The circumstances will be different to some extent in each traditional culture, but speaking for the one that I know best, that of the Slavic Balkans, I would find one of the most normal places for singing to be the house in a small village where neighbors gather for an evening and sit and talk and listen to a singer. Epics are sung also at weddings and to help celebrate the Slava, the family feast for its patron saint. Another informal setting is the coffeehouse in Moslem communities, where men gather, especially during Ramadan, and listen, after a day of fasting, to epic songs that may continue for a whole night. The singers and the listeners are all “insiders”; that is, they are part of the same tradition.

Let me explain what I mean by “tradition” in respect to epic song. For any individual singer the tradition consists of all the performances of all the songs of all the singers he has ever heard. All the singers encompasses the worst, the best, and all in between. Homer was the best of the traditional singers of whom we know in ancient Greece. He was not outside the tradition or “making use of the tradition”; he was part of it, in it. A tradition is dynamic and ongoing. It lasts as long as there are singers and listeners.

The singing of epic songs is very ancient. It is clear that it began before writing was invented. The ancient Greek tradition was very highly developed by Homer’s time. Though traditions start in the distant past and retain the strength of their roots, they are not of the past, until there are no longer any truly traditional singers and listeners. Traditions are subject to change: the reforming of old stories, the telling of new ones that may seem much like the old. A really living tradition has no need of “preservation” because it is always being preserved with every truly traditional performance by a truly traditional singer.

There are several categories of traditionality, that is, of elements that may persist over generations. I suggest five aspects of oral tradition, which I shall first enumerate, later returning to enlarge upon the second, third, and fifth categories, which call for special emphasis.

First, the practice of storytelling itself, be it in prose or verse, be it spoken, {3|4} sung, or chanted, and of singing songs of various kinds, can be traditional. This means that for generations in a given community or culture people have found a time, a place, and an audience for such a practice. Telling or singing has long had a place in their social behavior patterns. Laments, for example, are sung or chanted as part of the rituals practiced at times of death, and this custom has been kept since time immemorial.

Second, the art of composing songs and stories is itself handed down from one generation of creator-transmitters to the next. This is a crucial category for distinguishing some oral traditional songs or stories from their later literary—that is, “written literary”—counterparts. The traditional process of composition and transmission of oral traditional poetry or prose varies from genre to genre and is treated in detail when we look more closely, for example, at lyric, nonnarrative songs in Chapter 2. In general, lines are constructed with the help of “formulas,” and poems, or stories, or songs, are made up of “themes.”

Third, there is a category of traditional content of traditional literature. Here we find traditional story patterns, traditional generic secular and mythic narratives, and traditional generic types of nonnarrative songs, such as lyric or ritual songs.

Fourth, there are the specific works, the specific oral traditional stories, songs, and short literary forms in all their variants. By that I mean the ballad of “Barbara Allen,” the epic of “Marko Kraljević and Musa the Highwayman,” the tale of “The Three Princesses,” and so forth. I do not believe that this category needs elaboration, but it is necessary to insist that it contain all variants, recorded or not, of each work, because we cannot point to any one of them as the “correct” or “original” text.

The fifth category of traditionality is oral traditional poetics. It may be that from the beginning, some stories and songs were simple, brief, and ephemeral. They consisted of loosely structured, short-lived anecdotes and songs with a limited frame of reference. Yet it is certain that there came into being, as time went on, well-structured narratives and songs of wider reference and deeper meaning told or sung by skillful creator-storytellers or singers. In short, there emerged eventually an “oral literature” in the qualitative sense of the term. We can suppose that repetitions of sounds and patterns of words put together to be imitative and to have the power of magic came to set models of duplication and of balance and proportion which had an appeal to an innate human aesthetic sense.

The second line of the “theme” of the feast (which I shall analyze as a theme shortly) is another well-established formula: {5|6}

The “feast” at Vladimir’s is always “honorable.” Pochesten pir ‘honorable feast’ is clearly a common formula for Rjabinin (and for others), and one of the commonest verbs used with it is zavodil ‘held’ or, in the passive, zavodilsja ‘was held’. In No. 76 another verb is used, sobiral ‘he gathered’, and to pochesten ‘honorable’ is added for the sake of the meter another epithet, the ubiquitous slavnyj ‘glorious’. And in Nos. 80 and 81 an appositive is added, also for the sake of the meter, pirovan’ice ‘feasting’. These are clearly formulas, and taking two (or three) lines together, as one must, one has a cluster of formulas as well.

