Poetics of Repetition in Homer

Gregory Nagy
[[This article was originally published in 2004 in Greek Ritual Poetics (ed. D. Yatromanolakis and P. Roilos) 139–148. Hellenic Studies 3. Cambridge, MA and Washington, DC. In this online version, the original page-numbers of the printed version are indicated within braces (“{” and “}”). For example, “{139|140}” indicates where p. 139 of the printed version ends and p. 140 begins.]]
Repetition in Homeric poetry is a matter of performance, not only composition. I argue that this observation applies to the Homeric phenomenon of “repeated utterances.”
This argument is part of a larger project, which is, to show that Homeric poetry is a medium of oral poetry. That project is exemplified in a 1996 book entitled Poetry as Performance, where I argued that the text of the Homeric poems, that is, of the Iliad and the Odyssey, is derived from a system that needs to be distinguished from the text itself - or, better, from the Homeric textual tradition that editors try to reconstruct as a single Homeric text. [1] The text of Homer is merely the surface. Underneath that surface is the system that is Homeric poetry, which needs to be analyzed on its own terms, as a system. This system, I argued, reveals a medium of oral poetry. My task here is to take the argument further by showing that the phenomenon of repeated utterances in the Homeric textual tradition is explainable in terms of this system, that is, in terms of oral poetry.
Essential for my argument is the idea that Homeric composition comes to life in performance. In oral poetry, performance is essential to composition: the composition of oral poetry is a matter of recomposition-in-performance. [2] If it is true that the Homeric text derives from oral poetry, then the phenomenon of repeated utterances is itself a matter of performance, not only composition. [3]
In some cases of repeated utterances, it is obvious that the repetition is a matter of performance, as in Homeric descriptions of ensembles of mourning women whose laments echo the lament of the woman who initiates the mourning. [4] In other cases, however, the dimension of performance is less {139|140} obvious. Least obvious to the modern mind is the idea that the actual “quoting” of a speaker is itself a matter of performance in oral poetics. A most convincing demonstration of this idea is offered by Richard Martin in his 1989 book, The Language of Heroes. [5] From the standpoint of oral poetry, the “quoting” of a speaker’s utterance is not just a repetition of that utterance: it is also a performance - a reperformance - of that utterance. A repeated utterance, then, is a reperformance of a notionally original utterance. The same goes for any further repetition: it is a matter of further reperformance.
Here I find it most relevant to apply the intuitive formulation about repetition by Kierkegaard in his Repetition (1843): “The dialectic of repetition is easy, for that which is repeated has been - otherwise it could not be repeated - but the very fact that it has been makes the repetition into something new.” [6]
With this formulation in mind, I now proceed to analyze examples of repeated utterances in Homeric poetry. These examples, I argue, show that repetition in Homeric poetry is not just composed but also performed.
These same examples have already been analyzed in a work by Tormod Eide, who views the phenomenon of repeated utterances in terms of “reformulated repetitions.” [7] Although I disagree with Eide’s conclusions, I find his analysis a most enlightening point of departure for my own analysis (references to his work will be marked shorthand by way of his initials, “TE”).
TE summarizes his conclusions this way (p. 97): “strained grammar, narrative inconsistency, and anomalous word forms can sometimes be ascribed to the rephrasing of preexisting lines.” He makes a distinction between “reformulated repetition” and “formulaic repetition” (TE 98). He thinks that “reformulated repetitions” do not belong to “the poet’s “gemeinsames Versgut,” which he equates with oral poetry and which is characterized by “formulaic repetition.” I argue, by contrast, that the phenomena he describes can indeed be explained in terms of oral poetics.
