The Performance of Homeric Epic

Homer and South Slavic Epic

John Miles Foley
University of Missouri
E-mail: foley@showme.missouri.edu

'Performance is the enabling event, tradition the enabling referent.'
'Oral traditions work like languages, only more so.'
(two non-traditional proverbs)

South Slavic epic offers classicists an opportunity to do what we cannot do with archaic Greek epic: attend an actual performance by an oral epic singer. In a few minutes, thanks to the generosity of the Milman Parry Collection at Harvard University and its curator, Stephen Mitchell, we will be doing just that--listening to the beginning of an epic song performed by the South Slavic guslar Salih Ugljanin, and then both watching and listening to the 'greatest of singers' from that same tradition, Avdo Medjedovic. Perhaps only appropriately, then, I will leave the last word to the guslari.

As a proem to their performances, however, let me sketch in some background for the often cited but still poorly understood comparison between Homeric and South Slavic epic, first offering some observations on the guslar's actual performance style and then considering the manifold implications of that style for Ugljanin's, Medjedovic's, and (I would argue) Homer's art. For as interesting as the reality of oral traditional epic is from an ethnographic point of view, and as important as it is to meet the complex event of oral performance on its own terms, the telos of such investigations must eventually be the poems themselves. Performance is finally most significant in its impact on poiesis. (1)

Performance Style

It would be foolhardy--and prior scholarship has shown it to be dangerously misleading--to suppose that performances by guslar and aoidos can be simply equated, that observations can be transferred directly from the modern to the ancient event without translation. In short, any analogy has its limits. At best it is suggestive, heuristic, partially parallel, but as an analogy (rather than a mirror image) it also includes points of incongruity, places where the comparison just doesn't fit. Insistence on a one-to-one fit between the South Slavic bards and Homer has led both to inflated claims and to dismissive rebuttals, neither of which brings us closer to an understanding of either epic poetry. In concentrating on the South Slavic analogy, then, let me begin by stipulating that it represents one model of oral epic performance, and absolutely not the only model. In the study of oral epic, as in so many other areas, the kind of pluralism built into Steve Reece's planning for this session seems far the best course.

With that caveat as prelude, I will also contend that becoming a latter- day member of the South Slavic bard's audience can suggest various kinds of productive comparisons with Homer. The purpose of the opening section of this contribution is to outline some of those possibilities for comparison.

It is well to start by emphasizing the physical dimensions of South Slavic epic performance, those aspects that we customarily and unquestioningly delete in the ritual act of text-making. (2) Even the authoritative series Serbo-Croatian Heroic Songs, the official edition and translation record of the Parry Collection, masks all but the most book-friendly parameters of performance. Within the Moslem tradition, Parry's and Lord's principal focus because of its general homology in length and elaborateness with the Iliad and Odyssey, narratives were usually sung to the accompaniment of the single-stringed gusle, but they could also be recited for acoustic recording or dictated for written recording. Predictably, scholars following Parry and Lord, for whom a printed text was the primary goal, paid almost no attention to the vocal or instrumental music of the proceedings; besides BŽla Bart—k's partial transcription of The Captivity of Djulic Ibrahim (Ropstvo Djulic Ibrahima), very little scholarship on the musical dimensions of South Slavic appeared appeared before Stephen Erdely's 1995 study, Music of Southslavic Epics from the Bihac Region of Bosnia. Erdely has described formulaic structures at the level of music and rhythm, at the very least an attractive new perspective on the formulaic phraseology employed by the guslar as well as the aoidos.