Some of the above lines are not limited to the theme of a feast at Vladimir’s court. Rjabinin’s version of “Il’ja Muromec and Car’ Kalin” begins with the line

and continues:

Other examples of those two lines can be easily found.

At this point No. 84 parts company with the other three songs by telling that at the feast there were two widows, and the activity centers on their conversations and the results of them. But No. 76 continues with Nos. 80 and 81 for two more lines before diverging in its turn: {6|7}

The princes and boyars are “many” or “all,” and the bogatyrs are slavnyj ‘glorious’, of course—though only once—sil’nyj ‘mighty’, rus’skii ‘Russian’, and moguchii ‘powerful’.

I now move from formula to theme, using the remainder of the theme of a feast in Nos. 80 and 81 from Trofim Rjabinin to illustrate an example of that form of the theme in which a messenger, or ambassador, is chosen to undertake a dangerous mission. One is reminded of the council theme in The Song of Roland in which Ganelon is chosen to carry Charlemagne’s answer to Marsile.

But first the description of those invited to the feast continues for one more line: {7|8}

At this point the two stories begin to diverge, but they both present a speech from Vladimir. There is further setting for it in No. 81:

The speeches are, of course, different in each song. In No. 80, Vladimir needs someone to collect tribute; in No. 81 he wants someone to find him a wife. The reaction, however, is the same.

Now the two songs are on the same level again and the choice of messenger or ambassador is made for the separate undertakings. There is no need to follow in detail the ceremony of choice and acceptance with which the full theme is concluded. It is clear that the singer has a store of verses repeated more or less exactly to describe a feast at which a question is asked by the prince, messengers are proposed, and they are chosen with due ceremony. He needs only to fill in the specifics peculiar to each story. The process of composition of oral literature by using formulas and themes constitutes a traditional art, which one generation passes on to the next.

The Iliad and Odyssey depict the valiant actions of heroes, their prowess in combat, their courage in facing the unknown and the supernatural, their skill in overcoming obstacles. These are “the best of the Achaeans,” to quote the title of a distinguished book by my colleague Gregory Nagy. In war and in council they were preeminent. Most of them were mortals like the rest of us, but some, like Achilles, had one divine parent; and some, like Odysseus, had a god or goddess “on their side,” a divine protector. These connections with gods and goddesses may in part at least account for their being able to accomplish what they did. They should, however, be credited with being of the caliber to merit the assistance of the deities. Sometimes the heroes found themselves in narrative patterns created for gods and later assumed by humans—or half-humans. The god Marduk in the Babylonian creation epic fought with the primeval dragon Tiamat and with the help of supernatural weapons overcame her and created the universe from her carcass. Zeus subdued the monster Typhoeus. In the next generation or so in Mesopotamia, Gilgamesh, two-thirds god and one-third man, killed—with the help of his friend Enkidu, a fully human being—the monster Humbaba. The Greek hero Heracles, a son of Zeus and Alcmena, a mortal woman, was a famed monster slayer. Odysseus, with no immortal parents, had a loyal and powerful protector in Athena. It is not surprising that heroes, like the gods, came to have cults of their own.

The fifth category of traditionality, the poetics of oral literature, requires particular attention. In attempting to trace the unfolding of oral traditional aesthetics, one realizes that some storytellers or singers were more talented than others and that they influenced the way in which stories were told and songs sung by introducing what have later been called figures of speech, thus establishing artistic norms and enriching the tale or song. Rhyme, both end and internal; alliteration; and assonance were repetitions of sound which were soon appreciated for their own sake and became regular features of many oral styles. This was true also for such tropes as anaphora and epiphora, for example, which involve repetitions of words at the beginning or end of lines and are much beloved in Slavic oral traditional poetry and elsewhere.

