Aristarchus and the Epic Cycle

Gregory Nagy

To refer to this article, please cite it in this way:
Gregory Nagy, "Aristarchus and the Greek Epic Cycle," Classics@ Volume 6: Efimia D. Karakantza, ed. The Center for Hellenic Studies of Harvard University, edition of December 21, 2010.


This presentation centers on the concept of the epic Cycle as understood by Aristarchus of Samothrace, who was director of the Library of Alexandra in the mid third century BCE. The work of Aristarchus survives only indirectly, through the reportage of various ancient sources preserved mostly in scholia or notes found in ancient papyrus fragments and in medieval codices. I focus here on the scholia that reflect the hupomnēmata or commentaries of Aristarchus on the textual transmission of the Homeric Iliad and Odyssey. When Aristarchus evaluates Homeric textual variants in his hupomnēmata, he tries to determine, in each case, which of the variants are original to Homer and which of the variants stem from poets whom he considers to be neōteroi ‘the newer ones’. For Aristarchus, these ‘newer ones’ included the poets of the epic Cycle.

Here I begin to apply the results of an important work that focuses on the use of the term neōteroi by Aristarchus:

Severyns, A. 1928. Le cycle épique dans l’école d’Aristarque. Bibliothèque de la Faculté de Philosophie et Lettres de l’Université de Liége 40. Paris.

As Severyns demonstrates, the use of this term neōteroi in the Homeric scholia, especially in the scholia for the tenth century codex Venetus A, reflects the usage of Aristarchus himself, not only of the Aristarcheans who came after him—or of the scholiasts who report on the opinions of the Aristarcheans, as in the Venetus A. [1] In using this term neōteroi, meaning ‘newer’ or ‘neoteric’, Aristarchus had in mind poets whom he judged to be ‘newer’ than Homer. From here on, I will use the term neoteric in this sense, without prejudging whether the neoteric poets were really ‘newer’ than Homer.

The distinction made by Aristarchus between Homer and the neōteroi is in line with his basic editorial principle, Homēron ex Homērou saphēnizein ‘clarify Homer on the basis of Homer’. [2] This principle was founded on the premise that the real Homer must be systematically isolated from the supposedly non-Homeric accretions represented by the neōteroi or ‘newer’ poets—the neoterics. Aristarchus’ system, especially as reported by the Aristarchean scholar Didymus in the first century BCE, remains our main source for the working distinction between the neōteroi or ‘neoteric’ poets and the supposedly oldest poet of them all, Homer as the original poet. [3]

Severyns shows that Aristarchus relied on the poets of the Cycle as a main source for attestations of neoteric usage, and that the precision of his work on the Cycle and on other poetic sources is often blurred by the reportage of later Aristarcheans like Didymus; and that reportage is further blurred by the abbreviated summaries that we find in the Homeric scholia and in Eustathius. [4]

The neōteroi or ‘newer’ poets were imagined as authors of post-Homeric wordings—words or groups of words or whole verses—that somehow found their way into the Homeric texts. In the hupomnēmata of Aristarchus, these supposedly post-Homeric wordings were traced back to such ‘newer’ sources as Hesiod and the poets of the epic Cycle. [5]

In Homer the Classic, I argue that Orpheus too, like Hesiod and the poets of the Cycle, was evidently a ‘newer’ poet in the age of Aristarchus, though the question of Orphic accretions was evidently not a major concern for this editor. [6]

Orphic accretions were in fact a major concern, however, for a predecessor of Aristarchus, Zenodotus, who was director of the Library of Alexandria in the third century BCE. As I argue in Homer the Classic, the anti-Orphic editorial stance of Zenodotus can best be understood in the light of the generally anti-neoteric stance of Aristarchus a century later. In the case of Aristarchus, the category of poets he judged to be neōteroi ‘newer’ than Homer extended well beyond the archaic era of shadowy figures like Orpheus, all the way into the Hellenistic era. Even poets like Callimachus and his Alexandrian contemporaries were considered neōteroi, that is, ‘newer’ or ‘neoteric’, in their own right. [7] A most notable exponent of such neoteric poets in the Hellenistic era, besides Callimachus himself, was Apollonius of Rhodes, who succeeded Zenodotus of Ephesus as director of the Library of Alexandria: though Apollonius, unlike Zenodotus, did not produce an edition of Homer, he was most influential in reshaping the idea of Homer. Another notable exponent of such neoteric poets was Theocritus of Syracuse, likewise a contemporary of Callimachus and Apollonius. [8]

We have further evidence for the application of the term neōteroi in the sense of ‘neoterics’ to the scholar-poets who flourished in the age of Callimachus. [9] And the term is implicit in the reference made by Cicero (Tusculan Disputations 3.19.45) to the poetae novi ‘new poets’, whom he further describes as cantores Euphorionis ‘singers of Euphorion’; Euphorion of Chalkis, a scholar-poet who was made director of the Library of Antioch around 220 BCE, specialized in imitating Callimachus. [10]

After the age of Callimachus, in the age of Aristarchus, the Alexandrian editorial preoccupation with Orpheus and Orphic accretions had evidently waned, and Aristarchus was more preoccupied with isolating verses he judged to be Cyclic accretions in the Homeric text. For Aristarchus, it was the poets of the epic Cycle who became primary exponents of the category known to him as the neōteroi. [11] Occasionally, Aristarchus even criticized his predecessor Zenodotus for not being vigilant enough in isolating the neoterisms of the epic Cycle. [12] And just as Zenodotus seemed to be less preoccupied with the alleged neoterisms of the Cycle, so also the contemporary neoteric poets—among them, Callimachus—were less preoccupied with cultivating in their own poetry those aspects of Homeric poetry that seemed to be Cyclic. A salient example is the celebrated wording of Callimachus: ἐχθαίρω τὸ ποίημα τὸ κυκλικόν ‘I detest the Cyclic poem’ (Epigram 28.1). [13]

