Chapter 1. Homeric Responses [1]

In Odyssey 8.72-83, the first song of Demodokos, we see a link between the oracular clairvoyance of Apollo and the poetic composition of Homer. Such a link, where the god’s prophecy is equated with the plot of the poet’s narrative, is relevant to the word “responses” in my title, which is meant to capture the meaning of the ancient Greek word hupokrinesthai as we find it in the language of Homer. This word means more than simply “respond to a question.” It conveys the basic idea of responding by way of performing, and this idea links Homeric poetry with the clairvoyant language of seers. Such language has its own poetics, which I propose to call “mantic” or “oracular” poetry. [2]
I begin with a seemingly straightforward Homeric context. In Iliad 7, we hear a heroic speaker, Diomedes, making a speech in response to a preceding speech, a message that was delivered by Idaios, herald of the Trojans, to the Achaeans. This preceding message of the Trojans is “quoted” by the Homeric narrative (7.385-397), and so too is the corresponding message that Diomedes speaks in response, on behalf of all the Achaeans (7.400-402). Then the leader of the Achaeans, Agamemnon, is also “quoted” by the narrative: in his speech, he refers to the response of the Achaeans, as spoken by Diomedes: as Agamemnon says, the Achaeans are ‘responding’, hupokrinontai, to a message delivered from the Trojans (7.407). The Achaeans respond, hupokrinontai (7.407), by virtue of the fact that the words of their response have been performed by Diomedes (7.400-402). [3] Further, these words have been {21|22} performed by virtue of the fact that they are being “quoted” by Homeric performance. [4]
Each time the Iliad was performed, over the many centuries of Homeric performance traditions, let us imagine how a given audience would have heard these words of the Achaeans’ response, such as we have them in Iliad 7.400-402. For each new audience, it is notionally the exact same response all over again, because the original words of the original hero are supposedly being “quoted.” These words of response would be imagined to be the “same” each time, I argue, not because they are written down – and to that extent my use of the word “quoted” is imprecise – but simply because they are performed. In terms of performance, the original words of a Homeric response are being spoken by the original speaker – in this case, by the hero Diomedes. The mentality of “quotation” is basically performative. What matters ultimately, as we will see, is the actual performativity or “quotability” of hupokrinesthai in its Homeric contexts. Having made clear that I use “quotation” in a performative sense, I will hereafter omit quotation marks when I use the word with reference to Homeric contexts.
The notional “sameness” of a Homeric response on each occasion when Homeric poetry is being performed is part of an overall mentality of unchangeability in Homeric performance itself. Such a mentality, I am arguing, is revealed by the Homeric contexts of hupokrinesthai. In other words, Homeric poetry presents itself as the same thing each time it is performed, just as the words of heroes (and gods) that are quoted by the poetry are imagined to be the exact same words on each occasion of each new performance.
I hasten to add a clarification: such a built-in mentality, operating on the idea that every successive performance of Homeric poetry is supposedly the same, cannot be taken literally to mean that this poetry had in fact achieved some kind of permanent fixity. A similar mentality can be found in many different kinds of oral traditions in general. Claims of “sameness” by a given oral tradition, where each performance is notionally the same thing as each previous performance, are easily contradicted by empirical observation. The evidence collected from a wide variety of living oral traditions makes it clear that {22|23} the process of recomposition-in-performance results in various degrees of change during various phases in the evolution of these traditions. But the point is, the actual mentality of “sameness” is itself an empirical fact in its own right. [5]
Having noted this clarification, we are ready to survey some additional Homeric contexts of hupokrinesthai. Taken together, these contexts show that Homeric poetry views itself as a performance medium. Further, they show that this poetry has close links to another kind of performance medium, “mantic” or “oracular” poetry. In other words, Homeric poetry equates its own performance with that of a seer or mantis who performs oracular poetry in responding to questions about omens.
My first example of such an equation is Iliad 12.228, describing a hypothetical situation where hupokrinaito (ὑποκρίναιτο), meaning ‘he would respond’, has as its subject theopropos ‘seer’. The speaker is the hero Polydamas, a comrade of Hektor, and he has just interpreted a teras ‘omen’ (12.209): it is the vision of an eagle that drops a snake in midflight. This omen is quoted directly by the words of Homeric narrative at 12.200-209. Then the meaning of these poetic words is interpreted by the likewise poetic words of the hero Polydamas at 12.210-229. The hero is quoted as saying, at the conclusion of his words, that these same words of interpretation could be matched by the words of interpretation spoken by a hypothetical seer (12.228-229). In other words, what the hero performs in the Iliad corresponds to what a theopropos ‘seer’ would perform if he responded to the same vision that has just been narrated by the Iliad.
