7. Homeric Poetics and the Aeneid

Curtis Dozier
Everyone has always known that Vergil imitated Homer, but I do not think it is known whether Vergil thought he was imitating, on the one hand, the work of a blind bard who single-handedly recorded the best versions of the sack of Troy and the wandering of Odysseus, or, on the other hand, the result of a centuries-old oral tradition. Hence the question that Gregory Nagy’s Sather Seminar at the University of California—Berkeley in the spring of 2002 left me: How much of what we know about the way Homeric poetry was created and performed could Virgil have known? Our understanding of the artistry of Virgil’s poetry might very well hang on the answer. If Virgil thought he was imitating a single poet, he was (in Nagy’s view) wrong—he misunderstood his model in a fundamental way, since the oral nature of Homeric poetry shaped almost all of its features. In this case we need to ask how this misunderstanding conditions Virgil’s imitation. But if Virgil knew he was imitating an oral tradition, we must try to understand how a literate poet represents oral poetics. These two issues might be phrased in contrast with each other: How does a poet imitate formulaic language when he does not realize that it is formulaic? And how does a poet imitate formulaic language when he knows that his language is not formulaic?
These tantalizing questions lie behind much of the scholarship on Virgil’s imitation of Homer, but it may not be possible to divine what Virgil could have understood about formulaic language. In providing a partial answer, I want to sidestep this problem by asserting that, regardless of understanding, Virgil imitated whatever he saw when he read Homer, or whatever he heard when he heard him read. [1] And so I think it is safe to say that to Virgil, the formulaic language of Homer’s poetry must have looked repetitive. [2] In this essay I will examine some ways in which Virgil imitated this feature of Homeric language and how it shapes his poetry, and along the way I will propose what I think is a new way of conceptualizing the relationship of Virgil’s poem to the Homeric epics.
My discussion of Virgil’s language will depend on some aspects of Nagy’s view of how Homeric language works, which I summarize briefly here. To Nagy and his followers, formulaic language is not merely repetitive, but, one might say, pregnantly repetitive. That is, despite their formulaic nature, repetitions mean something. When Nagy offers the analogy “formula is to form as theme is to content” (Nagy 1996b:18), he is saying that formula is the basic unit of form just as theme is the basic unit of content, but he also means that in an oral tradition, formula and theme are inextricably linked. [3] Verbal repetition is thematic repetition. And so each particular formula invokes the other instances of that formula, [4] by which a network of meaning (often very intricate) is created. [5] Generations of unitarians and Nagy’s contemporary critics interpret these networks, which are marked by verbal echoes separated by immense numbers of lines, as proof of a literate, individual composer of the poems, [6] but Nagy sees them as an integral feature of oral poetics. [7] Such networks are an integral feature on the production side of epic poetry because a bard’s virtuosity is measured, in part, by how deftly he creates these networks through traditional language, and they are an integral feature of the poetry on the reception side because formulaic language is “hard-wired” in the minds of both performer and audience. [8] This language was omnipresent in their culture; phrases accumulated meaning for all members of the society, and people had heard so many performances, so many retellings of these episodes, that for many of them certain language was inextricably associated with certain narratives. [9] Thus when that language was used in a different context, all those associations came up in the mind of the listeners. They might not remember specific details of the times they had heard that language before, but the associations were still operative on the level of impression—the mood or tone of an episode could be summoned by the artful use of formulaic language to contrast or reinforce the mood of the episode being performed. [10] For the modern literary critic, then, there is no problem of authorial intention: if you can think it, then it is reasonable to imagine that some portion of the ancient audience could have thought it as well, more attuned to the language as they were.
The now-classic Structuralist model of langue and parole is instructive here: [11] the epic Tradition, with a capital T, forms a complete system of expression, the langue. All the performers and all the audience members are conversant in this langue. Particular instances of a formula in a particular performance are individual expressions of this system, that is, examples of parole. Thus the networks of meaning I have described can be seen as instances of the parole invoking the whole langue. This relationship is not so cut-and-dried when it comes to networks of meaning within the Homeric corpus, because while the langue must be described diachronically, over time, each parole exists as a synchronic event. [12] The Homeric corpus we read today is the product of a diachronic process; it describes the langue as it ultimately crystalized. [13] What we lack is a synchronic snapshot of this tradition—that is, we lack an example of any particular performance of Homer. [14] This lack of synchronic perspective makes discussing the network of meaning created by any parole difficult, because the performance in which that parole occurred must remain hypothetical. [15] In response to his critics, Nagy even acknowledges as much, saying that Homeric cross-references must be diachronic in nature (Nagy 1996b:9; Nagy 2000:40).
And so I return to Virgil and his treatment of what to us are Homeric formulas, and to him were repeated lines and phrases. These repetitions in Homer create networks of meaning, and the same is true of Virgilian poetry, which is rife with the same kinds of cross-references that abound in the Homeric corpus. They are unmistakably literary in character; there is no modern analytic debate about the Aeneid. [16] And they are not only intratextual [17] but intertextual; indeed, perhaps no other poet has so openly absorbed so many poetic traditions into a single work. [18] But I think it is uncontroversial to say that the most pervasive model for the Aeneid were the Homeric poems. [19] Virgil not only imitated Homeric repetitions but imitated the networks of meaning that they created.
In this essay I want to perform an experiment on the Aeneid: what if we see it as a parole for the Homeric langue? Or, to address the problem with Homeric cross-references that I summarized above, what if we take the Aeneid as a synchronic snapshot of the Homeric langue, as an individual performance of Homeric themes in Homeric language for a particular audience at a particular moment in time? Characterizing the Aeneid in this way will allow us to examine cross-references between a specific parole and the diachronic langue.