In Homer the Classic, I went out of my way to stress that the ‘Homeric’ elements of Homeric poetry do not necessarily represent the oldest layers of this poetry. Whatever elements in Homeric poetry seem to be Orphic and Cyclic may turn out to be genuine traces of still older layers of that poetry. Of these two older layers, the Cyclic elements were differentiated from ‘Homeric’ poetry more systematically than the Orphic elements. Whatever Cyclic elements Aristarchus may have found embedded in the Homeric textual tradition would have been more easy for him to isolate, whereas most of the Orphic elements had already been isolated by Zenodotus. For Aristarchus, the methods used by Zenodotus to isolate Orphic elements in the Homeric text must have seemed too radical, whereas the same editor’s methods in isolating Cyclic elements seemed too superficial. [14]

In short, Zenodotus and Aristarchus were dissimilar in their emphasis on what to isolate as neoteric. For the earlier editor, it was mainly the Orphic aspects of Homer that were targeted. For the later editor, it was mainly the Cyclic aspects. [15]

Here I need to stress once again a fundamental difference between Zenodotus and the neoteric poets of the Hellenistic era who imitated the neoterisms isolated by Zenodotus. Although these poets may have followed Zenodotus in treating accretive elements in Homer as examples of non-Homeric poetry, they were antithetical to Zenodotus in their cultivation of this poetry for its own sake. For them this kind of poetry was worth cultivating precisely because it was judged to be non-Homeric. For the likes of Callimachus and Apollonius, Homer was not the only Classic as an exponent of epic. The ‘newer’ poets—as represented by Cyclic and Hesiodic and Orphic poetry—were also Classics in their own right. [16]

Despite the anti-neoteric editorial stance of Zenodotus, the actual text of his Homeric edition must have seemed neoteric, retrospectively, by comparison with the corresponding text of Aristarchus. Despite the marginal marks and annotations devised by Zenodotus for the purpose of prescribing the rejection of neoteric elements—whether by athetesis or by outright deletion—his text of Homer included all these elements. In its inclusiveness, the Homeric text of Zenodotus could serve as an all-encompassing bible, as it were, for neoterics and anti-neoterics alike. [17]

By contrast, the Homeric edition of Aristarchus shaded over the neoteric elements, thereby highlighting what he considered to be the core text of Homer. These neoteric elements, like other variant readings, could not be studied in the scrolls containing the Homeric base text as presented by Aristarchus; instead, such study was relegated to his hupomnēmata ‘commentaries’, written in separate scrolls. [18]

The base text of Homer as edited by Aristarchus was adorned with critical signs that cross-referred to the hupomnēmata of the editor. It was in these hupomnēmata that the reader could find out about textual variants in general. Among these variants, the neoteric elements figured most prominently. In these hupomnēmata, Aristarchus offered a critical analysis of what he judged to be the rightness or wrongness of the core version as juxtaposed with all available textual variants found in other versions—and as juxtaposed also with any conjectures made by critics, including himself. Unfortunately, the corpus of variant readings assembled by Aristarchus has not survived as a corpus, since the Homeric hupomnēmata of Aristarchus have not been preserved in their textual integrity. [19] What survives from these commentaries is an unwieldy mass of excerpts and reports made by Aristonicus, Didymus, and other Aristarchean scholars, whose own works in turn survive only in the form of sporadic excerpts and reports made by scholiasts in the Homeric scholia. [20] The unfortunate outcome is that Aristarchus’ collection of neoteric elements, a most important component of the variant readings reported by him, has been for the most part irretrievably lost. [21] The Aristarchean editorial policy of shading over the supposedly peripheral neoteric elements in order to highlight the core text of Homer resulted ultimately—and unintentionally—in the relegation of this periphery into a permanent outer darkness. Against this background of darkness, the non-neoteric core text of Homer continued to shine forth to the public at large. [22]

This Homeric core text, surrounded by a periphery of neoteric elements that Aristarchus relegated to his hupomnēmata ‘commentaries’, was the Koine. The Homeric Koine, as approximated by the base text of Aristarchus in the second century BCE, stands in stark contrast with what I call the Homerus Auctus, as approximated by the base text of Zenodotus in the age of Callimachus in the third century BCE. By contrast with the base text of Aristarchus, the base text of Zenodotus was more inclusive—and encumbered—even though the editorial criteria of Zenodotus himself were in some ways more exclusive than those of Aristarchus. When it came to discriminating between ostensibly Homeric and non-Homeric verses, Zenodotus was more exclusive in that he tended to athetize more verses than did Aristarchus, at least when it came to verses judged to be Orphic accretions. An extreme example is the fact that Zenodotus athetized the entire narrative of the Shield of Achilles in Iliad XVIII. Conversely, Aristarchus tended to athetize more verses than did Zenodotus when it came to passages judged to be Cyclic accretions. Overall, the number of verses athetized by Aristarchus (but retained in his base text) on the grounds that they are Cyclic accretions tends to be exceeded, I argue, by the number of verses athetized by Zenodotus on the grounds that they are Orphic accretions. [23] In any case, both editors retained in their base texts the verses they athetized.

The notionally genuine Homer, as ‘corrected’ by the system of athetesis developed by Zenodotus, may have been more exclusive than the notionally genuine Homer of Aristarchus, but the point remains that the base text of Zenodotus was more inclusive, since it included not only the athetized verses marked with an obelos but also the plus verses marked with deletion signs and featuring brief comments written into the margins. In other words, the base text of Zenodotus must have encompassed the same kinds of augmentation encompassed by the base text of Crates, which was the equivalent of what I call the Homerus Auctus. [24] By contrast, the base text of Aristarchus retained only the verses he athetized, not the plus verses, and this base text was the equivalent of the Homeric Koine.

I emphasize that the Homeric Koine tends to include and accommodate variants that Aristarchus ascribes to the neōteroi. And, although Aristarchus isolates such variants as non-Homeric and even post-Homeric accretions—he continues to include and accommodate them in his base text, which is meant to correspond as closely as possible to the Koine as he understood it.

We can find specific cases where the Homeric base text of Aristarchus contains verses that he ascribes to the neoteric poets of the Cycle, judging these supposedly neoteric verses to be non-Homeric and even post-Homeric accretions in an evolving Homeric text. In such cases, Aristarchus follows the procedure of recording his judgment by athetizing the given verses in his base text and then by commenting on his judgment in his hupomnēmata.