A far more complex example of hupokrinesthai is Odyssey 19.535, where Penelope challenges Odysseus to respond to the omen of her dream about the killing of the geese in her courtyard by an eagle that swoops down on them: the verb hupokrinesthai is used here in the imperative, with the word for ‘dream’ in the accusative (ὄνειρον ὑπόκριναι). Primarily for this particular Homeric context of hupokrinesthai, the dictionary of Liddell and Scott offers the translation ‘expound, interpret, explain’. [6]
Before we examine this context any further, let us first have a look at an analogous context of hupokrinesthai at Odyssey 15.170, with reference to ‘interpreting’ {23|24} the ominous vision of a flying eagle that is holding a goose in its talons (15.160-165). I have elsewhere offered the following comment on the meaning of hupokrinesthai in this particular Homeric passage: “To interpret is really to formalize the speech-act that is radiating from the dream or omen.” [7] The vision, as quoted directly by the Homeric narration in 15.160-165, confronts the eyewitness Menelaos with an implicit question: what is the meaning of this vision? The hero ponders how to ‘respond’, hupokrinesthai, at 15.170. Before Menelaos can speak, however, Helen anticipates him (15.171: ὑποφθαμένη): she seizes the initiative and utters her own poetic words of response, quoted at 15.172-178. She describes herself as speaking the words of a mantis ‘seer’ (μαντεύσομαι, 15.172), since she is about to reveal how the course of narrated events will reach telos ‘fulfillment’ (ὡς τελέεσθαι ὀίω, 15.173). The words of this oracular performance by Helen, quoted at 15.172-178, give the meaning of the omen: just as the eagle seized the goose, so also Odysseus will deal with the suitors.
I should stress that the words of such an oracular performance are based on the actual vision of the given omen that is seen in “real life” or in a dream, and that this vision has to be performed first as a question - either by a character in the narration or simply by the narration itself - before its meaning can be performed as a response. Further, in the case of Odyssey 19.535, where Penelope challenges Odysseus to ‘respond’, hupokrinesthai, to the omen of her dream about the eagle and the geese, the dream is the omen. [8]
Once Penelope’s vision of the dream about the omen of the eagle and the geese is expressed in her own performance (19.536-553), the ambiguities of its meaning are ready to be resolved in the counter-performance of a response. Surprisingly, the response in this case is already quoted within the dream, when the protagonist of the vision, the eagle, begins to speak and proceeds to interpret, within the dream, the dream itself (19.546-550). In effect, then, Penelope already has a response built into her dream, but she nevertheless challenges the disguised Odysseus to give his own response (19.535).
The rhetoric of Odysseus’ response starts with a declaration that the meaning is so complex as to make any solution impossible. But then there follows a shift to a counter-declaration: that the meaning is in fact so simple that it is clearly visible. The solution, in other words, turns out to be self-evident. {24|25}
Let us look at the precise wording. At Odyssey 19.555, Odysseus starts his response to Penelope by saying that there is no way to ‘respond’, hupokrinesthai, to her dream - and again the word for ‘dream’ is in the accusative (ὑποκρίνασθαι ὄνειρον). But the rhetoric is not yet complete: the verses that follow, 556-558, go on to contradict verse 555: as it turns out, there is no way to respond if the speaker veers away from the words spoken by Odysseus in the dream, that is, only if the words are changed. These words of Odysseus had already earlier been quoted by Penelope herself in verses 546-553. It seems in this case that there can be only one way for Odysseus to respond, that is, to repeat the words already quoted by Penelope. For the meaning to be clarified, the quoted words would have be quoted again, that is, performed. We see at work here the poetic mentality of unchangeability: once the words of response have been performed as a speech act, they are ready to be quoted again as a fixed and unchangeable saying. I find it relevant that the figure of Odysseus is a master of changing the quotations of those he quotes, as he does in Iliad 9. [9]
In this passage of the Odyssey, of course, the time has not yet arrived for Odysseus to clarify his own identity, and so the ultimate clarification of the meaning of Penelope’s dream is postponed. The point remains, though, that the clarification had depended on the quotation of the words of the eagle - words that had in fact already interpreted or mediated the meaning of the omen seen by Penelope in her dream.
In both these Odyssean passages about interpreting the meanings of omens about eagles and geese, as expressed by hupokrinesthai, it is essential to keep in mind that the omen, as quoted, is a vision. In the case of Odyssey 15.168, when Peisistratos asks Menelaos to respond to his question about the teras ‘omen’ of the eagle killing the goose, he says that it had been sent as an epiphany by a god: ἔφηνε θεὸς τέρας ‘the god made as a vision the teras’. In the case of the omen that is Penelope’s dream, her words make it clear that she is quoting what she actually saw. At Odyssey 19.537, Penelope describes herself in her dream as seeing her geese (εἰσορόωσα) at the very moment when the eagle is about to swoop down on them; at 19.567, her words describe dreams in general as things that one sees (ὅτε κέν τις ἴδηται).
The visual orientation of oracular poetry is evident in the ultimate teras ‘omen’ of the Iliad, the vision interpreted by the seer Calchas. The Iliad quotes {25|26}the seer himself as saying at 2.324: ἡμῖν μέν τόδ᾿ ἔφηνε τέρας μέγα μητίετα Ζεύς ‘Zeus the Planner made as a vision for me this great teras’. The teras ‘omen’, which is framed by the narrative of Odysseus (2.284-332) - which in turn is framed by the overall Homeric narrative - is the vision of a snake that first devours eight young birds and then the mother bird (2.303-320). The vision happened in the first year of the Trojan War, during a sacrifice on the occasion of the assembling of Achaean forces at Aulis (2.303). Next, we hear that Zeus, ‘the god who made the vision’, (θεὸς ὅς περ ἔφηνε, 2.318), proceeded to make this vision permanent and unchanging - by changing the snake into stone (2.319). In the words of Odysseus, the whole vision is a sēma (2.308), which we may interpret not only as a mental ‘sign’ but also as a concretized ‘monument’, a landmark.