And standard printed editions are more than musically challenged. Left- and right-justified lines of English prose translation, or even the tidy linear march of the original-language version, give no hint of the substantial physical exertion required of the guslar. But, although we as modern readers of literary works may feel more comfortable with the epic event reduced to silence and imprisoned on the page, those who compose South Slavic epic understand their performance in much different terms, as when they use the verb turati ('to drive out, impel'), for instance, to describe the feat of oral epic performance. (3) As a result, singing sessions rarely last more than 30-40 minutes before a short break, which is taken at no special point but merely when the bard's voice needs a rest. guslari are expert at resuming the narrative after such breaks, frequently employing a standard proem for continuation and regularly backtracking to the last thematic (type-scene) boundary when not specifically prompted. Once again, there are implications for Homer. On the basis of this analogy, we should be suspicious of imagining any kind of correspondence between performance unit and book-division. With respect to this and many other issues, the kind of cultural or ideological intervention suggested by Richard Janko and, in other areas, by John D. Niles, seems a prudent, realistic hypothesis. Indeed, as last summer's international symposium on 'The Textualization of Oral Epics' at Turku University so thoroughly demonstrated, the written recording of oral epic is everywhere an exception to the rule, something out of the natural course of events.

More can certainly be said about performance style and its perhaps unexpected features from the guslar's perspective. In fact, an entire 'emic' or insider's glossary, comprising the South Slavic singers' own terms for what they do, could and should be assembled in order to foreground their-- rather than our--concerns. Entries would include telltale verbs like brojati ('to enumerate,' denoting the additive, paratactic quality of their composition), and pominjeti ('to keep on remembering,' an imperfective, continuing form describing the audience's ongoing responsibility). Another, crucially important entry would be the South Slavic word for 'word'--re--which in the epic names not a morpheme or a lexeme but a unit of utterance, a speech-act with more than simply a lexical meaning. Rei identify at minimum a line or metrical part-line, but singers also use the term to make reference to whole scenes or even entire song-performances. As speech-acts, these special 'words' index more than they contain meaning, pointing toward institutionalized connotation or referentiality and 'slotting' the immediate in the traditional according to an idiomatic code.

Generally speaking, the guslar's special performance language, or register, resembles Homeric language in a number of ways. I have tried to explain these correspondences--and where appropriate the accompanying differences--in an essay in the current TAPA. (4) The language of South Slavic epic is of course formulaic, although the disparity in prosodies and linguistic processes makes for a different set of traditional rules underlying its formation and maintenance. Instead of the hexameter, with its multiple caesurae, bridges, and colonic forms, the heroic decasyllable of the South Slavs consists of two regularly demarcated cola and different internal architecture, promoting a different kind of traditional morphology. Also like Homer, the guslar composes his tales in a performance idiom heavily dependent on different (geographically defined) dialects and on archaisms. Of special interest for Homerists is the subject of metrical irregularities in South Slavic epic phraseology. The epske pjesme reveal a relatively high incidence of vowel hiatus at points of formulaic juncture, just as in Homer, but the guslari have developed a compensating strategy of using non-lexical sounds to bridge these gaps and avoid intervening glottal stops. The fact that this practice is strictly performance-oriented, and not transcribed into texts even by Parry and Lord's native assistant Nikola Vujnovic, himself a guslar, begs the question of whether the Homeric tradition might have employed a similar strategy.

How consistent is the South Slavic singer's performance style? This is a question that goes right to the heart of the matter. As a language that style is quite consistent, although any given performance can vary substantially from any other. That is, the language employed by the guslar varies within limits: because it is the designated medium for a particular, highly marked activity, its spectrum is much narrower than, for example, the unmarked conversational standard. Like a dedicated word- processor, it has relinquished the responsibility of serving a broad range of functions in favor of serving one particular function with increased economy of expression. That idiomatic responsibility--and not merely the demands of prosody--is the real force behind the consistency and recurrent nature of the performance language. In effect, the most fundamental constraint on the epic register is not metri causa but artis causa.

On the other hand, precisely because it is a language, and not a handy compositional kit of readymade building blocks, the guslar's performance style is subject to expectable avenues of differentiation one encounters with any language. Regional varieties, well documented by the collecting procedures of Parry and Lord, constitute dialects within the traditional register, and individual idiolects are also demonstrable. Thus songs from a given region will identify themselves by the regular appearance of certain formulas, the recurrent details of a shared typical scene, or the consistent fleshing out of a story-pattern. As one moves outside the geographical area, differences tend to multiply, as formulaic phrases take variant characteristic forms, scenes feature different actions or objects, and so forth. Even within a given region different singers' performance styles will vary to some degree. To my mind this sort of areal and individual variability--relatively minor but still tangible--compares to the kind of variability one sees in the Homeric Hymns, for example, as opposed to the Iliad and Odyssey. Finally, as Parry and Lord long ago established, no two performances will ever be entirely superimposable. What remains the same throughout the South Slavic performances is precisely the language as a living idiom, an expressive vehicle that maintains its consistency via a built-in plasticity.