According to TE (p. 97) there are two categories of reformulated repetitions: (1) reported message and (2) the carrying out of a command, request, wish. To facilitate the discussion, I substitute for “carrying out” the expression “following through,” as a paraphrase for the Homeric expression tetelesmenon estai, in the sense of ‘it [= whatever has been said formally] will reach its due telos’. [8] {140|141}
TE elaborates (p. 97): “In both categories the repeated version entails changes primarily of the verbs and pronouns, changes which would often disturb the metrical pattern.” I draw attention to the use of the word “version” in this formulation. A problem here, I suggest, is that there is no way to decide which of any two or more “versions” is “the repeated version.” In terms of any formulaic system reflected by oral poetics, we cannot assume that any given “version” is the basic “version.” TE’s wording “changes which would often disturb the metrical pattern” invites questions. It is therefore most helpful that TE (p. 97) poses his own questions: “What does the poet do then, and how well does he do it? How often does the reformulation lead to grammatical irregularities? How often does the poet simply drop recalcitrant lines, and how does this affect narrative consistency? Does the poet’s treatment sometimes indicate recourse to writing?” TE’s own answer to the last question, as underlined, is “yes.”
I propose that this line of argumentation needs to be confronted with two observations about oral poetry: (1) formulaic language is a system and (2) this system generates forms for the present, for the here-and-now of composition.
Much depends on how oral poetry is understood in the first place. Let us take for example this formulation (TE pp. 97-98): “If the poet in the passages I discuss commits what to a modern reader would count as a blemish, it is not necessarily from lack of ability, but because he did not care, and did not need to care. The Homeric poems were not created for readers or scholars. Nor should we forget that the listeners’ thrill of hearing an impressive or dramatic passage repeated word for word would outweigh minor discrepancies introduced by the repetition.”
Such a formulation underestimates the potential artistry inherent in oral poetry. From the standpoint of oral poetics, the “thrill” is not only in hearing a notionally word-for-word repetition of what has just been heard but also in apprehending the changes that are being made within this repetition. For example, there are changes required by the shifting of first / second / third persons in pronouns and verbs as the composer recomposes. [9] The artistic bravura is not just a matter of recomposition: it is a matter of recomposition-in-performance. An aspect of this bravura is the skill of shifting from one person to another in performing a set of repeated sequences requiring the deployment of different persons in pronouns or verbs. [10] {141|142}
TE’s first test case is Iliad 3.67-75 ~ 86-94 ~ 136-138 ~ 252-258 (p. 99): “Paris suggests to Hector that the war be ended by a duel between Paris and Menelaus [passage A = 3.67-75]; Hector then passes on the suggestion to the men [passage B = 3.86-94]; the news is brought to Helen by Iris [passage C = 3.136-138], and, finally, the herald Idaeus summons Priam to descend on to the plain and conclude the pact [passage D = 3.252-258].”
A revealing aspect of TE’s analysis is where he says (p. 99): “Hector’s speech ends abruptly [passage B, after line 94], with no indication as to the future course of action that is to be sanctioned by the oath-taking, corresponding to the ending of Paris’ speech [passage A, 74-75]. Kirk’s explanation, that Paris’ last two verses may have been dropped ‘because Hektor does not feel it necessary or diplomatic to elaborate on terms of departure at this stage’[,] may seem too refined.” [11] I suggest that there are problems even in saying that the last two verses of passage A were “dropped” in passage B. It is as if we were dealing with some pre-existing text that became a new text in the process of repetition. I propose that the renovation is a matter of recomposed performance, not rewritten text. TE continues (p. 99): “Two manuscripts and a papyrus omit also v. 94 [in other words, Hector’s speech now ends after verse 93], and thus give a more natural ending to the speech. One may suspect that this was the invention of the original composer, and that the line [verse 94] was inserted later by a singer / rhapsode who noticed that one more of Paris’ lines could easily, from a metrical point of view, be included in Hector’s speech. Sometimes it is metrical adaptability rather than appropriateness that guides the composer in deciding what to include, what to leave out in the repetition” (underlines mine). I find it difficult to accept the distinction made here between “the original composer,” who judiciously cuts the speech after verse 93, and the less judicious later “composer” who may decide to keep on going with the speech merely because the meter makes it easy for him to keep going. I draw attention to the assumption that the speech as conceived in passage A must be the basic speech.