Just as this petrified vision of the story of Ilion - and of the Iliad itself - is imagined as permanent and unchanging, so too is the kleos or poetic ‘fame’ that radiates from the words of the seer Calchas as directly quoted by Odysseus in Iliad 2.323-332. In the words of the seer himself, as we already saw at 2.324, ἡμῖν μέν τόδ᾿ ἔφηνε τέρας μέγα μητίετα Ζεύς ‘Zeus the Planner made as a vision for me this great teras’. Then Calchas goes on to prophesy that the kleos of this teras will never perish:
ὄψιμον ὀψιτέλεστον, ὅου κλέος οὒποτ᾽ ὀλεῖται
[... this teras], late in coming, late in coming to fulfilment [telos], and its kleos will never perish.
Iliad 2.325
That was then, at Aulis, in the first year of the Trojan War (Iliad 2.303). Now, as Odysseus is quoting back those same words of the seer, it is the ninth year at Troy (2.295). Calchas was saying, back then, in the words that are now being quoted by Odysseus (2.323-332), that the nine birds are the nine years that the Achaeans will spend at Troy, and that the citadel will now be captured finally in the tenth year. After Odysseus quotes these words of the seer, he adds: τὰ δὴ νῦν πάντα τελεῖται ‘and they are now reaching their fulfilment [telos]’ (2.330).
In the poetics of the Iliad, the expression ὅου κλέος οὔποτ᾿ ὀλεῖται ‘and its kleos will never perish’ applies not only to the fame of the teras at 2.325 but also to the fame of the main hero of the Iliad itself. In the words of {26|27} Achilles, ὢλετό μοι κλέος ἐσθλόν ‘my kleos will perish’ (9.415), with the choice of a safe homecoming, nostos. The alternative is worded correspondingly: ὢλετο μέν μοι νόστος ‘my nostos will perish’ (9.413), with the choice of kleos. Ultimately, Achilles chooses kleos, which he prophesies will be permanent and unchanging: ἀτὰρ κλέος ἄφθιτον ἔσται ‘but it will be a kleos unwilting’ (9.413). [10] In other words, the epic of Achilles is its own self-fulfilling prophecy.
Conversely, the prophecy of the seer is not only fulfilled by the epic but also becomes the epic. The kleos of the vision interpreted by Calchas, described as permanent and unchanging, becomes coextensive with the kleos of the Iliad, which likewise describes itself as permanent and unchanging. Moreover, the Iliad prophesies - even at its very beginning - that its own ultimate telos ‘fulfillment’ will be the same thing as the irrevocable will of Zeus: Διὸς δ’ ἐτελείετο βουλή ‘and the will of Zeus was reaching fulfilment [telos]’ (1.5). So also the ultimate telos ‘fulfillment’ of the prophecy of Calchas will happen only when the tale of Troy is ended, when Troy is finally destroyed: only then will the prophecy ‘reach its telos’: teleitai (2.330).
There is a similar coextensiveness between the plot of the Odyssey and the prophecy of Helen concerning the omen of the eagle that kills the goose - a vision that challenges Menelaos to ponder how to ‘respond’, hupokrinesthai, at 15.170. As we already saw, it is Helen who seizes the initiative instead, and she is the one who utters the poetic words of response, quoted at 15.172-178. Speaking as a mantis ‘seer’ (μαντεύσομαι, 15.172), she prophesies how the course of narrated events will reach telos ‘fulfillment’ (ὡς τελέεσθαι ὀίω, 15.173). The words of this oracular performance by Helen, quoted at 15.172-178, express the meaning of the omen, and this meaning is equated with the outcome of the overall plot of the Odyssey: just as the eagle kills the goose, so also Odysseus will kill the suitors as the narrative reaches its telos.
As we see from the precise wording of Odyssey 15.168, the meaning of the Odyssey is formulated as an explicit oracular response to an explicit question about the vision of an omen. When Peisistratos refers to the teras ‘omen’ of the eagle killing the goose, sent as an epiphany by a god, he asks Menelaos {27|28} to respond to this question:
ἢ νῶϊν τόδ’ ἔφηνε θεὸς τέρας ἦε σοὶ αὐτῷ
Did the god made as a vision this teras for you or for the two of us?
Odyssey 15.168
Menelaos ponders how to respond, hupokrinesthai (15.170), to this question, but he fails, since Helen anticipates him (15.171: ὑποφθαμνέη). The response of Helen makes it clear that the meaning of the vision - and the plot of the Odyssey - concerns the agenda of Telemachus, as supported by Peisistratos. This meaning of the Odyssey is intended not for Menelaos and Helen, characters who are left over from the old world of the tale of Troy: rather, the outcome of the Odyssey is meant for the characters of the new world of narrative as represented by Telemachus and his companion Peisistratos. [11] Once again we see that the plot of Homeric narrative is coefficient with the prophecies contained by the narrative. [12]
Just as the poetic words of an oracular prophecy are expected to match exactly the realities of the future that is being prophesied, so also the poetic words of Homeric narrative are expected to match exactly the realities of the past. Since Homeric poetry figures itself as the fulfillment of the prophecies made in its own past, it is coextensive with oracular poetry: just as oracular poetry guarantees the future, Homeric poetry can guarantee the past. When a hero says, as we have seen in Iliad 12.228, that his quoted words are the same words that a theopropos ‘seer’ would have ‘responded’, hupokrinesthai, if he had seen the same vision, (ὧδε χ῾ ὑποκρίναιτο θεοπρόπος), this quotation is ostensibly reinforcing not just the credibility but also the exactness of the Homeric wording.