Performance and Reception

Before turning this presentation over to Salih Ugljanin and Avdo Medjedovic, I want to spend a few moments with the other key participant in the performance of South Slavic epic: the audience--both the original auditors and we latter-day readers who struggle to decode the silent page into a living event with an implied tradition. Because the Parry-Lord Moslem songs were produced and recorded under special conditions designed to promote uninterrupted composition, the information they provide on this score is limited chiefly to the rhetorical cues they preserve, chiefly in the proems, and the conversations that Nikola Vujnovic conducted with individual singers. But these are extremely important 'windows' on the cultural dynamics of the event, and they can be verified and supplemented by consulting the ethnographic reports of Matija Murko, Alois Schmaus, and other early investigators of South Slavic epic. To an extent I also draw on my own experience doing fieldwork in this area, as one member of a team focusing on Christian (rather than Moslem) epic and other oral traditional genres.

Telegraphically, we can say that the poetic cues, conversations, and other scholars' experiences harmonize in projecting an audience equipped to enter the performance arena, a 'place' defined by the speech-act rather than by actual geography. To join the bard in this epic space is to experience the verbal transaction in traditional context. That is, continuing with the concept of performance language alluded to above, the South Slavic oral epic seems to assume an audience that is fluent in the same traditional register employed by the singer. To put it as straightforwardly as possible: in performance, composition and reception represent two sides of the same coin. It is not that audience members know the language of performance as well as the singer, and can themselves compose; this overly romantic notion masks the truth that individual audience members might or (much more likely) might not be able to do so. Rather they know at least some of the story, understand the specialized language, are comfortable with the idiom, are fluent in the register. To enter the performance arena, then, is to receive on the singer's linguistic wavelength, and to grasp what is implied by the choice and continued use of that particular communicative channel. It is in this sense that auditors are partners in the exchange keyed by epic performance, and in this sense that even as latter-day readers we should aspire to enter the same arena.

To illustrate the referential power of performance as an event, and of its dedicated language as an expressive instrument, I quote from the first selection you are about to hear, lines 1-5 and 9-11 of Ugljanin's The Captivity of Djulic Ibrahim:

Ej! Dje sedimo, aj! da se veseljimo,
A da bi nas i Bog veseljijo,
Veseljijo, pa razgovorijo,
A ljepsu ni srecu dijeljijo
Na ovome mestu i svakome....
Davno bilo sada pominjemo.
Jednom vaktu bese josvanulo,
E! U Zadaru pucaju topovi
Dva zajedno ja trides ujedno.
Eh! Where we sit, ah! let us be happy,
And may God make us happy,
Make us happy, and foster conversation,
And may he grant us good fortune
In this place and every other....
Long ago it was, now we are remembering.
At one time it dawned,
Eh! In Zadar the cannon fired
Two together and thirty at once.

Heard against the background of the poetic tradition, these lines evoke much more than the reasonably accurate lexical rendering in the righthand column. First, the extensive instrumental prelude, a signal in itself that prescribes the channel for composition and reception, has been deleted, and this is to say nothing of the characteristic vocal melody once the guslar begins to sing. The extrametrical cues of Ej!, aj!, and E!, known to fluent audiences as markers of incipience, lose much of their indexical force when reduced to print symbols.