After entertaining an alternative possibility, that “the line [verse 93] was omitted by somebody who saw that the line as it stands was unsatisfactory without a continuation” (p. 99), TE proceeds to examine passage C, where we see one single verse conveying the same general idea that we see conveyed by two verses in passages A and B. TE adds, and I agree with him, that this single verse in passage {142|143} C “is more suitable to a private conversation between two women” than the two verses in the corresponding passages: “the mention of ktêmata panta is dropped, and the neutral gunaika is replaced by the more affectionate philê akoitis” (TE p. 100). I disagree, however, with the textual implications of TE’s analysis. I maintain that the adjustments required by different situations occasioned by repetition can be counted as examples of virtuosity in the art of recomposition-in-performance.
But what are we to do with the κε of passage C verse 138? Is it a lapse in artistry? This form is syntactically anomalous, in comparison with the corresponding κε of passage A verse 71 and of passage B verse 92. TE offers this explanation (p. 100): “We see that the constraints laid upon the poet when composing by this kind of recycling process may overrule both Greek grammar and narrative coherence.” My work in formulaic analysis leads me to a different interpretation: “this kind of recycling process” is not restricted to “reformulated repetitions” and in fact it pervades the entire formulaic system of Homeric poetry. In general, I hold that the formulaic system of Homeric poetry has a grammar of its own - with rules as well as licenses that differ in part from those of everyday language.
In passage D, the formal problems of adapting the “same” wording that we saw in passages A, B, and C seem to be compounded. Particularly problematic for TE is the case of “a line with no change in wording, but with a different syntactical relation to the preceding [line] in order to achieve the join.” For the first time, TE overtly applies to such a phenomenon the description “non-formulaic” (p. 101). We see here an explicit statement of what has in fact been the view of TE all along, that “reformulated repetitions” are symptomatic of non-formulaic composition. I see no justification for such a view.
Rather, I propose that the process of reformulation is in fact a reaffirmation of formulaic composition. In order to reaffirm or even clarify the meaning of something that has been “quoted,” the “quoted” words may need to be quoted again, that is, performed again. I am using the word “quote” not in a textual but in a performative sense. Once a given formulation has been performed as a speech act, that is, once it has been “quoted” in performance, it becomes eligible to be “quoted” anew.
The idea of performance as a speech act is especially relevant to cases where we can identify a ritual as the overt referent in a set of reformulated repetitions. One such case is Iliad 6. 87-101 ~ 269-278 ~ 297-310. Let us begin with the first in this set composed of three passages: in 6. 87-101, Helenus, who is speaking here as a seer, “instructs his brother Hector to withdraw to Troy and beg their mother to gather the Trojan women in Athena’s sanctuary to make offerings to the goddess in the hope that she will have pity on them” (TE p. 103): {143|144}
A Iliad 6. 87-101
Ἕκτορ ἀτὰρ σὺ πόλιν δὲ μετέρχεο, εἰπὲ δ' ἔπειτα
μητέρι σῇ καὶ ἐμῇ· ἣ δὲ ξυνάγουσα γεραιὰς
νηὸν Ἀθηναίης γλαυκώπιδος ἐν πόλει ἄκρῃ
οἴξασα κληῖδι θύρας ἱεροῖο δόμοιο
90 πέπλον, ὅς οἱ δοκέει χαριέστατος ἠδὲ μέγιστος
εἶναι ἐνὶ μεγάρῳ καί οἱ πολὺ φίλτατος αὐτῇ,
θεῖναι Ἀθηναίης ἐπὶ γούνασιν ἠϋκόμοιο,
καί οἱ ὑποσχέσθαι δυοκαίδεκα βοῦς ἐνὶ νηῷ
ἤνις ἠκέστας ἱερευσέμεν, αἴ κ' ἐλεήσῃ
95 ἄστύ τε καὶ Τρώων ἀλόχους καὶ νήπια τέκνα,
ὥς κεν Τυδέος υἱὸν ἀπόσχῃ Ἰλίου ἱρῆς
ἄγριον αἰχμητὴν κρατερὸν μήστωρα φόβοιο,
ὃν δὴ ἐγὼ κάρτιστον Ἀχαιῶν φημι γενέσθαι.