Similarly, when Calchas speaks as a theopropos ‘seer’ in Iliad 2.322 (θεοπροπέων), his words need to be remembered exactly and quoted back exactly - as they are at 2.323-333 - because the realities will turn out to be exactly the way he tells them. When the words of the seer’s oracular poetry {28|29} are quoted back exactly at 2.323-333, the quoting itself is a demonstration of the unchangeability of these poetic words. As Odysseus concludes after quoting the words, τὰ δὴ νῦν πάντα τελεῖται ‘these things, as I now see, are reaching their fulfillment [= verb teleō]’ (2.330). [13]
The recurring sameness of Homeric quotations, as signaled by such oracular words as hupokrinesthai ‘respond’, corresponds to the recurring sameness of the given vision that calls for the question that calls for the answer. In the case of the words of Calchas as quoted back by the Iliad at 2.323-333, the sameness of the original vision that is being retold is concretized in the image of petrification: Zeus turns the snake into stone at the critical moment when it has just devoured the nine birds (2.319). In doing so, the same god who had made the original epiphany has now made the centerpiece of that epiphany into a permanent landmark:
τὸν μὲ ἀρίζηλον θῆκεν θεὸς ὅς περ ἔφηνε
And it [the snake] was made a radiant thing by the very god that had made it visible [as an epiphany].
Iliad 2.318
The petrified snake, like some splendid statue sculpted by natural forces, radiates a permanent vision matching the permanent words that give it meaning. Such words provide the permanent response to the question posed by the permanent vision. It is this kind of definitive response, I repeat, that we see conveyed by the word hupokrinesthai.
I have just used the simile of a statue because I see a similar mentality of unchangeability at work in the wording inscribed on a bronze statue dated to somewhere between 490 and 480 BCE:
πασιν ισ’ ανθροποι|ς hυποκ|ρινομαι hοστις ε[ρ|ο]ται : |
hος μ᾿ ανεθεκ᾿ ανδ|ρον· Ἀντι|φανες δεκατεν {29|30}
I respond [hupokrinomai] the same things to all men, whoever asks me:
Who among men dedicated me? [14] Antiphanes did, as a tithe. [15]
CEG 286 [IG I ed. 3, 533]
The mentality of unchangeability, where the response to the question is always the exact same thing said in the exact same words, is signaled here again by the word hupokrinesthai. In this case, of course, it is also signaled by the fact that the response is written down in the inscription. The letters will not change, just as the words of oracular response will not change. Still, I will now argue that the actual writing down of the words is not at all the cause of the mentality of unchangeability in this inscription but simply an effect. So also with the word hupokrinesthai: I have been arguing all along that the usage of this word can be viewed as the effect, not the cause, of an overall mentality of unchangeability in oracular poetry and, by extension, in Homeric poetry.
Turning now to the inscription featuring the word hupokrinesthai, let us start with the obvious: the letters of this inscription give potential voice to the words of response. I say “potential” because the voice is not inherently built into the inscription. In the mentality of early Greek poetic inscriptions, including this one, the reader who happens to read a given inscription has to lend his or her own voice by reading the letters aloud, so that these letters may then transmit the words that the inscribed object is saying. [16] The speaker is not the inscription itself, nor is it the actual writing in a more general sense. Rather, the speaker is the dedicated object, which (or “who”) is conventionally marked as the “I” of the discourse. The words of this discourse are inherent in the dedicated object, and it is the actual vision of the object that leads to asking it any question in the first place: who are you and why are you here? The response of the object - I am such-and-such an object and I have been dedicated by so-and-so - is likewise dependent on the vision, which had set the framework of the question in the first place. The dedicated object can always give the same answer because it always gets the same question, given shape by the unchangeability of the vision that radiates from the object. {30|31}
In this case, then, the unchanging response is notionally spoken by the dedicated object, the statue, who is the “I” that responds to all the questions that come from all those who engage with the vision of statue. To repeat, I concede that the mentality of unchangeability is reinforced by the writing down of these words. Still, ideologically, these words of response would stay the same even without writing, since they are predicated on the overall vision of the statue.
The mentality of this inscription has been compared to a celebrated passage in Plato Phaedrus 275d, where the figure of Socrates remarks that an author’s writings will always ‘signify’ (sēmainein) the same thing each time you ask these writings a question. [17] The comparison is valid, to the extent that the writings of authors in books give a seemingly unchangeable response, much as the responses expressed by the word hupokrinesthai are notionally unchangeable. But there is also a fundamental difference: the “responses” of authors’ writings, as described by Plato’s wording here and elsewhere, are not expressed in terms of the word hupokrinesthai. [18] As I have argued from the contexts of hupokrinesthai, this word presupposes a response that is predicated on a specific question, which in turn is predicated on a specific vision. By contrast, the contexts evoked by Plato require no such predications.
