to the Graces’ status as a textual figure open to rewriting.

The differentiation which emerges from the citation feeds into the scholium's wider function. The exegesis elaborates the conditions implied by the text for a poem being ἐργμάτων χρονιώτɛρον, providing a purely internal account of the meaning of the glossed phrase. And yet the process of commentary is itself part of the text's status as ἐργμάτων χρονιώτɛρον, participating in and affecting its dissemination. The act of commentary is symptomatic of literary value, fulfilling a need for exegesis which presupposes interest in the text concerned, but also itself implicitly validating the text it comments on by marking it as worthy of scholarly attention. Although it does not explicitly spell out such a line of thinking, the scholium's stress on writing also points the reader in this direction, a suggestion strengthened by the movement from referring to the role of ‘spoken’ and ‘written’ discourses in a poem's dissemination (τὰ λɛγόμɛνα πάντα ἢ γραφόμɛνα χρονιώτɛρά ἐστι τῶν ἔργων) to emphasising writing alone (τὰ δὲ γραφόμɛνα αἰώνια παραδίδοται). This focus on writing draws attention to the form of the poem as a material document, and the contextual shifts attendant on this form. A tension emerges between what the scholium says about the Graces and its function in supplementing the text.

The exegesis generates different ways of thinking about literary value. The scholium articulates that values as bestowed by the Graces in the process of composition (ὡς ἐάν γɛ χωρὶς Χαρίτων γράφηται ὁ ὕμνος, ἀπόλωλɛ καὶ ἔσβɛσται, ἐὰν δὲ μɛτὰ Χαρίτων, μένɛι πρὸς αἰῶνα ɛὔδια). Yet the gloss also highlights the importance of the processes of reading and scholarly supplementation in reproducing and affirming a poem's value. Both these positions are anti-essentialist, seeing the literary as regulated by extratextual forces, and there is no reason for them not to be seen as complementary. We might suggest, for instance, that the latter process partakes of and extends the former; on this reading, the scholia's supplementation is dependent on the original act of composition μɛτὰ Χαρίτων. It is notable, however, that the scholium leaves open what the Graces mean or how they operate, an openness which corresponds to the invocational structure of Pindar's text, which does not confirm its own status as σὺν Χαρίτων τύχᾳ, expressing rather a wish for their assistance (N.4.6–9, followed by τό μοι θέμɛν … ὕμνου προκώμιον ɛἴη, ‘may I set forth a prelude to a hymn’, 9–11). This refusal to appropriate the Graces or to express certainty about their function is picked up by the scholium's articulation of the positive and negative possibilities attendant on the poetic composition (ἐάν γɛ χωρὶς Χαρίτων … ἐὰν δὲ μɛτὰ Χαρίτων), as well as by the absence of a definition of the Graces. The scholium's reflections reinforce the Graces’ powerful inscrutability.

The reconfigurative potential of literary and scholarly receptions is also at issue in the scholia's citation of the final lines of Callimachus’ Hymn to Zeus (H.1.95–6). This passage is cited twice, firstly to gloss the reflections on the correct use of wealth at O.2.53–6:

The citation occurs in the following scholium (Σ O.2.96f):

Callimachus’ lines are cited again at Σ P.5.1a, glossing the opening lines of the poem:

The two glosses are evidently related, as demonstrated by the citation of Sappho common to both:

In both cases, the Callimachean citation serves to exemplify the principle that wealth and excellence need to be conjoined in order to bring prosperity; one without the other is οὐχ ἡδύ and οὐκ ἀσφαλές. It is notable, however, that the citation decontextualises the quoted verses completely, saying nothing about their wider context. I shall return to this decontextualisation below.

The topos of the importance of wealth to a man's life is very frequent in Greek literature from the archaic period onwards, but while we should not assume that Callimachus is working with Pindar's use of the topos exclusively,Footnote ⁵³ the juxtaposition of the two passages invites a reading of them in relation to each other. Callimachus notably eschews the bold metaphors of ἀστὴρ ἀρίζηλος, ἐτυμώτατον | ἀνδρὶ φέγγος and ὁ μὰν πλοῦτος ἀρɛταῖς δɛδαιδαλμένος, and we might read his use of ἄνδρας ἀέξɛιν as a gloss on the complex phrasing of O.2.53–4, the act of paring down Pindaric complexity an aesthetically significant move. The negative phrasing (οὔτ’ … οὔτ’) also differentiates it from the Pindaric model. These small-scale differences operate against a wider intertextual backdrop: while the poem's construction of comportment towards a ruler brings Pindaric filiations into play, the greater prominence of Hesiodic and hymnic intertexts, as well as metre, situate Pindar as a less prominent model. Pindar's narrative of the gods’ division of the earth at O.7.54–68 is also one of the targets of the rationalising revision at 54–63, a passage which distinguishes Callimachus’ mythopoeia from that of his models.

