Cⓢ1. Five centuries of Homeric transmission

C§1. I propose to outline here the overall chronology of Homeric transmission as I have reconstructed it in this book. The basis for my overall reconstruction is the Homeric Koine, which I have equated with the Panathenaic Homer in the era of the Athenian democracy. This Koine was relatively unaugmented. To be contrasted is the Homerus Auctus, which I have defined as an augmented or expanded Homer. As I showed in Chapters 2 and 3, the themes that differentiate the poetry of the Homerus Auctus from the poetry of the Homeric Koine are characteristic of Cyclic, Orphic, and Hesiodic traditions. As I also showed, these traditions are pre-Homeric – that is, if we define pre-Homeric in terms of earlier periods when the Cyclic, Orphic, and Hesiodic traditions were as yet undifferentiated from what later became the Homeric tradition. In other words, the Cyclic, Hesiodic, and Orphic variants of the Homerus Auctus predate the Homeric Koine. I reconstruct a succession of five phases corresponding roughly to five successive centuries:

Phase A. The augmented Homer or Homerus Auctus took shape in an era that preceded the fifth century BCE. This version of Homer may be considered Panathenaic, but only in terms of the sixth century BCE. This phase is reconstructed in the twin book Homer the Preclassic.
Phase B (as introduced in Chapter 4). The Homerus Auctus was ultimately replaced as the Panathenaic Homer by an unaugmented Homer in the era of the democracy in the fifth century BCE. This new Panathenaic Homer is what I have been calling the Homeric Koine.
Phase C (as introduced in Chapter 3). The Homeric Koine persisted as the Panathenaic Homer into the age of Plato, in the fourth century BCE.
Phase D (as introduced in Chapter 2). The Homerus Auctus resurfaced, but not as a Panathenaic Homer, in the later age of Callimachus, in the third century BCE. {589|590}
Phase E (as introduced in Chapter 1). The Homerus Auctus was suppressed and replaced by the unaugmented Homeric Koine in the still later age of Aristarchus in the second century BCE. This unaugmented Koine of Homer, as edited by Aristarchus, was no longer simply a Panathenaic Homer. In editing the unaugmented Homer, Aristarchus reported in his commentaries a mass of non-Koine variants that had gone unreported in earlier editions of the unaugmented Koine of Homer, and many of these non-Koine variants eventually infiltrated the textual transmission of the unaugmented Homer.

C§2. In the age of Plato, the Athenian Koine was the dominant version of Homer. Later on, in the age of Callimachus, the Homerus Auctus became the dominant version – so much so that I never needed to use the word koinē in Chapter 2, which I devote to that later age. Still later, in the age of Aristarchus, the Athenian Koine became once again the dominant version of Homer.

C§3. What I just said about the age of Aristarchus is a one-sided formulation, however. In order to show the other side, I highlight a basic fact. The age of Aristarchus, director of the Library in Alexandria, was also the age of Crates, director of the Library in Pergamon. And while the base text of Homer used by Aristarchus was the Athenian Koine, the base text used by Crates was the Homerus Auctus.

Cⓢ2. An example of significant differences in theme between the Koine and the Homerus Auctus

C§4. There is a reference made by Callimachus to Scheria, the mythical island of the Phaeacians in the Odyssey. In his poetry, Callimachus equated this mythical place with a historical place, the island of Corcyra (Aetia Book 1 F 12, 13, 15). In making this equation, he was following a traditional theme already known to Thucydides, who says explicitly that the people of Corcyra claimed to be descended from the Phaeacians (1.25.4). [1] This theme differs from what we find in the text of Homer as we know it, which I identify with the Homeric Koine. This different theme, I argue, stems from the textual tradition that I identify with the Homerus Auctus.

