The senses of an ending: myth, ritual, and poetic exodia in performance

Richard P. Martin, Stanford University
I like to think it fitting that I first met Gregory Nagy, some 40 years ago, at a puppet show. Fred Christie, then a Jesuit scholastic and graduate student at Harvard, had taken two of us, rather sceptical students of Greek from Boston College High School, to see Peter Arnott’s one-man miniature production of the Choephoroi, which turned out to be a mesmerizing performance. Afterwards, we followed the crowd into Ticknor Lounge in Boylston Hall, gaping at our first-ever university reception, the presence of attractive young women (amidst the more intimidating faculty members), and the flow of wine. Though just 20 minutes on the Red Line, this was a long way from Dorchester.
It was Fred who made the introductions: Greg, very kindly in response to my interest in Homer, mentioned that I might try C. D. Buck’s The Greek Dialects and the Traité de grammaire comparée des langues classiques by Meillet and Vendryes. (The former I laid claim to immediately from our school library, entranced by its multi-colored frontispiece map and the idea that the distant Greeks, already so alien, were so different even to themselves; the latter had to wait a few years until I could manage enough French to puzzle it out.) For a 17-year-old, to feel language, literature, and performance all taken so seriously, but at the same time with such joy, was magic, as it was to discover a passionate and approachable teacher so generous with his advice. Performance was to become a major part of my scholarly life (as were Harvard and Greg, though I could hardly foresee any of it that evening). To honor that long-ago initiatory moment, as well as Greg’s pioneering work in poetry and performance, I offer these musings on endings.

Citharodic closings

Rare and unusual information about ancient performance traditions is contained in a fragment attributed to a scholar of the Hadrianic period, Aelius Dionysius the Atticist. He is probably the same man whose voluminous works on the history of music are mentioned in the Suda and several other sources, for these included much information on auletes, citharodes and poets, and the lexicon shows similar familiarity with technical terms of mousikê. [1] For example, Aelius Dionysius preserves for us the term epiporpama—a kind of jacket worn by citharodes—which he quotes from Plato the comic poet:

ἐπιπόρπαμα· ἡ τῶν κιθαρῳδῶν ἐφαπτίς. Πλάτων ἐν Ταῖς ἀφ’ ἱερῶν (PCG 10) ‘δότω τὴν κιθάραν τις ἔνδοθεν καὶ τοὐπιπόρπαμα’. [2]
Epiporpama: The citharodes’ upper garment. Plato in The Women from the Temples “Let someone from inside give the kithara and the epiporpama.”

The following fragment on musicians and poets is transmitted by the Byzantine scholar Photius:

Ἀλλ’ ἄναξ· ἐξοδίου κιθαρῳδικοῦ ἀρχή, ὥσπερ κωμικοῦ μὲν (adesp. fr. 48 Dem.) ἥδε “καλλιστέφανος,” ῥαψῳδῶν δὲ “νῦν δὲ θεοὶ μάκαρες τῶν ἐσθλῶν ἄφθονοί ἐστε” (Photius Lexicon α 987.)
“But, lord.” The beginning of a citharodic exodion [exit song], just as of the comic [exodion] there is this: “having a beautiful crown” and of the rhapsodes: “Now, blessed gods, be unstinting in fine things.”

Although he does not attribute it directly in his Lexicon to Aelius Dionysius, Photius elsewhere speaks of the utility of the Atticist’s specialized lexicon, Attika Onomata and recommends it—in both its first and second ekdoseis—to all who would care to learn Athenian epichoric terms for lawsuits and festivals, among much other useful information (Photius, Bibl. Cod. 152–153). Several other late sources enable us to fill out further this intriguing bit of antiquarian lore. The fullest version of the information, and the only to trace it directly back to Aelius Dionysius, comes in the 12th-century Homeric commentary by Eustathius, in partial explanation of a line in Book 2 of the Iliad:

(v. 360) Ἰστέον δὲ ὅτι ἐκ τοῦ “ἀλλ’ ἄναξ”, ὅπερ ἐνταῦθα παρὰ τῷ ποιητῇ κεῖται, ἀρχή τις ἐξοδίου κιθαρῳδικοῦ τὸ “ἀλλὰ ἀλλ’ ἄναξ”, ὡς ἱστορεῖ Αἴλιος Διονύσιος, ὥσπερ, φησί· κωμικοῦ μὲν ἥδε· “καλλιστέφανος”, ῥαψῳδοῦ δὲ αὕτη· “νῦν δὲ θεοὶ μάκαρες τῶν ἐσθλῶν ἄφθονοί ἐστε”, τραγικοῦ δέ “πολλαὶ μορφαὶ τῶν δαιμονίων”. (Eustathius Ad Hom. Iliadem Volume 1 page 364 line 6.) [3]
Note that from “but lord” which is here found in the poet, there is a beginning of a citharodic exodion, the [phrase] “but [other things?] lord,” as Aelius Dionysius relates, just as, he says, there is this of the comic [exodion]: “having a beautiful crown,” and this of the rhapsodes: “Now, blessed gods, be unstinting in fine things.” And of the tragic [exodion]: “many are the shapes of the divine.” [4]

Zenobius the paroemiographer (Cent. 5.99) appears to give a faulty version of the rhapsodic tag-line, but usefully adds a phrase to the citharodic closing:

Σὺν δὲ θεοὶ μάκαρες: τοῦτο ἐπιλέγουσιν οἱ ῥαψῳδοί· ὡς καὶ οἱ κιθαρῳδοὶ, Ἀλλ’ ἄναξ μάλα χαῖρε.
In addition [reading sun rather than the more plausible nun] the blessed gods: the rhapsodes make a formal close with this (epilegousin). In the same way, too, the citharodes “but lord, rejoice/farewell, indeed.” [5]

Diogenianus (Cent. 6.88), another collector of proverbs, attests to the more plausible phrasing “Now [nun] blessed gods” although his version (missing the particle de) is unmetrical for inclusion in hexameter verses. He indicates further that the phrase had a usage—perhaps outside of any poetic context—“said of those worthily avenged for what they did” (ἐπὶ τῶν ἀξίως τιμωρουμένων ἐφ’ οἷς ἔπραξαν.) An entry in the Suda (σ 1454) employs the Zenobian phrasing (Σὺν δὲ θεοὶ μάκαρες), and identifies it as rhapsodic then gives a slightly shorter form of the citharodic signature line (ἀλλ’ ἄναξ χαῖρε) and ends by explaining “these are finishing phrases (ἐπιφωνήματα) among the old-time poets.” No doubt the traditions of stylistic ornamentation among the teachers of rhetoric have influenced this description, as they may also have helped preserve the details about closing phrases, in the first place. We can note, for example, that the essay of Demetrius on style devotes a section to the definition and use of the epiphonema (De elocutione 106–111)—although he expands the idea to cover ornamental passages unneeded by the immediate argument (such as a phrase used to fill out a Homeric simile, or a gnomic utterance added to the narrative description of the arms in the hall of Odysseus). [6]

While we can see that the theorists of prose writing, as often, sought models in epic usage, and may have tacitly acknowledged the poetic performance practice of rhapsodes, citharodes, and others singing a stylized end-piece—the exodion—there is no sign that any of them (including Dionysius of Halicarnassus, whom the Suda [δ1174] identifies as ancestor of Aelius Dionysius the Atticist) elaborated on the specific form of these four types of poetic exodia or their wording. [7] But, then again, there are very few signs that the poetry itself, in the form that we now have it, featured any of the phrases that we find preserved by the lexicographic and paroemiographic traditions.
We might initially take comfort in seeing the lore about exodia confirmed by extant tragedies that do in fact end with the tag-line “many are the shapes of the divine.” The full tragic exodion as found in the Alcestis of Euripides runs (1159–63):

πολλαὶ μορφαὶ τῶν δαιμονίων,
πολλὰ δ’ ἀέλπτως κραίνουσι θεοί·
καὶ τὰ δοκηθέντ’ οὐκ ἐτελέσθη,
τῶν δ’ ἀδοκήτων πόρον ηὗρε θεός.
τοιόνδ’ ἀπέβη τόδε πρᾶγμα.
There are many shapes of divinity, and many things the gods accomplish against our expectation. What men expect is not brought to pass, but a god finds a way to achieve the unexpected. Such was the outcome of this story. (trans. D. Kovacs)
Three other tragedies by Euripides end with these exact lines (Andromache, Helen, Bacchae). Encouraged by this confirmation, we might then look at verses which are associated with rhapsodes or citharodes to see whether the conventional verse tag-lines assigned to their respective performers are found in our surviving texts. By extension, we could even hope to identify some compositions as having been part of certain genres or repertoires given the occurrence of the conventional exodia in them. But hopes of finding an exact repetition of either sort of rhapsodic or citharodic exodion are soon shattered by experiment. On the other hand, the pursuit becomes all the more interesting when we open up the possibility that our extant texts, while they may not quote exodia tag-lines, stylize, allude to, or present family resemblances to these apparently well-known phrases. It could be that performers, indeed, avoided direct use of the phrases precisely to undermine audience expectations, while still providing a touch that would recall to mind the poetic convention. Then, too, we can imagine the possibilities for signalling genre-crossing that such tacitly shared audience-performer knowledge could generate. All this requires us to think beyond the immediate text and to incorporate the context of live performance—which is itself a sort of text—into our interpretations. [8]
Let us begin with the alleged citharodic exodion, in the fullest form that we can reconstruct: *ἀλλὰ ἄναξ μάλα χαῖρε. [9] Two details of my reconstruction are worth pausing over. Although the sources transmit only the elided form ἀλλ’ ἄναξ, this combined with the rest of the phrase (preserved only in Zenobius) would be unmetrical for a hexameter. [10] It could of course be argued that citharodic nomoi were not regularly composed in hexameters. Our picture is complicated by the innovations of Timotheus and other proponents of the New Music at Athens in the later 5th century. [11] The testimonia do tell us that Terpander, the alleged 7th-century BC inventor of the names of the nomoi, and a composer of them, put Homeric hexameters, as well as his own hexameter verses, to melodic accompaniment (test.27 Gostoli = Heracl.Pont. fr.157 Wehrli). But as Greg Nagy and others have pointed out, lyric meters, some related to the hexameter, are also attested for such nomoi. [12] A second point to keep in mind is that an unelided form *ἀλλὰ ἄναξ would be linguistically older; Eustathius, offering what must be a mangled form of the phrase in the sequence “ἀλλὰ ἀλλ’ ἄναξ” (difficult to understand unless the second word represents a neuter plural object of some unexpressed verb), might nevertheless be preserving a memory of an original unelided phrase *ἀλλὰ ἄναξ in citharodic use. (I shall return shortly to Eustathius and this phrase, as it does appear unelided in the Homeric context upon which he comments.) Poets of a generation no longer aware of the original initial digamma in *wanax (as we see it attested in Mycenaean) would naturally avoid the apparent hiatus and elide to ἀλλ’ ἄναξ. At any rate, in looking for traces of this phrase, I have searched for both elided and unelided forms over the whole range of Greek epic and lyric verses, using as a basis of comparison both the short and long versions of the alleged exodion. There are some interesting results.
The logical place to look first is among verses identified as citharodic, meager as they are. Timotheus of Miletus, in his famous nomos, The Persians (PMG 791), uses the word ἀλλ’ at the very end of his citharodic composition, in combination with the sphrêgis that gives his name and polis affiliation (234–40):

Μίλητος δὲ πόλις νιν ἁ
θρέψασ’ ἁ δυωδεκατειχέος
λαοῦ πρωτέος ἐξ Ἀχαιῶν.
ἀλλ’ ἑκαταβόλε Πύθι’ ἁγνὰν
ἔλθοις τάνδε πόλιν σὺν ὄλβωι,
πέμπων ἀπήμονι λαῶι
τῶιδ’ εἰρήναν θάλλουσαν εὐνομίαι.
It was the city of Miletus that nurtured him, the city of a twelve-walled people that is foremost among the Achaeans. Come far-shooting Pythian, to this holy city and bring prosperity with you, conveying to this people, that they be untroubled, peace that floruishes in good civic order. (trans. Campbell)

Admittedly, it is not followed by ἄναξ or the rest of the full citharodic tag-line. Yet the line does contain a vocative addressing the god Apollo as Pythios and formulates a prayer for peace and prosperity. [13] (In this latter point it resembles the rhapsodic exodion, which is overtly a prayer for similar good things: esthla).