One such case is Iliad XXIV 601–620:

601                           νῦν δὲ μνησώμεθα δόρπου.
602   καὶ γάρ τ’ ἠΰκομος Νιόβη ἐμνήσατο σίτου,
603   τῇ περ δώδεκα παῖδες ἐνὶ μεγάροισιν ὄλοντο
604   ἓξ μὲν θυγατέρες, ἓξ δ’ υἱέες ἡβώοντες.
605   τοὺς μὲν Ἀπόλλων πέφνεν ἀπ’ ἀργυρέοιο βιοῖο
606   χωόμενος Νιόβῃ, τὰς δ’ Ἄρτεμις ἰοχέαιρα,
607   οὕνεκ’ ἄρα Λητοῖ ἰσάσκετο καλλιπαρῄῳ·
608   φῆ δοιὼ τεκέειν, ἣ δ’ αὐτὴ γείνατο πολλούς·
609   τὼ δ’ ἄρα καὶ δοιώ περ ἐόντ’ ἀπὸ πάντας ὄλεσσαν.
610    οἳ μὲν ἄρ’ ἐννῆμαρ κέατ’ ἐν φόνῳ, οὐδέ τις ἦεν
611   κατθάψαι, λαοὺς δὲ λίθους ποίησε Κρονίων·
612   τοὺς δ’ ἄρα τῇ δεκάτῃ θάψαν θεοὶ Οὐρανίωνες.
613   ἣ δ’ ἄρα σίτου μνήσατ’, ἐπεὶ κάμε δάκρυ χέουσα.
614   νῦν δέ που ἐν πέτρῃσιν ἐν οὔρεσιν οἰοπόλοισιν
615   ἐν Σιπύλῳ, ὅθι φασὶ θεάων ἔμμεναι εὐνὰς
616   νυμφάων, αἵ τ’ ἀμφ’ Ἀχελώϊον ἐρρώσαντο,
617   ἔνθα λίθος περ ἐοῦσα θεῶν ἐκ κήδεα πέσσει.
618   ἀλλ’ ἄγε δὴ καὶ νῶϊ μεδώμεθα δῖε γεραιὲ
619   σίτου· ἔπειτά κεν αὖτε φίλον παῖδα κλαίοισθα
620   Ἴλιον εἰσαγαγών· πολυδάκρυτος δέ τοι ἔσται.
601   But now the two of us [= Achilles speaking to Priam] must think of eating.
602   Even Niobe, the one with the beautiful hair, thought of eating grain,
603   the one who had twelve children, and all of them were killed in the palace,
604   six daughters and six sons in the bloom of youth.
605   Apollo killed the sons, shooting from his silver bow.
606   He was angry at Niobe—and the daughters were killed by Artemis, shooter of arrows—
607   angry because she [= Niobe] tried to make herself equal to Leto, the one with the beautiful cheeks.
608   She [= Niobe] said that she [= Leto] gave birth to two, while she herself produced many.
609   So the two of them [= Apollo and Artemis], only two though they were, destroyed the many.
610   They [= the children of Niobe] lay there in their gore for nine days, and there was no one
611   to bury them. The people had been turned into stone by the son of Kronos.
612   Then on the tenth day they [= the children of Niobe] were given a burial by the sky-dwelling gods themselves.
613   And she [= Niobe] thought of eating, since she was exhausted by her shedding of tears,
614   and now, somewhere amidst the rocks, on the desolate heights,
615   in Sipylos, where they say the goddesses have places to sleep
616   —the goddess nymphs, the ones who dance on the banks of the Akhelōios—
617   there does she [= Niobe], though she has been turned into stone, digest her sorrows inflicted by the gods.
618   So too now the two of us must think, radiant old man,
619   of eating grain. And then, after that, for your dear child you may weep again,
620   after you have brought him to Troy. And there will be many tears shed for him.

In the logic of this complex simile, Niobe weeps, then consumes grain, and then resumes her weeping as she continues to ‘digest’ her sorrows for all eternity: so also Priam weeps, then is invited to eat grain, and then he too will resume his own weeping. The point is, ‘even’ Niobe ate grain, though the sorrows she had to ‘digest’ were eternal. Her sorrows, in the rhetoric of the simile spoken by Achilles, are overwhelmingly greater than the sorrows of Priam over the death of Hector—or than the sorrows of Achilles over the death of Patroklos. The sorrows of Niobe are in fact so overwhelming that she continues to weep eternally even after the gods turn her into stone. A petrified figure should be drained of emotion, as we see from the logic of the narrative contained in the simile: when the population in the realm of Niobe is petrified, there can be no weeping, no mourning, and therefore no funeral, so that the gods themselves must conduct a funeral and bury the children of Niobe. But Niobe, even after she is petrified, is like a human figure in that she continues to dissolve into tears. So overwhelming are her sorrows. [25]

In this Koine version as it comes down to us by way of the medieval manuscript tradition, we see at verse 611 that all the people in the realm of Niobe were turned into stone after her sons and daughters were destroyed by Apollo and Artemis, so that they could not bury Niobe’s children; then, at verses 614–617, we see further that Niobe herself was turned into stone. Commenting on all five of these verses in his hupomnēmata, Aristarchus argues that verse 611 is genuinely Homeric while verses 614–617 are not, and he athetizes all four of these supposedly non-Homeric verses (scholia A for Iliad XXIV 614–617a 1). It would be a mistake to infer, however, that verses 614–617 had no prehistory of their own. Aristarchus reports that there was indeed a version of the story where Niobe herself became petrified, and he ascribes this version to poets later than Homer, referring to them as the neōteroi ‘newer ones’; by contrast, according to Aristarchus, the supposedly older poet Homer did not have such a version (scholia A for Iliad XXIV 613a 1).

As I show in Homer the Classic, there is another case where Aristarchus athetizes a verse that tells of petrifaction:

λᾶαν γάρ μιν ἔθηκε Κρόνου πάϊς ἀγκυλομήτεω·
For the son of Kronos, the one with the oblique plans, now made it into stone.

