• Oral Tradition and Internet Technology by John Miles Foley

Transferring an oral performance to the tAgora amounts to a kind of ritual killing. So much is lost: everything you see, everything you hear, everything you experience as an audience member, not to mention the unfolding, emergent character of the performance as an event. Real-time navigation gets replaced by a flat, predetermined script that will never vary – tomorrow it will present exactly the same itinerary as it did yesterday and today. It will foster repetition, and not recurrence. For all of its advantages, textual technology exacts a murderous price when we lead an unsuspecting oral performance into its merciless abattoir.

Textual ideology – that mindless celebration of whatever manages to get em-booked, regardless of the consequences – has blinded us to this inter-agora loss with great and persistent success. But even a moment’s evenhanded consideration will reveal how much we “write out of existence” by defaulting automatically to the book and page.

What performatives are

As one small but telling example of this phenomenon, I offer the counter-textual case of performatives in South Slavic oral epic. Sometimes small, uncomplicated illustrations can engender straightforward perspectives that lengthier, more complex examples can’t match.

Performatives are sounds that are added by singers to individual words without affecting their meaning. So let’s start by emphasizing that the sounds in question are in no way a recognized part of the sense-units to which they are attached, and that they therefore do not participate in the literal or traditional meaning of the poetic text. And here’s another clue: these additional sounds don’t occur anywhere else except in oral epic performances. Not in novels, newspaper articles, manuals, or the like; not even in other OT genres from the Former Yugoslavia. Maybe your tAgora alarm bells are already sounding. What possible function could these curious, apparently extraneous sounds have, then? If they don’t convey meaning and don’t turn up anywhere else, why do we need to pay attention to them?

But notice how I phrased their role: performatives don’t figure in the poetic text – that is, in the much-reduced, brick-and-mortar residue of oral performance. To put it another way, they don’t fulfill any textual function, and that’s a distinction worth our attention. We’re ironically conditioned to experience OT in general, and oral epic poetry in particular, as a much-reduced textual object. For that reason we only too easily conclude that textual function is all there is, and that everything else must always and everywhere be unimportant and negligible, merely static on the channel. As indefensible as our bias demonstrably is from the perspective of multiple agoras and multiple media-technologies, it can be a tricky process to negotiate a cultural blindspot we’re not even aware of. Alas, our unexamined and exclusive loyalty to tAgora thinking typically doesn’t make room for other-than-textual function. All the worse when performatives are in fact precisely that – non-textual.

How performatives work

Here’s how performatives do their work in the oAgora. Epic singers in this tradition regularly confront uncomfortable gaps of breath and voice caused by one word ending in a vowel and the next word beginning with a vowel. Left unattended to, these situations result in a stoppage of air – think of the common warning sound “Uh-oh,” which provides a familiar example of this phenomenon in everyday colloquial English. Since it’s impossible to sing continuously through such a stop-and-start gap, called hiatus by linguists, epic bards face the prospect of compromising the vocal quality of their performance with an unseemly, physically debilitating hesitation. It’s a kind of hiccup that would threaten to interrupt their fluent composition and the audience’s fluent reception.

Over the centuries during which the specialized language of the oral epic took shape, singers developed a clever strategy to deal with the “hiatus problem.” They learned to insert meaningless excrescent consonants, sounds that had no lexical role whatsoever, into the offending gaps. It’s as if we were to insert a “v” into the middle of “Uh-oh” to avoid having to hesitate between its two syllables; by adopting this strategy, we would produce continuous breathing and a much more singable version: “Uhvoh.” No stoppage of breath, no hiccup. (In this case, of course, we’d lose the hesitation that’s actually functional as we seek to register a warning, but that’s a feature of this particular signal in English and doesn’t figure in sung South Slavic epic.)

Other concerns aside, the South Slavic epic singers knew what they were doing. They weren’t modifying the sense-units of the epics in the least, neither what we textualists call words nor the oWords that are the minimal units of utterance in OT. They knew the rules of the game – namely, that these inserted sounds were performatively crucial but lexically non-functional. They did their job in the oAgora without interfering with word-structure at any level.

So the hiatus problem was solved, efficiently and globally. And it was a brilliant solution, one that arose from the singers’ understanding of the narratives they were composing as first and foremost living performances that had to be sung into existence.

Some samples to read (and hear)

The role of performatives is best understood through an actual sample, such as the seven lines below that begin a 1935 performance of The Wedding of Mustajbey’s Son Bećirbey as composed by the guslar Halil Bajgorić. This particular story happens to trace a familiar set of generic pathways through the South Slavic epic tradition, chronicling the adventures of an impetuous young man as he comes of age, seeks to marry, and runs head-on into a grand battle against the very enemy who has abducted his fiancée.

You can read this epic at its full 1030 lines in the free, open-access eEdition that appears online, as well as in a conventional book. More to the present point, at the website you can also listen to the audio, and in the process gain an acoustic appreciation of how Bajgorić deploys performatives in these lines and throughout his epic. I encourage you to add experience to analysis by visiting that site and listening to at least these seven lines.