Of course, it might be objected that almost all Greek hymns feature a send-off formula of some sort, even if ἀλλ[ὰ] ἄναξ μάλα χαῖρε never occurs verbatim. As Hordern points out (2002:246), in connection with the end of the Timotheus poem, “a short prayer is a common feature at the end of hymns” adducing epigraphically attested poems by Isyllus, Ismenius, Philodamos, Limenius, and Aristonoos. In effect, we are dealing with normal Greek prayer-style, of the sort represented even in a lower register by the writer of mimiambic verses, the Hellenistic poet Herodas, as in the opening of his poem about the visit to the shrine of Asclepius by two women wishing to sacrifice a cock (Mim. 4):

χαίροις, ἄν̣αξ Παίη̣ον, ὂς μ̣έδεις Τρίκκης
καὶ Κῶν γλυκεῖαν κἠπίδαυρον ὤικηκας,
σὺν καὶ Κορωνὶς ἤ σ’ ἔτικτε κὠπόλλων
χαίροιεν, ἦς τε χειρὶ δεξιῆι ψαύεις
῾Υγίεια, κὦνπερ οἴδε τίμιοι βωμοί
Πανάκη τε κἠπιώ τε κἰησὼ χαίροι

Hail, Lord Paieon, you who rule over Trikka
And have made Kos and Epidauros your dwelling-place;
Koronis, too, who bore you and Apollo,
May they fare well, and she whom you touch with the right hand, Hygieia,
And those for whom altars are revered, Panake and Epio and Ieso—Hail.

Because “fare well” (khaire) can be used at the start or end of any exchange in Greek, we hear it as the potential worshipper enters the sanctuary, or when the worshipper leaves the divinity’s space. A lyric poem from the shrine of Asclepius at Epidauros makes explicit the departure of the speaker (the Mother of the Gods, in this narrative) and joins this to the “send-off” formula and an address using the word ἄνασσα (PMG 935=IG IV2 131):

καὶ οὐκ ἄπειμι εἰς θεούς,
ἂν μὴ τὰ μέρη λάβω, (20)
τὸ μὲν ἥμισυ οὐρανῶ,
τὸ δ’ ἥμισυ γαίας,
πόντω τὸ τρίτον μέρος
χοὔτως ἀπελεύσομαι.
χαῖρ’ ὦ μεγάλα ἄνασ– (25)
σα Μᾶτερ Ὀλύμπου.
‘I shall not go off unless I get my portions, half of the heaven and half of the earth and a third portion, half of the sea: only then shall I go off’. Greetings, great Mother, queen of Olympus! (trans. Campbell)

So the individual elements of the alleged citharodic exodion—khaire-forms, forms of *wanax—are hardly surprising in a hymnic context. In addition to the examples above, and in several epigraphic hymns, it is worth noting that every one of the collection of “Homeric” hymns contains a khaire formula. [14] A few employ a multiform line that is quite close in wording to the citharodic exodion, as the following examples show:

In Herculem Line 9:
Χαῖρε ἄναξ Διὸς υἱέ· δίδου δ’ ἀρετήν τε καὶ ὄλβον.

In Solem Line 17:
Χαῖρε ἄναξ, πρόφρων δὲ βίον θυμήρε’ ὄπαζε·

In Lunam Line 17:
Χαῖρε ἄνασσα θεὰ λευκώλενε δῖα Σελήνη

Whether such lines are meant to actively summon up citharodic practice, in a medium that might have been recited rather than sung (see below), or whether the resemblance is a reflex of a more general hymnic-prayer rhetoric, it is hard to deny that the placement of these verses underlines their similarity with the citharodic exodion.

It may be slightly more surprising that citharodic compositions (so it would appear, if the information about the exodion is correct) were at some time period explicitly connected with a singer’s farewell to a god—a markedly hymnic closing for a genre that we might not immediately associate with hymns. If we take them at face value (perhaps unwisely) the words of Plato’s Athenian in the Laws (700b) make it clear that at some (perhaps irrecoverable) point, hymns were distinct from nomoi, and citharodic nomoi were distinct from other sorts:

Among us, at that time, music was divided into various classes and styles: [700b] one class of song was that of prayers to the gods, which bore the name of “hymns”; contrasted with this was another class, best called “dirges”; “paeans” formed another; and yet another was the “dithyramb,” named, I fancy, after Dionysus. “Nomes” also were so called as being a distinct class of song; and these were further described as “citharoedic nomes.” (νόμους τε αὐτὸ τοῦτο τοὔνομα ἐκάλουν, ᾠδὴν ὥς τινα ἑτέραν· ἐπέλεγον δὲ κιθαρῳδικούς).

Then again, paeans and dithyrambs, which Plato’s speaker places apart from humnoi, are clearly directed towards gods (Apollo and Dionysos, usually), just like hymns. Because the texts of identifiable citharodic nomes are so meager, the lore about citharodes must be consulted. It yields a number of details that do connect performance to the god Apollo. First of all, Apollo as depicted in the Hymn to Apollo, and in visual art, is himself the archetypal citharode. [15] Second, the chief festival at his shrine in Delphi had as its centerpiece in the mousikoi agônes a citharodic competition. [16] Before the reorganization of the Pythian games in the 6th century BC, the only contest was in citharody, which according to Pausanias, took the form of singing a hymn to Apollo. The first winner was Chrysothemis of Crete (10.7.3):

“The oldest contest and the one for which they first offered prizes was, according to tradition, the singing of a hymn to the god. The man who sang and won the prize was Chrysothemis of Crete, whose father Carmanor is said to have cleansed Apollo. After Chrysothemis, says tradition, Philammon won with a song, and after him his son Thamyris. But they say that Orpheus, a proud man and conceited about his mysteries, and Musaeus, who copied Orpheus in everything, refused to submit to the competition in musical skill.
They say too that Eleuther won a Pythian victory for his loud and sweet voice, for the song that he sang was not of his own composition. The story is that Hesiod too was debarred from competing because he had not learned to accompany his own singing on the harp. Homer too came to Delphi to inquire about his needs, but even though he had learned to play the harp, he would have found the skill useless owing to the loss of his eye-sight.”
A supplement to Pausanias’ information at this point comes from a tradition passed down by the Byzantine scholar Photius (Bibliotheca 320b) that the mortal singer Chrysothemis was the first to wear expensive clothing and to take up the kithara to sing a solo nomos or traditional song-pattern in imitation of Apollo (eis mimêsin Apollonos). This anecdote is nothing less than an aetiological story to explain the customary dress of both competitive citharists and their instruments. In sum, we can rest assured that the citharodic nomes featured a strong tie to Apollo as addressee, performance model, and probably theme.
Before turning to non-hymnic—or at least possibly covert hymnic—allusions to the citharodic art, we should note that the seemingly least important word of the exodion we have thus far been examining—ἀλλ’—is actually quite highly marked, in and of itself, as indicating a turn away from the narrative of a hymn and toward closure. [17] Consider, for example, the ending of the Homeric Hymn to Demeter as it turns from narrating the blessings of the older goddess to a direct address to the immortal pair (Demeter and her daughter Persephone):

Ἀλλ’ ἄγ’ Ἐλευσῖνος θυοέσσης δῆμον ἔχουσαι (490)
καὶ Πάρον ἀμφιρύτην Ἄντρωνά τε πετρήεντα,
πότνια ἀγλαόδωρ’ ὡρηφόρε Δηοῖ ἄνασσα
αὐτὴ καὶ κούρη περικαλλὴς Περσεφόνεια
πρόφρονες ἀντ’ ᾠδῆς βίοτον θυμήρε’ ὀπάζειν.
αὐτὰρ ἐγὼ καὶ σεῖο καὶ ἄλλης μνήσομ’ ἀοιδῆς.
Through phonic resonances the key indicator of the send-off (ἀλλ’) echoes twice (see boldfaced words above) as the poet begs for livelihood in exchange for the song—a perfect example of the do ut des economy in which ancient hymnic poetry operated. The final echo in ἄλλης … ἀοιδῆς functions to forecast future performances of this or a similar song; but at the same time, as has been pointed out by Greg Nagy, this ending allows the rhapsodic “hymn” to segue into a performance of another type of hymnos—namely Homeric epic. We can do no better in attempting to articulate this state of the evidence than quote at some length Greg Nagy’s own elegant formulation in Pindar’s Homer:

“That these Hymns are morphologically preludes, with the inherited function of introducing the main part of the performance, is illustrated by references indicating a shift to the performance proper, such as metabêsomai allon es humnon ‘I will shift to the rest of the song [humnos]’ at Homeric Hymns 5 (verse 293), 9 (verse 9), and 18 (verse 11). To sum up the essence of the prooimion, I quote the wording of Quintilian (Institutio oratoria 4.1.2): quod oimê cantus est, et citharoedi pauca illa, quae antequam legitimum certamen inchoent, emerendi favoris gratia canunt, prooemium cognominaverunt… ‘that oimê is song and that the kitharôidoi refer to those few words that they sing before their contest proper, for the sake of winning favor, as prooimion…’. Quintilian’s reference to ‘those few words’ sung by the kitharôidoi ‘lyre [kitharâ] singers’ is belied by the proportions of some of the larger Homeric Hymns, which had evolved into magnificent extravaganzas that rival epic in narrative power, as in the case of the Hymn to Apollo. It is in fact legitimate to ask whether the Homeric Hymns, especially the larger ones, were functional preludes….”