οὐδ' Ἀχιλῆά ποθ' ὧδέ γ' ἐδείδιμεν ὄρχαμον ἀνδρῶν,
100 ὅν πέρ φασι θεᾶς ἐξέμμεναι· ἀλλ' ὅδε λίην
μαίνεται, οὐδέ τίς οἱ δύναται μένος ἰσοφαρίζειν.
About 200 verses later, Hector conveys this mantic formulation to his mother Hecabe:
B Iliad 6.269-278
ἀλλὰ σὺ μὲν πρὸς νηὸν Ἀθηναίης ἀγελείης
270 ἔρχεο σὺν θυέεσσιν ἀολλίσσασα γεραιάς·
πέπλον δ', ὅς τίς τοι χαριέστατος ἠδὲ μέγιστος
ἔστιν ἐνὶ μεγάρῳ καί τοι πολὺ φίλτατος αὐτῇ,
τὸν θὲς Ἀθηναίης ἐπὶ γούνασιν ἠϋκόμοιο,
καί οἱ ὑποσχέσθαι δυοκαίδεκα βοῦς ἐνὶ νηῷ
275 ἤνις ἠκέστας ἱερευσέμεν, αἴ κ' ἐλεήσῃ
ἄστύ τε καὶ Τρώων ἀλόχους καὶ νήπια τέκνα,
αἴ κεν Τυδέος υἱὸν ἀπόσχῃ Ἰλίου ἱρῆς
ἄγριον αἰχμητὴν κρατερὸν μήστωρα φόβοιο.
Hecabe follows the instructions of A and B by proceeding to fetch a peplos from the store-room. Then the instructions are followed further, in the form of a ritual performed on the acropolis of Troy:
C Iliad 6.297-310
Αἳ δ' ὅτε νηὸν ἵκανον Ἀθήνης ἐν πόλει ἄκρῃ,
τῇσι θύρας ϊξε Θεανὼ καλλιπάρῃος
Κισσηῒς ἄλοχος Ἀντήνορος ἱπποδάμοιο· {144|145}
300 τὴν γὰρ Τρῶες ἔθηκαν Ἀθηναίης ἱέρειαν.
αἳ δ' ὀλολυγῇ πᾶσαι Ἀθήνῃ χεῖρας ἀνέσχον·
ἣ δ' ἄρα πέπλον ἑλοῦσα Θεανὼ καλλιπάρῃος
θῆκεν Ἀθηναίης ἐπὶ γούνασιν ἠϋκόμοιο,
εὐχομένη δ' ἠρᾶτο Διὸς κούρῃ μεγάλοιο·
305 πότνι' Ἀθηναίη ἐρυσίπτολι δῖα θεάων
ἆξον δὴ ἔγχος Διομήδεος, ἠδὲ καὶ αὐτὸν
πρηνέα δὸς πεσέειν Σκαιῶν προπάροιθε πυλάων,
ὄφρά τοι αὐτίκα νῦν δυοκαίδεκα βοῦς ἐνὶ νηῷ
ἤνις ἠκέστας ἱερεύσομεν, αἴ κ' ἐλεήσῃς
310 ἄστύ τε καὶ Τρώων ἀλόχους καὶ νήπια τέκνα.
TE (p. 104) argues that there are two important inconsistencies: “(a) in A and B Hecabe is requested to promise Athena a sacrifice of twelve cows, while in C the sacrifice will take place autika nun (308); (b) in A and B Hecabe herself is to give Athena the peplos and utter the prayer, while in C it is the priestess Theano who performs both services. ... Instead of asking, ‘Why does the poet change his mind?’ we ask, ‘Which of the three texts bears the mark of an adaptation?’ we will immediately see that it is A that has excited commentators.” In this case, TE (p. 105) supposes the poet “to be composing backwards,” using C as the basis for the variations in B and A. From the standpoint of oral poetics, however, A B C are all variants, and all three have to be treated as multiforms. [12] There is a crescendo effect here, as the ritual details become ever more explicit.