C§5. The wording of this different theme is actually attested at verse 158 of Odyssey xiii. In the scholia linked to this verse (at xiii 152), we learn that Aristophanes of Byzantium reported a reading that differed from the reading {590|591} he found in the Homeric Koine. [2] The difference in meaning has to do with an equation of the mythical Scheria, island of the Phaeacians, with the historical Corcyra. This equation was possible in terms of the variant reading, but it was impossible in terms of the reading found in the Homeric Koine, that is, in the Athenian Homer stemming from the new era of the democracy. In the Koine version of the Odyssey, the Phaeacians are cut off from the world outside their mythical past after Poseidon interposes a huge mountain that seals them off forever. In this particular version, t he wording at verse 158 of Odyssey xiii is μέγα δέ σφιν ὄρος πόλει ἀμφικαλύψαι ‘and make the huge mountain envelop their city’. In the non-Koine version favored by Callimachus, by contrast, Zeus enjoins Poseidon not to interpose the mountain. As we know from the testimony of Aristophanes of Byzantium, the variant wording is μηδέ σφιν ὄρος πόλει ἀμφικαλύψαι ‘but do not make the mountain envelop their city’. [3]

C§6. In terms of the Koine version heard by Athenians in the new era of the democracy, the mythical place of Scheria cannot be identified with the historical place of Corcyra, since Scheria had been sealed off forever. In terms of the non-Koine version favored by Callimachus, by contrast, the possibility of such an identification is not sealed off but left open. Thus the Phaeacians are saved from the fate of losing contact with the real world of their future, and they retain the alternative fate of becoming the ancestors of the people of Corcyra. [4]

C§7. But the Koine version of Homer negates such an identification of Scheria with Corcyra. I interpret this negation in terms of politics as well as poetics. The political terms correspond to the imperial design of Athens in the era of the democracy. If the mythical Scheria can be sealed off from the historical Corcyra, it is owned by Athens; if it is not sealed off, it is owned by Corcyra. [5] The Athenians may be said to own the mythical place of Scheria because of a political reality, that is, because they actually controlled the Homeric Koine in the era of the democracy. In the undifferentiated Homerus Auctus as emulated {591|592} by Callimachus and his contemporaries, by contrast, the imperial designs of the Athenians were not so clearly foregrounded. [6]

Cⓢ3. On the Homerus Auctus in the age of Virgil

C§8. In the twin book Homer the Preclassic, I argue that Virgil in the first century BCE drew on the Homerus Auctus as a source for his own epic, the Aeneid. [7] In drawing on the Homerus Auctus, he was emulating Callimachus and his contemporaries in the third century BCE. Like Callimachus, Virgil was a neoteric poet in that he based his poetry on the Homerus Auctus. Like Callimachus, he was drawing on a form of Homer that seemed to him newer than the supposedly real Homer. This seemingly newer Homer, however, was really the oldest available Homer of them all. As I also argue in Homer the Preclassic, the Homerus Auctus of Virgil preserved some of the oldest recoverable phases of Homer. [8]

C§9. As I argued in Chapters 1 and 2 of the present book, the Homerus Auctus was antithetical to the Homer of Aristarchus, whose edition represented the most limited and the most rigid of all Homers. This Homer of Aristarchus was the equivalent of scripture for the premier Aristarchean in the era of Virgil, Didymus. From the standpoint of an Aristarchean scholar like Didymus, the using of the Homerus Auctus by any poet would be the equivalent of abusing Homer. Such an attitude is not confined to the ancient world. Even today Virgil is occasionally criticized as an unworthy imitator of Homer on the grounds that he infuses his own epic with elements judged to be Cyclic, Hesiodic, and Orphic. Such criticism, I submit, is based on the same kinds of criteria that produced the most rigid Homer of them all, the quasi scriptural Homer of Aristarcheans like Didymus in the first century BCE. {592|593}

C§10. In Chapter 1, I was already saying that the Homer of Aristarchus was too rigid a Homeric model for Virgil. But now the question is, what kind of Homeric model could Virgil have chosen instead? Clearly, the most useful Homer for Virgil would have been the Homer of Zenodotus. This text of Homer was the text used by Callimachus and other neoteric poets, including Apollonius and Theocritus. [9] And Virgil’s poetry is modeled on their neoteric poetry.

C§11. It is common knowledge, of course, that Callimachus, Apollonius, and Theocritus – all three – represent major influences in the shaping of Virgil and his poetic repertoire. Virgil, after all, imitated these three poets as much as he imitated Homer, if not more so. To the extent that Virgil did indeed choose these neoteric poets as his models, he too was a neoteric poet.