Nagy concludes that “ the medium of the Homeric Hymns, which is poetry recited in dactylic hexameter, is several stages removed from the medium of kitharôidiâ, that is, song. [18]

With the mention of the mode for performance of the Homeric Hymns, in view of their citharodic relations, we need to pause for a moment to recognize a more complex interpretive situation than that involved in tracing prayer tropes and “sacral” language within hymns. Or, more accurately, we need to face the problem of explicating multiple speech-genre and song-related layers in early Greek poetry. That is to say, it is possible to foreground—but not to extricate—diction and rhetorical tropes that we can imagine being shared with non-poetic prayer (greetings to a god; the hypomnesis of a prayer, in which the god’s past intercession or the worshipper’s past invocations are recalled; the final wish for some boon, and send-off formulas). In fact, our earliest “prayers” in Greek are already all in poetic form—that is to say literary representations of prayer—and it would be difficult in practice to distinguish “poetic” prayer from “actual.” Except for a few scattered notices, we do not have anything in Greek like the corpus of Latin prayers preserved by the elder Cato in De Agricultura. Another way of approaching this problem is to admit the possibility that all prayers in early Greek were in high style, with formal metrical and phonic means that kept them distinct from “prose”—that in fact “poetry” was the medium with which to pray. It is a short step from here to the assertion that prayer and hymn in early Greek are in fact indistinguishable.
With “Homeric” hymns, however, we meet two degrees of further complexity. First, the hymnic function as it is represented by the Hymn to Apollo and Hymn to Hermes is unmistakably equivalent to citharodic performance, as Nagy has demonstrated. It features one performer, master of an instrument and singing to its music. Chronologically, the instrument moves from Hermes to Apollo (if we credit the tale of the Hymn to Hermes, itself highly colored with a particular musical ideology). I have argued (Martin 2010) that the pairing of the Hymn to Apollo and Hymn to Hermes allows us to trace a significant symbolic transfer that must be related to the relative standing of archaic Greek performance traditions, in which not only are sympotic lyric forms, accompanied with either barbitos or tortoise-shell lyre, set apart from public, citharodic competitive performances, but also citharodic performances of the agonistic type familiar at Delphi and Athens were distinguished from rhapsodic recitations—the latter, significantly, a feature of the Athenian Panathenaia but not the Delphic Pythia. [19] (As we shall see below, similar competing versions must undergird the lore about Archilochus and his acquisition or lack of instruments.)
However—a second point of complexity—Homeric hymns were themselves in all probability not just several stages removed from citharody, but had become embedded in the rhapsodic competition repertoire. One telling passage suggesting this comes at the end of the shorter Hymn to Aphrodite in the corpus, where the poet—using the khaire-diction familiar from the citharodic exodion—asks for victory “in this contest”: [20]

Χαῖρ’ ἑλικοβλέφαρε γλυκυμείλιχε, δὸς δ’ ἐν ἀγῶνι
νίκην τῷδε φέρεσθαι, ἐμὴν δ’ ἔντυνον ἀοιδήν.
αὐτὰρ ἐγὼ καὶ σεῖο καὶ ἄλλης μνήσομ’ ἀοιδῆς.
Hail, sweetly-winning, coy-eyed goddess! Grant that I may gain the victory in this contest, and order you my song. And now I will remember you and another song also. (HAphr 19-21, trans. Evelyn-White).

The deployment of the unmarked word “song” (ἀοιδῆς) here and elsewhere in hexameter verse that was most likely recited allows the competing poet to elide the actual conditions of performance and to affiliate himself with a tradition of musically accompanied verse, whether or not he himself plays or is accompanied by a musician. To put this another way, one type of poetry shaped for competitive performance in recitation (hexameter long narrative, including “Homeric” style hymns) clearly asserts a genealogical link to another type of poetry meant for singing (again, in competition) to the accompaniment of a cithara. What is the upshot? In short, we have a perfect setting for intertextual, inter-generic and inter-performative hybridization. Recall once more that the Panathenaia featured both rhapsodic and citharodic competitions: it was only natural that performers of one type would listen to and learn from those of the other, and that audiences for both would appreciate strategies, tropes, and even phrasings borrowed and shifting (in intergeneric competition) from one to the other. We might label this speculative, but—as with most philological work beyond the most basic steps—a certain amount of imaginative “reconstruction forward” has to be indulged in if we are to make progress.

To return to the analytical problem at a slightly more abstract level: any occurrence of sacral, prayer diction (khaire, anax) might be “hymnic” in a general application of that term. If we take seriously the evidence of Aelius Dionysius, at least one specific “sacral” usage was associated with citharodic songs (or, if we extrapolate from other lore about Delphic traditions, citharodic hymns to Apollo in particular). But, as I have noted already, the citharodic use must further be viewed within the framework of competitive citharoidia, at Athens, Sparta, and Delphi (to name just three of the premier venues for this contest). I propose now that the exodion ἀλλ[ὰ] ἄναξ μάλα χαῖρε was a sign and a product of that very competition. As a convention, it was by definition well-known, part of the audience’s horizon of expectation. Here we have the parallel of the tragic exodia to guide us. They must have developed at Dionysiac competitions in Athens and probably served the same function of indicating to judges and audience that the play was over—a curtain, as it were, in a theater that lacked such devices.
Given the conventional quality of citharodic exodia, I propose further that the “hymnic” or prayer style address (or elements thereof) involving address to a *wanax or the verb khaire might crop up even in rhapsodic usage—that is to say in the performance of Homeric, Hesiodic, or other recited poetry—and, when it did, could carry with it the distinct trace of a dual heritage—not just as (generic) “prayer” or “hymn” but as the sort of language used in citharodic contestation. In other words, it might repay the effort of hearing a few instances of such language within rhapsodic poetry played against a “citharodic” sound-track, to see whether these already rich texts might yield even more meaningful resonances.
In a previous work, I have pointed out briefly how one “rhapsodic” text that strives to assert its “citharodic” relationships makes use of the language found in the citharodic exordia. [21] This is the Hymn to Apollo, at lines 165–181.

ἀλλ’ ἄγεθ’ ἱλήκοι μὲν Ἀπόλλων Ἀρτέμιδι ξύν, (165)
χαίρετε δ’ ὑμεῖς πᾶσαι· ἐμεῖο δὲ καὶ μετόπισθε
μνήσασθ’, ὁππότε κέν τις ἐπιχθονίων ἀνθρώπων
ἐνθάδ’ ἀνείρηται ξεῖνος ταλαπείριος ἐλθών·
ὦ κοῦραι, τίς δ’ ὔμμιν ἀνὴρ ἥδιστος ἀοιδῶν
ἐνθάδε πωλεῖται, καὶ τέῳ τέρπεσθε μάλιστα; (170)
ὑμεῖς δ’ εὖ μάλα πᾶσαι ὑποκρίνασθ’ ἀμφ’ ἡμέων·
τυφλὸς ἀνήρ, οἰκεῖ δὲ Χίῳ ἔνι παιπαλοέσσῃ,
τοῦ πᾶσαι μετόπισθεν ἀριστεύουσιν ἀοιδαί.
ἡμεῖς δ’ ὑμέτερον κλέος οἴσομεν ὅσσον ἐπ’ αἶαν
ἀνθρώπων στρεφόμεσθα πόλεις εὖ ναιεταώσας· (175)
οἱ δ’ ἐπὶ δὴ πείσονται, ἐπεὶ καὶ ἐτήτυμόν ἐστιν.
αὐτὰρ ἐγὼν οὐ λήξω ἑκηβόλον Ἀπόλλωνα
ὑμνέων ἀργυρότοξον ὃν ἠΰκομος τέκε Λητώ.
ὦ ἄνα, καὶ Λυκίην καὶ Μῃονίην ἐρατεινὴν
καὶ Μίλητον ἔχεις ἔναλον πόλιν ἱμερόεσσαν, (180)
αὐτὸς δ’ αὖ Δήλοιο περικλύστου μέγ’ ἀνάσσεις.
And now may Apollo be favorable and Artemis; and farewell all you maidens. Remember me in after time whenever any one of men on earth, a stranger who has seen and suffered much, comes here and asks of you: “Whom think ye, girls, is the sweetest singer that comes here, and in whom do you most delight?” [170] Then answer, each and all, with one voice: “He is a blind man, and dwells in rocky Chios: his lays are evermore supreme.” As for me, I will carry your renown as far as I roam over the earth [175] to the well-placed cities of man, and they will believe also; for indeed this thing is true. And I will never cease to praise far-shooting Apollo, god of the silver bow, whom rich-haired Leto bare. O Lord, Lycia is yours and lovely Maeonia [180] and Miletus, charming city by the sea, but over wave-girt Delos you greatly reign your own self. (trans. Evelyn-White).

Clearly, all four elements of the citharodic exodion appear in these lines (ἀλλ’ at 165; ἄνα, 179; μάλα, 171; χαίρετε, 166). In retrospect, I now see more clearly that the conventional citharodic exodion has been dispersed and re-cast in two ways. First, there is an internal expansion (lines 166–78) that arises from the hymnist’s elaborate presentation of a dialogic episode within the narrative, in which the poet-speaker (“the man from Chios”) addresses the Delian maidens and contrasts their epichoric choral performance (marvelous for its fascinating imitative qualities) with his own mobile song-making—possibly Panhellenic in ambition. [22] If we did not have the intricate pronominal play of humeis/hemeis/egon (lines 166, 171, 174, 177), which is generated by this interplay (itself a vivid genre-contrast), and could in effect collapse the passage, suppressing the lines that intervene and that describe the respective musical accomplishments and tasks of the performing entities, line would 165 meet up with 179:

ἀλλ’ ἄγεθ’ ἱλήκοι μὲν Ἀπόλλων Ἀρτέμιδι ξύν, (165)
ὦ ἄνα, καὶ Λυκίην καὶ Μῃονίην ἐρατεινὴν (179)

That is, a version of the citharodic exodion would be prominently displayed, one element of it at the start of each line. I am not suggesting that any sort of mechanical expansion did in fact ever occur; rather, the entire idea of the Hymn to Apollo, with its multiple innovations and pairings, embodies an expansion aesthetic that has been organically employed to make a long and interesting narrative, the fruit of generations of reworking various traditions. Such an expansion might even have occurred in the context of rhapsodic contests, as I suggested some years ago (Martin 2000). A second sort of expansion in this particular passage of the Hymn takes the conventional sign-off formula with ἀλλ’ and lengthens it into a pivot passage: what sounds to be at first hearing a sign-off tag (at line 165) is then, after the internal expansion I have just outlined, re-doubled by the formulaic line αὐτὰρ ἐγὼν οὐ λήξω ἑκηβόλον Ἀπόλλωνα. Note that this, too, sounds at first like an exodion-style close: compare in the very same Hymn to Apollo the last lines (545-546): [23]

Καὶ σὺ μὲν οὕτω χαῖρε Διὸς καὶ Λητοῦς υἱέ·
αὐτὰρ ἐγὼ καὶ σεῖο καὶ ἄλλης μνήσομ’ ἀοιδῆς.

To sum up: the Hymn to Apollo, in its pursuit of a conscious modelling of its protagonists (Apollo and the Chian poet) on citharodes, makes “citharodic” sounds at the key switch-over point in the middle of the composition, which is both and ending and a beginning. It is further worth noting that αὐτάρ in this use is a perfect formulaic complement for ἀλλ(ά), because, while both are isosyllabic and of the same metrical shape, the latter ends with a vowel and can thus fit a range of line-initial metrical situations that cannot be accommodated by the former. The Hymn to Apollo passage shows us both complementary formulas at work, one backing up the other.

This is one passage, then, of a rhapsodic composition that gains highly relevant thematic resonance from a citharodic “feel” and rhetorical strategy. Before leaving the hymns to examine an epic moment, it is worthwhile pointing out another passage that may feature similar latent associations. In the Hymn to Aphrodite, the goddess encounters Anchises on Mt. Ida, alone in the cattle pen, while the other cowherds are out on the range. The goddess of love stands before him, disguised as a parthenos; as eros grips the young man, captivated by her wondrous, shining beauty, he greets her and runs through a litany of possible divinities she resembles, ending with a promise to build an altar and a prayer that he might receive from her fame, long, life and descendants. In light of the prayer, and the divine interlocutor, it is unremarkable that Anchises begins his speech with the words Χαῖρε ἄνασσ’ (92–99):

Χαῖρε ἄνασσ’, ἥ τις μακάρων τάδε δώμαθ’ ἱκάνεις,
Ἄρτεμις ἢ Λητὼ ἠὲ χρυσέη Ἀφροδίτη
ἢ Θέμις ἠϋγενὴς ἠὲ γλαυκῶπις Ἀθήνη
ἤ πού τις Χαρίτων δεῦρ’ ἤλυθες, αἵ τε θεοῖσι (95)
πᾶσιν ἑταιρίζουσι καὶ ἀθάνατοι καλέονται,
ἤ τις νυμφάων αἵ τ’ ἄλσεα καλὰ νέμονται,
ἢ νυμφῶν αἳ καλὸν ὄρος τόδε ναιετάουσι
καὶ πηγὰς ποταμῶν καὶ πίσεα ποιήεντα.
“Hail, lady, whoever of the blessed ones you are that are come to this house, whether Artemis, or Leto, or golden Aphrodite, or high-born Themis, or bright-eyed Athena. Or, maybe, you are one of the Graces come hither, who bear the gods company and are called immortal, or else one of the Nymphs who haunt the pleasant woods, or of those who inhabit this lovely mountain and the springs of rivers and grassy meads. (trans. Evelyn-White).