In this case, the execution of the formulation results in failure for the ritual, in that the goddess refuses the offering. I submit that the crescendo in detail provides an explanation for the failure: Hecuba had in the end chosen as her offering to the goddess something directly connected to the cause of the Trojan War: the peplos that she chose is connected to the story of Helen’s abduction by Paris. Thus the repetition of a formulation has in this case helped motivate the outcome of that formulation in the plot of the composition.
There are other instances as well where the plot of the composition is clarified by the poetics of repeated formulation. Let us consider Iliad 9.119-161 ~ 260-301, about which TE (p. 115) observes: “In the beginning of Book 9 Agamemnon describes ... the gifts he is prepared to offer Achilles to make him surrender his anger. Odysseus, in his speech to Achilles, follows Agamemnon’s words closely (264-99), and thus we have the opportunity of studying an unusually long and [illustrative] example of the reported message.” {145|146}
A Iliad 9.119-161 B Iliad 9.260-301
ἀλλ' ἐπεὶ ἀασάμην φρεσὶ λευγαλέῃσι πιθήσας,
120 ἂψ ἐθέλω ἀρέσαι δόμεναί τ' ἀπερείσι' ἄποινα.
ὑμῖν δ' ἐν πάντεσσι περικλυτὰ δῶρ' ὀνομήνω

ἕπτ' ἀπύρους τρίποδας, δέκα δὲ χρυσοῖο τάλαντα,
αἴθωνας δὲ λέβητας ἐείκοσι, δώδεκα δ' ἵππους
πηγοὺς ἀθλοφόρους, οἳ ἀέθλια ποσσὶν ἄροντο.
125 οὔ κεν ἀλήϊος εἴη ἀνὴρ ᾧ τόσσα γένοιτο,
οὐδέ κεν ἀκτήμων ἐριτίμοιο χρυσοῖο,
ὅσσά μοι ἠνείκαντο ἀέθλια μώνυχες ἵπποι.
δώσω δ' ἑπτὰ γυναῖκας ἀμύμονα ἔργα ἰδυίας
Λεσβίδας, ἃς ὅτε Λέσβον ἐϋκτιμένην ἕλεν αὐτὸς
130 ἐξελόμην, αἳ κάλλει ἐνίκων φῦλα γυναικῶν.
τὰς μέν οἱ δώσω, μετὰ δ' ἔσσεται ἣν τότ' ἀπηύρων
κούρη Βρισῆος· ἐπὶ δὲ μέγαν ὅρκον ὀμοῦμαι
μή ποτε τῆς εὐνῆς ἐπιβήμεναι ἠδὲ μιγῆναι,
ἣ θέμις ἀνθρώπων πέλει ἀνδρῶν ἠδὲ γυναικῶν.
135 ταῦτα μὲν αὐτίκα πάντα παρέσσεται· εἰ δέ κεν αὖτε
ἄστυ μέγα Πριάμοιο θεοὶ δώωσ' ἀλαπάξαι,
νῆα ἅλις χρυσοῦ καὶ χαλκοῦ νηησάσθω

Τρωϊάδας δὲ γυναῖκας ἐείκοσιν αὐτὸς ἑλέσθω,
140 αἴ κε μετ' Ἀργείην Ἑλένην κάλλισται ἔωσιν.
εἰ δέ κεν Ἄργος ἱκοίμεθ' Ἀχαιϊκὸν οὖθαρ ἀρούρης
γαμβρός κέν μοι ἔοι· τίσω δέ μιν ἶσον Ὀρέστῃ,
ὅς μοι τηλύγετος τρέφεται θαλίῃ ἔνι πολλῇ.
τρεῖς δέ μοί εἰσι θύγατρες ἐνὶ μεγάρῳ εὐπήκτῳ
145Χρυσόθεμις καὶ Λαοδίκη καὶ Ἰφιάνασσα,
τάων ἥν κ' ἐθέλῃσι φίλην ἀνάεδνον ἀγέσθω
πρὸς οἶκον Πηλῆος· ἐγὼ δ' ἐπὶ μείλια δώσω
πολλὰ μάλ', ὅσσ' οὔ πώ τις ἑῇ ἐπέδωκε θυγατρί·
ἑπτὰ δέ οἱ δώσω εὖ ναιόμενα πτολίεθρα
150 Καρδαμύλην Ἐνόπην τε καὶ Ἱρὴν ποιήεσσαν
Φηράς τε ζαθέας ἠδ' Ἄνθειαν βαθύλειμον
καλήν τ' Αἴπειαν καὶ Πήδασον ἀμπελόεσσαν.