Iliad II 319

What has been petrified here is a serpent that has just devoured nine birds, according to an embedded narration by Odysseus (II 299–332). Aristonicus, an Aristarchean scholar who flourished in first century BCE, has this to say about the verse, in the scholia A for Iliad II 319 (a1): ἀθετεῖται ‘it is athetized’—that is, this verse is athetized by Aristarchus. Once again, such an athetesis by Aristarchus should not be misunderstood to mean that verse 319 of Iliad II is a substandard verse. It is just the opposite: this verse, like the many hundreds of other verses athetized by Aristarchus, belongs to the standard version of Homer, the Koine. Moreover, Aristarchus featured this verse in the base text of his own edition of Homer. This verse would have looked like any other verse in the edition of Aristarchus—except for one small detail. Aristarchus placed a mark in the left-hand margin next to all verses that he athetized. That sign is the obelos (–). [26]

Here I turn to a formulation I offered in Homer the Classic concerning the editorial policy of Aristarchus. In cases where this editor was uncertain whether a given Homeric verse was genuinely Homeric, he had two choices, in terms of his own editorial system:

  1. If he found the given verse only weakly attested in the available manuscripts, he would omit the verse from his base text.
  2. If he found the given verse strongly attested, then he would keep the verse in his base text, marking that verse with an obelos (–), the sign of athetesis, in the left-hand margin.

As I have been arguing, the base text of Homer as established by Aristarchus was designed to reflect as accurately as possible the standard version of Homer, the Koine. Aristarchus confined to his hupomnēmata ‘commentaries’ whatever information and opinions he had about non-Koine variant readings, which in many cases he thought were more likely to be Homeric than the readings he featured in his own base text. [27]

Applying what I formulated in Homer the Classic about the editorial policies of Aristarchus, I reaffirm what I just said about the verse at Iliad II 319 as I quoted it. To put it in the most simple terms, this verse stems from the base text of Aristarchus. And, like all other verses in that base text, verse 319 of Iliad II stems from the Koine tradition.

The Homeric scholia report another relevant detail about verse 319 of Iliad II. So far, we have seen that this verse must have been strongly attested in the standard Homer texts available to Aristarchus. But now we will see that it was also attested in a Homer text that particularly interested Aristarchus—the Homer edition of Zenodotus, who predates Aristarchus by well over a century. According to Zenodotus (scholia A for Iliad II 318), the verse that we know as II 319 was connected in meaning to an epithet in the previous verse 318. That epithet is aridēlos (ἀρίδηλον) ‘most visible’, with reference to the petrified serpent. To quote the wording of Aristonicus (scholia A for II 318 1): ὅτι Ζηνόδοτος γράφει ἀρίδηλον καὶ τὸν ἐχόμενον προσέθηκεν ‘[Aristarchus disagrees with the reading of Zenodotus] because Zenodotus writes ἀρίδηλον in his text, and he [= Zenodotus] added the line that follows’. [28]

I must take a moment to offer some words of caution here about the wording of Aristonicus, which can be misleading. When Aristonicus says that Zenodotus had ‘added’ the next verse, that is, verse 319 as we know it, he is speaking retrospectively: by hindsight, Aristarcheans like Aristonicus in the first century BCE make it seem as if Zenodotus had ‘added’ such verses—as if the verses athetized by Aristarchus had come from outside the manuscript tradition of Homer. From the standpoint of Aristarchus himself, however, the verses that he athetized had not been added by previous editors like Zenodotus. [29] Rather, as I showed in Homer the Classic, Aristarchus was simply expressing his own editorial opinion that such verses should now be subtracted from the corpus of verses supposedly composed by Homer himself. Moreover, as I have already argued here, such editorial opinions of Aristarchus were confined to his hupomnēmata ‘commentaries’. Aristarchus did not actually subtract from his base text the verses that he athetized. To repeat, the base text of Aristarchus continued to reflect the standard manuscript tradition of Homer.

Returning to verse 318 of Iliad II, I repeat the testimony of Aristonicus concerning what Aristarchus found in the edition of Zenodotus: in that Homeric text, the epithet of the serpent was spelled ἀρίδηλον—which would have been pronounced ἀρίδδηλον. [30] In the standard Homeric texts consulted by Aristarchus, the epithet was spelled ἀρίζηλον, and it was evidently this spelling that Aristarchus featured in the base text of his own edition. These two phonological byforms aridēlos and arizēlos are parallel in morphology, but the meaning is transparent only in the phonological byform aridēlos—‘most visible’. [31]

It is essential to keep in mind that such information about the Homeric textual variation ἀρίζηλον / ἀρίδηλον, as mediated by the Homeric scholia, is derived ultimately from the hupomnēmata of Aristarchus. In these hupomnēmata, as we see further from the abridged reportage of the Homeric scholia, Aristarchus linked his discussion of the variants ἀρίζηλον and ἀρίδηλον at verse 318 with yet another variant, ἀίζηλον. Whereas arizēlos and aridēlos mean ‘most visible’, aïzēlos means the opposite, ‘invisible’. In this case, unfortunately, the abridgment in the scholia is so severe that the actual mention of the variant ἀίδηλον has dropped out, though the basic argument adduced by Aristarchus in favor of this reading has been preserved. I will confront the argument in a moment, but first I need to stress the morphological validity of the variant aïzēlos, which is a phonological byform of aïdēlos just as arizēlos ‘most visible’ is a phonological byform of aridēlos ‘most visible’ at verse 318: the form aïzēlos / aïdēlos must have the basic meaning ‘invisible’ (derived from earlier *awidēlos). [32]

Next I turn to the argumentation of Aristarchus. The basic meaning of aïdēlos / aïzēlos as ‘invisible’ helps explain why Aristarchus was interested in the variant reading ἀίζηλον ‘invisible’ as an alternative to ἀρίζηλον ‘most visible’ at verse 318 of Iliad II. Even in their abridged form, the Homeric scholia show clearly what Aristarchus said in his hupomnēmata about the meaning of this variant. I give here the wording of scholia A: λέγει μέντοι γε ὅτι ὁ φήνας αὐτὸν θεὸς καὶ ἄδηλον ἐποίησεν ‘he [= Homer] says that the same god that had made it [= the serpent] visible [= phainein] also made it invisible [adēlos]’ (scholia A for Iliad II 318 1). The point is reinforced by the wording of scholia T for verse 319: ἀθετεῖται· πιθανώτερον γὰρ αὐτὸν καθάπαξ πεποιηκέναι ἀφανῆ τὸν καὶ φήναντα θεόν ‘[This verse] is athetized [by Aristarchus]: for it is more plausible that the same god who made it [= the serpent] visible [= phainein] should straightaway make it disappear [= make it aphanēs ‘invisible’].