The cited lines occur in a passage which forges a connection between Zeus’ rule on Olympus and Ptolemy's rule of Egypt:

As scholars have noted, this passage draws on Hesiod's depiction of the just city in the Works and Days and on the connection between kings and the Muses elaborated in the Theogony.Footnote ⁵⁴ At 85–6, Ptolemy is made to exemplify the fact that Zeus does not apportion power equally, as his power exceeds that of other monarchs (πɛριπρὸ γὰρ ɛὐρὺ βέβηκɛν), a comparison which brings about an equivalence between Ptolemy and Zeus as pre-eminent powers in their respective realms.Footnote ⁵⁵ The figuration of Ptolemy as a (semi)divine being is particularly evident at 87–8, where the identity of thought and action recalls the frequent use of this topos of gods in earlier literature.Footnote ⁵⁶ Considered in the light of the Pindaric intertexts foregrounded by the above citations, the connection of Zeus and Ptolemy functions as a rupturing of previous modes of discourse, as Ptolemy's power is marked as exceeding the categories of action depicted in Pindar's texts. For all their political and cultural authority, the recipients of O.2 and P.5, Theron and Arcesilas, are not connected so overtly as Ptolemy with the divine, and this differentiation is sharpened by the Hesiodic intertexts and the generic figuration of Ptolemy as the secondary subject of a hymn. However, the intertextual overdetermination of the passage and the generic and historical transplantability of the οὔτ’ ἀρɛτῆς ἄτɛρ ὄλβος topos also work to situate Ptolemy more closely in the very tradition he exceeds, as the regrafting of these discourses entails approaching them through their generic and intertextual relations. The scholia's citations of Callimachus also have the effect, at least provisionally, of privileging Pindar as a model for Callimachus’ use of the topos, although this effect would doubtless have been mediated by readers’ sensitivity to the hymn's Homeric and Hesiodic filiations.

One consequence of this differentiation is a recalibration of the reader's understanding of these categories, and the Pindaric models of the idealised ruler. Callimachus’ stress on the unequal division of wealth among rulers (πᾶσι μέν, οὐ μάλα δ’ ἶσον, 84) and his exceeding of other rulers’ power, while doubtless primarily directed at pointing up Ptolemy's superiority to his contemporaries, also implicitly sets him above the rulers of the past. Reading Pindar alongside Callimachus entails an awareness of the historical contingency of the categories employed to understand the poetry of the past, and of the various ways in which the discourses of ‘classic’ poetry have been repositioned by their modern receptions.Footnote ⁵⁷ As in the case of the Graces at N.4.6–9 and Σ N.4.10a, the cited text's reconstruction of its subject matter sensitises the reader to shifts in the workings of the Pindaric corpus contingent on the processes of dissemination. By foregrounding the status of the Pindaric texts as literary models, the citations alert the reader to the fact that in the situations of reading and scholarly commentary elements such as addresses and advice to the victor are not, as in a performance, sets of communications between living people, however stylised, but textual figures removed from the historical circumstances in which they originally functioned.Footnote ⁵⁸ While the move of seeing Pindar's texts as traces of performances and historical situations is always open to the reader, and indeed encouraged by the conversational nature of addresses such as those of O.2 and P.5, the literary and scholarly appropriations just examined highlight Arcesilas’ and Theron's status as idealised textual figures and literary exempla as well as historical individuals. The poems’ disseminatory situation shifts the significance of these idealisations, the present tenses of which, for instance (e.g ὁ μὰν πλοῦτος … φέρɛι τῶν τɛ καὶ τῶν | καιρόν, O.2.53–4; σὲ δ’ ἐρχόμɛνον ἐν δίκᾳ πολὺς ὄλβος ἀμφινέμɛται, ‘great prosperity attends you as you move in justice’, P.5.14), can also be read as having a generalising force which elevates their subjects and makes them into transhistorical models.Footnote ⁵⁹ Callimachus’ appropriation of the trope of the ruler's use of wealth both validates and recontextualises the exemplary status of the Pindaric laudandi to which it applies.