C§12. But my argument goes further. Virgil had access not only to neoteric poets like Callimachus but also to the Homerus Auctus. To say this much is not the same thing as saying that the Homerus Auctus in the age of Virgil was represented by the base text once used by Zenodotus in the age of Callimachus. In the age of Virgil and already in the age of Aristarchus, that base text had been replaced by a narrower base text representing the unaugmented Athenian version. How, then, did Virgil have access to a Homerus Auctus? Here it is essential to recall the fact that the age of Aristarchus in Alexandria was also the age of Crates in Pergamon. This Crates, as I argued in Chapter 2, used for his edition of Homer a base text that still represented the Homerus Auctus. [10] Further, Crates as editor of this Homerus Auctus happened to privilege many of the same elements that neoteric poets in the age of Callimachus considered to be neoteric. What I am arguing, then, is that Virgil preferred the Homerus Auctus as represented by the Homer of Crates, not the Homeric Koine as represented by the Homer of Aristarchus. Virgil’s epic Aeneid was based on the inclusive Homer of Crates, not on the exclusive Homer of Aristarchus.

C§13. Just as the base text used by Aristarchus for his edition of Homer in Alexandria was inherited rather than created by that editor, the same can be said about the augmented base text used by Crates in Pergamon – and, a century earlier, by Zenodotus in Alexandria. The Homerus Auctus, which was the basis of the texts used by Crates and by Zenodotus before him, was not an editorial creation resulting from arbitrary additions of Cyclic, Hesiodic, and Orphic elements to an original Homer. Rather, as I argue in Homer the Preclassic, {593|594} the Homerus Auctus was a poetic creation resulting from organic accretions later judged to be non-Homeric and therefore neoteric, that is, Cyclic, Hesiodic, and Orphic.

Cⓢ4. The Shield of Aeneas in the Aeneid of Virgil

C§14. A prime example of Virgil’s use of the Homerus Auctus is his narrative of the Shield of Aeneas in Aeneid 8. This narrative is based not only on the narrative of the Shield of Achilles in Iliad XVIII. As Philip Hardie has shown, it is based also on the work of Crates himself in editing and commenting on this Iliadic narrative. [11] In the Epilegomena of Homer the Preclassic, I explore at some length the ways in which Virgil’s poetry not only emulates the Homerus Auctus but also refers to the edition and commentary of Crates. Here in this Appendix, I simply offer my own translation of relevant portions of Virgil’s narrative, along with a brief commentary of my own.

C§15. To narrate the Shield of Aeneas is to narrate the imperial power of Rome. [12] Such power is not only imperial but cosmic, thus transcending the conventional forms of epic narration. It is Virgil’s ultimate poetic formulation of cosmos and imperium: [13]

Virgil Aeneid 8.615-629, 729-731

615  dixit, et amplexus nati Cytherea petivit,
arma sub aduersa posuit radiantia quercu.
ille deae donis et tanto laetus honore
expleri nequit atque oculos per singula volvit,
miraturque interque manus et bracchia versat
620  terribilem cristis galeam flammasque vomentem,
fatiferumque ensem, loricam ex aere rigentem,
sanguineam, ingentem, qualis cum caerula nubes
solis inardescit radiis longeque refulget;
tum levis ocreas electro auroque recocto,
625  hastamque et clipei non enarrabile textum. {594|595}
illic res Italas Romanorumque triumphos
haud vatum ignarus venturique inscius aevi
fecerat ignipotens, illic genus omne futurae
stirpis ab Ascanio pugnataque in ordine bella.

Talia per clipeum Volcani, dona parentis,
730  miratur rerumque ignarus imagine gaudet
attollens umero famamque et fata nepotism.

615  She spoke, and the goddess of the island Cythera sought the embrace of her son,
and she placed the arms [arma], radiant, under the shade of an oak that faced her.
And he, rejoicing at the gifts of the goddess and at such a great honor,
cannot get his fill as he lets his eyes turn them over, over and over again, one by one, [14]
and he looks in wonder at them, handles them, and turns them over and over in his arms,
620  the helmet’s crest, with its look of terror, which belches bursts of flame,
and the sword that brings doom, and the rigid breastplate made of bronze,
blood-red and huge, like some bluish cloud
that is set on fire by the rays of the sun and reflects them from afar.
Then there were the polished shin-guards, made of amber and refined gold,
625  and the spear and the shield, the weaving [textus] of which is beyond all power to narrate.
It was there [= on the shield of Aeneas] that the story of Italy and the triumphs of the Romans {595|596}
had been made by one who is not at all without knowledge of what seers know, not without knowledge of the age that was yet to come, [15]
made by the master of fire, yes, there was the entire lineage of the future
descendants of Ascanius [/ Iulus], and the wars to be fought, to be fought in due order, those wars. [16]

Such things, through Vulcan’s shield given by his mother,
730  he gazes at with wonder [mirārī] and, having no knowledge of the universe that is there, he rejoices in the image [imāgō] [17]
and lifts it up on his shoulder. There is was, what was to be the fame and the destinies of his descendants.