So the variant verse at Iliad II 318 would look like this:

τὸν μὲν ἀίζηλον θῆκεν θεὸς ὅς περ ἔφηνε
And the god that had made it [= the serpent] visible [phainein] now made it invisible [aïzēlos].

Iliad II 318 (variant reading attested by Aristarchus)

This version of Iliad II 318 is evidently incompatible with II 319 as we have it, which tells about the petrifaction of the serpent. The theme of making the serpent disappear is evidently incompatible with the theme of making the serpent into stone:

τὸν μὲν ἀρίζηλον θῆκεν θεὸς ὅς περ ἔφηνε·
λᾶαν γάρ μιν ἔθηκε Κρόνου πάϊς ἀγκυλομήτεω·
The same god that had made it visible [phainein] now made it most visible [arizēlos].
For the son of Kronos, the one with the oblique plans, now made it into stone.

Iliad II 318–319

Some modern commentators have attempted to discredit the standard version I just quoted. One of these commentators claims that the contrast being made at verse 318 in this version is “rather pointless,” and he expresses his preference for the alternative version adduced by Aristarchus. [33] This commentator implies that the standard version conveys a redundancy, which we may paraphrase this way: “the snake was made visible by the god who had made it visible.” But such a paraphrase blunts the point of the intensifying prefix ari in the compound arizēlos ‘most visible’. I translate ari as ‘most’ rather than ‘very’ in order to convey a rhetoric of extreme intensification here, which serves to express a competitively outstanding quality. Hence my translation: ‘The same god that had made it visible [phainein] now made it most visible [arizēlos]’. I maintain, then, that the standard version of verses 318–319 of Iliad II makes just as much sense as the non-standard version of verse 318 minus verse 319.

Although Aristarchus personally preferred a configuration of textual variants that expresses the idea that Zeus made the serpent appear and then disappear, it is clear that he recognized the reality of the alternative configuration of variants expressing the idea that Zeus made the serpent appear and then made that appearance permanent by turning it into a physical landmark. It is also clear that he recognized that this alternative configuration of variants was considered the standard Homeric version even in his own time. The clearest indication of this recognition, as we saw earlier, is the fact that he included verse 319 of Iliad II in his base text instead of omitting it. In other words, the verse that signals the non-disappearance of the serpent occupies a place in the base text of Aristarchus, even though he athetizes it. Similarly at verse 318, Aristarchus featured in his base text the variant arizēlos ‘most visible’ as the epithet of the serpent, again indicating its non-disappearance, relegating to the hupomnēmata his comments on the variant aïzēlos ‘invisible’—which he linked in his commentary with his proposal to athetize verse 319. As we noted earlier, the linkage is self-evident: if the serpent disappears after its epiphany, then the subsequent verse describing the petrifaction of the serpent is out of place. But this non-Koine version, which features a scene of epiphany followed by disappearance, simply could not be formatted as the true Homeric alternative to the Koine version, which features a scene of epiphany followed by petrifaction. I must stress again: even though Aristarchus preferred the non-Koine version, he kept the Koine version in the base text of his edition.

As Montanari has shown, the reasoning of Aristarchus in this particular case was influenced by the earlier reasoning of Aristotle about these same verses 318–319 of Iliad II (Aristotle F 145 ed. Rose via the commentary of Porphyry on the Iliad (vol. I, pp. 32–33 ed. Schrader). [34] So Aristotle, like Aristarchus, was interested in isolating what he too considered to be Cyclic elements embedded in the Homeric Koine.

I conclude that the reasoning of Aristarchus about the epic Cycle in general can be traced back to the reasoning of Aristotle himself, for whom the Cycle was categorically non-Homeric. In his Poetics, Aristotle mentions two of the Cyclic epics he knew—the Cypria and the Little Iliad—and he makes clear his view that the authors of these epics were poets other than Homer; more than that, he chooses not even to name these poets (1459a37–b16). It is only in other sources, especially in the reportage of Proclus, that we find specific names and proveniences: for example, the author of the Cypria was supposedly Stasinus of Cyprus; of the Little Iliad, Lesches of Lesbos; of the Aithiopis and the Iliou Persis, Arctinus of Miletus. [35]

Aristotle viewed Homer as the author of only two epics, the Iliad and the Odyssey (again, Poetics 1459a37–b16; cf. 1448b38–1449a1). Plato, as we see in such works as the Ion, evidently held the same view. In general, the verses that Plato quotes from ‘Homer’ are taken from the Iliad and the Odyssey, not from the epic Cycle. [36] (The one exception is not an exception, because the source is specified as not Homer.)

This way of thinking is relevant to what I argue in Homer the Classic: that the Iliad and the Odyssey were the only two epics being performed at the Panathenaia in the age of Plato and Aristotle. [37] As I also argue there, the same situation holds in the classical period of the fifth century BCE: in that earlier age as well, the Iliad and the Odyssey were the only two epics being performed at the Panathenaia. As we keep moving further back in time to even earlier ages, however, we need to re-examine the idea of Homer as the author of only two epics, the Iliad and the Odyssey. There was a time when Homer could be viewed as the notional author of all epic, as represented by the concept of the epic Cycle before it became historically differentiated from the Iliad and Odyssey.

