Appendix A. Rhapsodes in Vase Painting; Rhapsōidia

Vase painting provides evidence for the nature of Homeric performance, though not as much as might be thought, and not in as much abundance as for other musical contests held at the Panathenaia. [1] The earliest depictions of performance on vases of Panathenaic shape pre-date Plato’s Ion by a century and a half, however, and so constitute precious evidence. Some of these have been interpreted as depicting rhapsodic performance. None of the vases are labeled “rhapsode,” however, and the identification of even the most famous “rhapsode” vase is disputed. That is the Kleophrades Painter’s depiction [2] shortly after 500 BCE of a bearded man standing on a platform, holding a knotted staff (a rhabdos?) and reciting (or singing). Beazley was confident this was a rhapsode. Herington calls this vase “the only pre-Platonic evidence for the manner in which the Greeks performed their epic poetry,” [3] and stresses that he is unable to find any other vase that “quite certainly represents an epic rhapsode.” This judgment, and the widespread reproduction of the vase, is due to the words painted as coming out of the performer’s mouth, ὧδε ποτ’ ἐν Τίρυνθι, ‘once upon a time in Tiryns’: a hexameter up to the trochaic caesura, but, unfortunately, not from the Iliad or the Odyssey (or any other identifiable poem) as we know them. On the reverse is an aulos player dressed more elaborately and also on a platform. Shapiro, [4] pointing to the Kleophrades Painter’s practice of spreading a single scene over the two sides of the vase, suggests that the two sides depict an aulode (a singer), rather than a rhapsode, and his accompanist, a pairing which other panathenaic-shaped vases show performing together on the same platform. Shapiro, however, unlike the even more skeptical Friis Johansen, [5] does allow that other vases do depict rhapsodes.
One of Shapiro’s criteria here seems to be the use of a staff: he rejects one figure previously identified as a rhapsode ostensibly because the performer is not carrying a staff. By contrast, Bundrick, while accepting that a rhapsode may be depicted with a distinctive staff, suggests two other vases as depicting rhapsodes even though they lack them. [6] The question of whether and when a rhapsode used a staff is difficult to answer on the basis of vase paintings, since sticks are held by audience members, young and old, and by figures identified as “judges” or “trainers.” [7] The “judge” and/or “trainer” figures are sometimes identified by the fact that their sticks are longer and forked, but this is not always the case for figures so identified. Shapiro notes that the “rhapsode” on a panathenaic-shaped vase in Oldenburg holds a staff with a curved handle that seems to differ from that held by his standing listeners. [8] He believes another panathenaic-shaped vase from Liverpool [9] is the earliest depiction of a rhapsode, from about 540 BCE. This man leans on a staff, looking slightly down, as he stands between two seated listeners, one of whom sniffs a flower in a pose familiar from other depictions of the audience in Panathenaic musical contests. [10] Although the figure is not presented on a platform, the shape of the vase and the typical pose of Athena between two columns on the obverse indicate a musical contest, and, as Shapiro says, it is difficult to see what else this man could be if not a rhapsode.
While kitharists and auletai are easily identifiable, rhapsodes need a staff to be recognized as such. Indeed, perhaps their lack of a distinctive prop is linked to their relative unpopularity on pots. [11] Yet the only other vase painting Shapiro regards as indisputably rhapsodic [12] shows a performer on a bema who seems to lack a staff, all the more striking in that his two listeners have them.
There are, then, on vases officially commemorating the Panathenaia and on other shapes associated with such musical content, depictions of solo performers unaccompanied by a musical instrument, some of whom carry staffs. This visual evidence suggests that, at least from the second half of the sixth century, around the time of the controversial “Panathenaic rule,” rhapsodes performed at the Panathenaia and, at least sometimes, used a staff. Whether Homeric performers in other locations and at earlier dates used a staff or musical accompaniment is another question.
Many discussions do not separate the depiction of Homeric performance from the depiction of “epic” (variously defined) or even more general categories. To be sure, any depiction of performance can stimulate thought: West mentions the representation (on a red-figure amphora from Vulci) of Musaeus carrying both a lyre and a tall laurel wand and relates this to Murko’s 1919 report that the Serbian bard “if necessary, holds a staff, or the long Turkish tobacco-pipe, or some other substitute, instead of his gusle.” [13] Two valuable points: actual performance need not conform to fixed rules, and performers like to hold things. Nevertheless, it is important to realize how scanty is the visual (and other) evidence for Homeric performance in proportion to the space it occupies in scholarly discourse.