πᾶσαι δ' ἐγγὺς ἁλός, νέαται Πύλου ἠμαθόεντος·
ἐν δ' ἄνδρες ναίουσι πολύρρηνες πολυβοῦται,
155 οἵ κέ ἑ δωτίνῃσι θεὸν ὣς τιμήσουσι
καί οἱ ὑπὸ σκήπτρῳ λιπαρὰς τελέουσι θέμιστας.
ταῦτά κέ οἱ τελέσαιμι μεταλήξαντι χόλοιο.
δμηθήτω· Ἀΐδης τοι ἀμείλιχος ἠδ' ἀδάμαστος,
τοὔνεκα καί τε βροτοῖσι θεῶν ἔχθιστος ἁπάντων·

160 καί μοι ὑποστήτω ὅσσον βασιλεύτερός εἰμι
ἠδ' ὅσσον γενεῇ προγενέστερος εὔχομαι εἶναι.
260 ... ἔα δὲ χόλον θυμαλγέα· σοὶ δ' Ἀγαμέμνων
ἄξια δῶρα δίδωσι μεταλήξαντι χόλοιο.
εἰ δὲ σὺ μέν μευ ἄκουσον, ἐγὼ δέ κέ τοι καταλέξω
ὅσσά τοι ἐν κλισίῃσιν ὑπέσχετο δῶρ' Ἀγαμέμνων·
ἕπτ' ἀπύρους τρίποδας, δέκα δὲ χρυσοῖο τάλαντα,
265 αἴθωνας δὲ λέβητας ἐείκοσι, δώδεκα δ' ἵππους
πηγοὺς ἀθλοφόρους, οἳ ἀέθλια ποσσὶν ἄροντο.
οὔ κεν ἀλήϊος εἴη ἀνὴρ ᾧ τόσσα γένοιτο
οὐδέ κεν ἀκτήμων ἐριτίμοιο χρυσοῖο,
ὅσσ' Ἀγαμέμνονος ἵπποι ἀέθλια ποσσὶν ἄροντο.
270 δώσει δ' ἑπτὰ γυναῖκας ἀμύμονα ἔργα ἰδυίας
Λεσβίδας, ἃς ὅτε Λέσβον ἐϋκτιμένην ἕλες αὐτὸς
ἐξέλεθ', αἳ τότε κάλλει ἐνίκων φῦλα γυναικῶν.
τὰς μέν τοι δώσει, μετὰ δ' ἔσσεται ἣν τότ' ἀπηύρα
κούρη Βρισῆος· ἐπὶ δὲ μέγαν ὅρκον ὀμεῖται
275 μή ποτε τῆς εὐνῆς ἐπιβήμεναι ἠδὲ μιγῆναι
ἣ θέμις ἐστὶν ἄναξ ἤτ' ἀνδρῶν ἤτε γυναικῶν.
ταῦτα μὲν αὐτίκα πάντα παρέσσεται· εἰ δέ κεν αὖτε
ἄστυ μέγα Πριάμοιο θεοὶ δώωσ' ἀλαπάξαι,
νῆα ἅλις χρυσοῦ καὶ χαλκοῦ νηήσασθαι
280 εἰσελθών, ὅτε κεν δατεώμεθα ληΐδ' Ἀχαιοί,
Τρωϊάδας δὲ γυναῖκας ἐείκοσιν αὐτὸς ἑλέσθαι,
αἵ κε μετ' Ἀργείην Ἑλένην κάλλισται ἔωσιν.
εἰ δέ κεν Ἄργος ἱκοίμεθ' Ἀχαιϊκὸν οὖθαρ ἀρούρης
γαμβρός κέν οἱ ἔοις· τίσει δέ σε ἶσον Ὀρέστῃ,
285 ὅς οἱ τηλύγετος τρέφεται θαλίῃ ἔνι πολλῇ.