This is not just a theoretical way of talking about allusions to Homer in the Aeneid, a subject that even the earliest commentators on Virgil studied. I actually want to treat the Aeneid as part of the epic cycle [20] for a day, as it were, which means that I will treat Virgilian allusions to Homer as if they were cross-references within a single poetic system. Research on allusion tends to treat it as a one-way street: a later author alludes to a passage in an earlier author. But cross-references within Homer lack this chronological element, for the networks of meaning are diachronic: one passage refers to a second, which refers to three other passages, one of which refers back to the original passage, and so on. Similarly, in this experiment, a Virgilian passage may refer to a Homeric passage, which then invokes another passage in Homer as well as a Virgilian passage. A metaphor of depth may be useful: upon the bedrock of the Homeric langue are built the myriad Homeric cross-references; they reach down, like roots, into the langue as far as one cares to follow them. The Aeneid, I propose, can be placed like a superstructure on these networks, sending down its own roots into the Homeric networks and the Homeric langue below.
The specific Virgilian network I will consider is that of the phrase arma virumque, the first two words of the Aeneid. [21] These two words recur together eleven times throughout the poem, and their programmatic significance is often cited by commentators, if not exhaustively analyzed. [22] Here I focus on their intertextual significance—they represent Virgil’s relationship to Homer in that arma is a metonymic description of the Iliad and virum is an exact translation of the first word of the Odyssey, Ἄνδρα. [23] We are to understand that Virgil will imitate both the general themes and the smallest details of the Homeric epics. [24] Nagy’s concept of formulaic poetics can be found in these two little words: verbal and thematic allusion together.
Probably the most striking repetition of arma virumque comes in Book 9, when Turnus kills Cretheus:
… amicum Crethea Musis,
Crethea Musarum comitem, cui carmina semper
et citharae cordi numerosque intendere nervis,
semper equos atque arma virum pugnasque canebat.
Cretheus, the Muses’ friend and companion, who always loved songs, and the lyre, and to play through the modes on its strings; he was always singing of horses, the arms of men, and battles.
Aeneid IX 774–777
A sympathetic connection between this unfortunate poet and Virgil himself has not been missed by many critics (Hardie 1994 on 777): Cretheus sings his own version of the Aeneid, a poem whose subject is arma virum. Our first network of meaning begins to take shape around this idea, for similar self-conscious references occur several times in the Homeric poems as well. For example, in Book 9 of the Iliad the embassy finds Achilles in his tent singing the κλέα ἀνδρῶν, “glorious deeds of men.” (Iliad IX 189). The Iliad, of course, is itself poetry on this subject, especially in Book 9, where the preceding eight books have shown us the aristeia of various heroes, Greek and Trojan. The theme of the Iliad as stated in its proem is μῆνις, but this rage springs from the culture portrayed in the poem, in which men are valued by the spoils they win by accomplishing the very things Achilles sings of in his tent while he sulks; if anything, κλέα ἀνδρῶν represents more of the Iliad than μῆνις does. And we should note that Achilles sings of κλέα ἀνδρῶν because he cannot participate in them, despite his status as the greatest of all ἀνδρῶν. Aeneas too, the vir of Virgil’s poem, is absent from battle as Cretheus sings, since he is seeking assistance from Evander. Virgil’s typology of Aeneas as Achilles is not based solely on similar actions; here we find the connection made through the recurrence of the proem form.
We find a similarly self-referential situation in the Odyssey, where two separate rhapsodes sing of νόστοι. In Book 1, the suitors force Phemius to sing for them: ὁ δ’ Ἀχαιῶν νόστον ἄειδε / λυγρόν, ὃν ἐκ Τροίης ἐπετείλατο Παλλὰς Ἀθήνη (“He sang of the baleful homecoming of the Achaeans, which Pallas Athena accomplished [for them] after Troy”; Od yssey i 326–327). As with Achilles’ κλέα ἀνδρῶν, this subject can be taken to represent the subject of the Odyssey at least as accurately as its stated subject, ἀνήρ. Demodocus sings songs that combine the above elements in Book 8: Μοῦσ’ ἄρ’ ἀοιδὸν ἀνῆκεν ἀειδέμεναι κλέα ἀνδρῶν, / οἴμης, τῆς τότ’ ἄρα κλέος οὐρανὸν εὐρὺν ἵκανε, / νεῖκος Ὀδυσσῆος καὶ Πηλεΐδεω Ἀχιλῆος (“Then the Muse inspired him to sing of the fame of the men from that song, whose fame then reached the wide heavens, namely of the strife between Odysseus and the son of Peleus, Achilles”; Odyssey viii 73–75). This is more the stuff of the Iliad than of the Odyssey, recalling as it does the poetry of Achilles in his tent, but Book 8 of the Odyssey is when Odysseus begins to tell of his adventures, adventures that contribute to his κλέος. Further, the νόστοι of the Greeks were part of the same poetic tradition that told the story of the Trojan War; indeed, they were considered part of the story of the Trojan War, a connection which is being shown by the repetition of Achilles’ κλέα ἀνδρῶν here. Clearly we should take this as a catch-phrase for all epic poetry, especially in that Demodocus here combines the heroes of both the Homeric epics into a single poem about the strife between them, a subject that does not come up anywhere else in extant Greek poetry, according to the commentaries. [25]
Cretheus’ use of arma virum recalls all three of these Homeric singers, and so extends the meaning of Virgil’s programmatic phrase. In the opening line of the Aeneid, virum was accusative, but in Cretheus’ song the word is genitive plural, ‘of men’—thus in the first line of the Aeneid the word translates exactly the first word of the Odyssey, and in Cretheus’ song it translates exactly the genitive plural ἀνδρῶν of κλέα ἀνδρῶν. The virum of arma virum thus translates the subject of the Odyssey or the subject of the Iliad, as conceived by Achilles, depending on whether it is taken as accusative or genitive. If genitive, the metonymy between arma and the Iliad tightens considerably—arma virum becomes a paraphrase for κλέα ἀνδρῶν, with arma filling in for the κλέος of the war. And since Demodocus conceives of κλέα ἀνδρῶν as a phrase for all epic poetry, arma virum signals not only Virgil’s imitation of the Homeric epics but his conception of the Aeneid as part of the same poetic milieu as Homeric poetry.