In this light, let us take a second look at the notion expressed in Plato Phaedrus 275d, that an author’s writing gives the same meaning each time you ask it to respond. This notion is actually linked with another notion, about the experience of vision: when you look at a painting, zōgraphia, and you ask it a question, its response is nothing but an august silence (275d). Such a mentality is quite different from the one conveyed by hupokrinesthai, where visualization can be translated directly into verbalization: the vision provides the question, which in turn provides the unchangeable response. When an oracle responds to a question about a vision, the unchangeability of that oracular response {31|32} is a sign of the vision’s responsiveness. By contrast, Plato’s notion of viewing a painting presupposes the notion of a vision’s unresponsiveness to the viewer. In terms of these Platonic notions, when you ask a painting what it means, it cannot answer you, and so also when you ask a piece of writing what it means, it can only answer you with the words that you read in the writing.
For Plato, such a response is of course insufficient, since it may not suit the question. For Plato, a response from a book cannot foresee every question. In the case of archaic contexts expressed by hupokrinesthai, by contrast, a response suffices to the extent that it has been made - in performance - to suit a question about a vision. That is, such a response suffices so long as the meaning of the vision is made clear.
When we examine the contexts of hupokrinesthai in historicized accounts of oracular responses, as in the History of Herodotus, we can see even more clearly the importance of the vision, of the visualization of the vision, as the basis for verbalizing any given oracular response. For example, Croesus of Lydia knows that he has had a vision when he ‘sees’ (ἰδόντι) horses devouring snakes, and he recognizes this vision as a teras ‘omen’ (1.78.1); he therefore orders his theopropoi ‘seers’ to consult the oracular exēgētai ‘interpreters’ of Telmessus with the question, ‘What does the omen [ teras ] mean [ sēmainein ]?’ (τὸ θέλει σημαίνειν τὸ τέρας 1.78.2). The theopropoi, once they get a response, do not have a chance to apangellein ‘announce’ it to Croesus because in the meantime he has already been defeated and captured by Cyrus (1.78.2). The content of the oracular response given by the interpreters of Telmessus is at this point quoted in indirect discourse, introduced by the expression ‘these were the things they had recognized’ (τάδε ἔγνωσαν, 1.78.3). The response makes clear the basic equation: that the snake is the same thing as an autochthonous native son, that is, Croesus the Lydian, while the horse is the same thing as an enemy newcomer, that is, Cyrus the Persian (1.78.3). Then, the content of the oracular response is concluded with the expression ‘these were the things they had said in responding [hupokrinesthai]’ (ταῦτα ὑπεκρίναντο, 1.78.3).
Here is another important example in Herodotus: Croesus sends messengers to the Oracle at Delphi, asking (εἰρωτᾶν/ἐπείρεωτᾶν, 1.90.4), among other things, why the Oracle had encouraged Croesus to initiate war against Cyrus. The response is given by the chief priestess of the Oracle, the Pythia, and the content of this oracular response is quoted in indirect discourse {32|33} (1.91.1-5). At one point in the course of this response, it is made explicit by the Pythia that Croesus had ‘asked’ (ἐπείρεσθαι, picked up later by ἐπανειρόμενος, 1.91.4) his original question wrongly. At a later point, referring to a verse of oracular poetry that is actually quoted earlier by the narrative (1.55.2), the Pythia makes clear an essential equation, declaring explicitly that Cyrus is the same thing as the hēmionos ‘mule’ that had been visualized by the riddling verse of that earlier oracular response. Then, the content of the current response is concluded with the expression ‘these were the things that the Pythia had said in responding [hupokrinesthai]’ (ταῦτα μέν ἡ Πυθίη ὑπεκρίνατο, 1.91.5). It is these words of response, ultimately, that the messengers of Croesus bring back to apangellein ‘announce’ to him. [19]
The visualization that is built into the verbalization of oracular responses must extend from the one who gives the response to the one who receives it. In the traditions of the Oracle at Delphi, a word that conventionally designates the recipient of the oracular response makes the idea of visualization explicit: it is theōros, meaning literally ‘he who sees [root hor-] a vision [thea]’, designating the messenger who is sent by his city to ask the question to which the Pythia will respond (Suda s.v. τὰ τρία; Theognis 805; Sophocles Oedipus Tyrannus 114). [20]
Such messengers can be theopropoi ‘seers’ in their own right, as we see from a story told by Herodotus about the consultation of the Delphic Oracle by the Athenians during the crisis of the Persian invasion (7.140-144). According to this story, the force of an oracular statement is not activated, the words do not become a completed speech-act, until they are performed before the audience for whom it was intended. Herodotus tells of Athenian emissaries who consulted the Delphic Oracle about their impending fate in the Persian War and who wrote down what the Oracle told them; having {33|34} written it down, they returned to Athens and ἀπήγγελλον ἐς τὸν δῆμον, ‘announced [apangellein] it to the people’ (7.142.1). [21]
In this narrative, the oracular poetry of the Delphic Oracle is delivered directly by the Pythia herself in hexameters; at one point, the narrative refers to her as the chief mantis ‘seer’ (7.141.2: ἡ πρόμαντις). [22] In other narratives about oracular responses, there may be an intermediary, the prophētēs (whence the English word prophet), who transforms the oracular words of the mantis ‘seer’ into formal poetry (Herodotus 8.135.3). [23]
There are many other variations in the usage of such words as mantis, prophētēs, and theōros, reflecting a wide variety of cultural differences in the traditions of oracular poetry. [24] Accordingly, we should not be surprised to find stories about disputes over the truth value of various historical instances of oracular poetry. [25] We see an example even in the story that we have just considered, concerning the consultation of the Delphic Oracle by the Athenians during the crisis of the Persian invasion (7.140-144): in this case, the responses of the Pythia, a set of oracular poems formulated in hexameter, provoke conflicts in interpreting the words of the Oracle, leading Themistocles to declare that the khrēsmologoi ‘oracle-sayers’ have not understood everything orthōs ‘correctly’ (7.143.1: οὗτο ὡνὴρ οὐκ ἔφη πᾶν ὀρθῶς τοὺς χρησμολόγους συμβάλλεσθαι).