But an echo of the citharodic exodion would be perfectly apt, since the audience at this stage already knows what Anchises was doing when first approached by Aphrodite—namely, playing the cithara (76–80):

τὸν δ’ εὗρε σταθμοῖσι λελειμμένον οἶον ἀπ’ ἄλλων
Ἀγχίσην ἥρωα θεῶν ἄπο κάλλος ἔχοντα.
οἱ δ’ ἅμα βουσὶν ἕποντο νομοὺς κάτα ποιήεντας
πάντες, ὁ δὲ σταθμοῖσι λελειμμένος οἶος ἀπ’ ἄλλων
πωλεῖτ’ ἔνθα καὶ ἔνθα διαπρύσιον κιθαρίζων. (80)
But she herself came to the neat-built shelters, and him she found left quite alone in the homestead—the hero Anchises who was comely as the gods. All the others were following the herds over the grassy pastures, and he, left quite alone in the homestead, was roaming hither and thither and playing thrillingly upon the cithara. (trans. Evelyn-White, slightly modified).

That is to say, the first words of Anchises to the goddess are also the last words of his song—his exodion. Perhaps, in a mise en abyme effect not unfamiliar to archaic Greek composers (one thinks of Iliad 9.186–91—Achilles singing to the phorminx while Patroclus waits), Anchises can be imagined as in the process of breaking off from singing to his cithara something like the Homeric Hymn to Aphrodite. He is, after all, as a rustic singer who encounters a goddess in the wild (like Hesiod or Archilochus), himself a multiform; perhaps he performs a tale of one, as well.

Does Homeric epic ever echo the citharodic exodion—and why? Word-searches of the relevant diction turn up only one passage, but a particularly interesting one. The stylistic background has to be stated, first, in order for us to appreciate the peculiarity foregrounded by the scene in question. Of the 93 times that ἄναξ appears in Homeric poetry, only 17 times does it function as vocative (i.e. as in the citharodic exodion). Seven out of those 17 occur in addresses to Agamemnon (not surprising in view of his regular epithet ἄναξ ἀνδρῶν Ἀγαμέμνων). [24] The only other mortals thus addressed are his brother Menelaus (Il. 23.588), Odysseus (Od. 11.71), Teiresias (Od. 11.144), and Ajax (Od. 11.561). Six times gods are called upon with ἄναξ (Apollo: Il. 16.514 and 523, Od. 8.339; Hypnos: Il. 14.233; an unknown river god: Od. 5.445 and 450). Only in one passage in all of Homeric verse is this vocative introduced with the word ἀλλὰ (as in the exodion). This occurs as Nestor offers his advice to his leader:

ἀλλὰ ἄναξ αὐτός τ’ εὖ μήδεο πείθεό τ’ ἄλλῳ· (360)
οὔ τοι ἀπόβλητον ἔπος ἔσσεται ὅττί κεν εἴπω·
κρῖν’ ἄνδρας κατὰ φῦλα κατὰ φρήτρας Ἀγάμεμνον,
ὡς φρήτρη φρήτρηφιν ἀρήγῃ, φῦλα δὲ φύλοις.
εἰ δέ κεν ὣς ἕρξῃς καί τοι πείθωνται Ἀχαιοί,
γνώσῃ ἔπειθ’ ὅς θ’ ἡγεμόνων κακὸς ὅς τέ νυ λαῶν (365)
ἠδ’ ὅς κ’ ἐσθλὸς ἔῃσι· κατὰ σφέας γὰρ μαχέονται.
γνώσεαι δ’ εἰ καὶ θεσπεσίῃ πόλιν οὐκ ἀλαπάξεις,
ἦ ἀνδρῶν κακότητι καὶ ἀφραδίῃ πολέμοιο.
But do thou, O King, thyself take good counsel, and hearken to another; the word whatsoever I speak, shalt thou not lightly cast aside. Separate thy men by tribes, by clans, Agamemnon, that clan may bear aid to clan and tribe to tribe. If thou do thus, and the Achaeans obey thee, thou wilt know then who among thy captains is a coward, and who among thy men, and who too is brave; for they will fight each clan for itself. So shalt thou know whether it is even by the will of heaven that thou shalt not take the city, or by the cowardice of thy folk and their witlessness in war.” (trans. A.T. Murray)

It is remarkable that this passage, which can be pinpointed nowadays through searching digitized texts, on the basis of its otherwise unattested sequence ἀλλὰ ἄναξ, was already singled out for comment in the 12th century (perhaps recalling earlier, no longer extant analyses) by Eustathius. For this is the precise point in his massive commentary when the Homeric scholar chooses to make the remark with which I began this paper:

(v. 360) Ἰστέον δὲ ὅτι ἐκ τοῦ “ἀλλ’ ἄναξ”, ὅπερ ἐνταῦθα παρὰ τῷ ποιητῇ κεῖται, ἀρχή τις ἐξοδίου κιθαρῳδικοῦ τὸ “ἀλλὰ ἀλλ’ ἄναξ”, ὡς ἱστορεῖ Αἴλιος Διονύσιος.
Note that from “but lord” which is here found in the poet, the beginning of a citharodic exodion [comes], the [phrase] “but [other things?] lord”, as Aelius Dionysius relates…

Perhaps as both bishop and master rhetorician, Eustathius was hypersensitive to forms of address and prayer-style. Or, it could be that his well-trained literary ear detected the unusual combination. He does not give an opinion as to whether the strong resemblance to the citharodic exodion in this (and only this) line in all of Homer means anything. But I believe we can indeed make a claim for several latent associations. First, it should be recognized that the overall structure of Nestor’s speech (Il. 2.336–68) places the vocative appeal at the proper location for an exodion-like close. Eight lines from the end of his intervention, it comes as he turns from comments more generally aimed at the reluctant troops of the Achaeans toward a final address to Agamemnon. The first part of his speech, moreover, functions like the hypomnesis of a prayer. The content, as well, has an overtly religious dimension, as the aged warrior reminds the Achaeans of the vows they undertook, with solemn libations, to conquer Troy. If they return to Argos now, it will be a denial of faith in the promises of Zeus, sealed with his own signs of thunder and lightning as they set out on the expedition (348–53). Unlike the usual prayer-style, Nestor does not match the hypomnesis with a wish for future blessings: rather than saying simply “just as you promised…so now act” he expands on the rhetorical formula to say “so now do not go until requiting Helen’s woes” (354–56). The final prayer or wish is then directed to Agamemnon (not the troops): he should marshall the troops, marking them off (κρῖν) by tribe and clan. This, too, has a religious function, for in this manner it will be possible to tell, says Nestor, whether failure to succeed comes through a divinity’s will (θεσπεσίῃ) or human insufficiency.

Kirk in his commentary notes that Nestor’s frequent tactical advice “tends to be expressed in untraditional language” (1985:154). One wonders whether Nestor’s style is generically different, and if that reflects in turn an affiliation with a different mode of singing the tales of past generations—something more akin to citharody. Nestor is, after all, a survivor of battles that are for the current Trojan War warriors the events of long ago; with these tales he entertains and instructs his most recent hearers. [25] Another angle of vision would allow us to see in the content and style of Nestor’s instructions to Agamemnon in Book 2 a more specifically Apolline directivity. The god who points out how things should be, whether the route of roads or the places for colonies, is also the quintessential god of nomos and properly honored with citharodic nomoi, just as he plays them himself while leading the chorus (cf. Pindar Nem. 5.24–25). [26] In the Hymn to Apollo the functions of leader of a chorus, giver of orders (to the Cretan sailors, his intended priests) and player of the kithara converge (HHA 513–43). Apollo’s concern for the phula of mortals who will come to his temple (cf. identical line-end formula in HHA 537 and 538: φῦλ’ ἀνθρώπων) is not quite the same as Nestor’s for the phula of the army which he would order into place, but the overlap in diction, given the other contextual markers, is interesting.
Also of interest is the afterlife of such language in similar religious contexts. Several centuries after the period in which the Homeric hymn likely crystallized into the form in which we have it, another Apolline figure—Isyllus of Epidauros—undertook to order the worship for Apollo Maleatas and Asclepius. In the hexameter portion of his foundational paeanic hymn, the diction of Nestor’s advice in the passage we have been analysing comes together with instructions for the proper celebration of his god:

Τόνδ’ ἱαρὸν θείαι μοίραι νόμον ηὗρεν Ἴσυλλος (10)
ἄφθιτον ἀέναον γέρας ἀθανάτοισι θεοῖσιν,
καί νιν ἅπας δᾶμος θεθμὸν θέτο πατρίδος ἀμᾶς,
χεῖρας ἀνασχόντες μακάρεσσιν ἐς οὐρανὸν εὐρύ[ν·
οἵ κεν ἀριστεύωσι πόληος τᾶσδ’ Ἐπιδαύρου
λέξασθαί τ’ ἄνδρας καὶ ἐπαγγεῖλαι κατὰ φυλὰς (15)
οἷς πολιοῦχος ὑπὸ στέρνοις ἀρετά τε καὶ αἰδώς,
τοῖσιν ἐπαγγέλλεν καὶ πομπεύεν σφε κομῶντας
Φοίβωι ἄνακτι υἱῶι τ’ Ἀσκλαπιῶι ἰατῆρι
εἵμασιν ἐν λευκοῖσι δάφνας στεφάνοις ποτ’ Ἀπόλλω,
ποὶ δ’ Ἀσκλαπιὸν ἔρνεσι ἐλαίας ἡμεροφύλλου (20)
ἁγνῶς πομπεύειν, καὶ ἐπεύχεσθαι πολιάταις
πᾶσιν ἀεὶ διδόμεν τέκνοις τ’ ἐρατὰν ὑγίειαν,
εὐνομίαν τε καὶ εἰράναν καὶ πλοῦτον ἀμεμφῆ,
τὰν καλοκαγαθίαν τ’ Ἐπιδαυροῖ ἀεὶ ῥέπεν ἀνδρῶν,
ὥραις ἐξ ὡρᾶν νόμον ἀεὶ τόνδε σέβοντας· (25)
οὕτω τοί κ’ ἀμῶν περιφείδοιτ’ εὐρύοπα Ζεύς. [27]

Obviously a careful student of epic-style hexameters, Isyllus may have heard in the unusual phrasing that the Homeric poet gives Nestor at Il. 2.360–63 a reminiscence of Apolline citharodic ethos.