τρεῖς δέ οἵ εἰσι θύγατρες ἐνὶ μεγάρῳ εὐπήκτῳ
Χρυσόθεμις καὶ Λαοδίκη καὶ Ἰφιάνασσα,
τάων ἥν κ' ἐθέλῃσθα φίλην ἀνάεδνον ἄγεσθαι
πρὸς οἶκον Πηλῆος· ὃ δ' αὖτ' ἐπὶ μείλια δώσει
290 πολλὰ μάλ', ὅσσ' οὔ πώ τις ἑῇ ἐπέδωκε θυγατρί·
ἑπτὰ δέ τοι δώσει εὖ ναιόμενα πτολίεθρα
Καρδαμύλην Ἐνόπην τε καὶ Ἱρὴν ποιήεσσαν
Φηράς τε ζαθέας ἠδ' Ἄνθειαν βαθύλειμον
καλήν τ' Αἴπειαν καὶ Πήδασον ἀμπελόεσσαν.
295 πᾶσαι δ' ἐγγὺς ἁλός, νέαται Πύλου ἠμαθόεντος·
ἐν δ' ἄνδρες ναίουσι πολύρρηνες πολυβοῦται,
οἵ κέ σε δωτίνῃσι θεὸν ὣς τιμήσουσι
καί τοι ὑπὸ σκήπτρῳ λιπαρὰς τελέουσι θέμιστας.
ταῦτά κέ τοι τελέσειε μεταλήξαντι χόλοιο.
300 εἰ δέ τοι Ἀτρεΐδης μὲν ἀπήχθετο κηρόθι μᾶλλον
αὐτὸς καὶ τοῦ δῶρα, σὺ δ' ἄλλους περ Παναχαιοὺς
τειρομένους ἐλέαιρε κατὰ στρατόν, οἵ σε θεὸν ὣς
τίσουσ'· ἦ γάρ κέ σφι μάλα μέγα κῦδος ἄροιο·
νῦν γάρ χ' Ἕκτορ' ἕλοις, ἐπεὶ ἂν μάλα τοι σχεδὸν ἔλθοι

305 λύσσαν ἔχων ὀλοήν, ἐπεὶ οὔ τινά φησιν ὁμοῖον
οἷ ἔμεναι Δαναῶν οὓς ἐνθάδε νῆες ἔνεικαν.
TE (p. 115) continues: “We see ... that the habit of reproducing as faithfully as possible (or necessary) the original [sic] speech, creates a strait-jacket that leads to some infelicities or inaccuracies. The infinitives nêêsasthai (279), helesthai (281), and agesthai (288), prompted by the (metrically equivalent) imperatives nêêsasthô, helesthô, and agesthô, are left with nothing to govern them. [...] The future indicatives dôsei, t[e]isei, for Agamemnon’s dôsô, t[e]isô, also strike a false note, as the third person form does not convey the notion of intention or will that the first person does. In the mouth of Odysseus these sentences become prophecies rather than promises.” But that is precisely the point. Ex post facto, from the standpoint of the rituals associated with the cults of later centuries, these sentences are indeed prophecies about Achilles as a cult hero. [13]
More important, the variation of 158-161 ~ 300-306 in Book 9 of the Iliad is essential to the master plot of the whole composition. In this case, all depends on the success or failure of Odysseus’ reformulated repetition of Agamemnon’s earlier formulation. If Odysseus’ reformulation had been successful, he would have persuaded Achilles to accept the offer of Agamemnon, but such a success would have resulted in the failure of Achilles’ own epic. [14] Odysseus’ reformulation failed, in that he failed to persuade Achilles. This way, the epic of Achilles could succeed. Achilles’ refusal to accept Agamemnon’s offer will cause his perpetual sorrow over the loss of his friend, Patroklos, but it also leads to the success of his own epic: he will achieve an epic fame that is perpetual, like his sorrow. [15] Achilles speaks about this fame already in Iliad 9.413, repeating a prophecy he says his mother had told him, but this prophecy cannot become a reality until the full story is told, that is, until the whole Iliad is performed. Such a performance already presupposes a repetition - in fact, an eternal series of repetitions. The words of Kierkegaard’s Repetition can be applied anew: “... repetition and recollection are the same movement, except in opposite directions, for what is recollected has been, is repeated backward, whereas genuine repetition is recollected forward.” [16]


[ back ] 1. G. Nagy, Poetry as Performance: Homer and Beyond (Cambridge 1996).