Visual depictions of musicians have been enlisted to support the contention that epic performance was accompanied by a kithara. Carter, arguing that epic grew out of ancestor cult (an independently interesting hypothesis), points to depictions of lyre players from the Bronze Age and afterward, claiming that “some of these must show occasions of epic recitation.” [14] But a connection between any given depiction of a lyre player and the Homeric poems is lacking. Many scholars, of course, interpret the depiction of bards within the Odyssey as a self-portrait of the poet-performer as someone who sings (cf. “sing, Muse …”) [15] accompanied by a stringed instrument; some would even add a particular melody. [16] This notion that we should base our idea of Homeric performance on Demodokos and Phemius is separate from the argument from visual evidence, but it leaks into many discussions without comment. [17]
The use of a staff has been approached from another direction, namely the origin and meaning of the word ῥαψωιδός, “rhapsode.” [18] Modern discussions focus on two passages from Pindar (Nemean 2.2 and Isthmian 4.63) and their scholia. Pindar himself does not use the word “rhapsode,” but his diction suggested to ancient commentators a connection to two different words, ῥάπτω, “to sew,” and ῥάβδος, “staff, wand, stitch, rivet.” At Nemean 2.2, Pindar sings of the Homeridai as ῥαπτῶν ἐπέων ... ἀοιδοί, “singers of stitched verses.” In Isthmian 4, [19] Pindar says that Homer has honored Ajax throughout humankind: erecting Ajax’s entire aretē, ἔφρασεν κατὰ ῥάβδον, “he pointed it out with his staff.” The scholia on these passages relate both ῥαπτῶν and ῥάβδος to ῥαψωιδός and ῥαψωιδία:
τοὺς ῥαψῳδοὺς οἱ μὲν ῥαβδῳδοὺς ἐτυμολογοῦσι διὰ τὸ μετὰ ῥάβδου δηλονότι τὰ Ὁμήρου ἔπη διεξιέναι. Καλλίμαχος (fr. 138)·
καὶ τὸν ἐπὶ ῥάβδῳ μῦθον ὑφαινόμενον
ἠνεκὲς ἀείδω δεδεγμένος.
οἱ δέ φασι τῆς Ὁμήρου ποιήσεως μὴ ὑφ’ ἓν συνηγμένης, σποράδην δὲ ἄλλως καὶ κατὰ μέρη διῃρημένης, ὁπότε ῥαψῳδοῖεν αὐτὴν, εἱρμῷ τινι καὶ ῥαφῇ παραπλήσιον ποιεῖν, εἰς ἓν αὐτὴν ἄγοντας.
Some etymologize “rhapsodes” as “rhabdodes” because of their going through the lines of Homer with a rhabdos. Callimachus: “And taking up the story on a rhabdos, I sing it, and it is woven continuously.” [20] Others say that since the poetry of Homer was not gathered into one, but was dispersed and divided into sections, then when they would rhapsodize it, they would be making a sort of concatenation or seam [ῥαφῇ], gathering it into one.
Scholion 1d (Drachmann) ad Nemean 2.1
Modern scholars generally reject an etymological connection to ῥάβδος in favor of ῥάπτειν, “stitch.” [21] The precise nuances of “stitching” have commanded more attention than a connection to the “staff” because of the stitching metaphor’s implications for composition, whether it applies to line-by-line composition or to the stitching together of sections [22] by later rhapsodes (as in the above scholion). It is nevertheless significant that the connection with ῥάβδος seems to be more popular in ancient testimony [23] and that this is explicitly connected with the practice of carrying staffs in performance. (For example, Σ Ion 530a, Herington App. 2 No. 29: “They were so called because in telling the Homeric epics they held rhabdous daphinas.”) The etymology is important, but so are folk etymologies. The scholia are late evidence, to be sure; but keep in mind Pindar’s own use of the phrase κατὰ ῥάβδον in describing Homer’s activity. Pindar’s phrase, in the context of ἔφρασεν and ὀρθώσαις, surely refers to the ῥάβδος as staff or wand, [24] and thus is a welcome early complement to the ceramic evidence that, though earlier than Pindar, does not refer explicitly to Homer. I do not defend a derivation of “rhapsode” from “rhabdos.” Nor does settling questions about this word necessarily settle issues of early Homeric performance. However, where the “stitching” interpretation of “rhapsode” has borne a heavy burden in arguments about composition, I am happy to take the folk-etymological rhabdos theory as emblematic of my own interpretation. Rhapsody qua ῥάπτειν stresses composition, and poiēsis; rhapsody qua rhabdos-wielding stresses performance, and (especially in the rhabdos’ transformative powers) presence/becoming or genesis.