The citation of Callimachus H.1.95–6 and the relations thus opened up with Pindar's poems draws together many of the issues of literary and scholarly reception on which I have focused, but the textual/metatextual totality constituted by the citational gloss also broaches a distinctive concern with the equation of wealth and excellence, and with the nature of political power. The Pindaric and Callimachean texts both dramatise the differences between power as actually exercised by historical individuals and its specifically textual form, and their respective textualities problematise the critical move of seeing textual figurations as mimetic of actual power relations. The slippage between subject and text, in which writing about power is itself an exercise of cultural authority, and the connections between textual figurations of power and receptive situations, make such passages charged sites of literary negotiation. In Callimachus’ Hymn to Zeus, as discussed above, the representation of power functions by staging power's ineffability. At 92–3, the narrator's confession of the difficulty of singing about Zeus’ power (τɛὰ δ’ ἔργματα τίς κɛν ἀɛίδοι; | οὐ γένɛτ’, οὐκ ἔσται· τίς κɛν Διὸς ἔργματ’ ἀɛίσɛι;) both ironically undermines his own previous treatment of Zeus’ deeds, and in pointing towards a fabula (τɛὰ δ’ ἔργματα) which exceeds representational capacities creates a vivid sense of the god's power. This undermining of representational modes picks up on the stress on the poem's fictionality and the role of the reader/audience at 64 (ψɛυδοίμην, ἀίοντος ἅ κɛν πɛπίθοιɛν ἀκουήν, ‘when I make fictions, may they be such as persuade the listener's ear’). A contrast emerges between the textual figuration of Zeus’ deeds and the deeds themselves, of which texts will only ever be, on this account, an imperfect, fictive representation. The extension of this analysis to past poets (οὐ γένɛτ’, 93) underscores the fictionality of their accounts: previous poetic treatments are figured simultaneously as privileged models and fictions to be contested.Footnote ⁶⁰

The citations at Σ O.2.96f and Σ P.5.1a can be read as indicating Callimachus’ indebtedness to and difference from Pindar, but they also serve to highlight the specifically textual nature of Pindar's subjects. It is important to recognise, however, that this process is already at work in Pindar's texts; even in the situation of performance the laudandus is not only an individual to whom the text is directed but a textual figure in the process of being constructed and idealised. This process is particularly notable in P.5, where Arcesilas is made to correspond to the supra-individual ideal articulated by means of the anonymous τις at P.5.2, and by the generalising comments about wise men at 12–13 (σοφοὶ δέ τοι κάλλιον | φέροντι καὶ τὰν θɛόσδοτον δύναμιν, ‘wise men bear better the power given by the gods’). ‘Arcesilas’ as a textual construct emerges in the movement from these generalisations to the explicit praise of his achievements at 14–23, as well as from the praise itself. A central part of his status is the way in which he corresponds to the generalised models enunciated by the narrator. His status as a figure being constructed as well as an addressee is articulated by the paraenesis at 23–5 (τῶ σɛ μὴ λαθέτω … παντὶ μὲν θɛὸν αἴτιον ὑπɛρτιθέμɛν, ‘so do not forget … to set god over all as cause’) the rhetoric of which marks its subject as (in part) a projection of the text by subordinating him to the ethical force of the injunction. This ethical force, moreover, should not be restricted in its operation to Arcesilas, as it also applies through the textualised Arcesilas to the wider community.Footnote ⁶¹ These microtextual effects are reinforced by the poem's wider intertextual filiations; the opening remarks about wealth, for instance, operate against the backdrop of the highly conflicted representation of wealth in archaic aristocratic discourses, particularly the poetry of Theognis, where it is often represented as socially destabilising.Footnote ⁶² As well as implicitly responding to such discourses by articulating an idealised convergence of wealth and excellence, Pindar's treatment also figures Arcesilas as a specifically (inter)textual figure which operates in relation to the (often anonymous) addressees of archaic sympotic discourse.

In Pindar's case, dissemination and its concomitant immortalisation of the victor further instantiates the figuration already at work in the texts, as the historical power of a living individual is exchanged for a virtualised immortality and literary exemplarity. Likewise literary appropriations: the dynamic of validation and supersession explored above in Callimachus’ use of Pindar picks up on the virtualisation of power already at work in the text. Moreover, the Callimachean citations at Σ O.2.96f and Σ P.5.1a themselves form a particular site of the operations of such intertextual and appropriative dynamics. We can now see more clearly that in these cases the Pindaric texts are, rather paradoxically, glossed with mention of a text which claims to exceed their own representations of power. This glossing functions, as mentioned above, through the decontextualising effects of the citations. Callimachus’ ironised representation of power and intertextual overdetermination is erased, and H.1.95–6 made to exemplify a transhistorical and extratextual principle (τὸν μɛτὰ ἀρɛτῆς πλοῦτόν φησιν ɛὐρυσθɛνῆ ɛἶναι. τὸ γὰρ ἕτɛρον ἄνɛυ τοῦ ἑτέρου οὐκ ἀσφαλὲς), an operation particularly ironic in the light Callimachus’ focus at H.1.79–90 on the historical and contextual specificities of political authority. The citations also involve a notable shift in the configuration of the speaking voice, as the self-ironising Callimachean narrator with his emphasis on fictionality and uncertainty, is substituted for the decontextualised, and implicitly authoritative, utterance of the commentary.