Rhapsodic practice—an intermezzo

As we turn to the alleged rhapsodic exodion (νῦν δὲ θεοὶ μάκαρες τῶν ἐσθλῶν ἄφθονοί ἐστε) preserved by Photius and Eustathius, it is not necessary to dwell on its similarity to hymnic or prayer formulations. The same strictures apply as with the citharodic: that is, more contextual clues need to be sought in order to face the objection that certain dictional items are generally hymnic, and not specifically attached to resonances of rhapsodic practice. We can, however, make one perhaps noteworthy distinction right off, in contrasting it with the citharodic closing. Whereas that tag-line called on one god (probably Apollo, given the lore about citharodic contests) and bade him “rejoice/ farewell” (khaire), leaving unspecified the further prayer for benefits, the rhapsodic exodion calls on all the gods (θεοὶ μάκαρες) and embeds in the same hexameter line the request for good things (esthla) in abundance. It is purely speculative to imagine that the context of rhapsodic contests called for a less specialized appeal, spreading out the set of addressees and pluralizing it as “blessed gods” rather than one. Of course, we might expect the Panathenaia to inspire poets to end their contest-pieces with appeals specifically to Athena—but these are not, to my knowledge, attested, apart from the two short Homeric Hymns addressed to her (#11 and #28, Allen-Sikes-Halliday). On the other hand, the very existence of such a varied range of hexameter prooimia as we find in the Hymns is itself a kind of confirmation that travelling, competing rhapsodes needed to be prepared for whatever divinity was most prominent in the cities they visited—a picture that fits well with that of the widely-ranging “Chian” bard in the Hymn to Apollo, as also with the portraits of Homer and Hesiod in the Lives and the Certamen.
It must be noted immediately that, just as with the citharodic exodion, the rhapsodic is in fact never attested verbatim in those extant texts that we have reason to believe rhapsodes may once have performed. This includes not just Homer and Hesiod, but also Archilochus, Stesichorus, and other “lyric” poets. By itself, the formula θεοὶ μάκαρες is a common one, occurring 30 times in Homeric poetry and 16 times in Hesiod (including transformations into other grammatical cases). Here there are some revealing details. Only four times does it occur (as in the exodion) in the vocative, all in the same formulaic line in the Odyssey (Ζεῦ πάτερ ἠδ’ ἄλλοι μάκαρες θεοὶ αἰὲν ἐόντες). At Od 5.7, Athena uses this line to preface her call for the gods to aid Odysseus; at Od. 8.306, Hephaestus calls the gods to witness his erring wife with Ares in bed; and at Od. 12.371 and 377, Helios calls for Odysseus’ crew to be punished for devouring his cattle. The usage is strikingly specific and consistent: in each passage, one divinity seeks justice from the rest of the community of gods. [28] Thinking of the possible resonances if we import into the poem what we know of this framing exodion, it is difficult to see in specific passages any tonal connection with the rhapsode’s closing call for esthla from the blessed gods. On the other hand, if we export to the rhapsodic exodion the specific tone and situation found through internal analysis of the epic, it could be significant that the call νῦν δὲ θεοὶ μάκαρες τῶν ἐσθλῶν ἄφθονοί ἐστε places in the mouth of a human singer what the audience will have recognized, at least in poems known to us, as divine speech. If this is the register of the line, then the competing rhapsodic poet will have elevated himself to a level above his audience, in what might be a relic of (or aspiration to) a vatic or sacral poetic function. [29]
Only once in hexameter verse before the late antique poet Nonnus does the adjective ἄφθονος appear. This occurs not in adjacency to the formula θεοὶ μάκαρες but in a telling passage that does have a relationship to gifts of the divine. The passage, I shall argue, takes on new shades of meaning, when we recall that the Hesiodic Works and Days was regularly performed by rhapsodes. [30] The description of the Golden Age emphasizes the abundance of grain available to the χρύσεον γένος that enjoyed a god-like existence under the reign of Kronos (Erga 109–12). Their death was like sleep, and the earth bore for them an unstinting harvest (117–26):

θνῇσκον δ’ ὥσθ’ ὕπνῳ δεδμημένοι· ἐσθλὰ δὲ πάντα
τοῖσιν ἔην· καρπὸν δ’ ἔφερε ζείδωρος ἄρουρα
αὐτομάτη πολλόν τε καὶ ἄφθονον· οἳ δ’ ἐθελημοὶ
ἥσυχοι ἔργ’ ἐνέμοντο σὺν ἐσθλοῖσιν πολέεσσιν. (120)
αὐτὰρ ἐπεὶ δὴ τοῦτο γένος κατὰ γαῖα κάλυψε,
τοὶ μὲν δαίμονες ἁγνοὶ ἐπιχθόνιοι τελέθουσιν
ἐσθλοί, ἀλεξίκακοι, φύλακες θνητῶν ἀνθρώπων,
οἵ ῥα φυλάσσουσίν τε δίκας καὶ σχέτλια ἔργα
ἠέρα ἑσσάμενοι πάντη φοιτῶντες ἐπ’ αἶαν, (125)
πλουτοδόται· καὶ τοῦτο γέρας βασιλήιον ἔσχον.
When they died, it was as though they were overcome with sleep, and they had all good things; for the fruitful earth unforced bare them fruit abundantly and without stint. They dwelt in ease and peace upon their lands with many good things…But after the earth had covered this generation—they are called pure spirits dwelling on the earth, and are kindly, delivering from harm, and guardians of mortal men; for they roam everywhere over the earth, clothed in mist and keep watch on judgements and cruel deeds, givers of wealth; for this royal right also they received. (trans. Evelyn-White, slightly modified).

As Greg Nagy eloquently argued thirty years ago, the Golden generation (along with the Silver) represents a poetically stylized version of heroes as they appear in cult. [31] Let us note, in relation to the rhapsodic exodion and the specific association we have seen of the vocative θεοὶ μάκαρες with just vengeance in the Odyssey, that the Golden Age daimones hagnoi as presented by Hesiod are specifically called defenders of justice (124). They are marked, further, by having good things and being good themselves (119, 123). In other words, we have overt or latent associations within this short passage of the Golden Age heroes, with the “gods,” givers of abundance and esthla, as addressed in the vocative of the exodion. [32] Finally, it is worth noting that—as Nagy has once more shown—the Golden Age is described in terms that make it a multiform of the Age of Heroes a few lines later in Hesiod (170–73): [33]

καὶ τοὶ μὲν ναίουσιν ἀκηδέα θυμὸν ἔχοντες (170)
ἐν μακάρων νήσοισι παρ’ Ὠκεανὸν βαθυδίνην,
ὄλβιοι ἥρωες, τοῖσιν μελιηδέα καρπὸν
τρὶς ἔτεος θάλλοντα φέρει ζείδωρος ἄρουρα.
And they live untouched by sorrow in the islands of the blessed along the shore of deep-swirling Ocean, happy heroes for whom the grain-giving earth bears honey-sweet fruit flourishing thrice a year. (trans. Evelyn-White)

The lucky heroes of this latter, fourth age, also beneficiaries of abundant crops, dwell blissfully in the Isles of the Blessed (also under the rule of Kronos). [34] In this multiform of the Golden Age, we therefore have the specific verbal link with θεοὶ μάκαρες of the exodion. What is to be made of these indications?

I submit that we can now distinguish the citharodic exodion more sharply from the rhapsodic, in a way that takes into account, and makes more sense of, performance contexts and the relative expressive positionings of the poet (singer or reciter). The citharodic exodion clearly calls upon a god—as we have seen, most likely Apollo—and probably was used in the context of festivals like the Pythia at Delphi and Karneia at Sparta. By extrapolation, the negotiation of the poetic performance is like that of sacrifice, prayer, and hymn: once praised, the god now should rejoice (and, implicitly, be favorable). The rhapsodic exodion is more specific about the overt request for abundance of good things, and its addressee is plural—the θεοὶ μάκαρες. But I suggest that these “blessed” ones are in fact the heroes whose exploits are described within epic. Functionally, citharodic and rhapsodic compositions work in a quite similar manner, to fashion a verbal agalma that attracts attention and triggers generous repayment from divinities (and perhaps from patrons—another dimension). But formally, the citharodic compositions will have focused on the exploits of the god—just as the “polukephalos nomos” that was a required contest-piece at the Pythia narrated the initiatory and foundational exploit of Apollo’s killing of the Python. [35] In contrast, the rhapsodic songs thematize mortal, heroic exploits—that is the deeds of heroes notionally “before” they become immortalized in song or cult. [36] But the framework for performance of such epics is, naturally, the post-epic world of the Iron Age in which the epic heroes, the daimones hagnoi—in the form of θεοὶ μάκαρες—can be worshipped and asked for blessings in abundance: νῦν δὲ θεοὶ μάκαρες τῶν ἐσθλῶν ἄφθονοί ἐστε. From the standpoint of the Greek audience for rhapsodic performances, the heroic protagonists of epic narrative exist still, as real cultic presences in the landscape, πλουτοδόται “wealth-providers.” To recite their exploits is to gain their favor. (A number of close parallels from ancient and modern Indic epic tradition might be adduced here.) [37] Even if a city-state sponsors the recitation (as at the Panathenaia), the heroic boon, now extended to a broader community, is still the aim and endpoint. As Greg Nagy continually reminds us, myth and poetry in ancient Greece are inextricable from ritual performance. In a very real sense, reciting Homeric epic is itself a ritual act. The long-neglected exodion of the rhapsodes encapsulates this fact.

End-piece: the “comic” exodion

As we have seen, the citharodic and rhapsodic exodia preserved by the lexicographical tradition are never found verbatim in poetic compositions, but tracing fragments of them can help us to see more links between text and context, ritual and performance. The exodion identified as “comic” offers at first sight a bleaker prospect for philological progress. To begin with, by both Photius and Eustathius, we are given only one word as a clue: “having a beautiful crown.”

ὥσπερ κωμικοῦ μὲν ἥδε “καλλιστέφανος,” ῥαψῳδῶν δὲ “νῦν δὲ θεοὶ μάκαρες τῶν ἐσθλῶν ἄφθονοί ἐστε” (Photius Lexicon α 987.)
ὡς ἱστορεῖ Αἴλιος Διονύσιος, ὥσπερ, φησί· κωμικοῦ μὲν ἥδε· “καλλιστέφανος”, ῥαψῳδοῦ δὲ αὕτη· “νῦν δὲ θεοὶ μάκαρες τῶν ἐσθλῶν ἄφθονοί ἐστε”, τραγικοῦ δέ “πολλαὶ μορφαὶ τῶν δαιμονίων”. (Eustathius Ad Hom. Iliadem Volume 1, 364.6.)

In our extant comic texts (from the 152 authors compiled in the Canon of the Thesaurus Linguae Graecae), the adjective appears once—or so it seems. Actually, the citation from Demianczuk’s Supplementum Comicum (1912) printed as #48 of his Fragmenta incertae comoediae and reading “ἥδε καλλιϲτέφανοϲ” is, of course, derived from the lexicographers, so we end up in a perfect textual circle. Furthermore, the old supplementum citation perpetuates a misunderstanding of the lexicographers’ syntax (a construction that may stem from Aelius Dionysius himself), a misconstrual also found in editions of Photius (but not Eustathius), by intruding into the quoted exodion the deictic ἥδε, which clearly was part of the scholars’ own prose transmitting the poetic tag-line, rather than a piece of the exodion. [38] So we are left empty-handed.