Andrew Ford concludes that ῥαψωιδία, though etymologically connected to composition (a “stitched” verse), refers to a mode of performance: “the solo presentation, in public, of a poetic text without musical accompaniment.” [25] Building on Patzer, Ford argues convincingly that the range of texts that were normally recited in this way were stichic, and so could be said to be “stitched” together line by line. ῥαψωιδία then “comprises the meters that, even when performed without musical accompaniment, remain vividly perceptible as verse.” [26] Is it only coincidence that this corresponds to performers who are free to hold a rhabdos? Here Ford’s etymological focus precludes full discussion of the staff. [27]
Ford notes that compositions normally performed as rhapsody, that is, unaccompanied, may be set to music. Since the evidence is scanty, one should admit the possibility that rhapsodes occasionally accompanied themselves, perhaps even when reciting Homer.
We have pursued the issue of the staff and musical accompaniment because for this, unlike many questions about Homeric performance, there is some evidence. But the question has implications for the interpretation of Homer. To know whether the bard held a lyre or a staff would be to know a little about his freedom of movement and his ability to gesture. This study’s analysis of certain passages explores the potential of a staff-wielding performer, and suggests that the script, at least in these passages, was developed with a staff in mind (and hand). But it is important to acknowledge that other modes of performance incorporate a histrionic dimension: there is no neat equation such as staff: kithara :: histrionic: non-histrionic. One should not look at photos of seated Serbian guslars and listen to recordings and shudder to think of acting. This Herington makes clear in his epilogue, recreating a competition in kitharody of Timotheus’ Persians. While certain gestures become impossible holding a lyre or kithara/guitar, others become possible, even inviting. Witnesses include singers who wield guitars, and, to take a modern epic example, Benjamin Bagby’s twenty-first-century performances of Beowulf. [28]
Nevertheless in our earliest (Platonic) sources, late as they are (Laws 658b; Ion 533b) rhapsōidia is distinguished from kitharōidia, and rhapsodes are linked with actors (Ion 536a; Aristotle Poetics 1462a6). Nagy argues for an increasing “theatricalization” of Homeric poetry, with the Homeristai mentioned by Athenaeus, Petronius, and other late authors [29] forming a kind of end point of this process. At least some of the Homeristai enacted scenes from Homer with multiple actors equipped with swords and shields. Nagy summarizes the thrust of Petronius’ treatment: “the histrionics of these performers are being ridiculed as an abstruse exercise in art, on display for pretentious but ludicrously ignorant connoisseurs.” [30]
But these performers seem to be outliers, rather than a point at the end of a historical development. Nagy’s schema, whereby the Homeric poems are increasingly “theatricalized,” is based in part on the existence of these Roman-period Homeristai and the fact that Plato and Aristotle portray rhapsodes as (overly) histrionic. But, granted the premise of increasing histrionics, it is not clear that the portrayal of singers within the Homeric poems is standing for the early end of this spectrum, a spectrum that parallels Nagy’s (and others’) evolutionary model for the Homeric text, whereby fixity of the text means a “script” which in turn entails “acting.” [31]
Even if we were to grant still further that the earliest performers of the Homeric poems sang to a stringed instrument like Demodokos, we cannot infer their performance style, mode, or aim from Yugoslav guslars; it is as plausible that they were like Bagby doing Beowulf to the Anglo-Saxon harp. And although actorly technique might be perfected using a fixed text, it is not obvious that in the early stages of the development of the poems the performer would be any less histrionic. Plato and Aristotle speak of histrionic performers, but also of a variety of performances. There must have been a difference between Ion in full glory before the masses and the rhapsodes heard “almost every day” by such aristocrats as Niceratus (Xenophon Symposium 3.5), even if this is an exaggeration. [32] Nor does the rhapsodes’ alleged stupidity [33] entail that the more creative poet, or monumental composer, or even “Homer,” was not aware of or interested in the effect of his own presence and performance mode. The contrary is in fact the case. Throughout this study, to repeat, “script” simply refers to the words that are actualized in performance, however fixed, because there is no better word.
Having said all of this, as I have repeatedly stressed, I am not arguing in this book for a high degree of histrionics, or really any particulars at all; only in a few places is the effect of a particular gesture brought to bear, and that need not be unsubtle. (I have not, however, hidden my suspicion that there was a staff involved by siphoning off all mentions of it to these Appendices.) What is important to realize is that the evidence for the histrionic presence of the bard is not all in. The wealth of internal evidence of the poems as scripts remains untapped until one stands up and begins performing. You can sit down if you like.