Even at the level most comfortable to most of us--'the words' within the decasyllable--unidiomatic glosses can misdirect the reader. The first five lines constitute more than a garden-variety induction; they re-create a recognizable scenario. Via traditional implication this poet, audience, and performance are being slotted into a familiar category, namely, the same performance arena where all prior groups have gone to transact their epic conversation and, presumably, the same virtual locus where future groups will do likewise. 'In this place and every other' says as much, quite explicitly. With the end of the proem in line 8--'Long ago it was, now we are remembering'--the poet finishes 'dialing the channel' and enters upon the story proper. Of course, here the dynamic of performance is very much the same, only now less generic and more focused: familiar phrases invoke the traditional referentiality of the recurrent scene, slotting this particular moment among many similar ones, as a South Slavic equivalent of the hrododaktulos eos phrase (line 9) resonantly marks the narrative onset and the cannon fire in expectable combination. At this point, the experienced auditor or reader already knows that a Christian ban (the enemy) is celebrating a victory over Turkish forces (the 'home team') and that the celebration probably involves the capture and incarceration of a beloved Turkish hero. All of these extralexical meanings are clearly and obviously present if not textually explicit; the performance medium and the ambient tradition engaged by performance guarantee the implications for anyone fluent in the epic register.

As I have tried to show in Immanent Art and The Singer of Tales in Performance, this kind of traditional referentiality is an important dimension of the art of South Slavic oral epic, and of Homeric epic as well. By hearing or reading the given 'word' against the poetic tradition, the experience of the guslar's song--and of the Iliad and Odyssey as well--is significantly enriched. The unique, idiosyncratic instance is often meaningfully glossed by its implication as a re or 'word,' a speech-act, as when the opening lines of The Captivity of Djulic Ibrahim 'slot' this performance in an arena familiar to poet and audience alike, or when the 'dawn' line echoes other dawns and rhetorically signals a beginning. Of course, the extant evidence limits us in collating instances of traditional words in archaic Greek hexameter poetry, and as a result there will always be resonances to which we are deaf. But progress can be, and has been, made in recovering performance-based meanings, in learning the idiomatic languages in which the guslar and the aoidos compose and according to whose strictures the audience must participate in the linked process of reception. Alongside the apparatus criticus so long a staple of classical scholarship we should contemplate at least a provisional version of an apparatus fabulosus. In my opinion this is a neo-philological enterprise that we must champion, for it promises insights that are achievable in no other way. It is true to the poetry because it is true to the performance tradition.

The Guslari

1. Audio selection: from Salih Ugljanin, The Captivity of Djulic Ibrahim, no. 4 in Serbo-Croatian Heroic Songs, vol. 2 (Novi Pazar: Srpskohrvatski tekstovi), ed. Albert B. Lord (Cambridge, Mass. and Belgrade: Harvard University Press and the Serbian Academy of Sciences, 1953), pp. 55-74; translation mine.

2. Avdo Medjedovic, 'Kino' [sic], Parry no. 12470, Bijelo Polje, recorded August 10, 1935, 19 lines, rec. no. 7014 (see Matthew W. Kay, The Index of the Milman Parry Collection 1933-1935: Heroic Songs, Conversations, and Stories [New York: Garland, 1995], p. 257).


(1) Behind these brief remarks stands a three-part study of Homer and comparative oral traditions: Traditional Oral Epic: The Odyssey, Beowulf, and the Serbo-Croatian Return Song (Berkeley: University of California Press, 1990, rept. 1993); Immanent Art: From Structure to Meaning in Traditional Oral Epic (Bloomington: Indiana University Press, 1991); and The Singer of Tales in Performance (Bloomington: Indiana University Press, 1995).

(2) For a more thorough consideration of the reductive act of text-making, see J.M. Foley, 'Folk Literature,' in Scholarly Editing: A Guide to Research, ed. David C. Greetham (New York: Modern Language Association, 1995), pp. 600-626.

(3) Cf. Old English wrecan, 'to drive out,' used for the same purpose.

(4) 'Guslar and Aoidos: Traditional Register in South Slavic and Homeric Epic,' Transactions of the American Philological Association 126 (1996): 11-41; see also Foley, The Singer of Tales in Performance, pp. 49-53, 82-92, 110-15, and 150-75.

John Miles Foley
University of Missouri
E-mail: foley@showme.missouri.edu

(John Miles Foley has written or edited twelve books and over 100 articles touching on the topic of comparative oral traditions.)