Reconstructing backwards, back to the earliest recoverable phases, I find that the idea of the epic Cycle was an idea of epic as a comprehensive totality: the term ‘Cycle’ or kuklos was sustained by metaphors of artistic completeness. [38] A related idea is conveyed in the meaning of the name of Homer, Homēros, who is mythologized as a prototypical author. [39] The farther back we go in time, the greater the repertoire attributed to this author, including all the so-called Cycle. [40] In fact, the very notion of Cycle had once served as a metaphor for all of Homer’s poetry. [41] In an earlier work, I offered this reconstruction: [42]

I propose that the metaphor of kuklos as the sum total of Homeric poetry goes back to the meaning of kuklos as ‘chariot-wheel’ (Iliad XXIII 340, plural kukla at V 722). The metaphor of comparing a well-composed song to a well-crafted chariot-wheel is explicitly articulated in the poetic traditions of Indo-European languages (as in Rig-Veda 1.130.6); more generally in the Greek poetic traditions, there is a metaphor comparing the craft of the master carpenter or ‘joiner’—the tektōn—to the art of the poet (as in Pindar Pythian 3.112–114). [43] Further, the root ar of arariskein ‘join, fit together’ (the verb refers to the activity of the carpenter in the expression ἤραρε τέκτων ‘the joiner [tektōn] joined together [ar]’ in Iliad IV 110, XXIII 712) is shared by the word that means ‘chariot-wheel’ in the Linear B texts, harmo (Knossos tablets Sg 1811, So 0437, etc.). Most important of all for my argument, the same root ar is evidently shared by the name of Homer, Homēros, the etymology of which can be explained as ‘he who joins together’ (homo plus ar). [44] Thus the making of the kuklos by the master poet Homer appears to be a global metaphor that pictures the crafting of the ultimate chariot-wheel by the ultimate carpenter or, better, ‘joiner’. This traditional pattern of thinking matches the classification of both the aoidos ‘singer’ and the tektōn ‘carpenter, joiner’ under the category of dēmiourgós or ‘itinerant artisan’ in Odyssey xvii (381–385).

In terms of this reconstruction, then, Homer is simply the poet of the epic Cycle. In later times, however, as best exemplified by the age of Plato and Aristotle, Cyclic poetry had became clearly differentiated from Homeric poetry, and the epic Cycle no longer represented any kind of totality. Newer ideas of completeness had replaced the older idea. [45]

These newer ideas were determined by the artistic measure of tragedy. In the days of Plato and Aristotle, epic totality was represented only by the Homeric Iliad and Odyssey, and their completeness was measured according to the standards of tragedy. Aristotle says explicitly that only the Homeric Iliad and Odyssey are comparable to tragedy because only these epics show a complete and unified structure, unlike the epics of the Cycle (Poetics 1459a37–b16). In the works of Plato as well, Homer is measured against the standards of tragedy, and Homer is imagined as a proto-tragedian in his own right (Theaetetus 152e; Republic 10.595c, 598d, 605c, 607a). For Plato and Aristotle, the Homeric Iliad and Odyssey measured up to the standards of tragedy, whereas the epics of the Cycle did not. [46]

This pattern of associating tragedy with epic, and epic with tragedy, reflects an institutional reality. The genre of epic, as performed at the festival of the Panathenaia, actually shaped and was shaped by the genre of tragedy as performed at the festival of the City Dionysia. In Athens, ever since the sixth century BCE, these two genres were “complementary forms, evolving together and thereby undergoing a process of mutual assimilation in the course of their institutional coexistence.” [47]

By the time of Plato and Aristotle, such a complementarity of epic and tragedy involved only the epics of the Homeric Iliad and Odyssey, no longer the epics of the Cycle. [48] This differentiation of the epic Cycle from the Homeric Iliad and Odyssey, as I show in Homer the Preclassic, can be linked with the obsolescence of performing the poetry of the epic Cycle at the Panathenaia after the age of the Peisistratidai. [49]

I argued a moment ago that the farther we go backward in time, the wider the repertoire of Homer becomes, including all the epic Cycle. Now I argue that the farther we go forward in time, by contrast, the less there is that Homer did himself. Not only is his repertoire becoming restricted to the Iliad and Odyssey: there are many parts even of these epics that now become suspect: for example, Homer surely could not have composed the Shield of Achilles in Iliad XVIII (483–608), in the opinion of Zenodotus (scholia A for Iliad XVIII 483a). And the original Homer of this more critical and suspicious age becomes all the more specific and even brittle in identity, reflecting ever more the critics’ personal interpretations of his archetypal creation, his text. [50]

According to the interpretation of Aristarchus himself, Homer was an Athenian who lived around 1000 BCE, in the time of the so-called Ionian Migration from Athens (Proclus F a 58–62 ed. Severyns; cf. Life of Homer p. 244.13, p. 247.8 ed. Allen; cf. scholia A for Iliad XIII 197); [51] moreover, the scholiastic tradition stemming ultimately from Aristarchus implies that Homer wrote his poems (scholia A for Iliad XVII 719) and that Hesiod actually had a chance to read them (scholia A for Iliad XII 22a). [52]

Even though Aristarchus, following the thought-patterns of myth, posited a Homeric original, he nevertheless accepted and in fact respected the reality of textual variants. He respected variants because, in terms of his own working theory, it seems that any one of them could have been the very one that Homer wrote (and Hesiod read). That is why he makes the effort of knowing the many different readings of so many manuscripts.

Even though Aristarchus did not and in fact could not reconstruct the supposedly original text of Homer dating back to around 1000 BCE, he made an effort by collecting and the evaluating all the variants that he could find. In that light, I once considered whether the absence of such an original text was really a loss for Aristarchus—and for us:

What, then, would Aristarchus have lost, and what would we stand to lose, if it really is true that the variants of Homeric textual tradition reflect for the most part the multiforms of a performance tradition? If you accept the reality of multiforms, you forfeit the elusive certainty of finding the original composition of Homer but you gain, and I think this is an important gain, another certainty, an unexpected one but one that may turn out to be much more valuable: you recover a significant portion of the Homeric repertoire. In addition, you recover a sense of the diachrony. [53]

Here I confront the fact that some Homer experts who accept Lord’s formulation of “oral” poetry seem ready at times to discount the value of Aristarchus’ editorial repertoire of variants, which go far beyond the Koine or “Vulgate” texts of Homer. There is an irony here. It would be more understandable for proponents of a “writing Homer” to reject variant X or Y, accepting Aristarchus’ implicit premise that only one variant can be right and that Homer could not have written X or Y for such-and-such reasons. [54] It is unnecessary, however, for proponents of an “oral Homer” to insist on one and only one right version, unless they are also willing to believe that the oral tradition ground to a dead halt sometime around the second half of the eighth century BCE, after the text was supposedly dictated. [55] In earlier work, presenting arguments that challenge the idea of an early dictation, I substituted an “evolutionary model” to account for the process of Homeric text-fixation. [56]