[ back ] 1. Shapiro 1993:95. Victors in musical contests, unlike athletes, did not receive amphorae as prizes. Vases depicting musical contests are not prize vases but souvenirs; they are panathenaic in shape and conventions but slightly smaller, and are referred to as “pseudo-panathenaic.” M. L. West’s 3 article “Rhapsodes” (unlike West 1997; see below) typifies the slippery way ceramic evidence is adduced: “Originally reciters of epic accompanied themselves on the lyre, but later they carried a staff instead ... Both are shown on vases.” West does not indicate here why he does not interpret lyre-players on pots simply as lyre-players.
[ back ] 2. London E270; ARV2 183, 15.
[ back ] 3. Herington 1985:14.
[ back ] 4. Shapiro 1993:96.
[ back ] 5. Friis Johansen (1967:236n324): the panathenaic-type amphorae “nowhere depict a contest between rhapsodes.” It is not clear whether this comment rules out the depiction of rhapsodes who are performing, but not competing, at the Panathenaia.
[ back ] 6. Bundrick 2015.
[ back ] 7. Beazley 1922:72–73.
[ back ] 8. Shapiro 1993:98; fig. 25. Shapiro earlier (1989:46) mooted the possibility that all three were rhapsodes, or that the figure at left, the only one with the forked staff, was a judge.
[ back ] 9. Liverpool, National Gallery 56.19.18; Shapiro 1993 figs. 26–27.
[ back ] 10. Shapiro 1993:100.
[ back ] 11. Bundrick 2015:25.
[ back ] 12. Dunedin E48.226: Shapiro 1989 pl. 22b–c; Beazley 302889.
[ back ] 13. West 1966 ad Theogony 30; emphasis mine.
[ back ] 14. Carter 1995:286.
[ back ] 15. But as Segal (1994:182) notes, not only do Alkinoos (11.368) and Eumaeus (17.518) compare Odysseus to a singer (aoidos), but the disguised Odysseus also refers to his story as singing (14.464). Nagy (1990b:21) writes: “the word aeidō [sing] (as in Iliad I 1) is a functional synonym, in contexts where the medium refers to its own performance, of the word e(n)nepō ‘narrate, recite’ (as in Odyssey I 1), which does not explicitly designate singing … Self-references in Archaic Greek poetry may be diachronically valid without being synchronically true.” Especially since Odysseus refers to his own narration as singing, it is not clear that Demodokos or Phemius is in any sense a self-portrait, even one that is only diachronically valid; it could be deliberately anachronistic or not referring to the current (current at any stage) poet’s own modality at all. Like Taplin (1992:30), I am “driven away from the model supplied by the bards of the Odyssey.” Cf. Collins 2004:168. But again, the answer to this question is not dispositive for the present inquiry.
[ back ] 16. Danek and Hagel 1995, whose reasoning, however, if applied to iambic trimeter, would show that trimeter too was sung. The movement from a Demodokos-like singer to a rhapsode as depicted on pots and in the Ion is a common unspoken assumption.
[ back ] 17. For example, Gentili (1988:6) cites Demodokos and Phemius as evidence that “the rhapsode’s mode of performance might or might not involve the use of song.” Perhaps, but Demodokos and Phemius are not called rhapsodes within the poems themselves, so this is confusing.
[ back ] 18. Chantraine 1968–1980 v. ῥαψωιδός; Meyer 1918; Fränkel 1925; Patzer 1952; Sealey 1957:317; Herington 1985 Appendix 2, Nos. 1, 28, 29; Gentili 1988:6–7; Ford 1988 (comprehensive discussion with bibliography); Nagy 1996a:61–76; West 1997:163–164; Collins 2004:179–184; West 2010; González 2013:§10.1.1.
[ back ] 19. 4.37–38 Race; 3/4.56 S-M.
[ back ] 20. Or “and the narrative woven around a staff I received and sing continuously”: Collins 2004:182.