Here is the passage, in the original South Slavic and in English translation, with the performatives highlighted in boldface type and underscoring (for the morphing book) or in red (for the wiki):

  wOj! Rano rani Djerdelez Alija, Oj! Djerdelez Alija arose early,  
  vEj! Alija, careva gazija, Ej! Alija, the tsar’s hero,  
  Na Visoko više Sarajeva, Near Visoko above Sarajevo,  
  Prije zore vi bijela dana – Before dawn and the white day –  
  Još do zore dva puna savata, Even two full hours before dawn, 5
  Dok se svane vi sunce vograne When day breaks and the sun rises  
  hI danica da pomoli lice. And the morning star shows its face.  

Within this brief sample Bajgorić inserts the performatives w, v, and h in order to bridge gaps between successive vowels, where stoppage of breath would otherwise result. For instance, he uses w before the expletive Oj! in order to get things started in line 1 (assuming no consonant to “push off from” before the performance starts), and has even more reason to insert v before Ej! in the second line because the Alija that closes the first line ends in a vowel. In line 4, the excrescent consonant v avoids an “Uh-oh” moment between zore (“dawn”) and i (“and”). Line 6 contains two performatives, one that mediates between svane (“breaks”) and i, and the other between sunce and ograne (“rises”). Finally, the last line in our sample illustrates how h can serve the same purpose as w and v. If we listened to the whole performance, we would also hear j, l, m, n, and rarely d used for bridging purposes. The performance is shot through with these sounds.

(Careful readers and listeners will notice the performative v where we might not expect it to occur – in the middle of a word in the fifth line above: savata [“hours”]. This deflection from the standard form sahata derives from the equivalent function of v and h as interchangeable performatives. Since v is more frequent in the guslar Bajgorić’s repertoire and style, it sometimes replaces even a lexical h occurring between two vowels. Every linguistic rule has outside-the-box implications!)

Once again, let’s remember that no typographically defined or OT-defined words are compromised by these added sounds. No word-meanings are altered. On the contrary, from the point of view of performance, the inserted bridges support and enable words of all kinds within the specialized language of epic singing. Their function is important and necessary.

Sung and heard, but not transcribed

But now a riddle arises. Nikola Vujnović, the third member of the Parry-Lord-Vujnović team that recorded more than 1500 epic performances during the 1933-35 period, and who was enlisted as the transcriber of the acoustic recordings some years later, left the performatives completely untranscribed. That is, he wholly eliminated them from his graphic representation of the songs he sought to transfer to a written format. Why would he do such a thing? Doesn’t that amount to an error of egregious proportions?

First, consider a preliminary answer that, while interesting and pertinent for the Pathways Project at large, really doesn’t go to the heart of this particular matter. Generally speaking, we can observe that Vujnović’s transcription reveals his personal remaking of Bajgorić’s performance in almost every line. Even though the committed scribe is listening to a fixed acoustic text on aluminum records, and even though he was instructed to render what he heard exactly as the singer performed it, his transcription differs in numerous ways from the actual recording. How do we explain the shortfall?

Well, the writing scribe Vujnović was also himself an epic singer, a navigator of South Slavic oral epic pathways, and had a foot in both worlds. At times his compositional competence as a guslar overrode his attempt to create a verbatim text of someone else’s original performance. Against all expectation, and with pen firmly in hand, Vujnović was resinging the song on the page.

While that general scenario represents an important inter-agora exchange that has significant implications for the distributed authorship and distributed editorship behind many oral-derived texts, it doesn’t directly address the wholesale absence of performatives in Vujnović’s transcription/resinging. After all, he didn’t leave only some of them out, or merely change certain sounds from one form to another; he eliminated every last performative.

To each agora its own

The more pertinent answer to the question of why he did so is simpler, but in its way just as far-reaching. However crucial to epic singing as a living oral tradition the performatives undoubtedly are, they just aren’t textual. They have no role in texts, or, more precisely put, they forfeit their original role as a result of the performance’s forced migration from the oAgora to the tAgora. From that point of view, it should come as no surprise that Vujnović didn’t attempt to bring the performatives across the gulf between agoras. As crucial and functional as they were in the marketplace of oral epic, they constituted only worthless currency in the textual marketplace.

To frame the issue fairly, we should note that Vujnović was hardly alone in his practice. Until this eEdition, all prior editors of South Slavic oral epic silently made the same distinction and cast the same judgment, seldom (and then very partially) acknowledging the acoustic presence of these vowel-to-vowel bridges. Indeed, more broadly we can observe that performatives convert only quite clumsily to default textual code (requiring italics or other typographical signals to differentiate them from the root word-stock memorialized in dictionaries, for example). And this lack of fit shouldn’t shock us either. Why not? Because by definition, textual code isn’t set up to deal with non-textual phenomena, and, cognitively speaking, performatives would represent foreign incursions into the tAgora. (They were included in the eEdition transcription to serve as a guide to audio recording and as evidence for their ubiquitous function in this oral tradition, in other words to increase attention to Bajgorić’s song as the oral performance it is.)

Most plainly put, performatives aren’t meant to facilitate reading; they’re meant to enable performing and listening. They survive and prosper in the oAgora, but the tAgora is where these living sounds go to die.