It is to these gifted tellers and singers that we owe also the elaboration of descriptions of heroes and maidens, of horses, and their trappings, of assemblies of men, of catalogs of chieftains, and of detailed accounts of battles. And they too over the centuries were responsible for working out what we know of as “ring composition” and “chiastic arrangement,” or chiasmus, for which they themselves, of course, had no terms. By “ring composition” is meant a {13|14} structure that can be diagramed as abcba; the elements a and b lead to a central element c, after which elements b and a recur in one form or another, but in reverse order, thus closing a “ring.” Ring composition differs from “chiastic arrangement” in that it “revolves” around a center, whereas in chiasmus there is no center, no c, but only abba. Such structures could be used within a single line, in a couplet, in a larger theme, in groups of themes, or in an entire song, forming ever expanding circles of relationships. In practice, chiasmus tends to be more common in short compass, especially in a single line, but ring composition belongs to longer passages.

Although Murat’s song as a whole can be analyzed in terms of ring composition, it is sufficient merely to outline a single series of episodes that converge in a center. In “The Capture of Temišvar” the serfs complain to the viziers in Buda that King Rakocija is oppressing them. The viziers write to him, asking him to desist. He replies that he will not do so but will gather the seven Christian kingdoms and drive the sultan from Stambol. The viziers write to the sultan, but their letter is intercepted and never reaches him. Now a “ring” begins (within a larger ring):

The recalling of the “exiled” grand vizier is the center of the ring, because this action is the turning point against the traitors. Their doom is sealed. In the first b, the exposing of the traitors is set in motion, and in the second b, the judgment on them is executed. The outer circle that frames this ring consists of the gathering of opposing forces in both as. {14|15}

Thus, in the course of time, developed the high quality of oral traditional literature which in the end was bequeathed to written literature; for all these elements came into being before writing was invented.

The stories and songs thus created were not only oral; they were also traditional. Young people learned from their elders how to tell stories and sing songs skillfully and with a sense of the special style that was theirs alone. One generation passed on to the next the technique, or art, of composing tales and songs, together with the appropriate story, ritual, or song material. Writing was not needed. The art was perfected without it and was never dependent on it. It is vastly important that this be understood. It is a technique that preserves essential patterns and associations and presents them in such a way as to make them most meaningful, most effective, and at the same time in a manner both pleasing and suitable, which happens also to be easy to remember.

Oral traditional poetics has some special characteristics because of the way in which it was created from the matrix of oral tradition. The opening lines of Hilferding No. 76 with their repetition of slavnyj ‘glorious’ would not be considered as good style in the poetics of written literature, but they are natural and right in the poetics of oral traditional literature:

Written literature can, of course, easily imitate this usage of noun-epithet formulas, which arises from the necessity of being able to use the needed noun in a variety of metrical circumstances, but it would be imitation of the oral traditional style. No poet in a written literary style would create such lines. Were he to do so, he would be severely criticized. We cannot employ the criteria of written poetics to such a passage without doing an injustice to the oral traditional poetics that formed it and that finds it normal and “right.”

Almost every time Vladimir appears, he has the epithet as above stol’ne-kievskoj ‘of the capital Kiev’; and Il’ja Muromec is always the ‘old Cossack’ (staroj kazak). Poets in written literature do not favor even more or less fixed, “stereotyped,” epithets. It is a different, oral, special poetics that does favor such epithets, understanding their necessity and feeling their appropriateness.

Moreover, to indicate another characteristic construction in Slavic oral traditional poetry, when in the same song Rjabinin sings: {15|16}

the repetition of the words at the end of one line at the beginning of the next would seem unnatural in written poetics. It is, however, an accepted, even preferred, device in oral traditional poetry, to which it is special, because it was created to allow three ideas to be expressed in two lines, namely, (1) ‘he shouted’ (krichit), (2) ‘with all his might’ (vo vsju golovu), and (3) ‘in a loud voice’ (gromkim golosom).

These are some of the devices of oral traditional poetics that are different from the poetics of written literature and stem from the method of oral composition by formula and theme. Associated with these peculiar differences is the impression that the oral style was inferior to the written. The examples given above indicate differences in devices but not in quality, although the critic of written literature might disagree.