What, then, is the word for what Themistocles is doing at the moment when he gives a different interpretation of the Oracle’s words? Galen in his Protrepticus (13) describes Themistocles at this moment as engaged in the process of hupokrinesthai; in the idiom, the words of the oracle become the direct object (ὑποκρινομένου τὸν χρησμόν). That is, Themistocles is intepreting the oracle in the process of quoting the words, performing them all over again.
Here we confront a basic problem. How is the performance of Themistocles, as expressed by the word hupokrinesthai, different from that of the Pythia, as also expressed by hupokrinesthai? The mind of the priestess, as chief {34|35} mantis, is notionally possessed by the god, inspired, at the time of her performance. Are we to say, by contrast, that Themistocles was not possessed at the time of his own performance? Here it is relevant to cite Plato’s description of the prophētēs as an oracular poet who, unlike the mantis, performs without being possessed or inspired:
τοῦ δὲ μανέντος ... οὐκ ἔργον τὰ φανέντα καὶ φωνηθέντα ὑφ᾿ ἑαυτοῦ κρίνειν ... ὅθεν δὴ καὶ τὸ τῶν προφητῶν γένος ἐπὶ ταῖς ἐνθέοις μαντείαις κριτὰς ἐπικαθιστάναι νόμος· οὓς μάντεις αὐτοὺς ὀνομάζουσίν τινες, τὸ πᾶν ἠγνοηκότες ὅτι τῆς δι᾿ αἰνιγμῶν οὖτοι φήμης καὶ φαντάσεως ὑποκριταί, καὶ οὔτι μάντεις, προφῆται δὲ μαντευομένων δικαιότατα ὀνομάζοιντ᾿ ἄν.
But when some person is in a state of mental possession, ... it is not that persons task to sort out [ krinein ] the visions that are made visible to him or the words that are voiced by him. ... For this reason it is customary to appoint the class of prophētai as kritai [= ‘those who sort out’ = ‘judges’] presiding over oracular utterances [manteiai] that had been made [by others] in the state of mental possession by the god. They [= the prophētai] are called by some, in ignorance, manteis. This is to ignore completely the fact that they [= the prophētai] are hupokritai , by ways of riddles [ainigmoi], of oracular utterance [phēmē] and oracular vision [phantasis]. So they would be most accurately called the prophētai of things that are uttered by those who function as manteis.
Plato Timaeus 72a-b
We see here the agent noun of the verb hupokrinesthai, hupokritēs, in the sense of a performer who is no longer possessed, no longer inspired by the god. But we must keep in mind that the context here reflects Plato’s own philosophical agenda, and that only in this context is the prophētēs assumed to be uninspired. In non-Platonic contexts, the prophētēs too is considered to be inspired (e.g. Pindar Nemean 1.60; Herodotus 3.37.2). [26] The distinction between mantis and prophētēs is not a question of being inspired or not inspired. Rather, it is more a question of different degrees of formalization: whereas the words of a mantis may or may not be poetic, those of a prophētēs {35|36} are predictably so: a salient example is Bacchylides Epinician 8.3, where the generic poet is the prophētēs of the Muse.
As with prophētēs, so also with hupokritēs: it does not need to be taken in the sense of a performer who is no longer possessed, no longer inspired by the god. Plato’s wording, as we have seen, describes hupokritēs as one who interprets, by way of performing, the utterance [phēmē] and the vision [phantasis] of a given oracle. There is no need to follow Plato, however, in assuming that such a hupokritēs must be disconnected from the visualization and the verbalization of the oracular vision. Just as the verb hupokrinesthai is predicated on the idea of a preexisting vision, so also the noun hupokritēs. In the case of hupokrinesthai, as we have seen, the idea of vision is evident from the word’s conventional associations with such other words as theōros, meaning literally ‘he who sees [root hor-] a vision [thea]’. In the case of hupokritēs, as we will now see, the word is associated with theatron ‘theater’, meaning literally ‘the vehicle for achieving vision [thea]’. In fact, hupokrinesthai and hupokritēs become the words for ‘perform’ and ‘performer’ in the language of Athenian State Theater.