C§16. The Shield of Aeneas is signaled here in Aeneid 8.616 with arma ‘armor, arms’ as the first word, which corresponds to the first word of the epic, at Aeneid 1.1: arma virumque cano ‘armor I sing, and the man’. But now, in the first reference to arma ‘armor’ in Aeneid 8.616, the description of the Shield as a shield has not yet happened. So far, only the armor in general is being described. But there is more to it: the word arma ‘armor’ here in Aeneid 8.616, by way of cross-referring to the initial use of arma in Aeneid 1.1, stands metonymically for the whole epic, not only for the ‘armor’ of Aeneas. The arma ‘armor’ at the beginning of Aeneid 1.1 can apply here in Aeneid 8.616 if we understand the deployment of arma in Aeneid 1.1 as a masterstroke of metonymy. What is being signaled by the arma in Aeneid 8.616 is a description of the Shield that becomes coextensive with the overall narration of the epic that is the Aeneid in its entirety. But when the actual description of the Shield begins in Aeneid 8.625, the wording makes it clear that this description defies any immediate narration: clipei non enarrabile textum ‘the shield, the {596|597} weaving [textus] of which is beyond all power to narrate’. To describe such a cosmic power will require an overall epic narration, from beginning to end, which cannot be successful until the story is fully told. Such a narration calls for a metaphor to substitute for the narration: instead of a tale that is being told, the narration is reconfigured as a web that is being woven, a textus. The story has to be told from beginning to end, just as a web has to be woven from beginning to end. Further, the metaphor of the woven web is itself a crossover from the world of weaving to the world of metalwork, as performed by the god Vulcan. There is a comparable crossover in Iliad XVIII 590, where the word poikillein ‘pattern-weave’ refers to the metalwork performed by the god Hephaistos in making the Shield of Achilles. [18] Virgil must have had this reference in mind in Aeneid 8.625, describing the ineffable textus or web that is the Shield of Aeneas. The ineffable power of this web, cosmic and imperial, is ultimately realized as the epic that is the Aeneid. It is simultaneously realized as the universal globe that is the Shield of the Aeneas, the description of which defies immediate narration but becomes coextensive with the ultimate narration of the Aeneid. As the description of the Shield reaches completion, the hero of the Aeneid shoulders its massive orb as his own burden.

C§17. As Hardie has shown, the Shield of Aeneas is not only a picturing of the globe that is the universe. The Shield is meant to be the globe itself. For Aeneas, carrying the Shield in Aeneid 8.730-731 is the same thing as carrying the weight of the whole world on his shoulder. The shouldering of this titanic burden by Aeneas matches the shouldering of the universal globe by Atlas the Titan. Here I find it relevant to show an evocative image. It is the statue of the Farnese Atlas. The sculpture pictures the Titan in the act of shouldering a celestial sphere, which is an idealization of the earthly sphere (Figure 18).

Figure 18. Sculpture: the “Farnese Atlas.” Roman marble “copy” of a Hellenistic Greek original. Naples, Museo Archeologico Nazionale, 6374.

C§18. Such a visualization of Atlas struggling underneath the massive burden of a cosmic and imperial globe was inspired by theories about a spherical world, and these theories were in turn inspired by allegorizing traditions stemming from the Homerus Auctus, that is, from a text that combined – or recombined – the world of Orpheus with the world of Homer. The burden that weighs so heavily on the shoulder of this primordial Atlas is analogous to the cosmic and imperial burden of an augmented and theoretically all-inclusive Homer. {598|599}

C§19. Aeneas does not yet know what it means for him to shoulder the Shield that was made for him by the divine metalworker – and, ultimately, by the poet of his own epic, which is the Aeneid. But the poetic description of what is pictured on the Shield has already prophesied that meaning. We may compare the poetic image of the petrified serpent in Iliad II verse 319, in the context of verses 299-332, as I analyzed it in Chapter 1: that stop-motion picture of terror and pity turns out to be a prophecy of the Troy story – once that story is fully told from beginning to end. So also the stop-motion picture of the Shield of Aeneas prophesies the story of Rome.