For now I consider only those aspects of my evolutionary model that concern the epic Cycle. I start with a concept, as developed in Homer the Preclassic, of a Panathenaic Regulation, which dates back to the Athenian festival of the Panathenaia in the later years of the Peisistratidai. Ultimately, this Regulation led to the restricting of the epic repertoire of the Panathenaia to the Homeric Iliad and Odyssey, performed by rhapsodes who took turns in narrating the entire sequence of these two epics. [57] This is not to say, however, that the epic repertoire of rhapsodes performing in relay at the Panathenaia was restricted to the Homeric Iliad and Odyssey already in the era of the Peisistratidai. Such a restriction, I argue, became a reality only later. Nor is it to say that the principle of rhapsodic relay that we see at work in the Panathenaic Regulation originated at the Panathenaia. As the work of Douglas Frame has shown, this principle, as represented by the Homēridai of Chios, was already operational in the late eighth and early seventh centuries BCE at the festival of the Panionia held at the Panionion of the Ionian Dodecapolis in Asia Minor. [58] All I am saying for the moment is that the Panathenaic Homer started taking shape in the later years of the Peisistratidai, with the introduction of the Panathenaic Regulation by way of the Homēridai.

With the introduction of the Panathenaic Regulation by way of the Homēridai, the epic repertoire at the Panathenaia was thereafter attributed exclusively to Homer. In the earlier years of the Peisistratidai, by contrast, the epic repertoire at the Panathenaia was not only Homeric: it still included traditions we can describe retrospectively as Cyclic, Hesiodic, and Orphic. [59]

I consider here only the Cyclic tradition. For a lengthy period of time in the evolution of the Panathenaia, the epic Cycle was not distinguished from the Homeric tradition of epic performance. During this time, the epics of the Cycle were not anti-Homeric or even non-Homeric: they were simply Homeric. Homer was considered to be the poet of an epic Cycle that included what we know as the Iliad and Odyssey. Only gradually did the Homeric Iliad and Odyssey become differentiated from the epic Cycle. In the course of this differentiation, the Iliad and Odyssey became the only epics that were truly Homeric, while the Cycle became non-Homeric. [60] The epics of the Cycle were then reassigned to poets other than Homer. For example, the Aithiopis and the Iliou Persis were reassigned to Arctinus of Miletus. Similarly, the Little Iliad was reassigned to Lesches of Lesbos. In the earlier years, by contrast, the entire epic Cycle had been assigned to Homer. [61]

Aristarchus, following Aristotle, viewed the epic Cycle as post-Homeric. This view, as I have argued, is Athenocentric, based on traditions of epic performance that stem from the Panathenaic Regulation as adopted at the festival of the Panathenaia in the later years of the Peisistratidai, toward the end of the sixth century BCE. Before that time, the epic Cycle could be viewed as Homeric, not post-Homeric, even in Athens. In the song culture of the Ionian world in general, the epic Cycle was more a concept than a fixed body of epics—a concept of epic poetry that was joined togeth er by the master joiner, Homer himself. Such a concept is older than the Panathenaic concept of Homer as the master poet of the Iliad and Odyssey—and of no other epic poetry.

It could even be said that the epic Cycle is pre-Homeric, at least as a concept. I say this because the concept of Homer as the poet of the Cycle predates not only the Panathenaic Regulation that evolved in the late sixth century BCE in the context of Homeric performances at the festival of the Panathenaia. It predates also something earlier—something that might be called the Panionian Regulation, which evolved in the late eighth and early seventh centuries BCE in the context of performances by the Homēridai at the festival of the Panionia held at the Panionion of the Ionian Dodecapolis in Asia Minor. As Douglas Frame has shown, such a regulation led to a Panionian concept of Homer as the master poet of the Iliad and Odyssey—and of no other epic poetry. [62] Such a Panionian concept, predating the corresponding Panathenaic concept, would in turn predate the more general concept of Homer as the poet of the Cycle.


BA = Nagy 1979.

Chantraine, P. 2009. Dictionnaire étymologique de la langue grecque: histoire des mots. Ed. J. Taillardat, O. Masson, and J.-L. Perpillou, with a supplement Chroniques d’étymologie grecque (ed. A. Blanc, Ch. de Lamberterie, and J.-L. Perpillou) 1–10. 2nd rev. ed. Paris.

Davison, J. A. 1955. “Peisistratus and Homer.” Transactions of the American Philological Association 86:1–21.

DELG = Chantraine 2009.

Frame, D. 2009. Hippota Nestor. Washington, DC.

HC = Nagy 2009.

HPC = Nagy 2010.

HQ = Nagy 1996b.

HTL = Nagy 2004.

Janko, R. 1992. The Iliad: A Commentary. Vol. IV, Books 13–16. General ed. G. S. Kirk. Cambridge.

Kirk, G. S. 1985. The Iliad: A Commentary. Vol. I, Books 1–4. Cambridge.

Montanari, F. 2008. “Aristotele, Zenodoto, Aristarco e il serpente pietrificato di Iliade II 319.” Studi offerti ad Alessandro Perutelli II. Ed. P. Arduini, S. Audano, A. Borghini, et al. 237–243. Rome.

Nagy, G. 1979. The Best of the Achaeans: Concepts of the Hero in Archaic Greek Poetry. 2nd ed. with new introduction published 1999. Baltimore.

———. 1990a. Pindar’s Homer: The Lyric Possession of an Epic Past. Baltimore.

———. 1990b. Greek Mythology and Poetics. Ithaca, NY.

———. 1996a. Poetry as Performance: Homer and Beyond. Cambridge.

———. 1996b. Homeric Questions. Austin, TX.

———. 2004. Homer’s Text and Language. Urbana, IL.

———. 2009. Homer the Classic. Washington, DC.

———. 2010. Homer the Preclassic. Berkeley, CA.

Pfeiffer, R. 1968. History of Classical Scholarship from the Beginnings to the End of the Hellenistic Age. Oxford.

PH = Nagy 1990a.

Porter, J. I. 1992. “Hermeneutic Lines and Circles: Aristarchus and Crates on the Exegesis Of Homer.” Homer’s Ancient Readers: The Hermeneutics Of Greek Epic’s Earliest Exegetes. Ed. R. Lamberton and J. J. Keaney. 67–114. Princeton, NJ.

PP = Nagy 1996a.