[ back ] 21. Herington (1985:167) considers the history and meaning of rhapsode “still unsolved.” Ford (1988:300) states: “We now reject as phonologically impossible the most popular ancient derivation, from ῥάβδος.” This seems likely (cf. González 2013: §10.1.1 “there is no linguistic path from ῥάβδος to ῥαψῳδός”), but the proof is not given. Ford and LSJ cite this scholion as though the rhabdos there is unqualifiedly a staff. But Callimachus, while he may also play on the meaning “staff,” clearly has in mind the “stitching” or fabric metaphor (compare ὑφαινόμενον ἠνεκὲς with Iliad 12.297 ῥάψε ... ῥάβδοισι διηνεκέσιν), or even, given the word’s wide application to sticklike objects, an unattested meaning such as “shuttle” or “distaff.” (Cf. the statement in the Nem. 2 scholia that a rhapsode was called στιχαοιδός because a rhabdos could also be called a stikhos; see Collins 2004:182.) The scholiast seems to be using the passage as evidence of the derivation rhapsode < rhabdode < rhabdos, “staff” (cf. the scholiast’s διεξιέναι). Fränkel 1925 suggests the origin of the term should be sought in epic usage of ῥάπτειν and leaves aside ῥάβδος, except when (as in the Iliad passage just cited) it means “rivet.” Note that this very passage in the links ῥάπτειν and ῥάβδος, whether ῥάβδος means “stitch” or “rivet.” This complicates things considerably, and scholars are perhaps too eager to separate these words (contrast Chantraine’s more cautious ῥάπτω entry).
[ back ] 22. See Burgess 2004:13n51.
[ back ] 23. González 2013: §10.1.1.
[ back ] 24. The entire phrase is typically cryptic. May we link ὀρθώσαις as well as ἔφρασεν with κατὰ ῥάβδον, lending the phrase a necromantic flavor (cf. Appendix B below)? Are we to render ἐπέων closely with ῥάβδον? LSJ does so, yet blandly renders ῥάβδον ἐπέων “measure of his verses,” perhaps having in mind the one Pindar scholium (on Isthmian 4.63) explaining κατὰ ῥάβδον by κατὰ στίχον. Cf. Meyer 1918:330; Patzer 1952; Ford 1988:306; González 2013:§10.1.1.
[ back ] 25. Ford 1988:303.
[ back ] 26. Ford 1988:303.
[ back ] 27. One might argue that the use of the staff qua prop extends beyond the range of the rhapsodic as Ford traces it, beginning with the Callimachus fragment quoted in the scholia above, such that the rhabdos would extend beyond the traditional range of rhapsōidia. But even here, the Callimachus fragment is elegiac, not strophic. If, as Ford emphasizes, traditionally “rhapsodic” (non-melic) poetry may be set to music, it may also be true that melic poetry may be rhapsodically recited. See Ford 2002:25. The rhabdos is, as it were, the instrument taking the place of the kithara or the aulos in the rhapsōidia compound: so would go the folk etymology.
[ back ] 28. Failing a live performance, see Bagby 2006.
[ back ] 29. “Theatricalization”; “ever more theatrical and mimetic over time”: Nagy 1996a:162, 165; Collins 2004:203–218. On Homeristai, Nagy (pp. 158–168) discusses Athenaeus 620b–c; Achilles Tatius, Clitophon and Leucippe 3.20.4, 8.9.2–3; Artemidorus 4.2 ed. Pack; Petronius, Satyricon 59.2–6; and papyri from the second and third centuries CE. Similarly, Danek and Hagel (1995:15–17): improvised “Epengesang zur Phorminx” breaks off at a certain point, after which we see only the endpoints of two different lines of development—uncreative rhapsodes without music, and setting fixed texts to new music. For a skeptical view of Athenaeus’ testimony on earlier Homeristai, see Boyd 1994, esp. 116n18.
[ back ] 30. Nagy 1996a:166.
[ back ] 31. González 2013 repeatedly invokes the “dramatic” or “mimetic potential” of Homeric epic. “In adopting under the influence of drama some of the accoutrements of the acting trade, he was only bringing out the extraordinary mimetic potential already inherent in the Iliad and the Odyssey” (González 2013, Conclusion). If the mimetic potential was there, and “extraordinary,” what spark does either evolution or the influence of drama add?
[ back ] 32. M. L. West 2001:19; Pelliccia (2003:111) suggests these rhapsodes were text-teachers specifically hired to teach Niceratus the text verbatim.
[ back ] 33. Plato Ion; Xenophon Memorabilia 4.2.10; Xenophon Symposium 3.6. Cf. Redfield (1973:144): “they possessed the stupidity which is often found among, and even recommended for, actors. The performer is not a creator; he responds to another’s creation …” My point is simply that one cannot assume rhapsode: stupid: actorly :: creative bard: non-stupid: wooden. Cf. Collins 2004:135–146.