There are other signs of this positioning as well. Demodocus’ third song in Odyssey viii is about the Trojan horse (Od yssey viii 499–520), and Cretheus’ poetry is not limited to arma virum but to equi as well. In other words, Cretheus does not seem just to know about the current mission of the Trojans in Italy, but about the whole story of Troy, including, we might imagine, the Iliad itself. He is a poet who brings the Iliad and the Aeneid together into one poem. The often observed connection between Cretheus and Virgil is justified, for Virgil is a poet who accomplishes the same feat. In fact the above connections indicate that the phrase arma virum marks passages in which Homeric precedents may be particularly keenly felt.
Now to a slightly more complex example: In Book 11 of the Aeneid the Amazon warrior Camilla slays many Trojans in the course of her aristeia. Just how badly things are going for the Trojans is made clear when Orchilochus’ death under Camilla’s spear is described in the language of the Aeneid’s proem: tum validam perque arma viro perque ossa securim … congeminat (“Then she hacked powerfully again and again with her axe, through the man’s arms, and then through his bones”; Aeneid XI 696–698). But the tide begins to turn against Camilla when the Etruscan general Tarchon, in order to spur on his troops to come to the aid of the Trojans, stages a dazzling coup. He snatches his opponent Venulus off his horse and wrestles with him on his own horse as he leads his men into battle: volat igneus aequore Tarchon arma virumque ferens (“Like a fire Tarchon flies across the plain, carrying arms and the man”; Aeneid XI 746–747). Thus Tarchon carries as his battle standard a physical manifestation of the phrase arma virumque that embodies the Trojans’ destiny and their necessary victory. There is no Homeric precedent for the spectacular feat of wrestling one’s opponent off his horse and onto one’s own [26] and here we are to imagine that Tarchon knew he would have to do something truly unheard of in order to rally his troops. Tarchon’s charge, in particular his creation of the imagery of the proem of the Aeneid, inspires the Trojans to redouble their efforts in the battle against Camilla; Arruns immediately starts plotting her downfall, which sets into motion the final stages of the conflict and the poem.
What does have a Homeric precedent in this passage is the simile of the eagle and the snake, which describes Tarchon and Venulus wrestling on the horse:
utque volans alte raptum cum fulva draconem
fert aquila implicuitque pedes atque unguibus haesit,
saucius at serpens sinuosa volumina versat
arrectisque horret squamis et sibilat ore
arduus insurgens, illa haud minus urget obunco
luctantem rostro, simul aethera verberat alis.
And just as when a tawny eagle, flying on high, carries off a snake he has caught by winding his feet around it and clutching it with his claws, but when the wounded snake bends his winding coils and bristles with raised scales and hisses with its mouth, striking forcefully, no less does he press it as it struggles with his curved beak as he beats the air with his wings.
Aeneid XI 751–756
The event described in this simile is found in the Iliad in Book 12, when the Trojans, hesitating before pressing on toward the Greek ships, see an eagle carrying a snake. The snake bites the eagle and the eagle drops it and flies away (Il iad XII 200–207). Polydamas interprets this to mean that the Trojans may reach the ships but will not be able to return home with their booty; Hector dismisses him as superstitious and presses on. Polydamas is correct that the portent is ominous for the Trojans; they do not, in the end, destroy the ships, and although many do escape, probably just as many die. Tarchon creates a similar oracle to encourage the Trojans, the difference being that he does not drop his prey, Venulus, even though he fights back fiercely. The Trojans of the Iliad were doomed, and the dropped snake of their oracle reflects that, but the Trojans of the Aeneid must triumph; Tarchon knows that as long as he can hold on to his snake, as it were, his homemade oracle will be effective in portending Trojan victory.
In fact, we never find out what happens to Tarchon and Venulus; after the simile the narrative shifts to Arruns and Camilla. This lack of resolution in a passage that refers to an eagle seizing prey recalls another Homeric omen that ends with an unresolved image. At Iliad II 308–319 the Greeks, debating their future at Aulis, see a snake eat nine sparrows and their mother, after which Zeus turns the snake to stone as a monument for all time. Calchas interprets this to mean that the Greeks will take Troy in the tenth year; the omen is their assurance that Troy will soon fall. Tarchon and Venulus may not be turned to stone in the narrative of the poem, but the only image of them that the poetry provides is of them wrestling, without end. This limited perspective is a type of petrification—the words on the page describe the state in which Tarchon apparently remains, triumphantly clutching arma virumque. It is a portentious image, as the simile of the snake and the eagle, in connection with their Homeric precedent, shows, and it is a frozen image, like that of the snake in Book 2 of the Iliad. And just as Odysseus reminds the Greeks in Iliad II that after nine years only one remains, the Trojans here may sense that they are nearing the end of their conflict; Turnus’ death and the end of the war are only a book away.