Rengakos, A. 2000. “Aristarchus and the Hellenistic Poets.” Seminari Romani de Cultura Greca 3:325–335.

Schmitt, R. 1967. Dichtung und Dichtersprache in indogermanischer Zeit. Wiesbaden.

Schrader, H., ed. 1880–1882. Porphyrii Quaestionum Homericarum ad Iliadem pertinentium reliquiae. Leipzig.

Severyns, A. 1928. Le cycle épique dans l’école d’Aristarque. Bibliothèque de la Faculté de Philosophie et Lettres de l’Université de Liége 40. Paris.


[ back ] 1. Severyns 1928:33–34n4.

[ back ] 2. The wording comes from Porphyry, Homeric Questions [Iliad] 297.16 ed. Schrader; see also scholia D for Iliad V 385. On the Aristarchean provenience of the wording, see Porter 1992:70–74 (who effectively addresses the skepticism of Pfeiffer 1968:225–227).

[ back ] 3. HC 2§188.

[ back ] 4. Severyns 1928:81.

[ back ] 5. The wording of this paragraph corrects some imprecisions in the wording of the first printed version (2009) of HC 2§189. On Hesiod as neōteros according to the Aristarcheans, see Severyns 1928:39, 89; on the poets of the Cycle as neōteroi, see especially p. 63, where Severyns argues that Aristarchus considered the Cycle to be a major component of this neoteric category.

[ back ] 6. HC 2§190.

[ back ] 7. Rengakos 2000.

[ back ] 8. HC 2§192.

[ back ] 9. HC 2§193. For a survey of references dating back to the Hellenistic era, I cite again Rengakos 2000, especially p. 333 on the criticism leveled at Callimachus Aetia Book 1 F 13 by Apollodorus of Athens FGH 244 F 157 (ed. Jacoby II B p. 1089 line 34); via Strabo 1.2.37 C44: among the neoterisms of Callimachus mentioned here by Apollodorus is the poet’s equation of Corcyra with the island of the Phaeacians.

[ back ] 10. Pfeiffer 1968:150.

[ back ] 11. Severyns 1928:63.

[ back ] 12. See especially Severyns 1928:44, 46, 98–99.

[ back ] 13. HC 2§217.

[ back ] 14. HC 2§219.

[ back ] 15. HC 2§220.

[ back ] 16. HC 2§221.

[ back ] 17. HC 2§223.

[ back ] 18. HC 2§224.

[ back ] 19. For an overall assessment of Aristarchus’ methods in collecting Homeric textual variants, see HTL ch. 5.

[ back ] 20. I offer an overall survey in HTL ch. 1.

[ back ] 21. For a tracing of the broad outlines of this Aristarchean collection of neoteric variants, as indirectly reflected in the usage of the Hellenistic poets, see Rengakos 2000.

[ back ] 22. HC 2§225.

[ back ] 23. HC 2§226.

[ back ] 24. In displaying the Homerus Auctus, the base text of Zenodotus was different from the base text of Crates in one important way: Zenodotus followed an editorial policy of marking for deletion the plus verses that Crates later chose to leave undeleted.

[ back ] 25. HC I§34.

[ back ] 26. HC I§21.

[ back ] 27. HC I§22.

[ back ] 28. In the Homeric A scholia, the use of ὅτι ‘because’ as the word that introduces information derived from Aristonicus is a conventional way of indicating that ‘Aristarchus writes a marginal sign in the margin because …’. In this case ὅτι was keyed to the marginal sign of the diplē periestigmenē (>:) placed at the left of Iliad II 318 in the base text of Aristarchus, and this placement of the sign is still attested in the Venetus A manuscript of the Iliad. Then there was the marginal sign of an obelos (–) placed at the left of Iliad II 319, and this placement is also still attested in the Venetus A.

[ back ] 29. In making this point, I agree with Montanari 2008.

[ back ] 30. On the pronunciation ἀρίδδηλον, see Chantraine DELG s.v. ἀρίζηλος, with further citations.

[ back ] 31. Again, Chantraine DELG s.v. ἀρίζηλος.

[ back ] 32. Chantraine DELG s.v. ἀίδηλος. He discusses both the morphology and the meaning of aïdēlos, reconstructing an active sense of ‘causing someone or something to become invisible’ alongside the intransitive sense of ‘invisible’. I propose that the meaning ‘invisible’ is attested also in the figurative sense of ‘inconspicuous, undistinguished’ in Homeric contexts where aïdēlos is applied as an insult to morally undistinguished characters (as in Odyssey xxii 165).

[ back ] 33. Kirk 1985:149.

[ back ] 34. Montanari 2008.

[ back ] 35. HC 3§13.

[ back ] 36. HC 3§14.

[ back ] 37. HC 3§15.

[ back ] 38. HQ 38, 89.

[ back ] 39. PH 52–81.

[ back ] 40. PH 70–79.

[ back ] 41. Pfeiffer 1968:73; HQ 38.

[ back ] 42. PP 74–75.

[ back ] 43. BA 297–300, interpreting the evidence assembled by Schmitt 1967:296–298.

[ back ] 44. BA 300.

[ back ] 45. HC 3§16.

[ back ] 46. HC 3§17.

[ back ] 47. PP 81.

[ back ] 48. HC 3§20.

[ back ] 49. HPC I§§169 and following.

[ back ] 50. PP 150–151.

[ back ] 51. Cf. Davison 1955:21, Pfeiffer 1968:228, Janko 1992:32 (n53), 71.

[ back ] 52. Porter 1992:83.

[ back ] 53. PP 151–152.

[ back ] 54. PP 152.

[ back ] 55. For the theory that the Iliad and Odyssey were indeed dictated in the eighth century, see Janko 1992:22, 26.

[ back ] 56. Summary, with bibliography, in HQ ch. 2.

[ back ] 57. HC I§167.

[ back ] 58. Frame 2009 ch. 11.

[ back ] 59. HPC I§168.

[ back ] 60. HPC I§169.

[ back ] 61. PH 5–81.

[ back ] 62. Frame 2009 ch. 11. Also, I agree with Frame’s argument (pp. 583–584) that the Homeric Iliad and Odyssey, as epic traditions, “reached Athens almost immediately after they took root on Chios, and that even earlier they may have begun to be known in Athens directly from the Panionia.”