There are plenty of crowns in comedy—the noun stephanos (uncompounded) and related verb forms occur at least 100 times—but these, too, offer barely a foothold for analysis. As one of the most abundant and visible semiotic markers in ancient social life, crowns crop up at symposia, as signs of festivity, at sacrifices, as dedicatory objects, and even as funeral bier adornments. Worshippers at the Thesmophoria, the Dionysia, and the Eleusinian mysteries, were said to wear crowns of myrtle. The Pythia’s tripod at Delphi was adorned with a laurel crown. Brides and grooms wore crowns: since comedy sometimes ends with a wedding, should we look to such final marriage scenes? (Unfortunately, even if crowns were part of the actors’ costumes at, for instance, the wedding marking the end of the Birds, the word is never used in the text, and we would have to fill in stage directions ourselves.) And of course winning athletes wore crowns. This last context, given the competitive nature of all the poetic events featuring exodia, deserves further investigation.
As it turns out, in three Aristophanic passages, the connection of crown and victory can help us explore more relevant background. It has to be admitted, however, that only two are exodia passages, while only one features a crown (not the passage in the exodion position). The one involving a send-off that features a crown and a victory song comes in the Knights some 200 lines before the play’s end. The Sausage-Seller has finally defeated Paphlagon to become the new favorite of the Demos, who in turn regrets having ever crowned and endowed his previous demagogic favorite and tells him to surrender his stephanos (Eq. 1225–28). After some further comic exchange about the fulfillment of a fatal oracle, Paphlagon exits on the ekkuklema in paratragic fashion (Eq. 1250–56):

ὦ στέφανε, χαίρων ἄπιθι· καί σ’ ἄκων ἐγὼ (1250)
λείπω· σὲ δ’ ἄλλος τις λαβὼν κεκτήσεται,
κλέπτης μὲν οὐκ ἂν μᾶλλον, εὐτυχὴς δ’ ἴσως.
Ἑλλάνιε Ζεῦ, σὸν τὸ νικητήριον.
ὦ χαῖρε, καλλίνικε, καὶ μέμνησ’ ὅτι
ἀνὴρ γεγένησαι δι’ ἐμέ·

Begone and farewell my crown; against my will do I abandon you.
“Some other man will take you as his own, no greater thief but luckier perhaps.”
Zeus of the Hellenes, yours the prize of victory!
Hail, fair victor, and bear in mind that you became a big shot thanks to me. (trans. Henderson)

The Acharnians ends with a similar scene of defeat, when Dicaeopolis (Ach. 1224–25) calls to be taken to the judges of the Choes drinking-contest (and by metatheatrical extension, those judging the Lenaia competition in the hic et nunc of perfromance), while his opponent Lamachus is taken offstage. He returns triumphant having drained his cup first, and breaks into a victory cry:

Δι. ὁρᾶτε τουτονὶ κενόν. τήνελλα καλλίνικος.
Χο. τήνελλα δῆτ’, εἴπερ καλεῖς γ’, ὦ πρέσβυ, καλλίνικος.
Δι. καὶ πρός γ’ ἄκρατον ἐγχέας ἄμυστιν ἐξέλαψα.
Χο. τήνελλά νυν, ὦ γεννάδα· χώρει λαβὼν τὸν ἀσκόν. (1230)
Δι. ἕπεσθέ νυν ᾄδοντες· “ὦ τήνελλα καλλίνικος”.
Χο. ἀλλ’ ἑψόμεσθα σὴν χάριν
τήνελλα καλλίνικον ᾄ-
δοντες σὲ καὶ τὸν ἀσκόν.
: Look, this pitcher’s empty cup! Hail the Champion!
: Hail then—since you bid me, old sir—the Champion!
 And what’s more, I poured the wine neat and chugged it straight down!
Then Hail, old chap! Take the wineskin and go.
Dicaeopolis: Then 
Follow me, singing “Hail the Champion”
Chorus: Yes, we’ll follow in your honor singing “Hail the Champion” for you and your wineskin. (trans. Henderson).

Dicaeopolis announces his victory with the traditional cry that we have already heard accompanying the transfer of the crown in the Knights. There is no crown mentioned at this moment in the Acharnians (although we might imagine that Dicaeopolis as a symposiast at the Anthesteria is wearing one). What starts as a triumphal interjection is then turned by both the protagonist and the chorus into a song as they exit the orchestra (1231: ᾄδοντες). This same song-phrase τήνελλα καλλίνικος also concludes the Birds, as Peisetaerus leads his bride Basileia (beautifully crowned?) offstage to the triumphal paean of the chorus (Av. 1763–65):

Χο. ἀλαλαλαί, ἰὴ παιών,
τήνελλα καλλίνικος, ὦ δαιμόνων ὑπέρτατε.

Thus far we have seen three send-off scenes, two of them at the position of an exodion, all marked with the language of the victory song (and one—although not an exodion—with an overt reference to crowning). But such circumstantial evidence does not seem quite enough to make a firm connection with the alleged exodion phrase καλλιστέφανος. The scholia to the end of the Birds, however, provide some further information that can help narrow the gap between our texts and the lexicographic evidence. From the commentary we learn—with some confusion—the deeper history of the τήνελλα καλλίνικος song:

Τὸ τήνελλα μίμησίς ἐστι φωνῆς κρούματος αὐλοῦ ποιᾶς ἀπὸ τοῦ ἐφυμνίου οὗ εἶπεν Ἀρχίλοχος εἰς τὸν Ἡρακλέα μετὰ τὸν ἆθλον Αὐγέου
“τήνελλα ὦ καλλίνικε,
χαῖρε ἄναξ Ἡράκλεες,
αὐτός τε κἰόλαος, αἰχμητὰ δύω.”
δοκεῖ δὲ πρῶτος Ἀρχίλοχος νικήσας ἐν Πάρῳ τὸν Δήμητρος ὕμνον ἑαυτῷ τοῦτον ἐπιπεφωνηκέναι.
“Tenella” is an imitation of a type of sound from the playing of the pipe, from the hymn-refrain (ephumnion) which Archilochus addressed to Heracles after the Augean contest:
Tenella, O fair-victory
Hail lord Herakles
Yourself and Iolaos, two spear-fighters
It seems that Archilochus first having been victor in the Demeter hymn in Paros added this interjection for himself.

Worth noting, first off, is the detail that the τήνελλα song is marked as having a double performance heritage, even in this compact rendition of its history. It was both a declaration celebrating personal victory in a hymn contest (a scenario similar to its use by Dicaeopolis in Acharnians) and part of a praise-song for Heracles. The time dimensions, relative and absolute, are vague. Is Archilochus imagined to have been present at the victory of Heracles, or simply to have recalled it? And was he the first ever to win the Demeter hymn-contest (as Chrysothemis had been first at Delphi to win the Apollo-hymn event)? Or, did he win the hymn-contest first, and then compose the praise verses? And do we get here a glimpse of an Archilochus as citharode? (If so the phrase χαῖρε ἄναξ might recall the citharodic exodion). The association of Archilochus with Parian Demeter cult is attested in other sources, as well, so this may be the place to note that the second-earliest occurence of the rare adjective καλλιστέφανος is as an epithet of Demeter in the Homeric Hymn dedicated to her (Hy. Dem. 251, 295). (The very earliest would be the late 8th-century BC Nestor’s Cup inscription, where the epithet modifies the name of Aphrodite.) If we can imagine Archilochus’s own hymn to Demeter as featuring this epithet, then the phrase καλλίνικε applied to himself would have been a nice complement and ending to the body of the hymn, in effect playing on her crown within the narrative to draw attention to his own (crowned?) victory after singing the praise narrative.

The scholia to Pindar present us with even more dizzying possibilities. They comment extensively on the first lines of Pindar’s ode Olympian 9, which begins with a reference to the Archilochean composition:

Τὸ μὲν Ἀρχιλόχου μέλος
φωνᾶεν Ὀλυμπίᾳ,
καλλίνικος τριπλόος κεχλαδώς
ἄρκεσε Κρόνιον παρ’ ὄχθον ἁγεμονεῦσαι
κωμάζοντι φίλοις Ἐφαρμόστῳ σὺν ἑταίροις·

The resounding strain of Archilochus,
the swelling thrice-repeated song of triumph,
sufficed to lead Epharmostus to the hill of Cronus,
in victory-procession with his dear companions. (trans. Svarlien)

In addition to what we have learned from the scholia to the Birds, the Pindaric poem pins down the place of performance to Olympia, refers to a triplex singing, and uses the apparently impromptu short song as a foil for Pindar’s own epinician praises. Note also that the short song is said to be sufficient for the victor Epharmostos as he celebrates in a komos with his friends. We shall return to this setting shortly.

Clearly the song was of great interest to ancient scholars whose notes were absorbed into the Byzantine commentary tradition. The key information contained within the scholia vetera to Olympian 9.1 can be summed up as follows, with my further notations in brackets, pointing to some curious gaps:
Archilochus having come to Olympia made the verses to Heracles [unstated: why and when]. Because he lacked a citharode, Archilochus spoke the word τήνελλα and thereby imitated the tune (or rhythm and sound) of the kithara in the midst of the chorus, while the chorus said the rest of the words καλλίνικε χαῖρε ἄναξ Ἡράκλεις; thereafter anyone lacking a citharode used the τήνελλα [this assumes that normal celebratory song at Olympia was citharodic]. Alternatively: during an evening kômos celebrating the victor, if an aulete could not be found one of the victor’s companions would sing τήνελλα [the confusion over which instrument was imitated recalls the use in the Birds scholion of the term krouma for the sound of the aulos, a term usually applied to kithara playing]. The victor acted as exarchos for the song as he processed in kômos with friends to the altar of Zeus. The phrase and short song came to be used to celebrate all victors.
A few more lines in the scholia vetera attribute to Eratosthenes (apparently, the Alexandrian librarian: p. 226 Bernh.) more specific information about genre. The Hellenistic scholar said that the melos of Archilochus was not an epinikion but a hymn to Heracles; that Pindar called it triploos because the phrase kallinike used to be repeated three times; and that the word τήνελλα was said by the exarchos “apart from the tune” (ἔξω τοῦ μέλους) whenever either citharist [note: not citharode] or aulete was not present. The chorus of komasts then added their “fair-victory” cry so that the whole line was created through this collaboration: τήνελλα καλλίνικε. The actual arkhê of the hymn was the next line (χαῖρε ἄναξ Ἡράκλεες). It seems that Eratosthenes is trying to reconcile competing claims about exactly which instrumental sound was replaced by τήνελλα, the epiphonêma of Archilochus. I would add the suggestion that these competing claims most likely went back to performers’ lore about the invention of their own repertoires, and were shaped further by the competing claims of festivals such as the Pythia, Karneia, and Panathenaia. We know about the Pythia, for instance, that stories circulated concerning why certain major poets could not compete: Hesiod, for example, had not learned how to sing to the kithara (Paus. 10.7.3—see also above). In short, a pre-Hellenistic genre-sorting process that must have aligned recited verse (Hesiodic, Archilochean) in contradistinction to sung poetry probably underlies the “evolution” of Archilochus in lore from (apparently citharodic) hymn singer to impromptu kômos leader lacking an instrument to accompany him. At the same time, as the Hellenistic-era Mnesiepes inscription commemorates, there Archilochus could be remembered as a lyre-player and singer, who acquired his instrument directly from the Muses. [39]
Let us return to the problem of the “comic” exodion. By now it has taken a slight twist. Tracing the roots of the καλλίνικε song segment that marks a turning point in the Knights and provides an exodion-like moment for the end of Acharnians and Birds, we can see that—even without mentioning crowns—the song-event thereby evoked within comedy is in fact komastic. To put this another way, we might read into the wording of Aelius Dionysius and the lexicographical tradition (κωμικοῦ μὲν ἥδε· καλλιστέφανος) the root-meaning “belonging to the kômos.” Of course, in context, the “comic” is meant (clearly contrasted with the “tragic” closing also mentioned in the lexica), but this καλλίνικε-song practice, when we put it into the context of Olympian competitive culture with its musical and poetic traditions, reveals itself to be, rather, “komic.” It is as if, in the use of the καλλίνικε-ending, Aristophanes—at least in some productions (just as Euripides confined the “tragic” exodion to a few plays, of which we know)— gestured toward the pre-dramatic origins of his own art-form, affirming the affiliation between Archilochean iambus and Old Comedy. [40]
At the same time, we need to consider whether the καλλίνικε-ending always carried with it a latent association with athletic games—specifically (as far as the sources tell us) the Olympian. Does it mean something for characters in an Aristophanic comedy—Dicaeopolis and Peisetaerus—to be escorted from the orchestra at play’s end by a komastic chorus to the sound of a victory-song that the audience must have well-known to be for Olympic victors? Or for the Sausage-Seller to be greeted as if he had won the pancration or other event? Once more, we face the analytical dilemma of how to prioritize layers, as when we dealt with hymnic language above: were certain phrasings, like καλλίνικε, more vaguely triumphal, and only thought of as athletic when an obvious Olympic context called for it? We might flip the question the other way round: could it be that the proclamation by Dicaeopolis, after his victory in the compulsory drinking contest of the Choes, points to a genre-specific transmission of the καλλίνικε formula, without reference to the Olympic games? Even in that case, there would still be an interesting overlay of meanings. The Anthesteria (the Dionysiac wine festival in which the “Pots” segment, Choes, took place on the second day) is in this way being consciously melded into the Dionysia and Lenaia (distinct festivals at Athens), and the “pre-dramatic” but highly komastic rites of the former are being signaled, stylized, and incorporated through the snatch of the exodion-like triumph-song, into the hic et nunc of dramatic production (in the case of Acharnians, at the Lenaia of 425 BC).
A third layering of signification cannot be ruled out. Indeed, there is one remaining body of evidence that may prove relevant to the question of the missing καλλιστέφανος exodion for comedies. It might, in fact, bring together the komastic, Olympic, and dramatic celebrations, with an Athenian focus. One fact that can be discovered though an otherwise unrewarding search for significant uses of καλλιστέφανος in the corpus of Greek literature has to do with an unusually specific naming practice, reported by scholiasts, Pausanias, and Aristotle. The scholion (1d Drachmann) to Pindar’s Olympian 8, glossing the word khrusostephanos, informs us that the olive tree from which leaves were used to crown Olympic victors was named “fair-crown”: ἡ διδομένη τοῖς νικῶσιν ἐλαία καλλιστέφανος καλεῖται. Similar information comes from Pausanias, who further provides the exact location of the famous tree (5.15.3):