This network of passages is perhaps less than spectacular in its depth, for it only involves one or two Homeric precedents for a Virgilian scene. Still, it shows some ways that we can build an interpretation of a particular Virgilian passage such as that of Camilla and Tarchon out of the network of Homeric passages invoked by those scenes. A potentially more fruitful network takes shape in Aeneid VI: while traveling through the underworld, Aeneas and the Sibyl reach “the most distant fields, separate [from the rest of the underworld], which the famous war heroes populate” (Aeneid VI 477–478). Aeneas stops and talks to the Trojan heroes, who in turn are eager to speak with him. The reaction of the Greeks, however, is quite different.
at Danaum proceres Agamemnoniaeque phalanges
ut videre virum, fulgentiaque arma per umbras,
ingenti trepidare metu; pars vertere terga,
ceu quondam petiere rates, pars tollere vocem
exiguam; inceptus clamor frustratur hiantis.
But the Greek noblemen, and Agamemnon’s soldiers, as they saw the man and his arms flashing through the shadows, shook with great fear. Some turned and fled, as when they had run for their ships, some lifted a faint shout; but the cries, only begun, stuck in their gaping throats.
Aeneid VI 489–493
This variation of arma virum has a Homeric quality to it—the epics make clear that one aspect of rhapsodic virtuosity was to reshape formulaic language into different metrical units for different narrative situations. The Greeks feel fear on several occasions in the Iliad, but this passage refers specifically to the Greeks’ flight to their ships in Book 15 shortly before Patroclus asks Achilles for his armor: “The Achaeans fled because of their weakness, for Apollo sent fear upon them, and gave honor to the Trojans and Hector” (XV 325–327). The implication is that Aeneas appears to the Greeks in the underworld as Hector appeared to them on that day.
But even as they connect Aeneas to the heroism of his ancestors, these passages recall several Homeric scenes that work together to foreshadow Aeneas’ future heroism. The phrase fulgentia arma is something of a formulaic phrase in the Aeneid, appearing ten times in various contexts. By comparison, the Homeric equivalent, τεύχεα παμφανόωντα, appears only twice. One of those appearances comes in Thetis’ plan to request arms for Achilles from Hephaestus in Iliad XVIII: εἶμι παρ’ Ἥφαιστον κλυτοτέχνην, αἴ κ’ ἐθέλῃσιν / υἱεῖ ἐμῷ δόμεναι κλυτὰ τεύχεα παμφανόωντα (“I will go to the famous smith Hephaestus, to see if he would be willing to give to my son glorious shining armor”; XVIII 143–144). Here the network of connections takes the kind of turn I am interested in, one that shows how Virgil integrated the Homeric structure into his own; the Virgilian passage, via fulgentia arma, recalls a Homeric passage, Thetis’ request for Achilles’ arms, which then evokes a Virgilian passage, Venus’ request to Vulcan for Aeneas’ arma: “arma rogo, genetrix nato. te filia Nerei, te potuit lacrimis Tithonia flectere coniunx” (“I ask for arms, a mother for her son. The daughter of Nereus and Tithonus’ wife were able to convince you [to do this] with their tears”; Aeneid VIII 383–384). I call attention to this passage because Virgil explicitly invokes the Iliad and the epic cycle; the wife of Tithonus refers to Aurora, who in the Aethiopis requests armor for her son Memnon. If more of the Aethiopis had survived than the four lines we have, we could follow this reference into that poem as well. As it is, I can only observe that Virgil had not only Homer but the whole epic cycle as the structure under his work.
The other instance of τεύχεα παμφανόωντα in the Iliad comes in Book 5: ἤριπε δ’ ἐξ ὀχέων, ἀράβησε δὲ τεύχε’ ἐπ’ αὐτῷ / αἰόλα παμφανόωντα, παρέτρεσσαν δέ οἱ ἵπποι / ὠκύποδες (“He fell from the chariot, and his arms rattled on him, shimmering and flashing, and the swift-footed horses swerved in fear”; Iliad V 294–296). This is the death of Lycaon at the hands of Diomedes, which may at first seem like an interpretive dead end, but the passage connects back to the Aeneid in a startling way: it is Aeneas who arrives to defend the body of Lycaon, and for the first but not last time in the Iliad, he only narrowly escapes death with divine aid and is thus able to go on to found the Roman race. So while this Homeric passage does not actually invoke any Virgilian passages, it nevertheless is an extremely important passage for the story of Aeneas.
To return to Aeneas’ encounter with the Greeks in the underworld, we find a similar network in an Odyssean passage alluded to there, namely the appearance of Heracles in the Nekuia: “Around him there was the screaming of the dead, as of birds fleeing terrified in every direction” (Odyssey xi 605). We may conclude that in some sense Aeneas in the underworld is a Heracles figure. But we can go further, because this Homeric passage gestures back, as it were, to a Virgilian passage: we make the connection between the Virgilian Aeneas and the Odyssean Heracles, and can then make a further connection between Aeneas and the Virgilian Hercules of Aeneid VIII. Hercules is a legendary hero for Rome before it was Rome, and so will Aeneas be a legendary hero for the next stage of the history of the site of Rome, the stage in which Rome reaches the height of its glory. And without the Homeric connection of Aeneas to Heracles, the resonance of this connection between Aeneas and Hercules in Book 8 could not be felt. And I have only scratched the surface of this network—as I mentioned earlier, the phrase fulgentia arma is repeated nine more times in the Aeneid. Aeneas’ encounter with the Greeks in the underworld refers to all nine, and each of those scenes marks the start of yet another path through the Homeric and Virgilian worlds.