C§20. The Shield of Achilles, as a globe, was simply Homeric as far as Crates was concerned. Since it was evidently the model for the Shield of Aeneas, it must have been Homeric for Virgil as well. Evidently, Virgil’s Homer was the Homerus Auctus of Crates, to be distinguished from the narrower Homer of Aristarchus, for whom the Ōkeanos was a fresh-water river encircling an earth that was flat. Virgil’s Homer is also to be distinguished from the supposedly real Homer as edited by Zenodotus, for whom the verses about the Ōkeanos – and in fact all the verses about the Shield of Achilles – were Orphic accretions that needed to be athetized in his base text of Homer.

C§21. From all we have just seen, I conclude that the idea of cosmos and imperium in Virgil’s Aeneid was derived from the Homerus Auctus – as mediated by the Homeric edition and the Homeric commentaries of Crates in Pergamon. This Cratetean Homer was the source for the imperial design of Virgil’s Aeneid. {599|600}


[ back ] 1. Nagy 2001c:89, with reference to Thucydides 3.70.4 as well as 1.25.4.

[ back ] 2. Nagy 2001c:93n65.

[ back ] 3. Nagy 2001c:83-84.

[ back ] 4. Nagy 2001c:84-91.

[ back ] 5. Douglas Frame notes that we may see traces here of converging Athenian and Panionian agenda.

[ back ] 6. Nagy 2001c:84-91. The politics and poetics of identifying Corcyra with the land of the Phaeacians are comparable, I think, to the politics and poetics of identifying the historical Ithaca of the first millennium BCE with the homeland of Odysseus. An essential place-marker for the latter identification was the “Cave of the Nymphs” in the Bay of Polis on the northwestern coast of Ithaca, which was a stopping point for travelers sailing from the Corinthian Gulf to Corcyra and back. On the links between this cave, which is the location of a hero-cult for Odysseus, and the Cave of the Nymphs in Odyssey xiii 363, where we see Odysseus in the act of hiding the tripods given to him by the Phaeacians, see Malkin 1998:64-67, 95-107. On the epic poetry of the Odyssey as a form of “merchandise,” see the pertinent analysis of Dougherty 2001:53-68. For an earlier localization of Ithaca, dating back to the second millennium BCE, see Bittlestone 2005.

[ back ] 7. HPC E§§145 and following.

[ back ] 8. HPC E§§154 and following.

[ back ] 9. See ch. 2 on the neoterism of Callimachus and his contemporaries.

[ back ] 10. See also LP (Nagy 1998) 215, 222-223.

[ back ] 11. Hardie 1986.

[ back ] 12. In Virgil Aeneid 8.640, the wording Iovis ante aram ‘in front of the altar of Jupiter’ may imply a mental association with the Altar of Zeus at Pergamon.

[ back ] 13. For this term, see especially Hardie 1986:339.

[ back ] 14. The visual sequencing of the metalwork is made parallel to the verbal sequencing of the poetry.

[ back ] 15. In Aeneid 8.627, Vulcan is haud vatum ignarus ‘not at all without knowledge of what seers know’. Thus the metalworker is conscious of an equation between the metalworking that creates the Shield and the songmaking that creates the description that defies narration. The choice of the word vates instead of poeta is appropriate to the Augustan poetics of a romanized Homerus Auctus, containing elements that critics like Zenodotus would have considered non-Homeric and therefore Orphic.

[ back ] 16. See my earlier remark on the visual sequencing of the metalwork.

[ back ] 17. The character of Aeneas cannot yet ‘read’ the prophecy woven into the description that cannot be narrated as a narration, only as a weaving.

[ back ] 18. HPC II 413. Also, in HPC E§148, I argue that the use of the word triplax ‘three-fold’ or ‘triple’ with reference to the Shield of Achilles at Iliad XVIII 480 shows another comparable pattern of crossover from weaving to metalwork.