[ back ] 2. Nagy, Poetry as Performance, chapters 1-3.
[ back ] 3. This observation about Homeric poetry applies also to Hesiodic poetry and beyond. In a future project, I hope to undertake such an application.
[ back ] 4. I draw special attention to the repetition marked by the Homeric expression epi de stenakhonto gunaikes ‘and the women lamented in response’ (Iliad 18.315; 22.429, 515; 24.722, 746). See also in general M. Alexiou, The Ritual Lament in Greek Tradition (first ed. Cambridge 1974; second ed., revised by P. Roilos and D. Yatromanolakis, Lanham, Maryland 2002) 64, 84-5, 90, 93-4, 96-8, 135-6, 156-9, 166-75, 168-9, 180.
[ back ] 5. R. P. Martin, The Language of Heroes: Speech and Performance in the “Iliad” (Ithaca 1989). See also E. J. Bakker, Poetry in Speech: Orality and Homeric Discourse (Ithaca 1997).
[ back ] 6. S. Kierkegaard, Fear and Trembling, Repetition (trans. H. V. Hong and E. H. Hong, Princeton 1983) 149. For an application of Kierkegaard’s formulation to oral poetics, see Nagy, Poetry as Performance p. 52 (cf. also pp. 100-2).
[ back ] 7. T. Eide, “Reformulated Repetitions in Homer,” Symbolae Osloenses 74 (1999) 97-139.
[ back ] 8. See for example Iliad 2.257, in the ritualized context of a prophecy uttered by the seer Calchas.
[ back ] 9. On the linguistic term shifter as applied to “shifting” from one person to another in the three persons of pronouns and verbs, see R. Jakobson, “Shifters, Verbal Categories and the Russian Verb,” Selected Writings II (The Hague 1971) 130-147.
[ back ] 10. For a particularly valuable collection of examples, see S. H. Blackburn, P. J. Claus, J. B. Flueckiger, and S. S. Wadley, eds., Oral Epics in India (Berkeley and Los Angeles 1989); see especially P. J. Claus, “Behind the Text: Performance and Ideology in a Tulu Oral Tradition,” pp. 55-74. At p. 74, Claus examines “a transition from a story about a spirit, to one told to a spirit, to one told by a spirit,” and he adds: “Accompanying these transitions are shifts in verbal style: from the third person pronominal referent, to the second, to the first. There are also changes in the behavior of the performers and the audience.” Another kind of performative bravura in shifting involves the movement from nominative to vocative. In a forthcoming study, I will examine the phenomenon of narratives addressed to the second person agent of the narration, as in the case of Patroklos and Eumaios in the Iliad and Odyssey respectively. See also p. xxii in my essay “The Sign of the Hero: A Prologue,” in J. K. Berenson Maclean and E. B. Aitken, eds., Flavius Philostratus, Heroikos (Atlanta 2001) xv-xxxv.
[ back ] 11. For the Iliad commentary used by TE, see G. S. Kirk, ed., The Iliad: A Commentary I. Books 1-4 (Cambridge 1985).
[ back ] 12. For more on the terms “variant” and “multiform,” see G. Nagy, “Homeric Poetry and Problems of Multiformity: The ‘Panathenaic Bottleneck’.” Classical Philology 96 (2001) 109-119.
[ back ] 13. Cf. D. L. Page, History and the Homeric Iliad (Berkeley and Los Angeles 1959) 133 and 166.
[ back ] 14. G. Nagy, Homeric Questions (Austin 1996) 142-143.
[ back ] 15. Ibid.
[ back ] 16. Kierkegaard (see n. 8) 131.