One final example shows how dense the network of meaning can be, how the trail of reference can lead from the Aeneid into Homer and then back to the Aeneid, connecting Virgilian passages that would have no connection without the Homeric intermediary. Again we start with arma virum, as Dido, unable to convince Aeneas to stay in Carthage, resolves to kill herself and gives Anna instructions:
tu secreta pyram tecto interiore sub auras
erige, et arma viri thalamo quae fixa reliquit
impius exuviasque omnis lectumque iugalem,
quo perii, super imponas.
Build a pyre secretly in the inner parts of the house, in the open air, and place on it the man’s arms, which he wickedly left hanging in the wedding chamber, along with all his spoils, and the wedding bed, on which I was destroyed.
Aeneid IV 494–497
Dido is trying to convince her sister that burning Aeneas’ possessions will quench her love, but in reality she is planning her own death—the whole tragedy of Book 4, the inability of Aeneas to be what Dido wants him to be, is contained in her contemptuous use of the phrase arma viri, which is here used with Aeneas as the referent of vir for the first time in the narrative. The phrase recurs in reference to another pyre, this time in Book 6. Before Aeneas can enter the underworld, he must bury Misenus, who lies dead on the beach, killed when he challenged Triton to a trumpet-playing match. After his men perform most of the preparations for the pyre, Aeneas adds the finishing touch: At pius Aeneas ingenti mole sepulcrum imponit suaque arma viro remumque tubamqu e (“But Aeneas dutifully placed the tomb on a huge pile, along with the man’s arms, his oar, and his trumpet”; Aeneid VI 232–233). This pyre for a deceased companion calls to mind many Homeric scenes; the burial of Elpenor, killed when he fell from the roof of Circe’s house in a drunken stupor; the burial of Patroclus, killed and stripped by Hector; the burial of Hector himself; the burial of Greeks and Trojans during the temporary truce of Iliad VII; Agamemnon’s description of the burial of Achilles in the underworld in Odyssey xxiv; and Andromache’s burial of her father, Eetion. Of these pyres I want to highlight Elpenor’s and Eetion’s because, like Misenus’ and Dido’s pyres, they include arms, Elpenor’s because he did not die in battle and Eetion’s because Achilles did not take them.
First, Elpenor: αὐτὰρ ἐπεὶ νεκρός τ’ ἐκάη καὶ τεύχεα νεκροῦ, / τύμβον χεύαντες καὶ ἐπὶ στήλην ἐρύσαντες / πήξαμεν ἀκροτάτῳ τύμβῳ εὐῆρες ἐρετμόν (“But when the corpse and the dead man’s arms were burned, after raising a mound and dragging a gravestone upon it, we planted a well-shaped oar on the top of the mound”; Odyssey xii 13–15). This passage is itself a Homeric doublet: Elpenor requests to be buried this way at Odyssey xi 74. Heubeck ad loc. notes that Elpenor may have had arms, but that he was not a warrior at all, but merely an oarsman. Misenus, trumpeting on the beach while his companions made camp, is a similar sort of man. We could ask then what it means for the programmatic phrase of the Aeneid to appear in a description of a pyre that commemorates a man manifestly different from the heroes of the Trojan war, Achilles, Patroclus, and Hector, whose arms could not be included in their funeral, and that rather honors a man who died an unnecessary, natural (and therefore somewhat cowardly) death. Misenus’ funeral-with-arma also recalls that of Eetion, described by his daughter Andromache in Iliad VI as she begs Hector not to return to the fighting.
… κατὰ δ’ ἔκτανεν Ἠετίωνα,
οὐδέ μιν ἐξενάριξε, σεβάσσατο γὰρ τό γε θυμῷ,
ἀλλ’ ἄρα μιν κατέκηε σὺν ἔντεσι δαιδαλέοισιν
ἠδ’ ἐπὶ σῆμ’ ἔχεεν·
[Achilles] killed Eetion. And he did not strip his armor, because he honored him in his heart. But he burned him with his cunningly wrought arms and placed over him a tomb.
Iliad VI 416–418
Another reference to Eetion comes in Iliad XXII, as Andromache, lamenting the death of Hector, calls her father δύσμορος (XXII 481). And in this same speech Andromache describes what will be the pyre for her absent spouse: “But in your house there are clothes, fine and pleasing, made by the hands of women. But I will surely burn all of these things in a blazing fire—they are no use to you, since you will not lie in them” (XXII 510–514). Through this connection we come back to Dido’s pyre, which twists the details of the Homeric scene: Aeneas, the spouse, is absent, as Hector was, but Dido, unlike Andromache, burns herself: “We were both born for a single destiny,” mourns Andromache at Iliad XXII 477—one of suffering. Dido, too, was born for suffering, but the problem is that she and Aeneas were born for separate destinies. We might infer that she had wanted Aeneas to be another Hector, and she had wanted to be his Andromache. Now she sees the impossibility of this arrangement, that Aeneas cannot be another Hector. The damage has been done, though, and ironically she can only perish in anguish, like Andromache. Another relevant feature of Andromache’s speech is her lament for her son: “For even if he escapes the sorrowful war with the Achaeans, there will always be toil for him, and troubles in the future” (Iliad XXII 487–488). She goes on to describe the loneliness Astyanax will face, the alienation, the roaming. This passage reminds me, at least, of another passage from Aeneid IV, which would not have been recalled by the Aeneid’s pyre scene alone, Dido’s raving wish that she and Aeneas had produced a son: “At least if some offspring had been conceived in me from you before you fled, if some little Aeneas were playing in my courtyard, whose face would remind me of you, I would not seem to have been so completely deceived and abandoned” (Aeneid IV 328–330). The irony, of course, is that Aeneas does have a son, Ascanius, whose promised future is a factor in the decision to leave Carthage; Mercury tells Aeneas, “If the glory of such great things does not move you, think of Ascanius, growing up, and the hope of your heir Iulus” (Aeneid IV 272–274). Ascanius’ future as the first king of Alba Longa and the forefather of a distinguished race is quite the opposite of what Andromache expects for Astyanax, but like Andromache Dido is left with no future. This is a particularly striking effect because it is generated by a network between passages that starts in Aeneid IV, moves by allusion to Odyssey xii, to Iliad VI, to Iliad XXII, and then back to a different passage in Aeneid IV, a move not possible if we rigidly take the Aeneid to refer only to the Iliad as an earlier text, rather than as part of the same network of meaning created by the Homeric poems. And we can even form further links between these passages and those previously discussed. As we learned in Iliad IX, Achilles got the lyre with which he sang the κλέα ἀνδρῶν when he sacked the city of Eetion (Iliad IX 188), whose pyre (after Achilles killed him) forms one of the crucial links in the network under discussion here.