I find it relevant and even essential to consider at this point the verb hupokrinesthai in the sense of ‘perform in theater’, with the object of performance in the accusative. A case in point is Demosthenes 19.246: τὴν Ἀντιγόνην Σοφοκλέους ὑποκέκριται as ‘he has performed the Antigone of Sophocles’. The definition of hupokrinesthai in the dictionary of Liddell and Scott (LSJ s.v., B II) is suggestive: “Speak in dialogue, hence play a part on the stage, the part played being put in [accusative].” By metonymy, the part can become the whole (as reflected in the italics that I used here for Antigone), so that the part being played stands for the whole play. We may compare such accusative constructions as τὸ δρᾶμ ...ὑπεκρίναντο ‘they performed the drama’ (again, Demosthenes 19.246). We may compare also ὑπεκρἰνω plus accusative of the words being spoken (at Demosthenes 19.250). Another relevant example is Aristotle Rhetoric 3.1403b23: ὑπεκρίνοντο γὰρ αὑτοὶ τὰς τραγῳδίας οἱ ποιηταὶ τὸ πρῶτον ‘in the beginning the poets themselves used to perform their tragedies’. [27]
Besides the verb hupokrinomai in the contexts just surveyed, I find it just as essential to consider the noun hupokritēs, which is the standard word for {36|37} ‘performer in theater’, ‘actor’ (Plato Republic II 373b, Symposium 194b, etc.). This word is used in parallel contexts with a word that refers to the performer of Homeric poetry, the rhapsōidos ‘rhapsode’, as in Plato Ion 532d7. [28]
The concept of the rhapsode is basic for understanding the central argumentation of this essay - and of this whole book. Just as the rhapsōidos performs Homer, so too the hupokritēs performs drama and other forms of poetry. Conversely, just as the hupokritēs acts out his given role in drama, so too the rhapsōidos acts out his given role as the master narrator of the Homeric Iliad and Odyssey.
When a rhapsode like Plato’s Ion performs Homeric poetry, he is not only “quoting” the words of heroes and gods, thereby “acting” both their words and their personalities: he is also “quoting” the words of Homer, which are the narrative frame of heroic song. That is, the rhapsode is also “acting” both the words and the persona of Homer himself. [29]
Not only is the hupokritēs parallel to the rhapsōidos: “the term hupokritēs, normally used of the dramatic actor, could also be used of the rhapsode.” [30] The performance of Homeric poetry by rhapsodes, just like the performance of dramatic poetry by actors, requires a responsive mentality, as conveyed by the word hupokrinesthai.
The connection between the meanings of hupokritēs and rhapsōidos is relevant to the collocation of hupokrinesthai and hupolambanein, as in Aristotle Politics V.1310a10. Let us consider the definition for hupolambanein in the dictionary of Liddell and Scott: “In discourse, take up what is said, interpret or understand it in a certain way” (e.g. Plato Republic I.338d). [31] In dramatized dialogue, hupolambanein marks the response of one speaker to the previous speaker: ephē hupolabōn ‘he said in response’ (e.g. Plato Republic I 331d, etc.; Herodotus 1.11.5, 1.27.4, etc.). The noun hupolēpsis, derived from hupolambanein, is in fact a technical rhapsodic term (“Plato” Hipparkhos 228b-c). [32]
The dictionary definition of hupolambanein as ‘interpret’ calls for further comment. The very act of performing can be considered an act of interpretation, as we see from such modern usages as French interpréter in the sense of {37|38} ‘sing’ or ‘play’ a given musical composition. In this light, let us consider the idea inherent in usages of krinein, from which the compound form hupokrinomai is derived. This verb krinein, in the active voice, can be translated as ‘interpret’ when combined with the noun opsis ‘vision’ as its object (Herodotus 7.19.12) or with enupnion ‘dream’ as its object (Herodotus 1.120). [33] It is a question of interpreting-in-performance. The middle voice of hupokrinomai suggests that the performer is interpreting for himself as well as for others. [34] The basic idea of hupokrinesthai, then, is to see the real meaning of what others see, and to “quote” this vision for them.
Whereas the art of the rhapsode, as conveyed by the word hupokrinesthai and related forms, is associated with the idea of responsiveness, Plato seems to associate this art with the idea of unresponsiveness. We may consider in this regard another compound of krinein, that is, anakrinein, which means ‘interrogate [judicially]’, as in Thucydides 1.95. In Plato Phaedrus 277e, we read of logoi that are rhapsōidoumenoi ‘performed rhapsodically’ and are exempt from anakrisis ‘interrogation’ (οἱ ῥαψῳδούηενοι ἄνευ ἀνακρίσεως).
I come back to the central point, that the performance of Homeric poetry by rhapsodes requires a responsive mentality, as conveyed by the word hupokrinesthai. I repeat also a related point: that the framing words of Homer require the same responsive mentality as required by the framed words of heroes and gods. The performance of Homer as a speaker mirrors the performances of the heroes and gods whose speeches he frames. [35] Homer as the framing narrator mirrors the poetic virtuosity of his framed epic characters, especially Achilles. [36] The responsive mentality of speakers in Homeric song extends ultimately to Homer himself, who becomes reenacted again and again in the traditions of rhapsodic performance. {38|39}


[ back ] 1. An earlier and shorter version of this chapter is Nagy 2002b.
[ back ] 2. (An earlier and shorter version of this chapter is in Nagy 2002b). Cf. Koller 1857 on a basic insight: that the prehistory of oracular discourse in early Greek civilization is relevant to the prehistory of the word hupokrinesthai. (His article also provided a bibliography of earlier research on hupokrinesthai and its derivatives.)
