21. Conclusions and Prospects

Nearly fifty years ago Raphael Sealey cautioned his readers that in regard to the Homeridae, the fifth-century clan from Chios who at one time claimed exclusive descent from Homer: For Sealey, and many before and after him, Phemius and Demodocus in the Odyssey represent the poets (aoidoi) who compose while they perform, while Ion, the rhapsode (rhapsôidos) featured in Plato’s dialogue by that name, represents the opposite extreme of the largely recitational performer. The case for creativity among rhapsodes has not been made easier by the prejudices of Plato (as evidenced in the Ion) and Xenophon, who ranks them among the stupidest of men (Symposium 3.6, Memorabilia 4.2.10). For Plato and Xenophon, although rhapsodes may recite Homer’s words correctly, they simply do not know what they mean.

Even in the largely defamatory treatment of rhapsodes in Plato’s dialogue Ion, however, we may detect a hint of the importance of improvisation in the performance of Homeric poetry. When Ion boasts of his victory at a rhapsodic contest at Epidauros, he says: The verb κοσμεῖν ‘embellish, adorn’, as others have noted, [2] elsewhere in the Ion refers to adornment with regard to clothing (530b5, 535d1), and in itself cannot be translated as ‘improvise’. However, given the improvisational skills of rhapsodes that we have seen, I suggest that Ion’s “embellishment” of Homer be interpreted broadly to include the range of rhapsodic performance techniques: mimetic and gestural elements, vocal range, and especially improvisation and modification of verses. Verbal improvisation against tradition was integral, if admittedly not exclusive, to the popular appeal of rhapsodic competition in performance, and we must see that such competition was essentially a poetic game. The master of that game, like Ion, was the one who most deftly displayed his rhapsodic abilities in live performance.

The negative, conventional view of rhapsodes should be taken primarily to reflect the preoccupations of Xenophon and Plato. Xenophon has Socrates say that the problem with rhapsodes is that they do not know the underlying thought (ὑπόνοια) of Homer. [3] Plato had taken this line of criticism a step further, when he sought to vitiate the claim that by knowing the thought (διάνοια) of Homer about a given subject, a rhapsode could translate that into direct experience. [4] Such hostile views are simply not commensurate with the widespread evidence for public interest in rhapsodic performance attested from the sixth century BCE down to the third century CE. This evidence surely bespeaks the popularity of rhapsôidia as a mode of live performance, and it is the hold of this type of competitive performance over the imagination of the Greeks that we should seek to explain.

So why, to put it simply, were rhapsodic performances so engaging? One answer, as I have outlined it here, is that the damning opinions of Plato and Xenophon have overshadowed a degree of creative improvisation and verbal dexterity, in addition to sheer mnemonic ability, that electrified audiences. In the context of competition these skills allowed for spontaneity and audience engagement against the backdrop of an extremely well-known body of poetry. [5] And we cannot forget that audiences, as Plato’s Ion reminds us, were always the ultimate arbiters of a performer’s victory (Ion 535e). [6] Moreover, Ion’s statement cited earlier also suggests that his creative embellishment, rather than the mere recitation of Homeric poetry, would prompt the Homeridae to reward him (Plato, Ion 530d6–9). Thus a rhapsode’s ability to embellish was central to his technique.

The most important practical implication to be derived from this perspective is that by incorporating a more fluid model of live performance into our understanding of the performance tradition of Homer, we may be able more effectively to account for variations in the manuscript tradition, including the eccentric papyri. [7] This is not to say, of course, that all variations in the Ptolemaic papyri or in Homeric manuscripts generally are to be attributed to rhapsodes. But I have tried to demonstrate that some of them could be, and that the competitive demands of rhapsodic performance accounts for the types of variations that we find better than models that emphasize error, imperfections of memory, and the like. We ought also to dismiss the idea that rhapsodic (or homeristic) variations were meant to compete with “Homer,” an idea that inevitably leads to the conclusion that their innovations are inferior. [8] Until we remove the stigma attached to rhapsodes by the likes of Plato and Xenophon, [9] we will not make any headway in understanding the specific effects achieved by their variations. And yet these variations, such as they are, may give us direct access to how Homer was actually performed and interpreted in performance—information that simply cannot be recovered from any one Homeric text alone.