If we return to our original passage, Dido’s lament, we can follow this network in yet another direction. When Dido uses the words lectum iugalem (‘marriage bed’, Aeneid IV 496), she calls to mind the marriage bed of Odysseus, which Penelope uses as the final test to determine whether Odysseus is really who he says he is (Odyssey xxiii 205–206). But while the marriage bed of Odysseus and Penelope serves as a symbol of their undying love and a verification of their reunion, the lectum iugalem of Aeneas and Dido is a symbol of the impossibility of their having such a relationship; indeed, Aeneas would deny that it was iugalis at all. Consequently Dido burns it on a pyre reminiscent of the pyre on which Andromache burned Hector’s possessions, which she described as οὐδὲν σοί γ’ ὄφελος ('No use to you'; Iliad XXII 513). But through this connection Aeneas is not completely absent. Odysseus describes how he built the olive-tree bed at Odyssey xxiii 200–201: “inlaying it with gold and silver and ivory,” among other things. Dido’s first view of Aeneas after he emerges from Venus’ cloud is described in these same terms: “for his mother had shed upon her son beauty … the beauty the hand gives to ivory, or when silver is set in yellow gold” (Aeneid I 592–593). Dido’s ironic, and pitiful, invocation of the marriage bed of Odysseus thus becomes an invocation of Aeneas himself.
These four sample passages—Cretheus, Camilla, Aeneas in the underworld, and Dido’s pyre—generate their own networks of meaning, and yet the four are not unconnected; indeed, they are all part of the network of meaning around arma virum, a phrase that appears several more times in the poem. That is, all four of these examples can be seen as part of the same network of meaning, which, I am arguing, is as interpretively robust as any that has been found in Homeric poetry. Along the way I have indicated some ways I think these networks could be interpreted; as for the overarching network around arma virum, it would be interesting to see how the meaning of that phrase changes throughout the poem, as Aeneas and the Trojans fulfill their destiny.
And so I return to my original question about whether Virgil understood the nature of Homeric poetry. The connections between repeated language and themes in the Aeneid can stand up to the same scrutiny as those in Homer, and Homer’s poetry itself is a critical part of these networks. As for Virgil’s understanding, I see three possibilities. First, perhaps Virgil understood nothing, and it just looks like he did because the networks I have described are in fact a feature of all language. In this case the distinction between oral and written poetry is perhaps not so defined as some might like. The second possibility is that perhaps Virgil did understand Homer as a formulaic poet, along with all the resonances that an Archaic Greek audience would have felt. In this case (and I do not think it is too much of a stretch to claim) Virgil was the most attentive, versatile, and hard-wired audience that Homer ever had. A third possibility is that he did not understand anything about formulaic poetry, but that it did not matter: his imitation of Homer was so thorough that he unwittingly, as it were, created a piece of literature that functioned, in this respect at least, as a truly formulaic poem would.


Albrecht, E. 1881. “Wiederholte Verse und Verstheile bei Vergil.” Hermes 16:393–444.
Bloch, Alfred. 1970. “Arma Virumque als heroisches Leitmotiv.” Museum Helveticum 27:206–211.
Clay, Jenny Strauss. 1983. The Wrath of Athena: Gods and Men in the Odyssey. Princeton.
Conte, Gian Biagio. 1994. Latin Literature: A History. Trans. J. B. Solodow, D. Fowler, and G. W. Most. Baltimore.
Goold, George P. 1977. “The Nature of Homeric Composition.” Illinois Classical Studies 2:1–34.
Hainsworth, Bryan. 1993. The Iliad: A Commentary. Vol. III, Books 9–12. Cambridge.
Hardie, Philip, ed. 1994. Virgil: Aeneid IX. Cambridge.
Knauer, Georg Nicolaus. 1964. Die Aeneis und Homer; Studien zur poetischen Technik Vergils, mit Listen der Homerzitate in der Aeneis. Hypomnemata 7. Gottingen.
Kopff, E. Christian. 1981. “Virgil and the Cyclic Epics.” ANRW 2.31.2:919–947.
Levitan, William. 1993. “Give Up the Beginning?: Juno’s Mindful Wrath (Aeneid 1.37).” Liverpool Classical Monthly 18.1:14.
Lord, Albert. 2000. The Singer of Tales. Harvard Studies in Comparative Literature 24. 2nd ed., with introduction by S. Mitchell and G. Nagy. Cambridge, MA.
Miller, D. Gary. 1982. Homer and the Ionian Epic Tradition. Innsbrucker Beiträge zur Sprachwissenschaft 38. Innsbruck.
Moskalew, Walter. 1982. Formular Language and Poetic Design in the Aeneid. Mnemosyne Supplement 73. Leiden.