[ back ] 3. For analogous attestations of hupokrinesthai, see Odyssey 2.111 and Homeric Hymn to Apollo 171. In the Hymn to Apollo, the quoted words of response, verses 172-174, are a description of “Homer” – as quoted by “Homer” himself: see PH 376.
[ back ] 4. The “quoting” of words in Homeric poetry is equivalent to performing the “quotation”: see Nagy 2002a:21-23 with reference to Martin 1989, esp. his p. 117.
[ back ] 5. HQ 40, with bibliography.
[ back ] 6. LSJ s.v.
[ back ] 7. PH 168 n. 95.
[ back ] 8. On the poetics of perception in quoted dreams, I value the insights of Murnaghan 1987:52.
[ back ] 9. See further in Chapter 3 below, where I discus Odysseus’s rewording of the offer to Achilles by Agamemnon, as reformulated by Odysseus in his quoted speech from the Embassy Scene. Cf. Martin 1989:116-117, 123.
[ back ] 10. PH 244-245. On the epithet aphthiton in Iliad 9.413, see Chapter 2 below (originally published as Nagy 2000d). There I stress that attributive and predicative usages of adjectives can be explained as syntactical variants within one formulaic system, from the standpoint of the generative grammar. See also Volk 2002.
[ back ] 11. On Telemachus as a representative of the “post-heroic” age, see Martin 1993.
[ back ] 12. On the involvement of the figure Peisistratos, as notional ancestor of the Peisistratidai of Athens, in the teleology of Homeric narrative, see HQ 43 n. 58.
[ back ] 13. The particle δὴ here has an “evidentiary” force (‘Aha, now I see that…’); for more on this sense of δὴ, see Bakker 1997:74-80. As for the usage of teleō ‘come to fulfillment’ here at Iliad 2.330, we may compare Odyssey 19.547, where the talking eagle of Penelope’s dream prophesies that this dream (hupar) will come to fulfillment – that it will be tetelesmenon (verb teleō).
[ back ] 14. On the use of a relative pronoun for an indirect question, cf. PH 221 n. 34.
[ back ] 15. I no longer agree with my translation as given in HQ 35 n. 25: now I take the genitive-plural construction ‘among men’ to go with “who,” not with “Antiphanes.”
[ back ] 16. Svenbro 1988:33-52 (= 1993:26-43), esp. pp. 36-38 (= 29-31); cf. also Day 1989.
[ back ] 17. Svenbro 1988:36 (= 1993: 28-29).
[ back ] 18. In his discussion of Plato Phaedrus 275d, Svenbro cites analogous passages in Plato Protagoras 329a (when you ask them a question, books have nothing to answer, apokrinesthai [not hupokrinesthai], nor any questions to ask), Hippias Minor 365c-d (not directly relevant to books per se: Socrates says that it is impossible to ask Homer what he really meant to say, so that Hippias should answer Socrates on Homer’s behalf as well as on his own; cf. Republic 2.378d), Laws 12.968d-e (only tangentially relevant; no direct references to questions and answers), Letters 7.343a (on the idea of the ametakinēton, the ‘unchangeable’, in writings).
[ back ] 19. Moving from Herodotus, we may note the usage of hupokrinesthai in Thucydides. At Thucydides 7.44.5, we read that the Athenians did not know how to respond, hupokrinointo (ὑποκρίνοιντο), to the xunthēma ‘watchword’ of the Syracusans. From 7.44.4, it becomes clear that both the Athenian and the Syracusans had their own watchwords, which were formulated in the form of erōtēmata ‘questions’. To save your life, you had to have the right response (hupokrinesthai) to the right question. (The Oxford Classical Texts edition [1901+] of H. S. Jones simply prints the M manuscript reading ἀποκρίνοιντο, without indicating the lectio difficilior ὑποκρίνοιντο, as found in other manuscripts.) The Athenians, in the darkness, would keep asking, erōtōntes, their password, thereby making is saphes ‘clear’ for the enemy. The Syracusans were understanding, epistamenoi, the xunthēma of the Athenians, while the Athenians did not correspondingly (homoiōs) understand, ēpistanto, that of the Syracusans (7.44.5).
[ back ] 20. For further discussion, see PH 164-167.
[ back ] 21. PH 168.
[ back ] 22. Note, too, the context of promantis in Herodotus 6.66.2.
[ back ] 23. See further in PH 168 n. 95.
[ back ] 24. For a general discussion, see PH 164-169.
[ back ] 25. Of particular interest are stories suggesting that the Pythia was on various historical occasions “persuaded” to declare oracular pronouncements that were prejudiced in favor of special-interest groups: see, for example, Herodotus 6.66.3, 6.75.3 (cf. PH 163).
[ back ] 26. PH 163-164.
[ back ] 27. Nagy 2002a:10.
[ back ] 28. Ibid.:28.
[ back ] 29. PP 86, 220.
[ back ] 30. Murray 1996:110, referring to Koller (1957:104), who cites collocations of hupokritēs and rhapsōidos in Plato Ion 532d, 535e.
[ back ] 31. LSJ s.v.
[ back ] 32. Nagy 2002a:10-11.
[ back ] 33. Koller 1957:101.
[ back ] 34. Ibid.:102.
[ back ] 35. This point is argued most effectively by Martin (1989; se esp. pp. 231-239).
[ back ] 36. Martin 1989:220-230.