κατὰ δὲ τὸν ὀπισθόδομον μάλιστά ἐστιν ἐν δεξιᾷ πεφυκὼς κότινος· καλεῖται δὲ ἐλαία Καλλιστέφανος, καὶ τοῖς νικῶσι τὰ Ὀλύμπια καθέστηκεν ἀπ’ αὐτῆς δίδοσθαι τοὺς στεφάνους. Τούτου πλησίον τοῦ κοτίνου πεποίηται Νύμφαις βωμός· Καλλιστεφάνους ὀνομάζουσι καὶ ταύτας.
About opposite the rear chamber [of the Temple of Zeus] a wild olive is growing on the right. It is called the olive of the Beautiful Crown, and from its leaves are made the crowns which it is customary to give to winners of Olympic contests. Near this wild olive stands an altar of Nymphs; these too are styled Nymphs of the Beautiful Crowns. (trans. Jones & Ormerod)

We should note at this point that tree and nymphs share the epithet—just as we have seen it applied to a goddess (Demeter in the Hymn to her). In the work attributed to Aristotle called On Marvellous Things Heard, an account is given of the progenitor of this Olympic sacred olive tree (De mirab. auscult. 51):

Ἐν τῷ Πανθείῳ ἐστὶν ἐλαία, καλεῖται δὲ καλλιστέφανος· ταύτης πάντα τὰ φύλλα ταῖς λοιπαῖς ἐλαίαις ἐναντία πέφυκεν· ἔξω γὰρ ἀλλ’ οὐκ ἐντὸς ἔχει τὰ χλωρά. ἀφίησί τε τοὺς πτόρθους ὥσπερ ἡ μύρτος εἰς τοὺς στεφάνους συμμέτρως. ἀπὸ ταύτης φυτὸν λαβὼν ὁ Ἡρακλῆς ἐφύτευσεν Ὀλυμπίασιν, ἀφ’ ἧς οἱ στέφανοι τοῖς ἀθληταῖς δίδονται. ἔστι δὲ αὕτη παρὰ τὸν Ἰλισσὸν ποταμόν, σταδίους ἑξήκοντα τοῦ ποταμοῦ ἀπέχουσα· περιῳκοδόμηται δέ, καὶ ζημία μεγάλη τῷ θιγόντι αὐτῆς ἐστίν. ἀπὸ ταύτης δὲ τὸ φυτὸν λαβόντες ἐφύτευσαν Ἠλεῖοι ἐν Ὀλυμπίᾳ, καὶ τοὺς στεφάνους ἀπ’ αὐτῆς ἔδωκαν.
In the Pantheon there is an olive-tree, which is called that of the beautiful crowns. But all its leaves are contrary in appearance to those of other olive-trees; for it has the pale-green outside, instead of inside, and it sends forth branches, like those of the myrtle, suitable for crowns. From this Heracles took a shoot, and planted it at Olympia and from it are taken the crowns which are given to the combatants. This tree is near the river Ilissus, sixty stadia distant from the river. It is surrounded by a wall, and a severe penalty is imposed on any one who touches it. From this the Eleians took the shoot, and planted it in Olympia, and from it they took the crowns which they bestowed. (trans. L. Dowdall)

This story fits, in general outline, with the myth attested in Pindar’s Olympian 3, that Heracles was the first to plant olive trees at Olympia. But the remarkable feature of this report, a bit of lore not found elsewhere, is that the famous Olympian olive was actually a transplanted shoot of an older tree that grew in Athens. In the Pindaric ode, Heracles travels to the land of the Hyperboreans to get saplings. As Pavlos Sfyroeras has pointed out, however, there are signs that Pindar in Olympian 3 appears to be manipulating traditions about the birthplace of the olive tree in an indirect challenge to Athenian claims about their polis having been the first place such trees grew, as gifts of Athena. [41] If this is right, then the erasure in non-Athenian literature might have extended to epichoric Athenian stories about which spot the Olympian Καλλιστέφανος supposedly originated from. The site of the “Pantheon” near the Ilissos is not discussed elsewhere, as far as I can find, in the ancient descriptions of Athens. (The building called “of all the gods” in Pausanias is a later Hadrianic construction, near his Library and the Agora.) The foundations of a few as yet unidentified temples have been found near the Ilissos, southeast of the Olumpieion. Further complication is introduced by the version of the story in [pseudo-]Aristotle that we find in the scholion to Aristophanes Plutus (ad 586): in an otherwise very similar account, the tree is said to be near the Ilissos yet some distance from a temple “σταδίους ξʹ τοῦ ἱεροῦ ἀπέχουσα.” It is unclear whether the commentator had in mind a different temple, or placed the tree apart from the so-called Pantheon. Another scholiast, on Pindar Olympian 3, elucidates line 12 with a note that conflates the alsos at Olympia with an (otherwise unattested) Pantheon in Olympia itself: πρὸς αὐτὸ τὸ τῆς Πίσης ἄλσος διαλέγεται· ἢ πρὸς τὸ Πάνθειον, ὅπου αἱ ἐλαῖαι φύονται.

Let me conclude this string of lore and extrapolations with the possibility that once upon a time Attic comedies did in fact end with Καλλιστέφανος. If this was actually just the first word of a longer utterance, the whole might have run “Fair-crowned is Demeter” or even Athena, or Dionysos. But even the single word might have evoked deeper associations. Like καλλίνικε, it would make audiences think of victory—dramatic or komastic. In regard to the latter, it is not amiss to recall that the Anthesteria itself was held at the shrine of Dionysos “in the Marshes” (another site presumably near the marshy verge of the Ilissos, also unidentified at present). To have the chorus and actors shout Καλλιστέφανος with a gesture toward the southeast of the Theater of Dionysos would once more merge pre-dramatic ritual for the wine-god with the highly evolved Dionsyaic celebrations centered around tragedy, comedy and the dithyramb. Finally, this exodion—comic or “komic” as it may have been—will have been a proud assertion that even Olympia, the ultimate site of victory in the Panhellenic world, owed its coveted crowns to an ancient sacred olive tree growing not far from an archetypal Athenian river.


Barker, A. 1984. Greek Musical Writings. Vol. 1: The Musician and His Art. Cambridge.
Blackburn, S. et al., ed. 1989. Oral Epics in India. Berkeley: University of California Press.
Claus, P. 1989. “Behind the Text: Performance and Ideology in a Tulu Oral Tradition.” In Blackburn, S. et al.: 55–74.
Clay, D. 2004. Archilochus Heros. Washington, DC: Center for Hellenic Studies.
Demianczuk, J. 1912. Supplementum Comicum. Cracow: Nakladem Akademii. Repr. Hildesheim: Olms, 1967.
Erbse, H. 1950. Untersuchungen zu den attizistischen Lexika [=Abhandlungen der deutschen Akademie der Wissenschaften zu Berlin, Philosoph.-hist. Kl.]. Berlin: Akademie Verlag.
Ford, A. 2004. The Origins of Criticism: Literary Culture and Poetic Theory in Classical Greece. Princeton.
Frame, D. 2009. Hippota Nestor. Hellenic Studies 37. Washington, DC: Center for Hellenic Studies.
Furley, W. and J. Bremer. 2001. Greek Hymns. Vol. II: Greek Texts and Commentary.
Geffcken, J. 1902. Die Oracula Sibyllina. Die griechischen christlichen Schriftsteller 8. Leipzig: Hinrichs.
Gostoli, A. 1990. Terpandro. Rome.
Kassel, R. and C. Austin, ed. 1989. Poetae Comici Graeci (PCG): Vol. 7. Berlin.
———. 1995. Poetae Comici Graeci (PCG): Vol.8. Berlin.
Kirk, G. 1985. The Iliad. A Commentary. Vol. 1: Books 1–4. Cambridge.
Knudsen, R. A. 2009. The Artificer of Discourse: Homeric Speech and the Origins of Rhetoric. Diss. Stanford.
Kothari, K. 1989. “Performers, Gods, and Heroes in the Oral Epics of Rajasthan.” In Blackburn, et al.: 102–117.
LeVen, P. 2008. The Many-Headed Muse: Tradition and Innovation in Fourth-Century BC Greek Lyric Poetry. Diss. Paris IV-Sorbonne.
Martin, R. P. 2000. “Synchronic Aspects of Homeric Performance: The Evidence of The Hymn to Apollo.” In Una nueva visión de la cultura griega antigua hacia el fin del milenio, ed. A. M. González de Tobia. La Plata, Argentina. 403–432.
———. 2010. “Apolo ejecutante.” In Mito y performance, ed. A. M. González de Tobia. La Plata, Argentina.
———. 2011. “Poseidon’s Crash Test and Apollo’s Cithara.” Paper delivered November 2010 at Athens Dialogues. Available on CHS website at:
Mondi, R. J. 1978. The Function and Social Position of the Kêrux in Early Greece. Diss. Harvard.
Nagy, G. 1990. Pindar’s Homer. Baltimore. Cited in the notes as PH and with chapter and paragraph numbering as from online version.
———. 1999. Best of the Achaeans. 2nd ed. rev. Baltimore. Cited in the notes as BA.
Pavlou, M. 2010. “Pindar Olympian 3: Mapping Acragas on the Periphery of the Earth.” Classical Quarterly 60: 313–326.
Peponi, A-E. 2009. “Choreia and Aesthetics in the Homeric Hymn to Apollo: The Performance of the Delian Maidens (lines 156–64).” Classical Antiquity 28: 39–70.
Powell, J. U. 1925. Collectanea Alexandrina. Oxford. Repr. 1970.
Power, T. 2010. The Culture of Kitharôidia. Washington, DC.
Reynolds, D. 1995. Heroic Poets, Poetic Heroes: The Ethnography of Performance in an Arabic Oral Epic Tradition. Ithaca, NY.
Schwabe, E. 1890. Aelii Dionysii et Pausaniae Atticistarum Fragmenta. Leipzig.
Sfyroeras, P. 2003. “Olive trees, north wind, and time: a symbol in Pindar, Olympian 3.” Mouseion 3: 313–24.
Theodoridis, C. 1982. Photius Lexicon. Vol. 1. Berlin.
Van der Valk, M. 1971. Commentarii ad Homeri Iliadem Pertinentes ad Fidem Codicis Laurentiani Editi: Praefationem et Commentarios ad Libros A-D Complectens. Vol 1. Leiden.
West, M. L. 1978. Hesiod: Works & Days. Oxford.
———. 1992. Ancient Greek Music. Oxford.