Nagy, Gregory. 1990a. Pindar’s Homer: The Lyric Possession of an Epic Past. Baltimore.
———. 1990b. Greek Mythology and Poetics . Ithaca.
———. 1996a. Poetry as Performance: Homer and Beyond. Cambridge.
———. 1996b. Homeric Questions. Austin.
———. 1999. The Best of the Achaeans: Concepts of the Hero in Archaic Greek Poetry . 2nd ed. with new introduction. Baltimore.
———. 2003. Homeric Responses. Austin.
Oliensis, Ellen. 2004. “Sibylline Syllables: The Intratextual Aeneid.” Proceedings of the Cambridge Philological Society 50:29–45.
Schlunk, Robin. 1974. The Homeric Scholia and the Aeneid. Ann Arbor.
Scodel, Ruth. 2004. “The Modesty of Homer.” Oral Performance and Its Context (ed. C.  J. Mackie) 1–19. Mnemosyne Supplement 248. Leiden.
West, Martin L. 1985. The Hesiodic Catalogue of Women. Oxford.
Wills, Jeffrey. 1996. Repetition in Latin Poetry: Figures of Allusion. Oxford.
Zetzel, James E. G. 1984. “Re-Creating the Canon: Augustan Poetry and the Alexandrian Past.” Canons (ed. R. von Hallberg) 107–130. Chicago.


[ back ] 1. I thank Leslie Kurke for hosting a graduate colloquium at which I presented this paper; still more thanks are due to my fellow Berkeley graduate students who attended that event and offered invaluable advice. Will Shearin and Lauri Reitzammer read and improved a much earlier version of this work. I thank Ralph Hexter for discussing with me how (i.e. with what technology) Virgil read Homer. He outlined his ideas on this subject at a lecture, still unpublished but in circulation here in Berkeley, at Stanford University some years ago. Also still relevant is Schlunk 1974.
[ back ] 2. I have not made a detailed study of Jeffrey Wills 1996 because his main interest seems to lie in repetitions within a single poetic line, as he indicates at the start of his sections on gemination (46) and on polyptoton (189n1). I am interested in phrases repeated in different sections of a single work. Still, Wills’s subtitle shows that he and I are ultimately interested in the same poetic effects. I have probably erred in neglecting for similar reasons Moskalew 1982. Both works have much to teach us, of course.
[ back ] 3. “Formulas do not have a life independent of themes” (Nagy 1996b:24). See further Miller 1982:44: “Formulas attach to and are triggered by the repetition of certain themes (arming, battles, counsel, dining, etc.).” Lord 2000:32 makes a similar point about the relationship of meter and theme: “rhythm and thought are one.”
[ back ] 4. Anyone who asserts that the repetitions of formulas are meaningful comes face to face with the Homeric Question: how can there be “intricate verbal parallelism” (as Miller puts it [1982:8]) in an oral poet? “Poets have a plan in their head, and the plan carries numerous thematic repetitions that may or may not be symmetrical, and with the individual themes go certain formulas. It is simply a misrepresentation of the oral art to think in terms of poets ‘remembering’ words or phrases over a distance of thousands of lines” (Miller 1982:8, with further bibliography). See also Nagy 1999:xi: “My reading of Homer, especially of the passages in Odyssey viii and Iliad IX, has occasionally been disputed on the grounds that it gives the impression of literary rather than oral poetics.” See also the discussion of the phrase ἄριστος Ἀχαιῶν starting at Nagy 1999:26, in which every instance of this phrase is seen as connected.
[ back ] 5. I use this phrase “network of meaning” instead of Nagy’s (and others’) “cross-reference” because the latter term implies, to me, an allusion between two particular passages. I want to emphasize the way one passage can refer to many different passages, which in turn refer to other passages, and so on.
[ back ] 6. Goold 1977 is a good example of this kind of argument. Nagy attempts to refute Goold’s (and others’) assumptions about what oral poets are and are not capable of at Nagy 1996b:26.
[ back ] 7. Nagy 1996b:19: “unity … is a result of performance tradition, not a cause effected by a composer.” See also Nagy 1999:6–7, where he argues for the audience’s role in the creation of a poetic whole using as an example the two halves of the hymn to Apollo which indicate a “social fusion of two distinct audiences.”
[ back ] 8. In his Sather Seminar, Nagy frequently employed this term “hard-wired”’ to describe the degree to which ancient audiences and performers had internalized traditional language. I do not know where, or if, he uses it in his published work.
[ back ] 9. Nagy 1999:15, where he calls tradition “the culmination of perhaps a thousand years of performer-audience interaction.”
[ back ] 10. Lord 2000:32–33 on the “unconscious” role form plays in relation to the “conscious” theme. In this passage he describes various possibilities for understanding how traditional language is received: even on the level of the sounds of words connections can be made. See also Nagy 1996b:25.
[ back ] 11. Nagy 1990a:17n2 and Nagy 1996a:1. The terms langue and parole go back to Saussure in his Course in General Linguistics and their application beyond linguistics is usually attributed to Levi-Strauss in The Structural Study of Myth.
[ back ] 12. On the terms “diachronic” and “synchronic” see Nagy 1996b:17, with further bibliography, in particular Nagy 1990b:20–21.
[ back ] 13. Nagy 1996a:108–110 describes this as the “relatively rigid” stage of Homeric transmission in comparison with earlier periods.
[ back ] 14. The vulgate text of Homer, which I described as a description of the diachronic tradition, is, of course, a synchronic artifact as well, but in the context of Nagy’s work can hardly be called the record of a particular performance of the Homeric poems.