Marjanović made many changes in the manuscript. He left out lines and added lines; he left out blocks of five to ten lines. He changed all eleven-syllable lines to ten syllables, and sometimes he combined two lines. His edited texts do not represent the exact words of the singer who dictated them. Marjanović brought to the editing criteria different from those of the singer. Sometimes he omitted “awkward” lines, such as “Then you should see Beg Mustajbeg,” lines the singer used frequently in performance and, interestingly enough, continued to use in dictating the text for a scribe. This fact, incidentally, speaks eloquently for the high quality of collecting involved. Marjanović’s collection is one of the most scrupulously gathered ones that I have seen. {16|17}

Finally, the last line was changed by omitting jer ‘because’ at the beginning and adding je ‘it’ after razmota ‘opened’, literally, ‘unwound’. This change is different from the other two, and its acceptability could be argued. The singer introduces many lines with jer ‘because’ quite ungrammatically and illogically. From the point of view of oral traditional poetics, his doing so could legitimately be held to be poor style. In short, oral traditional poetics has its own standards as well; it is not true that anything done by singers is necessarily good. But the criteria for making such judgments should be those of oral traditional poetics, not those of written poetics. We are in fact dealing with two separate poetics. It is our task to set forth the particulars of their differences and especially to describe oral traditional poetics as specifically as possible, as it is the less well known.

The anaphora in lines 4-6 of the passage can be matched in written literature, of course. Such figures of speech, however, are not the property of written literature alone, nor did they originate in written literature. They were used in oral literature long before writing was invented. This passage of investiture, moreover, shows a typical traditional pattern of positioning of verbs and of alteration of a group in final position with one in initial position. The close of the passage is marked, again traditionally, by internal rhyme. I have underlined the verbs to make this patterning clear. Makić put those words in those positions following a traditional poetics, not an individual one of his own making. The configurations in the passage are part and parcel of the formulaic style and are learned together with the art of making lines and clusters of lines and the technique of moving forward from one line to another. The poetics Makić was using was that of the traditional process of composition. Therefore, though the two literatures, oral and written, share many figures of speech, each has its own particular characteristics.

Notable also in the above passage is the paratactic construction. The necessary enjambment between lines one and two is followed by a series of twelve independent lines, each containing a complete thought. It is the whole constellation of elements found in the lines quoted, their combination into a special structure, that constitutes the particular quality of oral traditional poetics. {19|20}

It is now time to turn to the genres themselves: oral traditional epics, oral traditional ballads, oral traditional lyrics and ritual poetries, oral traditional praise poems, oral traditional wisdom poetry (for example, proverbs and riddles), oral traditional verse and prose used in games, and oral traditional prose stories, such as folk tales. In some of these genres the poem or tale is conceived of as a more or less fixed entity, with its own wording, and in these cases the poem—more rarely also a story—is itself transmitted as a verbal entity. This applies particularly to wisdom poetry and prose, but it is also applicable to ritual, lyric, and game songs. There may be variation, and generally there is, but there is also on the part of the singers or reciters themselves a sense that such poems have, or should have, a given set of words. These poems, of which there are many, are ordinarily comparatively short and consist primarily of oral traditional nonnarrative songs. It is worth repeating that these poems are transmitted from one generation to another in a traditional society as verbal entities with a certain distinctive verbal content; each has its own more or less stable set of words, its own identity, and, if this be true, it would be perfectly correct to say here that songs, meaning texts of songs, are transmitted.

It may seem superfluous to insist on reminding ourselves at this point that memory does not always involve conscious exact memorization of a text fixed in only one form. In the world of everyday communication we remember with many degrees of exactness; we even say, “I could not repeat it word for word, but I can tell you the gist of it.” That means that we did not bother to memorize the exact words of what was heard or read, but we remembered its essential meaning. It also means that we had a way through everyday speech of expressing what we had heard. I pay considerable attention to this phenomenon in the realm of the forms of verbal art in which oral literature is couched, attempting to define as precisely as possible and as specifically as possible what the “more or less” means.

Questions about such concepts as textuality, fixity or fluidity of text, and {20|21} memory demand that we consider the composition of nonnarrative genres in some detail. By textuality I mean that awareness, on the part of the composer, of the words he is using, of a text as such—as against content. One of the important concerns that arises is the “span” of textuality. How many lines, for example, how much text, does the composer’s sense of textuality cover? In stanzaic poetry does it go beyond the stanza, or can it embrace several stanzas? Is it, perhaps, limited to couplets? The only way to answer these questions is to look at specific examples of repeated texts in lyric poetry.