[ back ] 1. Cf. on the historian of music Suda σ 1171: Διονύσιος, Ἁλικαρνασεύς, γεγονὼς ἐπὶ Ἀδριανοῦ Καίσαρος, σοφιστής, καὶ Μουσικὸς κληθεὶς διὰ τὸ πλεῖστον ἀσκηθῆναι τὰ τῆς μουσικῆς. ἔγραψε δὲ Ῥυθμικῶν ὑπομνημάτων βιβλία κδʹ, Μουσικῆς ἱστορίας βιβλία λϛʹ· ἐν δὲ τούτοις αὐλητῶν καὶ κιθαρῳδῶν καὶ ποιητῶν παντοίων μέμνηται· Μουσικῆς παιδείας ἢ διατριβῶν βιβλία κβʹ, Τίνα μουσικῶς εἴρηται ἐν τῇ Πλάτωνος Πολιτείᾳ βιβλία εʹ. Schwabe (1890:6) is agnostic on the question of whether the Atticist was the same man as the musicologist, but cites Müller’s belief that they were. On the lexicographical work in relation to other such work see Erbse (1950).
[ back ] 2. Erbse (1950) s.v. I have changed the cited text of Plato Comicus to that in Kassel-Austin (1989), which the editors base on the similar notice in Pollux 10.190. Callimachus uses the related form ἐπιπορπίς to describe the dress of the archetypal citharode Apollo in his hymn to the god (Ap. 32).
[ back ] 3. I disagree with the punctuation provided by the latest editor, Theodoridis (1982:106, lemma #987); he includes ἥδε within the quoted exodion. The correct syntax (and thus punctuation) is clear from the otherwise similar Eustathius passage (κωμικοῦ μὲν ἥδε vs. ῥαψῳδοῦ δὲ αὕτη). Schwabe (1890:102) prints the phrase correctly and notes further (103n2) that lexicographers also collected proemial phrases, e.g. Hesychius α 3944 ἀμφὶ ἄνακτα ἀρχὴ νόμου κιθαρῳδικοῦ (cf. Terpander. fr. 2 Gostoli=PMG 697).
[ back ] 4. On the evidence of the other attestations, I assume the repetition in Eustathius (ἀλλὰ ἀλλ’ ἄναξ) is a garbled transmission of an original ἀλλ’ ἄναξ, although van der Valk (1971) ad loc. does not entertain this possibility.
[ back ] 5. For epilegô as “make formal close” I rely on the usage of such later genres as Attic oratory, in which epilogos was a recognized and stylized final portion of the defense speech. See LSJ s.v.
[ back ] 6. Cf. Hermogenes De inv. 4.9 for an even more expansive definition.
[ back ] 7. On the employment in later analysts of rhetorical practice, especially Aristotle, of Homeric poetry, see now the dissertation by Rachel Ahern Knudsen (2009).
[ back ] 8. I am inspired by Reynolds (1995), a meticulous ethnographic study stressing the need to view “context” as the real “text” (and vice versa) in oral performances.
[ back ] 9. Power (2010:188–89) assumes the phrase was dactylic, but interprets it as more likely “the lead-off verse…of the end of the prooimion rather than of the song proper,” comparing it to the conclusion of many Homeric Hymns (e.g. 21.5, 31.17).
[ back ] 10. See also for the elision Hesychius α 3113: ἀλλ’ ἄναξ· ἐξόδιον κιθαρῳδῶν τοῦτο, καθάπ<ερ ῥαψ>ῳδῶν καὶ τὸ ‘νῦν <δὲ θεοί ..>’
[ back ] 11. Power (2010:336–350) looks at the testimonia concerning Terpander and other early figures in the light of Timotheus and the New Music.
[ back ] 12. It is worth quoting Nagy PH (online version) Chap. 3 para.13n1 regarding a prime source of our information, [Plutarch] On Music (1132de): “In this source, the nomoi of Terpander are understood anachronistically as equivalent to the nomoi of Timotheus of Miletus, a virtuoso composer of the late fifth century, who is said to have composed his earliest nomoi in dactylic hexameters: ‘Plutarch’ On Music 1132e. At 1132de (see Barker 1984.209n25) the source infers that Terpander too composed primarily hexameters (though it would be more accurate to say, on the basis of Terpander PMG 697, that Terpander composed in meters related to the hexameter: Gentili and Giannini 1977.35–36).”
[ back ] 13. Gostoli (1990:147) notes the dictional resemblance between these lines and the exodion, but prefers to see the similarity as the result of widespread, traditional formulaic closings and salutations.
[ back ] 14. Cf. also the Paean Erythraeus in Aesculapium lines 19–20 (Χαῖρέ μοι, ἵλαος δ’ ἐπινίσεο τὰν ἐμὰν πόλιν εὐρύχορον) and the Paean in Apollinem et Aesculapium (=IG 3.1.171b) lines 23–24 (Χαῖρε, βροτοῖς μέγ’ ὄνειαρ, δαῖμον κλεινότατε, ὢ [ἰὲ Παιάν]) both in Powell (1925).
[ back ] 15. See PH Chap. 12 para. 33 on what Nagy calls “The archetypal virtuoso performance of Apollo.”
[ back ] 16. On the Pythian agôn see Power (2010:371–78).
[ back ] 17. This would seem to be a long-lasting convention: cf. the ending of a Sibylline Oracle of the Imperial period (Geffcken #11, lines 322–24), which is even closer to the citharodic exodion (and also embeds a final phrase from the shorter Homeric Hymn to Aphrodite [#6]): ἀλλά, ἄναξ, νῦν παῦσον ἐμὴν πολυήρατον αὐδήν / οἶστρον ἀπωσάμενος καὶ ἐτήτυμον ἔνθεον ὀμφήν / καὶ μανίην φοβεράν, δὸς δ’ ἱμερόεσσαν ἀοιδήν.
[ back ] 18. PH Chap. 12 para. 33–35.
[ back ] 19. See also Martin 2011. Power (2010:468–75) discuss the “lyric politics” underlying the depiction of the transfer in the Hymn to Hermes. His analysis of the Hymn’s melding of sympotic with competitive citharody does not cover the contrast I would draw between Athenian (Panathenaic) and Delphic (Pythian) festival contexts as they are ideologically expressed in Hy.Herm vs. Hy.Ap. Elsewhere (250–57), Power does note the rivalry between citharodes and rhapsodes expressed variously in ancient sources. I would simply extend this insight to the contrastive stances of the two Homeric Hymns just mentioned.
[ back ] 20. See further Martin 2000.
[ back ] 21. The idea was expressed in one paragraph in my paper on “Apollo Citharode” delivered at Delphi in the conference on Apolline Politics and Poetics in July 2003. This was later reworked and modified in a talk at the International Symposium on Myth and Poetics in La Plata, Argentina in June 2009, which resulted in the publication Martin (2010). A longer English version (in which the exodion idea remained only in a footnote, #38 to section 3.15) was delivered at the Athens Dialogues in November 2010 and is available on the CHS website at:
[ back ] 22. For full analysis of the Delian maidens’ performance, see Peponi (2009).
[ back ] 23. See similar phrasing at the end of the long Hymns to Demeter (495) and Hermes (580) and at the end of nine of the shorter Hymns in the collection (Pan; Muses and Apollo; Artemis etc.)
[ back ] 24. Unusual as is the sequence ἀλλὰ ἄναξ, it might have been generated formulaically on the basis of two relevant templates (or their “preverbal Gestalt” to use Nagler’s helpful concept). The phrase “But, up now” occurs four times in Homer (e.g. Il. 6.331: ἀλλ’ ἄνα μὴ τάχα ἄστυ πυρὸς δηΐοιο θέρηται). This association would have provided a familiar sound-sequence upon which an innovative ἀλλὰ ἄναξ might be composed in performance, especially since the ἀλλ’ ἄνα phrase features a homonym (ἄνα) of the rarer vocative form of ἄναξ (ἄνα) which occurs three times in Homer (always as vocative of Zeus, e.g. Il. 3.351: Ζεῦ ἄνα δὸς τίσασθαι ὅ με πρότερος κάκ’ ἔοργε). A more immediate source for the unique expression in Il. 2.360 may be lines of the type attested twice in the Odyssey, in which the feminine form is used, addressing a goddess. At Od. 3.380, Nestor (again) is the speaker, in a prayer: ἀλλά, ἄνασσ’, ἵληθι, δίδωθι δέ μοι κλέος ἐσθλόν. At Od. 6.175, Odysseus uses the same vocative phrase to address Nausicaa, whom he has comapred to a goddess. In neither scene is a citharodic flavor detectable (although the Nausicaa scene is undoubtedly colored by poetry about choruses of young women, and Alcman may have composed a similar scene: cf. scholia HQ ad Od. 6.244ff [Dindorf]: αἲ γὰρ ἐμοὶ—ἐνθάδε ναιετάων] ἄμφω μὲν ἀθετεῖ Ἀρίσταρχος, διστάζει δὲ περὶ τοῦ πρώτου, ἐπεὶ καὶ Ἀλκμὰν αὐτὸν μετέβαλε παρθένους λεγούσας εἰσάγων “Ζεῦ πάτερ, αἲ γὰρ ἐμὸς πόσις εἴη.”
[ back ] 25. On the depth of tradition hidden in the stories by and about Nestor, see now Frame (2009).
[ back ] 26. On the ancient associations between the two meanings of nomos (“law” and “musical mode”) see Ford (2004:260).
[ back ] 27. IG IV2 128; Furley and Bremer (2001:180–192). For a recent analysis of the innovations in this poem, see LeVen (2008:268–80).
[ back ] 28. The context reminds us that Diogenianus (Cent. 6.88), mentioned above, cites as proverbially said in the case of vengeance the phrase νῦν θεοὶ μάκαρες.
[ back ] 29. Cf. Mondi (1978) for comparative evidence concerning Indic masters of speech (heralds) and their function as sacrificial officials.
[ back ] 30. The main evidence for rhapsodic performance of Hesiod is Plato Ion 531.
[ back ] 31. See BA 151–222 for the extended argument.
[ back ] 32. For gods (generally) as givers of esthla, see Hy.Dem. 224: καὶ σὺ γύναι μάλα χαῖρε, θεοὶ δέ τοι ἐσθλὰ πόροιεν (the ironic reply of the goddess to Metaneira).
[ back ] 33. BA 167–73.
[ back ] 34. On the positioning here and import of the lines concerning Kronos found in some MSS see West (1978:195–96).
[ back ] 35. Details in West (1992:213–215).
[ back ] 36. It is not possible here to enter into the much broader question of the relationship between the development of Greek hero-cult and of epic, the question on which Greg Nagy has done the most fundamental work throughout his career.
[ back ] 37. See e.g. Claus (1989) and Kothari (1989) on the relation of local epics, within the Tulu and Rajasthan traditions, to communication with the deified dead.
[ back ] 38. Kassel and Austin correct the error: PCG vol.8 (1995) 67, #181.
[ back ] 39. Clay (2004:14–16).
[ back ] 40. Rosen (1988) remains fundamental for this connection.
[ back ] 41. Sfyroeras (2003). See also Pavlou (2010).