[ back ] 15. See Clay 1983:243 for a forceful statement of this objection. Her point is that Nagy calls on a tradition that was “so conservative and fixed as to become a text”—that is, to be, for all intents and purposes, a literary, not oral, artifact.
[ back ] 16. I would like to suggest in passing that we might do well to reconsider the several “inconsistencies” in the Aeneid; I refer, for example, to the episode in Book 2 in which Aeneas sees Helen watching the sack of Troy from the roofs in apparent contradiction to Deiphobus’ claim in Book 6 that she was with him during the sack, the attribution of Celaeno’s prophecy from Book 3 to Anchises in Book 7, and Aeneas’ reference to an otherwise unknown prophecy that Palinurus would reach Italy safely in Book 6. These inconsistencies seem to me to share something with similar inconsistencies in Homer, which analytic scholars invoke against unitarian arguments, such as the problem of the duals in the embassy of Iliad IX. Critics usually explain away such Virgilian contradictions by, like analytic Homeric critics, expunging lines (as in the Helen episode) or by saying that Virgil would have corrected these had he had time to revise the poem. But I wonder if Virgil, having found such inconsistencies in the Homeric epics, might have intentionally included some of his own in the Aeneid.
[ back ] 17. One famous example of an intratextual cross-reference is Virgil’s use of the same line for the death of Camilla (Aeneid XI 831) as for the death of Turnus (XII 952). Oliensis 2004 offers an intratextual reading of the Aeneid with emphasis on the “distresses of the textual condition”—that is, the way in which violent episodes of opening in the Aeneid give us “a meditation on the susceptibility of the written text to scattering.” Clearly she and I are approaching the poem from different directions, but we do discuss many of the same passages.
[ back ] 18. See Zetzel 1984:109–111 on the blending of genres in the Aeneid. Also Kopff 1981:938.
[ back ] 19. The standard, monumental, study is Knauer 1964; I have relied extensively on Knauer’s lists of Virgilian citations of Homer. I have also made use of Albrecht 1881:393–444, which lists the repeated lines in Virgil.
[ back ] 20. Kopff 1981:928 rejects the idea that Virgil ranks himself as a cyclic author, but Conte 1994:277 highlights the way the Aeneid can be seen as a continuation of the Homeric narrative.
[ back ] 21. Part of the appeal of Nagy’s approach to formulaic language is that because he views the Homeric corpus as a complete system, any line works as well as any other as an example of how formulaic language works. And so when I started this project I let the sortes Vergilianae determine which Virgilian passage I would use to study Virgil’s treatment of Homeric repetitions—that is, I closed my eyes, opened my OCT of Virgil, and let my finger fall on a random passage. The gods were smiling on me that day, because my finger fell on the death of Cretheus with its striking repetition of arma virum. I mention this because one could object that, as the programmatic phrase of the whole poem, arma virumque forms something of a special case in Virgilian language and that for this reason the networks of meaning around arma virum should not be taken as representative of the networks formed by other phrases. My response is that, by the same token, arma virumque is the perfect phrase for this study because its engagement with the Homeric poems is so deep—after all, my purpose is to understand how Virgil represented the kinds of repetitions he found in Homer in his own work, and what role Homeric language itself plays in that representation.
[ back ] 22. I would have thought that this phrase would have attracted more interest, especially from New Critics, but Hardie 1994 on Aeneid IX 57 cites only Bloch 1970:206–211 and Norden’s commentary on Aeneid VI. More recently Oliensis 2004:31 treats the Cretheus episode.
[ back ] 23. Corinne Crawford has pointed out to me that, pace the orthodox interpretation, arma is not really a very good metonymy for the Iliad since it leaves out so many important features of the Homeric poem. And Levitan 1993:14 shows that the first word of the Iliad, μῆνιν, is in fact to be found in the Aeneid, in the elided opening of Juno’s first speech at Aen. I 37: men(e) incepto …
[ back ] 24. The word cano carries significance as well, for while the Homeric proems ask the Muse or Muses to tell their respective stories, Virgil says “I sing.” This is usually taken as an allusion to the opening line of the cyclic Little Iliad: Ἴλιον ἀείδω καὶ Δαρδανίην ἐύπωλον, / ἧς πέρι πόλλ’ ἔπαθον Δαναοί, θεράποντες Ἄρηος (“I sing of Ilios, and Dardania with its beautiful colts, over which the Danaans, the servants of Ares, suffered many things”; Little Iliad fr. 1 Allen). Space does not permit a discussion of the relationship between the Aeneid and the Trojan epic cycle using the parole and langue model I have proposed here, but see Kopff 1981 for a survey of the parallels. For now it is enough to say that Virgil is further highlighting the completeness of his imitation of Homer—not just the events of the Iliad and the Odyssey but the whole tradition of epic poetry dealing with the Trojan War. For a recent treatment of the relationship between the Homeric epics and the epic cycle, see Scodel 2004, an article that offers many tempting angles from which to approach Virgil’s attitude toward the Trojan Cycle. Hainsworth 1993 on Iliad 524–605, Phoenix’s telling of the Meleager myth, points out that the Homeric epics frequently show “knowledge” of the other epic cycles (Theban, Aetolian/Elean, Iolkos). These are described by West 1985:137. It would be interesting to determine what knowledge Virgil had of these cycles and how he employed it.
[ back ] 25. Nagy 1999:42–58 discusses this passage at some length.
[ back ] 26. The closest parallel is the simile in Iliad XV that compares Ajax jumping from ship to ship to a man who can jump from horse to horse while they run (Iliad XV 679). But Ajax does not actually jump from horse to horse—he is only capable of this metaphorically.